Tutorial Galgas Montaje

Tech Note TN-501-2
MICRO-MEASUREMENTS
Noise Control in Strain Gage Measurements
T
E
C
H

N
O
T
E
Strain Gages and Instruments
For technical support, contact
micro-measurements@vishaypg.com
www.micro-measurements.com
1
Document Number: 11051
Revision: 01-Nov-2010
Introduction
Strain measurements must often be made in the presence
of electric and/or magnetic elds which can superimpose
electrical noise on the measurement signals. If not
controlled, the noise can lead to inaccurate results and
incorrect interpretation of the strain signals; and, in severe
cases, can obscure the strain signals altogether. In order to
control the noise level, and maximize the signal-to-noise
ratio, it is necessary first to understand the types and
characteristics of electrical noise, as well as the sources of
such noise. With this understanding, it is then possible to
apply the most effective noise-reduction measures to any
particular instrumentation problem.
This technical note identies some of the more common
noise sources, and describes the routes by which the
noise is induced into strain gage circuits. It should be
noted that the treatment here is limited to noise from
external electrical and magnetic sources. This note does
not cover effects from nu clear or thermal sources, nor
does it consider the effects of variable wiring or contact
resistance caused by slip rings, connectors, switches, etc.
Following the discussion of noise sources, specic methods
are given, varying with the noise-coupling mechanism, for
noise avoidance. The information in this technical note
is equally applicable to both analog and digital systems
employing dc ampliers. It also applies to systems using
carrier excitation and carrier ampliers.
Noise Sources and Pickup Media
Virtually every electrical device that generates, consumes,
or transmits power is a potential source for causing noise in
strain gage circuits. And, in general, the higher the voltage
or current level, and the closer the strain gage circuit to
the electrical device, the greater will be the induced noise.
Following is a list of common electrical noise sources:
ac power lines arc welders
motors and motor starters vibrators
transformers fuorescent lamps
relays radio transmitters
generators electrical storms
rotating and soldering irons
reciprocating machinery
Electrical noise from these sources can be categorized into
two basic types: electrostatic and magnetic. The two types
of noise are fundamentally different, and thus require
different noise-reduction measures. Unfortunately, most
of the common noise sources listed produce combinations
of the two noise types, which can complicate the noise-
reduction problem.
Electrostatic fields are generated by the presence of
voltagewith, or without current f low. Alternating
electrical fields inject noise into strain gage systems
through the phenomenon of capacitive coupling, by which
charges of correspondingly alternating sign are developed
on any electrical conductors subjected to the field
(Figure 1). Fluorescent lighting is one of the more common
sources of electrostatic noise.
Magnetic fields are ordinarily created either by the
fow of electric current or by the presence of permanent
magnetism. Motors and transformers are examples of the
former, and the earths magnetic eld is an instance of the
latter. In order for noise voltage (emf) to be developed
in a conductor, magnetic lines of f lux must be cut
by the conductor. Electric generators function on this
basic principle. In the presence of an alternating eld,
such as that surrounding a 50/60-Hz power line, voltage
will be induced into any stationary conductor as the
magnetic eld expands and collapses (Figure 2). Similarly,
a conductor moving through the earths magnetic eld has
a noise voltage generated in it as it cuts the lines of fux.
Figure 1. Electrostatic noise coupling.
T
E
C
H

N
O
T
E
For technical questions, contact
TN-501-2
Micro-Measurements
2
Since most irons and steels are ferro-magnetic, moving
machine members redirect existing lines of fux, and may
cause them to be cut by adjacent sensitive conductors. As
a result, signal conductors in the vicinity of moving or
rotating machinery are generally subject to noise voltages
from this source.
Detecting and Troubleshooting
In order to effectively assess the presence and magnitude
of noi se, the strai n gage i nstrument selected for
use should incorporate a simple, but very significant
featureprovision for removing the excitation from the
Wheatstone bridge. With such a control, the instrument
output can be easily checked for noise, independently of
any strain signal. This represents a very powerful tool for
evaluating the effectiveness of shields and grounding, and
for experimentally modifying these methods to minimize
the effects of noise. All Micro-Measurements strain gage
instruments and data systems are equipped with this
important feature.
The following procedure can be used to troubleshoot a
system for noise:
1. If not already known, determine the tolerable levels of
noise in output units (millivolts, inches of defection,
etc.) as observed on a readout such as an oscilloscope or
data system display.
2. Consideration should be given rst to noise sources
affecting the measurement system, isolated from all
external circuits. For this purpose, disconnect any
strain gage leads, and terminate the S+/S amplier
inputs with about the same input impedance that the
amplier normally senses (typically between 120 and
1000 ohms). If excessive noise exists:
a) Check for ground loops (more than one connection
of the system to ground).
b) Check for line- (mains-) radiated noise.
c) If feasible, reduce amplier gain and compensate by
increasing bridge voltage.
3. Having eliminated or satisfactorily minimized noise
pickup by the instrument, turn next to the external
circuitry. With the excitation set to off, connect the
gage or transducer circuit (including leadwires) to
the instrument, and observe noise. Of course, any
additional noise picked up in this step is attributed to
leadwire and/or gage pickup. If the output changes
when the instrument chassis is touched with a nger,
this is an indication of a poor ground and/or radio-
frequency interference.
4. Apply a load to the part under test (with excitation still
off). If additional noise is observed, the noise is due to
something associated with the loading system such as
a motor creating a magnetic eld, or the motion of the
gage or wiring (generating emf).
5. If possible, remove the load from the test part and apply
excitation voltage to the bridge circuit. After balancing
the bridge, any subsequent change in output, if gradual,
is zero-shift, not noise. This may be due to gage self-
heating effects (see Tech Note TN-502, Strain Gage
Excitation Levels)or other time-dependent resistance
changes.
The following sections of this Tech Note give recommended
noise-reduction procedures for electrostatic noise, and for
magnetic noise.
Electrostatic Noise Reduction
The simplest and most effective barrier against electrostatic
noise pickup is a conductive shield, sometimes referred to
as a Faraday cage. It functions by capturing the charges that
would otherwise reach the signal wiring. Once collected,
these charges must be drained off to a satisfactory
ground (or reference potential). If not provided with a
low-resistance drainage path, the charges can be coupled
into the signal conductors through the shield-to-cable
capacitance (Figure 3).
Figure 2. Electromagnetic noise coupling.
Figure 3. Electrostatic shielding.
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
3
The two most popular types of cable shields are braided
wire and conductive foil. The braided-shield construction
provides about 95 percent coverage of the cable, and is
characteristically low in resistance. Although commonly
higher in resistance, foil shields give 100 percent cable
coverage, and are also easier to terminate. Following
are commercially available examples of the two types of
shielded cable:
braided: Micro-Measurements Type 430-FST
(four conductors, twisted)
foil: Micro-Measurements Type 422-DSV
When long reaches of multiple conductors are run adjacent
to each other, problems with crosstalk between conductors
can be encountered. With runs of 50 feet [15 m] or more,
signicant levels of noise can be induced into sensitive
conductors through both magnetic and electrostatic
coupling. Even though bridge-excitation conductors may
carry only a millivolt of noise, there can be signicant
coupling to signal conductors to produce potentially
troublesome microvolt-level noise in those conductors.
The noise transfer can be minimized by employing an
instrumentation cable composed of individually shielded
pairsone pair for excitation, and one pair for the
signal. This type of construction is embodied in Micro-
Measurements Type 422-DSV cable. When using such
cable (those having separate shields), both shields should
be grounded at the same, usually instrument, end of the
cable. Electromagnetic coupling between excitation and
signal pairs can be reduced somewhat by using a cable that
has its conductor pairs twisted on separate axes. Belden
No. 8730 cable has the conductor pairs separately twisted,
including one pair shielded with foil.
The shield-to-conductor capacitance can also become
signi f icant for long runs, si nce the capacitance is
proportional to the cable length. Therefore, a signicant
portion of the residual noise can be coupled from even
a well-grounded shield to the sensitive conductors. To
minimize this effect, some strain gage instruments (for
example, Micro-Measurements 2300 System) incorporate
a feature called a driven guard. A driven guard (also
known as a driven shield) functions by maintaining the
shield at a voltage equal to the average signal, or common-
mode voltage. Since, with this arrangement, the voltage
difference between the conductors and shield is essentially
zero, the effective capacitance is decreased, and there is
minimal noise transfer. The result is a very quiet shield. It
is important to note that, for proper operation, the driven
shield is connected at only one end to the driven-guard pin
on the instrument input connector. The driven shield is
ordinarily surrounded by a second shield, which should be
grounded at one end.
In a fully guarded amplier system (for example, Micro-
Measurements Model 2200 System), the common-mode
voltage of the bridge excitation supply and the signal
input terminals foat to the level on the guard shield.
Connecting the shield to the test structure or source of
common-mode voltage at the gage installation site can
provide very effective noise reduction since the voltage
between signal conductors and the shield is minimized.
Another often-overlooked source of noise is leakage
to ground through the strain gage and/or the cabling.
If excessive, this leakage can cause noise transfer from
the specimen to the gage circuit, since even supposedly
well-grounded specimens may carry some noise. It is not
uncommon to have strain gages installed on nominally
grounded test objects that, in fact, have noise levels
expressible in volts. And, of course, any strain gage
installation on a conductive specimen forms a classic
capacitor which can couple noise from the specimen to the
gage. In light of these considerations, it is always a good
practice to make certain that the specimen is properly
grounded and that leakage between the gage circuit and
the specimen is well within bounds.
Prior to connecting leadwires to the strain gage, the
insulation resistance from the gage to the specimen should
be measured with a megohm meter such as the Micro-
Measurements Model 1300 Gage Installation Tester.
A reading of 10 000 meg ohms is normally considered a
minimum for satisfactory system operation. Readings
below this level are indicative of a possibly troublesome
gage installation which can deteriorate with time and
strain. It should also be kept in mind, for gage installations
which will operate at elevated temperatures, that leakage
resistance tends to decrease as the temperature increases.
After cable placement and connection at the gage-end of
the cable, the following resistance measurements should
be made, preferably from the instrument-end of the cable:
conductor-to-ground, shield-to-ground, and conductor-to-
shield. Because of distributed leakage, these resistances may
be somewhat lower than the gage-to-specimen resistance;
but cables with signicantly lower resistances should be
investigated, and the excessive leakage eliminated to avoid
potential noise problems.
Electromagnetic Noise Reduction
The most effective approach to minimizing magnetically
induced noise is not to attempt magnetic shielding of the
sensitive conductors; but, instead, to ensure that noise
voltages are induced equally in both sides of the amplier
input (Figure 4). When analyzed, all conventional strain
gage bridge arrangementsquarter bridge (two- or
three-leadwire), half bridge, and full bridgereduce to
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
4
the same basic circuit shown in Figure 4. This is also true
for systems that employ the rotated or nonsymmetrical
bridge circuit. Achievement of noise cancellation by the
method shown in Figure 4 requires that the amplifier
exhibit good common-mode rejection characteristics.
Attention must also be given, however, to the strain gage
wiring, and to the effects of nearby power lines. For
example, it is evident from Figure 2 that a gradient in
magnetic eld intensity exists with respect to distance from
the current-carrying power line. The series noise voltages
(V
1
and V
2
) induced in the signal wires will therefore depend
greatly upon their distances from the current-carrying
conductors. Twisting the signal conductors together tends
to make the distances equal, on the average, thereby
inducing equal noise voltages which will cancel each other.
Correspondingly effective, the magnetic eld strengths
radiated by power lines can be reduced by twisting the
power conductors.
In theory, at least, the more twists per unit conductor
length, the better. Standard twisted-conductor cables,
such as Belden No. 8771, have sufficient twisting for
most applications. How ever, in environments with high
magnetic field gradients, such as those found close to
motors, generators, and transformers, tighter twisting
may be required. For particularly severe applications,
conventional twisting may be inadequate, and it may be
necessary to use a special woven cable as described later.
When attaching leadwires to a strain gage for operation in
a magnetic eld, connections should be made directly to
the solder tabs on the gage, rather than through auxiliary
terminals. Micro-Measurements CEA-Series gages, with
copper-coated solder tabs, are particularly suited to
this type of application. As shown in Figure 5, the gage
selection and the wiring arrangements can greatly affect
the sensitivity to magnetic pickup. It will be noticed that
the preferred arrangement minimizes the susceptible loop
area between the wires. For the same reason, fat ribbon
cable is very prone to noise pickup, and its use in magnetic
fields should be avoided. When necessary to use this
type of cable, optimal conductor allocation, as shown in
Figure 6, can help reduce the pickup. In addition, excess
lengths of input cable should be eliminated; and under
no circumstances should the extra length be disposed
of by winding into a coil as illustrated in Figure 7a. If
excess cable length cannot be avoided, it should be folded
in half and coiled as indicated in Figure 7b so that each
clockwise current loop is intimately accompanied by a
counterclockwise loop. The same cabling considerations
apply to both the excitation leads and the signal leads, and
to power cables.
Figure 5. Gage selection and wiring technique.
Figure 6. Cable comparison.
Figure 7. Handling excess cable.
Figure 4. Noise cancellation by
amplier common-mode rejection.
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
5
Unlike the case for electrostatic noise, a simple, grounded
conductive shield does not function as a barrier to magnetic
noise. Magnetic shields operate on a different principle,
and serve to bend or shunt the magnetic eld around the
conductor rather than eliminate it. Magnetic shields are
made from high-permeability materials such as iron and
other ferro-magnetic metals. At the relatively low 50/60-Hz
power line frequencies often encountered in magnetic
noise problems, shield thicknesses (using common iron for
example) on the order of 0.1 in [2.5 mm] are needed before
signicant noise reduction is achieved. Heavy-walled iron
conduit can also be used to provide some reduction in
magnetic noise pickup. However, there are special high-
permeability alloys (mu-metal
, for instance) that have been

developed specically for magnetic shielding purposes.
These are effective in much thinner shields than with iron.
When faced with the apparent necessity for magnetic
shielding, attention should always be given to reducing
the noise at its source. As an example, transformers can
readily be designed to minimize the leakage fux.
Severe Noise Environments
The preceding two sections have treated the standard
methods of noise reduction applicable to the majority
of instrumentation problems. This section describes
techniques that may become necessary when very high
noise levels are anticipated or experienced.
Electrostatic Fields
General ly, when shieldi ng agai nst audio-frequency
electrostatic noise (below 20 kHz), it is not good practice
to ground the shield at more than one point. The reason for
this is that the ground points may be at different voltage
levels, causing current to fow through the shield. Current
fow in such ground loops can induce noise in the signal-
carrying conductors through the same phenomenon that
occurs in a transformer.
However, for long cables in severe noise environments, the
shield impedance from one end to the other can become
signicant, particularly with high-frequency noise sources.
When this occurs, the noise charges captured by the shield
no longer nd a low-resistance drain to ground, and the
result is a noisy shield. Improved shield performance
under such circumstances can often be obtained by
grounding the shield at both ends, and/or at intermediate
pointspreferably at points near any localized sources
of electrostatic noise. Multiple-point ground connections
may also be necessary when radio-frequency interference
(RFI) problems are encountered. At these frequencies
the shield, or segments of the shield between grounded
points, can display antenna behavior. By experimentally
grounding the shield at numerous points along its length,
the optimum grounding scheme can be determined.
Although the leadwires are ordinarily the dominant
medium for noise induction in a strain gage circuit, noise
pickup can also occur in the gage itself. When needed, a
simple electrostatic shield can be fabricated by forming
an aluminum-foil box over the gage and the unshielded
leadwire terminations. If the gaged specimen is small and
electrically conductive, aluminum tape with conductive
adhesive should be used to connect the cable shield, the
gage shield, and the specimen together. Conductive epoxy
compounds can also be used for this purpose.
On the other hand, when gages are installed on machinery
or other large, conductive test objects, care must be
exercised to prevent the occurrence of ground current
loops in the shield. In such cases, the foil should be
electrically insulated from the machine. But the machine
should be grounded with a heavy-gauge copper wire (at
least 14 gauge or heavier depending upon application)
connected to the single-point ground near the instrument.
Care must also be taken to make certain that the shield
does not form a short circuit to the gage wiring. If the
cable has two shields, then, ideally at least, a double-foil
shield should be used over the strain gage. The two shields
should be connected together only at the instrument end
of the cable.
A word about ground connections is in order. It is important
to remember that all conductors are characterized by
resistance, inductance, and shunt capacitance. As a result,
attention should always be given to the quality of the
ground connections. To be effective, a connection to
ground should be made with heavy-gauge copper wire,
and should be as short as practicable. If the nearest earth
ground is too remote, a 6-ft [2-m] copper rod can be driven
into the earth to establish a local ground.
Electromagnetic Fields
As with electrostatic noise pickup, the leadwires commonly
represent the principal source of magnetic noise induction
in strain gage circuits. In intense electromagnetic elds
with steep gradients (near motors, generators, and similar
equipment), ordinary wire-twisting techniques may prove
inadequate. An end view of a conventionally twisted
pair can reveal the reason for pickup. As indicated in
Figure 4, even if the induced noise were precisely equal
in both wires, the amplier noise output would be zero
only if the amplier had innite common-mode rejection
characteristicsan impossibility. In order to minimize
common-mode noise voltages, a special, woven, four-
wire cable has been designed that, as seen from the wire
Allegheny Ludlum Steel Co.

T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
6
end, eliminates the spiral inductive loops (Figure 8). For
maximum cancellation of electrostatic elds, pairs of wires
(composed of one wire from each plane) are connected in
parallel. Referring to the gure, wires 1 and 2 are paralleled
to form one conductor; and wires 3 and 4 to form the other.
So connected, this type of cable is largely insensitive to
magnetic eld gradients, both parallel and perpendicular
to the cable length. The cable is known as Inter-8 Weave,
and is available from: Magnetic Shield Corp., 740 N.
Thomas Drive, Bensensville, Illinois 60106.
Even though the strain gage is much less frequently the
significant medium for magnetic noise induction than
the leadwires, different gage patterns have differing
sensitivities to noise pickup. For instance, if the gage has
both solder tabs at one end, the net noise pickup is less than
for a gage with one tab at each end. As shown in Figure 5,
the difference in noise sensitivity results from the relative
size of the inductive loop area in each case. It is also worth
noting that smaller gages, with more closely spaced grid
lines, are intrinsically quieter than large gages.
In severe magnetic fields, especially those with steep
gradients in eld intensity, additional measures may be
required. For this purpose, Micro-Measurements has
developed a special gage configuration, the H-Series,
consisting of two identical grids, with one stacked directly
above, and insulated from, the other. By connecting the
upper and lower gage elements in series so that the current
fows in opposite directions through the two grids, the noise
induced in the assembly tends to be self-cancelling. This
arrangement is particularly effective against magnetic eld
gradients and their components parallel to the test surface.
The dual-element gage is intended to function as one arm
of a Wheatstone bridge circuit; and the bridge is usually
completed with another gage of the same type, or with a
xed precision resistor. Standard practices are followed
when instal ling the gages; but the Micro-Measurements
M-Bond 600/610 adhesive system is recommended for
bonding, since this will result in the thinnest glueline,
and placement of the grids as close as possible to the
specimen surface. Available from Micro-Measurements
are two types of dual-element, noninductive stacked
gageslinear H06A-AC1-125-700 and a three-gage rosette
H06A-AD3-125-700. See our Precision Strain Gages Data
Book for details.
In addition to the strain gage size and pattern, the selection
of the gage grid alloy should be given careful consideration.
If the grid alloy is magnetic, it will be subject to extraneous
physical forces in a magnetic eld; and, if magnetoresistive,
will undergo spurious resistance changes. Similarly, if
the alloy is magnetostrictive, the grid will try to change
length in the magnetic eld. Isoelastic alloy, for example,
should not be used in magnetic fields, since it is both
strongly magnetoresistive and magnetostrictive. Stemming
from their comparative freedom from magnetic effects,
constantan and Karma-type alloys are usually selected
for such applications. Constantan, however, at cryogenic
temperatures and in high magnetic fields (7-70 Tesla)
becomes severely magnetoresistive. The Karma-type alloy
is ordinarily preferred for cryogenic service because of its
generally superior performance in magnetic elds at very
low temperatures.
When necessary, strain gages can also be shielded from
electromagnetic fields to some degree with a magnetic
shielding material such as mu-metal. Two or more layers
of the shielding material may be required to effect a
noticeable improvement. Of course, even this will be
ineffective if the source of the magnetic eld is beneath the
strain gage. When high-frequency elds are encountered,
be sure that the material is suitable (high permeability) at
the anticipated frequency.
Suggested Additional Reading
Coffee, M.B., Common-mode Rejection Techniques for
Low-Level Data Acquisition. Instrumentation Technology
24, No. 7: 45-49 (1977).
Ficchi, R.F., Practical Design for Electromagnetic Com-
patibility. New York: Hayden Book Company, 1971.
Freynik, H.S., et. al., Nickel-Chromium Strain Gages for
Cryogenic Stress Analysis of Super-Conducting Structures
Figure 8. Woven cable to reduce severe electromagnetic
radiation and pickup.
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
7
in High Magnetic Fields. Proceedings of the Seventh
Symposium on Engineering Problems of Fusion Research,
October, 1977.
Hayt, W.H., Jr., Engineering Electromagnetics. New York:
McGraw-Hill Book Company, 1967.
Klipec, B.E., How to Avoid Noise Pickup on Wire and
Cable. Instruments & Control Systems 50, No. 12: 27-30
(1977).
Krigman, Alan, Sound and Fury: The Persistent Problem
of Electrical Noise. In-Tech 32, No. 1: 9-20 (1985).
(Extensive bibliography).
McDermott, Ji m, EMI Shieldi ng and Protective
Components. EDN 24, No. 16: 165-176 (1979).
Morrison, Ralph, Grounding and Shielding Techniques in
Instrumentation, 2nd Ed. New York: John Wiley & Sons,
Inc., 1977.
Severinsen, J., Gaskets that Block EMI. Machine Design
47, No. 19: 74-77 (1975).
Sitter, R.P., RFI What It Is and How to Control It, Part
II: Reduction of Interference. Instrumentation Technology
25, No. 10: 59-65 (1978).
*Stein, Peter K., Spurious Signals Generated in Strain
Gages, Thermocouples and Leads. Lf/MSE Publication
No. 69, April 1977.
*Stein, Peter K., The Response of Transducers to Their
Environment, The Problem of Signal and Noise. Lf/MSE
Publication No. 17, October 1969.
Strain Gages Operate in 50 000-Gauss Magnetic Fields For
Fusion Research. Epsilonics (published by Measurements
Group, Inc.) II, No. 3: 4 (1982).
Whit e, D.R. J., El ectromagneti c Interference and
Compatibility, Vol. 3, Germantown, Maryland: Don White
Consultants, 1973.
*Available from: Stein Engineering Services, 5602 E.
Monte Rosa, Phoenix, Arizona 85018.
Introduction
A common request in strain gage work is to obtain the
recommended value of bridge excitation voltage for a
particular size and type of gage. A simple, denitive answer
to this question is not possible, unfortunately, because
factors other than gage type are involved. The problem is
particularly difcult when the maximum excitation level is
desired.
This Tech Note is intended to outline the most signicant
considerations that apply, and to suggest specic ap proaches
to optimizing excitation levels for various strain gage
applications.
It is important to realize that strain gages are seldom
damaged by excitation voltages considerably in excess of
proper values. The usual result is performance degradation,
rather than gage failure; and the problem therefore becomes
one of meeting the total requirements of each particular
installation.
Thermal Considerations
The voltage applied to a strain gage bridge creates a power
loss in each arm, all of which must be dissipated in the
form of heat. Only a negligible fraction of the power input
is available in the output circuit. This causes the sensing
grid of every strain gage to operate at a higher temperature
than the substrate to which it is bonded. With exceptions,
which are discussed later, it can be considered that the
heat generated within a strain gage must be transferred by
conduction to the mounting surface. The heat ow through
the specimen causes a temperature rise in the substrate,
which is a function of its heat-sink capacity and the gage
power level.
Consequently, both sensing grid and substrate operate at
temperatures higher than ambient. When the temperature
rise is excessive, gage performance will be affected as
follows:
1. A loss of self-temperature-compensation (S-T-C) occurs
when the grid temperature is considerably above the
specimen temperature. All manufacturers data on
S-T-C are necessarily obtained at low excitation levels.
2. Hysteresis and creep effects are magnied, since these
are dependent on backing and glueline temperatures.
A gage backing normally rated at +250F [+120C] in
transducer service might have to be derated by 20 to
50F [10 to 30C] under high-excitation conditions.
3. Zero (no-load) stability is strongly affected by excessive
excitation. This is particularly true in strain gages with
high thermal output characteristics, and when inherent
half-bridge or full-bridge compensation is relied upon
to meet a low zero-shift vs. temperature specication.
The zero-shift occurs because of variation in heat-sink
conditions between gages in the bridge circuit.
Another point should be emphasized. Any tendency for
localized areas of the grid to operate at higher temperatures
than the rest of the grid will restrict the allowable excitation
levels. Creep and instability are particularly susceptible to
these hot-spot effects, which are usually due to voids or
bubbles in the glueline or discontinuities in the substrate.
Imperfections in the gage itself can cause hot spots to
develop, and only gages of the highest quality should be
considered for high-excitation applications.
When other factors are constant, the power-dissipation
capability of a strain gage varies approximately with the
area of the grid (active gage length x active grid width). The
amount or type of waterproong compound or encapsulant
is relatively unimportant. Open-face gages mounted on
metal show only 10 to 15% less power-handling capacity
than fully encapsulated gages with the same grid area. Note,
however, that proper waterproong materials must always
be applied to open-face gages to prevent loss of performance
through grid corrosion.
It is sometimes stated that gage adhesives of high thermal
conductivity can considerably improve the power-handling
capability of strain gage installations. Generally, this is not
correct. These adhesives incorporate high-conductivity llers
such as aluminum oxide and metal powders. This produces
an adhesive of high viscosity, resulting in excessively thick
gluelines and a longer thermal path from gage to substrate.
Any net gain in thermal conductivity is more than offset by
the performance degradation due to thicker gluelines. It is
much better, for high gage excitation as well as normal gage
applications, to use high-functionality adhesives that permit
thin, void-free gluelines. On smooth mounting surfaces,
ideal glueline thicknesses range from 0.0001 to 0.0003 in
[0.0025 to 0.0075 mm].
Tech Note TN-502
MICRO-MEASUREMENTS
Optimizing Strain Gage Excitation Levels
T
E
C
H

N
O
T
E
11
Factors Affecting Optimum Excitation
Following are factors of primary importance in determining
the optimum excitation level for any strain gage
application:
1. Strain gage grid area (active gage length x active grid
width).
2. Gage resistance. High resistances permit higher voltages
for a given power level.
3. Heat-sink properties of the mounting surface. Heavy
sections of high-thermal-conductivity metals, such as
copper or aluminum, are excellent heat sinks. Thin
sections of low-thermal-conductivity metals, such as
stainless steel or titanium, are poor heat sinks. Also, the
shape of the gaged part may create thermal stresses in
portions of the structure due to gage self-heating. Long
warm-up times and apparent gage instability can result.
The situation often arises in low-force transducers,
where thin sections and intricate machining are fairly
common.
Strai n measurement on plastic requi res special
consideration. Most plastics act as thermal insulators
rather than heat sinks. Extremely low values of
excitation are required to avoid serious self-heating
effects. The modulus of elasticity of the common
plastics drops rapidly as temperature rises, increasing
viscoelastic effects. This can signicantly affect the
material properties in the area under the strain gage.
Plastics that are heavily loaded with inorganic llers
in powder or brous form present a lesser problem,
because such fillers reduce expansion coefficients,
increase the elastic modulus, and improve thermal
conductivity.
4. Environmental operating temperature range of the gage
installation. Creep in the gage backing and adhesive
will occur at lower ambient temperatures when grid
and substrate temperatures are raised by self-heating
effects. Thermal output due to temperature will also
be altered when grid and substrate temperatures are
signicantly different.
5. Required operational specications. Gages for normal
stress analysis can be excited at a higher level than
under transducer conditions, where the utmost in
stability, accuracy, and repeatability are needed.
A signicant distinction exists between gages used
in dynamic strain measurement and those used in
static measurement applications. All the various
performance losses due to gage self-heating affect
static characteristics of the gage much more seriously
than the dynamic response. Therefore, it is practical
to drive the dynamic installations much harder, and
thus take advantage of the higher signal-to-noise ratio
that results.
6. Installation and wiring technique. If the gage is damaged
during installation, if solder tabs are partially unbonded
due to soldering heat, or if any discontinuities are
formed in the glueline, high levels of excitation will
create serious problems. Proper technique is essential
in obtaining consistent performance in all strain
gage work, but particularly under high-excitation
conditions.
In addition to the preceding, secondary factors can affect
maximum permissible excitation levels. Poor grid design,
such as improper line-to-space ratio, will reduce the heat
transfer effectiveness. The type of gage matrix, in terms of
resin and ller, determines the thermal conductivity of the
backing. The backing is usually more important than the
adhesive selected because the adhesive layer is thinner than
the backing in proper installations.
Stacked Rosette Gages
These represent a special case, because the thermal path
length is much greater from the upper grid to the substrate,
and because the temperature rise of the lower grids adds
directly to those above. For a three-element stacked rosette
in which the three grids are completely superimposed, the
top grid will have six times the temperature rise of a similar
single gage, if all grids receive the same input power. To keep
the temperature rise of the top grid equal to that of a similar
single gage, the three rosette sections should each receive 1/6
of the power applied to the single gage. This corresponds to
a reduction factor of 2.5 for bridge excitation voltage, since
power varies as the square of the applied voltage. For two-
element stacked rosettes, the comparable de rating factor is 3
for power, and 1.7 for bridge voltage.
This discussion is based on rosettes of square grid geometry,
where each grid covers essentially all of the grid(s) in the
assembly. When substantial areas of the grids are not
superimposed, the derating factors mentioned above will be
somewhat conservative.
Cryogenic Gage Applications
Many strain gage measurements are made under direct
submersion in liqueed gases such as nitrogen, hydrogen, and
helium. Since these liquids are electrically nonconductive,
open-face gages have been used occasionally without a
protective or waterproong coating. An interesting effect
has been reported under these conditions. If excitation
voltages are not kept sufficiently low, grid self-heating
will cause gas bubbles to form on the gridlines, and thus
partially insulate the grid from the cold liquid. Larger
bubbles are then created by increased grid temperatures,
and bubbles periodically break loose and rise toward the
surface. The relative motion of these insulating bubbles with
respect to the gridlines produces local temperature changes,
which appear in the output signal as noise. Grid alloys that
display very high values of thermal output at cryogenic
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
12
temperatures (most constantan alloys, for example) are
particularly susceptible to this effect. The remedy is to utilize
very low excitation levels, and/or to use protective coatings
over the grid to prevent direct liquid contact. Such coatings
must necessarily retain sufficient flexibility at cryogenic
temperatures to prevent cracking of the protective layer.
Experimental Determination of
Maximum Gage Excitation
To be certain that the excitation level chosen for a given
strain gage application is not excessive, it is necessary to
run performance tests at the maximum environmental
temperature. In many cases, however, this rather complicated
procedure can be greatly simplied by gradually increasing
the bridge excitation under zero-load conditions until a
denite zero instability is observed. The excitation should
then be reduced until the zero reading becomes stable
again, without a signicant offset from the low-excitation
zero reading. For most applications in experimental stress
analysis, this value of bridge voltage is the highest that can
be used safely without signicant performance degradation.
Conducting this test at the maximum operating temperature
instead of room temperature will increase the likelihood that
the maximum safe bridge voltage has been established.
The rigid operating requirements for precision transducers
make the above procedure useful primarily as a first
approximation; and further verication is usually required.
The performance tests most sensitive to excessive excitation
voltage are: (1) zero-shift vs. temperature and (2) stability
under load at the maximum operating temperature.
Excitation Levels for
Resistance Temperature Sensors
It has become increasingly common to measure specimen
temperatures in strain gage work by the use of bondable
nickel-grid temperature sensors such as the ETG-50 and
WTG-50. These sensors are fabricated in the same manner
as strain gages, and consequently experience environmental
temperature changes in the same way. By eliminating
many of the measurement errors often encountered with
thermocouples, temperature sensors are ideal for correcting
strain gage data under rapidly changing temperature
conditions.
Like strain gages, temperature sensors are adversely
affected by excessive excitation levels. Variation in heat-sink
conditions and accuracy requirements make universally
applicable excitation recommendations impossible, but a
simple test procedure is available. The excitation level should
be increased until the readout device indicates an excessive
grid temperature rise; it should then be reduced as necessary.
Since the readout in this case shows temperature measurement
error directly, the determination is straightforward.
Since temperature sensors are most often used with
linearization networks of the LST type, it is not normally
necessary to check for errors due to excitation level. These
networks greatly attenuate the input bridge voltage, and the
sensors are therefore operated at very low power levels.
Typical Strain Gage Excitation Values
The data curves on pages 15-17 represent general
recommendations or starting points for determining optimum
excitation levels. These curves are plots of bridge excitation
voltage vs. grid area (active gage length x grid width) for
constant power-density levels in watts/in
2
[or kilowatts/m
2
].
A large number of standard Micro-Measurements single-
element gage patterns are listed at the various grid areas they
represent. Separate plots are provided for gage resistances of
120, 350, and 1000 ohms. For other grid areas and/or other
gage resistances, calculations can be made according to the
following formulas for recommended power-density levels:
Power Dissipated in Grid (watts) = =
E
R
P
B
G
G
2
4
Power Density in Grid (watts/in
2
or kW/m
2
) = =
P
A
P
G
G
G
where: R
G
= Gage resistance in ohms
A
G
= Grid area (active gage length x grid width)
E
B
= Bridge excitation in volts
Note that bridge voltage (E
B
) is based on an equal-
arm bridge arrangement, where the voltage across
the active arm is one-half the bridge voltage.
When grid area (A
G
), gage resistance (R
G
), and grid power
density (P'
G
) are known:

E R P A
B G G G
= 2 x x
Grid Power-Density Curves
Selecting the most appropriate power-density lines on the
following charts depends, primarily, on two considerations:
degree of measurement accuracy required, and substrate
heat-sink capacity. A series of general recommendations
follows, but should be veried by the procedures previously
described for critical applications.
Typical Power-Density Levels in Watts/in
2
[kW/m
2
]
It is of interest that some commercial strain measurement/
instrumentation utilizes constant excitation voltage of 3
to 5 volts. The power densities created in gages of various
sizes and resistances by these bridge voltages can be taken
directly from the charts and compared with Table 1. For very
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
13
E
R A
R
R R
P
B
G G
G
G D
G
2
2
4
= (desired)
R
E R
A P
R
D
B G
G G
G
=
=
2
19 5 . ohms
Table 1 Heat-Sink Conditions
watts/in
2
[kilowatts/m
2
]
R R
R
D G
G
+
=
+
=
19 5 120
120
1 16
.
.
small gages, it is evident that commercial instruments may
require voltage reduction for proper results. A simple circuit
modication, which can be utilized when the instrument
voltage is not adjustable, involves the insertion of dead
resistance in the form on high-precision resistors of VPG
type in series with the active and dummy gages in the bridge
circuit. Power density is then reduced by (multiplied by) the
factor [R
G
/(R
D
+ R
G
)]
2
, where R
D
is the inactive series
resistance in ohms, and R
G
is the active gage resistance in
ohms. Note that the adjacent bridge arm must be increased
by the same R
D
to maintain bridge balance under these
conditions. The sensitivity of the bridge will be decreased
by this procedure, and the readings must be multiplied by
the ratio (R
D
+ R
G
)/R
G
to correct for this desensitization.
Alternately, a shunt calibration resistor can be connected
directly across the dummy gage, and the instrument gage
factor setting adjusted to display the proper calibration
level.
Examples of Chart Use
Case 1: What excitation level can safely be applied to an
EA-09-125AD-120 strain gage, mounted on a 1/2 x 1/16 x 6
in [12.5 x 1.5 x 150 mm] stainless-steel bar, for a static stress
analysis test with moderate accuracy (3 to 5%)?
From Table 1, determine the typical power-density level, 1 to
2 W/in
2
[1.6 to 3.1 kW/m
2
], corresponding to fair heat-sink
condition in stainless steel. Refer to Chart 1:
Enter on the horizontal axis at the arrowhead for the
125AD gage 0.0156 in
2
[10.06 mm
2
].
Mark the intersection of the vertical line with the 1 and
2 W/in
2
[1.6 and 3.1 kW/m
2
] sloped lines.
Read horizontally to the left ordinate for Bridge
Excitation of 2.7 and 3.8 volts, respectively. A strain
indicator with a maximum bridge excitation of 3.8 volts
can be used.
Case 2: Can an instrument with a xed 4.5V excitation be
used? If not, what correction of data points must be made?
To determine the power density in a 125AD gage at a given
excitation level, refer to Chart 1:
Enter the left ordinate at 4.5 volts until intersecting the
abscissa value equivalent to the 125AD gage. The power
density is 2.7 W/in
2
[4.2 kW/m
2
], which is in excess of the
maximum power determined in Case 1. If low accuracy
(i.e., 5 to 10% data) is acceptable, the higher P'
G
can be
used. If greater accuracy must be maintained, several
alternatives are available: (1) select a higher resistance
gage, (2) select a gage with a larger area, or (3) reduce
the bridge voltage with an inactive series resistor, R
D
.
The inactive resistor (R
D
) required to reduce the power
density to the desired 2 W/in
2
(3.1 kW/m
2
), with a given
EB, can be determined from the following relationship:

E
R A
R
R R
P
B
G G
G
G D
G
2
2
4
= (desired)
For a desired P'
G
of 2 W/in
2
(3.1 kW/m
2
)

R
E R
A P
R
D
B G
G G
G
=
=
2
19 5 . ohms
Select the nearest precision resistor value greater than
19.5 ohms for R
D
.
For actual strain values, accounting for the inserted
inactive resistor, all indicated strain readings must be
multiplied by:
R R
R
D G
G
+
=
+
=
19 5 120
120
1 16
.
.
Accuracy
Requirements
EXCELLENT
Heavy Aluminum
or
Copper Specimen
GOOD
Thick Steel
FAIR
Thin
Stainless Steel
or Titanium
POOR
Filled Plastic
such as
Fiberglass/Epoxy
VERY POOR
Unlled Plastic
such as Acrylic
or Polystyrene
High
25
[3.17.8]
12
[1.63.1]
0.51
[0.781.6]
0.10.2
[0.160.31]
0.010.02
[0.0160.031]
Moderate
510
[7.816]
25
[3.17.8]
12
[1.63.1]
0.20.5
[0.310.78]
0.020.05
[0.0310.078]
Low
1020
[1631]
510
[7.816]
25
[3.17.8]
0.51
[0.781.6]
0.050.1
[0.0780.16]
High
510
[7.816]
510
[7.816]
25
[3.17.8]
0.51
[0.781.6]
0.010.05
[0.0160.078]
Moderate
1020
[1631]
1020
[1631]
510
[7.816]
12
[1.63.1]
0.050.2
[0.0780.31]
Low
2050
[3178]
2050
[3178]
1020
[1631]
25
[3.17.8]
0.20.5
[0.310.78]
S
T
A
T
I
C
D
Y
N
A
M
I
C
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
14
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
15
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
16
T
E
C
H

N
O
T
E
TN-502
Micro-Measurements
17
I. Residual Stresses and Their Measurement
Residual (locked-in) stresses in a structural material or
component are those stresses that exist in the object without
(and usually prior to) the application of any service or
other external loads. Manufacturing processes are the
most common causes of residual stress. Virtually all
manufacturing and fabricating processes casting, welding,
machining, molding, heat treatment, etc. introduce
residual stresses into the manufactured object. Another
common cause of residual stress is in-service repair or
modication. In some instances, stress may also be induced
later in the life of the structure by installation or assembly
procedures, by occasional overloads, by ground settlement
effects on underground structures, or by dead loads which
may ultimately become an integral part of the structure.
The effects of residual stress may be either beneficial
or detrimental, depending upon the magnitude, sign,
and distribution of the stress with respect to the load-
induced stresses. Very commonly, the residual stresses are
detrimental, and there are many documented cases in which
these stresses were the predominant factor contributing
to fatigue and other structural failures when the service
stresses were superimposed on the already present residual
stresses. The particularly insidious aspect of residual stress
is that its presence generally goes unrecognized until after
malfunction or failure occurs.
Measurement of residual stress in opaque objects cannot be
accomplished by conventional procedures for experimental
stress analysis, since the strain sensor (strain gage,
photoelastic coating, etc.) is totally insensitive to the history
of the part, and measures only changes in strain after
installation of the sensor. In order to mea sure residual stress
with these standard sensors, the locked-in stress must be
relieved in some fashion (with the sensor present) so that the
sensor can register the change in strain caused by removal of
the stress. This was usually done destructively in the past
by cutting and sectioning the part, by removal of successive
surface layers, or by trepanning and coring.
With strain sensors judiciously placed before dissecting the
part, the sensors respond to the deformation produced by
relaxation of the stress with material removal. The initial
residual stress can then be inferred from the measured
strains by elasticity considerations. Most of these techniques
are limited to laboratory applications on at or cylindrical
specimens, and are not readily adaptable to real test objects
of arbitrary size and shape.
X-ray diffraction strain measurement, which does not
re quire stress relaxation, offers a nondestructive alternative
to the foregoing methods, but has its own severe limitations.
Aside from the usual bulk and complexity of the equipment,
which can preclude eld application, the technique is limited
to strain measurements in only very shallow surface layers.
Although other nondestructive techniques (e.g., ultrasonic,
electromagnetic) have been developed for the same purposes,
these have yet to achieve wide acceptance as standardized
methods of residual stress analysis.
The Hole-Drilling Method
The most widely used modern technique for measuring
residual stress is the hole-drilling strain-gage method of
stress relaxation, illustrated in Figure 1.
Briey summarized, the measurement procedure involves
six basic steps:
A special three- (or six-) element strain gage rosette is
installed on the test part at the point where residual
stresses are to be determined.
The gage grids are wired and connected to a multi-
channel static strain indicator, such as the Micro-
Measurements Model P3 (three-element gage), or
System 5000 (six-element gage).
Figure 1. Hole-Drilling Strain Gage Method
Strain Gage
Rosette
Drilled Hole
* Drilling implies all methods of introducing the hole (i.e., drilling,
milling, air abrasion, etc).
Tech Note TN-503
MICRO-MEASUREMENTS
Measurement of Residual Stresses
by the Hole-Drilling* Strain Gage Method
T
E
C
H

N
O
T
E
19
A precision milling guide (Model RS-200, shown in
Figure 1) is attached to the test part and accurately
centered over a drilling target on the rosette.
After zero-balancing the gage circuits, a small, shallow
hole is drilled through the geometric center of the
rosette.
Readings are made of the relaxed strains, corresponding
to the initial residual stress.
Usi ng special data-reduction relationships, the
principal residual stresses and their angular orientation
are calculated from the measured strains.
The foregoing procedure is relatively simple, and has been
standardized in ASTM Standard Test Method E 837.
1

Using commercially available equipment and supplies, and
adhering to the recommendations in the ASTM standard,
the hole-drilling method can be applied routinely by
any qualied stress analysis technician, since no special
expertise is required for making the measurements. The
method is also very versatile, and can be performed in
either the laboratory or the eld, on test objects ranging
widely in size and shape. It is often referred to as a semi-
destructive technique, since the small hole will not, in many
cases, signicantly impair the structural integrity of the
part being tested (the hole is typically
1
32 to
3
16 in [0.8 to
4.8 mm] in both diameter and depth). With large test objects,
it is sometimes feasible to remove the hole after testing is
completed, by gently blending and smoothing the surface
with a small hand-held grinder. This must be done very
carefully, of course, to avoid introducing residual stresses in
the process of grinding.
NOTE 1: In its current state of development, the hole-
drilling method is intended primarily for applications in
which the residual stresses are uniform throughout the
drilling depth, or essentially so. While the procedures for
data acquisition and reduction in such cases are well-
established and straightforward, seasoned engineering
judgment is generally required to verify stress uniformity and
other criteria for the validity of the calculated stresses. This
Tech Note contains the basic information for understanding
how the method operates, but cannot, of course, encompass
the full background needed for its proper application in all
cases. An extensive list of technical references is provided in
the Bibliography as a further aid to users of the method.
NOTE 2: Manual calculation of residual stresses from the
measured relaxed strains can be quite burdensome, but there
is available a specialized computer program, H-DRILL, that
completely eliminates the computational labor.
II. Principle and Theory of the
Hole-Drilling Strain Gage Method
The introduction of a hole (even of very small diameter)
into a residually stressed body relaxes the stresses at that
location. This occurs because every perpendicular to a
free surface (the hole surface, in this case) is necessarily a
principal axis on which the shear and normal stresses are
zero. The elimination of these stresses on the hole surface
changes the stress in the immediately surrounding region,
causing the local strains on the surface of the test object to
change correspondingly. This principle is the foundation for
the hole-drilling method of residual stress measurement,
rst proposed by Mathar.
2
In most practical applications of the method, the drilled
hole is blind, with a depth which is: (a) about equal to
its diameter, and (b) small compared to the thickness of
the test object. Unfortunately, the blind-hole geometry is
sufciently complex that no closed-form solution is available
from the theory of elasticity for direct calculation of the
residual stresses from the measured strains except by
the introduction of empirical coefcients. A solution can
be obtained, however, for the simpler case of a hole drilled
completely through a thin plate in which the residual stress
is uniformly distributed through the plate thickness. Because
of this, the theoretical basis for the hole-drilling method
will rst be developed for the through-hole geometry, and
subsequently extended for application to blind holes.
Through-Hole Analysis
Depicted in Figure 2a (following) is a local area within a thin
plate which is subject to a uniform residual stress, o
x
. The
initial stress state at any point P (R, o) can be expressed in
polar coordinates by:

+ ( )
( )
r
x
x
r
x
2
1 2
2
1 2
2
2
cos
cos
sin

Figure 2b represents the same area of the plate after a
small hole has been drilled through it. The stresses in
the vicinity of the hole are now quite different, since o
r

and t
r
must be zero everywhere on the hole surface. A
solution for this case was obtained by G. Kirsch in 1898, and
yields the following expressions for the stresses at the point
P (R, o):
3

( )

j
(
\
,
+ +
j
(
\
,
+
j
(
\
,
+
j
(
\
,
+
j
(
\
,

r
x x
x x
r
x
o
o
o
r r r
r r
r r
r
R
R
R R
R
R
2
1
1
2
1
3 4
2
2
1
1
2
1
3
2
2
1
3 2
2
2 4 2
2 4
4 2
cos
cos
sin

hole radius
arbitrary radius from hole center arbitrary radius from hole center

( )

j
(
\
,
+ +
j
(
\
,
+
j
(
\
,
+
j
(
\
,
+
j
(
\
,

r
x x
x x
r
x
o
o
o
r r r
r r
r r
r
R
R
R R
R
R
2
1
1
2
1
3 4
2
2
1
1
2
1
3
2
2
1
3 2
2
2 4 2
2 4
4 2
cos
cos
sin

hole radius
(1a)
(1b)
(1c)
(2a)
(2b)
(2c)
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
20
Measurement of Residual Stresses by the Hole-Drilling Strain Gage Method

where:

( )

j
(
\
,
+ +
j
(
\
,
+
j
(
\
,
+
j
(
\
,
+
j
(
\
,

r
x x
x x
r
x
o
o
o
r r r
r r
r r
r
R
R
R R
R
R
2
1
1
2
1
3 4
2
2
1
1
2
1
3
2
2
1
3 2
2
2 4 2
2 4
4 2
cos
cos
sin

hole radius

Subtracting the initial stresses from the nal (after drilling)
stresses gives the change in stress, or stress relaxation at point
P (R, o) due to drilling the hole. That is:

r r r
r r r

Substituting Equations (1) and (2) into Equations (3) yields
the full expressions for the relaxed (or relieved) stresses. If
the material of the plate is homogeneous and isotropic in its
mechanical properties, and linear-elastic in its stress/strain
behavior, these equations can then be substituted into the
biaxial Hookes law to solve for the relieved normal strains at
the point P (R, o). The resulting expressions are as follows:
r
o
o
o
r
x
E r r r

+
+
+ ( )
,
,
( ) 1
2
1 3
2
4
1
2
2 4 2
cos cos
]]
]
]
+
+
+ ( )
( )
r
o
o
x
E r r r
1
2
1 3
2
4
1
2 4 2
cos coss2o
,
,
]
]
]
The preceding equations can be written in a simpler
form, demonstrating that along a circle at any radius
R (R R
o
) the relieved radial and tangential strains vary in a
sinusoidal manner:
Comparison of Equations (5) with Equations (4)
demonstrates that coefcients A, B, and C have the following
denitions:
Thus, the relieved strains also vary, in a complex way, with
distance from the hole surface. This variation is illustrated in
Figure 3 on page 22, where the strains are plotted along
the principal axes, at o = 0 and o = 90. As shown
by the figure, the relieved strains generally decrease as
distance from the hole increases. Because of this, it is
desirable to measure the strains close to the edge of the
hole in order to maximize the strain gage output signal.
On the other hand, parasitic effects also increase in the
immediate vicinity of the hole. These considerations, along
with practical aspects of strain gage design and application,
necessitate a compromise in selecting the optimum radius
(R) for gage location. Analytical and experimental studies
have established a practical range of 0.3 < r < 0.45 where
r = R
o
/R and R is the radius to the longitudinal center of
the gage.
It can be noticed from Figure 3 that for o = 0 (along the
axis of the major principal stress) the relieved radial strain,
r
r
, is considerably greater than the tangential strain, r
, in
the specied region of measurement. As a result, commercial
strain gage rosettes for residual stress analysis are normally
designed with radially oriented grids to measure the relieved
radial strain, r
r
. This being the case, only Equation (5a) is
directly relevant for further consideration in this summary.
It is also evident from the gure that the relieved radial
strain along the major principal axis is opposite in sign to
the initial residual stress. This occurs because the coefcients
Z
R
P(R, )
Y
x
o
( )

j
(
\
,
+ +
j
(
\
,
+
j
(
\
,
+
j
(
\
,
+
j
(
\
,

r
x x
x x
r
x
o
o
o
r r r
r r
r r
r
R
R
R R
R
R
2
1
1
2
1
3 4
2
2
1
1
2
1
3
2
2
1
3 2
2
2 4 2
2 4
4 2
cos
cos
sin

hole radius

+ ( )
( )
r
x
x
r
x
2
1 2
2
1 2
2
2
cos
cos
sin

+ ( )
( )
r
x
x
r
x
2
1 2
2
1 2
2
2
cos
cos
sin
R
P(R, )
Y
x
o
X
R
o
( )

j
(
\
,
+ +
j
(
\
,
+
j
(
\
,
+
j
(
\
,
+
j
(
\
,

r
x x
x x
r
x
o
o
o
r r r
r r
r r
r
R
R
R R
R
R
2
1
1
2
1
3 4
2
2
1
1
2
1
3
2
2
1
3 2
2
2 4 2
2 4
4 2
cos
cos
sin

hole radius
X
(a)
(b)
Figure 2. Stress states at P (R, a),
before and after the introduction of a hole.
(3a)
(3b)
(3c)
(4a)
(4b)
r o o
r o o
r x
x
A B
A C
+ ( )
+ ( )
cos
cos
2
2
(5a)
(5b)
A
E r
B
E r r

+
j
(
,
\
,
(

+
+
j
(
,
\
,
(
,
1
2
1
1
2
4
1
1 3
2
2 4
,
]
]
]

+
+
j
(
,
\
,
(
+
,
,
]
]
]
C
E r r
1
2
4
1
1 3
2 4

(6a)
(6b)
(6c)
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
21
A and B in Equation (5a) are always negative, and (for
o = 0) cos 2o = +1.
The preceding treatment considered only the simplest case,
uniaxial residual stress. In practice, however, residual stresses
are more often biaxial, with two nonzero principal stresses.
This condition can readily be incorporated in the analysis by
employing the superposition principle, which is applicable to
linear-elastic material behavior. Referring to Figure 2 again,
it is apparent that had the uniaxial residual stress been along
only the Y axis instead of the X axis, Equations (1) and (2)
would still apply, with cos 2o replaced by cos 2(o + 90), or
by the equivalent, cos 2o. Thus, the relieved radial strain at
the point P(R, o) due to uniaxial residual stress in only the
Y direction can be written as:
And, employing the corresponding notation, Equation (5a)
becomes:
When both residual stresses are present simultaneously,
the superposition principle permits algebraic addition of
Equations (7) and (8), so that the general expression for the
relieved radial strain due to a plane biaxial residual stress
state is:
Or, in a slightly different form,
Equations (9) represent the basic relationship underlying
the hole-drilling method of residual stress analysis. This
relationship must be inverted, of course, to solve for the two
principal stresses and the angle a in terms of the measured
strains that accompany stress relaxation. Since there are
three unknown quantities, three independent measurements
of the radial strain are required for a complete solution.
These three measurements can be substituted successively
into Equation (9a) or Equation (9b) to yield three equations
which are then solved simultaneously for the magnitudes
and directions of the principal stresses.
The common procedure for measuring the relieved strains is
to mount three resistance strain gages in the form of a rosette
around the site of the hole before drilling. Such a rosette is
shown schematically in Figure 4, where three radially oriented
strain gages are located with their centers at the radius R from
the center of the hole site. Although the angles between gages
can be arbitrary (but must be known), a 45-degree angular
increment leads to the simplest analytical expressions, and
thus has become the standard for commercial residual stress
rosettes. As indicated in Figure 4, o
1
is the acute angle from
the nearer principal axis to gage no. 1, while o
2
= o
1
+45 and
o
3
= o
1
+ 90, with positive angles measured in the direction
of gage numbering. It should be noted that the direction of
gage numbering for the rosette type sketched in Figure 4 is
clockwise, since gage no. 2, although physically at position
2a, is effectively at position 2b for gage numbering purposes.
Equations (9) can be used to verify that both locations for
Figure 3. Variation of relieved radial and tangential
strains with distance (along the principal axes) from the
center of the drilled hole uniaxial residual stress.
R
o
1 2 3 4 5
= 0
= 90
R
/
R
o
5

4
3
2
1
+1 0 1
R /R
o
+1
0
1
r o o
r
y
y
A B ( ) cos2
(7)
(8)
r
x
x
A B + ( ) cos2
r o o o o
r o o o
r x y
r x y x
A B A B
A B
+ ( ) + ( )
+
( )
+
cos cos 2 2
( )
o o
y
cos2
(9a)
r o o o o
r o o o
r x y
r x y x
A B A B
A B
+ ( ) + ( )
+
( )
+
cos cos 2 2
( )
o o
y
cos2
(9b)
1
R
45
R
o
1
2b
2a
3
45
Figure 4. Strain gage rosette arrangement
for determining residual stress.
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
22
gage no. 2 produce the same result providing the residual
stress is uniform over the area later occupied by the hole.
For general-purpose applications, location 2a is usually
preferred, because it minimizes the possible errors caused by
any eccentricity of the drilled hole. When space for the gage
is limited, as in measuring residual stresses near a weld or
abutment, location 2b permits positioning the hole closest to
the area of interest.
Equation (9b) can now be written three times, once for each
gage in the rosette:
r
1
= A(o
x
+ o
y
) + B(o
x
o
y
) cos 2o (10a)
r
2
= A(o
x
+ o
y
) + B(o
x
o
y
) cos 2(o+ 45) (10b)
r
3
= A(o
x
+ o
y
) + B(o
x
o
y
) cos 2(o+ 90) (10c)

When Equations (10) are solved simultaneously for the
principal stresses and their direction, the results can be
expressed as:
where o is the angle from the nearer principal axis to gage
no. 1 (in the direction of gage numbering, if positive; or
opposite, if negative).
Reversing the sense of o to more conveniently dene the
angle from gage no. 1 to the nearer axis, while retaining the
foregoing sign convention,
(11c)
The following important comments about Equations
(11) should be carefully noted. These equations are very
similar in appearance to the data-reduction relationships
for conventional strain gage rosettes, but the differences are
signicant. The coefcients A and B not only incorporate
the elastic properties of the test material, but also reect
the severe attenuation of the relieved strains relative to
the relaxed stress. It can be observed, in addition, that
the signs between terms in Equations (11a) and (11b) are
opposite to those in the conventional rosette equations. This
occurs because A and B are always negative; and thus, since
Equation (11a) is algebraically greater than Equation (11b),
the former must represent the maximum principal stress.
Equation (11c) is identical to that for a conventional
three-element rectangular rosette, but must be interpreted
differently to determine which principal stress is referred to
gage no. 1. The following rules can be used for this purpose:
r
3
> r
1
: o refers to o
max
r
3
< r
1
: o refers to o
min
r
3
= r
1
: o = 45
r
2
< r
1
: o
max
at +45
r
2
> r
1
: o
max
at 45
Careful consideration must also be given to determining
the appropriate values for coefcients A and B. As dened
algebraically in Equations (6), they apply only when
the conditions imposed by the Kirsch solution are met.
This solution gives the stress distribution at points with
coordinates (r, o) around a circular hole through a thin,
wide plate subjected to uniform plane stress. However,
comparison of Figures 3 and 4 illustrates that, since the
strain gage grids in the rosette have nite areas, they sense
varying strain distributions such as those plotted in Figure 3.
Thus, the output of each gage tends to represent the average
strain over the area of the grid. Moreover, because the grids
are usually composed of parallel lines, those lines which are
not directly on the centerline of a radially oriented grid are
not radial. Therefore, the gages are slightly sensitive to the
tangential strain, as well as the radial strain. As a result,
more accurate values for the coefcients can be obtained
by integrating Equations (4) over the areas of the respective
gage grids. The coefcients thus determined, which account
for the nite strain gage area, are designated here by A and
B to distinguish them from the values at a point as dened
by Equations (6). An alternative method for obtaining
A and B is to measure them by experimental calibration.
The procedure for doing so is described in Section III,
Determining Coefficients A and B. When performed
correctly, this procedure is potentially the most accurate
means for evaluating the coefcients.
When employing conventional strain gage rosettes for
exper imental stress analysis, it is usually recommended that
the strain measurements be corrected for the transverse
sensitivity of the gages. Correction relationships for this
purpose are given in Tech Note TN-509. These relationships
are not directly applicable, however, to the relieved strains
mea sured with a residual stress rosette by the hole-drilling
method.
In the residual stress case, the individual gages in the rosette
are effectively at different locations in a spatially varying
strain eld. As a result, the relieved axial and transverse
strains applied to each gage are not related in the same
manner as they are in a uniform strain field. Rigorous
correction would require evaluation of the coefcient C
[actually, its integrated or calibrated counterpart, C see
Equations (6)], for both the through-hole and blind-hole
geometries. Because of the foregoing, and the fact that the
transverse sensitivities of Micro-Measurements residual
stress rosettes are characteristically very low (approximately
1%), it is not considered necessary to correct for transverse
tan2o
r r r
r r
1 2 3
1 3
2
o
r r
r r r r r
o
r
max
min
A B
+
( ) + + ( )
1 3
3 1
2
3 1 2
2
4
1
4
2
11 3
3 1
2
3 1 2
2
4
1
4
2
+
+ ( ) + + ( )
r
r r r r r
A B
(11a)
(11b)
tan2o
r r r
r r
1 2 3
3 1
2
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
23
sensitivity. Kabiri
4
, for example, has shown that the error
due to ignoring transverse sensitivity (in the case of uniaxial
residual stress) is negligible compared to the remaining
uncertainties in the measurement and data-reduction
procedures.
Blind-Hole Analysis
The theoretical background for the hole-drilling method
was developed in the preceding treatment on the basis of a
small hole drilled completely through a thin, wide, at plate
subjected to uniform plane stress. Such a conguration is far
from typical of practical test objects, however, since ordinary
machine parts and structural members requiring residual
stress analysis may be of any size or shape, and are rarely
thin or at. Because of this, a shallow blind hole is used in
most applications of the hole-drilling method.
The introduction of a blind hole into a eld of plane stress
produces a very complex local stress state, for which no exact
solution is yet available from the theory of elasticity. Fortu -
n ately, however, it has been demonstrated by Rendler and
Vigness
5
that this case closely parallels the through-hole
condition in the general nature of the stress distribution.
Thus, the relieved strains due to drilling the blind hole still
vary sinusoidally along a circle concentric with the hole, in
the manner described by Equations (9). It follows, then, that
these equations, as well as the data-reduction relationships
in Equations (11), are equally applicable to the blind-hole
implementation of the method when appropriate blind-hole
coefcients A and B are em ployed. Since these coefcients
cannot be calculated directly from theoretical considerations,
they must be obtained by empirical means; that is, by
experimental calibration or by numerical procedures such as
nite-element analysis.
Several different investigators [e.g., (20)(23)] have published
nite-element studies of blind-hole residual stress analysis.
The most recently developed coefficients by Schajer are
incorporated in ASTM standard E 837, and are shown
graphically for the case of uniform stress in Figure 8 of this
Tech Note. The computer program H-DRILL uses these
coefcients.
Compared to the through-hole procedure, blind-hole
analysis involves one additional independent variable;
namely, the dimensionless hole depth, Z/D (see Figure 5).
Thus, in a generalized functional form, the coefcients can
be expressed as:
A = f
A
(E, , r, Z/D) (12a)
B = f
B
(E, , r, Z/D) (12b)
For any given initial state of residual stress, and a xed
hole diameter, the relieved strains generally increase (at
a decreasing rate) as the hole depth is increased. Therefore,
in order to maximize the strain signals, the hole is normally
drilled to a depth corresponding to at least Z/D = 0.4 (ASTM
E 837 species Z/D = 0.4 for the maximum hole depth).
The general variation of relieved strain with depth is illustrated
in Figure 5, where the strains have been normalized, in this
case, to 100% at Z/D = 0.4. The data include experimental
results from two different investigators demonstrating the
manner in which the relieved-strain function is affected by the
ratio of hole diameter to gage circle diameter (D
o
/D). Both
cases involve uniform uniaxial (plane) stress, in specimens
that are thick compared to the maximum hole depth. The
curves plotted in the gure are considered representative
of the response to be expected when the residual stress is
uniform throughout the hole depth.
An important contribution of the Rendler and Vigness
work is the demonstration that, for any given set of material
properties, E and , coefficients A and B are simply
geometric functions, and thus constants for all geometrically
similar cases. This means that once the coefcients have
been determined for a particular rosette conguration, the
rosette size can be scaled upward or downward and the same
coefcients will still apply when the hole diameter and depth
are similarly scaled (assuming, of course, the same material).
Several different approaches have been taken in attempting
to remove the material depen dency from A and B, leaving
only the geometric dependence. One of these, proposed by
Schajer,7 is adopted in this Tech Note. Schajer introduced
two new coefcients, denoted here as a and b, and dened
as follows:

(13a)
(13b)
Z/D
100
80
60
40
20
0
0 0.1 0.2 0.3 0.4
GAGE #1
D
o
/D
P
E
R
C
E
N
T

S
T
R
A
I
N

R
E
L
I
E
V
E
D
120
0
.
4
0
K
e
l
s
e
y
(
R
e
f
.
6
)
0
.
2
9
R
e
n
d
l
e
r
&
V
i
g
n
e
s
s
(
R
e
f
.
5
)
D
o
Gage
Z
D
3
1
2
Figure 5. Relieved strain versus ratio of hole depth to gage circle
diameter (strains normalized to 100% at Z/D = 0.4).
a
EA
b EB

+

2
1
2
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
24
By comparison with Equations (6), it can be seen that
for the through hole, at least a is material-independent,
and b depends only weakly on Poissons ratio. Schajer
has determined from nite-element calculations that for
blind holes, a and b vary by less than 2% for the range of
Poissons ratio from 0.25 to 0.35.
III. Determining Coefcients A and B
Whether the residual stress analysis application involves
through-hole or blind-hole drilling, the coefcients A and
B (or a and b ) must be determined to calculate the stresses
from the relieved strains. In the case of the through hole,
reasonably accurate values of the coefcients can be obtained
for any particular case by analytical means, if desired. This
is done by integrating, over the area of the gage grid, the
component of strain parallel to the primary strain-sensing
axis of the gage. Given the details of the grid geometry
(line width and spacing, number of lines, etc.), Slightly
greater accuracy may be obtained by integrating along the
individual grid lines. This method cannot be applied to
blind-hole analysis because closed-form expressions relating
the relieved strains to the residual stress, in terms of hole
depth, are not available.
Experimental Calibration
The needed coefcients for either through-hole or blind-
hole analysis can always be determined by experimental
calibration. This procedure is particularly attractive since it
automatically accounts for the mechanical properties of the
test material, strain gage rosette geometry, hole depth and
diameter, and the strain-averaging effect of the strain gage
grid. When performed correctly, with sufcient attention
to detail, it is potentially the most accurate means for
determining the coefcients. Its principal disadvantage is
that the calibration must be repeated each time a different set
of geometric parameters is involved.
Calibration for A and B is accomplished by installing a
residual stress strain gage rosette on a uniaxially stressed
tensile specimen made from the same material as the test
part. The rosette should be oriented to align grid no. 1
parallel to the loading direction, placing grid no. 3 along
the transverse axis of the specimen. Care must be taken that
the tensile stress is uniform over the cross section of the test
specimen; i.e., that bending stress is negligible. To minimize
edge and end effects, the specimen width should be at least
ten times the hole diameter, and the length between machine
grips, at least ve times the width. When determining A
and B for blind-hole applications, a specimen thickness of
ve or more times the hole diameter is recommended. For
through-hole calibration, the thickness of the calibration
specimen is preferably the same as that of the test part. It
is also important that the maximum applied stress during
calibration not exceed one-half of the proportional limit
stress for the test material. In any case, the applied stress plus
the initial residual stress must be low enough to avoid the
risk of local yielding due to the stress-concentrating effect
of the hole.
Basically, the calibration procedure involves measuring the
rosette strains under the same applied load or calibration
stress, o
c
, both before and after drilling the hole. Such a
procedure is necessary in order to eliminate the effect of
the strain relief that may occur due to the relaxation of
any initial residual stress in the calibration specimen. With
this technique, the observed strain difference (before and
after hole drilling) is caused only by the applied calibration
stress, and is uniquely related to that stress. The steps in the
calibration procedure can be summarized briey as follows,
noting that the strains in only grid no. 1 and grid no. 3 need
to be measured, since these grids are known to be aligned
with the principal axes of the specimen.
1. Zero-balance the strain gage circuits.
2. Apply a load, P, to the specimen to develop the desired
calibration stress, o
c
.
3. Measure strains r
1
and r
3
(before drilling).
4. Unload the specimen, and remove it from the testing
machine.
5. Drill the hole, as described in Section V, Experimental
Techniques.
6. Replace the specimen in the testing machine, re-zero
the strain gage circuits, and then reapply exactly the
same load, P.
7. Measure strains r
1
and r
3
(after drilling).
The calibration strains corresponding to the load, P, and the
stress, o
c
, are then:
r
c1
= r
1
r
1
r
c3
= r
3
r
3

Calibration reliability can ordinarily be improved by loading
the specimen incrementally and making strain measurements
at each load level, both before and after drilling the hole.
This permits plotting a graph of o
c
versus r
c1
and r
c3
, so that
best-t straight lines can be constructed through the data
points to minimize the effect of random errors. It will also
help identify the presence of yielding, if that should occur.
The resulting relationship between the applied stress and
the relieved strain is usually more representative than that
obtained by a single-point determination.
Since the calibration is performed with only one nonzero
principal stress, Equation (5a) can be used to develop
expressions for the calibrated values of A and B. Successively
substituting o = 0 (for grid no. 1) and o = 90 (for grid no.
3) into Equation (5a):
r
c1
= o
c
[A + B cos (0)] = o
c
(A + B)
r
c3
= o
c
[A + B cos (2 x 90)] = o
c
(A B)
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
25
Solving for A and B,
(14a)

(14b)
The procedure described here was applied to a through-
hole specimen made from Type 304 stainless steel, and the
calibration data are plotted in Figure 6. It can be seen from the
gure that for this geometry (D
o
/D = 0.35) and material, r
c1

and r
c3
are 90r and +39r, respectively, when o
c
is 10 000 psi
[69 MPa]. Substituting into Equations (14),
A = 0.25 10
8
psi
1
[0.36 10
12
Pa
1
]
B = 0.65 10
8
psi
1
[0.94 10
12
Pa
1
]
Although the preceding numerical example referred to the
through-hole coefcients, the same procedure is followed
in calibrating for full-depth blind-hole coefcients. Once A
and B have been obtained in this manner, the corresponding
material-independent coefcients, a and b, can be calculated
from Equations (13) if the elastic modulus and Poissons ratio
of the test material are known. If desired, the procedure can
then be repeated over the practical range of D
o
/D to permit
plotting curves of a and b for all cases of interest.
It should be noted that the values for the basic coefcients A
and B obtained from a particular calibration test are strictly
applicable only for residual-stress measurement conditions
that exactly match the calibration conditions:
material with the same elastic properties;
same rosette geometry (but rosette orientation is
arbitrary);
same hole size;
same hole form (through hole or full-depth blind hole);
uniform stress with depth;
nominally uniform in-plane stress at the hole.
Coefcients for Micro-Measurements
Residual Stress Rosettes
Micro-Measurements supplies special strain gage rosettes
for residual stress analysis in four basic configurations,
illustrated and described in Figure 7. Among other features,
these rosette designs incorporate centering patterns for
positioning the boring tool precisely at the center of the gage
circle. All RE and UL designs have geometrically similar
grid congurations, with the gage-circle diameter equal to
A
B
c c
c
c c
c
r r
o
r r
o
1 3
1 3
2
2
Figure 6. Stress versus relieved strain for calibration of
coefcients A and B on 304 Stainless Steel (through-hole).
c
= 10 000 psi
S
T
R
E
S
S

[
M
P
a
]
c
1

=

9
0
c
3

=

+
3
9
CALIBRATION MICROSTRAIN (
c
)
S
T
R
E
S
S

(
1
0
0
0

p
s
i
)
GAGE #1 GAGE #3
100
80
60
40
20
100 80 60 40 20 0 20 40
10

5
1
3
2
EA-XX-062RE-120
This geometry conforms to the early Rendler
and Vigness design
5
and has been used in
most reported technical articles (see refer-
ences). It is available in a range of sizes to
accommodate applications requiring different
hole diameters or depths.
CEA-XX-062UL-120
This rugged, encapsulated de sign incorpo-
rates all practical advantages of the CEA
strain gage series (integral copper solder-
ing tabs, conformability, etc.). Installation
time and expense are greatly reduced, and
all solder tabs are on one side of the gage to simplify leadwire
routing from the gage site. It is compatible with all methods of
introducing the hole, and the strain gage grid geometry is identi-
cal to the 062RE pattern.
CEA-XX-062UM-120
Another CEA-Series strain gage, the
062UM grid widths have been reduced to
facilitate positioning all three grids on one
side of the measurement point as shown.
With this geometry, and appropriate trim-
ming, it is possible to position the hole closer to welds and other
irregularities. The user should be reminded, however, that the
data reduction equations are theoretically valid only when the
holes are well removed from free boundaries, discontinuities,
abrupt geometric changes, etc. The UM design is compatible
with all methods of introducing the hole.
N2K-XX-030RR-350/DP
The K-alloy grids of this open-faced six-
element rosette are mounted on a thin,
high-performance laminated polyimide
lm backing. Solder tabs are duplex cop-
per plated for ease in making solder joints
for lead attachment. Diametrically op-
posed circumferential and radial grids are
to be wired in half-bridge congurations.
Figure 7. Residual stress strain gage
rosettes (shown approximately 2X).
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
26
3.25 times the active gage length. The 062RE rosette, for
example, has a gage-circle diameter of 0.202 in [5.13 mm].
Because of this similitude, the same material-independent
coefcients a and b apply to all sizes of the RE rosette,
and to the UL rosette, for geometrically similar holes (i.e.,
for the same Do/D and Z/D ratios). The 062UM rosette
conguration has the same ratio of gage circle to grid length,
but the grids are narrower to permit their close grouping on
one side of the hole. As a result, the sensitivity of the gage to
the relieved strains is slightly greater, and coefcients specic
to the 062UM are required for data reduction.
The 030RR rosette is fundamentally different from the other
rosettes illustrated in Figure 7. To begin with, this rosette
includes both radially and circumferentially oriented grids
which are to be connected as half-bridge pairs. The 030RR
rosette incorporates a number of features that contribute
to its greater output and higher accuracy compared to
conventional three-element rosettes: (a) the individual
gridlines in the radial elements are purely radial, instead
of being simply parallel to the central gridline as in the
other rosettes; (b) for a given maximum hole diameter,
the outermost radius of the grids is considerably less than
for the corresponding conventional rosettes, and thus the
grids sense slightly greater average released strains; and (c)
since the radial and circumferential grids get connected in
a half-bridge conguration, the bridge output is augmented
correspondingly, and the circuit is intrinsically self-
temperature compensating. As a result of these features, the
030RR rosette yields about twice the output of the three-
element rosettes for a given state of residual stress, while
displaying better stability and accuracy.
Since the sign of the residual stress is of primary importance
in determining its effect on the structural integrity of any
mechanical component, the user of the six-element rosette
(030RR) must exercise care in connecting the rosette grids
into Wheatstone bridge circuits. To obtain the correct sign
in the instrument output signal, the radially oriented grids
should always be connected between the positive excitation
and the negative signal terminals, while the tangentially
oriented grids are to be connected between the negative
signal and negative excitation terminals.
The a and b coefcients for Micro-Measurements residual
stress rosettes are provided graphically in Figure 8 on page
28, where the solid lines apply to full-depth blind holes and
the dashed lines to through holes assuming, in both cases,
that the initial residual stress is uniform with depth. Both
the through-hole and full-depth blind-hole coefficients
plotted in Figure 8 were determined by a combination of
nite-element analysis and experimental verication. These
coefcients are also supplied numerically in tabular form in
ASTM E 837-99, where RE/UL rosettes are designated as
Type A, UM rosettes as Type B, and RR rosettes as Type
C. For the blind-hole coefcients in the ASTM standard,
full depth corresponds to a value of 0.40 for the depth to
rosette-mean-diameter-ratio, Z/D.
IV. Measuring Nonuniform Residual Stresses
The coefficients given in this Tech Note and in ASTM
E 837-99 are strictly applicable only to situations in which
the residual stresses do not vary in magnitude or direction
with depth from the test-part surface. In reality, however,
residual stresses may often vary signicantly with depth,
due, for example, to different manufacturing processes such
as casting, forging, heat treatment, shot peening, grinding,
etc.
Numerous nite-element studies have been made in attempts
to treat this situation [see, for instance, references (20)
through (23)]. The results of the fnite element work by
Schajer have been incorporated in a computer program,
H-DRILL, for handling stress variation with depth. When
the measured strains from hole drilling do not fit the
reference curves in Figure 10, or when there is any other
basis for suspecting signicant nonuniformity, the program
H-DRILL or some other finite-element-based program
is necessary to accurately determine the stresses from the
measured relaxed strains.
V. Experimental Techniques
As in all experimental methods, proper materials, application
procedures, and instrumentation are essential if accurate
results are to be obtained. The accuracy of the hole-drilling
method is dependent chiey upon the following technique-
related factors:
strain gage selection and installation.
hole alignment and boring.
strain-indicating instrumentation.
understanding the mechanical properties of the test
material.
Strain Gage Selection and Installation
Installing three individual strain gages, accurately spaced
and oriented on a small circle, is neither easy to do nor
advisable, since small errors in gage location or orientation
can produce large errors in calculated residual stresses. The
rosette con gurations shown in Figure 7 have been designed
and developed by Micro-Measurements specifically for
re sid ual stress measurement. The rosette designs incorporate
centering marks for aligning the boring tool precisely at the
center of the gage circle, since this is critical to the accuracy
of the meth od.
9,10,11
All congurations are available in a
range of temperature compensations for use on common
structural metals. However, only the RE design is offered
in different sizes (031RE, 062RE, and 125RE), where the
three-digit prex represents the gage length in mils (0.001 in
[0.0254 mm]). The RE design is available either open-faced
or with Option SE (solder dots and encapsulation).
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
27
C O E F F I C I E N T S A N D
1
.
2
1
.
1
1
.
0
0
.
9
0
.
8
0
.
7
0
.
5
0
.
4
0
.
3
0
.
4
0
0
.
4
5
0
.
5
0
0
.
5
5
0
.
6
0
(
c
)

R
R

R
O
S
E
T
T
E
D
o
/
D

a
b a
0
.
6
b
l
i
n
d

h
o
l
e
t
h
r
o
u
g
h

h
o
l
e
1
.
3
S
U
G
G
E
S
T
E
D

L
I
M
I
T
S

b
D
0
.
4
D
D
o
S
U
G
G
E
S
T
E
D

L
I
M
I
T
S
C O E F F I C I E N T S A N D
0
.
8
0
.
7
0
.
6
0
.
5
0
.
4
0
.
3
0
.
2
0
.
10
0
.
3
0
0
.
3
5
0
.
4
0
0
.
4
5
0
.
5
0
(
a
)

R
E

A
N
D

U
L

R
O
S
E
T
T
E
S
D
o
/
D
b
l
i
n
d

h
o
l
e
t
h
r
o
u
g
h

h
o
l
e
1 1
1
D
D
o
2
0
.
4
D
3

b
a
b a
0
.
8
0
.
7
0
.
6
1 1
0
.
5
0
.
4
0
.
3
0
.
2
0
.
1
0
0
.
3
0
0
.
3
5
0
.
4
0
0
.
4
5
0
.
5
0
(
b
)

U
M

R
O
S
E
T
T
E
D
o
/
D
1 1

b

a
C O E F F I C I E N T S A N D b a
b
l
i
n
d

h
o
l
e
t
h
r
o
u
g
h

h
o
l
e
S
U
G
G
E
S
T
E
D

L
I
M
I
T
S
1
D
D
o
2
0
.
4
D
3
F
i
g
u
r
e

8
.

F
u
l
l
-
d
e
p
t
h

d
a
t
a
-
r
e
d
u
c
t
i
o
n

c
o
e
f
c
i
e
n
t
s

a

a
n
d

b

v
e
r
s
u
s

d
i
m
e
n
s
i
o
n
l
e
s
s

h
o
l
e

d
i
a
m
e
t
e
r

(
t
y
p
i
c
a
l
)

f
o
r

M
i
c
r
o
-
M
e
a
s
u
r
e
m
e
n
t
s

r
e
s
i
d
u
a
l

s
t
r
e
s
s

r
o
s
e
t
t
e
s
,

i
n

a
c
c
o
r
d
a
n
c
e

w
i
t
h

A
S
T
M

E

8
3
7
.
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
28
The UL and UM configurations are supplied in
1
16 in
[1.6 mm] gage length, and both types are fully encapsulated.
Both configurations have integral, copper-coated
solder tabs, and offer all advantages of the popular
C-Feature strain gage series. These residual stress rosettes
are constructed with self-temperature-compensated
constantan foil, mounted on a exible polyimide carrier.
Gage resistance is 120 ohms 0.4%. The 030RR six-
element rosette incorporates self-temperature-compensated
K-alloy (modied Karma) foil on a laminated polyimide
film backing. Solder tabs are duplex copper plated for
ease in making solder connections. Gage resistance is
350 ohms 0.4%.
Surface preparation for installing the rosettes is basically
standard, as described in Application Note B-129. Caution
should be observed, however, in abrading the surface of
the test object, since abrasion can alter the initial state of
residual stress.
12
In general, it is important that all surface-
preparation and gage-installation procedures be of the
highest quality, to permit accurate measurement of the small
strains typically registered with the hole-drilling method. As
evidenced by the calibration data in Figure 6, the relieved
strains corresponding to a given residual stress magnitude
are considerably lower than those obtained in a conventional
mechanical test at the same stress level. Because of the small
measured strains, any drift or inaccuracy in the indicated
gage output, whether due to improper gage installation,
unstable instrumentation, or otherwise, can seriously affect
the calculated residual stresses.
Strain-Measurement Instrumentation
The residual stress rosettes described in Figure 7 impose no
special instrument requirements. When measurements are
to be made in the eld, a portable, battery-operated static
strain indicator, supplemented by a precision switch-and-
balance unit, is ordinarily the most effective and convenient
instrumentation. The Model P3 Strain Indicator and Recorder
is ideally suited for this type of application. In the laboratory
it may be convenient to use a computerized automatic data
system such as System 5000, which will rapidly acquire and
record the data in an organized, readily accessible form.
A special offline, Windows
-based computer program

H-DRILL is also available to perform the calculations and
determine the residual stress magnitudes in accordance with
ASTM E 837. The database for the program includes values
of the coefcients a and b for blind holes, and covers the full
range of recommended hole dimensions applicable to all
Micro-Measure ments residual stress rosettes.
Alignment
Rendler and Vigness observed that the accuracy of the
(hole-drilling) method for eld applications will be directly
related to the operators ability to position the milling cutter
precisely in the center of the strain gage rosette. More recent
studies have quantied the error in calculated stress due to
eccentricity of the hole. For example, with a hole that is 0.001
in [0.025 mm] off-center of the 062RE or 062UL rosette, the
error in calculated stress does not exceed 3% (for a uniaxial
stress state).
9,10,11
In practice, the required alignment precision
to within 0.001 in [0.025 mm] is accomplished using the
RS-200 Milling Guide shown in Figure 9. The milling
guide is normally secured to the test object by bonding its
three foot pads with a quick-setting, frangible adhesive. A
microscope is then installed in the unit and visual alignment
is achieved with the aid of the four X-Y adjustment screws
on the exterior of the guide.
Boring
Numerous studies on the effects of hole size and shape
and machining procedures have been published. Rendler
and Vigness
5
specied a specially dressed end mill which
is compatible with the residual stress rosettes of Figure 7.
The end mill is ground to remove the side cutting edges,
and then relieved immediately behind the cutting face to
avoid rubbing on the hole surface. It is imperative that the
milling cutter be rigidly guided during the drilling operation
so that the cutter pro gresses in a straight line, without side
pressure on the hole, or friction at the noncutting edge.
These end mills generate the desired flat-bottomed and
square-cornered hole shape at initial surface contact, and
maintain the appropriate shape until the hole is completed.
In doing so, they fulll the incremental drilling requirements
as stipulated in ASTM E 837. Specially dressed end mills
offer a direct and simple approach when measuring residual
stresses on readily machinable materials such as mild steel
and some aluminum alloys. Figure 9b shows the RS-200
Milling Guide with the microscope removed and the end-
mill assembly in place. The end mill is driven through the
universal joint at the top of the assembly, by either a hand
drill or variable-speed electric drill.
In 1982, Flaman
13
rst reported excellent results for residual
stress measurement using a high-speed (up to 400 000 rpm)
air turbine and carbide cutters. This technique maintains
all of the advantages (good hole shape, adaptability to
incremental drilling, etc.) of the specially dressed end mill
while providing for easier operation and more consistent
results. Further, the air turbine is highly recommended
for use with test materials that are difcult to machine, such
as Type 304 stainless steel. Carbide cutters are not effective
for penetrating glass, most ceramics, very hard metals, etc.;
but diamond cutters have shown promise on these kinds of
test materials. Figure 9c shows the air turbine/carbide cutter
assembly installed in the same basic RS-200 Milling Guide.
Bush and Kromer
14
reported, in 1972, that stress-free holes
are achieved using abrasive jet machining (AJM). Modi-
cations and improvements were made to AJM by Procter
and Beaney,
10
and by Bynum.
15
Wnuk
16
experienced good
Windows is a registered trademark of Microsoft, Inc.
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
29
results by mechanically adapting AJM apparatus for use
in the RS-200 Milling Guide. The principal advantage of
AJM is its reported ability to generate stress-free holes in
virtually all materials. Its chief limitations center about
the considerable changes in hole shape as a function of
hole depth. The initial shape is saucer-like, and the nal is
cylindrical with slightly rounded corners. During drilling
there is also uncertainty as to the actual hole depth at any
stage. These factors make AJM a less practical technique for
determining the variation of relieved strain with hole depth,
as recommended in ASTM E 837.
Mechanical Properties
As in any form of experimental stress analysis, the accuracy
of residual stress measurement is limited by the accuracies to
which the elastic modulus and Poissons ratio are known. But
typical uncertainties in the mechanical properties of common
steel and aluminum alloys are in the neighborhood of 1 to
3% and, as such, are minor contributors to potential errors in
residual stress analysis. Much larger errors can be introduced by
deviations from the assumptions involved in the basic theory,
as described in Section II. A key assumption, for instance,
is linear-elastic material behavior. If the stress/ strain
relationship for the test material is nonlinear (as is the case
for cast iron), due to yielding or other causes, the calculated
residual stresses will be in error.
When the initial residual stress is close to the yield strength
of the test material, the stress concentration caused by the
presence of the hole may induce localized yielding. It is
therefore necessary to establish a threshold level of residual
stress below which yielding is negligible. This problem
has been studied both experimentally and analytically,
and there is substantial agreement among the different
investigations.
10,17,18
That is, errors are negligible when the
residual stress is less than 70% of the proportional limit of
the test material for both blind holes and through holes.
On the other hand, when the initial residual stress is equal
to the proportional limit, errors of 10 to 30% (and greater)
have been observed. The error magnitude obviously depends
on the slope of the stress/strain diagram in the post-yield
region; and tends to increase as the curve becomes atter,
approaching the idealized perfectly plastic behavior.
18
VI. Data Reduction and
Interpretation Blind Hole
As recommended in ASTM E 837, it is always preferable
to drill the hole in small increments of depth, recording
the observed strains and measured hole depth at each
increment. This is done to obtain data for judging whether
the residual stress is essentially uniform with depth, thus
validating the use of the standard full-depth coefcients a
and b for calculating the stress magnitudes. If incremental
measurements are not taken, there is no means for making
Figure 9. RS-200 Milling Guide, used for
machining a precisely located at-bottomed hole.
(a) Alignment Setup
(b) End-Mill Drilling Setup
(c) Hi-Speed Drilling Setup
Eyepiece
Microscope
Tube
Locking
Collar
Illuminator
X-Y
Adjustments
(4)
Vertical Height
Adjustments
(3)
Locking
Nuts
Cap
Pad
Milling Bar with
Universal Joint At-
tached
Micrometer
Adjustment
Locking
Collar
Micrometer
Lock
Micrometer
End Mill
To Air Supply
Air Turbine
Assembly Spring
Assembly
Grooved Nylon
Collar
Anti-Rotation
Ring Adapter
Basic RS-200
Milling Guide
Carbide Cutter
Mount
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
30
inferences about stress uniformity, and the calculated residual
stress may be considerably in error. In such cases, when
the stress varies with depth, it should be realized that the
calculated stress is always lower than the actual maximum.
There is currently no absolute criterion for verifying
stress uniformity from the surface of the test piece to the
bottom of a full-depth hole. However, the incremental data,
consisting of relieved strain versus hole depth, can be used
in two different ways to aid in detecting a nonuniform stress
distribution. The rst of these is to calculate, for each depth
increment, the sums and differences of the measured strain
data, r
3
+ r
1
and r
3
r
1
respectively.
1
Express each set of data
as fractions of their values when the hole depth equals 0.4
times the mean diameter of the strain gage circle. Plot these
percent strains versus normalized hole depth. These graphs
should yield data points very close to the curves shown in
Figure 10. Data points which are removed from the curves in
Figure 10 indicate either substantial stress nonuniformity or
strain measurement errors. In either case, the measured data
are not acceptable for residual stress calculations using the
full-depth coefcients a and b.
When a principal residual stress direction is closer to the
axial direction of gage no. 2 in Figure 4 than to either gage
nos. 1 or 3, the strain sum r
3
+ r
1
2r
2
will be numerically
larger than r
3
r
1
. In such a case, the percent strain data
check should be done using r
3
+ r
1
2r
2
instead of r
3
r
1
.
NOTE: This graphical test is not a sensitive indicator of
stress field uniformity. Specimens with significantly
nonuniform stress elds can yield percentage relieved strain
curves substantially similar to those shown in Figure 10. The
main purpose of the test is to identify grossly nonuniform
stress fields. Further, the graphical comparison test using
r
3
r
1
or r
3
+ r
1
2r
2
, for example, becomes ineffective when
the residual stress eld approaches equal biaxial tension or
compression (r
1
r
2
r
3
) as expected in surface blasting and
heat treating procedures. Comparison to the r
3
+ r
1
plot is
ineffective when r
3
= r
1
(pure shear); however, this condition
is relatively uncommon in the practical industrial setting.
Limitations and Cautions
Finite-element studies of the hole-drilling method by
Schajer and by subsequent investigators
20,21,22,23
have shown
that the change in strain produced in drilling through any
depth increment (beyond the rst) is caused only partly
by the residual stress in that increment. The remainder
of the incremental relieved strain is generated by the
residual stresses in the preceding increments, due to the
increasing compliance of the material, and the changing
stress distribution, as the hole is deepened. Moreover, the
relative contribution of the stress in a particular increment
to the corresponding incremental change in strain decreases
rapidly with distance from the surface. As a result, the
total relieved strain at full-hole depth is predominantly
inuenced by the stresses in the layers of material closest
Figure 10. Percent strain versus normalized hole depth
for uniform stress with depth for different rosette
types, after ASTM E 837.1
RE AND UL ROSETTE
P
E
R
C
E
N
T

R
E
L
I
E
V
E
D

S
T
R
A
I
N
0.0 0.1 0.2 0.3 0.4
Z/D
100
80
60
40
20
0
3
+
1
3

1
or
3
+
1
2
2
D
o
Z
D
Gage
UM ROSETTE
P
E
R
C
E
N
T

R
E
L
I
E
V
E
D

S
T
R
A
I
N
0.0 0.1 0.2 0.3 0.4
Z/D
100
80
60
40
20
0
D
o
Z
D
Gage
3
+
1
3

1
or
3
+
1
2
2
RR ROSETTE
P
E
R
C
E
N
T

R
E
L
I
E
V
E
D

S
T
R
A
I
N
0.0 0.1 0.2 0.3 0.4
Z/D
100
80
60
40
20
0
D
o
Z
D
Gage
3
+
1
3

1
or
3
+
1
2
2
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
31
to the surface say, in the upper third, or perhaps half,
of the hole depth. At hole depths corresponding to
Z/D > 0.2, the stresses in these increments have very little
effect on the observed strains. This behavior is conrmed
(for uniform stress) by the shape of the normalized strain
graph in Figure 5, where about 80% of the total strain
relief normally occurs in the rst half of the hole depth.
Because of these characteristics, little, if any, quantitative
interpretation can safely be made of the incremental strain
data for increments beyond Z/D = 0.2, irrespective of the
analytical method employed for data reduction.
To summarize, the ideal application of the hole-drilling
method is one in which the stress is essentially uniform with
depth. For this case, the data-reduction coefcients are well-
established, and the calculated stresses sufciently accurate
for most engineering purposes assuming freedom from
significant experimental errors. Incremental drilling and
data analysis should always be performed, however, to
verify the stress uniformity. If the graph of percent-strain-
relieved versus Z/D (see Figure 10) suggests that the stress
is nonuniform with hole depth, then the procedure specied
by ASTM E 837 is not applicable, and a program such as
H-DRILL must be used to calculate the stresses.
Error and uncertainty are always present, in varying degrees,
in all measurements of physical variables. And, as a rule,
their magnitudes are strongly dependent on the quality of the
experimental technique as well as the number of parameters
involved. Since residual stress determination by the hole-
drilling method involves a greater number and variety of
techniques and parameters than routine experimental stress
analysis, the potential for error is correspondingly greater.
Because of this, and other considerations briey outlined in
the following, residual stresses cannot usually be determined
with the same accuracy as stresses due to externally applied
static loads.
Introduction of the small hole into the test specimen is
one of the most critical operations in the procedure. The
instruction manual for the RS-200 Milling Guide contains
detailed directions for making the hole; and these should be
followed rigorously to obtain maximum accuracy. The hole
should be concentric with the drilling target on the special
strain gage rosette. It should also have the prescribed shape
in terms of cylindricity, at bottom, and sharp corner at the
surface. It is particularly necessary that the requirements on
hole conguration be well-satised when doing incremental
drilling to examine stress variation with depth. Under
these same circumstances, it is important that the hole
depth at each drilling increment be measured as accurately
as possible, since a small absolute error in the depth can
produce a large relative error in the calculated stress. Because
of practical limitations on measuring shallow hole depths,
the rst depth increment should ordinarily be at least 0.005
in [0.13 mm]. Accurate measurement of the hole diameter
is also necessary. Finally, it is imperative that the hole be
drilled (milled) without introducing signicant additional
residual stresses. To the degree that any of the foregoing
requirements fail to be met, accuracy will be sacrificed
accordingly.
Strains relieved by drilling the hole are measured
conventionally, with static strain instrumentation. The
indicated strains are characteristically much smaller,
however, than they would be for the same stress state in
an externally loaded test part. As a result, the need for
stable, accurate strain measurement is greater than usual.
With incremental drilling, the strains measured in the rst
few depth increments can be especially low, and errors of
a few microstrain can cause large percentage errors in the
calculated stresses for those depths.
Beyond the above, it is also necessary that the underlying
theoretical assumptions of the hole-drilling method be
reasonably satised. In full-depth drilling per ASTM E 837,
the stress must be essentially uniform with depth, both in
magnitude and direction, to obtain accurate results. With
nite-element and other procedures for investigating stress
variation in subsurface layers, it is required only that the
directions of the principal stresses not change appreciably
with depth. As for all conventional strain-gage rosette
measurements, the data-reduction relationships assume
that the stress is uniformly distributed in the plane of the
test surface. However, for residual stress measurements the
effective gage-length is the hole diameter rather than the
relatively large dimensions of the overall rosette geometry.
Consequently, uncertainties introduced by in-plane surface
strain gradients are generally lower for residual stress
determination than for conventional static load testing. No
generalization can currently be made about the effects of
steeply varying, nonlinear stress distributions in subsurface
planes parallel to the rosette.
BIBLIOGRAPHY
1. Determining Residual Stresses by the Hole-Drilling
Strain-Gage Method. ASTM Standard E 837.
2. Mathar, J., Determination of Initial Stresses by
Measuring the Deformation Around Drilled Holes.
Trans., ASME 56, No. 4: 249-254 (1934).
3. Timoshenko, S. and J.M. Goodier, Theory of Elasticity,
New York: McGraw-Hill (1951).
4. Kabiri, M., Measurement of Residual Stresses by
the Hole-Drilling Method: Inuences of Transverse
Sensitivity of the Gages and Rel i eved Strai n
Coeffcients, Experimental Mechanics 25: (252-256)
(Sept. 1984).
5. Rendler, N.J. and I. Vigness, Hole-drilling Strain-gage
Method of Measuring Residual Stresses. Proc., SESA
XXIII, No. 2: 577-586 (1966).
T
E
C
H

N
O
T
E
TN-503-6
Micro-Measurements
32
6. Kelsey, R.A., Measuring Non-uniform Residual
Stresses by the Hole-drilling Method. Proc., SESA
XIV, No. 1: 181-194 (1956).
7. Schaj er, G. S., Appl i cation of Fi nite El ement
Calculations to Residual Stress Measurements.
Journal of Engineering Materials and Technology 103:
157-163 (1981).
8. Redner, S. and C.C. Perry, Factors Affecting the
Accuracy of Residual Stress Measurements Using the
Blind-Hole Drilling Method. Proc., 7th International
Conference on Experimental Stress Analysis. Haifa,
Israel: Israel Institute of Technology, 1982.
9. Sandifer, J.P. and G.E. Bowie, Residual Stress by Blind-
hole Method with Off-Center Hole. Experimental
Mechanics 18: 173-179 (May 1978).
10. Procter, E. and E.M. Beaney, Recent Developments
i n Centre-hol e Techni que for Resi dual -stress
Measurement. Experimental Techniques 6: 10-15
(December 1982).
11. Wang, H.C., The Alignment Error of the Hole-Drilling
Method. Experimental Mechanics 17: 23-27 (1979).
12. Prevey, P.S., Residual Stress Distributions Produced
by Strain Gage Surface Preparation. Proc., 1986
SEM Conference on Experimental Mechanics (1986).
13. Flaman, M.T., Brief Investigation of Induced Drilling
Stresses in the Center-hole Method of Residual-stress
Measurement. Experimental Mechanics 22: 26-30
(January 1982).
14. Bush, A.J. and F.J. Kromer, Simplication of the Hole-
drilling Method of Residual Stress Measurements.
Trans., ISA 112, No. 3: 249-260 (1973).
15. Bynum, J.E., Modications to the Hole-drilling Tech-
nique of Measuring Residual Stresses for Improved
Accur-acy and Reproducibi l ity. Experi mental
Mechanics 21: 21-33 (January 1981).
16. Wnuk, S.P.. Residual Stress Measurements in the Field
Using the Airbrasive Hole Drilling Method. Presented
at the Technical Committee for Strain Gages, Spring
Meeting of SESA, Dearborn, Michigan, June, 1981.
17. Delameter, W.R. and T.C. Mamaros, Measurement of
Residual Stresses by the Hole-drilling Method. Sandia
National Laboratories Report SAND-77-8006 (1977),
27 pp. (NTIS).
18. Flaman, M.T. and B.H. Manning, Determination
of Residual Stress Variation with Depth by the Hole-
Drilling Method. Experimental Mechanics 25: 205-
207 (1985).
19. Niku-Lari, A.J. Lu and J.F. Flavenot, Measurement of
Residual Stress Distribution by the Incremental Hole-
Drilling Method. Experimental Mechanics 25: 175-185
(1985).
20. Flaman, M.T., B.E. Mills, and J.M. Boag, Analysis of
Stress-Variation-With-Depth Measurement Procedures
for the Centre Hole Method of Residual Stress
Measurements. Experimental Techniques 11: 35-37
(June 1987).
21. Schajer, G.S., Measurement of Non-Uniform Residual
Stresses Using the Hole Drilling Method, Journal of
Engineering Materials and Technology, 110, No. 4: Part
I, 338-343; Part II, 344-349 (1988).
22. Ajovalasit, A., Measurement of Residual Stresses
by the Hole-Drilling Method: Inf luence of Hole
Eccentricity. Journal of Strain Analysis 14, No. 4: 171-
178 (1979).
23. Beaney, E.M. and E. Procter, A Critical Evaluation
of the Centre-hole Technique for the Measurement of
Residual Stresses. Strain, Journal of BSSM 10, No. 1:
7-14 (1974).
24. Nawwar, A.M., K. McLachlan, and J. Shewchuk, A
Modified Hole-Drilling Technique for Determining
Residual Stresses in Thin Plates. Experi mental
Mechanics 16: 226-232 (June 1976).
25. Witt, F., F. Lee, and W. Rider, A Comparison of
Residual Stress Measurements Using Blind-hole,
Abrasive-jet and Trepan-ring Methods. Experimental
Techniques 7: 41-45 (February 1983).
26. Schajer, G.S., Judgment of Residual Stress Field
Uniformity when Using the Hole-Drilling Method,
Proceedings of the Inter national Conference on
Residual Stresses II, Nancy, France. November 23-25,
1988, 71-77.
27. Flaman, M.T. and J.A. Herring, SEM/ASTM Round-
Robin Residual-Stress-Measurement Study Phase
1, 304 Stai nless-Steel Speci men, Experi mental
Techniques, 10, No. 5: 23-25.
28. Yavelak, J.J. (compiler), Bulk-Zero Stress Standard
AISI 1018 Carbon-Steel Specimens, Round Robin
Phase 1, Experimental Techniques, 9, No. 4: 38-41
(1985).
29. Schajer, G.S., Strain Data Averaging for the Hole-
Drilling Method. Experimental Techniques. Vol. 15,
No. 2, pp. 25-28, 1991.
30. Schajer, G.S. and E. Altus, Stress Calculation Error
Analysis for Incremental Hole-Drilling Residual Stress
Measurements. Journal of Engineering Materials and
Technology. Vol. 118, No. 1, pp. 120-126, 1996.
31. Schajer, G.S., Use of Displacement Data to Calculate
Strain Gauge Response in Non-Uniform Strain Fields.
Strain. Vol. 29, No. 1, pp. 9-13, 1993.
32. Schajer, G.S. and Tootoonian, M., A New Rosette
Design for More Reliable Hole-drilling Residual Stress
Measurements. Experimental Mechanics. Vol. 37, No.
3, pp. 299-306, 1997.
T
E
C
H

N
O
T
E
TN-501-2
Micro-Measurements
33
Tech Note TN-504-1
MICRO-MEASUREMENTS
Strain Gage Thermal Output and
Gage Factor Variation with Temperature
T
E
C
H

N
O
T
E
35
Revision 11-Jul-2012
1.0 Introduction
Ideally, a strain gage bonded to a test part would respond
only to the applied strain in the part, and be unaffected
by other variables in the environment. Unfortunately, the
resistance strain gage, in common with all other sensors,
is somewhat less than perfect. The electrical resistance
of the strain gage varies not only with strain, but with
temperature as well. In addition, the relationship between
strain and resistance change, the gage factor, itself varies
with temperature. These deviations from ideal behavior
can be important under certain circumstances, and can
cause signicant errors if not properly accounted for. When
the underlying phenomena are thoroughly understood,
however, the errors can be controlled or virtually eliminated
by compensation or correction.
In Section 2.0 of this Tech Note, thermal output (sometimes
referred to as temperature-induced apparent strain) is
dened, and the causes of this effect are described. Typical
magnitudes of the thermal output are then given, followed
by the commonly used methods for compensation and
correction. Section 3.0 treats gage factor variation with
temperature in a similar but briefer manner since this error
source is generally much less signicant. Methods for the
simultaneous correction of both thermal output and gage
factor errors are given in Section 4.0, accompanied by
numerical examples.
2.0 Thermal Output
Once an installed strain gage is connected to a strain indicator
and the instrument balanced, a subsequent change in the
temperature of the gage installation will normally produce
a resistance change in the gage. This temperature-induced
resistance change is independent of, and unrelated to, the
mechanical (stress-induced) strain in the test object to which
the strain gage is bonded. It is purely due to temperature
change, and is thus called the thermal output of the gage.
Thermal output is potentially the most serious error source
in the practice of static strain measurement with strain
gages. In fact, when measuring strains at temperatures
remote from room temperature (or from the initial balance
temperature of the gage circuit), the error due to thermal
output, if not controlled, can be much greater than the
magnitude of the strain to be measured. At any temperature,
or in any temperature range, this error source requires
careful consideration; and it is usually necessary to either
provide compensation for thermal output or correct the
strain measurements for its presence.
Thermal output is caused by two concurrent and algebraically
additive effects in the strain gage installation. First, the
electrical resistivity of the grid conductor is somewhat
temperature dependent; and, as a result, the gage resistance
varies with temperature. The second contribution to thermal
output is due to the differential thermal expansion between
the grid conductor and the test part or substrate material
to which the gage is bonded. With temperature change, the
substrate expands or contracts; and, since the strain gage
is rmly bonded to the substrate, the gage grid is forced to
undergo the same expansion or contraction. To the extent
that the thermal expansion coefcient of the grid differs
from that of the substrate, the grid is mechanically strained
in conforming to the free expansion or contraction of the
substrate. Because the grid is, by design, strain sensitive,
the gage exhibits a resistance change proportional to the
differential expansion.
Each of the two thermally induced resistance changes may
be either positive or negative in sign with respect to that of
the temperature change, and the net thermal output of the
strain gage is the algebraic sum of these. Thus, expressed
in terms of unit resistance change, the thermal output
becomes:
R
R
F
K
K
T
T O
G G
t
t
S G
0 0
1
1
j
(
,
\
,
(
+
+
j
(
,
\
,
(
( )
,
,
,
]
]
]
]
/

(1)

where, in consistent units:
R
R
TO
0
/
= unit change in resistance from the initial
reference resistance, R
0
, caused by change in
temperature resulting in thermal output.

G
= temperature coefficient of resistance of the
grid conductor.
F
G
= gage factor of the strain gage.
K
t
= transverse sensitivity of the strain gage.

0
= Poissons ratio (0.285) of the standard test material
used in calibrating the gage for its gage factor.
In this Tech Note, the gage factor of the strain gage is identied as
F
G
, to distinguish it from the gage factor setting of the measuring
instrument, denoted here by F
I
. This distinction is important, since
the gage factor setting of the instrument may sometimes, as a matter
of convenience or utility, be different from that of the gage.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
Revision: 11-Jul-2012
36
Strain Gage Thermal Output and Gage Factor Variation with Temperature
(
S

G
) = difference in thermal expansion coefcients of
substrate and grid, respectively.
T = temperature change from an arbitrary initial
reference temperature.
The correction factor for transverse sensitivity [(1 + K
t
)/
(1
0
K
t
)] is included in Equation (1) to account for the fact
that the strain in the gage grid due to differential thermal
expansion is equal-biaxial, while the gage factor, F
G
, refers
to the strain sensitivity as calibrated in a uniaxial stress state,
with a principal strain ratio of 1/(0.285).
It should not be assumed from the form of Equation
(1) that the thermal output is linear with temperature
change, because all of the coefcients within the brackets are
themselves functions of temperature. The equation clearly
demonstrates, however, that thermal output depends not
only on the nature of the strain gage, but also on the material
to which the gage is bonded. Because of this, thermal output
data are meaningful only when referred to a particular type
of strain gage, bonded to a specied substrate material.
For convenience in correcting measured strain data for
thermally induced resistance changes, the thermal output of
the gage is usually expressed in strain units. Thus, dividing
Equation (1) by the gage factor setting of the instrument,

TO
TO
I
G G
t
t
R
R
F
F
K
K
/
/
j
(
,
\
,
(
+
+
j
(
,
\
,
(
0
0
1
1
SS G
I
T
F
( )
,
,
]
]
]

(2)

where:
T/O
= thermal output in strain units; that is, the
strain magnitude registered by a strain indi-
cator (with a gage factor setting of F
I
), when
the gage installation is subjected to a tem-
perature change, T, under conditions of
free thermal expansion for the substrate.
When measuring stress-induced strains at a temperature
different from the initial balance temperature, the thermal
output from Equation (2) is superimposed on the gage
output due to mechanical strain, causing the measurement
to be in error by that amount. Many factors affect the
thermal output of strain gages. Some of the more important
are: test specimen material and shape, grid alloy and lot,
gage series and pattern, transverse sensitivity of the gage,
bonding and encapsulating materials, and installation
procedures. It is never possible for Micro-Measurements
to predict exactly what the thermal output of any gage will
be when the user has bonded it to a test structure. Even in
cases where applications involve the same material as that
used by Micro-Measurements in its tests, differences can be
expected since structural materials vary in thermal expansion
characteristics from lot to lot. The best practice is always to
evaluate one or more gages under thermal conditions as
nearly like those to be encountered in the testing program
as possible.
Figure 1 shows the variation of thermal output with
temperature for a variety of strain gage alloys bonded to
steel. These data are illustrative only, and not for use in
making corrections. It should be noted, in fact, that the curves
for constantan and Karma are for non-self-temperature-
compensated alloys. With self-temperature compensation
(Section 2.1.2), as employed in Micro-Measurements strain
gages, the thermal output characteristics of these alloys are
adjusted to minimize the error over the normal range of
working temperatures.
As indicated by Figure 1, the errors due to thermal output
can become extremely large as temperatures deviate from the
arbitrary reference temperature (ordinarily, room temperature)
with respect to which the thermal output is measured. The
illustration shows distinctly the necessity for compensation
or correction if accurate static strain measurements are to be
made in an environment involving temperature changes.
With respect to the latter statement, it should be remarked
that if it is feasible to bring the gaged test part to the test
temperature in the test environment, maintaining the test part
completely free of mechanically or thermally induced stresses,
0 50 +50 +100 +150 +200 +250
+4000
+1000
+2000
+3000
0
1000
2000
3000
4000
100 0 +100 +200 +300
T
H
E
R
M
A
L

O
U
T
P
U
T

(
F
I

=

2
.
0
)
+400 +500
TEMPERATURE C
TEMPERATURE F
ALLOYS BONDED TO
STEEL SPECIMEN
CONSTANTAN
(FULL HARD)
KARMA (FULL HARD)
NICHROME V
ISOELASTI C
+24C
+75F
Figure 1. Thermal output variation with temperature
for several strain gage alloys (in the as-rolled
metallurgical condition) bonded to steel.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
37
and balance the strain indicator for zero strain under these
conditions, no thermal output error exists when subsequent
strain measurements are made at this temperature. In other
words, when no temperature change occurs between the
stress-free and stressed conditions, strain measurements can
be made without compensating or correcting for thermal
output. In practice, however, it is rare that the foregoing
requirements can be satised, and the stress analyst ordinarily
finds it necessary to take full account of thermal output
effects.
Also, in the case of purely dynamic strain measurements,
where there is no need to maintain a stable zero-strain
reference, thermal output may be of no consequence. This is
because the frequency of the dynamic strain signal is usually
very high with respect to the frequency of temperature
change, and the two signals are readily separable. If, however,
there is combined static/dynamic strain, and the static
component must also be measured, or if the frequency
of temperature change is of the same order as the strain
frequency, thermal output effects must again be considered.
2.1 Compensation for Thermal Output
2.1.1 Compensating (Dummy) Gage
In theory, at least, the error due to thermal output can be
completely eliminated by employing, in conjunction with
the active strain gage, but connected in an adjacent arm of
the Wheatstone bridge circuit, an identical compensating or
dummy gage mounted on an unstrained specimen made
from the identical material as the test part, and subjected
always to the same temperature as the active gage. Under
these hypothetical conditions, the thermal outputs of the
two gages should be identical. And, since identical resistance
changes in adjacent arms of the Wheatstone bridge do not
unbalance the circuit, the thermal outputs of the active
and dummy strain gages should cancel exactly leaving
only the stress-induced strain in the active strain gage to be
registered by the strain indicator. For this to be precisely
true requires additionally that the leadwires to the active and
dummy gages be of the same length and be routed together
so that their temperature-induced resistance changes also
match identically.
The principal problems encountered in this method of
temperature compensation are those of establishing and
maintaining the three sets of identical conditions postulated
above. To begin with, it is sometimes very difcult to arrange
for the placement of an unstrained specimen of the test
material in the test environment; and even more difcult to
make certain that the specimen remains unstrained under all
test conditions. There is a further difculty in ensuring that
the temperature of the compensating gage on the unstrained
specimen is always identical to the temperature of the active
gage. This problem becomes particularly severe whenever
there are temperature gradients or transients in the test
environment. And, as indicated in the preceding paragraph,
the same considerations apply to the leadwires. Finally, it
must be recognized that no two strain gages even from the
same lot or package are precisely identical. For most static
strain measurement tasks in the general neighborhood of
room temperature, the difference in thermal output between
two gages of the same type from the same lot is negligible;
but the difference may become evident (and significant)
when measuring strains at temperature extremes such as
those involved in high-temperature or cryogenic work.
In these instances, point-by-point correction for thermal
output will usually be necessary. With non-self-temperature-
compensated gages, the gage-to-gage differences in thermal
output may be so great as to preclude dummy compensation
for temperatures which are remote from room temperature.
In general, when the three identity criteria already mentioned
can be well satised, the method of compensating with a
dummy gage is a very effective technique for controlling
the thermal output error. There is, moreover, a special class
of strain measurement applications which is particularly
adaptable to compensation of thermal output with a second
gage. This class consists of those applications in which the
ratio of the strains at two different but closely adjacent (or at
least thermally adjacent) points on the test object are known
a priori. Included in this class are bars in pure torsion, beams
in bending, columns, diaphragms, etc., all stressed within
their respective proportional limits. In these applications, the
compensating gage can often be located strategically on the
test member itself so as to provide two active gages which
undergo the same temperature variations while sensing
strains that are preferably opposite in sign and of known
ratio. The two gages in adjacent arms of the Wheatstone
bridge circuit then function as an active half bridge.
For example, when strain measurements are to be made on
a beam which is thin enough so that under test conditions
the temperatures on the two opposite surfaces normal to
the plane of bending are the same, the two strain gages can
be installed directly opposite each other on these surfaces
(Figure 2a). The active half bridge thus formed will give
effective temperature compensation over a reasonable range
of temperatures and, since the strains sensed by the gages
are equal in magnitude and opposite in sign, will double
the output signal from the Wheatstone bridge. Similarly,
for a bar in pure torsion (Figure 2b), the two gages can
be installed adjacent to each other and aligned along the
principal axes of the bar (at 45 to the longitudinal axis). As
in the case of the beam, excellent temperature compensation
can be achieved, along with a doubled output signal.
When making strain measurements along the axis of a
column or tension link, the compensating gage can be
mounted on the test member adjacent to the axial gage and
aligned transversely to the longitudinal axis to sense the
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
38
Poisson strain (Figure 2c). The result, again, is compensation
of the thermal output, accompanied by an augmented
output signal [by the factor (1 + ) in this case]. It should
be borne in mind in this application, however, that the
accuracy of the strain measurement is somewhat dependent
upon the accuracy with which the Poissons ratio of the test
material is known. The percent error in strain measurement
is approximately equal to /(1 + ) times the percent error in
Poissons ratio. A further caution is necessary when strain
gages are mounted transversely on small-diameter rods
(or, for that matter, in small-radius llets or holes). Hines
has shown (see Appendix) that under these conditions the
thermal output characteristics of a strain gage are different
than when the gage is mounted on a at surface of the same
material.
In all strain-measurement applications which involve
mounting the compensating gage on the test object itself,
the relationship between the strains at the two locations must
be known with certainty. In a beam, for example, there must
be no indeterminate axial or torsional loading; and the bar
in torsion must not be subject to indeterminate axial or
bending loads. This requirement for a priori knowledge of
the strain distribution actually places these and most similar
applications in the class of transducers. The same method of
compensation is universally employed in commercial strain
gage transducers; such transducers, however, ordinarily
employ full-bridge circuits and special arrangements of the
strain gages to eliminate the effects of extraneous forces or
moments.
2.1.2 Self-Temperature-Compensated Strain Gages
The metallurgical properties of certain strain gage alloys
in particular, constantan and modied Karma (Micro-
Measurements A and K alloys, respectively) are such that
these alloys can be processed to minimize the thermal output
over a wide temperature range when bonded to test materials
with thermal expansion coefficients for which they are
intended. Strain gages employing these specially processed
alloys are referred to as self-temperature-compensated.
Since the advent of the self-temperature-compensated
strain gage, the requirement for a matching unstrained
dummy gage in the adjacent arm of the Wheatstone bridge
has been relaxed considerably. It is now normal practice when
making strain measurements at or near room temperature to
use a single self-temperature-compensated gage in a quarter-
bridge arrangement (with a three-wire hookup), completing
the bridge circuit with a stable xed resistor in the adjacent
arm (Figure 3). Such bridge-completion resistors, with
very low temperature coefcients of resistance, are supplied
by Micro-Measurements and are incorporated in most
modern strain indicators.
Figure 4 illustrates the thermal output characteristics of
typical A- and K-alloy self-temperature-compensated strain
gages. As demonstrated by the gure, the gages are designed
to minimize the thermal output over the temperature range
from about 0 to +400F [20 to +205C]. When the self-
temperature-compensated strain gage is bonded to material
M
A
M
C
T
P
T
P
ACTIVE
GAGE
L1
L3
L2
e
o
L4
COMPENSATING
GAGE
C
A
C
A
C A
Figure 2. Examples illustrating the use of a second
(compensating) strain gage in an adjacent Wheatstone bridge
arm to cancel the effect of thermal output.
ACTIVE
GAGE L1
L3
L2
BRIDGE
COMPLETION
RESISTOR
e
o
Figure 3. A single self-temperature-compensated strain gage in a
three-wire quarter-bridge circuit exemplies modern strain gage
practice for stress analysis measurements.
(b)
(c)
(a)
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
39
having the thermal expansion coefcient for which the gage
is intended, and when operated within the temperature range
of effective compensation, strain measurements can often be
made without the necessity of correcting for thermal output.
If correction for thermal output is needed, it can be made as
shown in the following sections.
Self-temperature-compensated strain gages can also be
used in the manner described in Section 2.1.1. That is, when
circumstances are such that a pair of matched gages can be
used in adjacent arms of the bridge circuit, with both gages
maintained at the same temperature, and with one of the
gages unstrained (or strained at a determinate ratio to the
other gage), excellent temperature compensation can be
achieved over a wide temperature range.
The designations of Micro-Measurements self-temperature-
compensated strain gages include a two-digit S-T-C number
identifying the nominal thermal expansion coefcient (in
ppm/F) of the material on which the gage will exhibit
optimum thermal output characteristics as shown in Figure
4. Micro-Measurements constantan alloy gages are available
in the following S-T-C numbers: 00, 03, 05, 06, 09, 13, 15, 18,
30, 40, and 50. S-T-C numbers of 30 and higher are intended
primarily for use on plastics. In K alloy, the range of S-T-C
numbers is more limited, and consists of 00, 03, 05, 06, 09,
13, and 15. For reference convenience, Table 1 lists a number
of engineering materials, and gives nominal values of the
Fahrenheit and Celsius expansion coefficients for each,
along with the S-T-C number which would normally be
selected for strain measurements on that material. The table
also identies those test materials used in determining the
published thermal output curves for Micro-Measurements
self-temperature-compensated strain gages.
+500
+400
+300
+200
+100
0
100
200
300
400
500
600
700
800
900
1000
200 100 0 +100 +200 +300 +400 +500
100 50 0 +50 +100 +150 +200 +250
T
H
E
R
M
A
L

O
U
T
P
U
T

(
F
I

=

2
.
0
)
TEMPERATURE F
TEMPERATURE C
TEST SPECIMEN 1018 STEEL
+24 C
+75 F
K-ALLOY
A-ALLOY
Figure 4. Typical thermal output variation with temperature for
self-temperature-compensated constantan (A-alloy) and modied
Karma (K-alloy) strain gages.
TABLE 1NOMINAL THERMAL EXPANSION COEFFICIENTS
OF ENGINEERING MATERIALS
MATERIAL
DESCRIPTION
EXPANSION
COEFFICIENTS**
RECOMMENDED
S-T-C NUMBER
Per F [Per C]
ALUMINA, red 3.0 [5.4] 03
ALUMINUM Alloy,
2024-T4*, 7075-T6
12.9 [23.2] 13*
BERYLLIUM, pure 6.4 [11.5] 06
BERYLLIUM COPPER,
Cu 75, Be 25
9.3 [16.7] 09
BRASS, Cartridge,
Cu 70, Zn 30
11.1 [20.0] 13
BRONZE, Phosphor,
Cu 90, Sn 10
10.2 [18.4] 09
CAST IRON, gray 6.0 [10.8] 06
COPPER, pure 9.2 [16.5] 09
GLASS, Soda, Lime, Silica 5.1 [9.2] 05
INCONEL, Ni-Cr-Fe alloy 7.0 [12.6] 06
INCONEL X, Ni-Cr-Fe alloy 6.7 [12.1] 06
INVAR, Fe-Ni alloy 0.8 [1.4] 00
MAGNESIUM Alloy*,
AZ-31B
14.5 [26.1] 15*
MOLYBDENUM*, pure 2.7 [4.9] 03*
MONEL, Ni-Cu alloy 7.5 [13.5] 06
NICKEL-A, Cu-Zn-Ni alloy 6.6 [11.9] 06
QUARTZ, fused 0.3 [0.5] 00
STEEL Alloy, 4340 6.3 [11.3] 06
STEEL, Carbon,
1008, 1018*
6.7 [12.1] 06*
STEEL, Stainless,
Age Hardenable
(17-4PH)
6.0 [10.8] 06
STEEL, Stainless,
Age Hardenable
(17-7PH)
5.7 [10.3] 06
STEEL, Stainless,
Age Hardenable
(PH15-7Mo)
5.0 [9.0] 05
STEEL, Stainless,
Austenitic (304*)
9.6 [17.3] 09*
STEEL, Stainless,
Austenitic (310)
8.0 [14.4] 09
STEEL, Stainless,
Austenitic (316)
8.9 [16.0] 09
STEEL, Stainless,
Ferritic (410)
5.5 [9.9] 05
TIN, pure 13.0 [23.4] 13
TITANIUM, pure 4.8 [8.6] 05
TITANIUM Alloy,
6AL-4V*
4.9 [8.8] 05*
TITANIUM SILICATE*,
polycrystalline
0.0 [0.0] 00*
TUNGSTEN, pure 2.4 [4.3] 03
ZIRCONIUM, pure 3.1 [5.6] 03
* Indicates type of material used in determining thermal output
data supplied with Micro-Measurements strain gages.
** Nominal values at or near room temperature for temperature
coefcient of expansion values.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
40
If a strain gage with a particular S-T-C number is installed
on a material with a nonmatching coefcient of expansion,
the thermal output characteristics will be altered from
those shown in Figure 4 by a general rotation of the curve
about the room-temperature reference point (see Section
2.2.5). When the S-T-C number is lower than the material
expansion coefcient, the rotation is counterclockwise; and
when higher, clockwise. Rotation of the thermal output
curve by intentionally mismatching the S-T-C number
and expansion coefcient can be used to bias the thermal
output characteristics so as to favor a particular working
temperature range.
2.2 Correction for Thermal Output
Depending upon the test temperature and the degree of
accuracy required in the strain measurement, it will sometimes
be necessary to make corrections for thermal output, even
though self-temperature-compensated gages are used. In any
case, when making strain measurements at a temperature
different from the instrument balance temperature, the
indicated strain is equal to the sum of the stress-induced
strain in the test object and the thermal output of the gage
(plus the strain equivalent of any other resistance changes in
the gage circuit). With the thermal output expressed in strain
units, as in Equation (2), correction for this effect is made
by simply subtracting (algebraically, with sign) the thermal
output from the indicated strain.
As an aid to the user in correcting for temperature-dependent
properties, the Engineering Data Sheet in each package of
Micro-Measurements A- and K-alloy strain gages includes
a graph showing the thermal output and gage-factor
variation with temperature. Figure 5 is typical (for A alloy)
of the graphs supplied with the gages. In addition to plots
of thermal output and gage factor variation, polynomial
equations are provided (in both Fahrenheit and Celsius
units) for the thermal output curve. Also given on the graph
are two other important items of information: (1) the lot
number of the strain gages, and (2) the test material used in
measuring the thermal output characteristics. It should be
noted that the thermal output data are specically applicable
to only gages of the designated lot number, applied to the
same test material.
2.2.1 Simple Procedure
Approximate correction for thermal output can be
accomplished most directly and easily using the formula
(Figure 5) on the gage package data label. This simple
method of correction is based on the fact that the gage
factors of A- and K-alloy gages are close to 2.0, which
is the standardized gage-factor setting employed in
calibrating the gages for thermal output. Adjustment of
the thermal output data for a different instrument gage-
factor setting is described in Section 2.2.2.

The rst step in the correction procedure is to refer to the
graph and read the thermal output corresponding to the
test temperature. Then, assuming that the strain indicator
was balanced for zero strain at room temperature (the
reference temperature with respect to which the thermal
output data were measured), subtract the thermal output
given on the graph from the strain measurements at the
test temperature, carrying all signs. This procedure can be
expressed analytically as follows:

e e e =
T O /
(3)
where:

e
= uncorrected strain measurement, as registered
by the strain indicator.

e
= partially corrected strain indicationthat is,
corrected for thermal output, but not for gage
factor variation with temperature (see Sections
3.0 and 4.0).

T/O
= thermal output, in strain units, from the gage
package data label.
As an example, assume that, with the test part under
no load and at room temperature, the strain indicator
was balanced for zero strain. At the test temperature of
+250F [+121C], the indicated strain is +2300. Referring
to Figure 5, assuming that the graph was the one in the
gage package, the thermal output at test temperature is
100. From Equation (3), the corrected strain is thus 2300
(100) = 2400. Had the indicated strain been negative,
the corrected strain would be: 2300 (100) = 2200.
If the instrument were balanced for zero strain at some
temperature other than +75F [+24C], the value of
T/O
for
use in Equation (3) is the net change in thermal output in
going from the balance temperature to the test temperature.
That is,
T/O
=
T/O
(T
2
)
T/O
(T
1
), carrying the sign of the
thermal output in each case.
TEMPERATURE IN CELSIUS
THERMAL OUTPUT
TESTED ON:
TEMPERATURE IN FAHRENHEIT
TEST PATTERN: CODE: ENG.:
0
+4.0%
0
4.0%
2.0%
A48AF28
2024-T4 ALUMINUM 250BG 101171 GU
+2.0%
+24C
+75F
50 0 +50 +100 +150 +200 +250
0 100 +200 +300 +400 +500
+400
+300
+200
+100
100
400
500
200
300
T
H
E
R
M
A
L

O
U
T
P
U
T

m
/
m
(
B
a
s
e
d

o
n

I
n
s
t
r
u
m
e
n
t

G
.
F
.

o
f

2
.
0
0
)
T
e
m
p
.

C
o
e
f
f
.

o
f

G
a
g
e

F
a
c
t
o
r

=

(
+
1
.
1

0
.
2
)
%
/
1
0
0
C
V
A
R
I
A
T
I
O
N

O
F

G
A
G
E

F
A
C
T
O
R
W
I
T
H
T
E
M
P
E
R
A
T
U
R
E
GAGE FACTOR
THERMAL OUTPUT
+100
TO =8.82x10
1
+2.71x10
0
T2.53x10
2
T
2
+6.72x10
5
T
3
4.03x10
8
T
4
(F)
TO =2.52x10
1
+2.33x10
0
T6.19x10
2
T
2
+3.62x10
4
T
3
4.23x10
7
T
4
(C)
Figure 5. Graph showing typical thermal output
and gage factor variation with temperature.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
41
2.2.2 Adjusting Thermal Output for Gage Factor
It should be noted that the instrument gage factor setting
employed in recording thermal output data is standardized
at 2.0 for all Micro-Measurements A- and K-alloy gages. If,
during strain measurement, the users instrument is set at a
gage factor different from 2.0, the thermal output component
of the indicated strain will differ accordingly from that given in
Figure 5. This difference is usually no more than several
percent when the instrument gage factor is set to that of an
A- or K-alloy gage. A modest improvement in the accuracy
of the thermal output correction can thus be made by
adjusting the data from Figure 5 (taken at F
I
= 2.0) to the
current gage factor setting of the instrument. This is done
as follows:
= e e
T O T O
I
F
/ /
. 2 0
(4)
where:
T/O
= thermal output adjusted for instrument
gage factor setting.

T/O
= thermal output from gage package data
sheet (F
I
= 2.0).
F
I
= instrument gage factor setting during
strain measurement.
Continuing the numerical example, and assuming that the
data sheet gives a room-temperature gage factor of 2.10
for the gage, and that the instrument is set at this same
gage factor, the adjusted thermal output is calculated from
Equation (4):

= =
T O /
.
.
100
2 0
2 1
95 x

And the corrected strain measurements become:
2300 (95) = 2395
and,
2300 (95) = 2205
As shown in Figure 5, the gage factor of the strain gage varies
slightly with temperature. When this effect is signicant
relative to the required accuracy in strain measurement, the
gage factor of the strain gage can be corrected to its test-
temperature value (Section 3.1), and the gage factor of the
instrument set accordingly. The resulting instrument gage
factor is substituted into Equation (4) to obtain the adjusted
thermal output, which is then subtracted algebraically from
the indicated strain to yield the stress-induced strain.
2.2.3 Extensive Data Acquisition
If desired, for extensive strain measurement programs, the
thermal output curve in Figure 5 can be replotted with the
gage factor adjustment either room-temperature or test-
temperature already incorporated. Upon completion, the
thermal output read from the replotted curve can be used
directly to correct the indicated strain. This procedure may
be found worth the effort if many strain readings are to be
taken with one gage or a group of gages from the same lot.
For convenience in computerized correction for thermal
output, Micro-Measurements supplies, for each lot of
A-alloy and K-alloy gages, a regression-tted (least-squares)
polynomial equation representing the thermal output curve
for that lot. The polynomial is of the following form:
T/O
= A
0
+ A
1
T + A
2
T
2
+ A
3
T
3
+A
4
T
4
(5)
where: T = temperature.
If not included directly on the graph, as shown in Figure
5, the coefficients A
i
for Equation (5) can be obtained
from Micro-Measurements on request by specifying the lot
number.
It should be borne in mind that the regression-fitted
equations, like the data from which they are derived, are
based on an instrument gage factor of 2.0; and, for greatest
accuracy, the thermal output values calculated from the
equations must be adjusted to the gage factor setting of
the instrument if other than 2.0. As an alternative, the A
i

coefcients in Equation (5) can be multiplied by the ratio
2.0/F
I
, where F
I
is the instrument gage factor used for
strain measurement. Another consideration which should
not be overlooked is that the supplied thermal output
data and equations are applicable only to the specied lot
of gages, bonded to the identical material used by Micro-
Measurements in performing the thermal output tests.
2.2.4 Accuracy and Practicality
First-Hand Measurement of Thermal Output
There is a limit as to just how far it is practical to go
in adjusting the manufacturers thermal output data in
an attempt to obtain greater accuracy. In the rst place,
the thermal output curve provided on the gage package
data label (or by the polynomial equation) represents
an average, since there is some variation in thermal
output characteristics from gage to gage within a lot.
And the width of the scatterband increases as the test
temperature departs further and further from the room-
temperature reference.
Furthermore, the thermal output data given in the gage
package were necessarily measured on a particular lot of
a particular test material (see Table 1). Different materials
with the same or closely similar nominal expansion
coefcients, and even different lots and forms of the same
material, may have signicantly different thermal expansion
characteristics.
From the above considerations, it should be evident that in
order to achieve the most accurate correction for thermal
output it is generally necessary to obtain the thermal output
data with the actual test gage installed on the actual test part.
For this purpose, a thermocouple or resistance temperature
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
42
sensor is installed immediately adjacent to the strain gage.
The gage is then connected to the strain indicator and, with
no loads applied to the test part, the instrument is balanced
for zero strain. Subsequently, the test part is subjected to
the test temperature(s), again with no loads applied, and
the temperature and indicated strain are recorded under
equilibrium conditions. If, throughout this process, the part
is completely free of mechanical and thermal stresses, the
resulting strain indication at any temperature is the thermal
output at that temperature. If the instrument gage factor
setting during subsequent strain measurement is the same
as that used for thermal output calibration, the observed
thermal output at any test temperature can be subtracted
algebraically from the indicated strain to arrive at the
corrected strain. Otherwise, the thermal output data should
be adjusted for the difference in gage factor settings, as
described in Section 2.2.2, prior to subtraction.
In order to correct for thermal output in the manner described
here, it is necessary, of course, to measure the temperature at
the strain gage installation each time a strain measurement
is made. The principal disadvantage of this procedure is that
two channels of instrumentation are preempted for each
strain gage one for the strain gage proper, and one for the
thermocouple or resistance temperature sensor.
2.2.5 S-T-C Mismatch
When a strain gage is employed on a material other than
that used in obtaining the manufacturers thermal output
data for that lot of gages, an S-T-C mismatch occurs.
In such cases, the thermal output of the gage will differ
from the curve supplied in the gage package. Consider,
for example, strain measurements made at an elevated
temperature on Monel with a strain gage of 06 S-T-C
number, calibrated for thermal output on 1018 steel (Table
1). The thermal expansion characteristics of Monel are
somewhat different from 1018 steel, and the strain gage will
produce a correspondingly different thermal output. Thus,
if accurate strain measurement is required, the thermal
output characteristics of the gage bonded to Monel must
be measured over the test temperature range as described
in Section 2.2.4. For small temperature excursions from
room temperature, the effect of the difference in expansion
properties between Monel and 1018 steel is not very
signicant, and would commonly be ignored.
On the other hand, when the difference in thermal expansion
properties between the thermal output calibration material
and the material to which the gage is bonded for stress
analysis is great, the published thermal output curve cannot
be used directly for making corrections. Examples of this
occur in constantan strain gages with S-T-C numbers of
30, 40, and 50. The principal application of these gages
would normally be strain measurement on high-expansion-
coefcient plastics. But the thermal (and other) properties
of plastics vary signicantly from lot to lot and, because
of formulation differences, even more seriously from
manufacturer to manufacturer of nominally the same
plastic. This fact, along with the general instability of
plastics properties with time, temperature, humidity, etc.,
creates a situation in which there are no suitable plastic
materials for use in directly measuring the thermal output
characteristics of gages with S-T-C numbers of 30 and
above. As an admittedly less-than-satisfactory alternative,
the thermal output data provided with these gages are
measured on 1018 steel specimens because of the stability
and repeatability of this material.
As a result of the foregoing, it is always preferable when
measuring strains on plastics or other materials with 30,
40 or 50 S-T-C gages (at temperatures different from the
instrument balance temperature) to first experimentally
determine the thermal output of the gage on the test material
as described in Section 2.2.4. Using these data, corrections
are then made as usual by subtracting algebraically the
thermal output from the measured strain.
For use as a quick first approximation, the thermal
output characteristics of 30, 40, or 50 S-T-C gages on a
plastic (or on any other material) of known coefficient
of expansion can be estimated by reversing the clockwise
rotation of the thermal output curve which occurred
when measuring the characteristics on a steel specimen.
Assume, for example, that a 30 S-T-C gage is to be used
for strain measurements on a plastic with a constant
expansion coefcient of 35 x 10
-6
/F (63 x 10
-6
/C) over
the test temperature range. Assume also that the expansion
coefcient of 1018 steel is constant at 6.7 x 10
-6
/F (12.1 x
10
-6
/C) over the same temperature range. With the strain
indicator set at the gage factor of the strain gage, so that F
I

=F
G
, and noting that the ratio (1 + K
t
)/(1
0
K
t
) is normally
close to unity for A-alloy gages, Eq. (2) can be rewritten in
simplied (and approximate) form as follows:

T O
G
G
G S
F
T T
/
=
+

(6)
(Note: Although the remainder of this example is
carried through in only the Fahrenheit system to avoid
overcomplicating the notation, the same procedure produces
the equivalent result in the Celsius system.)
When specically applied to 6.7 and 35 x 10
-6
/F materials,
Equation (6) becomes:

T O
G
G
G
F
T T
/ .
.
6 7
6 7
( )
=
+
(7a)
and,

T O
G
G
G
F
T T
/ 35
35
( )
=
+

(7b)
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
43
Solving Eq. (7a) for:

T O
G
G
G
F
T T
/ 35
35
( )
=
+
T, and substituting
into Equation (7b),

T/O(35)
=
T/O(6.7)
+ (35 6.7)T (8)
In words, Equation (8) states that the thermal output curve
for the 30 S-T-C gage mounted on 1018 steel can be converted
to that for the same gage mounted on a 35 x 10
-6
/F plastic
by adding to the original curve the product of the difference
in expansion coefficients and the temperature deviation
from room temperature (always carrying the proper sign
for the temperature deviation). Figure 6 shows the thermal
output curve for a 30 S-T-C gage as originally measured on
a 1018 steel specimen, and as rotated counterclockwise to
approximate the response on a plastic with an expansion
coefcient of 35 x 10
-6
/F.
The procedure just demonstrated is quite general, and can
be used to predict the approximate effect of any mismatch
between the expansion coefcient used for obtaining the
thermal output curve on the gage package data sheet and
the expansion coefcient of some other material on which
the gage is to be installed. Although generally applicable, the
procedure is also limited in accuracy because the expansion
coefficients in Equation (6) are themselves functions of
temperature for most materials. A further limitation in
accuracy can occur when measuring strains on plastics or
other materials with poor heat transfer characteristics. If,
due to self-heating, the temperature of the strain gage is
signicantly higher than that of the test part, the thermal
output data supplied in the gage package cannot be applied
meaningfully.
It should be borne in mind that the foregoing procedure gives,
at best, a rough approximation to the actual thermal output
when there is a mismatch between the expansion coefcient of
the test material and the selected S-T-C number of the strain
gage. When accurate correction for thermal output is required,
direct measurement, as described in Section 2.2.4, is highly
recommended.
3.0 Gage Factor Variation with Temperature
The alloys used in resistance strain gages typically exhibit a
change in gage factor with temperature. In some cases, the
error due to this effect is small and can be ignored. In others,
depending upon the alloy involved, the test temperature, and
the required accuracy in strain measurement, correction for
the gage factor variation may be necessary.
It can be seen from Figure 7 that the effect in the A alloy
is essentially linear, and quite small at any temperature,
typically being in the order of 1% or less per 100F [2%
or less per 100C]. Thus, for a temperature range of, say,
100F [50C], about room temperature, correction may
not be necessary. At more extreme temperatures, when
justied by accuracy requirements, the correction can be
made as shown in Section 3.1, or combined with the thermal
output correction as in Section 4.0.
The variation of gage factor in the D alloy, while very
modest and at between room temperature and +200F
[+95C], steepens noticeably outside of this range. However,
even for temperatures where the gage factor deviation is
several percent, correction may not be practical. This is
because D alloy is used primarily for purely dynamic strain
measurement, under which conditions other errors in the
measurement system may greatly overshadow the gage
factor effect.
+100 +200 +300 +400 +500
A S-T-C 30 FOIL BONDED TO 1018 STEEL ( = 6.7 x 10
6
/F) AS MEASURED.
B APPROXIMATE RESPONSE EXPECTED FROM S-T -C 30 FOIL BONDED TO MATERIAL
WITH = 35 x 10
6
/F.
TEMPERATURE F
TEMPERATURE C
T
H
E
R
M
A
L

O
U
T
P
U
T
,

(
B
a
s
e
d

o
n

I
n
s
t
r
u
m
e
n
t

G
.
F
.

o
f

2
.
0
0
)
+8000
+6000
+4000
+2000
0
2000
4000
6000
8000
10 000
50 0 +50 +100 +150 +200 +250
0 100
A
B
+24C
+75F
Figure 6. Rotation of the thermal output function [from (A) to (B)]
when a strain gage is installed on a material of higher thermal
expansion coefcient than that used by the manufacturer
in S-T-C calibration.
TEMPERATURE C
TEMPERATURE F
+24C
+75F
50 0 +50 +100 +150 +200 +250
G
A
G
E

F
A
C
T
O
R

V
A
R
I
A
T
I
O
N

%
F
R
O
M

+
7
5
F

(
+
2
4
C
)

V
A
L
U
E
100 0 +100 +200 +300 +400 +500
3
4
2
1
0
+1
+2
+3
5
A
D
A-ALLOY
D-ALLOY
Figure 7. Gage factor variation with temperature for constantan
(A-alloy) and isoelastic (D-alloy) strain gages.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
44
As shown in Figure 8, the gage factor variation with
temperature for modified Karma (K alloy) is distinctly
different from that of the A and D alloys. The gage factor
variation is nearly linear with temperature, as it is for A
alloy, but the slope is negative and is a function of the S-T-C
number, becoming steeper with higher numbers.
3.1 Correcting Strain Measurements for
Gage Factor Variation with Temperature
The standard procedure for measuring the gage factor of
a lot of any particular type of strain gage is performed at
room temperature. It is this value of the gage factor, along
with its tolerance, which is given on the gage package data
label of Micro-Measurements strain gages. Thus, at any
temperature other than room temperature the gage factor
is different, and a correction may be needed, according
to the circumstances. Also given on each data sheet is the
applicable graph of gage factor variation with temperature,
such as those in Figures 7 and 8. This information is all that
is required to make the correction.
In general, any strain measurement data can be corrected (or
adjusted) from one gage factor to another with a very simple
relationship. Assume, for instance, that a strain, e
1
, was
registered with the gage factor setting of the strain indicator
at F
1
, and it is desired to correct the data to a gage factor of
F
2
. The corrected strain, e
2
, is calculated from:
e e
2 1
= x
1
2
F
F
(9)
When correcting for gage factor variation with temperature,
F
1
can be taken as the package-data room-temperature gage
factor at which the strain indicator may have been set, and
F
2
the gage factor at the test temperature. Of course, when
the test temperature is known with reasonable accuracy
in advance, the gage factor control of the strain indicator
can be set at F
2
, initially, and no correction is necessary. It
should be noted in this case, however, that if thermal output
corrections are to be made from the graph (or polynomial
equation) on the gage package data label, the thermal output
data must be adjusted from a gage factor of 2.0 (at which the
thermal output was measured) to the test temperature gage
factor, F
2
, being used for strain measurement.
The following relationship is used to determine the gage
factor at the test temperature from the tabular and graphical
data supplied in the gage package:
F F
F
2 1
= +
1
100
(%)
(10)
where: F(%) = percent variation in gage factor with
temperature as shown in Figures 7 and
8. (Note: The sign of the variation must
always be included.)
As a numerical example, using Equations (9) and (10),
assume that the room-temperature gage factor of a 13 S-T-C,
K-alloy gage is 2.05 and, with the instrument set at this value,
the strain indication at +450F [+230C] is 1820e. Referring
to Figure 8, F(%) for this case is 3, and, from Equation (10),
F
2
= 2.05 (1 0.03) = 1.99
Substituting into Eq. (9),

F
2
x
=
= =
2 05 1 0 03 1 99
1820
2 05
1 99
1875
2
. ( . ) .
.
.

Since gage factor variation with temperature affects both the
thermal output and the stress-induced strain, and because
confusion may arise in making the corrections individually
and then combining them, the following section gives
equations for performing both corrections simultaneously.
4.0 Simultaneous Correction of
Thermal Output and Gage Factor Errors
Relationships are given in this section for correcting indicated
strains for thermal output and gage factor variation with
temperature. The forms these relationships can take depend
upon the measuring circumstances primarily upon the
strain indicator gage factor setting and the temperature at
which the instrument was balanced for zero strain.
The strain indicator gage factor can be set at any value
within its control range, but one of the following three is
most likely:
1. Gage factor, F*, used by Micro-Measurements in
determining thermal output data (F* = 2.0).
TEMPERATURE C
TEMPERATURE F
S-T-C 03
06
09
1
3
+24C
+75F
50 0 +50 +100 +150 +200 +250
G
A
G
E

F
A
C
T
O
R

V
A
R
I
A
T
I
O
N

%
F
R
O
M

+
7
5
F

(
+
2
4
C
)
V
A
L
U
E
100 0 +100 +200 +300 +400 +500
3
4
2
1
0
+1
+2
+3
Figure 8. Variation of K-alloy gage factor with
temperature and S-T-C number.
The instrument gage factor setting should not be changed during
a test (after zero-balance), since this may introduce a zero shift.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
45
2. Room-temperature gage factor as given on the gage
package data label.
3. Gage factor of gage at test temperature or at any
arbitrary temperature other than room or test
temperature.
No single gage factor is uniquely correct for this situation;
but, of the foregoing, it will be found that selecting the rst
alternative generally leads to the simplest form of correction
expression. Because of this, the procedure developed here
requires that the gage factor of the instrument be set at
F
I
= F* = 2.0, the gage factor at which the thermal output
data were recorded.
Similarly, the strain indicator can be balanced for zero strain
at any one of several strain gage temperatures:
1. Room temperature
2. Test temperature
3. Arbitrary temperature other than room or test
temperature
The second and third of the above choices can be used
for meaningful strain measurements only when the test
object is known to be completely free of mechanical and
thermal stresses at the balancing temperature. Because this
requirement is usually difcult or impossible to satisfy, the
rst alternative is generally preferable, and is thus selected
for the following procedure.
As an example, assume that the strain indicator is balanced
with the gage at room temperature, and with the gage factor
control set at F*, the value used by Micro-Measurements
in recording the thermal output data. Assume also that a
strain

e
1
is subsequently indicated at a temperature T
1
which
is different from room temperature. The indicated strain
e
1
is generally in error due both to thermal output and to
variation of the gage factor with temperature and hence
the double caret over the strain symbol.
Consider rst the correction for thermal output. Since the
gage factor setting of the strain indicator coincides with that
used in measuring the thermal output, this correction can be
made by direct subtraction of the thermal output (as given
on the gage package data label) from the indicated strain.
That is,

e
1
=
e
1
e
T/O
(T
1
)
where:

e
1
= indicated strain, uncorrected for either thermal
output or gage factor variation with temperature.

e
1

= semi-corrected strain; i.e., corrected for thermal
output only.
e
T/O
(T
1
) = thermal output at temperature T
1
(functional
notation is used to avoid double and triple sub-
scripts).
Next, correction is made for the gage factor variation with
temperature. Because the strain measurement was made at a
gage factor setting of F*, the correction to the gage factor
at the test temperature is performed with Equation (9) as
follows:
e e
1 1
1
=
F
F T
*
( )

where:
1
= strain magnitude corrected for both thermal
output and gage factor variation with
temperature.
F(T
1
) = gage factor at test temperature.
Combining the two corrections,

[ ]

1 1
=
T O 1
1
T
F
F T
/
( )
*
( )
(11)
When the prescribed conditions on the gage factor setting
and the zero-balance temperature have been met, the strain
e
1
from Equation. (11) is the actual strain induced by
mechanical and/or thermal stresses in the test object at the
test temperature. As a numerical example of the application
of Equation (11), assume the following:
Strain gage WK-06-250BG-350
Test material Steel
Room-temperature gage factor, F
0
2.07
Test temperature 50F [45C]

e
1
, indicated strain at test temperature
(with instrument gage factor set at F*) 1850e
e
T/O
(T
1
), thermal output at
test temperature 200e
F(T
1
), deviation at test
temperature from
room-temperature gage factor +0.6%
Using Equation (10) to obtain F(T
1
), the gage factor of the
gage at test temperature,
Substituting into Eq. (11), with F* = 2.0,

F T F
F T
( )
.
. .
( ) .
( )
.
.
1 0
1
1
1
0 6
100
2 07 1 006
2 08
1850 200
2 0
2 08
1587
= +
=
=
= [ ] =
x

From engineering data on gage package.
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
46
For what might appear to be a more complex case, consider a
strain-gage-instrumented centrifugal compressor, operating
rst at speed N
1
, with the temperature of the gage installation
at T
1
. Under these conditions, the indicated strain is
e
1
. The
compressor speed is then increased to N
2
, with a resulting
gage installation temperature of T
2
, and an indicated strain
e
2
. The engineer wishes to determine the change in stress-
induced strain caused by the speed increase from N
1
to N
2
.
This problem is actually no more difcult than the previous
example. Applying Equation (11) to each condition:
e e e
e e e
1 1 1
1
2 2 2
2
=
[ ]
=
[ ]
T O
T O
T
T
F
F T
F
F T
/
/
( )
( )
*
( )
*
( )

The same numerical substitution procedure is followed as
before, and the results subtracted to give (e
2
e
1
), the change
in stress-induced strain caused by the speed increase. The
subtraction can also be done algebraically to yield a single
equation for the strain change:

e e
e e e e
2 1
2 2
2
1 1
1
=

F
F T
T
F T
T O T O
T
*
( )
( )
( )
/ /
( )

When computerized data reduction is used, analytical
expressions for the functions e
T/O
(T) and F(T) can be
introduced into the program to permit direct calculation of
corrected strains from indicated strains.
Appendix
Surface Curvature Effects on Thermal Output
Frank F. Hines has demonstrated that when a strain gage
is installed on a sharply curved surface, the thermal output
manifested by such an installation is different than for the
same gage mounted on a at surface. The curvature-indicated
change in thermal output, referred to here as the incremental
thermal output, is due to the fact that the strain-sensitive
grid of the gage is above the surface of the test member by
the thickness of the gage backing and adhesive layer. It can
be shown that under these conditions a temperature change
causes a different strain in the grid than would occur with
the grid bonded to a plane surface. The result is an altered
thermal output from the data provided in the gage package.
The curvature-induced incremental thermal output is a second-
order effect which can ordinarily be ignored; but it can become
signicant when the radius of curvature is very small. As a rule
of thumb, the incremental thermal output can be neglected
when the radius of curvature is 0.5 in (13 mm) or greater. With
smaller radii, correction may be necessary, depending upon the
required strain-measurement accuracy.
Employing the same basic approach and approximations
used by Hines in his derivation, but generalizing the treatment
to allow for any combination of adhesive and backing
properties, an expression for estimating the incremental
thermal output can be written as follows:
(A-1)

T /O
A B A A B B A B S A B
R
h h h h T
=
+ ( ) + ( ) + ( )
[ ]

1
1 2 2
where, in consistent units,
e
T/O
= curvature-induced incremental thermal output.
R = radius of curvature of test surface at gage site.

AB
= average Poissons ratio of adhesive and backing.
h
A
,h
B
= adhesive and backing thickness, respectively.

A
,
B
= thermal expansion coefcients of adhesive and
backing, respectively.

S
= thermal expansion coefcient of substrate (spec-
imen material).
T = temperature change from reference temperature.
Approximate values for the adhesive and backing parameters
in Equation (A-1) are given in Table A-1 for representative
Micro-Measurements adhesives and gage series. The sign of
the incremental thermal output is obtained from Equation
(A-1) when the signs of T and R are properly accounted
for that is, an increase in temperature from the reference
temperature is taken as positive, and a decrease negative;
and correspondingly, a convex curvature is positive, while
a concave curvature is negative. The calculated result
from Equation (A-1) is then added algebraically to the
thermal output data supplied in the gage package to give
the curvature-corrected thermal output for use in making
thermal output corrections as shown in this Tech Note.
Proceedings, Western Regional Strain Gage Committee, Nov. 9,
1960, pp. 39-44.
TABLE A-1
Adhesive and Backing Parameters for Use with Equation (A-1)
Adhesive Type
M-Bond 200
M-Bond AE-10/15
M-Bond 600/610
h
A
, h
B
A
,
B
in [mm] per F [per C]
0.0006
0.001
0.0002

0.0012
0.001
0.0015
[0.015]
[0.025]
[0.005]
[0.030]
[0.025]
[0.038]
45 x 10
-6
45
45
50 x 10
-6
10
10
[81 x 10
-6
]
[81]
[81]
[90 x 10
-6
]
[18]
[18]
Gage Series
(Backings)
EA, CEA, EP, ED
SA, SK, SD, S2K
WA, WK, WD
AB
= 0.35 for all combinations
T
E
C
H

N
O
T
E
TN-504-1
Micro-Measurements
47
Because the adhesive and backing parameters given in Table
A-1 are approximate, and are affected by gage installation
tech-nique and other variables, the curvature correction
dened by Equation (A-1) is limited in accuracy. When the
surface curvature is severe enough so that the curvature-
induced incremental thermal output may be important,
the actual thermal output should be measured as described
in Section 2.2.4 of the text. In other words, the strain gage
should be bonded to the test part as for strain measurement,
a thermocouple or resistance temperature sensor should
be installed adjacent to the gage, and the test part then
subjected to test temperatures (while free of mechanical and
thermal stresses) to record the true thermal output.
As an aid in judging the approximate magnitude of the
curvature-induced thermal output, Equation (A-1) has been
evaluated for several representative combinations of Micro-
Measurements adhesives and gage series. Parameters from
Table A-1 were substituted into the equation, along with
S

= 6.0 x 10
6
(assuming a steel test material), and the results
plotted in Figure A-1. Note, in the gure, that the ordinate
gives the incremental thermal output per unit of temperature
change from the initial reference temperature that is,
e
T/O
/T.
0.08 0.10 0.15 0.20 0.30 0.40 0.50 0.60 0.80 1.00
2.0 3.0 4.0 5.0 6.0 8.0 10.0 15.0 20.0 25.0
1
2
3
4
0
1
2
3
4
5
6
7
0.06 0.05
1.5
I
N
C
R
E
M
E
N
T
A
L

T
H
E
R
M
A
L

O
U
T
P
U
T
P
E
R

U
N
I
T

O
F
T
E
M
P
E
R
A
T
U
R
E

C
H
A
N
G
E
,

T
/
O
/
T
RADIUS, R in
RADIUS, R mm
"W" BACKING & 600/610 ADHESIVE
GAGES MOUNTED ON STEEL :
S
= 6 x 10
-6
/F (10.84 x 10
-6
/ C)
P
E
R

C
P
E
R

F
"E" BACKING & 600/610 ADHESIVE
"E" BACKING & AE-10/15 ADHESIVE
Figure A-1. Equation (A-1) evaluated and plotted for various
standard Micro-Measurements strain gage backing materials
when bonded to a steel substrate.
Tech Note TN-505-4
MICRO-MEASUREMENTS
Strain Gage Selection:
Criteria, Procedures, Recommendations
T
E
C
H

N
O
T
E
49
1.0 Introduction
The initial step in preparing for any strain gage installation
is the selection of the appropriate gage for the task. It
might at rst appear that gage selection is a simple exercise,
of no great consequence to the stress analyst; but quite
the opposite is true. Careful, rational selection of gage
characteristics and parameters can be very important in:
optimizing the gage performance for specied environmental
and operating conditions, obtaining accurate and reliable
strain measurements, contributing to the ease of installation,
and minimizing the total cost of the gage installation.
The installation and operating characteristics of a strain
gage are affected by the following parameters, which are
selectable in varying degrees:
strain-sensitive alloy
backing material (carrier)
grid resistance
gage pattern
self-temperature compensation number
gage length
options
Basically, the gage selection process consists of determining
the particular available combination of parameters which
is most compatible with the environmental and other
operating conditions, and at the same time best satisfies
the installation and operating constraints. These
constraints are generally expressed in the form of
requirements such as:
accuracy test duration
stability cyclic endurance
temperature ease of installation
elongation environment
The cost of the strain gage itself is not ordinarily a prime
consideration in gage selection, since the signicant economic
measure is the total cost of the complete installation, of
which the gage cost is usually but a small fraction. In many
cases, the selection of a gage series or optional feature
which increases the gage cost serves to decrease the total
installation cost.
It must be appreciated that the process of gage selection
generally involves compromises. This is because parameter
choices which tend to satisfy one of the constraints or
requirements may work against satisfying others. For
example, in the case of a small-radius fillet, where the
space available for gage installation is very limited, and
the strain gradient extremely high, one of the shortest
available gages might be the obvious choice. At the
same time, however, gages shorter than about 0.125 in
[3 mm] are generally characterized by lower maximum
elongation, reduced fatigue life, less stable behavior, and
greater installation difculty. Another situation which often
inuences gage selection, and leads to compromise, is the
stock of gages at hand for day-to-day strain measurements.
While compromises are almost always necessary, the
stress analyst should be fully aware of the effects of such
compromises on meeting the requirements of the gage
installation. This understanding is necessary to make the best
overall compromise for any particular set of circumstances,
and to judge the effects of that compromise on the accuracy
and validity of the test data.
The strain gage selection criteria considered here relate
primarily to stress analysis applications. The selection
criteria for strain gages used on transducer spring elements,
while similar in many respects to the considerations
presented here, may vary signicantly from application to
application and should be treated accordingly. The Micro-
Measurements Transducer Applications Department can
assist in this selection.
2.0 Gage Selection Parameters
2.1 Strain-Sensing Alloys
The principal component which determines the operating
characteristics of a strain gage is the strain-sensitive alloy
used in the foil grid. How ever, the alloy is not in every case
an independently selectable parameter. This is because each
Micro-Measurements strain gage series (identifed by the frst
two, or three, letters in the alphanumeric gage designation)
is designed as a complete system. That system is comprised
of a particular foil and backing combination, and usually
incorporates additional gage construction features (such as
encapsulation, integral leadwires, or solder dots) specifc to
the series in question.
Micro-Measurements supplies a variety of strain gage alloys as
follows (with their respective letter designations):
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
50
Strain Gage Selection: Criteria, Procedures, Recommendations
A: Constantan in self-temperature-compensated form.
P: Annealed constantan.
D: Isoelastic.
K: Ni ckel- chromium al loy, a modi f i ed Karma i n
self-temperature-compensated form.
2.1.1 Constantan Alloy
Of all modern strain gage alloys, constantan is the oldest,
and still the most widely used. This situation reects the
fact that constantan has the best overall combination of
properties needed for many strain gage applications. This
alloy has, for example, an adequately high strain sensitivity,
or gage factor, which is relatively insensitive to strain level
and temperature. Its resistivity is high enough to achieve
suitable resistance values in even very small grids, and its
temperature coefficient of resistance is not excessive. In
addition, constantan is characterized by good fatigue life
and relatively high elongation capability. It must be noted,
however, that constantan tends to exhibit a continuous drift at
temperatures above +150F [+65C]; and this characteristic
should be taken into account when zero stability of the
strain gage is critical over a period of hours or days.
Very importantly, constantan can be processed for self-
temperature-compensation (see box at right) to match a
wide range of test material expansion coefcients. A alloy is
supplied in self-temperature-compensation (S-T-C) numbers
00, 03, 05, 06, 09, 13, 15, 18, 30, 40 and 50, for use on test
materials with corresponding thermal expansion coefcients
(expressed in ppm/F).
For the measurement of very large strains, 5% (50 000)
or above, annealed constantan (P alloy) is the grid material
normally selected. Constantan in this form is very ductile;
and, in gage lengths of 0.125 in [3 mm] and longer, can be
strained to >20%. It should be borne in mind, however,
that under high cyclic strains the P alloy will exhibit some
permanent resistance change with each cycle, and cause
a corresponding zero shift in the strain gage. Because of
this characteristic, and the tendency for premature grid
failure with repeated straining, P alloy is not ordinarily
recommended for cyclic strain applications. P alloy is
available with S-T-C numbers of 08 and 40 for use on metals
and plastics, respectively.
2.1.2 Isoelastic Alloy
When purely dynamic strain measurements are to be
made that is, when it is not necessary to maintain a
stable reference zero isoelastic (D alloy) offers certain
advantages. Principal among these are superior fatigue life,
compared to A alloy, and a high gage factor (approximately
3.2) which improves the signal-to-noise ratio in dynamic
testing.
D alloy is not subject to self-temperature-compensation.
More over, as shown in the graph (see box), its thermal
output is so high (about 80/F [145/C]) that this
alloy is not normally usable for static strain measurements.
There are times, however, when D alloy nds application in
special-purpose transducers where a high output is needed,
and where a full-bridge arrangement can be used to achieve
reasonable temperature compensation within the circuit.

Self-Temperature-Compensation
An important property shared by constantan and
modied Karma strain gage alloys is their responsiveness
to special processing for self-temperature-compensation.
Self-temperature-compensated strain gages are designed
to produce minimum thermal output (temperature-
induced apparent strain) over the temperature range
from about 50 to +400F [45 to +200C]. When
selecting either constantan (A-alloy) or modified Karma
(K-alloy) strain gages, the self-temperature-compensation
(S-T-C) number must be specifed. The S-T-C number is the
approximate thermal expansion coefficient in ppm/F of
the structural material on which the strain gage will display
minimum thermal output.
The accompanying graph illustrates typical thermal output
characteristics for A and K alloys. The thermal output of
uncompensated isoelastic alloy is included in the same graph
for comparison purposes. In normal practice, the S-T-C
number for an A- or K-alloy gage is selected to most closely
match the thermal expansion coefcient of the test material.
However, the thermal output curves for these alloys can be
rotated about the room-temperature reference point to favor
a particular temperature range. This is done by intentionally
mismatching the S-T-C number and the expansion coeffcient
in the appropriate direction. When the selected S-T-C number
is lower than the expansion coefcient, the curve is rotated
counterclockwise. An opposite mismatch produces clockwise
rotation of the thermal output curve. Under conditions of
S-T-C mismatch, the thermal output curves for A and K alloys
do not apply, of course, and it will generally be necessary to
calibrate the installation for thermal output as a function of
temperature.
For additional information on strain gage temperature effects,
see Tech Note TN-504.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
51
Other properties of D alloy should also be noted when
considering the selection of this grid material. It is, for
instance, magnetoresistive; and its response to strain is
somewhat nonlinear, becoming signicantly so at strains
beyond 5000.
2.1.3 Karma Alloy
Modified Karma, or K alloy, with its wide areas of
application, represents an important member in the family
of strain gage alloys. This alloy is characterized by good
fatigue life and excellent stability; and is the preferred choice
for accurate static strain measurements over long periods
of time (months or years) at room temperature, or lesser
periods at elevated temperature. It is recommended for
extended static strain measurements over the temperature
range from 452 to +500F [269 to +260C]. For short
periods, encapsulated K-alloy strain gages can be exposed
to temperatures as high as +750F [+400C]. An inert
atmosphere will improve stability and extend the useful gage
life at high temperatures.
Among its other advantages, K alloy offers a much
flatter thermal output curve than A alloy, and thus
permits more accurate correction for thermal output
errors at temperature ex tremes. Like constantan, K
alloy can be self-temperature-compensated for use on
materials with different thermal expansion coefficients.
The available S-T-C numbers in K alloy are limited,
however, to the following: 00, 03, 05, 06, 09, 13, and 15.
K alloy is the normal selection when a temperature-
compensated gage is required that has environmental
capabilities and performance characteristics not attainable
in A-alloy gages.
Due to the difculty of soldering directly to K alloy, the
duplex copper feature, which was formerly offered as an
option, is now standard on all Micro-Measurements open-
faced strain gages produced with K alloy. The duplex copper
feature is a precisely formed copper soldering pad (DP) or
dot (DD), depending on the available tab area. All K-alloy
gages which do not have leads or solder dots are specied
with DP or DD as part of the designation (in place of, or
in addition to, the option specifer). The specifc style of
copper treatment will be advised when the Customer Service
Department is contacted. Open-faced K-alloy gages may also
be ordered with solder dots.
2.2 Backing Materials
Conventional foil strain gage construction involves a photo-
etched metal foil pattern mounted on a plastic backing or
carrier. The backing serves several important functions:
provides a means for handling the foil pattern during
installation
presents a readily bondable surface for adhering the
gage to the test specimen
provides electrical insulation between the metal foil
and the test object
Backing materials supplied on Micro-Measurements
strain gages are of two basic types: polyimide and glass-
fiber-reinforced epoxy-phenolic. As in the case of the
strain-sensitive alloy, the backing is not completely an
independently speciable parameter. Certain backing and
alloy combinations, along with special construction features,
are designed as systems, and given gage series designations.
As a result, when arriving at the optimum gage type for
a particular application, the process does not permit the
arbitrary combination of an alloy and a backing material,
but requires the specication of an available gage series.
Micro-Measurements gage series and their properties are
described in the following Section 2.3. Each series has its
own characteristics and preferred areas of application; and
selection recommendations are given in the tables that follow.
The individual backing materials are discussed here, as the
alloys were in the previous section, to aid in understanding
the properties of the series in which the alloys and backing
materials occur.
The Micro-Measurements polyimide E backing is a tough
and extremely exible carrier, and can be contoured readily
to t small radii. In addition, the high peel strength of the
foil on the polyimide backing makes polyimide-backed gages
less sensitive to mechanical damage during installation.
With its ease of handling and its suitability for use over the
temperature range from 320 to +350F [195 to +175C],
polyimide is an ideal backing material for general-purpose
static and dynamic stress analysis. This backing is capable
of large elongations, and can be used to measure plastic
strains in excess of 20%. Polyimide backing is a feature of
Micro-Measurements EA-, CEA-, EP-, EK-, S2K-, N2A-,
and ED-Series strain gages.
For outstanding performance over the widest range of
temperatures, the glass-fiber-reinforced epoxy-phenolic
backing material is the most suitable choice. This backing
can be used for static and dynamic strain measurement
from 452 to +550F [269 to +290C]. In short-term
applications, the upper temperature limit can be extended
to as high as +750F [+400C]. The maximum elongation
of this carrier material is limited, however, to about 1 to
2%. Reinforced epoxy-phenolic backing is employed on the
following gage series: WA, WK, SA, SK, WD, and SD.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
52
STANDARD STRAIN GAGE SERIES SELECTION CHART
Gage
Series
Description and Primary Application Temperature Range Strain Gage
Fatigue Life
Strain
Level
in
Number
of Cycles
EA Constantan foil in combination with a tough, flexible,
polyimide backing. Wide range of options available.
Primarily intended for general-purpose static and
dynamic stress analysis. Not recommended for highest
accuracy transducers.
Normal:
100 to +350F [75 to +175C]
Special or Short-Term:
320 to +400F [195 to +205C]
3% for gage
lengths under
1/8 in [3.2 mm];
5% for 1/8 in
and over
1800
1500
1200
10
5
10
6
10
8
CEA
Universal general-purpose strain gages. Constantan
grid completely encapsulated in polyimide, with large,
rugged copper-coated tabs. Primarily used for general-
purpose static and dynamic stress analysis. C-Feature
gages are specially highlighted throughout the gage
listings of our Precision Strain Gages Data Book.
Normal:
100 to +350F [75 to +175C]
Stacked rosettes limited to
+150F [+65C]
3% for gage
lengths under 1/8
in [3.2 mm];
5% for 1/8 in
and over
1500
1500
10
5
10
6*
*Fatigue life improved
using low-modulus
solder.
N2A Open-faced constantan foil gages with a thin, laminated,
polyimide-film backing. Primarily recommended
for use in precision transducers, the N2A Series is
characterized by low and repeatable creep performance.
Also recommended for stress analysis applications
employing large gage patterns, where the especially flat
matrix eases gage installation.
Normal Static Transducer Service:
100 to +200F [75 to +95C]
3% 1700
1500
10
6
10
7
WA Fully encapsulated constantan gages with high-
endurance leadwires. Useful over wider temperature
ranges and in more extreme environments than EA
Series. Option W available on some patterns, but
restricts fatigue life to some extent.
Normal:
100 to +400F [75 to +205C]
320 to +500F [195 to +260C]
2%
2000
1800
1500
10
5
10
6
10
7
SA Fully encapsulated constantan gages with solder dots.
Same matrix as WA Series. Same uses as WA Series
but derated somewhat in maximum temperature and
operating environment because of solder dots.
Normal:
100 to +400F [75 to +205C]
320 to +450F [195 to +230C]
2%
1800
1500
10
6
10
7
EP
Specially annealed constantan foil with tough, high-
elongation polyimide backing. Used primarily for
measurements of large post-yield strains. Available
with Options E, L, and LE (may restrict elongation
capability).
100 to +400F [75 to +205C]
10% for gage
lengths under
1/8 in [3.2 mm];
20% for 1/8 in
and over
1000 10
4
EP gages show zero
shift under high-cyclic
strains.
ED Isoelastic foil in combination with tough, flexible
polyimide backing. High gage factor and extended
fatigue life excellent for dynamic measurements. Not
normally used in static measurements due to very high
thermal-output characteristics.
Dynamic:
320 to +400F [195 to +205C]
2%
Nonlinear at
strain levels
over 0.5%
2500
2200
10
6
10
7
WD
Fully encapsulated isoelastic gages with high-endur-
ance leadwires. Used in wide-range dynamic strain
measurement applications in severe environments.
Dynamic:
320 to +500F [195 to +260C]
1.5%
Nonlinear at
strain levels
over 0.5%
3000
2500
2200
10
5
10
7
10
8
SD Equivalent to WD Series, but with solder dots instead
of leadwires.
Dynamic:
320 to +400F [195 to +205C]
1.5%
See above note
2500
2200
10
6
10
7
EK K-alloy foil in combination with a tough, flexible polyimide
backing. Primarily used where a combination of higher
grid resistances, stability at elevated temperature, and
greatest backing flexibility are required.
Normal:
320 to +350F [195 to +175C]
452 to +400F [269 to +205C]
1.5% 1800 10
7
WK Fully encapsulated K-alloy gages with high-endurance
leadwires. Widest temperature range and most extreme
environmental capability of any general-purpose gage
when self-temperature-compensation is required.
Option W available on some patterns, but restricts both
fatigue life and maximum operating temperature.
452 to +550F [269 to +290C]
452 to +750F [269 to +400C]
1.5%
2200
2000
10
6
10
7
SK Fully encapsulated K-alloy gages with solder dots.
Same uses as WK Series, but derated in maximum
temperature and operating environment because of
solder dots.
Normal:
452 to +450F [269 to +230C]
452 to +500F [269 to +260C]
1.5%
2200
2000
10
6
10
7
S2K K-alloy foil laminated to 0.001 in [0.025 mm] thick,
high-performance polyimide backing, with a laminated
polyimide overlay fully encapsulating the grid and
solder tabs. Provided with large solder pads for ease of
leadwire attachment.
Normal:
100 to +250F [75 to +120C]
300 to +300F [185 to +150C]
1.5%
1800
1500
10
6
10
7
The performance data given here are nominal, and apply primarily to gages of 0.125-in [3-mm] gage length or larger.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
53
2.3 Gage Series
As noted in Sections 2.1 and 2.2, the strain-sensing alloy and
backing material are not subject to completely independent
selection and arbitrary combination. Instead, a selection
must be made from among the available gage systems, or
series, where each series generally incorporates special design
or construction features, as well as a specic combination of
alloy and backing material. For convenience in identifying the
appropriate gage series to meet specied test requirements,
the information on gage series performance and selection is
presented in the two tables above, in condensed form.
The frst table gives brief descriptions of all general-purpose
Micro-Measurements gage series including in each
case the alloy and backing combination and the principal
construction features. This table denes the performance of
each series in terms of operating temperature range, strain
range, and cyclic endurance as a function of strain level.
It must be noted, however, that the performance data are
nominal, and apply primarily to gages of 0.125 in [3 mm] or
longer gage length.
The second table gives the recommended gage series
for specific test profiles, or sets of test requirements,
categorized by the following criteria:
50 to +150F [45 to +65C]
50 to +400F [45 to +205C]
452 to +450F [269 to +230C]
<600F [<315C]
<700F [<370C]
50 to +150F [45 to +65C]
0 to +500F [20 to +260C]
452 to +500F [269 to +260C]
100 to +150F [75 to +65C]
320 to +500F [195 to +260C]
50 to +150F [45 to +65C]
50 to +200F [45 to +95C]
50 to +300F [45 to +150C]
320 to +350F [195 to +175C]

CYCLIC

TYPICAL SELECTION

TYPE OF TEST

OPERATING

TEST

ACCURACY
ENDURANCE REQD

OR

TEMPERATURE

DURATION

REQUIRED**

Maximum Number

Gage Series

M-Bond

APPLICATION

RANGE

IN HOURS

Strain, of Cycles

Adhesive
GENERAL
STATIC OR
STATIC-
DYNAMIC
STRESS
ANALYSIS*
HIGH-
ELONGATION
(POST-
YIELD)
DYNAMIC
(CYCLIC)
STRESS
ANALYSIS
TRANSDUCER
GAGING
* This category includes most testing situations where some degree of stability under static test conditions is required. For absolute stability with
constantan gages over long periods of usage and temperatures above +150F [+65C], it may be necessary to employ half- or full-bridge congu-
rations. Protective coatings may also inuence stability in cases other than transducer applications where the element is hermetically sealed.
** It is inappropriate to quantify accuracy as used in this table without consideration of various aspects of the actual test program and the instru-
mentation used. In general, moderate for stress analysis purposes is in the 2 to 5% range, high in the 1 to 3% range, and very high 1% or
better.
<10
6
<10
6
>10
6
>10
6
<10
6
<10
6
>10
6
<10
6
<10
6
1
1
1
1
1
10
7
10
7
10
7
<10
5
<10
6
<10
6
10
6
10
6
10
6
1300
1300
1600
2000
1600
2000
2000
1800
1500
50 000
100 000
200 000
15 000
10 000
2000
2400
2000
2300
1300
1300
1500
1600
1800
<10
4
>10
4
>10
4
>10
4
<10
3
>10
3
>10
3
<10
2
<10
<10
>10
3
>10
3
<10
2
<10
3
<10
4
<10
4
<10
4
<10
4
<10
4
<10
6
<10
4
<10
4
<10
4
CEA, EA
CEA, EA
WA, SA
WK, SK
WA, SA
WK, SK
WK, SK
WK
WK
CEA, EA
EP
EP
SA, SK, WA, WK
SK, WK
ED
WD
WD
WD
CEA, EA
CEA
N2A
WA, SA
WK, SK
200 or AE-10
AE-10 or AE-15
AE-15 or 610
AE-15 or 610
600 or 610
600 or 610
610
610
610
AE-10
AE-15
A-12
610
600 or 610
200 or AE-10
AE-10 or AE-15
600 or 610
600 or 610
AE-10 or AE-15
AE-15
600, 610, or 43B
610
610
Moderate
Moderate
Very High
High
Moderate
High
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
1 to 5%
1 to 5%
Better than
0.2%
0.2 to 0.5%
Better than
0.5%
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
54
type of strain measurement (static, dynamic, etc.)
operating temperature of gage installation
test duration
accuracy required
cyclic endurance required
This table provides the basic means for preliminary selection
of the gage series for most conventional applications. It
also includes recommendations for adhesives, since the
adhesive in a strain gage installation becomes part of the
gage system, and correspondingly affects the performance
of the gage. This selection table, supplemented by the
information in the first table, is used in conjunction
with our Precision Strain Gages Data Book, to arrive
at the complete gage selection. The procedure for
accomplishing this is described in Section 3.0 of this
Tech Note.
When a test prole is encountered that is beyond the ranges
specied in the above table, it can usually be assumed that
the test requirements approach or exceed the performance
limitations of available gages. Under these conditions, the
interactions between gage performance characteristics
become too complex for presentation in a simple table.
In such cases, the user should consult our Applications
Engineering Department for assistance in arriving at the
best compromise.
As indicated in the previous table, the CEA Series is
usually the preferred choice for routine strain-measurement
situations, not requiring extremes in performance or
environmental capabilities (and not requiring the very
smallest in gage lengths, or specialized grid confgurations).
CEA-Series strain gages are polyimide-encapsulated
A-alloy gages, featuring large, rugged, copper-coated tabs
for ease in soldering leadwires directly to the gage (below).
These thin, exible gages can be contoured to almost any
radius. In overall handling characteristics, for example,
convenience, resistance to damage in handling, etc., CEA-
Series gages are outstanding.
2.4 Gage Length
The gage length of a strain gage is the active or strain-
sensitive length of the grid, as shown below. The end
loops and solder tabs are considered insensitive to strain
because of their relatively large cross-sectional area and low
electrical resistance. To satisfy the widely varying needs of
experimental stress analysis and transducer applications,
Micro-Measurements offers gage lengths ranging from
0.008 in [0.2 mm] to 4 in [100 mm].
Gage length is often a very important factor in determining
the gage performance under a given set of circumstances.
For example, strain measurements are usually made at
the most critical points on a machine part or structure
that is, at the most highly stressed points. And, very
commonly, the highly stressed points are associated with
stress concentrations, where the strain gradient is quite steep
and the area of maximum strain is restricted to a very small
region. The strain gage tends to integrate, or average, the
strain over the area covered by the grid. Since the average
of any nonuniform strain distribution is always less than the
maximum, a strain gage which is noticeably larger than the
maximum strain region will indicate a strain magnitude that
is too low. The sketch above illustrates a representative strain
GAGE LENGTH
BACKING

ENCAPSULATION
COPPER-COATED TABS
PEAK
STRAIN
INDICATED
STRAIN
S
T
R
A
I
N
X
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
55
distribution in the vicinity of a stress concentration, and
demonstrates the error in strain indicated by a gage which is
too long with respect to the zone of peak strain.
As a rule of thumb, when practicable, the gage length should
be no greater than 0.1 times the radius of a hole, llet, or
notch, or the corresponding dimension of any other stress
raiser at which the strain measurement is to be made. With
stress-raiser confgurations having the signifcant dimension
less than, say, 0.5 in [13 mm], this rule of thumb can lead to
very small gage lengths. Because the use of a small strain
gage may introduce a number of other problems, it is often
necessary to compromise.
Strain gages of less than about 0.125 in [3 mm] gage length
tend to exhibit degraded performance particularly in
terms of the maximum allowable elongation, the stability
under static strain, and endurance when subjected to
alternating cyclic strain. When any of these considerations
outweigh the inaccuracy due to strain averaging, a larger
gage may be required.
When they can be employed, larger gages offer several
advantages worth noting. They are usually easier to handle
(in gage lengths up to, say, 0.5 in or 13 mm) in nearly
every aspect of the installation and wiring procedure than
miniature gages. Furthermore, large gages provide improved
heat dissipation because they introduce, for the same
nominal gage resistance, lower wattage per unit of grid area.
This consideration can be very important when the gage
is installed on a plastic or other substrate with poor heat
transfer properties. Inadequate heat dissipation causes high
temperatures in the grid, backing, adhesive, and test specimen
surface, and may noticeably affect gage performance and
accuracy (see Tech Note TN-502, Optimizing Strain Gage
Excitation Levels).
Still another application of large strain gages in this
case, often very large gages is in strain measurement
on nonhomogeneous materials. Consider concrete, for
example, which is a mixture of aggregate (usually stone) and
cement. When measuring strains in a concrete structure it
is ordinarily desirable to use a strain gage of sufcient gage
length to span several pieces of aggregate in order to measure
the representative strain in the structure. In other words, it
is usually the average strain that is sought in such instances,
not the severe local uctuations in strain occurring at the
interfaces between the aggregate particles and the cement.
In general, when measuring strains on structures made of
composite materials of any kind, the gage length should
normally be large with respect to the dimensions of the
inhomogeneities in the material.
As a generally applicable guide, when the foregoing
considerations do not dictate otherwise, gage lengths in the
range from 0.125 to 0.25 in [3 to 6 mm] are preferable. The
largest selection of gage patterns and stock gages is available
in this range of lengths. Furthermore, larger or smaller sizes
generally cost more, and larger gages do not noticeably
improve fatigue life, stability, or elongation, while shorter
gages are usually inferior in these characteristics.
2.5 Gage Pattern
The gage pattern refers cumulatively to the shape of the grid,
the number and orientation of the grids in a multiple-grid
gage, the solder tab conguration, and various construction
features which are standard for a particular pattern.
All details of the grid and solder tab congurations are
illustrated in the Gage Pattern columns of our strain gage
data book. The wide variety of patterns in the list is designed
to satisfy the full range of normal gage installation and
strain measurement requirements.
With single-grid gages, pattern suitability for a particular
application depends primarily on the following:
Solder tabs These should, of course, be compatible
in size and orientation with the space available at the
gage installation site. It is also important that the
tab arrangement be such as to not excessively tax the
prociency of the installer in making proper leadwire
connections.
Gri d wi dth When severe st rai n gradi ent s
perpendicular to the gage axis exist in the test specimen
surface, a narrow grid will minimize the averaging
error. Wider grids, when available and suitable to the
installation site, will improve the heat dissipation and
enhance gage stability particularly when the gage
is to be installed on a material or specimen with poor
heat transfer properties.
Gage resistance In certain instances, the only
difference between two gage patterns available in
the same series is the grid resistance typically 120
ohms vs. 350 ohms. When the choice exists, the higher-
resistance gage is preferable in that it reduces the heat
generation rate by a factor of three (for the same applied
voltage across the gage). Higher gage resistance also
has the advantage of decreasing leadwire effects such
as circuit desensitization due to leadwire resistance,
and unwanted signal variations caused by leadwire
resistance changes with temperature f luctuations.
Similarly, when the gage circuit includes switches, slip
rings, or other sources of random resistance change, the
signal-to-noise ratio is improved with higher resistance
gages operating at the same power level.
In experimental stress analysis, a single-grid gage would
normally be used only when the stress state at the point of
measurement is known to be uniaxial and the directions
of the principal axes are known with reasonable accuracy
(5).
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
56
These requirements severely limit the meaningful applicability
of single-grid strain gages in stress analysis; and failure to
consider biaxiality of the stress state can lead to large errors
in the stress magnitude inferred from measurements made
with a single-grid gage.
For a biaxial stress state a common
case necessitating strain measurement
a two- or three-element rosette is
required in order to determine the
principal stresses. When the directions
of the principal axes are known in
advance, a two-element 90-degree (or
tee) rosette can be employed with the
gage axes aligned to coincide with the principal axes. The
directions of the principal axes can sometimes be determined
with sufcient accuracy from one of several considerations.
For example, the shape of the test object and the mode of
loading may be such that the directions of the principal
axes are obvious from the symmetry of the situation, as in
a cylindrical pressure vessel. The principal axes can also be
defned by PhotoStress
testing.
In the most general case of surface
stresses, when the directions of the
principal axes are not known from
other considerations, a three-element
rosette must be used to obtain the
principal stress magnitudes. The
rosette can be installed with any
orientation, but is usually mounted
so that one of the grids is aligned with some signicant
axis of the test object. Three-
element rosettes are available in both
45-degree rectangular and 60-degree
delta configurations. The usual
choice is the rectangular rosette since
the data-reduction task is somewhat
simpler for this conguration.
When a rosette is to be employed, careful consideration should
always be given to the difference in characteristics between
single-plane and stacked rosettes. For any given gage length,
the single-plane rosette is superior to the stacked rosette
in terms of heat transfer to the test
specimen, generally providing better
stability and accuracy for static strain
measurements. Furthermore, when
there is a signicant strain gradient
perpendicular to the test surface (as
in bending), the single-plane rosette
will produce more accurate strain
data because all grids are as close as possible to the test
surface. Still another consideration is that stacked rosettes
are generally less conformable to contoured surfaces than
single-plane rosettes.
On the other hand, when there are large strain gradients in
the plane of the test surface, as is often the case, the single-
plane rosette can produce errors in strain indication because
the grids sample the strain at different points. For these
applications the stacked rosette is ordinarily preferable.
The stacked rosette is also advantageous when the space for
mounting the rosette is limited.
90-degree rosette
45-degree rosette

Stacked rosette
Micro-Measurements offers a selection of optional features
for its strain gages and special sensors. The addition of
options to the basic gage construction usually increases the
cost, but this is generally offset by the benets. Examples
are:
Signifcant reduction of installation time and costs
Reduction of the skill level necessary to make depend-
able installations
Increased reliability of applications
Simplifed installation of sensors in diffcult locations on
components or in the eld
Increased protection, both i n handl i ng duri ng
installation and shielding from the test environment
Achievement of special performance characteristics
Availability of each option varies with gage series and
pattern. Standard options are noted for each sensor in our
strain gage data book.
Shown below is a summary of the optional features offered.
Option Brief Description
W Integral Terminals and Encapsulation
E Encapsulation with Exposed Tabs
SE Solder Dots and Encapsulation
L Preattached Leads
LE Preattached Leads and Encapsulation
STANDARD CATALOG OPTIONS
2.6 Optional Features
60-degree rosette
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
57
Option W Series Availability: EA, EP, WA, ED, WD, EK, WK
General Description: This option provides encapsulation, and thin, printed circuit terminals at the tab end
of the gage. Beryllium copper jumpers connect the terminals to the gage tabs. The terminals are 0.0014 in
[0.036 mm] thick copper on polyimide backing nominally 0.0015 in [0.038 mm] thick. Option W gages are
rugged and well protected, and permit the direct attachment of larger leadwires than would be possible with
open-faced gages. This option is primarily used on EA-Series gages for general-purpose applications. Solder:
+430F [+220C] tin-silver alloy solder joints on E-backed gages, +570F [+300C] lead-tin-silver alloy solder
joints on W-backed gages. Temperature Limit: +400F [+200C] for E-backed gages, +500F [+260C] for
W-backed gages. Grid Protection: Entire grid and part of terminals are encapsulated with polyimide. Fatigue
Life: Some loss in fatigue life unless strain levels at the terminal location are below 1000. Size: Option W
extends from the soldering tab end of the gages and thereby increases gage size. With some patterns width is slightly greater.
Strain Range: With some gage series, notably E-backed gages, strain range will be reduced. This effect is greatest with EP
gages, and Option W should be avoided with them if possible. Flexibility: Option W adds encapsulation, making gages
slightly thicker and stiffer. Conformance to curved surfaces will be somewhat reduced. In the terminal area itself, stiffness
is markedly increased. Resistance Tolerance: On E-backed gages, resistance tolerance is normally doubled.
Option E Series Availability: EA, ED, EK, EP
General Description: Option E consists of a protective encapsulation of polyimide lm approximately 0.001 in
[0.025 mm] thick. This provides ruggedness and excellent grid protection, with little sacrifce in fexibility. Sol-
dering is greatly simplied since the solder is prevented from tinning any more of the gage tab than is deliber-
ately exposed for lead attachment. Option E protects the grid from ngerprints and other contaminating agents
during installation and, therefore, contributes signifcantly to long-term gage stability. Heavier leads may be
attached directly to the gage tabs for simple static load tests. Supplementary protective coatings should still be
applied after lead attachment in most cases. Temperature Limit: No degradation. Grid Protection: Entire grid
and part of tabs are encapsulated. Fatigue Life: When gages are properly wired with small jumpers, maximum
endurance is easily obtained. Size: Gage size is not affected. Strain Range: Strain range of gages will be reduced because
the additional reinforcement of the polyimide encapsulation can cause bond failure before the gage reaches its full strain
capability. Flexibility: Option E gages are almost as conformable on curved surfaces as open-faced gages, since no internal
leads or solder are present at the time of installation. Resistance Tolerance: Resistance tolerance is normally doubled when
Option E is selected.
Option SE Series Availability: EA, ED, EK, EP
General Description: Option SE is the combination of solder dots on the gage tabs with a 0.001-in [0.025-mm]
polyimide encapsulation layer that covers the entire gage. The encapsulation is removed over the solder dots
providing access for lead attachment. These gages are very exible, and well protected from handling damage
during installation. Option SE is primarily intended for small gages that must be installed in restricted areas,
since leadwires can be routed to the exposed solder dots from any direction. The option does not increase
overall gage dimensions, so the matrix may be feld-trimmed very close to the actual pattern size. Option SE
is sometimes useful on miniature transducers of medium- or low-accuracy class, or in stress analysis work on
miniature parts. Solder: +570F [+300C] tin-silver alloy. To prevent loss of long-term stability, gages with
Option SE must be soldered with noncorrosive (rosin) fux, and all fux residue should be carefully removed with Rosin Sol-
vent after wiring. Protective coatings should then be used. Temperature Limit: No degradation. Grid Protection: Entire gage
is encapsulated. Fatigue Life: When gages are properly wired with small jumpers, maximum endurance is easily obtained.
Size: Gage size is not affected. Strain Range: Strain range of gages will be reduced because the additional reinforcement
of the polyimide encapsulation can cause bond failure before the gage reaches its full strain capability. Flexibility: Option
SE gages are almost as conformable on curved surfaces as open-faced gages. Resistance Tolerance: Resistance tolerance is
normally doubled when Option SE is selected.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
58
Option LE Series Availability: EA, ED, EK, EP
General Description: This option provides the same conformable soft copper lead ribbons as used in Option L,
but with the addition of a 0.001-in [0.025-mm] thick encapsulation layer of polyimide flm. The encapsulation
layer provides excellent protection for the gage during handling and installation. It also contributes greatly to
environmental protection, though supplementary coatings are still recommended for feld use. Gages with Op-
tion LE will normally show better long-term stability than open-faced gages which are waterproofed only after
installation. A good part of the reason for this is that the encapsulation layer prevents contamination of the grid
surface from ngerprints or other agents during handling and installation. The presence of such contaminants
will cause some loss in gage stability, even though the gage is subsequently coated with protective compounds.
Leads: Nominal ribbon size for most gages is 0.012 wide x 0.004 in thick [0.30 x 0.10 mm] copper ribbons. Leads
are approximately 0.8 in [20 mm] long. Solder: +430F [+220C] tin-silver alloy. Temperature Limit: +400F
[+200C]. Grid Protection: Entire gage is encapsulated. A short extension of the backing is left uncovered at the
leadwire end to prevent contact between the leadwires and the specimen surface. Fatigue Life: Fatigue life will
normally be degraded by Option LE. This occurs primarily because the copper ribbon has limited cyclic endurance. Option LE
is not often recommended for very high endurance gages such as the ED Series. Size: Matrix size is unchanged. Strain Range:
Strain range will usually be reduced by the addition of Option LE. Flexibility: Gages with Option LE are not as conformable
as standard gages. Resistance Tolerance: Resistance tolerance is normally doubled by the addition of Option LE.
Leadwire Orientation for Options L and LE
These illustrations show the standard orientation of leadwires relative to the gage pattern geometry for Options L and LE.
The general rule is that the leads are parallel to the longest dimension of the pattern. The illustrations also apply to leadwire
orientation for WA-, WK- and WD-Series gages, when the pattern shown is available in one of these series.
(3 or 4 tabs)
Option L Series Availability: EA, ED, EK, EP
General Description: Option L is the addition of soft copper lead ribbons to open-faced polyimide-backed
gages. The use of this type of ribbon results in a thinner and more conformable gage than would be the case
with round wires of equivalent cross section. At the same time, the ribbon is so designed that it forms almost
as readily in any desired direction. Leads: Nominal ribbon size for most gages is 0.012 wide x 0.004 in thick
[0.30 x 0.10 mm]. Leads are approximately 0.8 in [20 mm] long. Solder: +430F [+220C] tin-silver alloy.
Temperature Limit: +400F [+200C]. Fatigue Life: Fatigue life will normally be degraded by Option L. This
occurs primarily because the copper ribbon has limited cyclic endurance. When it is possible to carefully
dress the leads so that they are not bonded in a high strain eld, the performance limitation will not apply.
Option L is not often recommended for very high endurance gages such as the ED Series. Size: Matrix size is
unchanged. Strain Range: Strain range will usually be reduced by the addition of Option L. Flexibility: Gages
with Option L are not as conformable as standard gages. Resistance Tolerance: Not affected.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
59
As in other aspects of strain gage selection, the choice of
options ordinarily involves a variety of compromises. For
instance, an option which maximizes a particular gage
performance parameter such as fatigue life may at the same
time require greater skill in installing the gage. Because
of the many interactions between installation attributes
and performance parameters associated with the options,
the relative merits of all standard options are summarized
qualitatively in the chart on page 60 as an aid to option
selection. For comparison purposes, the corresponding
characteristics of the CEA Series are given in the right-most
column of the table.
Since, in strain measurement for stress analysis, the standard
options are most frequently applied to EA-Series strain
gages, the information supplied in this section is directed
primarily toward such option applications.
When contemplating the application of an EA-Series gage
with an option, the rst consideration should usually be
whether there is an equivalent CEA-Series gage that will
satisfy the test requirements. Comparing, for example, an
EA-Series gage equipped with Option W and a similar CEA-
Series pattern, it will be found that the latter is characterized
by lower cost, greater exibility and conformability, and
superior fatigue life. The only possible advantages for the
selection of Option W are the wider variety of available
patterns and the occasional need for large soldering
terminals.
It should also be noted that many standard strain gage types,
without options, are normally available from stock; while
gages with options are commonly manufactured to order,
and may thus involve a minimum order requirement.
Option P Series Availability: EA, N2A
General Description: Option P is the addition of preattached leadwire cables to many patterns of EA- and N2A-
Series strain gages. Encapsulation seals small jumper leadwires at gage end, and cable insulation protects solder
joints at cable end. Option P virtually eliminates need for soldering during gage installation. Leads: A pair of 1-in
[25-mm] M-LINE 134-AWP (solid copper, polyurethane enamel) single conductor jumper leadwires. Cable: 10
ft [3.1 m] of color-coded, fat, three-conductor 26-gauge [0.404 mm dia.], stranded, tinned copper with vinyl in-
sulation (similar to M-LINE 326-DFV). Solder: +430F [+220C] tin-silver alloy solder joints, jumper to gage.
Cable conductors and jumpers joined with +430F [+220C] solder beneath cable insulation. Exposed leadwires
on unattached end of cable are pretinned for ease of hookup. Temperature Limit: 60 to +180F [50 to +80C];
limited by vinyl insulation on cable. Grid Encapsulation: Entire grid and tabs are encapsulated with Option E.
Fatigue Life: Fatigue life will normally be degraded by Option P, primarily because the copper jumper wires
have limited cyclic endurance. Pattern Availability: Most EA- and N2A-Series single-grid patterns that are 0.062
in [1.5 mm] or greater gage length, with parallel solder tabs on one end of the grid, and suitable for encapsulation.
(Consult our Applications Engineering Department for availability of Option P on other gage series/patterns, and
for nonstandard cable lengths.) Size: Matrix size is unchanged. Strain Range: Strain range will usually be reduced
by the addition of Option P. Flexibility: E-backed gages with Option P are not as conformable as standard gages.
Resistance Considerations: Each conductor of the cable has a nominal resistance of 0.04 ohm/ft [0.13 ohm/m]. Gage resistance
is measured at gage tabs. Gage Factor: Gage factor is determined for gages without preattached cable.
Option P2 Series Availability: CEA
General Description: Option P2 is the addition of preattached leadwire cables to CEA-Series strain gages. Op-
tion P2 virtually eliminates need for soldering during gage installation. Cable: 10 ft [3.1 m] of color-coded,
fat, three-conductor 30-gauge [0.255 mm], stranded, tinned copper with vinyl insulation (similar to M-LINE
330-DFV). Solder: +361F [+180C] tin-lead alloy solder joints. Exposed leadwires on unattached end of
cable are pretinned for ease of hookup. Temperature Limit: 60 to +180F [50 to +80C]; limited by vinyl
insulation on cable. Grid Encapsulation: Entire grid is encapsulated. (Solder tabs are not encapsulated.) Fatigue
Life: Fatigue life will normally be unchanged by Option P2. Pattern Availability: Most CEA-Series single- and
multiple-grid patterns. Size: Matrix size is unchanged. Strain Range: Standard for CEA-Series gages. Flexibil-
ity: No appreciable increase in stiffness. Resistance Considerations: Each conductor of the cable has a nominal
resistance of 0.11ohm/ft [0.36 ohm/m]. Gage resistance is measured at gage tabs. Gage Factor: Gage factor is
determined for gages without preattached cable.
2.7 Characteristics of Standard Catalog Options on EA-Series Gages
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
60
In the following table, the respective performance parameters
for an open-faced EA-Series gage without options are
arbitrarily assigned a value of 5. Numbers greater than 5
indicate a particular parameter is improved by addition of
the option, while smaller numbers indicate a reduction in
performance.

Installation Attribute

Standard Options

CEA

Or Performance Parameter

W E SE L LE

Series

Overall Ease of Gage Installations 8 7 6 5 6 10
Ease of Leadwire Attachment 10 8 7 7 8 10
Protection of Grid from Environmental Attack 8 8 8 5 8 8
Cyclic Strain Indurance 2 7 8 3 4 4
Elongation Capability 2 3 3 4 3 3
Resistance Tolerance 3 3 3 5 3 3
Reinforcement Effects 2 3 3 5 3 3
3.0 Gage Selection Procedure
The performance of a strain gage in any given application is
affected by every element in the design and manufacture of
the gage. Micro-Measurements offers a great variety of gage
types for meeting the widest range of strain measurement
needs. Despite the large number of variables involved, the
process of gage selection can be reduced to only a few basic
steps. From the following diagram that explains the gage
designation code, it is evident that there are but ve parameters
to select, not counting options. These are: the gage series, the
S-T-C number, the gage length and pattern, and the
resistance.
Of the preceding parameters, the gage length and pattern are
normally the rst and second selections to be made, based
on the space available for gage mounting and the nature
of the stress eld in terms of biaxiality and expected strain
gradient. A good starting point for initial consideration of
gage length is 0.125 in [3 mm]. This size offers the widest
variety of choices from which to select remaining gage
parameters such as pattern, series and resistance. The
gage and its solder tabs are large enough for relatively easy
handling and installation. At the same time, gages of this
length provide performance capabilities comparable to those
of larger gages.
The principal reason for selecting a longer gage would
commonly be one of the following: (a) greater grid area
for better heat dissipation; (b) improved strain averaging
on inhomogeneous materials such as fiber-reinforced
composites; or (c) slightly easier handling and installation
(for gage lengths up to 0.50 in [13 mm]). On the other hand,
a shorter gage length may be necessary when the object is
to measure localized peak strains in the vicinity of a stress
concentration, such as a hole or shoulder. The same is true,
of course, when the space available for gage mounting is very
limited.
In selecting the gage pattern, the first consideration is
whether a single-grid gage or rosette is required (see Section
2.5). Single-grid gages are available with different aspect
(length-to-width) ratios and various solder tab arrangements
for adaptability to differing installation requirements. Two-
element 90-degree rosettes, when applicable, can also be
selected from a number of different grid and solder tab
confgurations. With three-element rosettes (rectangular or
delta), the primary choice in pattern selection, once the gage
length has been determined, is between planar and stacked
construction, as described in Section 2.5.
The format of our strain gage data book is designed to
simplify selection of the gage length and pattern. Similar
patterns available in each gage length are grouped together,
and listed in order of size. The strain gages in the main
EA-06-250BF-350 OPTION LE
GAGE LENGTH OPTION (IF ANY)
GAGE PATTERN RESISTANCE
GAGE SERIES S-T-C NUMBER
Strain Gage Selection Steps
Step 4
Step 5
Step 6
Step 1
Step 2
Step 3
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
61
listing (large pictures) are the most widely used for stress
analysis applications. This section should always be reviewed
rst to locate an appropriate gage.
With an initial selection of the gage size and pattern
completed, the next step is to select the gage series, thus
determining the foil and backing combination, and any
other features common to the series. This is accomplished by
referring to the earlier chart, which gives the recommended
gage series for specific test profiles, or sets of test
requirements. If the gage series is to have a standard option
applied, the option should be tentatively specied at this
time, since the availability of the desired option on the
selected gage pattern in that series requires verification
during the procedure outlined in the following paragraph.
After selecting the gage series (and option, if any), reference
is made again to our Precision Strain Gages Data Book to
record the gage designation of the desired gage size and
pattern in the recommended series. If this combination is not
listed as available in the catalog, a similar gage pattern in the
same size group, or a slightly different size in an equivalent
pattern, can usually be selected for meeting the installation
and test requirements. In extreme cases, it may be necessary
to select an alternate series and repeat this process. Quite
frequently, and especially for routine strain measurement,
more than one gage size and pattern combination will be
suitable for the specied test conditions. In these cases, it is
wise to select a gage from the main listings (large pictures)
to eliminate the likelihood of extended delivery time or a
minimum order requirement.
As noted under the gage pattern discussion on page 55, there
are often advantages from selecting the 350-ohm resistance
if this resistance is compatible with the instrumentation
to be used. This decision may be inuenced, however, by
cost considerations, particularly in the case of very small
gages. Some reduction in fatigue life can also be expected
for the high-resistance small gages. Finally, in recording
the complete gage designation, the S-T-C number should
be inserted from the list of available numbers for each alloy
given in this Tech Note.
This completes the gage selection procedure. In each step of
the procedure, the Strain Gage Selection Checklist provided
in Section 4.0 should be referred to as an aid in accounting
for the test conditions and requirements which could affect
the selection.
4.0 Strain Gage Selection Checklist
This checklist is provided as a convenient, rapid means for
helping make certain that no critical requirement of the test
profile which could affect gage selection is overlooked.
It should be borne in mind in using the checklist that
the considerations listed apply to relatively routine and
conventional stress analysis situations, and do not embrace
exotic applications involving nuclear radiation, intense
magnetic elds, extreme centrifugal forces, and the like.
CONSIDERATIONS FOR
PARAMETER SELECTION
Selection Step: 1 strain gradients
Parameter: Gage Length area of maximum strain
accuracy required
static strain stability
maximum elongation
cyclic endurance
heat dissipation
space for installation
ease of installation
Selection Step: 2 strain gradients (in-plane
Parameter: Gage Pattern and normal to surface)
biaxiality of stress
heat dissipation
space for installation
gage resistance availability
Selection Step: 3 type of strain measurement
Parameter: Gage Series application (static, dynamic,
post-yield, etc.)
operating temperature
test duration
cyclic endurance
accuracy required
Selection Step: 4 type of measurement (static,
Parameter: Options dynamic, post-yield, etc.)
installation environment
laboratory or eld
stability requirements
soldering sensitivity of
substrate (plastic, bone, etc.)
space available for installation
installation time constraints
Selection Step: 5 heat dissipation
Parameter: Gage Resistance leadwire desensitization
signal-to-noise ratio
Selection Step: 6 test specimen material
Parameter: S-T-C Number operating temperature range
accuracy required
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
62
5.0 Gage Selection Examples
In this section, three examples are given of the gage-
selection procedure in representative stress analysis
situations. An attempt has been made to provide the
principal reasons for the particular choices which are
made. It should be noted, however, that an experienced
stress analyst does not ordinarily proceed in the same
step-by-step fashion illustrated in these examples. Instead,
simultaneously keeping in mind the test conditions and
environment, the gage installation constraints, and the test
requirements, the analyst reviews our strain gage data book
and quickly segregates the more likely candidates from
among the available gage-pattern and series combinations in
the appropriate sizes. The selection criteria are then rened
in accordance with the particular strain-measurement task
to converge on the gage or gages to be specied for the test
program. Whether formally or otherwise, the knowledgeable
practitioner does so in the light of parameter selection
considerations such as those itemized in the preceding
checklist.
A. Design Study of a Pressure Vessel
Strain measurements are to be made on a scaled-down
plastic model of a pressure vessel. The model will be tested
statically at, or near, room temperature; and, although
the tests may be conducted over a period of several
months, individual tests will take only a few hours to run.

Gage Selection:
1. Gage Length Very short gage lengths should be
avoided in order to minimize heat dissipation problems
caused by the low thermal conductivity of the plastic.
The model is quite large, and apparently free of severe
strain gradients; therefore, a 0.25-in [6.3-mm] gage
length is specied, because the widest selection of gage
patterns is available in this length.
2. Gage Pattern In some areas of the model, the
directions of the principal axes are obvious from
considerations of symmetry, and single-grid gages can
be employed. Of the patterns available in the selected
gage length, the 250BF pattern is a good compromise
because of its high grid resistance which will help
minimize heat dissipation problems.
In other areas of the model, the directions of the
principal axes are not known, and a three-element
rosette will be required. For this purpose, a planar
rosette should be selected, since a stacked rosette would
contribute significantly to reinforcement and heat
dissipation problems. Because of its high-resistance
grid, the 250RD pattern is a good choice.
3. Gage Series The polyimide (E) backing is preferred
because its low elastic modulus wi l l mi ni mi ze
reinforcement of the plastic model. Because the normal
choice of grid alloy for static strain measurement at
room temperature is the A alloy, the EA Series should
be selected for this application.
4. Options Excessive heat application to the test model
during leadwire attachment could damage the material.
Option L (preattached leads) is therefore selected so
that the instrument cable can be attached directly to
the leads without the application of a soldering iron to
the gage proper. Option L is preferable over Options LE
and P because the encapsulation in the latter options
would add reinforcement.
5. Resistance In this case, the resistance was determined
in Step 2 when the higher resistance alternative was
selected from among the gage patterns; i.e., in selecting
the 250BF over the 250BG, and the 250RD over the
250RA. The selected gage resistance is thus 350 ohms.
6. S-T-C Number Ideally, the gages should be self-
temperature-compensated to match the model mate-
rial, but this is not always feasible, since plastics
particularly reinforced plastics vary widely
in thermal expansion coefficient. For unreinforced
plastic, S-T-C 30, 40 or 50 should usually be selected.
If a mismatch between the model material and the
S-T-C number is necessary, S-T-C 13 should be selected
(because of stock status), and the test performed at
constant temperature.
Gage Designations:
From the above steps, the strain gages to be used are:
EA-30-250BF-350/Option L (single-grid)
EA-30-250RD-350/Option L (rosette)
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
63
B. Dynamic Stress Analysis Study
of a Spur Gear in an Hydraulic Pump
Strain measurements are to be made at the root of the gear
tooth while the pump is operating. The llet radius at the
tooth root is 0.125 in (or about 3 mm) and test temperatures
are expected to range from 0 to +180F [20 to +80C].
Gage Selection:
1. Gage Length A gage length which is small with
respect to the llet radius should be specied for this
application. A length of 0.015 in [0.38 mm] is preferable,
but reference to our strain gage data book indicates
that such a choice severely limits the available gage
patterns and grid alloys. Anticipating problems which
would otherwise be encountered in Steps 2 and 3, a
gage length of 0.031 in [0.8 mm] is selected.
2. Gage Pattern Because the gear is a spur gear, the
directions of the principal axes are known, and single-
grid gages can be employed. A gage pattern with both
solder tabs at the same end should be selected so that
leadwire connections can be located in the clearance
area along the root circle between adjacent teeth. In
the light of these considerations, the 031CF pattern is
chosen for the task.
3. Gage Series Low strain levels are expected in this
application; and, furthermore, the strain signals
must be transmitted through slip rings or through a
telemetry system to get from the rotating component
to the stationary instrumentation. Isoelastic (D alloy)
is preferred for its higher gage factor (nominally 3.2,
in contrast to 2.1 for A and K alloys). Because the
gage must be very flexible to conform to the small
llet radius, the E backing is the most suitable choice.
The maximum test temperature is not a consideration
in this case, since it is well within the recommended
temperature range for any of the standard backings.
The combination of the E backing and the D alloy
denes the ED gage series.
4. Options For protection of the gage grid in the test
environment, Option E, encapsulation, should be
specied. Because of the limited clearance between the
outside diameter of one gear and the root circle of the
mating gear, a particularly thin gage installation must
be made; and very small leadwires will be attached to
the gage tabs at 90 to the grid direction, and run over
the sides of the gear for connection to larger wires.
This requirement necessitates attachment of the small
leadwires after gage bonding, and prevents the use of
preattached leads.
5. Resistance In the ED-Series version of the 031CF
gage pattern, our strain gage data book lists the
resistance as 350 ohms. The higher resistance should
usually be selected whenever the choice exists, and
will be advantageous in this instance in improving the
signal-to-noise ratio when slip rings are used.
6. S-T-C Number D alloy is not subject to self-
temperature-compensation, nor is compensation
needed for these tests since only dynamic strain is to
be measured. In the ED-Series designation the two-
digit S-T-C number is replaced by the letters DY for
dynamic.
Gage Designation:
Combining the results of the above selection procedure, the
gage to be employed is:
ED-DY-031CF-350/Option E
C. Flight-Test Stress Analysis
of a Titanium Aircraft Wing Tip Section
With, and Without, a Missile Module Attached
The operating temperature range for strain measurements
is from 65 to +450F [55 to +230C], and will be a
dominant factor in the gage selection.
Gage Selection:
1. Gage Length Preliminary design studies using the
PhotoStress photoelastic coating technique indicate
that a gage length of 0.062 in [1.6 mm] represents the
best compromise in view of the strain gradients, areas
of peak strain, and space for gage installation.
T
E
C
H

N
O
T
E
TN-505-4
Micro-Measurements
64
2. Gage Pattern With information about the stress
state and directions of principal axes gained from the
photoelastic coating studies, there are some areas of
the wing tip where single-grid gages and two-element
tee rosettes can be employed. In other locations,
where principal strain directions vary with the nature
of the fight maneuver, 45-degree rectangular rosettes
are required.
The strain gradients are sufciently steep that stacked
rosettes should be selected. From our strain gage data
book, the foregoing requirements suggest the selection
of 060WT and 060WR gage patterns for the stacked
rosettes, and the 062AP pattern for the single-grid
gage. In making this selection, attention was given to
the fact that all three patterns are available in the WK
Series, which is compatible with the specifed operating
temperature range.
3. Gage Series The maximum operating temperature,
along with the requirement for static as well as dynamic
strain measurement, clearly dictates use of K alloy for
the grid material. Either the SK or WK Series could be
selected, but the WK gages are preferred because they
have integral leadwires.
4. Options For ease of gage installation, Option W,
with integral soldering terminals, is advantageous.
This option is not applicable to stacked rosettes,
however, and is therefore specifed for only the single-
grid gages.
5. Resistance When available, as in this case, 350-
ohm gages should be specied because of the benets
associated with the higher gage resistance.
6. S-T-C Number The titanium alloy used in the wing
tip section is the 6Al-4V type, with a thermal expansion
coefficient of 4.910
6
per F [8.810
6
per C].
K alloy of S-T-C number 05 is the appropriate choice.
Gage Designations:
WK-05-062AP-350/Option W
WK-05-060WT-350
WK-05-060WR-350
Tech Note TN-506-3
MICRO-MEASUREMENTS
Bondable Resistance Temperature
Sensors and Associated Circuitry
T
E
C
H

N
O
T
E
65
1.0 Introduction
Micro-Measurements manufactures a line of resistance
temperature sensors that are constructed much like wide-
temperature-range strain gages, but which utilize high-purity
nickel-foil sensing grids. These temperature sensors are
bonded to structures using standard strain gage installation
techniques, and can measure surface temperatures from
320 to approximately +500F [195 to +260C].
This Tech Note discusses the operational characteristics of
nickel temperature sensors, as well as various methods of
data readout. The standard line of temperature sensors and
matching networks for use with strain indicators is listed in
the Precision Strain Gages Data Book.
The resistance of high-purity nickel increases rapidly
with temperature, following a repeatable and stable curve
to over +500F [+260C]. As shown in Figure 1, the
resistance changes are quite large, resulting in high signal
levels. Figure 1 also includes a curve for Balco
alloy,
discussed in Section 6. Note that a reference value of 50.0
ohms occurs at a temperature of +75.0F [+23.9C] in
Figure 1. All standard Micro-Measurements TG temperature
sensors are manufactured to this nominal value, but gages
of other resistance values are available on special order.
The resistance-versus-temperature characteristic of sensors
having nonstandard reference resistances can be expressed
directly as a percentage of the nominal +75.0F [+23.9C]
value by multiplying the ordinate in Figure 1 by a factor
of 2.0.
Tables are given later in this Tech Note for obtaining
resistance values of TG sensors at various temperatures to a
higher accuracy than is readable in Figure 1. These tables are
also referenced to a value of 50.0 ohms at +75.0F [+23.9C]
and can be multiplied by a factor of 2.0 to obtain resistances
of nonstandard sensors as a percentage of their +75.0F
[+23.9C] value.
2.0 Sensor Installation
TG temperature sensors are installed with the same techniques
and materials used for installation of wide-temperature-
range strain gages. M-Bond 600 or 610 adhesives are usually
employed because they are useful over the entire temperature
range of the sensor itself. Surface preparation techniques are
given in Instruction Bulletin B-129, and specic installation
procedures are included in the selected adhesive kit.
Leadwires are normally handled in the same way as those
for strain gages, with one signicant difference. The three-
wire system used with strain gages to eliminate errors due
to temperature-induced resistance changes in the leadwires
is not generally effective with TG sensors. This subject is
treated at greater length in Sections 3 and 5.
3.0 Readout Methods
One method of reading temperature with TG temperature
sensors is to connect the sensor to a Wheatstone resistance
bridge, and convert the resistance readings to equivalent
temperatures with the tables given in this Tech Note. But,
the leadwires can cause two different errors in a Wheatstone
bridge. First, the resistance of the leads, which can be
appreciable with remote gages, produces an initial offset
error, and desensitizes the arm of the bridge containing
-200 -100 0 +100 +200 +300 +400 +500
140
130
120
110
100
90
80
70
60
50
40
30
20
10
R
e
s
i
s
t
a
n
c
e

i
n

O
h
m
s
Temperature in F
-100 0 +100 +200
Temperature in C
5
0
-
o
h
m

p
u
r
e

n
i
c
k
e
l

-

L
o
t

T
0
4
5
0
-
o
h
m

B
a
l
c
o

-

L
o
t

B
0
3
5
0
-
o
h
m
n
ic
k
e
l s
e
n
s
o
r
lin
e
a
r
iz
e
d
b
y
1
8
7
.5
o
h
m

x
e
d
s
h
u
n
t
r
e
s
is
t
o
r
Figure 1. Variation of resistance with temperature
for 50-ohm sensors mounted on 1018 steel.
Balco is a Trademark of W.B. Driver Co.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
66
Bondable Resistance Temperature Sensors and Associated Circuitry
the temperature sensor. The second error is the result
of resistance change in the leads caused by temperature
variations. Except under unusual conditions, errors of
this type are very small. In cases where long leadwires are
necessary, special calibration techniques or compensation
systems can be used.
1,2*
A variation of the above method is capable of providing
accurate compensation for leadwire resistances with a three-
wire system. The circuit is shown in Figure 2, in which a
precision decade box is used in place of one resistor in the
Wheatstone bridge.
The decade box in this circuit is varied to keep the output
at null, and the indicated decade resistance is therefore
the same as the sensor resistance. Provided leads 1 and 3
are of the same length and size, resistance changes in the
leadwire circuit caused by temperature changes common
to all wires will not create errors in the reading. Three-wire
compensation is effective in this case because this is a true
null-balance system in which the bridge arm adjacent to the
sensor (the decade box) is always set to the sensor resistance
at the time of readout.
Excitation power can be either dc or ac, depending on the
null indicator chosen. Excessive excitation can create errors
due to self-heating in the sensor, but this error is easily
avoided or corrected as discussed in Section 5.5.
A more sophisticated readout system, which eliminates
the need for manual rebalance, is shown in Figure 3. This
arrangement eliminates leadwire errors by use of a four-wire
system.
If the digital voltmeter has a high enough input impedance,
the readings will be a known function of sensor resistance,
regardless of resistance change in any of the leadwires.
A current level of one mA will allow the voltmeter to
read sensor resistance in terms of mV (50.0 ohms reads as
50.0 mV).
3.1 Linearization of Sensor Response
The readout methods shown in Figures 2 and 3 can be quite
accurate, but are somewhat awkward to use in that they read
sensor resistance directly, and the nonlinear characteristic
of sensor resistance-versus-temperature requires the
use of tabulated data to convert resistance values to the
equivalent temperatures. A very simple method exists for
converting the nonlinear sensor to a linear resistance change
with temperature with good practical accuracy. This is
accomplished by shunting the sensor with a xed resistor
equal in value to 3.75 times the +75.0F [+23.9C] value of
sensor resistance, or 187.5 ohms for standard TG sensors.
The resultant resistance change with temperature has a
lower slope, but is quite linear as shown in Figure 1. Figure
4 is a plot of deviation from linearity for this circuit, and
provides much higher error readability than Figure 1.
A useful application of shunt linearization is to provide an
asymmetrical bridge circuit that has a linear output voltage
with temperature. This circuit can be used with a digital
voltmeter for direct temperature readings (expressed in mV)
or to drive one axis of an X-Y recorder for directly plotting
temperature against another variable. A simplied version
of this circuit is shown in Figure 5.
This linearization circuit requires a constant-voltage
excitation source, and this can be conveniently provided
*Superscript numbers refer to references appended to this Tech Note.
Figure 2. Three-wire null-balance circuit for TG sensor.
Figure 3. Four-wire circuit for TG sensor.
Figure 4. Deviation from linearity for shunted TG sensor of Figure 1.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
67
by a single silver-oxide battery. The output factor is 0.5
mV/V/F [0.9 mV/V/C], which can be scaled down by use
of a high-resistance voltage divider at the output terminals.
A balance potentiometer is incorporated for balancing
out the tolerance on the nominal sensor resistance and
the offset caused by leadwire resistance. (When the Celsius
temperature scale is used, it is more convenient to balance
at +24C, rather than +23.9C, so that the readings are in
round numbers.) Calibration can be checked by substituting
a precision decade box for the TG sensor and dialing in
resistances equivalent to various temperatures from the
resistance-versus-temperature tables.
Because of the asymmetrical bridge, a three-wire lead
system cannot be used effectively with the circuit in Figure 5
to eliminate the errors from a temperature-induced leadwire
resistance change. However, it is possible to use the three-
wire method for the more limited purpose of compensating
the initial offset error caused by the leadwire resistance. This
is accomplished by adding the third wire (shown dashed)
and breaking the connection at the point marked X.
4.0 LST Matching Networks
Commercial single- or multiple-channel strain indicators are
excellent readout devices for TG temperature sensors, and
are particularly convenient when combinations of strain and
temperature are to be measured or recorded simultaneously.
This readout method requires the use of an interface signal
conditioning network, referred to as an LST network, to
match the temperature sensor response to that of the
strain indicator. The arrangement is shown in Figure 6.
The LST network is a small, completely encapsulated unit
consisting of four special precision resistors. When used
with standard 50-ohm nickel TG temperature sensors, it
performs the following three functions:
1. Provides a linear resistance change at the output
terminals.
2. Attenuates the resistance change slope to the equivalent
of 10 or 100/deg for a gage factor setting of 2.00 on
the strain indicator. It is usually most practical to use
10/deg networks when a large temperature range is
involved, and 100/deg networks for high resolution
of small temperature spans.
3. Presents a complete 350-ohm half-bridge circuit to the
strain indicator.
Differential temperature measurements can be made by
combining two TG sensors and two LST networks. The
arrangement in Figure 7a provides a half-bridge circuit to
the strain indicator in which the ACTIVE arm responds to
sensor 1, and the DUMMY arm to sensor 2. An alternate
method is to arrange the networks to form a full-bridge
circuit as shown in Figure 7b, which may be preferred for
some types of readout equipment.
Calibration and balancing of circuits containing LST
networks can be handled in various ways. Initial balance
is usually obtained by setting the BALANCE of the strain
indicator channel so that the instrument reading corresponds
to the initial temperature of the TG sensor. When it is not
possible to control or measure the initial temperature of the
sensor, it can be temporarily replaced by a precision 50.0-ohm
resistor. The BALANCE is then set to create an equivalent
readout of +75.0F (Fahrenheit networks) or +23.9C
(Celsius networks). The sensor can then be reconnected to
the network. The rst procedure has the obvious advantage
of correcting for the error due to tolerance limits on initial
TG-sensor resistance.
Resistance shunt calibration
3
can be applied to the output
terminals of the LST network in order to verify linearity
and span accuracy of the associated instrumentation. It
is also useful in setting GAGE FACTOR or in correcting
for leadwire desensitization when the LST network is near
the sensor and remote from the readout instrument. In this
latter case, note that shunt resistors must be placed across
terminals 2 and 3 at the network, not at the instrument.
Because of signicant resistance changes in the ACTIVE arm
with temperature, shunt calibration across network terminals
Figure 5. Linearization circuit. Figure 6. LST network circuit.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
68
2 and 3 will be correct only when the sensor temperature is
near +75F [+24C]. So, it is preferable that calibration
resistors be shunted across the DUMMY arm (network
terminals 1 and 2). When it is desired to calibrate across both
arms to simulate both plus and minus temperature changes,
and it is not convenient to stabilize sensor temperature near
+75F [+24C], the sensor can be replaced by a precision
50.0-ohm resistor during calibration.
When applying shunt calibration to the differential
temperature measurement circuit shown in Figure 7a, it will
be necessary to calibrate across the ACTIVE arms (network
terminals 2 and 3). Shunting network 1 will simulate a
temperature decrease for sensor 1 or an increase for sensor
2. Under these conditions the nominal or common-mode
temperature for both sensors must be near +75F [+24C] if
it is desired to calibrate across both networks. If this cannot
be conveniently arranged, the sensors should be temporarily
replaced by 50.0-ohm resistors during calibration. If the
alternate full-bridge connection of Figure 7b is used for a
differential operation, shunt calibration steps can be applied
to either DUMMY arm, terminals 1 and 2, regardless of
temperature.
For best accuracy, it is always advisable to select shunt
calibration values that are close to the temperature span of
greatest interest. When the instrument readings are not in
agreement with the simulated calibration temperature, the
GAGE FACTOR can be adjusted to eliminate the error.
Custom LST networks can be tailored for special temperature
ranges or output slopes, and for impedance matching.
Contact our Applications Engineering Department for
details.
5.0 Sources of Error
5.1 Deviations from Linearity
Linearization of TG-Series sensors with passive resistance
circuits characteristically leaves small deviations from true
linearity. To assure optimum correction for nonlinearity of
the nickel, the user may select from four temperature ranges:
200 to +500F, use LST-10F/100F-350D
320 to +100F, use LST-10F/100F-350C
150 to +260C, use LST-10C/100C-350D
200 to +25C, use LST-10C/100C-350C
Deviations from linearity for these networks are shown in
Figure 8.
The expansion coefficient of the structure to which the
temperature sensor is bonded will affect linearity, and Figure
9 shows the deviation of TG sensors with LST-10F-350D
conditioning networks on 1018 steel and 2024 aluminum
alloy. For highest accuracy it is necessary to calibrate sensor/
LST systems on the material to be used. In most work,
however, the average curves of Figure 8 are satisfactory.
5.2 Leadwire Effects and Related Errors
Leadwires are a source of error in all circuits using TG
sensors, except those of the type shown in Figures 2 and
3. To minimize these errors, leadwires between the sensor
and the readout device (or LST network) should be of
low resistance and no longer than necessary. A total two-
wire resistance of 0.5 ohm will introduce a shift or offset
of about +4F [+2C] at room temperature. This leadwire
resistance corresponds to 25 ft [7.5 m] of AWG No. 20 [0.8-
mm diameter] copper double leads, or 100 ft [30 m] of AWG
No. 14 [1.6-mm diameter] double leads.
Changes in leadwire temperature are normally a minor
source of error. A change of 50F [28C] over the entire
length of a 0.5-ohm copper leadwire circuit will create an
offset error of approximately 0.4F [0.2C] when the sensor
temperature is near +75F [+24C]. This error decreases at
higher sensor temperatures and increases at lower sensor
temperatures. Accurate measurements in the cryogenic
temperature region may require the approach of Figures 2
or 3 when long lengths of small diameter leadwire must be
employed.
Initial zero errors or offsets due to the tolerances applicable
to LST networks and the TG sensors themselves can be
eliminated by stabilizing the sensor installation at any known
temperature close to +75F [+24C], and then setting the
instrument BALANCE so that the reading corresponds
(a)
(b)
Figure 7. Differential temperature measurement
using two TG sensors and two LST networks
(a) half-bridge arrangement and (b) full-bridge arrangement.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
69
to this known temperature. This procedure also eliminates
offset error caused by initial leadwire resistance.
In certain circumstances it may be necessary to locate the
instrumentation at a long distance from the sensor installations.
When LST networks are employed under these conditions, it
is preferable to position the network close to its associated
sensor and use a three-wire lead circuit between the network
and the remote indicator (this should not be done if the
test temperature exceeds the temperature capability of the
LST network). This type of hookup will eliminate first-
order offset errors due to leadwire resistance and leadwire
temperature changes. Desensitization or slope-change error
is greatly reduced and can be eliminated by setting the strain
indicator GAGE FACTOR properly. The correct setting
can be calculated on the basis of known leadwire resistance,
or directly determined by applying shunt calibration to the
remote network terminals.
5.3 Strain Sensitivity Errors
The strain sensitivity of pure nickel can create error signals
when TG sensors are installed in areas of high mechanical
strain. The magnitude of this effect is fairly small, however,
as shown in Figure 10.
The shape of this curve is caused by the nonlinear response
of pure nickel. The strain-sensitivity coefcient has a high
negative value in the central portion of the elastic region and
tends toward a much smaller positive value on either side
of this region. It will be observed that compressive strains
result in smaller error signals, and this strain eld orientation
should therefore be selected for sensor placement when
possible. The center of symmetry of this curve is located at
approximately +750, because the manufacturing process
leaves the sensor with a residual compression of this value.
It is important to realize that the center of symmetry can
be shifted by installing the gage on materials of different
Figure 9. Typical deviation from linearity for TG-Series sensors
mounted on 1018 steel and on 2024 aluminum alloy.
Figure 10. Typical error signal caused by strain applied to TG
sensor. Data applies to sensor temperatures near +75F [+24C].
LST-10F/100F-350D LST-10C/100C-350D
LST-10F/100F-350C LST-10C/100C-350C
Figure 8. Typical deviation from linearity for TG-Series sensors and LST networks.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
70
thermal expansion coefcients and/or with different adhesive
cure temperatures. It is for this reason that gage response
when mounted on aluminum alloy will differ slightly from
that obtained when mounted on steel. The tables on the
following pages demonstrate the change in resistance-
versus-temperature characteristic created by these two
mounting conditions.
It has been shown that repeatability of properly installed TG
sensors can be better than 0.05% of applied temperature
span. To take full advantage of this repeatability, and of
the other intrinsic features of TG temperature sensors, it is
always advisable to conduct a calibration run of the sensor
mounted on a specic material when highest measurement
accuracy is required.
5.4 Stability
In common with most other organic resin systems, the
matrix of TG sensors will slowly sublimate and lose strength
when aged at elevated temperatures. When properly installed,
life will be essentially innite below +250F [+120C], and
will be approximately 10 000 hours at +400F [+205C]. At
+500F [+260C], life can be estimated at 1000 hours in the
presence of air, and will be considerably extended if an inert
atmosphere is used.
The sensing grids are quite stable under the aging conditions
described above. If exposed to temperatures above +500F
[+260C], however, a slight shift in resistivity will occur,
together with a small change in temperature coefcient.
For example, if the WTG-Series sensor is exposed to a
temperature of +600F [+315C] for one hour, the +75.0F
[+23.9C] resistance will shift from 50 ohms to approximately
50.6 ohms. On a normalized basis, the resistance increase
from +75.0 to +450F [+23.9 to +232C] will become 140%
instead of the previous 143%. Operation at temperatures
below +550F [+290C] will thereafter be stable and
repeatable.
5.5 Self-Heating
In order to obtain a useful output from passive transducers
such as TG temperature sensors, it is necessary to apply
an excitation voltage, which results in self-heating of the
sensors. This will cause a certain temperature rise in the
surface to which the sensor is bonded, thus creating an error
signal. Since TG sensors have a high temperature coefcient
of resistance, it is not necessary to utilize high excitation
levels to develop large outputs, and self-heating errors can
easily be kept to insignicant values. When it is necessary to
use high excitation levels to obtain maximum output signals,
it should be noted that the largest practical sensor grid size
should be chosen. The thermal conductivity and thermal
capacity of the specimen will then determine the highest
excitation level that can be used for a given self-heating
error.
It is usually very simple to measure self-heating errors
directly with TG sensors because the excitation can be
varied under constant ambient temperature conditions
while observing the change in output indication in degrees.
A bridge excitation of 0.25V or less will usually produce
self-heating errors of a fraction of one degree for standard
sensors mounted on metallic specimens. Special attention
should be given to self-heating errors when accurate mea-
surements must be made on low thermal conductivity
materials such as plastic or glass.
The attenuation factor incorporated into LST networks
greatly reduces the excitation voltage from strain gage
instrumentation, and self-heating errors are seldom
encountered when this readout method is used with TG
sensors.
6.0 Special Sensors
In addition to the standard line of TG sensors described
in the Micro-Measurements Precision Strain Gages Data
Book, we can furnish almost any sensor pattern desired in a
wide range of resistances. Setup charges will be minimized
when the special pattern design corresponds to one of the
patterns in the line of EA-Series strain gages, although
resistances available are sometimes limited.
Balco, an alloy of nickel and iron with a high temperature-
coefficient-of-resistance and a resistivity 2.4 times that
of pure nickel, can also be furnished on special order for
measurement or control functions. This alloy is frequently
used for temperature compensation of transducer gage
circuits.
Two xed bondable Balco resistors are commonly inserted
in the excitation leadwires of a transducer bridge circuit to
provide automatic compensation for the combined effects of
elastic modulus variation (in the transducer) and gage factor
variation (in the strain gage) with temperature. A series
of adjustable, bondable resistor patterns is available, and
permits trimming to the exact value of resistance required.
Both resistor types are also available in nickel.
While Balco has a slightly lower temperature coefcient of
resistance than high-purity nickel, its lower cost and higher
resistivity permit smaller size and better economy. Figure
1 shows the resistance-versus-temperature characteristic of
Balco compared to that for nickel.
7.0 Calibration
Nickel foil lots are calibrated with specially designed test
equipment consisting of a carefully controlled, uniform
temperature bath, and a platinum resistance standard having
calibration results traceable to the National Institute of
Standards and Technology. Accuracy of calibration during
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
71
these tests is +0.5F [+0.3C]. The temperature range used is
320 to +500F [195 to +260C]. The test data from TG
sensors shows that their behavior can be described generally
by a polynomial equation of the form:
R = A + BT + CT
2
+ DT
3
+ ET
4
+ FT
5
+ GT
6
(1)
where: R = Resistance of the gage
T = Temperature
A through G = Constants determined using
regression analysis curve tting
When the RTDs are being used to indicate temperature, the
equation must be entered in the transposed form:
T = A + BR + CR
2
+ DR
3
+ ER
4
+ FR
5
+ GR
6
(2)
Values of the constants for Eqs. (1) and (2) for 50-ohm (at
+75F [+23.9C]) nickel TG sensors are listed in Tables 1A
and 1B.
Notes
1. The constants A through G in Eq. (2) are not the same
as A through G in Eq. (1).
2. Curve-t Eqs. 1 and 2 are best-t equations, and may
not produce exactly the data in Tables 2 through 5.
8.0 Resistance-versus-Temperature Tables
The following section gives the resistance of standard Micro-
Measurements 50-ohm TG nickel temperature sensors from
the boiling point of liquid nitrogen (320.4F [195.8C] to
+500F [+260C]). Tables are provided for both Fahrenheit
(in two-degree increments) and Celsius (in one-degree
increments) scales, and for sensors mounted on both steel
and aluminum alloy. It is worth noting that different types
of substrate material (as discussed in Section 5.1), or even
the same type of substrate material obtained from different
sources, may have different thermal expansion properties,
and may affect sensor response.
References
1. Watson, Robert B., James Dorsey, James E. Starr,
Conditioning Circuits for Bondable RTDs, 30th
International Instrumentation Symposium, Instrument
Society of America, Denver, Colorado, May 1984.
2. Amer i can Nat i onal St andard ASTM E- 644,
Standard Methods for Testing Industrial Resistance
Thermometers. American Society for Testing and
Materials. 1916 Race Street, Philadelphia, Pennsylvania
19103, 1978.
3. Micro-Measurements, Tech Note TN-514, Shunt
Calibration of Strain Gage Instrumentation, 1988.
Table 1ARegression Analysis Constants
for Calculating Resistance
(Lot T04 Nickel)
Constants
Material
1018
Steel
2024-T4
Aluminum Alloy
A
3.946795 x 101
[4.384157 x 101]
3.939097 x 101
[4.378630 x 101]
B
1.33972 x 101
[2.50996 x 101]
1.34525 x 101
[2.52373 x 101]
C
8.31445 x 105
[2.84885 x 104]
8.63411 x 105
[2.95290 x 104]
D
4.72396 x 108
[3.00807 x 107]
5.02142 x 108
[2.88453 x 107]
E
4.93933 x 1011
[2.03720 x 1010]
1.21108 x 1011
[2.13625 x 1010]
F
2.16840 x 1013
[2.95460 x 1012]
2.51606 x 1013
[2.97514 x 1012]
G
3.15935 x 1016
[1.07663 x 1014]
4.88934 x 1016
[1.65922 x 1014]
Coefcients in brackets are C.
Table 1BRegression Analysis Constants
for Calculating Temperature
(Lot T04 Nickel)
Constants
Material
1018
Steel
2024-T4
Aluminum Alloy
A
3.87148 x 10
2
[2.32852 x 10
2
]
3.85275 x 10
2
[2.31809 x 10
2
]
B
1.437356 x 10
1
[79.8384 x 10
-1
]
1.423312 x 10
1
[79.0495 x 10
1
]
C
20.6576 x 10
2
[11.4693 x 10
2
]
1.96463 x 10
1
[1.09006 x 10
1
]
D
3.47019 x 10
3
[1.92684 x 10
3
]
3.14370 x 10
3
[1.74321 x 10
3
]
E
3.70193 x 10
5
[2.05640 x 10
5
]
3.23587 x 10
5
[1.79412 x 10
5
]
F
2.05767 x 10
7
[1.14362 x 10
7
]
1.75570 x 10
7
[9.73545 x 10
8
]
G
4.55192 x 10
10
[2.53118 x 10
10
]
3.82500 x 10
10
[2.12139 x 10
10
]
Coefcients in brackets are C.
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
72
Table 2Resistance versus Temperature in Degrees Fahrenheit
Type TG Nickel Sensors Mounted on 1018 SteelLot No. T04AH
F R F R F R F R F R F R
+500
+498
+496
+494
+492
+490
+488
+486
+484
+482
+480
+478
+476
+474
+472
+470
+468
+466
+464
+462
+460
+458
+456
+454
+452
+450
+448
+446
+444
+442
+440
+438
+436
+434
+432
+430
+428
+426
+424
+422
+420
+418
+416
+414
+412
+410
+408
+406
+404
+402
+400
+398
+396
+394
+392
+390
+388
+386
+384
+382
+380
+378
+376
+374
+372
+370
+368
+366
+364
134.27
133.73
133.19
132.66
132.13
131.60
131.07
130.54
130.02
129.49
128.97
128.45
127.93
127.42
126.90
126.39
125.88
125.37
124.86
124.36
123.85
123.35
122.85
122.35
121.85
121.36
120.86
120.37
119.88
119.39
118.90
118.42
117.93
117.45
116.97
116.49
116.01
115.53
115.06
114.59
114.11
113.64
113.17
112.71
112.24
111.78
111.31
110.85
110.39
109.93
109.48
109.02
108.57
108.12
107.66
107.21
106.77
106.32
105.87
105.43
104.99
104.54
104.10
103.67
103.23
102.79
102.36
101.92
101.49
+362
+360
+358
+356
+354
+352
+350
+348
+346
+344
+342
+340
+338
+336
+334
+332
+330
+328
+326
+324
+322
+320
+318
+316
+314
+312
+310
+308
+306
+304
+302
+300
+298
+296
+294
+292
+290
+288
+286
+284
+282
+280
+278
+276
+274
+272
+270
+268
+266
+264
+262
+260
+258
+256
+254
+252
+250
+248
+246
+244
+242
+240
+238
+236
+234
+232
+230
+228
+226
101.06
100.63
100.20
99.78
99.35
98.93
98.50
98.08
97.66
97.24
96.82
96.41
95.99
95.58
95.16
94.75
94.34
93.93
93.52
93.12
92.71
92.31
91.90
91.50
91.10
90.70
90.30
89.90
89.50
89.11
88.71
88.32
87.93
87.54
87.15
86.76
86.37
85.98
85.60
85.21
84.83
84.45
84.07
83.69
83.31
82.93
82.55
82.17
81.80
81.43
81.05
80.68
80.31
79.94
79.57
79.20
78.83
78.47
78.10
77.74
77.37
77.01
76.65
76.29
75.93
75.57
75.21
74.86
74.50
+224
+222
+220
+218
+216
+214
+212
+210
+208
+206
+204
+202
+200
+198
+196
+194
+192
+190
+188
+186
+184
+182
+180
+178
+176
+174
+172
+170
+168
+166
+164
+162
+160
+158
+156
+154
+152
+150
+148
+146
+144
+142
+140
+138
+136
+134
+132
+130
+128
+126
+124
+122
+120
+118
+116
+114
+112
+110
+108
+106
+104
+102
+100
+ 98
+ 96
+ 94
+ 92
+ 90
+ 88
74.15
73.79
73.44
73.09
72.73
72.38
72.03
71.69
71.34
70.99
70.64
70.30
69.96
69.61
69.27
68.93
68.59
68.25
67.91
67.57
67.23
66.89
66.56
66.22
65.89
65.56
65.22
64.89
64.56
64.23
63.90
63.57
63.24
62.92
62.59
62.27
61.94
61.62
61.29
60.97
60.65
60.33
60.01
59.69
59.37
59.05
58.74
58.42
58.10
57.79
57.48
57.16
56.85
56.54
56.23
55.92
55.61
55.30
54.99
54.68
54.37
54.07
53.76
53.46
53.15
52.85
52.55
52.25
51.95
+86
+84
+82
+80
+78
+76
+75.0
+74
+72
+70
+68
+66
+64
+62
+60
+58
+56
+54
+52
+50
+48
+46
+44
+42
+40
+38
+36
+34
+32
+30
+28
+26
+24
+22
+20
+18
+16
+14
+12
+10
+ 8
+ 6
+ 4
+ 2
+ 0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
51.65
51.35
51.05
50.75
50.45
50.15
50.00
49.86
49.56
49.27
48.97
48.68
48.38
48.09
47.80
47.51
47.22
46.93
46.64
46.35
46.07
45.78
45.49
45.21
44.92
44.64
44.35
44.07
43.79
43.51
43.23
42.94
42.67
42.39
42.11
41.83
41.55
41.28
41.00
40.73
40.45
40.18
39.90
39.63
39.36
39.09
38.82
38.55
38.28
38.01
37.74
37.47
37.21
36.94
36.67
36.41
36.15
35.88
35.62
35.36
35.10
34.83
34.57
34.31
34.05
33.80
33.54
33.28
33.02
50
52
54
56
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
100
102
104
106
108
110
112
114
116
118
120
122
124
126
128
130
132
134
136
138
140
142
144
146
148
150
152
154
156
158
160
162
164
166
168
170
172
174
176
178
180
182
184
186
32.77
32.51
32.26
32.01
31.75
31.50
31.25
31.00
30.75
30.50
30.25
30.00
29.75
29.50
29.26
29.01
28.76
28.52
28.27
28.03
27.79
27.55
27.31
27.06
26.82
26.58
26.35
26.11
25.87
25.63
25.40
25.16
24.93
24.69
24.46
24.23
24.00
23.76
23.53
23.30
23.08
22.85
22.62
22.39
22.17
21.94
21.72
21.49
21.27
21.04
20.82
20.60
20.38
20.16
19.94
19.72
19.51
19.29
19.07
18.86
18.64
18.43
18.22
18.00
17.79
17.58
17.37
17.16
16.95
188
190
192
194
196
198
200
202
204
206
208
210
212
214
216
218
220
222
224
226
228
230
232
234
236
238
240
242
244
246
248
250
252
254
256
258
260
262
264
266
268
270
272
274
276
278
280
282
284
286
288
290
292
294
296
298
300
302
304
306
308
310
312
314
316
318
320
320.4
16.74
16.54
16.33
16.13
15.92
15.72
15.51
15.31
15.11
14.91
14.71
14.51
14.31
14.11
13.92
13.72
13.53
13.33
13.14
12.95
12.75
12.56
12.37
12.18
12.00
11.81
11.62
11.44
11.25
11.07
10.88
10.70
10.52
10.34
10.16
9.98
9.80
9.63
9.45
9.28
9.10
8.93
8.76
8.58
8.41
8.24
8.08
7.91
7.74
7.58
7.41
7.25
7.09
6.92
6.76
6.60
6.44
6.29
6.13
5.97
5.82
5.67
5.51
5.36
5.21
5.06
4.91
4.88
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
73
Table 3Resistance versus Temperature in Degrees Fahrenheit
Type TG Nickel Sensors Mounted on 2024-T4 Aluminum AlloyLot No. T04AH
F R F R F R F R F R F R
+500
+498
+496
+494
+492
+490
+488
+486
+484
+482
+480
+478
+476
+474
+472
+470
+468
+466
+464
+462
+460
+458
+456
+454
+452
+450
+448
+446
+444
+442
+440
+438
+436
+434
+432
+430
+428
+426
+424
+422
+420
+418
+416
+414
+412
+410
+408
+406
+404
+402
+400
+398
+396
+394
+392
+390
+388
+386
+384
+382
+380
+378
+376
+374
+372
+370
+368
+366
+364
134.14
133.59
133.04
132.49
131.95
131.40
130.86
130.33
129.79
129.26
128.72
128.19
127.67
127.14
126.62
126.09
125.57
125.05
124.54
124.02
123.51
123.00
122.49
121.99
121.48
120.98
120.48
119.98
119.48
118.99
118.49
118.00
117.51
117.02
116.54
116.05
115.57
115.09
114.61
114.13
113.65
113.18
112.71
112.23
111.77
111.30
110.83
110.37
109.90
109.44
108.98
108.53
108.07
107.61
107.16
106.71
106.26
105.81
105.36
104.92
104.47
104.03
103.59
103.15
102.71
102.28
101.84
101.41
100.98
+362
+360
+358
+356
+354
+352
+350
+348
+346
+344
+342
+340
+338
+336
+334
+332
+330
+328
+326
+324
+322
+320
+318
+316
+314
+312
+310
+308
+306
+304
+302
+300
+298
+296
+294
+292
+290
+288
+286
+284
+282
+280
+278
+276
+274
+272
+270
+268
+266
+264
+262
+260
+258
+256
+254
+252
+250
+248
+246
+244
+242
+240
+238
+236
+234
+232
+230
+228
+226
100.54
100.12
99.69
99.26
98.84
98.41
97.99
97.57
97.15
96.73
96.31
95.90
95.48
95.07
94.66
94.25
93.84
93.43
93.03
92.62
92.22
91.82
91.41
91.01
90.61
90.22
89.82
89.43
89.03
88.64
88.25
87.86
87.47
87.08
86.69
86.31
85.92
85.54
85.16
84.77
84.39
84.02
83.64
83.26
82.89
82.51
82.14
81.76
81.39
81.02
80.65
80.28
79.92
79.55
79.19
78.82
78.46
78.10
77.74
77.38
77.02
76.66
76.30
75.94
75.59
75.24
74.88
74.53
74.18
+224
+222
+220
+218
+216
+214
+212
+210
+208
+206
+204
+202
+200
+198
+196
+194
+192
+190
+188
+186
+184
+182
+180
+178
+176
+174
+172
+170
+168
+166
+164
+162
+160
+158
+156
+154
+152
+150
+148
+146
+144
+142
+140
+138
+136
+134
+132
+130
+128
+126
+124
+122
+120
+118
+116
+114
+112
+110
+108
+106
+104
+102
+100
+ 98
+ 96
+ 94
+ 92
+ 90
+ 88
73.83
73.48
73.13
72.78
72.43
72.09
71.74
71.40
71.06
70.71
70.37
70.03
69.69
69.35
69.01
68.68
68.34
68.01
67.67
67.34
67.00
66.67
66.34
66.01
65.68
65.35
65.02
64.70
64.37
64.04
63.72
63.39
63.07
62.75
62.43
62.11
61.79
61.47
61.15
60.83
60.51
60.20
59.88
59.57
59.25
58.94
58.62
58.31
58.00
57.69
57.38
57.07
56.76
56.45
56.15
55.84
55.53
55.23
54.93
54.62
54.32
54.02
53.71
53.41
53.11
52.81
52.51
52.22
51.92
+86
+84
+82
+80
+78
+76
+75.0
+74
+72
+70
+68
+66
+64
+62
+60
+58
+56
+54
+52
+50
+48
+46
+44
+42
+40
+38
+36
+34
+32
+30
+28
+26
+24
+22
+20
+18
+16
+14
+12
+10
+ 8
+ 6
+ 4
+ 2
+ 0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
51.62
51.33
51.03
50.73
50.44
50.15
50.00
49.86
49.56
49.27
48.97
48.68
48.38
48.09
47.80
47.51
47.22
46.93
46.64
46.35
46.07
45.78
45.49
45.21
44.92
44.64
44.35
44.07
43.79
43.51
43.23
42.94
42.67
42.39
42.11
41.83
41.55
41.28
41.00
40.73
40.45
40.18
39.90
39.63
39.36
39.09
38.82
38.55
38.28
38.01
37.74
37.47
37.21
36.94
36.67
36.41
36.15
35.88
35.62
35.36
35.10
34.83
34.57
34.31
34.05
33.80
33.54
33.28
33.02
50
52
54
56
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
100
102
104
106
108
110
112
114
116
118
120
122
124
126
128
130
132
134
136
138
140
142
144
146
148
150
152
154
156
158
160
162
164
166
168
170
172
174
176
178
180
182
184
186
32.77
32.51
32.26
32.01
31.75
31.50
31.25
31.00
30.75
30.50
30.25
30.00
29.75
29.50
29.26
29.01
28.76
28.52
28.27
28.03
27.79
27.55
27.31
27.06
26.82
26.58
26.35
26.11
25.87
25.63
25.40
25.16
24.93
24.69
24.46
24.23
24.00
23.76
23.53
23.30
23.08
22.85
22.62
22.39
22.17
21.94
21.72
21.49
21.27
21.04
20.82
20.60
20.38
20.16
19.94
19.72
19.51
19.29
19.07
18.86
18.64
18.43
18.22
18.00
17.79
17.58
17.37
17.16
16.95
188
190
192
194
196
198
200
202
204
206
208
210
212
214
216
218
220
222
224
226
228
230
232
234
236
238
240
242
244
246
248
250
252
254
256
258
260
262
264
266
268
270
272
274
276
278
280
282
284
286
288
290
292
294
296
298
300
302
304
306
308
310
312
314
316
318
320
320.4
16.74
16.54
16.33
16.13
15.92
15.72
15.51
15.31
15.11
14.91
14.71
14.51
14.31
14.11
13.92
13.72
13.53
13.33
13.14
12.95
12.75
12.56
12.37
12.18
12.00
11.81
11.62
11.44
11.25
11.07
10.88
10.70
10.52
10.34
10.16
9.98
9.80
9.63
9.45
9.28
9.10
8.93
8.76
8.58
8.41
8.24
8.08
7.91
7.74
7.58
7.41
7.25
7.09
6.92
6.76
6.60
6.44
6.29
6.13
5.97
5.82
5.67
5.51
5.36
5.21
5.06
4.91
4.88
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
74
Table 4Resistance versus Temperature in Degrees Celcius
Type TG Nickel Sensors Mounted on 1018 SteelLot No. T04AH
C R C R C R C R C R C R C R
+260
+259
+258
+257
+256
+255
+254
+253
+252
+251
+250
+249
+248
+247
+246
+245
+244
+243
+242
+241
+240
+239
+238
+237
+236
+235
+234
+233
+232
+231
+230
+229
+228
+227
+226
+225
+224
+223
+222
+221
+220
+219
+218
+217
+216
+215
+214
+213
+212
+211
+210
+209
+208
+207
+206
+205
+204
+203
+202
+201
+200
+199
+198
+197
+196
+195
134.27
133.78
133.30
132.82
132.34
131.86
131.39
130.91
130.44
129.96
129.49
129.02
128.56
128.09
127.62
127.16
126.70
126.24
125.78
125.32
124.86
124.41
123.95
123.50
123.05
122.60
122.15
121.70
121.26
120.81
120.37
119.93
119.49
119.05
118.61
118.17
117.74
117.31
116.87
116.44
116.01
115.58
115.15
114.73
114.30
113.88
113.46
113.03
112.61
112.20
111.78
111.36
110.95
110.53
110.12
109.71
109.30
108.89
108.48
108.07
107.66
107.26
106.86
106.45
106.05
105.65
+194
+193
+192
+191
+190
+189
+188
+187
+186
+185
+184
+183
+182
+181
+180
+179
+178
+177
+176
+175
+174
+173
+172
+171
+170
+169
+168
+167
+166
+165
+164
+163
+162
+161
+160
+159
+158
+157
+156
+155
+154
+153
+152
+151
+150
+149
+148
+147
+146
+145
+144
+143
+142
+141
+140
+139
+138
+137
+136
+135
+134
+133
+132
+131
+130
+129
105.25
104.85
104.46
104.06
103.67
103.27
102.88
102.49
102.10
101.71
101.32
100.93
100.55
100.16
99.78
99.39
99.01
98.63
98.25
97.87
97.49
97.12
96.74
96.37
95.99
95.62
95.25
94.88
94.51
94.14
93.77
93.40
93.04
92.67
92.31
91.94
91.58
91.22
90.86
90.50
90.14
89.78
89.43
89.07
88.71
88.36
88.01
87.66
87.30
86.95
86.60
86.26
85.91
85.56
85.21
84.87
84.52
84.18
83.84
83.50
83.16
82.82
82.48
82.14
81.80
81.46
+128
+127
+126
+125
+124
+123
+122
+121
+120
+119
+118
+117
+116
+115
+114
+113
+112
+111
+110
+109
+108
+107
+106
+105
+104
+103
+102
+101
+100
+ 99
+ 98
+ 97
+ 96
+ 95
+ 94
+ 93
+ 92
+ 91
+ 90
+ 89
+ 88
+ 87
+ 86
+ 85
+ 84
+ 83
+ 82
+ 81
+ 80
+ 79
+ 78
+ 77
+ 76
+ 75
+ 74
+ 73
+ 72
+ 71
+ 70
+ 69
+ 68
+ 67
+ 66
+ 65
+ 64
+ 63
81.13
80.79
80.46
80.12
79.79
79.46
79.13
78.80
78.47
78.14
77.81
77.48
77.16
76.83
76.51
76.18
75.86
75.54
75.21
74.89
74.57
74.25
73.93
73.61
73.30
72.98
72.66
72.35
72.03
71.72
71.41
71.09
70.78
70.47
70.16
69.85
69.54
69.23
68.93
68.62
68.31
68.01
67.70
67.40
67.10
66.79
66.49
66.19
65.89
65.59
65.29
64.99
64.69
64.40
64.10
63.80
63.51
63.21
62.92
62.62
62.33
62.04
61.75
61.46
61.16
60.87
+62
+61
+60
+59
+58
+57
+56
+55
+54
+53
+52
+51
+50
+49
+48
+47
+46
+45
+44
+43
+42
+41
+40
+39
+38
+37
+36
+35
+34
+33
+32
+31
+30
+29
+28
+27
+26
+25
+24
+23.9
+23
+22
+21
+20
+19
+18
+17
+16
+15
+14
+13
+12
+11
+10
+ 9
+ 8
+ 7
+ 6
+ 5
+ 4
+ 3
+ 2
+ 1
+ 0
1
2
60.59
60.30
60.01
59.72
59.43
59.15
58.86
58.58
58.29
58.01
57.73
57.44
57.16
56.88
56.60
56.32
56.04
55.76
55.48
55.20
54.93
54.65
54.37
54.10
53.82
53.55
53.28
53.00
52.73
52.46
52.19
51.92
51.65
51.38
51.11
50.84
50.57
50.30
50.03
50.00
49.77
49.50
49.24
48.97
48.71
48.44
48.18
47.92
47.66
47.39
47.13
46.87
46.61
46.35
46.09
45.84
45.58
45.32
45.06
44.81
44.55
44.30
44.04
43.79
43.53
43.28
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
43.03
42.78
42.53
42.27
42.02
41.77
41.52
41.28
41.03
40.78
40.53
40.29
40.04
39.79
39.55
39.30
39.06
38.82
38.57
38.33
38.09
37.85
37.61
37.37
37.13
36.89
36.65
36.41
36.17
35.93
35.70
35.46
35.23
34.99
34.76
34.52
34.29
34.05
33.82
33.59
33.36
33.13
32.90
32.67
32.44
32.21
31.98
31.75
31.52
31.30
31.07
30.85
30.62
30.40
30.17
29.95
29.72
29.50
29.28
29.06
28.84
28.62
28.40
28.18
27.96
27.74
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
27.52
27.31
27.09
26.87
26.66
26.44
26.23
26.01
25.80
25.59
25.37
25.16
24.95
24.74
24.53
24.32
24.11
23.90
23.70
23.49
23.28
23.08
22.87
22.66
22.46
22.26
22.05
21.85
21.65
21.45
21.25
21.04
20.85
20.65
20.45
20.25
20.05
19.85
19.66
19.46
19.27
19.07
18.88
18.69
18.49
18.30
18.11
17.92
17.73
17.54
17.35
17.16
16.97
16.79
16.60
16.41
16.23
16.04
15.86
15.68
15.49
15.31
15.13
14.95
14.77
14.59
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
195.8
14.41
14.23
14.05
13.88
13.70
13.53
13.35
13.18
13.00
12.83
12.66
12.49
12.32
12.15
11.98
11.81
11.64
11.47
11.31
11.14
10.97
10.81
10.65
10.48
10.32
10.16
10.00
9.84
9.68
9.52
9.36
9.21
9.05
8.89
8.74
8.58
8.43
8.28
8.13
7.98
7.83
7.68
7.53
7.38
7.23
7.09
6.94
6.80
6.65
6.51
6.37
6.22
6.08
5.94
5.80
5.67
5.53
5.39
5.26
5.12
4.99
4.88
T
E
C
H

N
O
T
E
TN-506-3
Micro-Measurements
75
Table 5Resistance versus Temperature in Degrees Celcius
Type TG Nickel Sensors Mounted on 2024-T4 Aluminum AlloyLot No. T04AH
C R C R C R C R C R C R C R
+260
+259
+258
+257
+256
+255
+254
+253
+252
+251
+250
+249
+248
+247
+246
+245
+244
+243
+242
+241
+240
+239
+238
+237
+236
+235
+234
+233
+232
+231
+230
+229
+228
+227
+226
+225
+224
+223
+222
+221
+220
+219
+218
+217
+216
+215
+214
+213
+212
+211
+210
+209
+208
+207
+206
+205
+204
+203
+202
+201
+200
+199
+198
+197
+196
+195
134.14
133.64
133.15
132.65
132.16
131.68
131.19
130.70
130.22
129.74
129.26
128.78
128.30
127.82
127.35
126.88
126.41
125.94
125.47
125.00
124.54
124.08
123.61
123.15
122.70
122.24
121.78
121.33
120.88
120.43
119.98
119.53
119.08
118.64
118.20
117.75
117.31
116.88
116.44
116.00
115.57
115.13
114.70
114.27
113.84
113.42
112.99
112.56
112.14
111.72
111.30
110.88
110.46
110.04
109.63
109.21
108.80
108.39
107.98
107.57
107.16
106.75
106.35
105.94
105.54
105.14
+194
+193
+192
+191
+190
+189
+188
+187
+186
+185
+184
+183
+182
+181
+180
+179
+178
+177
+176
+175
+174
+173
+172
+171
+170
+169
+168
+167
+166
+165
+164
+163
+162
+161
+160
+159
+158
+157
+156
+155
+154
+153
+152
+151
+150
+149
+148
+147
+146
+145
+144
+143
+142
+141
+140
+139
+138
+137
+136
+135
+134
+133
+132
+131
+130
+129
104.74
104.34
103.94
103.55
103.15
102.76
102.36
101.97
101.58
101.19
100.80
100.42
100.03
99.64
99.26
98.88
98.50
98.12
97.74
97.36
96.98
96.61
96.23
95.86
95.48
95.11
94.74
94.37
94.00
93.64
93.27
92.90
92.54
92.18
91.82
91.45
91.09
90.73
90.38
90.02
89.66
89.31
88.95
88.60
88.25
87.90
87.55
87.20
86.85
86.50
86.15
85.81
85.46
85.12
84.77
84.43
84.09
83.75
83.41
83.07
82.74
82.40
82.06
81.73
81.39
81.06
+128
+127
+126
+125
+124
+123
+122
+121
+120
+119
+118
+117
+116
+115
+114
+113
+112
+111
+110
+109
+108
+107
+106
+105
+104
+103
+102
+101
+100
+ 99
+ 98
+ 97
+ 96
+ 95
+ 94
+ 93
+ 92
+ 91
+ 90
+ 89
+ 88
+ 87
+ 86
+ 85
+ 84
+ 83
+ 82
+ 81
+ 80
+ 79
+ 78
+ 77
+ 76
+ 75
+ 74
+ 73
+ 72
+ 71
+ 70
+ 69
+ 68
+ 67
+ 66
+ 65
+ 64
+ 63
80.73
80.40
80.06
79.73
79.41
79.08
78.75
78.42
78.10
77.77
77.45
77.12
76.80
76.48
76.16
75.84
75.52
75.20
74.88
74.56
74.25
73.93
73.62
73.30
72.99
72.68
72.36
72.05
71.74
71.43
71.12
70.82
70.51
70.20
69.90
69.59
69.28
68.98
68.68
68.37
68.07
67.77
67.47
67.17
66.87
66.57
66.27
65.98
65.68
65.38
65.09
64.79
64.50
64.21
63.91
63.62
63.33
63.04
62.75
62.46
62.17
61.88
61.59
61.31
61.02
60.73
+62
+61
+60
+59
+58
+57
+56
+55
+54
+53
+52
+51
+50
+49
+48
+47
+46
+45
+44
+43
+42
+41
+40
+39
+38
+37
+36
+35
+34
+33
+32
+31
+30
+29
+28
+27
+26
+25
+24
+23.9
+23
+22
+21
+20
+19
+18
+17
+16
+15
+14
+13
+12
+11
+10
+ 9
+ 8
+ 7
+ 6
+ 5
+ 4
+ 3
+ 2
+ 1
+ 0
1
2
60.45
60.16
59.88
59.60
59.31
59.03
58.75
58.47
58.19
57.91
57.63
57.35
57.07
56.79
56.52
56.24
55.96
55.69
55.41
55.14
54.86
54.59
54.32
54.05
53.77
53.50
53.23
52.96
52.69
52.42
52.16
51.89
51.62
51.35
51.09
50.82
50.56
50.29
50.03
50.00
49.77
49.50
49.24
48.97
48.71
48.44
48.18
47.92
47.66
47.39
47.13
46.87
46.61
46.35
46.09
45.84
45.58
45.32
45.06
44.81
44.55
44.30
44.04
43.79
43.53
43.28
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
43.03
42.78
42.53
42.27
42.02
41.77
41.52
41.28
41.03
40.78
40.53
40.29
40.04
39.79
39.55
39.30
39.06
38.82
38.57
38.33
38.09
37.85
37.61
37.37
37.13
36.89
36.65
36.41
36.17
35.93
35.70
35.46
35.23
34.99
34.76
34.52
34.29
34.05
33.82
33.59
33.36
33.13
32.90
32.67
32.44
32.21
31.98
31.75
31.52
31.30
31.07
30.85
30.62
30.40
30.17
29.95
29.72
29.50
29.28
29.06
28.84
28.62
28.40
28.18
27.96
27.74
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
27.52
27.31
27.09
26.87
26.66
26.44
26.23
26.01
25.80
25.59
25.37
25.16
24.95
24.74
24.53
24.32
24.11
23.90
23.70
23.49
23.28
23.08
22.87
22.66
22.46
22.26
22.05
21.85
21.65
21.45
21.25
21.04
20.85
20.65
20.45
20.25
20.05
19.85
19.66
19.46
19.27
19.07
18.88
18.69
18.49
18.30
18.11
17.92
17.73
17.54
17.35
17.16
16.97
16.79
16.60
16.41
16.23
16.04
15.86
15.68
15.49
15.31
15.13
14.95
14.77
14.59
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
195.8
14.41
14.23
14.05
13.88
13.70
13.53
13.35
13.18
13.00
12.83
12.66
12.49
12.32
12.15
11.98
11.81
11.64
11.47
11.31
11.14
10.97
10.81
10.65
10.48
10.32
10.16
10.00
9.84
9.68
9.52
9.36
9.21
9.05
8.89
8.74
8.58
8.43
8.28
8.13
7.98
7.83
7.68
7.53
7.38
7.23
7.09
6.94
6.80
6.65
6.51
6.37
6.22
6.08
5.94
5.80
5.67
5.53
5.39
5.26
5.12
4.99
4.88
Tech Note TN-507-1
Errors Due to Wheatstone
Bridge Nonlinearity
T
E
C
H

N
O
T
E
77
Revision 19-Nov-2010
MICRO-MEASUREMENTS
1.0 Introduction
Commerci al st at i c st rai n i ndi cators and si gnal
conditioners vary considerably in their circuit details; and,
although most of them are based upon some form of the
Wheatstone bridge circuit, the bridge circuit is employed
in differing ways in different instruments. Because of
the many variations in instrument design, a completely
general treatment of instrument nonlinearities is not
practicable within the scope of this publication. There is,
however, a large class of static strain indicators and signal
conditioners with a more-or-less characteristic circuit
arrangement (employing the unbalanced Wheatstone
bridge), and displaying a characteristic nonlinearity. This
Tech Note has been prepared to provide a simple means
for determining the magnitudes of the nonlinearity errors
and for making corrections when necessary. Note that the
error and correction relationships given here apply only to
instruments having the characteristics dened in Section
2.0. For other instruments, the nonlinearity errors, if they
exist, will have to be determined by direct calibration or
from manufacturers specications.
The nonlinearity error occurs because, when strain mea-
surements are made with an unbalanced Wheatstone
bridge circuit (as described in Section 2.0), there are certain
conditions under which the output of the bridge circuit is
a nonlinear function of the resistance change(s) producing
that output. The error due to the nonlinearity, when
present, is ordinarily small, and can usually be ignored
when measuring elastic strains in metals. However, the
percentage error increases with the magnitude of the strain
being measured, and can become quite signicant at large
strains (for example, the error is about 0.1% at 1000, 1%
at 10 000, and 10% at 100 000; or, as a convenient rule
of thumb, the error, in percent, is approximately equal to
the strain, in percent).
2.0 The Unbalanced
Wheatstone Bridge Circuit
Most static strain indicators and signal conditioners
for use with resistance strain gages use a form of the
Wheatstone bridge circuit in which the bridge arms consist
of one to four active gages. The classical Wheatstone
bridge arrangement has been used for many years for the
accurate measurement of a single unknown resistance;
and, in such instruments, the bridge is balanced at the time
of measurement by adjusting the resistances of the other
arms. The bridge circuit found in most strain indicators,
on the other hand, is unbalanced by the varying gage
resistance(s) at the time of making the measurement, and
is therefore commonly referred to as the unbalanced
Wheatstone bridge.
The output voltage obtained from the unbalanced
Wheatstone bridge is a function of the amount of unbalance,
and is therefore directly related to the strain applied to the
strain gage. However, under certain conditions frequently
encountered in actual practice, the bridge output voltage
is, as noted earlier, a nonlinear function of the resistance
change in the bridge arms. When this occurs, the strain
readings will be somewhat in error.
Figure 1 shows two of the circuit arrangements most
commonly employed in commercial strain indicators and
signal conditioners. In circuit (A), the bridge output voltage
is amplied and displayed on an indicating instrument,
frequently a digital voltmeter. In circuit (B), the bridge
output voltage is nulled by an equal and opposite voltage
injected into the measurement circuit. In both cases, the
Display
E
R1 R2
R3 R4
Amplier
Voltage Injection
Circuit
R1 R2
R3 R4
E
Amplier Display
Balance
Control
Figure 1. Circuit arrangements for commercial instruments.
(A)
(B)
T
E
C
H

N
O
T
E
TN-507-1
Micro-Measurements
78
Errors Due to Wheatstone Bridge Nonlinearity
nonlinearity errors are identical if the amplifiers have
high input impedances, and if the power supplies are of
the constant-voltage type. Note also that in both circuits
the balance control is used only to establish initial
bridge balance before the gages are strained, and that the
balance controls do not form part of the readout circuit.
This type of balance circuit is normally provided with a
limited range so as not to cause problems in resolution and
setting-stability; and, therefore does not greatly inuence
the nonlinearity errors as described in this Tech Note.
To permit a rigorous treatment of the errors without
introducing other considerations, it is assumed throughout
Bridge/Strain Description Bridge Output, E
O
/E Nonlinearity, Corrections
Arrangement mV/V Where (Note 3, 4)
(Note 1) (Notes 2, 3) E
O
/E = K x 10
3
(1- )
(Notes 2, 3)
Single active gage in
uniaxial tension or com-
pression.
Two active gages in uni-
axial stress eld one
aligned with maximum
principal strain, one
Poisson gage.
Two active gages with
equal and opposite
strains typical of
bending-beam arrange-
ment.
Two active gages with
equal strains of same
sign used on oppo-
site sides of column with
low temperature gradi-
ent (bending cancella-
tion, for instance).
Four active gages in
unaxial stress eld two
principal strain, two
Poisson gages (col-
umn).
Four active gages in uni-
axial stress eld two
principal strain, two
Poisson gages (beam).
Four active gages with
pairs subjected to equal
and opposite strains
(beam in bending or
shaft in torsion).
E
Eo
E
Eo
Eo E

Eo E
Eo E
Eo E
Eo E
E
E
F
F
o
=

10
4+2
3
10
6
E
E
F
F
o
=
+
( )

( )

1
1 10
6
10
4+2
3
E
E
F
o
=

2
10
3
E
E
F
F
o
=

10
2+
3
10
6
E
E
F
F
o
=
+
( )

( )

1
1 10
6
10
2+
3
E
E
F
o
=
+
( )
1 10
2
3
E
E
F
o
= 10
3
K
F
F
F
=
=

4
10
10
6
6

2+
K
F
F
F
=
+
=

( )

( )
( )
1
4
1 10
1 10
6
6
2+
K
F
= =
2
0 ;
K
F
F
F
=
=

2
10
10
6
6

2+
K
F
F
F
=
+
=

( )

( )
( )
1
2
1 10
1 10
6
6
2+
K
F
=
+
=
( )
;
1
2
0
K F = = ; 0
=

2
10
6
i
i
F 2

=

2
10
6
i
i
F 2(1+ ) (1 )

=
i
2
=

2
10
6
i
i
F 4

=

2
10
6
i
i
F 4(1+ ) (1 )
=
i
2 1+ ) (
=
i
4
TABLE 1
E
E
F
F
o
=

10
4+2
3
10
6
Notes: 1. (R
1
/R
4
)
nom
= 1; (R
2
/R
3
)
nom
= 1 when two or less active arms are used.
2. Constant voltage power supply is assumed.

3. and
i
(strains) are expressed in microstrain units (in/in x 10
6
) where
i
is the strain indicated by your instrument
and is the actual strain under a single active gage.
4. Expressions in this column correct for Wheatstone bridge nonlinearity (if present) and for the number of active
gages in the circuit.
3
2 1
4
2
3
4
5
6
7
1
T
E
C
H

N
O
T
E
TN-507-1
Micro-Measurements
79
the following discussion that the balance
circuit is either completely disconnected, or
that the control is left at the midpoint of its
range. It is also assumed that the bridge arms
are nominally resistively symmetrical about an
axis joining the output corners of the bridge;
i.e., that:
(R
1
/R
4
)
nom
= 1 = (R
2
/R
3
)
nom
As a result of the ci rcuit arrangements
described above, obtaining a reading from the
instrument (whether or not the process involves
nulling a meter) has no effect on the state
of resistive balance within the Wheatstone
bridge circuit. Even if the Wheatstone bridge
is initially balanced resistively so that R
1
/R
4
=
R
2
/R
3
, this will no longer be true, in general,
when one or more of the strain gages in the
bridge arms are strained. Consequently, the
Wheatstone bridge is ordinarily operated in
a resistively unbalanced state. In this mode
of operation, resistance changes in the bridge
arms may cause changes in the currents
through the arms, depending upon the signs
and magnitudes of the resistance changes in
all four arms. When current changes occur, the
voltage output of the bridge is not proportional
to the resistance changes, and thus the output
is nonlinear with strain, and the instrument
indication is in error.
3.0 Nonlinearities and Corrections
For the class of instruments described in Section 2.0,
Table 1 illustrates the nonlinearities to be expected, and
includes correction relationships. The table gives the
nondimensional output voltage (E
o
/E) as a function of the
applied strain (and gage factor) for a variety of commonly
encountered strain states and different arrangements of
gages on the structural member and within the Wheatstone
bridge. It can readily be seen that the output expressions
for cases 1, 2, 4, and 5 are intrinsically nonlinear, while
those for cases 3, 6, and 7 are linear. Examination of the
column of bridge/strain arrangements demonstrates that
only when the resistance changes due to strain are such
that the currents through the bridge arms remain constant
that is, R
1
/R
1
+ R
4
/R
4
= 0 and R
2
/R
2
+ R
3
/R
3
= 0
is the output a linear function of the strain.
For each of the cases in Table 1, the nondimensional circuit
output can also be expressed in the following form:
E
o
/E = K x 10
3
(1), mV/V
In this relationship, K is a constant, determined by the
gage factor of the strain gage(s) and the number of active
arms in the bridge circuit; and (when not zero) represents
the nonlinearity caused by changes in the currents through
the bridge arms. One hundred times the fraction /1 is then
the percentage nonlinearity in the circuit output.
The nonlinearity column in Table 1 gives the mathematical
expression for calculating as a function of the applied
strain and other relevant parameters. It can be noticed
from the table that on a percentage basis the nonlinearity
magnitudes are identical for cases 1 and 4, and for 2 and
5, although the circuit outputs differ. For convenience in
quickly judging nonlinearity magnitudes, the relationships
in cases 1, 2, 4, and 5 are plotted in Figure 2, assuming
positive (tensile) strains, a gage factor of 2.0, and Poissons
ratio (where involved) of 0.30. The nonlinearity for
compressive strain is opposite in sign and somewhat
different in magnitude, but can always be calculated from
the relationships given in Table 1.
The last (right most) column in Table 1 provides the
relationships for converting the indicated strain,
i
, as
registered by a strain indicator or other instrument system
to the actual surface strain under a single active gage, .
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
1000 10 000 100 000
Case #1, #4
Case #2, #5
P
e
r
c
e
n
t

N
o
n
l
i
n
e
a
r
i
t
y
Strain,
Figure 2. Nonlinearity errors for tensile strain in bridge circuits.
T
E
C
H

N
O
T
E
TN-507-1
Micro-Measurements
80
The expressions in this column correct for Wheatstone
bridge nonlinearity (when present) and for the number of
active gages in the circuit.
4.0 Numerical Examples
For si mpl icity i n presentation, the fol lowi ng three
numerical examples assume, in each case, a quarter-bridge
circuit
with a single active gage, and a gage factor of 2.0.
The procedures employed in the examples are, however,
quite general, and apply to all circuit arrangements in
Table 1.
As a rst example, assume that the quarter-bridge circuit
was initially balanced resistively with no load applied
to the test piece. Subsequently, the test piece was loaded
until the strain indicator registered 15 000 in tension.
From Figure 2 for case 1, the nonlinearity in output is
approximately 1.5 percent or about 225. The actual
surface strain for an indicator reading of 15 000 can be
obtained from the relationship in the corrections column
of Table 1. Substituting F = 2.0 and
i
= 15 000 into the
expression = (2
i
)/(2 F
i
x 10
6
) gives = 15 228. If the
strain indicator had read 15 000 for compression, the
same procedure would yield a surface strain of 14 778.
As demonstrated by these calculations, Wheatstone bridge
nonlinearity causes indicated tensile strains to be too low,
and indicated compressive strains too high.
It was assumed in the previous example that the Wheatstone
bridge was initially in a state of resistive balance. In the
practice of experimental stress analysis with strain gages,
this may not always be the case. For instance, during the
bonding of a strain gage the resistance of the gage may be
altered signicantly from the manufactured value by poor
installation technique. It may also happen that the gage is
strained to the plastic range by assembly or preload stresses
before subsequent strain measurements are to be made.
The initial resistive unbalance, unless it is known to be
insignicant, should be measured and properly accounted
for in making nonlinearity corrections. When great
enough to warrant consideration, the initial unbalance
(expressed in strain units) must be added algebraically to
any subsequent observed strains so that the nonlinearity
correction is based on the total (or net) unbalance of
the Wheatstone bridge at any stage in the measurement
process.
For this example, assume that the strain indicator displays
an initial unbalance of 4500 in the installed gage, with
no load applied to the test object. This is an indicated
unbalance, and therefore includes a small nonlinearity
error which will be corrected for in this case to illustrate
the procedure. Substituting as before into the correction
relationship for case 1 in Table 1, the actual initial unbalance
(in strain units) is: = 4480. After taking this reading
(but not resistively balancing the Wheatstone bridge arms),
the test object is subjected to its specied load. The change
in strain indication corresponding to the applied load is
8000. The total indicated unbalance in the Wheatstone
bridge is then 12 500. The calculated correction for this
strain indication (case 1, Table 1) yields 12 346 for the
actual total unbalance. The actual applied strain is thus
12 346 (4480) = 7866.
As a nal example, consider a case in which the indicated
initial unbalance after installing the strain gage was
2500. Then the gaged member was installed in a
structure with an indicated additional assembly strain of
45 500. After taking this reading, subsequent loading
of the structure produced an indicated strain change of
3000 in the tension direction. What corrections should
be made to determine the actual tensile strain caused by
loading the structure?
Prior to loading the structure, the Wheatstone bridge was
unbalanced by an indicated 48 000. Substituting into
the correction expression in Table 1 for this case, the actual
resistive unbalance prior to loading was 45 802 in strain
units. After loading the structure, the indicated unbalance
in the Wheatstone bridge was 48 000 + 3000 = 45 000.
The calculated correction for this indicated strain yields
43 062. The applied tensile strain due to loading the
structure was thus 43 062 (45 802) = +2740. This
example demonstrates that even with relatively modest
working strains the nonlinearity error can be very signicant
(about 10% in this instance) if the Wheatstone bridge is
operated far from its resistive balance point.
5.0 Nonlinearities in Shunt Calibration and
Dynamic Strain Measurement
The nonlinearity error described in the preceding sections
of this Tech Note should always be kept in mind during the
shunt calibration of a strain indicator or signal conditioner.
In the conventional practice of shunt calibration, the
strain gage is momentarily shunted by a large resistor of
a magnitude selected to produce a decreased resistance in
the Wheatstone bridge arm corresponding to, and precisely
simulating, a predetermined compressive strain in the gage
(at a specied gage factor). As an alternative, the internal
dummy resistance in the adjacent arm of the bridge circuit
can be shunted to simulate a tensile strain in the gage.
During shunt calibration, the strain indicated by the
instrument will be the same as that for a strain gage at
the same level of strain, if the proper calibration resistor
is employed. To be precise, the calibration resistor for
simulating a specic tensile strain is slightly different from
that for the same level of compressive strain because the
nonlinearities are different in tension and compression.
When simulating small strains (less than, say, 2000), the
T
E
C
H

N
O
T
E
TN-507-1
Micro-Measurements
81
tension/compression difference is negligible, and standard
shunt calibration resistors can be used in either mode. For
accurate calibration at large strains, resistors specically
intended for tension or compression simulation should be
selected. Tech Note TN-514, Shunt Calibration of Strain
Gage Instruments, gives the necessary relationships for
calculating the appropriate shunt resistances for both
cases.
The usual practice, after shunting the active or dummy
gage, is to adjust the gage factor or gain control of the
instrument to exactly register the simulated strain level.
Subsequent strain measurements at or near the calibrated
strain level do not require correction for nonlinearity.
However, measurement at a signicantly different level
will be somewhat in error due to the different nonlinearity
at a different strain level. Tech Note TN-514 also gives
relationships for adjusting indicated strains to account for
calibration at one level, followed by strain measurement
at a different level. Although not treated here, leadwire
resistance is another factor to be considered in shunt
calibration, and Tech Note TN-514 provides relationships
to correct for leadwire effects.
Errors can also arise when the initial state of Wheatstone
bridge unbalance is different during shunt calibration
than it is when strain measurements are to be made. If
this situation exists, it is necessary to measure the initial
unbalance and determine the actual simulated strain
following the procedures demonstrated in the second and
third of the preceding numerical examples.
Whenever dynamic strain measurements are made with a
Wheatstone bridge circuit, the bridge is always operated in
the unbalanced mode. Therefore, the nonlinearities listed
in Table 1 of this Tech Note apply to every such dynamic
strain measurement assuming, again, that the bridge is
initially balanced resistively. Under these conditions, the
error due to the nonlinearity is ordinarily small at typical
working strain levels, as illustrated by Figure 2. However,
if the bridge is initially unbalanced, the nonlinearity error
can be much greater; and, with large initial unbalances,
may result in signicantly inaccurate strain indications.
6.0 Summary
The nonlinearity errors occurring in conventional strain
gage bridge circuits are normally small enough to ignore
when measuring modest strain magnitudes such as those
encountered in the elastic range of metals (if the bridge is
initially balanced resistively). Large resistive unbalances
can, on the other hand, lead to sizable errors in strain
indication. The relationships and procedures presented
in this Tech Note can be used when necessary to correct
for such errors. It also follows that for accurate strain
measurements, it is imperative to select strain gages with
tightly controlled resistance tolerances, and to minimize
resistance shifts during gage bonding by carefully following
recommended installation techniques.
Tech Note TN-508-1
MICRO-MEASUREMENTS
Fatigue Characteristics of
Micro-Measurements Strain Gages
T
E
C
H

N
O
T
E
83
All metals are subject to fatigue damage when strained
cy cli cally at sufciently high amplitudes; and the foils used
in strain gages are no exception. Fatigue damage in a strain
gage is rst evidenced as a permanent change in unstrained
resistance of the gage, ordinarily expressed in terms of
equivalent indicated strain, and referred to as zero-shift.
As damage increases in strain gages, cracks eventually begin
to develop and these can result in data that is seriously
in error.
Micro-Measurements monitors three parameters of strain
gages during fatigue life testing: super-sensitivity, gage
factor change, and zero-shift. Super-sensitivity results from
cracks that are just forming, and that are open only during
the tension portion of the loading cycle. If output of a strain
gage is monitored continuously on an oscilloscope during a
fatigue test, the waveform observed for an undamaged gage
will be a sine wave. As cracking starts, the sine wave will
distort during the tension portion of the cycle. Monitoring
for onset of cracking is necessary because, at zero load, cracks
often close and their presence can be hidden. Experimental
stress analysts who unexpectedly encounter large strain
signals from strain gages in cyclic applications should check
signal waveform for any indications of super-sensitivity.
Fatigue damage in strain gages can also cause gage factor
changes, although substantial differences are rare. If
cracking has started, however, it will cause an apparent
increase in tension gage factor, easily detected because the
compression value will be much lower.
Procedures for fatigue testing of Micro-Measurements
gages start with the very best installations that can be made
(see Appendix A). No strain gage application requirement
is more demanding than cyclic endurance. Micro-
Measurements uses NAS-942 test rigs with modied beams
to achieve strain levels of different magnitudes*. Zero of
each gage is recorded along with output for both tension
and compression static strains of the same magnitude as
will be encountered during cycling. Throughout dynamic
loading, gage outputs are monitored for super-sensitivity.
Strain gages can be considered to fail over a wide range
of damage levels, depending on the application and
the required accuracy. For example, in static/dynamic strain
measurement, varying with the particular situation, there is
some level of damage at which zero-shifts may impair the
utility of strain gages for that application. Such zero-shifts
thus represent failure under those conditions, even though
the strain gages could still endure many thousands or millions
of additional cycles before fatigue cracks became evident.
On the other hand, for purely dynamic strain measurement
zero-shift is relatively incidental, and strain gages can be
considered functionally adequate until fatigue damage has
progressed almost to the stage of super-sensitivity.
Normal behavior of strain gages is illustrated by the
following graphs in Figure 1 of data from WK-05-250BG-350
gages. Note that as the strain level is reduced, and life
extended, the spread in data increases markedly. Prediction
of gage life in high cycle fatigue (over 5 x 10
5
cycles) is
difficult because the test data displays large variations.
Using the data in Figure 1, a nominal fatigue life curve can
be drawn for WK-05-250BG-350 gages as shown in Figure
2, based on a 100 zero-shift failure criterion. Other typical
data are shown in Appendix B.
Table 1 is a summary of the fatigue characteristics of
Micro-Measurements strain gages. Fatigue Life in the
table generally refers to the approximate number of cycles
at which a zero-shift of 100 can be expected. WD gages,
normally used only for dynamic testing, have values for a
300 zero-shift (approximately).
The data in Table 1 and graphs in Appendix B were obtained
using 1/4-in [6.4-mm] gage length strain gages. Gage selection
is an important criterion in achieving maximum cyclic life
and many different parameters affect the endurance of
strain gages. A study of Table 1 will provide information
on how the different series (backing/alloy combinations)
relate to each other in fatigue life. In addition, a number
of other criteria have marked effects and each compromise
made deteriorates life to some degree. While there are
many aspects to selection of appropriate gages for any
application, there are some general rules concerning what
will improve or deteriorate fatigue life. (Refer to Micro-
Measurements Tech Note TN-505, Strain Gage Selec tion
Criteria, Procedures, Recommendations.) The larger the
grid area of foil strain gage, the higher its fatigue life; but,
the higher the gages resistance, the lower its fatigue life.
* The NAS-942 beam is designed for constant stress, producing a
single 1500 strain level.
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
84
Fatigue Characteristics of Micro-Measurements Strain Gages
Encap sulation (such as E or SE on E-backed gages) is
helpful while solder and copper have a negative effect; so
CEA-Series gages or gages with Options L, LE, or W on
EA-Series gages should be used with caution. Some very
small Micro-Measurements gages, and some very high
resistance gages, are made using ultra-thin foil. These gages
should be avoided where cyclic endurance is important.
While such gages cant be identied from catalogs, estimated
fatigue life for any Micro-Measurements gages can be
obtained from our Applications Engi neering Depart ment.
Figure 2 Average cyclic endurance of WK strain gages.
Figure 1 Strain gage fatigue test results.
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
85
Table 1
Nominal Fatigue Life 100 Zero-Shift
M-M
Gage Series
Strain Level,
Number
of Cycles
M-M
Gage Series
Strain Level,
Number of
Cycles
CEA
1500
1300
10
5
10
6
SA
1800
1500
10
6
10
7
EA
1500
1200
10
6
10
8
SK
2200
2000
10
6
10
7
ED
2500
2200
10
6
10
7
S2K
1800
1500
10
6
10
7
EK
1800
1500
10
7
10
8
SD
2500
2200
10
6
10
8
EP
1500
1000
10
2
10
4
TA
1700
1500
10
6
10
7
J2A
1700
1500
10
6
10
7
TK
2200
2000
10
7
10
8
J5K
2000
1800
10
7
10
8
TD
2400 10
7
N2A
1700
1500
10
6
10
7
WA
1800
1500
10
6
10
7
N2K
1800
1500
10
7
10
8
WK
2400
2200
10
6
10
7
N3K
1500 10
8
WD
3000
2500
10
5*
10
7*
*300 zero-shift (approximately) for WD strain gages.
The above fatigue life data is based on fully reversed strain levels. As a generalized approximation, this table can be
used for unidirectional strains, or various mean-strains, by taking the indicated peak-to-peak amplitude and derating by
10 percent. As an example, 1500 would be approximately equivalent in gage fatigue damage to strain levels of
+2700 0 +2500
0
2700
200
A mean-strain which increases in a tensile direction during cycling will lead to much earlier failure, however.
References
Strain Gages, Bonded Resistance, Classication Specication NAS-942 (National Aerospace Standard 942) 1963. Aerospace
Industries Association of America, Inc. Published by National Standards Association, Inc., 1200 Quince Orchard Boulevard,
Gaithersburg, MD 20878.
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
86
When designing structures for maximum fatigue endurance,
special care should be employed to avoid rapid changes in
section, stress concentrations, and excess inertial loading, to
name a few considerations. When stress concentrations are
unavoidable, they should be in a minimum strain eld. The
same considerations apply to strain gage installations in a
fatigue environment.
Ideally, bonding agents used to adhere gages should be
unlled, thin-lm-setting adhesives such as M-Bond 200,
M-Bond 600 or 610, or M-Bond AE-10 or AE-15. All
excess adhesive around the edge of gages should be removed
to prevent this unnecessarily thick film of cement from
cracking and initiating failure of the gage bond.
Solder connections present the greatest stress concentrat
ions at the gage location. Solder is poor in fatigue and,
therefore, must be applied sparingly. Controlling solder
ow on gage tabs can be accomplished by carefully masking
the tab with drafting tape. The masked area should expose
a section of the tab only slightly wider than the leadwire
diameter. An alternative for limiting solder ow is selection
of gages with solder dots and encapsulation. EA-Series
gages with Option SE, and SA/SK-Series gages provide very
controlled solder masses for leadwire at tach ment. (Refer to
Micro-Measurements Application Note TT-606, Soldering
Techniques for Lead Attachment to Strain Gages with Solder
Dots.) Also, specialty soft solder such as 50-50% Tin-Indium*
alloy minimizes stress concentration factors at the solder/tab
interface, and will maximize the fatigue life of any user-
soldered gage installation. Data presented in this Tech Note
for gages without leads was obtained using 63-36.7-0.3%
Lead-Tin-Antimony alloy solder (Micro-Measurements
Type 361A-20R). This solder is recommended for its ease of
use and minimum shrinkage.
All heavy instrument leadwires must be soldered to
intermediate terminal strips; terminal strips and gages
should be interconnected with small-diameter wires. Proper
handling procedures for using terminal strips are described in
Micro-Measurements Application Note TT-603, The Proper
Use of Bondable Ter minals in Strain Gage Applications. Any
leadwires located in high strain elds should not be rigidly
constrained because cyclic stress will fail soft copper wires
well before gages fail; a exible restraint should be provided
to prevent peeling of the leadwires and gage tabs from the
gage backing. If the strain eld at gage locations has been
dened, it is good practice to route leads onto solder tabs
in the minimum strain direction. After soldering the leads,
a exible restraint should be positioned on them as close as
possible to the gage backing. This restraint may be a piece
of drafting or aluminum tape (PDT-1 or FA-2, respectively),
a small drop of adhesive such as M-Bond AE-10, or a
protective coating like M-Coat J.
The WK- and WD-Series, with preattached beryllium
copper leads, exhibit the highest fatigue life of any Micro-
Measurements strain gages. When using these gages in a
cyclic strain eld, refer to Micro-Measurements Application
Note TT-604, Leadwire Attachment Tech niques for Obtain ing
Maximum Fatigue Life of Strain Gages, for the recommended
method of leadwire attachment. An EA-Series gage with
Option LE should not be used in a fatigue environment.
Most fatigue failures in the strain gage occur in the solder
tab and transition area the area between the tab and the
outer gridline. When the strain eld has been predened, it is
good practice to position the gage tabs in the lowest possible
strain area.
CEA-Series gages may require special attention, although
these installations can have the same life as reported for an
open-faced EA-Series gage if each installation is carefully
made. But some characteristics which make the former gages
so convenient for static tests must be carefully managed
when maximum fatigue life is required. The large, soft-
copper-coated tabs, bonded in the strain eld, are subject
to cracking at fewer cycles than the high endurance A-foil
alone. The act of tinning the tabs can affect the ultimate
life of the installation, because solder introduces additional
loading on the tabs.
When presented with the large copper-coated tabs of CEA
gages, the natural tendency is to apply solder to the entire
tab. This markedly deteriorates fatigue life. When cyclic loads
are applied to the structure, copper, having lower fatigue
endurance than the A-foil, begins to crack. These cracks
effectively create notch stress concentrations at the copper and
foil interface, which cause the foil to fail prematurely. Solder
has greater stiffness and even poorer fatigue endurance than
copper; hence, the notch concentration effect may cause even
earlier failure of the installation. The encapsulating layer on
CEA gages presents a natural solder stop. Adding solder
up to the encapsulation creates a stress concentration in the
softer copper coating, assuring failure at the encapsulant/
solder/copper interface. By placing a minimum amount of
solder in only the bottom one-third of the CEA tab area,
cyclic life may be markedly improved.
Such solder control may be achieved by masking the solder
location with a piece of drafting tape in the lower one-third
tab area. This should be followed by the application of a
second piece of drafting tape below the rst piece, making
a gap about the diameter of the solder wire. This exposes a
sufcient area of the copper-coated tab for tinning and lead
attachment.
Appendix A Installation Recommendation for Maximum Strain Gage Fatigue Life
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
87
Available solders have a variety of characteristics. While
melt temperatures are most familiar, solders also vary in
their stiffness, shrinkage and their resistance to cyclic loads.
The higher the melt temperature, the greater the solders
stiffness and shrinkage. Eutectic solders, regardless of
composition, provide for the best cyclic endurance for that
composition. The solders listed in Micro-Measurements
Strain Gage Accessories Data Book are all eutectic or
nearly eutectic, and may be selected based on a given
fatigue tests thermal conditions. When the test is to be
run below +350F [+175C], 361A-20R offers the least
stiffness and shrinkage. With lower stiffness, less stress
concentration occurs at the solder/copper interface; and,
the less shrinkage a solder has, the lower the peel stress that
is imposed on the tab during solder solidication. When
the maximum fatigue life of a CEA installation is required,
INDALLOY
50/50 Tin-Indium solder should be used. This

is a very soft solder which melts at +242F [+117C] and
has very low stiffness and shrinkage. The solder requires
some practice to apply as its wetting and ow characteristics
are poor.
Regardless of the solder chosen, only a minimum amount of
solder should be applied to tin the exposed tab area between
the pieces of drafting tape.
Soldering iron temperature should be set no higher than
necessary to melt the solder being used, as excessively high
soldering temperatures may cause the strain gage tabs to lift
from the gage backing.
When auxiliary soldering ux is used with (for example) solid
solder wire, the ux should be applied sparingly because
these uxes contain volatile solvents that may contribute to
tab lifting. Dipping the solder wire into the ux, and then
letting the excess ux drain from the solder for a few seconds
before wiping the uxed solder across the gage solder tabs,
will normally apply the proper amount of ux. The gage
tabs may then be tinned following the procedures outlined
in Micro-Measurements Application Note TT-609, Strain
Gage Soldering Techniques.
Gage-to-terminal wiring may use either solid copper or
stranded leads, and should not be larger than 34-AWG
[0.0063 in [0.160 mm] diameter]. All leads should be
supported as close to the solder connection as practicable.
This prevents inertial loads from peeling the gage tabs from
the gage backing. Our Strain Gage Accessories Data Book
lists various types of wires that can be chosen for the intra-
bridge connections.
* Indium solders, available from Indium Company of America, are
difcult to apply and expensive.
Registered Trademark of Indium Company of America.

Appendix A Installation Recommendation for Maximum Strain Gage Fatigue Life
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
88
Appendix B
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
89
Appendix B
T
E
C
H

N
O
T
E
TN-508-1
Micro-Measurements
90
Appendix B
Tech Note TN-509
MICRO-MEASUREMENTS
Errors Due to Transverse Sensitivity
in Strain Gages
T
E
C
H

N
O
T
E
91
Revision: 28-Jun-2011
Transverse Sensitivity
Transverse sensitivity in a strain gage refers to the behavior
of the gage in responding to strains which are perpendicular
to the primary sensing axis of the gage. Ideally, it would
be preferable if strain gages were completely insensitive
to transverse strains. In practice, most gages exhibit some
degree of transverse sensitivity; but the effect is ordinarily
quite small, and of the order of several percent of the axial
sensitivity.
In plane wire strain gages, transmission of strain into the
wire from a direction perpendicular to the wire axis is nearly
negligible. As a result, the transverse sensitivity of these
gages is due almost entirely to the fact that a portion of the
wire in the end loop lies in the transverse direction. Because
of this, the sign of the transverse sensitivity for a plane
wire gage will always be positive, and the magnitude of the
effect can be calculated quite closely from the geometry of
the grid. This statement does not apply to the small wrap-
around gages having the wire wound on a attened core.
Such gages often exhibit negative transverse sensitivities.
In foil strain gages, on the other hand, the transverse
sensitivity arises from much more complex phenomena,
and it is affected by almost every aspect of grid design and
gage construction. In addition to end loop effects, the foil
gridlines, having a large ratio of width to thickness, are
strained signicantly by transverse strains. The magnitude
of transverse strain transmission into the gridlines is
determined by the relative thicknesses and elastic moduli of
the backing and foil, by the width-to-thickness ratio of the foil
gridlines, and, to a lesser degree, by several other parameters,
including the presence or lack of an encapsulating layer over
the grid.
Depending upon the foil material and its metallurgical
condition, the contribution to transverse sensitivity from
the transmission of transverse strain into the gridlines can
be either positive or negative. Because of this, the overall
transverse sensitivity of a foil strain gage can also be either
positive or negative. While the transverse sensitivity of a
foil gage is thus subject to a greater degree of control in the
design of the gage, the compromises necessary to optimize
all aspects of gage performance generally limit the attainable
reduction in transverse sensitivity.
Errors Due to Transverse Sensitivity
Errors in strain indication due to transverse sensitivity are
generally quite small since the transverse sensitivity itself
is small. However, in biaxial strain elds characterized by
extreme ratios between principal strains, the percentage
error in the smaller strain can be very great if not corrected
for transverse sensitivity. On the other hand, in the particular
case of uniaxial stress in a material with a Poissons ratio of
0.285, the error is zero because the gage factor given by the
manufacturer was measured in such a uniaxial stress eld
and already includes the effect of the Poisson strain. It is
important to note that when a strain gage is used under any
conditions other than those employed in the gage-factor
calibration, there is always some degree of error due to
transverse sensitivity. In other words, any gage which is: (a)
installed on a material with a different Poissons ratio; or
(b) installed on steel, but subjected to other than a uniaxial
stress state; or (c) even installed on steel with a uniaxial
stress state, but aligned with other than the maximum
principal stress, exhibits a transverse-sensitivity error which
may require correction.
The historical practice of quoting gage factors which, in
effect, mask the presence of transverse sensitivity, and which
are correct in themselves for only a specic stress eld in a
specic material, is an unfortunate one. This approach has
generally complicated the use of strain gages, while leading
to errors and confusion. Although the uniaxial stress eld is
very common, it is not highly signicant to the general eld
of experimental stress analysis. There is no particular merit,
therefore, in combining the axial and transverse sensitivities
for this case.
In general, then, a strain gage actually has two gage factors,
F
a
and F
t
, which refer to the gage factors as determined in
a uniaxial strain eld (not uniaxial stress) with, respectively,
the gage axes aligned parallel to and perpendicular to the
strain eld. For any strain eld, the output of the strain gage
can be expressed as:

R
R
F F
a a t t
= + (1)
where:
a
,
t
= strains parallel to and perpendicular to
the gage axis, or the gridlines in the gage.
F
a
= axial gage factor.
F
t
= transverse gage factor.
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
92
Errors Due to Transverse Sensitivity in Strain Gages
Or,

R
R
F K
a a t t
= + ( ) (2)
where: K
F
F
t
t
a
= = transverse sensitivity coefcient,
referred to from here on as the
transverse sensitivity.
When the gage is calibrated for gage factor in a uniaxial
stress eld on a material with Poissons ratio,
0
,

t a
=
0
Therefore,

R
R
F K
a a t a
= ( )
0
or,

R
R
F K
a t a
= ( ) 1
0
(3)
The strain gage manufacturers commonly write this as:

R
R
F =

(3a)
where: F = manufacturers gage factor, which is
deceptively simple in appearance, since, in
reality:
F F K
a t
= ( ) 1
0
(4)
Furthermore, is actually
a
, the strain along the gage axis
(and only one of two strains sensed by the gage during
calibration) when the gage is aligned with the maximum
principal stress axis in a uniaxial stress (not uniaxial strain)
eld, on a material with
0
= 0.285. Errors and confusion
occur through failure to fully comprehend and always
account for the real meanings of F and as used by the
manufacturers.
It is imperative to realize that for any strain eld except that
corresponding to a uniaxial stress eld (and even in the latter
case, with the gage mounted along any direction except the
maximum principal stress axis, or on any material with
Poissons ratio other than 0.285), there is always an error
in strain indication if the transverse sensitivity of the strain
gage is other than zero. In some instances, this error is small
enough to be neglected. In others, it is not. The error due to
transverse sensitivity for a strain gage oriented at any angle,
in any strain eld, on any material, can be expressed as:

n
K
K
t
t
a
t
=
+

0
1
100
0

(5)
where: n
= the error as a percentage of the actual

strain along the gage axis.

0
= the Poissons ratio of the material on
which the manufacturers gage factor, F,
was measured (usually 0.285).

a
,
t
= respectively, the actual strains parallel and
perpendicular to the primary sensing axis
of the gage.*
From the above equation, it is evident that the percentage
error due to transverse sensitivity increases with the absolute
values of Kt and
t
/
a
, whether these parameters are positive
or negative. Equation (5) has been plotted in Figure 1 for
convenience in judging whether the magnitude of the error
may be signicant for a particular strain eld. Figure 1 also
yields an approximate rule-of-thumb for quickly estimating
the error due to transverse sensitivity that is,

n K
t
t
a
x (percent) 100
As Equation (5) shows, this approximation holds quite well
as long as the absolute value
t
/
a
is not close to
0
. For an
example, assume the task of measuring Poisson (transverse)
strain in a uniaxial stress eld. In this case, the Poisson strain
is represented by
a
, the strain along the gage axis, and the
longitudinal strain in the test member by
t
, since the latter
is transverse to the gage axis (see sketch and footnote below).

a t
t a
=
=
1
t
a
P P
* Subscripts (a) and (t) always refer to the axial and transverse direc-
tions with respect to the gage (without regard to directions on the
test surface), while subscripts (x) and (y) refer to an arbitrary set of
orthogonal axes on the test surface, and subscripts ( p) and (q) to the
principal axes.
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
93
If the test specimen is an aluminum alloy, with = 0.32, then
t
/
a
= 1/ = 3.1. Assuming that the transverse sensitivity
of the strain gage is 3% (i.e., K
t
= 0.03*), the rule of
thumb gives an approximate error of +9.3%. The actual
error, calculated from Equation (5), is +8.5%.
Correcting for Transverse Sensitivity
The effects of transverse sensitivity should always be
considered in the experimental stress analysis of a biaxial
stress eld with strain gages. Either it should be demonstrated
that the effect of transverse sensitivity is negligible and can
be ignored, or, if not negligible, the proper correction should
be made. Since a two- or three-gage rosette will ordinarily
be used in such cases, simple correction methods are given
here for the two-gage 90-degree rosette, the three-gage
rectangular rosette, and the delta rosette. Unless otherwise
noted, these corrections apply to rosettes in which the
transverse sensitivities of the individual gage elements in
the rosettes are equal to one another, or approximately so.
Generalized correction equations for any combination of
transverse sensitivities are given in the Appendix.
Consider first the two-gage 90-degree rosette, with the
gage axes aligned with two orthogonal axes, x and y, on
the test surface. When using this type of rosette, the x
and y axes would ordinarily be the principal axes, but this
need not necessarily be so. The correct strains along any
two perpendicular axes can always be calculated from the
following equations in terms of the indicated strains along
those axes:

x
t x t y
t
K K
K
=
( )
( )
1
1
0
2

(6)

y
t y t x
t
K K
K
=
( )
( )
1
1
0
2

(7)
where:

x a
=
1
= the indicated (uncorrected) strain from
gage no. 1.

y a
=
2
= the indicated (uncorrected) strain from
gage no. 2.

x y
,
= corrected strains along the x and y axes,
respectively.
The (1K
t
2
) term in the denominators of Equations (6)
and (7) is generally in excess of 0.995, and can be taken as
unity:

x t x t y
K K = ( )
( )
1
0

(6a)

y t y t x
K K = ( )
( )
1
0

(7a)
Data reduction can be further simplied by setting the gage
factor control on the strain-indicating instrumentation at F
a

instead of F, the manufacturers gage factor. Since,

F
F
K
a
=
1
0 1
Equations (6a) and (7a) can be rewritten:

x x t y
y y t x
K
K
=
=

(6b)
(7b)
where:

x y
= strains as indicated by
instrumentation with
gage factor control set at
Figure 1
t
/
a
= 5
5 =
t
/
a
* For substitution into any equation in this Tech Note, K
t
must always
be expressed decimally. Thus, the value of K
t
(in percent) from the
gage package data sheet must be divided by 100 for conversion to its
decimal equivalent.
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
94

F
K
t
1
0
As an alternative to the preceding methods, a quick graphical

correction for the transverse sensitivity can be made through
the use of Figure 2. To use the graph, the rst step is to
calculate:

t
a
y
x
t
= =
1
2
1
(Gage No. 1)
a
x
y
= =
2
1
2
(Gage No. 2)
Having done this, it is only necessary to enter the graph
at the approximate value of K
t
, move upward to the line
(or interpolated line) representing the observed (indicated)
strain ratio, /
t a
( ) for that particular rosette element, and
horizontally to the vertical scale on the left to read the
correction factor.
Then,

x
C = =
1 1 1
Similarly,

y
C = =
2 2 2
Following is a numerical example utilizing rst Equations

(6a) and (7a), and then Figure 2.
Assume that the indicated strains for rosette elements (1)
and (2) along the x and y axes are, respectively:

1
2
1530
920
= +
= +
Assume also that K
t
= 0.06. Substituting into Equations
(6a) and (7a), with
0
= 0.285,

x
= (1 + 0.285 x 0.06) (1530 + 0.06 x 920) = 1612

y
= (1 + 0.285 x 0.06) (920 + 0.06 x 1530) = 1029
For use with the correction graph, Figure 2,

. .
t
a
t
a
= =
1
920
1530
0 601 0 6
22
1530
920
1 663 1 65 = = . .
Following the line for K
t
= 0.06 upward, interpolating
the location of
/
t a
( )
1
= 0.6, and
/
t a
( )
2
= 1.65, and
reading the respective values of the correction factor,
C
1
= 1.06; C
2
= 1.12
From which,

x x
y y
C
C
= = =
= = =
1
2
1 06 1530 1620
1 12 920
.
.
x
x 11030
Correction For Shear Strain
A two-gage, 90-degree rosette, or T-rosette, is sometimes
used for the direct indication of shear strain. It can be shown
that the shear strain along the bisector of the gage axes, is,
in this case, numerically equal to the difference in normal
strains on these axes. Thus, when the two gage elements of
the rosette are connected in adjacent arms of a Wheatstone
bridge, the indicated strain is equal to the indicated shear
strain along the bisector, requiring at most correction for
Figure 2
/
t a
5 =
= 5 /
t a
K
t
IN%
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
95
the error due to transverse sensitivity. The latter error can
be corrected for very easily if both gages have the same
transverse sensitivity, since the error is independent of the
state of strain. The correction factor for this case is:
C
K
K
t
t
1
1
0

(8)
The actual shear strain is obtained by multiplying the
indicated shear strain by the correction factor. Thus,

= =
( )
=

( )
C C
K
K
x y
t
t
x y

1
1
0
For convenience, the shear strain correction factor is plotted
in Figure 3 against K
t
, with
0
= 0.285. Since this correction
factor is independent of the state of strain, it can again
be incorporated in the gage factor setting on the strain-
indicating instrumentation if desired. This can be done by
setting the gage factor control at:

F F
K
K
t
t
1
1
0
(9)
With this change, the strain indicator will indicate the actual
shear strain along the bisector of the gage axis, already
corrected for transverse sensitivity in the strain gages.
Three-Gage Rectangular (45) Rosette
When the directions of the principal axes are unknown,
three independent strain measurements are required to
completely determine the state of strain. For this purpose,
a three-gage rosette should be used, and the rectangular
rosette is generally the most convenient form.
If the transverse sensitivity of the gage elements in the
rosette is other than zero, the individual strain readings will
be in error, and the principal strains and stresses calculated
from these data will also be incorrect.
Correction for the effects of transverse sensitivity can be
made either on the individual strain readings or on the
principal strains or principal stresses calculated from these.
Numbering the gage elements consecutively, elements (1)
and (3) correspond directly to the two-gage, 90-degree
rosette, and correction can be made with Equations (6) and
(7), or (6a) and (7a), or (by properly setting the gage factor
control on the strain indicator) with Equations (6b) and (7b).
The center gage of the rosette requires a special correction
relationship since there is no direct measurement of the
strain perpendicular to the grid. The correction equations
for all three gages are listed here for convenience:

1
0
2
1 3
2
0
2
2
1
1
1
1
=

( )
=

K
K
K
K
K
t
t
t
t
t

KK
K
K
K
t
t
t
t

1 3 2
3
0
2
3 1
1
1
+ ( )

[[ ]

(10)

(11)

(12)

where:

, ,
1 2 3
= indicated strains from the respective gage
elements.

1 2 3
, ,
= corrected strains along the gage axes.
It should be noted that Equations (10), (11), and (12) are
based upon the assumption that the transverse sensitivity
is the same, or effectively so in all gage elements, as it is in
stacked rosettes. This may not be true for planar foil rosettes,
Figure 3
C
K
K
t
t
=
1
1
0 285
0
0
.
K
t
IN%
C
o
r
r
e
c
t
i
o
n

F
a
c
t
o
r

C
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
96
since the individual gage elements do not all have the same
orientation with respect to the direction in which the foil was
rolled. It is common practice, however, to etch the rosette in
a position of symmetry about the foil rolling direction, and
therefore the transverse sensitivities of gage elements (1) and
(3) will be nominally the same, while that of element (2) may
differ. Correction equations for rosettes with nonuniform
transverse sensitivities among the gage elements are given in
the Appendix.
Delta Rosettes
A delta strain gage rosette consists of three gage elements
in the form of an equilateral triangle or a Y with equally
spaced branches. The delta rosette offers a very slight
potential advantage over the three-gage rectangular rosette in
that the lowest possible sum of the strain readings obtainable
in a particular strain eld is somewhat higher than for a
three-gage rectangular rosette. This is because the three gage
elements in the delta rosette are at the greatest possible angle
from one another. However, the data reduction for obtaining
the principal strains or correcting for transverse sensitivity
is also more involved and lengthy than for rectangular
rosettes.
As in the case of rectangular rosettes, plane foil delta rosettes
are manufactured symmetrically with respect to the rolling
direction of the foil. Thus, two of the gage elements will
ordinarily have the same nominal transverse sensitivity, and
third may differ. Correction equations for this condition are
given in the Appendix. In the stacked delta rosette, all three
gages have the same nominal sensitivity.
The individual strain readings from a delta rosette can
be corrected for transverse sensitivity with the following
relationships when a single value of K
t
can be used for the
transverse sensitivity:

1 2 1 2 3
1
1
1
3
2
3
0
+
j
(
,
\
,
(
+ ( )
, K
K
K
K
t
t
t
t

,
]
]
]
+
j
(
,
\
,
(
+

2 2 2 3
1
1
1
3
2
3
0
K
K
K
K
t
t
t
t

1
3 2 3
1
1
1
3
2
3
0
( )
,
,
]
]
]
+
j
(
,
\
,
(
K
K
K
t
t
t
KK
t

1 2
+ ( )
,
,
]
]
]

(13)

(14)

(15)

As before, simplification can be achieved by treating
(1 K
t
2
) as unity, and by incorporating the quantity
(1
0
K
t
) into the gage factor setting for the strain
instrumentation. When doing this, the gage-factor control
is set at:

F
F
K
a
t
=
1
0
Correction of Principal Strains

With any rosette, rectangular, delta, or otherwise, it is
always possible (and often most convenient) to calculate
the indicated principal strains directly from the completely
uncorrected gage readings, and then apply corrections to
the principal strains. This is true because of the fact that
the errors in principal strains due to transverse sensitivity
are independent of the kind of rosette employed, as long
as all gage elements in the rosette have the same nominal
transverse sensitivity. Since Equations (6) and (7) apply to
any two indicated orthogonal strains, they must also apply
to the indicated principal strains. Thus, if the indicated
principal strains have been calculated from strain readings
uncorrected for transverse sensitivity, the actual principal
strains can readily be calculated from the following:

(16)

(17)

p
t
t
p t q
q
t
t
q
K
K
K
K
K
=

( )
=

1
1
1
1
0
0
2
2

KK
t p

( )

Furthermore, Equations (16) and (17) can be rewritten to
express the actual principal strain in terms of the indicated
principal strain and a correction factor. Thus,

(18)

(19)

p p
t
t
t
q
p
K
K
K

j
(
,
\
,
(

j
(
,
\
,
(
,
,
,
1
1
1
0
2
]]
]
]
]
j
(
,
\
,
(

j
(
,
\
,

q q
t
t
t
p
q
K
K
K

1
1
1
0
2
((
,
,
,
]
]
]
]
Since Equations (18) and (19) are the same relationship used
to plot the correction graph of Figure 2, this graph can be
used directly to correct indicated principal strains by the
procedure described earlier, merely noting that:

t
a
q
p
= when correcting
p
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
97
and

t
a
p
q
= when correcting
q
In fact, the indicated strains from three gages with any
relative angular orientation dene an indicated Mohrs
circle of strain. When employing a data-reduction scheme
that produces the distance to the center of Mohrs circle
of strain, and the radius of the circle, still another simple
correction method is applicable. To correct the indicated
Mohrs circle to the actual Mohrs circle, the distance to
the center of the indicated circle should be multiplied by
(1
0
K
t
)/(1 + K
t
), and the radius of the circle by
(1
0
K
t
)/(1 - K
t
). The maximum and minimum principal
strains are the sum and difference, respectively, of the
distance to the center and the radius of Mohrs circle
of strain.
Bibliography
ASTM Standard E251, Part III. Standard Test Method
for Performance Characteristics of Bonded Resistance
Strain Gages.
Avril, J. LEffet Latral des Jauges lectriques. GAMAC
Conference. April 25, 1967.
Baumberger, R. and F. Hines. Practical Reduction
Formulas for Use on Bonded Wire Strain Gages in Two-
Dimensional Stress Fields. Proceedings of the Society for
Experimental Stress Analysis II: No. 1, 113-127, 1944.
Bossart, K. J. and G. A. Brewer. A Graphical Method
of Rosette Analysis. Proceedings of the Society for
Experimental Stress Analysis IV: No. 1, 1-8, 1946.
Campbell, W, R, Performance Tests of Wire Strain Gages:
IV Axial and Transverse Sensitivities. NACA TN1042,
1946.
Gu, W. M. A Simplied Method for Elminating Error
of Transverse Sensitivity of Strain Gage. Experimental
Mechanics 22: No. 1 16-18, January 1982.
Meier, J.H. The Effect of Transverse Sensitivity of SR-4
Gages Used as Rosettes. Handbook of Experimental Stress
Analysis, ed. by M. Hetnri, John Wiley & Sons, pp. 407-
411, 1950.
Meier, J. H. On the Transverse-strain Sensitivity of Foil
Gages. Experimental Mechanics 1: 39-40, July 1961.
Meyer, M.L. A Unified Rational Analysis for Gauge
Factor and Cross-Sensitivity of Electric-Resistance Strain
Gauges. Journal of Strain Analysis 2: No. 4, 324-331, 1967.
Meyer, M. L. A Simple Estimate for the Effect of Cross
Sensitivity on Evaluated Strain-gage Measurement.
Experimental Mechanics 7: 476-480, November 1967.
Murray, W.M. and P. K. Stein. Strain Gage Techniques.
Massachusettes Institute of Technology, Cambridge,
Massachusetts, pp. 56-81, 1959.
Nasudevan, M. Note on the Effect of Cross-Sensitivity
in the Determination of Stress. STRAIN 7: No. 2, 74-75,
April 1971.
Starr, J.E. Some Untold Chapters in the Story of the
Metal Film Strain Gages. Strain Gage Readings 3: No. 5,
31, December 1960 January 1961.
Wu, Charles T. Transverse Sensitivity of Bonded Strain
Gages. Experimental Mechanics 2: 338-344, November
1962.
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
98
APPENDIX
The following relationships can be used to correct for transverse sensitivity when the gage elements in a rosette do not all
have the same value of K
t
. In each case,
0
is the Poissons ratio of the material on which the manufacturers gage factor was
measured (usually 0.285).
Two-Gage, 90-Degree rosette

1
1 2
2
2
1 1
1
0 0
1 1
1 2
=

( )

( )
K K K
K K
t t t
t t

2 1
1
1 1
1
0 0
2 2
1 2
( )

( )
K K K
K K
t t t
t t

where:

,
1 2 = indicated strains from gages (1) and (2),
uncorrected for transverse sensitivity.
K
t
1
, K
t
2
= transverse sensitivities of gages (1) and (2).

1 2
, = actual strains along gage axes (1) and (2).
Three-Gage Rectangular (45-Degree) Rosette

1
1 3
3
2
1 1
1
0 0
1 1
1 3
=

( )

( )
K K K
K K
t t t
t t

2 1
1 3
1
1
1 1
0 0
2
2
2
( )

( )

( )
+ K
K
K K K
t
t
t t t

33
3 1
3
1 1
1 1
0
1 3 2
( )

( )
( )

( )
=
K K
K K K
t t
t t t

3 1
1
1 1
1
0 0
3 3
1 3
( )

( )
K K K
K K
t t t
t t
When the transverse sensitivities of the orthogonal gages (1) and (3) are nominally the same, let
K
t
1

= K
t
3
= K
t
13

(20)

(21)
(22)
(23)
(24)
T
E
C
H

N
O
T
E
TN-509
Micro-Measurements
99
Then:

1
13
13
2 1
13
3
2
1
1
1
0
0
2
=

( )
=

( )
K
K
K
K
t
t
t
t

11 1
1
13
2
13
1 3
2
13
0
+
( )

( )
+ ( )
+
( )
K K K
K
t t t
t

11
1
1
2
3
13
13
2 3
13
1
0
( )
=

( )
K
K
K
K
t
t
t
t

where:

, ,
1 2 3
= indicated strains from gages (1), (2), and (3), uncorrected for transverse sensitivity.
K
t
1
,
K
t
2
, K
t
3
= transverse sensitivities of gages (1), (2), and (3).
K
t
13
= transverse sensitivity of orthogonal gages (1) and (3).

1 2 3
, ,
= actual strains along gage axes (1), (2), and (3).
Delta Rosette

1
1
1 2 3 2 3 1
2
1 3 2 1
0
=

( )

( )
K K K K K K
t t t t t t

0 0
2 3
3
3 2
1
1 1 1
3
K K K K
K
t t t t
t
( )

( )
+
( )

( )
KK K K K K K K K K K K
t t t t t t t t t t t
2 3 1 2 2 3 1 3 1 2 3
3 +
22
2
2 3 1 3 1 2
3
1 3 2 1
0
=

( )

( )
K K K K K K
t t t t t t

0 0
3 1
1
1 3
1
1 1 1
3
K K K K
K
t t t t
t
( )

( )
+
( )

( )
2 3 1 2 2 3 1 3 1 2 3
3 +
33
3
3 1 2 1 2 3
1
1 3 2 1
0
=

( )

( )
K K K K K K
t t t t t t

0 0
1 2
2
2 1
1
1 1 1
3
K K K K
K
t t t t
t
( )

( )
+
( )

( )
2 3 1 2 2 3 1 3 1 2 3
3 +
When two of the gages, for example, (1) and (3), have the same nominal transverse sensitivity,

1
1
13 2 13 13 2 13
1 3 2
0
=

( )

( )
K K K K K K
t
t t t t t

2
2 13
3
13 2
1 1 1 1
0 0
( )

( )
+
( )

( )
K K K K
t t t t
+
=
3 2 2 3
13 13
2
2
2
13 2 13 2
2
K K K K K K K
t t t t t t t

2
2 13 2
1 3
13
1 3 2 1
0 0
( )
+
( )
+ ( )
(
K K K K
t t t t

))
+
=

( )

K K K K
K
t t t t
t
13 13 2 2
3
3
13
3 3
1 3
0
KK K K K K K K
t t t t t t t
2 13 13 2 13
1
13 2
2 1 1
0

( )

( )

(( )
+
( )

( )

2
2 13
2
2
2
1 1
3
0
13 13
K K
K K K
t t
t t t
+ 2 2 3
13 2 13 2
K K K K
t t t t
The subscripts in Equations (28) through (33) have the same signicance as in Equations (22) through (27), except that the two
gages with common transverse sensitivity, K
t
13
, are not orthogonal.
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
(33)
Tech Note TN-510-1
MICRO-MEASUREMENTS
Design Considerations for
Diaphragm Pressure Transducers
T
E
C
H

N
O
T
E
101
The following notes are intended only as general guidance
for the prel i mi nary design of diaphragm pressure
transducers. The actual design and development process
involves arriving at the best compromise (relative to
the performance specications) of sensitivity, linearity,
and frequency response, as determined primarily by the
diaphragm diameter and thickness.
The formulas included here are based upon the following
assumptions:
Uniform diaphragm thickness
Small defections
Infnitely rigid clamping around the diaphragm
periphery
Perfectly elastic behavior
Negligible stiffening and mass effects due to the
presence of the strain gage on the diaphragm.
To the degree that the actual transducer fails to satisfy all
of the above assumptions, the formulas will be inaccurate.
Because of this, the formulas should be used only in the
initial stages of transducer development to determine the
approximate proportions of the transducer.
Sensitivity
The strain distribution in a rigidly clamped diaphragm
under uniform pressure distribution is shown in Figure 1.
The radial and tangential strains at the center of the
diaphragm are identical, and expressed by:

Rc T
o
c
PR
t E
= =
( )
3 1
8
2 2
2

(1)

where:
U.S.
Customary
Units
Metric
(Si)
Units
P = Pressure psi Pa
R
o
= Diaphragm Radius in mm
t = Diaphragm thickness in mm
= Poissons ratio dimensionless
E = Modulus of elasticity psi Pa

The radial strain decreases rapidly as the radius increases,
becoming negative, and (at the edge) equal to twice the
Tangential Strain
Radial Strain
t
R
o

R T
o
PR
t E
= =

( )
3 1
8
2 2
2

R
o
o
PR
t E
=

( )
3 1
4
2 2
2
Fig.1 Strain distribution in clamped diaphragm
T
E
C
H

N
O
T
E
TN-510-1
Micro-Measurements
102
Design Considerations for Diaphragm Pressure Transducers
center strain. The tangential strain decreases to zero at the
periphery of the diaphragm. Thus,

R
o
o
PR
t E
=
( )
3 1
4
2 2
2

(2)

T
o
=
0

(3)
Diaphragm Strain Gages
Micro-Measurements manufactures two different strain
gage congurations (Figure 2a and b) for use on diaphragm
pressure transducers.
(a) Circular Pattern
(b) Linear Pattern
Fig. 2 Micro-Measurements
diaphragm strain gages for pressure transducers.
The traditional circular pattern as shown in Figure 2a has
been designed to take advantage of the orientation of the
tangential and radial strain elds described above. Taking
account of the sign difference in the strains sensed by
the radial and tangential grid elements, and dividing the
elements into symmetrical pairs, permits incorporating a
full bridge into a single strain gage. In terms of optimizing
the strain gage design, the solder tabs have been located in
a region of low strain.
The linear gage conguration shown in Figure 2b functions
in the same manner as the circular version with only minor
differences in total gage output (eo)
1
. The main advantages
of using a linear design are ease of installation (less surface
area to bond) and generally lower gage cost.
For either the circular or linear patterns, averaging the
strain over the region covered by each sensing element
(assuming a gage factor of 2.0), and averaging the outputs
of all sensing elements, the total gage output (e
o
) in
millivolts per volt can be expressed approximately by the
following formula:

e
PR
t E
o
o
= 0 75
1
10
2 2
2
3
.
( )
,
x mV/V

(4)
Linearity
The preceding equations for diaphragm strain and output
indicate that the output is proportional to the applied
pressure. This precise linearity applies, however, only
for vanishingly small def lections. In the case of finite
defections, the diaphragm pressure transducer is inherently
nonlinear, and becomes more so, as defection increases.
As a general rule, the defection of the diaphragm at the
center must be no greater than the diaphragm thickness;
and, for linearity in the order of 0.3%, should be limited to
one quarter the diaphragm thickness.
Following is the formula for diaphragm defection, based
upon small-defection theory:

Y
PR
t E
c
o
=
3 1
16
4 2
3
( )

(5)
where: Y
c
= Center defection, in (mm)
Frequency Response
In order to faithfully respond to dynamic pressures, the
resonant frequency of the diaphragm must be considerably
higher than the highest applied frequency. Depending
strongly upon the degree of damping in the diaphragm
strain gage assembly and in the fuid in contact with the
diaphragm, the resonant frequency should be at least three
to ve times as high as the highest applied frequency. The
subject of proper design for accurate dynamic response is
too complex and extensive to be included here. However,
for transducers subject to high frequencies or to sharp
pressure wave fronts involving high-frequency components,
careful consideration must be given to frequency response,
both in terms of amplitude and phase-shift.
T
E
C
H

N
O
T
E
TN-510-1
Micro-Measurements
103
For reference purposes only, and subject to the assumptions
listed earlier, the undamped resonant frequency of a
rigidly clamped diaphragm can be expressed using U.S.
Customary Units as follows:

f
t
R
gE
n
o
=
0 469
1
2 2
.
( )
, Hz

(6)
where: g= Acceleration of gravity, 386.4 in/sec
2
= Weight density, lbs/in
3
Since in the metric system (SI), density is derived without
the effect of gravity, Equation (6) must be slightly modied
when using SI Units as follows:

f
t
R
E
n
o
=
0 469
1
2 2
.
( )
, Hz

(7)
where: = Mass density, g/cm
3
Construction
For maximum accuracy and minimum hysteresis, it is
common practice to design pressure transducers so that
the diaphragm is an integral part of the transducer body
(Figure 3).
Fig. 3 Typical diaphragm arrangement
for pressure transducer.
Nominal Diaphragm Diameter
Alternate
Strain Gage
Locations
It is neither necessary nor desirable to try to machine the
body of the transducer to a sharp internal corner at the
junction with the diaphragm. The presence of the llet
radius, however, is merely one of the ways in which practical
transducer construction differs from the idealized concept
corresponding to the earlier assumptions and the equations
given here. Because of this and the other differences, the
transducer behavior will necessarily differ from the ideal;
and experimental development will obviously be required
to optimize the performance of a particular transducer.
Wiring
As shown in Figure 4a, the internal circuit of the circular
pattern strain gage has two adjacent corners of the full
bridge left open. The open bridge corners are left for the
introduction of zero-shift versus temperature correction,
and subsequent restoration of zero balance. The linear
pattern (Figure 4b) has a sl ightly di fferent ci rcuit
arrangement but the purpose is the same.
NOTE: See Micro-Measurements Transducer-Class Strain
Gages Data Book for circular pattern strain gages, and
linear pattern strain gages for pressure transducers.
References
1. For a more detailed analysis of circular versus linear
diaphragm strain gages, request a copy of Inuence
of Grid Geometry on the Output of Strain-Gage-Based
Diaphragm Pressure Transducers by R.B. Watson
(available from Micro-Measurements).
(b) Linear Pattern
Fig. 4 Internal circuit of Micro-Measurements strain gages
for diaphragm pressure transducers.
2 1
6
5
3
4
T

R
R

T
2 1
6 3
4
T

R
R

T
5
(a) Circular Pattern
T
E
C
H

N
O
T
E
TN-510-1
Micro-Measurements
104
Assume that a diaphragm pressure transducer is to be designed for a maximum rated pressure of 1000 psi [6.89 MPa],
under which pressure the output (e
o
) from a steel diaphragm should be 2 mV/V. If the diaphragm diameter is to be 0.670 in
[17.02 mm], nd the following:
(a) Diaphragm thickness (c) Resonant Frequency
(b) Center defection (d) Approximate maximum diaphragm strain level
CONSTANTS*
U.S. Customary Metric (SI)
P = 1000 lbs/in
2
= 0.283 lbs/in
3
P = 6.89 MPa R
o
= 8.51 mm = 8.51 x 10
-3
m
E = 30 x 10
6
psi g= 386.4 in/sec
2
= 0.285 e
o
= 2 mV/V = 2 x 10
-3
V/V
R
o
= 0.335 in e
o
= 2 mV/V = 2 x 10
-3
V/V E = 207 GPa = 7.83 g/cm
3
= 7.83 x 10
3
kg/m
3

= 0.285
(a) From Eq. (4), solve for t,
with e
o
in units of V/V
t = 0.036 in t = 9.11 x 10
-4
m = 0.911 mm
(b) From Eq. (5),
Y
c
= 0.0016 in Y
c
= 3.98 x 10
-5
m = 0.0398 mm
(c) From Eq. (6), From Eq. (7),
f
n
= 31 766 Hz f
n
= 31 647 Hz
NUMERICAL EXAMPLE - U.S. Customary and Metric (SI) Units
t
PR
e E
o
o
=
( )
0.75
2
1
2

t =
( ) ( )
( )
0.75 1000 0.335 1 0.285

2 2
3
2 10 330 10
6
( )
t =

( )

( )
( )
0.75 6 89 10 8 51 10 1 0 285
6 3
2
2
. . .
( )

( )
2 10 207 10
3 9
Y
PR
t E
c
o
=
( )
3 1
16
4 2
3

Y
c
=
( ) ( )
( )
3 1000 0 335 1 0 285
16 0 036
4 2
3
. .
.
( )
30 10
6
Y
c
=

( )

( )
( )
3 6 89 10 8 51 10 1 0 285
1
6 3
4
2
. . .
66 9 11 10 207 10
4
3
9

( )

( )
.
f
t
R
gE
n
=
( )
0 469
1
0
2 2
.

f
t
R
gE
n
=
( )
0 469
1
0
2 2
.

f
n
=

( )

( )
0 469 0 036
0 335
386 4 30 10
0 283 1
2
6
. .
.
.
. 00 285
2
. ( )
( )
f
n
=

( )
( )
0 469 9 11 10
8 51 10
207 10
7 83
4
3
2
9
. .
.
. ( )
10 1 0 285
3 2
.
T
E
C
H

N
O
T
E
TN-510-1
Micro-Measurements
105
(d) From Eq. (2),

R
0
= 1989 in/in
R
0
= 2001 m/m
*The small differences occurring in comparable U.S. Customary and Metric results arise from rounding numbers in both sets of calculations.

R
o
o
PR
t E
=
( )
3 1
4
2 2
2
R
o
=
( ) ( )
( )
3 1000 0 335 1 0 285
4 0 036
2 2
. .
.
22 6
30 10
( )

R
o
=

( )

( )
( )
3 6 89 10 8 51 10 1 0 285
6 3
2
2
. . .

( )

( )
4 9 11 10 207 10
4
2
9
.
NUMERICAL EXAMPLE - U.S. Customary and Metric (SI) Units
Tech Note TN-511
MICRO-MEASUREMENTS
Errors Due to Misalignment of Strain Gages
T
E
C
H

N
O
T
E
107
Single Gage in a
Uniform Biaxial Strain Field
When a gage is bonded to a test surface at a small
angular error with respect to the intended axis of strain
measurement, the indicated strain will also be in error
due to the gage misalignment. In general, for a single gage
in a uniform biaxial strain eld, the magnitude of the
misalignment error depends upon three factors (ignoring
transverse sensitivity):
1. The ratio of the algebraic maximum to the algebraic
minimum principal strain,
p
/
q
.
2. The angle between the maximum principal strain
axis and the intended axis of strain measurements.
3. The angular mounting error, , between the gage
axis after bonding and the intended axis of strain
measurement.
These quantities are dened in Figures 1 and 2 for the
particular but common case of the uniaxial stress eld.
Figure 1 is a polar diagram of strain at the point in
question, and Figure 2 gives the concentric Mohrs circles
for stress and strain for the same point. In Figure 1, the
distance to the boundary of the diagram along any radial
line is proportional to the normal strain along the same
line. The small lobes along the Y axis in the diagram
represent the negative Poisson strain for this case. It can
be seen qualitatively from Figure 1 that when is 0 or
90, a small angular misalignment of the gage will produce
a very small error in the strain indication, since the polar
strain diagram is relatively at and passing through zero-
slope at these points.
However, for angles between 0 and 90, Figure 1 shows
that the error in indicated strain due to a small angular
misalignment can be surprisingly large because the slope
of the polar strain diagram is very steep in these regions.
More specically, it can be noticed from Figure 2 (on page
108), when = 45, or 2 = 90, that the same small angular
misalignment will produce the maximum error in indicated
strain, since is changing most rapidly with angle at this
point. The same result could be obtained by writing the
analytical expression for the polar strain diagram, and
setting the second derivative equal to zero to solve for the
angle at which the maximum slope occurs. In fact, the
general statement can be made that in any uniform biaxial
strain eld the error due to gage misalignment is always
greatest when measuring strain at 45 to a principal axis,
and is always least when measuring the principal strains.*
The error in strain indication due to angular misalignment
Figure 1 Polar Strain distribution
corresponding to uniaxial stress,
illustrating the error in indicated strain
when a gage is misaligned by
from the intended angle .
* The exception to this statement is the singular case when
p

q
, as
on the surface of a pressurized sphere. In this instance, the strain is
everywhere the same and independent of directions.
T
E
C
H

N
O
T
E
TN-511
Micro-Measurements
108
of the gage can be expressed as follows:

n =
( )

(1)
where: n = Error,

= Strain along axis of intended measurement

at angle from principal axis,

()
= Strain along gage axis with angular
mounting error of ,
Or,

n
p q
=

( )

2
2 2 cos cos

(2)
where:

p
,
q
= Maximum and minimum principal strains,
respectively
The error can also be expressed as a percentage of the
intended strain measurement,
:

=

( )
n

100

(3)

=
( )
+
+
n
R
R
cos cos
cos
2 2
1
1
2

100

(4)
where: R
p
q
=
However, from Equation (3) it can be seen that n becomes
unmeaningfully large for small values of
, and innite
when
vanishes. In order to better illustrate the order of

magnitude of the error due to gage misalignment, Equation
(2) will be evaluated for a more-or-less typical case.
In a uniaxial stress field,
q
=
p
. And, for steel,
= 0.285.
Assume
p
= 1000
Then,
q
= 285
And, n = 642.5 [cos2( ) cos2] (5)
Figure 2 Mohrs circles of stress and strain
for uniaxial stress, an alternative representation
of the misalignment errors.
Figure 3 Error in indicated strain due to gage
misalignment for the special case of a uniaxial stress
eld in steel.
p
= 1000,
q
= 285.
STRAIN
CIRCLE
STRESS
CIRCLE
T
E
C
H

N
O
T
E
TN-511
Micro-Measurements
109
Equation (5) is plotted in Figure 3 over a range of from 0
to 90, and over a range of mounting errors from 1 to 10.
In order to correct for a known misalignment by reading
the value of n from Figure 3, it is only necessary to solve
Equation (1) for
and substitute the value of n; including

the sign as given by Figure 3. This gure is given only as an
example, and applies only to the case in which
q
= 0.285
p
(uniaxial stress in steel). Equation (2) can be used to
develop similar error curves for any biaxial strain state.
Two-Gage Rectangular Rosette
While the above analysis of the errors due to misalignment
of a single gage may help in understanding the nature
of such errors, the 90-degree, two-gage rosette is of
considerably greater practical interest.
A two-gage rectangular rosette is ordinarily used by stress
analysts for the purpose of determining the principal
stresses when the direction of the principal axes are known
from other sources. In this case, the rosette should be
bonded in place with the gage axes coincident with the
principal axes. Whether there is an error in orientation
of the rosette with respect to the principal axes, or in the
locations of the principal axes themselves, there will be a
corresponding error in the principal stresses as calculated
from the strain readings.
In Figure 4, a general biaxial strain eld is shown, with
the axes of a two-gage rosette, misaligned by the angle
, superimposed. The percentage errors in the principal
stresses and maximum shear stress due to the misalignment
are:

n
p
p p
p
=

x 100

(6)

n
R v
R v
p
=
( ) ( ) ( )
+ ( )
1 1 1 2
2
cos
x 100

(7)

n
q
q q
q
=

x 100

(8)

n
R v
vR
q
=
( ) ( ) ( )
+ ( )
1 1 1 2
2 1
cos
x 100

(9)

n
MAX MAX
MAX
MAX
x 100 =

(10)

n

MAX
x 100 = ( ) cos 1 2

(11)
where:

, ,
p q MAX
are the principal stresses and
maximum shear stress inferred from the indicated
strains when the rosette is misaligned by the angle .
R

=
p
/
q
, the ratio of the algebraic maximum to
the algebraic minimum principal strain, as before.
Figure 4 Biaxial strain eld with rosette axes misaligned by the angle from the principal axes.
T
E
C
H

N
O
T
E
TN-511
Micro-Measurements
110
When the principal strain ratio is replaced by the principal
stress ratio, where:

R
R
R
=
+
+ 1
=
p
q

(12)
Or,

R
R
R
1
=
p
q

(13)

n
R
R
P
=

( )
1
2
1 - cos2 x 100

(14)

n
R
q

=

( )
1
2
1 - cos2 x 100

(15)

Equations (11), (14), and (15) will now be applied to an
example in order to demonstrate the magnitudes of the
errors encountered.
Consider rst a thin-walled cylindrical pressure vessel. In
this case, the hoop stress or circumferential stress is twice
the longitudinal stress, and of the same sign.
Thus,

p
q
R = = 2
And Equations (11), (14), and (15) become:

n
MAX
= ( ) cos 1 1 2 x 100

(11a)

n
P
= ( ) / cos 1 4 1 2 x 100

(14a)

n
q
= ( ) / cos 1 2 1 2 x 100

(15a)
Equations (11a), (14a), and (15a) are plotted in Figure 5.
From the gure, it can be seen that the errors introduced by
rosette misalignment in this instance are quite small. For
example, with a 5 mounting error,
MAX
,
p
, and
q
are in
error by only 1.5%, 0.38%, and 0.75%, respectively.
In order to correct for a known misalignment by reading
the value of n from Figure 5, or any similar graph derived
from the basic error equations [Equations (7), (9), (11), (14),
(15)], it is only necessary to solve Equations (6), (8), and (10)
for
p
,
q
, and
MAX
, respectively, and substitute the value
of n from Figure 5, including the sign. That is,

p
p
n
p
=
+
1
100

(16

q
q
n
q
=
+
1
100

(17)

MAX
MAX
M
=
+
1
100
n
AX

(18)
Figure 5 Percentage erros in principal stresses
and maximum shear stress for a biaxial
stress eld with
p
/
q
= 2.0.
8
7
6
5
4
3
2
1
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5 6 7 8 9 10
S
T
R
E
S
S

E
R
R
O
R

%
ROSETTE MOUNTING ERROR
n
q
n
p
n
MAX
+
p
/
q
= 2.0
T
E
C
H

N
O
T
E
TN-511
Micro-Measurements
111
where:

p = maximum principal stress as calculated
from gage readings

q = minimum principal stress as calculated
from gage readings

MAX = maximum shear stress as calculated from

MAX
=

P q
2
While the errors in the above case were very small, this
is not true for stress elds involving extremes of R. In
general, n
p
becomes very large for |R|<<1.0, as does n
q

for |R|>>1.0. The error in shear stress is independent of
the stress state.
The above generalities can be demonstrated by extending
the previous case of the pressurized cylinder. Consider an
internally pressurized cylinder with an axial compressive
load applied externally to the ends. If, for example, the load
were 0.8 r
2
p, where r is the inside radius of the cylinder,
and p is the internal pressure, the principal stress ratio
would become,
R =10
Equations (14) and (15) become:
n
p
= 0.45(1 cos2) x 100 (14b)

n
q
= 4.5(1 cos2) x 100 (15b)
For this case, a 5 error in mounting the rosette produces a
0.68% error in
p
and a 6.75% error in
q
.
The errors dened and evaluated in the foregoing occur,
in each case, due to misalignment of a single strain gage or
of an entire rosette. The effect of misalignment among the
individual gages within a rosette is the subject of a separate
study.
Tech Note TN-512-1
MICRO-MEASUREMENTS
Plane-Shear Measurement with Strain Gages
T
E
C
H

N
O
T
E
113
Revision 12-Sep-2010
Introduction
Loading a specimen as shown in Figure 1a produces shear
stresses in the material. An initially square element of the
material, having vertical sides parallel to the direction of
loading, is distorted by these stresses into a diamond shape
as illustrated in Figure 1b, where the distortion is greatly
exaggerated for pictorial clarity. Shear strain is dened as
the magnitude of the change in the initial right angle of the
element at the X-Y origin. That is,

=
2

(1)
Since shear strain is a change in angle, its natural units are
radians, although it can also be expressed in terms of in/in
[m/m] and percent. From Equation (1), the sign of the shear
strain is positive when the initial right angle of the element
is reduced ( < / 2). Reversing the directions (sign) of the
shear stresses in Figure 1 causes the initial right angle to
increase and results in a negative shear strain.
Normal strains cause dimensional changes in the grid
of a strain gage, changing its electrical resistance. Pure
shear strains merely rotate the grid, and do not cause the
elongation or contraction necessary to vary the resistance.
Fortunately, shear and normal strains are related through
mechanics principles, allowing strain gages to provide a
direct indication of shear strain. Properly orienting gages
on a strained surface and properly connecting them in a
Wheatstone bridge circuit yields an instrument indication
that is directly proportional to the surface shear strain.
This Tech Note rst develops an expression for determining
the surface shear strain in any given direction from two
normal-strain measurements. Next is a discussion of strain
gage and Wheatstone bridge arrangements for direct
indication of shear strains. The shear-strain magnitude
varies sinusoidally around a point in a biaxial strain eld.
Since the maximum strain values are usually of primary
interest in stress analysis, Mohrs strain circle is used to
obtain an expression for the maximum shear strain at a
point in a biaxial strain eld. Practical examples of shear
measurement with strain gages are given for both isotropic
materials (e.g., metals) and orthotropic materials such as
wood and ber-reinforced composites.
Shear Strain from Normal Strains
Consider an array of two strain gages oriented at arbitrarily
different angles with respect to an X-Y coordinate system
which, in turn, is arbitrarily oriented with respect to
the principal axes, as in Figure 2 (following page). From
elementary mechanics of materials, the strains along the
gage axes can be written as:
(b)
Y
X
(a)
Figure 1 (a) Shear loads applied to specimen;
(b) enlarged shear deformation of an initially
square element of the material.
T
E
C
H

N
O
T
E
TN-512-1
Micro-Measurements
Revision: 12-Sep-2010
114

1 1 1
2 2
2
2
2 =
+
+

+
x y x y xy
cos sin

(2)

2 2 2
2 2
2
2
2 =
+
+

+
x y x y xy
cos sin

(3)
Subtracting (3) from (2) and solving for
xy

xy
x y
=
( )
( )
( )
2 2 2
2
1 2 1 2
1
cos cos
sin siin2
2

(4)

It is now noticeable that if cos2
1
cos2
2
, the term in
x
and
y
vanishes, and

xy
=
( )
2
2 2
1 2
1 2
sin sin

(5)
Since the cosine function is symmetrical about the zero
argument, and about all integral multiples of ,
cos2
1
cos2
2
when, for an arbitrary angle ,

1 2
2 0 2
2
+ = = / , , / , ...
n

(6)
It is thus evident that, if the gage axes are oriented
symmetrically with respect to, say, the X axis (Figure 3),

1 2
= =
and,

xy
=

=

=

1 2
2
1 2
1
1 2
2 2 2 sin sin sin

(7)
The preceding results can be generalized as follows: The
difference in normal strain sensed by any two arbitrarily
oriented strain gages in a uniform strain eld is proportional
to the shear strain along an axis bisecting the strain gage
axes, irrespective of the included angle between the gages.
When the two gages are 90 degrees apart, the denominator
of Equation (7) becomes unity and the shear strain along
the bisector is numerically equal to the difference in
normal strains. Thus, a conventional 90-deg two-gage
rosette constitutes an ideal shear half bridge because the
required subtraction,
1

2
, is performed automatically
for two gages in adjacent arms of the bridge circuit (Figure
4a). When the gage axes of a two-gage 90-deg rosette
are aligned with the principal axes, the output of the
half bridge is numerically equal to the maximum shear
strain. A full shear-bridge (with twice the output signal) is
then composed of four gages as shown in Figure 4b. The
gages may have any of several congurations, including
the cruciform arrangement and the compact geometry
illustrated in the gure.
Principal Strains
It should be kept in mind that with the shear-bridges
described above, the indicated shear strain exists along
the bisector of any adjacent pair of gage axes, and it is
Figure 2 Arbitrarily oriented
strain gages in a biaxial strain eld.
Figure 3 Dening X axis as bisector of angle
between gage axes,
xy
= (
1

2
)/sin 2.
T
E
C
H

N
O
T
E
TN-512-1
Micro-Measurements
115
not possible to determine the maximum shear strain or
the complete state of strain from any combination of
gage outputs unless the orientation of the gage axes with
respect to the principal axes is known. In general, when the
directions of the principal axes are unknown, a three-gage
45-deg rectangular rosette can be used.
Referring to Mohrs circle for strain (Figure 5) it is apparent
that the two shaded triangles are always identical for a
45-deg rosette, and therefore the maximum shear strain
is equal to the vector sum of the shear strains along
any two axes which are 45 degrees apart on the strained
surface. Looking at the 45-deg rosette as shown in Figure
6, it can be seen that the shear strains along the bisectors of
the gage pairs and are in fact 45 degrees apart
and, thus, the maximum shear strain is,

MAX
= +
A B
2 2
and, considering Equation (7),

MAX
o o
=

1 2
2
2 3
2
45 45 sin sin
or,

MAX
= ( ) + ( ) 2
1 2
2
2 3
2

(8)

and, from Mohrs circle again, the principal normal
strains are obviously:

p q ,
=
+
( ) + ( )
1 3
1 2
2
2 3
2
2
1
2

(9)
Correction for Transverse Sensitivity
Up to this point, the effect of transverse sensitivity on shear-
strain indication has been ignored. However, correction
for this effect is particularly simple when all strain gage
grids, in either a two-element tee rosette or a three-element
45-deg rectangular rosette, have the same transverse
sensitivity. For such cases, correction consists of merely
multiplying the indicated shear strain by the factor (1
0
K
t
)/(1K
t
), where K
t
represents the common transverse
sensitivity of the rosette grids and
0
is the Poissons ratio
of the beam material on which the manufacturer measured
the gage factor of the gages. When the rosette grids do not
have the same transverse sensitivity, the error is a function
of the strain state, and the indicated strain from each
grid must be corrected separately. Relationships for this
purpose are provided in our Tech Note TN-509, Errors
Due to Transverse Sensitivity in Strain Gages.
Figure 6 45-deg rosette used to determine
A
and
B
, the
shear strains in the

and directions, respectively.
Figure 4a 90-deg rosette for direct
indication of shear strain
xy.
Figure 4b Full shear-bridge.
Figure 5 Mohrs circle for strain used to
determine maximum shear strain.

MAX
= +
A B
2 2
T
E
C
H

N
O
T
E
TN-512-1
Micro-Measurements
116
Applications
The area of application for shear strain measurement can
be divided into two categories by the type of material
(isotropic or orthotropic) on which the measurement is
made. As a rule the same categorization also divides the
applications according to the purpose of the measurement.
In the case of materials which can be treated as isotropic
(e.g., the structural metals), the usual reason for measuring
shear strain is to determine the magnitude of an applied
shear stress or load. In contrast, shear strain measurements
on materials such as wood and unidirectionally reinforced
plastics (orthotropic materials) are most commonly made
for the purpose of determining a mechanical property of
the material namely, its shear modulus or modulus of
rigidity.
An example of shear strain measurement on metals occurs
in shear-buckling studies. When thin panels of steel or
aluminum alloy are loaded in shear, there is ordinarily a
critical load at which the material buckles, forming one or
more waves, generally parallel to the maximum principal
stress direction. In studies evaluating the relative merits of
different structural congurations, a common practice is
to install strain gage rosettes near the center of the panel
to determine the maximum sustainable shear stress or
applied load prior to buckling.
A more common application for shear measurement on
metals is the torque transducer. A cylindrical shaft in
torsion is a case of essentially pure shear, and the applied
torque can be readily determined from two tee rosettes
positioned diametrically on the shaft surface, and oriented
so that their gridlines are at 45 degrees to the shaft axis.
Specially congured tee rosettes (Figure 7) are ordinarily
used for this purpose. When the rosettes are connected to
form a full Wheatstone bridge as indicated in the gure,
the bridge output is doubly sensitive to shaft torque, but
insensitive to bending and axial loads.
The most frequent application of shear strain measurement
in transducers is the shear-beam load cell, indicated
schematically in Figure 8. The load cell consists of a
short, stiff cantilever beam with the material recessed
in one area to form a thin shear web. Tee rosettes are
installed on both sides of this web to produce an output
proportional to the vertical shear force on the beam. Since
the vertical shear force is necessarily constant throughout
the length of the beam, the transducer output tends to
be independent of the position (along the beam axis) of
the applied load. The tee rosettes are connected in a full-
bridge circuit as indicated to render the output insensitive
to side loads and off-axis load components on the beam.
E
e
o
1
3
4
1
2
T
T
T
1
2
2
3
4
Figure 7 Torque transducer based on shear measurement in shaft.
Figure 8 Schematic representation of shear-beam load cell.
E
e
o
1
3
4
2
2
1
2
3
4
P
P
1
T
E
C
H

N
O
T
E
TN-512-1
Micro-Measurements
117
Experimental determination of the shear modulus of a
metal is uncommon, since this property can usually be
calculated with sufcient accuracy from the relationship:
G = E/[2(1+)]. In orthotropic materials, however, the
shear modulus is an independent mechanical property
which must be measured for each different material. The
usual procedure for doing so is to establish a specimen
geometry and loading system which produces a state
of pure shear with respect to the principal material
directions. Tee rosettes are installed on both sides of the
specimen for determining the shear strain () under load.
The corresponding shear stress () is obtained from the
measured load divided by the cross-sectional area in shear.
The shear modulus is then calculated from: G
12
=
12
/
12
,
where the 1-2 subscripts refer to the principal material
directions.
Although many different shear test methods have been
used on composite materials, the Iosipescu method (Figure
9) is widely favored and is the subject of an ASTM standard
(D 5379). The standard species measurement of the shear
strain by tee rosettes installed on the horizontal centerline
of the specimen, while the shear stress is calculated from
the load divided by the cross-sectional area between the
notches. It has been demonstrated, however, that the shear
stress distribution between the notches is far from uniform.
Moreover, the nonuniformity in shear stress distribution
varies markedly according to whether the length of the
specimen is parallel or perpendicular to the major principal
material direction (i.e., a 0-deg or 90-deg specimen). As a
result, it is necessary in each case to adjust the calculated
shear modulus with a different empirical correction
factor.
Actually, the true average shear stress on the cross section
between the notch roots is, by definition, equal to the
load divided by the cross-sectional area. If the average
shear strain over the same area were measured, the true
shear modulus of the material could be obtained directly,
without the need for correction factors. But every strain
gage gives an output corresponding to the average strain
under its grid length. Thus, the appropriate tee rosette for
the Iosipescu specimen is one which spans the complete
depth, from notch root to notch root as shown in Figure
10. When these special shear gages (Micro-Measurements
C032 and C085 patterns) are employed with the Iosipescu
specimen, the observed shear modulus is the same for the
0-deg and 90-deg specimens (without correction factors), as
it should be for the validity of the mechanics of orthotropic
materials.
h
L = 76.2 mm
h = 19.05 mm
d = 3.81 mm
R = 1.27 mm
w = 11.43 mm
thickness = 2.54 5.08 mm
(12.7 mm max)
90
d
w
d
L
R
0 Fibers
90 Fibers
Figure 9 Iosipescu V-notch specimen for
measurement of shear modulus.
E
e
o
1
3
4
2
1 2 1 2 3 4
3 4
FRONT BACK
Figure 10 Iosipescu specimen with tee rosette shear gages installed on both faces.
T
E
C
H

N
O
T
E
TN-512-1
Micro-Measurements
118
Some Cautions and Limitations
Since shear strains are inferred from normal strain
measurements, accurate results require the same care and
attention to detail as for any other strain measurement
task. There are, however, several additional considerations
when determining shear strains or analyzing data from
strain gage rosettes. The data-reduction relationships used
with rosettes are derived from the strain-transformation
equations [Equations (2), (3)], which are, in turn, based on
small-strain theory. The latter equations are precise only
for innitesimal strains, but they are adequately accurate
for typical strain measurements on metal test objects where
the strains are usually less than 1 percent. With plastics
and composite materials, however, larger strains may be
encountered, and the rosette data-reduction relationships
can become increasingly inaccurate at strain levels greater
than 1 percent.
Another consideration is the uniformity of the strain eld
under the area covered by the rosette, which is necessarily
greater than for a single grid of the same gage length.
For typical stress analysis purposes (i.e., to obtain the
stress at a point in a homogeneous material) the state of
strain under the rosette should be nearly uniform. When a
signicantly nonuniform strain eld is expected, the rosette
should be small enough relative to the strain gradient to
satisfactorily approximate measurement of the strain at a
point. This restriction does not apply when there is a valid
reason to integrate or average the strain over a specic
area, as in the previously cited case of the Iosipescu shear-
test specimen. Similarly, for any strain measurements on
composite materials, the rosette dimensions should usually
be large relative to the distance between inhomogeneities
(reinforcing ber spacing) to provide an accurate indication
of the macroscopic strain state.
Since the determination of shear strain is accomplished
by measuring two or three normal strains, it is always
necessary to make certain that the indicated shear strain
is unaffected by nonshearing load components such as
bending, twisting, or axial loads. This requirement can
usually be satised quite easily in the case of geometrically
symmetric test objects and specimens. With shear-test
specimens, for example, it is imperative that strain gage
rosettes be installed on both sides of the specimen at
corresponding points. When the rosettes are connected in
the Wheatstone bridge circuit as indicated in Figure 10,
strains due to out-of-plane bending or twisting are canceled
within the bridge circuit. Cancellation of undesired strain
components within the Wheatstone bridge circuit is a
widely applicable technique, and standard practice in the
technology of strain gage based transducers (see Figures
7 and 8). However, the intrinsic capacity for removing
undesirable strain components through bridge circuits
should not be used in place of good load alignment.
Carried to the extreme, shear strains could theoretically
be extracted from even predominantly off-axis loading. To
achieve the best accuracy, however, it is always preferable
to load the test specimen through its centroidal axis.
Tech Note TN-513-1
MICRO-MEASUREMENTS
Measurement of Thermal Expansion
Coefcient Using Strain Gages
T
E
C
H

N
O
T
E
119
The thermal expansion coefcient is a very basic physical
property which can be of considerable importance in
mech an ical and structural design applications of a
material. Al though there are many published tabulations
of expansion coefficients for the common metals and
standard alloys, the need occasionally arises to measure
this property for a specific material over a particular
temperature range. In some cases (e.g., new or special
alloys, composites, etc.), there is apt to be no published
data whatsoever on expansion coefcients. In others, data
may exist (and eventually be found), but may encompass
the wrong temperature range, apply to somewhat different
material, or be otherwise unsuited to the application.
Historically, the classical means for measuring expansion
coefcients has been the dilatometer. In this type of
instru ment, the difference in expansion between a rod
made from the test material and a matching length of
quartz or vitreous silica is compared
1,2
. Their differential
expansion is measured with a sensitive dial indicator,
or with an electrical dis place ment transducer. When
necessary, the expansion properties of the quartz or silica
can be calibrated against the accurately known expansion
of pure platinum or copper. The instrument is normally
inserted in a special tubular furnace or liquid bath to obtain
the required temperatures. Making measurements with the
dila tometer is a delicate, demanding task, however, and
is better suited to the materials science laboratory than
to the typical experimental stress analysis facility. This
Tech Note provides an alternate method for easily and
quite accur ately measuring the expansion coefcient of a
test material with respect to that of any reference material
having known expansion characteristics.
The technique described here uses two well-matched strain
gages, with one bonded to a specimen of the reference
material, and the second to a specimen of the test material.
The specimens can be of any size or shape compatible with
the available equipment for heating and refrigeration (but
speci mens of uniform cross section will minimize potential
prob lems with temperature gradients). Under stress-free
conditions, the differential output between the gages on
the two specimens, at any common temperature, is equal to
the dif ferential unit expansion (in/in, or m/m). Aside from
the basic simplicity and relative ease of making thermal
expansion measurements by this method, it has the distinct
advantage of requiring no specialized instruments beyond
those normally found in a stress analysis laboratory. This
technique can also be applied to the otherwise difcult
task of determining directional expansion coefcients of
materials with anisotropic thermal properties.
Because typical expansion coefficients are measured
in terms of a few parts per million, close attention to
procedural detail is required with any measurement
method to obtain accurate results; and the strain gage
method is not an exception to the rule. This Tech Note has
been prepared as an aid to the gage user in utilizing the full
precision of the modern foil strain gage for determining
expansion coefcients. Given in the rst of the following
sections is an explanation of the technical principles
underlying the method. The next section describes,
in some detail, the strain-gage-related materials and
procedures in making the measurement. Basically, the
latter consists of essentially the same techniques required
for any high-precision strain measurement in a variable
thermal environment. Suggested renements for achieving
maximum accuracy are then given in the following section;
after which, the principal limitations of the method are
described.
Principle of The Measurement Method
When a resistance strain gage is installed on a stress-free
specimen of any test material, and the temperature of
the material is changed, the output of the gage changes
correspondingly. This effect, present in all resistance strain
gages, was formerly referred to as temperature-induced
apparent strain, but is currently dened as thermal output
3
.
It is caused by a combination of two factors. To begin with,
in common with the behavior of most conductors, the
resistivity of the grid alloy changes with temperature. An
additional resistance change occurs because the thermal
expansion coefcient of the grid alloy is usually different
from that of the test material to which it is bonded.
Thus, with temperature change, the grid is mechanically
strained by an amount equal to the difference in expansion
coefcients. Since the gage grid is made from a strain-
sensitive alloy, it produces a resistance change proportional
to the thermally induced strain. The thermal output of the
gage is due to the combined resistance changes from both
sources. The net resistance change can be expressed as
the sum of resistivity and differential expansion effects as
follows:

= + ( )
R
R
F T
G s G G

(1)
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
120
Measurement of Thermal Expansion Coefcient Using Strain Gages
where:
R/R = unit resistance change

G
= thermal coefcient of resistivity of
grid material

s

G
= difference in thermal expansion coefcients
between specimen and grid, respectively
F
G
= gage factor of the strain gage
T = temperature change from arbitrary initial
reference temperature
The indicated strain due to a resistance change in the gage
is:

i
I
R R
F
=
/

(2)
where: F
I
= instrument gage factor setting
Then, the thermal output in strain units can be expressed
as:

T O G S
G S G G
I
F T
F
/ ( / )
=
+ ( )

(3)
where:

T/O(G/S)
= thermal output for grid alloy G on
specimen material S
Or, in the usual case, with the instrument gage factor set
equal to that of the strain gage, so that F
I
= F
G
,

T O G S
G
G
S G
F
T
/ / ( )
= + ( )

(4)
It should not be assumed from the form of Equation (4)
that the thermal output is linear with temperature, since
all of the coefcients within the brackets are themselves
functions of temperature. As an example, typical thermal
output characteristics for a Micro-Measurements A-alloy
gage (self-temperature-compensated constantan grid),
bonded to steel, are represented by the solid curve in
Figure 1. The lot of foil identied in the upper right corner
of the graph was specially processed to minimize the
thermal output over the temperature range from about 50
to +300F [45 to +150C]. Strain gages fabricated from
this lot of foil are intended for use only on material such
as steel with a coefcient of expansion of approximately
6 x 10
-6
/F [11 x 10
-6
/C]. If the gages are installed on some
other material with a different coefcient of expansion, the
result is to effectively rotate the curve in Figure 1 about its
reference point at +75F [+24C]. Installation on a material
with a higher coefcient of expansion than steel will rotate
the curve counterclockwise, while a material with a lower
expansion coefficient than steel will cause clockwise
rotation. For example, the broken curve labeled A in the
gure illustrates the general effect of installing a gage from
the subject lot on a beryllium alloy having an expansion
coefcient of about 9 x 10
-6
/F [16 x 10
-6
/C]. Similarly,
if a gage from this lot were bonded to a titanium alloy
with a somewhat lower expansion coefcient than steel,
the thermal output would be shifted in the manner of the
broken curve labeled B.
The principle of measuring expansion coefcients with
strain gages then becomes evident from Figure 1, since
the rotation from one thermal output curve to the other is
due only to the difference in thermal expansion properties
between the materials represented by the two curves. An
algebraic demonstration of the principle can be obtained
by rewriting Equation (4) twice; once for the gage installed
on a specimen of the test material of unknown expansion
coefcient
S
, and again for the same type of gage installed
on a standard reference material with a known expansion
coefcient
R
:

T O G S
G
G
S G
F
T
/ / ( )
= + ( )

(5a)

T O G R
G
G
R G
F
T
/ / ( )
= + ( )

(5b)
Figure 1 Rotation of the thermal output from
a strain gage when installed on materials with
differing thermal expansion coefcients.
TEMPERATURE IN FARENHEIT
+400
-50 0 +50
24C
A
B
STD
75F
+100 +150 -200 +250
+300
+200
+100
0
-100
-100 0 +100 +200 +300 +400 +500
-200
-300
-400
-500
A
P
P
A
R
E
N
T

M
I
C
R
O
S
T
R
A
I
N
(
B
a
s
e
d

o
n

I
n
s
t
r
u
m
e
n
t

G
.
F
.

o
f

2
.
0
0
)
Lot No. A38AD497
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
121
Subtracting Equation (5b) from (5a), and rearranging,

S R
T O G S T O G R
T
=
( ) / ( / ) / ( / )

(6)
Thus, the difference in expansion coefcients, referred
to a particular temperature range, is equal to the unit
difference in thermal output for the same change in
temperature. Although this technique for measuring
expansion coefficients is widely applicable, and often
the most practical approach, there is relatively little
information about it in the technical literature. Rep re-
sentative applications are described in the bibliography to
this Tech Note
4,5
.
Measurement Procedures
Reference Material
Selection of the material to be used as a reference standard
is naturally an important factor in the accuracy of
the method, as it is for any other form of differential
dilatometry. In principle, the reference material could
be any substance for which the expansion properties are
accurately known over the temperature range of interest.
In practice, however, it is often advantageous to select
a material with expansion properties as close to zero as
possible. Doing this will provide an output signal that
closely corresponds to the absolute expansion coefcient
of the test material, and permits a more straightforward
test procedure. The thermal expansion of the reference
material should also be highly repeatable, and stable with
time at any constant temperature. In addition, the elastic
modulus of the material should be great enough that
mechanical reinforcement by the strain gage is negligible.
An excellent reference material with these and the other
desirable properties is ULE Titanium Silicate Code 7972,
available from Corning Glass Company, Corning, NY
14831.* As illustrated in Figure 2, this special glass has an
extremely low thermal expansion coefcient, particularly
over the temperature range from about 50 to +350F
[45 to +175C)]. It should be noted, however, that the
material has a low coefcient of thermal conductivity,
making it slow to reach thermal equilibrium. For optimum
results, a dwell time of at least 45 minutes should be used at
each new temperature point before taking data. Another
potential disadvantage of titanium silicate as a reference
material is its brittleness, since it will fracture readily if
dropped on a hard surface. Because of the foregoing, a
low-expansion metal (such as Invar or a similar alloy) may
offer a preferable alternative if the alloy has repeatable
and accurately known expansion properties over the
temperature range of interest.
Strain Gage Selection
The type of strain gage selected for use in measuring
expansion coefcients is also an important consideration,
just as it is for stress analysis and transducer applications.
Gage selection usually requires weighing a variety of
factors which can directly or indirectly affect the suitability
of a particular gage type to a specied measurement task.
To assist gage users in this process, our Tech Note TN-505
provides extensive background data for gage selection,
along with procedures, recommendations, and application
examples
6
. The subject Tech Note should serve as the
primary reference on gage selection, supplemented here by
special considerations applicable to the measurement of
expansion coefcients.
For good accuracy, combined with ease of installation, a
gage from Micro-Measurements CEA Series is ordinarily
a suitable choice. This assumes that the temperature
extremes for the measurements fall within the range of
Thermal Expansion: 80 to +150C
Temperature (C)
L
/
L

(
p
p
m
)

8
6
4
2
0
-2
80 60 40 20 0 +20 +40 +60 +80 +100 +120 +140
Thermal Expansion: 0 to 150C
Temperature (C)
Code 7972
L
/
L

(
p
p
m
)

6
4
2
0
0 10 30 50 70 90 110 130 150
-2
Figure 2 Thermal expansion characteristics of the titanium silicate reference material (data source: Corning Glass Company).
* Also available from Micro-Measurements as Part No. TSB-1.
See Appendix for specimen dimensions.
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
122
greatest stability and precision for the constantan foil in
this type of gage [about 50 to +150F (45 to +65C)].
If a wider temperature range is involved, a gage from the
WK Series becomes the preferred choice. The latter gage
type is somewhat stiffer, however, and consideration of
reinforcement effects may be necessary if the test material
has a low modulus of elasticity, or the test specimen is thin
and narrow.
In each of the foregoing cases, a 350 gage is preferable in
order to minimize self-heating by the excitation current.
The 350 gage is also advantageous in reducing the effects
of small imbalances which may occur due to unsymmetric
resistance changes in the leadwires with temperature. In
addition, it is good practice, when feasible, to employ a
medium gage length say, 1/8 in [3 mm] or larger for
more stable operation and improved heat transfer to the
substrate.
Another gage parameter to be specified is the self-
temperature-compensation (S-T-C) number. In principle,
as indicated by Equation (6), it should not matter what
S-T-C number is selected. Only the difference in thermal
output, for the same gage type on two different materials,
is involved in the expansion calculations. Practically,
however, there are two considerations which may inuence
the choice. One of these is the availability of the selected
gage in the desired series, gage pattern, and resistance.
As a rule, the greatest selection of gages is available in the
06 and 13 S-T-C groups, since these are the most widely
used compensations for stress analysis and transducer
applications. It will often be expedient, therefore, to
specify one of the above for the S-T-C number.
When expansion measurements must be made over an
ex tended temperature range, or at high or low temperature
ex tremes, the S-T-C number should be carefully selected
to obtain the best measurement accuracy. It is evident
from Figure 1 that, with excessive mismatch between the
S-T-C number of the gage and the expansion coefcient
of the specimen, the slope of the thermal output curve can
become very steep at one or both extreme temperatures.
Under such circumstances, a small error in temperature
(or temperature deviation between the reference and test
materials) can produce a large error in the thermal output
signal. Judicious selection of the S-T-C mismatch can be
used to simultaneously keep the slopes of the thermal
output curves for both the test and reference materials
under reasonably good control in the temperature range
of interest.
Almost any single-element linear grid pattern can
be employed for measuring expansion coefficients. As
indicated earlier, however, the two gages one on the
reference specimen, and one on the test material
must always be well-matched. That is, the gages must
be identically the same type, and must be from the same
manufacturing lot to assure closely related thermal output
characteristic. Both requirements can be met by simply
using a pair of gages taken from the same package. Gages
of the identical type taken from different packages, but
having the same lot number, will be equally close in their
thermal outputs. When a still closer relationship is desired
for greater measurement accuracy, a dual-grid gage pattern
such as the 125MG (Figure 3) can be selected, and the grids
cut apart to form two individual gages. The resulting gages
are, in effect, identical twins, and will provide the closest
possible match in thermal output characteristics (as in all
other properties).
Gage Installation
As noted, one of the advantages of this method is that
the specimens of the reference and test materials can be
of any convenient size or configuration suitable to the
available heating or refrigeration equipment. In fact, the
two specimens can even be different in size or shape if
there is a reason to have them so. In general, however,
specimens should be uniform in cross section to minimize
temperature gradients induced during heating or cooling;
and the use of flat specimens will make for easier and
higher-quality gage and temperature sensor installations.
The specimens should also be large enough in cross section
so that the strain gage stiffness is negligible compared
to the overall section stiffness. Beyond the foregoing,
selection of the specimen dimensions for about the same
thermal inertia will be helpful in most quickly achieving
the same temperature when both specimens are heated or
cooled together.
Specimen surfaces should be thoroughly cleaned and
prepared for bonding as described in Micro-Measurements
Instruc tion Bulletin B-129, which includes specic step-by-
step procedures for a wide variety of materials
7
. For best
accuracy, bonding should be done with a high performance
adhesive such as M-Bond 600 or 610. Both adhesives are
capable of forming thin, hard gluelines for maximum
delity in transmitting strains from the specimen surface to
the gage. These adhesives are intended for use on relatively
smooth, nonporous surfaces, and should not be used where
Figure 3 Micro-Measurements
type 125MG dual-grid strain gage pattern.
~3.5x actual size
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
123
the adhesive is required to ll surface irregularities or to
seal pores. For the latter conditions, the recommended
adhesive is M-Bond AE-10 or AE-15. In all cases, complete
instructions for applying and curing the adhesive are
included in the package with the material.
Extra care is required in the selection of leadwires and
their attachment to the gages, in order to obtain the
most accurate results. Ther mally produced resistance
changes in the leadwires will generate circuit outputs
which are indistinguishable from the thermal outputs
being measured. If these differ in any way between the
reference and test specimens, the indicated differential
expansion data will be in error accordingly. To minimize
such effects, leadwire resistance should be kept as low as
possible by employing a generous wire size, and by keeping
the leads short. The wiring should also be the same for both
specimens in size, length, and routing. If measurements
are to be made on both specimens in the same chamber or
liquid bath at the same time, the leadwires should be kept
physically together throughout as much of their length
as practical. Leadwire insulation must be selected, of
course, for compatibility with the temperature range and
environment encountered in the measurements.
In attaching leadwires to the gage solder tabs or to solder
terminals, the solder joints should be smooth, bright, and
free of spikes or excess solder. The joints should also be as
uniform as possible; and the leadwires should be dressed
the same on both specimens. After lead attachment, the
gage installations must be thoroughly cleaned with rosin
solvent to remove all traces of soldering ux and residues.
The nal step in the installation is to apply a protective
coating system which is appropriate to the expected
test envi ronment. Si nce these tests are normal ly
conducted under short-term laboratory conditions, a
coating is selected for basic pro tec tion against moisture,
dew point condensation in cold tests and minimum/
maximum operating temperature range. The coating
recommendations in the following table also take into
consideration low reinforcement of the specimen. Fur-
ther details on these and other coatings can be found in
Micro-Measurements Strain Gage Accessories Data
Book.
The process of gage installation has been summarized
very briey here, since detailed instructions are supplied
elsewhere in our technical publications. It should be
appreciated, however, that proper gage installation is a
basic requirement for accurate measurement of expansion
coefcients. In general, gage installations should be of the
highest quality comparable to those found in precision
strain gage transducers. Care should also be taken that the
two gage installations, on the reference and test specimens,
are as uniform as possible to minimize small physical
differences which could affect the differential thermal
response. If installation questions or problems arise, the
user should consult the our Applications Engineering
Department for assistance. Figure 4 is a photograph of
a properly installed strain gage on a metal specimen for
thermal expansion measurements. A bondable resistance
temperature sensor (see Figure 6) is installed adjacent
to the gage to monitor the specimen temperature. This
photograph shows the installation just prior to application
of the protective coating over the gage and temperature
sensor.
Strain and Temperature Instrumentation
Basically, any stable precision strain indicator can be used
for the strain measurements needed in this procedure.
Satis factory instruments for this purpose include the
Model P3 and Model 3800 Strain Indicators produced by
the Instruments Division of Micro-Measurements. Beyond
the necessity for instrument precision and stability, it is
important that the gage excitation voltage be kept low
enough to avoid the effects of self-heating in the gage.
Both the Models P3 and 3800 are high-gain instruments
with low excitation voltages. Using these strain indicators,
there is ordinarily no self-heating problem with a gage
such as the 125MG pattern installed on a metal specimen
PROTECTIVE COATING
Operating
F
Temperature
Range
C
Coating
+60 to +250
0 to +150
100 to +500
452 to +400
[+15 to +120]
[20 to +65]
[75 to +260]
[269 to+200]
M-Coat A or C
W-1 Wax
3140 or 3145 RTV
Two coats M-Bond 43B
Figure 4 Strain gage (half of the 125MG dual-gage
pattern, at top) and resistance temperature sensor,
installed side-by-side on a specimen of test material
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
124
with reasonably good heat-dissipating characteristics.
When measurements are made with other instruments
having higher excitation voltages, or with gages installed
on specimens of low thermal conductivity, self-heating
may be excessive, and the voltage applied to the gage must
be reduced. Compre hen sive background information and
guidelines for setting exci ta tion voltages are provided in
Tech Note TN-502
8
.
Either of two basic circuit arrangements can be used in
measuring expansion coefcients. One of these, shown
in Figure 5a, employs separate, three-wire, quarter-
bridge circuits for the gages on the reference and test
specimens. With this arrangement, the gage outputs
are read individually, and subsequently subtracted to
determine the differential strain for use with Equation (6).
Since the separate circuits permit monitoring the gages
independently, it is relatively simple to identify the cause
of any improper or anomalous strain readings which may
occur when conducting the test. A disadvantage of this
ap proach is that it requires a switch-and-balance unit
(when used with a single-channel strain indicator) or a
two-channel instrument.
The second arrangement (Figure 5b) uses the properties
of the half-bridge circuit to perform the subtraction
electrically. When the two gages are connected as adjacent
arms of the bridge circuit, the instrument output is equal
to the difference in the individual thermal outputs. The
circuit is obviously simpler in terms of both wiring and
instrumentation, and is direct-reading. Its pri mary
disadvantage lies in the difculty of isolating the gage
which may be malfunctioning when improper operation is
suspected.
In both of the foregoing circuit arrangements, the leadwires
to the gages should be as short as possible, and should be of
the same wire size and length. Since leadwires #1 and #3 are
always in adjacent arms of the bridge circuit, they should
be particularly well-matched and maintained physically
together throughout their lengths, to minimize differential
resistance changes which could appear in the instrument
output. With a half-bridge circuit such as shown in Figure
5b, it is also necessary that leadwire #2 be connected at the
midpoint of the jumper between the gages. This is done to
place half of the jumper resistance in series with each gage
in its respective bridge arm, and thus avoid a false output
signal due to the thermally induced resistance change in
the jumper wire. It is worth noting that a 6-in [~150-mm]
dissymmetry in the wiring whether in leadwires #1 and
#3, or in the jumper in AWG 30 [0.25 mm] wire size will
cause a false output of about 17 per 100F [per 55C].
Temperature measurement also requi res care and
consideration to obtain accurate expansion data. Typically,
a temperature-sensing probe is placed immediately adjacent
to the gage, and in intimate contact with the specimen
surface, to indicate the spec i men/gage temperature. This
procedure assumes that previous verication has been made,
by multiple temperature measurements on the specimen as
necessary, to assure uniform specimen temperature under
conditions of thermal equilibrium in the test chamber.
Since the materials in the reference and test specimens
normally differ in their thermal conductivity and specic
heat, it is necessary that the temperatures at both gage sites
be measured. The temperature must be the same, of course,
whenever paired strain readings are made.
Dependi ng pri mari ly on personal preference and
instrumentation availability, temperatures can be measured
either with thermocouples or with resistance temperature
sensors. If a thermocouple is employed on each specimen,
type J (iron-constantan) is preferred, assuming that the
test temperature range is compatible with this type. The
sensing junction should be small, as should the leadwires
(in the range of AWG 30 to AWG 26 [0.25 to 0.4 mm]), and
premium grade thermocouple wire should be selected.
Heat transfer from the specimen to the junction can be
improved by taping the rst 2 to 3 in [50 to 75 mm] of the
extension wires to the specimen surface.
An alternate approach is to use resistance temperature
sensors such as Micro-Measurements TG-Series (Figure
6). The temperature sensor looks like a strain gage, and
has essentially the same construction except that the
grid is made from high-purity nickel foil. It is installed
Figure 5 Strain gage circuits for measuring
thermal expansion coefcients: (a) separate
quarter-bridge circuits; (b) half-bridge circuit.
Test Material
Reference Matl
Test Material
Reference Matl
1
2
3
1
2
3
1
2
3
e
o
e
o
1
2
3
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
125
with standard strain gage installation procedures, and
should be mounted side-by-side with the strain gage on the
specimen surface. Because it is physically like the strain
gage, and is attached to the specimen in the same way,
the temperature sensor has about the same heat-transfer
characteristics and thermal time constant as the strain
gage. When used in conjunction with a specially designed
passive resistance network for linearization and signal
scaling (Micro-Measurements Type LST), it permits direct
measurement of temperature with any conventional strain
indicator. The small size and low stiffness of the TG-Series
temperature sensor present minimum mechanical restraint
to the free thermal expansion and contraction of the
specimen.
Making Expansion Measurements
For any method of dilatometry, it is always necessary
that the reference and test specimens be exposed to at
least two di fferent temperatures i n measuri ng the
expansion coefcient. The actual means of achieving the
desired temperatures in a particular case depends on the
temperatures involved, and on the available facilities.
These may consist, for instance, of ovens, or liquid baths,
or various other forms of environmental chamber. The
strain gage method imposes no special restrictions on the
nature or design of the chamber. On the contrary, the size
and shape of the specimen can usually be adapted to suit
the existing facilities. Since the available equipment varies
widely from one laboratory to the next, the following
remarks are limited to the general requirements for any
dilatometric temperature chamber.
Two of the most desirable features of a chamber for
measuring expansion coefficients are uniformity and
stability of temperature. To avoid errors due to the
development of thermal stresses in the specimen, the
temperature should be uniform throughout the specimen
at the ti me of measurement. This condition can be
established only if the chamber temperature at equilibrium
is essentially uniform at least in the region containing
the specimens. Temperature stability in the chamber is also
necessary to permit measuring specimen temperatures and
strains under static, nonvarying conditions.
Thermal equilibrium in the specimen can be achieved in
a chamber equipped with a forced convection system to
vigorously circulate the heat-transfer medium past the
specimen surfaces. Heating and cooling rates should also be
kept low to minimize temperature gradients perpendicular
to the specimen surface. The required condition of uniform
temperature throughout the specimen is difcult to judge,
however, and is not necessarily assured by observing
equal temperature readings at different points on the
surface. One of the most effective ways to test for control
over the uniformity of specimen temperature is to make a
continuous plot of strain gage output versus temperature
over the working temperature range in both the heating
and cooling directions. In this process, the temperature
is changed incrementally; and, at each test temperature,
after the specimen is evidently in thermal equilibrium,
the temperature and thermal output are recorded and
plotted. If uniformity of specimen temperature is actually
achieved, the heating and cooling legs of the plotted curve
should very nearly coincide. If, on the other hand, the two
portions of the curve are signicantly separated to form a
hysteresis loop, a likely cause is nonuniform temperature
distribution through the thickness of the specimen. In the
latter case, the heating and cooling rates must be lowered,
or thermal stabilization times increased, or other measures
taken to essentially eliminate the temperature gradients.
Means must be provided for supporting the specimens in
the chamber so that friction cannot impede expansion or
contraction. In some cases, a simple way to accomplish
this is to suspend the specimens from one end. Although
the specimen may be strained slightly by its own weight,
the strain is constant (as long as the elastic modulus is
essentially constant), and does not affect the change in
thermal output with temperature. If the elastic modulus
of the test material changes signicantly over the range
of temperatures to be encountered, the error due to this
effect must be evaluated to determine the suitability of
the method. Another approach is to lay the specimens on
the oor of the chamber or compartment, supported by a
layer of berglass cloth or some other low-friction medium.
When this method is used, its effectiveness should be
veried by observing the behavior of the thermal output
as the specimen is cycled through the working temperature
range. Erratic output, hysteresis, or lack of repeatability
may indicate excessive friction.
Before performing actual measurements to determine the
coefcient of expansion, the entire system, including both
specimens (with gages installed and power applied), should
be stabilized by cycling several times to temperatures at
least 10F [5C] above the highest, and below the lowest,
Figure 6 Micro-Measurements TG-Series
ETG-50B/W bondable temperature sensor.
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
126
test temperatures. One of the reasons for this procedure
is that residual stresses are generally present in all of the
components the reference and test specimens, the gages
as manufactured and installed, the leadwires, etc. Thermal
cycling is intended to relax and/or redistribute any residual
stresses which might otherwise change during the test and
cause the data to be nonrepeatable. The cycling procedure
should be performed at low enough rates of temperature
change to minimize thermal stresses in the specimens due
to temperature gradients. Otherwise, the thermal stress,
superimposed on the residual stress, may cause yielding,
and thus defeat the purpose of the cycling.
Normally, after the second or third stabilizing cycle,
the thermal output at any given temperature should be
highly repeatable. If not, and if the lack of repeatability
is signicant compared to the accuracy required from the
test, the sources of the variability must be found. In such
cases, the problem may be associated with the temperature,
or the strain, or both. Careful re-reading of this Tech Note
may provide the clue for nding and correcting the trouble.
Further assistance, if needed, can be obtained from our
Applications En gi neering Department.
Following stabilization, veried by reproducible strain
indi ca tions throughout the temperature range, the user is
ready to per form the nal measurements for determining
the thermal expansion properties of the test material.
When the oven or other chamber is such that only a single
specimen can be ac commodated, the two specimens are
tested one-at-a-time, using the circuit of Figure 5a. The
resulting two sets of thermal output data are subtracted
(and the difference divided by the temperature change) as
indicated by Equation (6) to give the differential thermal
expansion coefcient. With the preferable ar rangement,
having both specimens together in the chamber, the
measurements can be made separately as in Figure. 5a,
or the differential thermal output can be read directly as
shown in Figure 5b.
Special Precautions and Renements
for Improving Accuracy
When attempting to achieve greater and greater accuracy
with the strain gage method (or with any method), it is
necessary to examine ever smaller effects which may
introduce errors. In some instances, these second-order
errors are well-dened, systematic in nature, and responsive
to routine procedures for correction or elimination. In
others, the cause-and-effect relationship is more nebulous,
and error reduction is accomplished primarily by technique
renement i.e., by removing or minimizing all of the
known possible sources of error.
An example of a readily correctable inaccuracy (in certain
cases) is the error due to transverse sensitivity. This error
arises because the strain eld induced in the gage grid by
the difference in thermal expansion between the specimen
and grid [Equation (1)] is generally different from that
employed in gage factor calibration
9
. When both the
reference and test materials are isotropic in their thermal
expansion properties, the transverse-sensitivity error,
which is ordinarily quite small, can be corrected for rather
easily. Although not derived here, correction can be made
by multiplying the difference in thermal outputs [Equation
(6)] by the factor (1 0.285 K
t
)/(1 + K
t
), where K
t
is the
decimalized transverse sensitivity of the gage in use. This
correction factor is not applicable to orthotropic materials,
for which case differential thermal outputs between a
reference gage and two perpendicularly oriented specimen
gages are required to correct for transverse sensitivity.
Another minor error source is the variation of gage factor
with temperature. The gage factor specied for Micro-
Measurements strain gages is measured at +75F [+24C].
At any other temperature it is slightly different. With
constantan gages, for example, the gage factor varies
directly with temperature, at a rate of about 0.5% per 100F
[0.9% per 100C]. In contrast, the gage factor of K-alloy
(modied Karma) gages varies inversely with temperature.
The rate of change depends on the S-T-C number of the
gage, but is generally in the range from 0.5 to 1.0% per
100F [0.9 to 1.8% per 100C]. Represen tative plots of
gage factor variation with temperature are illustrated in
Figure 7 for both types of gages. The technical data sheet
contained in each gage package includes data for the gage
factor variation applicable to that gage type.
Complete elimination of the small error introduced by
gage factor variation is not always feasible, but rst-order
correction, to remove most of the error, is relatively simple.
When expansion measurements are made incrementally
Figure 7 Gage factor variation with temperature
(typical) for A- and K-alloy strain gages.
+2.0%
0 +50
24C
03
09
A-alloy
K-alloy
75F
+100 +150 -200 +250
+1.0%
0
0 +100 +200 +300 +400 +500
1.0%
2.0%
G
A
G
E

F
A
C
T
O
R

V
A
R
I
A
T
I
O
N
(
T
y
p
i
c
a
l
)
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
127
across the working temperature range, the differential
thermal output for each increment in temperature can be
corrected individually. This is done by multiplying the
difference in indicated thermal outputs from the specimen
and reference gages by the factor 1/(1 + F
G
). The term F
G

in the foregoing is the decimalized change in gage factor
(with sign) corresponding to the middle temperature of
each measurement increment.
Sometimes, the average differential expansion coefcient
is to be determined over the full temperature range by
making only two sets of measurements, at the temperature
extremes. The same correction procedure can be applied,
using the F
G
for the mid-range temperature, but it will
be much less effective because the thermal output is a
nonlinear function of temperature.
When the leadwire resistance can be kept very low,
as recommended in the preceding section, the signal
attenuation (desensiti zation) caused by the i nert
resistance in series with the gage should be negligible. If, on
the other hand, the series resistance is greater than about 1
percent of the gage resistance, the user who is striving for
maximum accuracy may wish to perform a correction. For
this purpose, the indicated thermal outputs are multiplied
by the factor (R
G
+ R
L
)/RG, where R
G
is the gage resistance,
and R
L
is the leadwire resistance in series with the gage in
the same arm of the bridge circuit. An alternative, for
direct reading of corrected strains, is to set the gage factor
control of the instrument at F
G
x R
G
/(R
G
+ R
L
), where F
G
is
the specied gage factor of the gages in use.
The supposition is made, in the strain gage method of
measuring expansion coefcients, that if the two gages
(and gage circuits) behave identically, then any difference in
their outputs can be due only to the difference in expansion
properties be tween the reference and test specimens. It
is obvious, therefore, that the highest accuracy will be
achieved by minimizing all differences in gage behavior.
For this reason, as noted earlier, the thermal output
characteristics of the gages should be as nearly the same as
possible. However, two nominally identical gages from the
same manufacturing lot do not especially have identical
thermal outputs. Instead, as shown in Figure 8, there
is a tolerance on the thermal output.* Almost all of the
tolerance can be removed by splitting a dual-element gage
(such as the 125MG pattern) to make a pair of twin gages,
and this procedure is always recommended when high
accuracy is the goal. The same reasoning underlies the
repeated emphasis in this Tech Note on the uniformity of
gage installations. Identical installation procedures should
be used for both gages; and, ideally, there should be no
visible differences in the completed installations.
The remaining areas of possible renement for improved
accuracy are primarily associated with the measurements
procedures. Each of the items in the following checklist can
be considered, and steps taken as necessary to satisfy the
desired conditions:
a) stable, accurate instrumentation, for both temperature
and strain.
b) high-quality, stable gage installations, exhibiting
negligible drift over the operating temperature range.
c) gage excitation at a level low enough to avoid self-
heating effects.
d) thermal stabilization of specimens, gages, and wiring
prior to making expansion measurements.
e) assurance of thermal equilibrium in the specimens
when measurements are made.
f) avoidance of signicant thermal stresses during
heating and cooling.
g) elimination of frictional effects preventing free
expansion and contraction.
Except for the absolute accuracy of the instrumentation,
the degree to which the foregoing conditions have been met
can be judged quite well by the repeatability of the data.
Highly reproducible data generally indicate that the system
is functioning properly, and that random error sources are
well-controlled.
After it has been demonstrated that the measurement
system and procedures are suitable for obtaining closely
reproducible data from a single specimen, consideration
should be given to the question of variation in thermal
properties from specimen to specimen. The usual purpose
of expansion-coefficient measurements is to determine
the nominal value which is representative of a particular
material. But the thermal and other physical properties
of any material tend to vary randomly from specimen to
+400
-50 0 +50
24C
A
75F
+100 +150 -200 +250
+300
+200
+100
0
-100
-100 0 +100 +200 +300 +400 +500
-200
-300
-400
-500
T
H
E
R
M
A
L

O
U
P
U
T

m
/
m
(
B
a
s
e
d

o
n

I
n
s
t
r
u
m
e
n
t

G
.
F
.

o
f

2
.
0
0
)
U
n
c
e
r
t
a
i
n
t
y
:

0
.
2
7

(
m
/
m
)

C
-
1
* See Tech Note TN-504.
Figure 8 Tolerance band for the thermal output of randomly
selected A-alloy strain gages from the same manufacturing lot.
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
128
specimen within a lot, and still more widely from lot to
lot. Since such variation is not subject to the control of
the user, it becomes necessary to use statistical sampling
techniques, with a sample size large enough to provide an
adequate estimate of the mean and standard deviation.
Variability in thermal properties is apt to be particularly
great in materials such as plastics and composites.
The mechanical and thermal properties of some materials
(e.g., graphite, titanium 6A14V, composites with oriented
ber reinforcement, etc.) are highly directional. In such
cases, orientation of the strain gage on the specimen (with
respect to the natural axes of the material, as determined
by the rolling direction, ber orientation or otherwise)
is critical if the directional expansion coefficient is to
be measured. When it is impossible to determine the
directions of the natural material axes, it may be necessary
to make measurements over a wide range of angles to
dene the distribution of the expansion coefcient, or to
obtain a rough, integrated average value.
Limitations
The strain gage method of differential dilatometry has
very few special limitations. Of these, the principal one for
some types of studies may be the allowable temperature
range. Con stantan gages, for instance, should be used for
high-accuracy measurements only within a temperature
range from about 50 to +150F (45 to +65C). Higher
temperatures normally require the use of K-alloy gages,
which can provide accurate strain measurements from
approximately 50 to +400F (45 to +205C). With
special techniques, these temperature ranges can sometimes
be extended, depending on the circumstances. Users
should consult with the Micro-Measurements Appli cations
Engineering Department for recommendations.
Mechanical reinforcement of the specimen by the strain
gage can also be a limitation in some instances. When
the test specimen is made from a material such as plastic,
with a very low modulus of elasticity, the stiffness of the
gage may perturb the local strain eld and introduce a
sizeable error. With metal specimens, the reinforcement
effect is ordinarily negligible unless the specimen is so thin
and narrow that the gage stiffness represents a signicant
fraction of the overall section stiffness.
Other limitations are generally those common to all
methods of differential dilatometry. For example, the
expansion coefficient of the test material can never be
determined to greater accuracy than that of the reference
material. Similarly, the measurements can be no more
accurate than the instrumentation used to indicate the
temperatures and strains.
Summary
This Tech Note has described a simple, straightforward
means of measuring the expansion coefcient of a test
material relative to that of any reference material having
known expansion properties. The method is particularly
well-suited to the stress analysis laboratory, since it
usually requires no special instrumentation, techniques,
or materials not already available in such a facility.
Considerable attention has been given here to procedural
details aimed at extracting the utmost accuracy from
the method. Most of the recommended procedures,
however, should represent standard practices for a stress
laboratory which is accustomed to making precision
strain measurements in a variable thermal environment.
Even when expedience dictates somewhat less rigorous
procedures, the method can be used to quickly and easily
measure thermal expansion coefficients with sufficient
accuracy for many engineering purposes.
References
1. American Society for Testing and Materials, Standard
Test Method for Linear Expansion of Metals, ASTM
Standard No. B95-39.
2. American Society for Testing and Materials, Linear
Ther mal Expansion of Rigid Solids with a Vitreous
Silica Dila tometer, ASTM Standard No. E228-71.
3. Micro-Measurements, Tech Note TN-504, Strain
Gage Thermal Output and Gage Factor Variation with
Temperature, 1989.
4. Finke, T. E., and T. G. Heberling, Determination of
Thermal Expansion Characteristics of Metals Using
Strain Gages, Proceedings, SESA (now, SEM), Vol.
XXV, No. 1, 1978, pp. 155-158.
5. Poore, M. W., and K. F. Kesterson, Measuring the
Thermal Expansion of Solids with Strain Gages,
Journal of Testing and Evaluation, ASTM, Vol. 6, No. 2
(March 1978), pp. 98-102.
6. Micro-Measurements, Tech Note TN-505, Strain Gage
Selection Criteria, Procedures, Recommendations,
1989.
7. Micro-Measurements, Bulletin B-129, Surface Pre-
paration for Strain Gage Bonding, 1976.
8. Micro-Measurements, Tech Note TN-502, Optimizing
Strain Gage Excitation Levels, 1979.
9. Micro-Measurements, Tech Note TN-509, Errors Due
to Transverse Sensitivity in Strain Gages, 1982.
T
E
C
H

N
O
T
E
TN-513-1
Micro-Measurements
129
APPENDIX
REFERENCE INFORMATION
I. Specication for CORNING GLASS WORKS Titanium Silicate, Code 7972 ULE thermal expansion coeffecient

Control limit: +40 to +95F [+5 to +35] 0.00 0.017 x 10
-6
/F [0.00 0.03 x 10
-6
/C]
Typical values: +32 to +390F [0 to +200C] 0.017 0.017 x 10
-6
/F [0.03 0.03 x 10
-6
/C]
150 to +390F [100 to +200C] 0.017 0.017 x 10
-6
/F [0.0c 0.03 x 10
-6
/C]
Tolerance within one specimen purchased from Micro-Measurements (Part No. TSB-1):
+40 to +95F [+5 to +35C] 0.000.008 x 10
-6
/F [0.00 0.015 x 10
-6
/C]
This tolerance also applies to typical values noted above.
Micro-Measurements specimen size: 6 x 1 x 0.25 in [155 x 30 x 6.5 mm]
Micro-Measurements specimen nish: 80 grit
II. Thermal Output Scatter of Micro-Measurements Strain Gages
All data are based on a 2 or 95% condence level over the temperature range of +32 to +350F [0 to +175C]
Single-element A-alloy gages: 0.15 in/in/F [0.27 m/m/C]
Single-element K-alloy gages: 0.25 in/in/F [0.45 m/m/C]
EA-XX-125MG-120 with one grid on Code 7971 and other on unknown material: 0.03 in/in/F [0.05 m/m/C]
WK-XX-125MG-350 used as described for the EA gage: 0.06 in/in/F [0.10 m/m/C]
III. Correction for Transverse Sensitivity
With K
t
, in decimal form, multiply the parenthetic expression [
T/O(G/S)

T/O(G/R)
] in Equation (6), by
(1 0.285 K
t
) /(1 + K
t
) for isotropic materials only.
IV. Correction for Gage Factor vs. Temperature
For any temperature increment, multiply the parenthetic expression [
T/O(G/S)

T/O(G/R)
] in Equation (6), by
1/(1 + F
G
). The term F
G
in decimal form, corresponds to the midpoint of the temperature increment over which
thermal output measurements are made.
V. Correction for Leadwire Resistance (R
L
) for a Single Gage in a Three-Wire Conguration
R
L
is the resistance of a single leadwire in the three-wire connection to the instrument. To avoid the tedious task of
correcting all individual readings by the factor (R
G
+ R
L
)/R
G
, it is much simpler to adjust the gage factor setting of
the instrument to F
1
= F
G
x R
G
/(R
G
+ R
L
).
To evaluate the need for this correction, the approximate lead resistances for typical Micro-Measurements
cables are:
326-DFV, 326-DTV: 0.043 ohms/ft [0.141 ohms/m]
330-DFV, 330-FFE, 330-FJT, 330-FTE: 0.108 ohms/ft [0.354 ohms/m]
Tech Note TN-514
MICRO-MEASUREMENTS
Shunt Calibration of Strain Gage Instrumentation
T
E
C
H

N
O
T
E
131
I. Introduction
The need for calibration arises frequently in the use
of strain gage instrumentation. Periodic calibration is
required, of course, to assure the accuracy and/or linearity
of the instrument itself. More often, calibration is necessary
to scale the instrument sensitivity (by adjusting gage factor
or gain) in order that the registered output correspond
conveniently and accurately to some predetermined input.
An example of the latter situation occurs when a strain
gage installation is remote from the instrument, with
measurable signal attenuation due to leadwire resistance.
In this case, calibration is used to adjust the sensitivity
of the instrument so that it properly registers the strain
signal produced by the gage. Calibration is also used to set
the output of any auxiliary indicating or recording device
(oscillograph, computer display, etc.) to a convenient scale
factor in terms of the applied strain.
There are basically two methods of calibration available
direct and indirect. With direct calibration, a precisely
known mechanical input is applied to the sensing elements
of the measurement system, and the instrument output is
compared to this for verication or adjustment purposes.
For example, in the case of transducer instrumentation, an
accurately known load (pressure, torque, displacement, etc.)
is applied to the transducer, and the instrument sensitivity
is adjusted as necessary to register the corresponding
output. Direct calibration of instrument systems in this
fashion is highly desirable, but is not ordinarily feasible
for the typical stress analysis laboratory because of the
special equipment and facilities required for its valid
implementa tion. The more practical and widely used
approach to either instrument verication or scaling is by
indirect calibration; that is, by applying a simulated strain
gage output to the input terminals of the instrument. It is
assumed throughout this Tech Note that the input to the
instrument is always through a Wheatstone bridge circuit
as a highly sensitive means of detecting the small resistance
changes which characterize strain gages. The behavior
of a strain gage can then be simulated by increasing or
decreasing the resistance of a bridge arm.
As a rule, strain gage simulation by increasing the resistance
of a bridge arm is not very practical because of the small
resistance changes involved. Accurate calibration would
require inserting a small, ultra-precise resistor in series with
the gage. Furthermore, the electrical contacts for inserting
the resistor can introduce a significant uncertainty in
the resistance change. On the other hand, decreasing
the resistance of a bridge arm by shunting with a larger
resistor offers a simple, potentially accurate means of
simulating the action of a strain gage. This method, known
as shunt calibration, places no particularly severe tolerance
requirements on the shunting resistor, and is relatively
insensitive to modest variations in contact resistance. It is
also more versatile in application and generally simpler to
implement.
Because of its numerous advantages, shunt calibration is
the normal procedure for verifying or setting the output
of a strain gage instrument relative to a predetermined
mechanical input at the sensor. The subject matter of this
Tech Note encompasses a variety of commonly occurring
bridge ci rcuit arrangements and shunt-cal ibration
procedures. In all cases, it should be noted, the assumptions
are made that the excitation for the bridge circuit is
provided by a constant-voltage power supply,
1
and that
the input impedance of any instrument applied across the
output terminals of the bridge circuit is effectively innite.
The latter condition is approximately representative of
most modern strain-measurement instruments in which
the bridge output is balanced by injecting an equal and
opposite voltage developed in a separate network. It is
also assumed that there are no auxiliary resistors (such
as those commonly used in transducers for temperature
compensation, span adjustment, etc.) in either the bridge
circuit proper or in the circuitry supplying bridge power.
Although simple in concept, shunt calibration is actually
much more complex than is generally appreciated. The
full potential of this technique for accurate instrument
calibration can be realized only by careful consideration
of the errors which can occur when the method is misused.
Of primary concern are: (1) the choice of the bridge arm
to be shunted, along with the placement of the shunt
connections in the bridge circuit; (2) calculation of the
proper shunt resistance to simulate a prescribed strain
level or to produce a prescribed instrument output; and
(3) Wheatstone bridge nonlinearity (when calibrating at
high strain levels). Because of the foregoing, different
1
In general, the principles employed here are equally applicable to
constant-current systems, but the shunt-calibration relationships will
differ where nonlinearity considerations are involved.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
132
shunt-calibration relationships are sometimes required for
different sets of circumstances. It is particularly important
to distinguish between two modes of shunt calibration which
are referred to in this Tech Note, somewhat arbitrarily, as
instrument scaling and instrument verication.
In what is described as instrument scaling, the reference
is to the use of shunt calibration for simulating the strain
gage circuit output which would occur during an actual test
program when a particular gage in the circuit is subjected to a
predetermined strain. The scaling is normally accomplished
by adjusting the gain or gage-factor control of the
instrument in use until the indicated strain corresponds
to the simulated strain. The procedure is widely used to
provide automatic correction for any signal attenuation
due to leadwire resistance. In the case of half- and full-
bridge circuits, it can also be employed to adjust the
instrument scale factor to indicate the surface strain under
a singe gage, rather than some multiple thereof. When
shunt calibration is used for instrument scaling, as dened
here, the procedure is not directly related to verifying the
accuracy or linearity of the instrument itself.
By instrument verication, in this context, is meant the
process of using shunt calibration to synthesize an input
signal to the instrument which should, for a perfectly
accurate and linear instrument, produce a predetermined
output indication. If the shunt calibration is performed
properly, and the output indication deviates from the correct
value, then the error is due to the instrument. In such cases,
the instrument may require repair or adjustment of internal
trimmers, followed by recalibration against a standard
such as the our Model 1550A Calibrator. Thus, shunt
calibration for instrument verication is concerned only
with the instrument itself; not with temporary adjustments
in gain or gage factor, made to conveniently account for a
particular set of external circuit conditions.
It is always necessary to maintain the distinction between
instrument scaling and verification, both in selecting a
calibration resistor and in interpreting the result of shunting.
There are also several other factors to be considered in
shunt calibration, some of which are especially important
in scaling applications. The relationships needed to
calculate calibration resistors for commonly occurring
cases are given in the remaining sections of the Tech Note
as follows:
Section Content
II. Basic Shunt Calibration
Derivation of fundamental shunt-calibration
equations.
III. Instrument Scaling for Small Strains
Simple quarter-bridge circuit downscale, upscale
calibration. Half- and full-bridge circuits.
IV. Wheatstone Bridge Nonlinearity
Basic considerations. Effects on strain measurement
and shunt calibration.
V. Instrument Scaling for Large Strains
Quarter-bridge circuit downscale, upscale
calibration. Half- and full-bridge circuits.
VI. Instrument Verication
Small strains. Large strains.
VII. Accuracy Considerations
Maximum error. Probable error.
For a wide range of practical applications, Sections II,
III, and VI should provide the necessary information
and relationships for routine shunt calibration at modest
strain levels. When large strains are involved, however,
reference should be made to Sections IV and V. Limitations
on the accuracy of shunt calibration are investigated in
Section VII. The Appendix to this Tech Note contains a
logic diagram illustrating the criteria to be considered in
selecting the appropriate shunt-calibration relationship for
a particular application.
II. Basic Shunt Calibration
Illustrated in Figure 1 is the Wheatstone bridge circuit in
its simplest form. With the bridge excitation provided by
the constant voltage E, the output voltage is always equal
to the voltage difference between points A and B.

E E
R
R R
E E
R
R R
A
B

+
j
(
,
\
,
(

+
j
(
,
\
,
(
1
1
4
4 3
1
1 2
And,

e E E E
R
R R
R
R R
O A B

+

+
j
(
,
\
,
(
1
1 2
4
4 3

(1)
Or, in more convenient, nondimensional form:

e
E
R R
R R
R R
R R
O
+

+
1 2
1 2
4 3
4 3
1 1
/
/
/
/

(1a)
It is evident from the form of Equation (1a) that the output
depends only on the resistance ratios R
1
/R
2
and R
4
/R
3
, rather
than on the individual resistances. Furthermore, when
R
1
/R
2
= R
4
/R
3
, the output is zero and the bridge is described
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
133
as resistively balanced. Whether the bridge is balanced or
unbalanced, Equation (1a) permits calculating the change
in output voltage due to decreasing any one of the arm
resistances by shunting. The equation also demonstrates
that the sign of the change depends on which arm is shunted.
For example, decreasing R
1
/R
2
by shunting R
1
, or increasing
R
4
/R
3
by shunting R
3
will cause a negative change in output.
Correspondingly, a positive change in output is produced
by shunting R
2
or R
4
(increasing R
1
/R
2
and decreasing R
4
/
R
3
, respectively).
Equation (1a) is perfectly general in application to
constant-voltage Wheatstone bridges, regardless of the
values of R
1
, R
2
, R
3
and R
4
. In conventional strain gage
instrumentation, however, at least two of the bridge arms
normally have the same (nominal) resistance; and all four
arms are often the same. For simplicity in presentation,
without a significant sacrifice in generality, the latter
case, known as the equal-arm bridge, is assumed in the
following, and pictured in Figure 2. The diagram shows a
single active gage, represented by R
1
, and an associated
calibration resistor, R
C
, for shunting across the gage to
produce an output signal simulating strain. The bridge
is assumed to be in an initial state of resistive balance;
and all leadwire resistances are assumed negligibly small
for this introductory development of shunt-calibration
theory. Methods of accounting for leadwire resistance (or
eliminating its effects) are given in Section III.
When the calibration resistor is shunted across R
1
, the
resistance of the bridge arm becomes R
1
R
C
/(R
1
+ R
C
), and
the change in arm resistance is:

R
R R
R R
R
C
C

1
1
1
+

(2)
Or,

R
R
R
R R
C 1
1
1

+

(3)
Reexpressing the unit resistance change in terms of strain
yields a relationship between the simulated strain and the
shunt resistance required to produce it. The result is usually
written here in the form R
C
= f(
s
), but the simulated strain
for a particular shunt resistance can always be calculated
by inverting the relationship.
The unit resistance change in the gage is related to
strain through the denition of the gage factor, F
G
(see
Footnote 2).

R
R
F
G
G

(4)
where:
R
G
= the nominal resistance of the strain gage
(e.g., 120 ohms, 350 ohms, etc.).
Combining Eqs. (3) and (4), and replacing R
1
by R
G
, since
there is no other resistance in the bridge arm,

F
R
R R
G s
G
G C

+
Or,

S
G
G G C
R
F R R

+ ( )

(5)
where:
s
= strain (compressive) simulated by shunting
R
G
with R
C
. Solving for R
C
,
Figure 1 Basic Wheatstone bridge circuit.
Figure 2 Shunt calibration of single active gage.
2
In this Tech Note, the symbol F
G
represents the gage factor of the
strain gage, while F
I
denotes the setting of the gage factor control on
a strain indicator.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
134

R
R
F
R
C
G
G S
G

(6)
Since the simulated strain in this mode of shunt calibration
is always negative, it is common practice in the strain gage
eld to omit the minus sign in front of the rst term in
Equation (6), and write it as:

R
R
F
R
R
F
R
C
G
G S
G
G
G S
G

x 0
6
1
( )

(7)
where:
s
( ) = simulated strain, in microstrain units.
When substituting into Equation (7), the user must
always remember to substitute the numerical value of the
compressive strain, without the sign.
The relationships represented by Equations (5) through (7)
are quite general, and accurately simulate the behavior of
a strain gage for any magnitude of compressive strain. For
convenient reference, Table 1 lists the appropriate shunt-
calibration resistors for simulating strains up to 10000
in 120-, 350-, and 1000-ohm gage circuits, based on a gage
factor of 2.000. Precision resistors (0.02%) in these and
other values are available from Micro-Measurements, and
are described in our Strain Gage Accessories Data Book.
If the gage factor is other than 2.000, or if a nonstandard
calibration resistor is employed, the simulated strain
magnitude will vary accordingly. The true magnitude of
simulated strain can always be calculated by substituting
the exact values of F
G
and R
C
into Equation (5).
While Equations (5) through (7) provide for accurately
simulating strain gage response at any compressive strain
level (as long as the gage factor remains constant), this may
not be sufcient for some calibration applications. It is
always necessary to consider the effects of the Wheatstone
bridge circuit through which the instrument receives its
input signal from the strain gage. If the nondimensional
output voltage of the bridge (
o
/E) were exactly proportional
to the unit resistance change R/R
G
, a perfectly accurate
instrument should register a strain equal to the simulated
strain (at the same gage factor). In fact, however, the
Wheatstone bridge circuit is slightly nonlinear when a
resistance change occurs in only one of the arms (see
Reference 1: Our Tech Note TN-507). Because of this,
the instrument will register a strain which differs from
the simulated (or actual) strain by the amount of the
Table 1 Shunt Calibration Resistors
GAGE
CIRCUIT
RESISTANCE
IN OHMS
EQUIVALENT
MICROSTRAIN
3
120-OHM
599 880
119 880
59 880
29 880
19 880
14 880
11 880
5880
100
500
1000
2000
3000
4000
5000
10 000
350-OHM
349 650
174 650
87 150
57 983
43 400
34 650
17 150
500
1000
2000
3000
4000
5000
10 000
1000-OHM
999 000
499 000
249 000
165 666
124 000
99 000
49 000
500
1000
2000
3000
4000
5000
10 000
3
The Equivalent Microstrain column gives the true compressive
strain, in a quarter-bridge circuit, simulated by shunting each calibra-
tion resistor across an active strain gage arm of the exact indicated
resistance. These values are based on a circuit gage factor setting of
2.000.
Small versus Large Strain
With respect to shunt calibration, at least, the
distinction between small and large strains is purely
relative. Somewhat like beauty, it resides primarily in
the eye of the beholder or the stress analyst.
Errors due to Wheatstone bridge nonlinearity vary
with the circuit arrangement, and with the sign and
magnitude of the simulated strain. As shown in TN-507,
the percentage error in each case is approximately
proportional to the strain. Thus, if the error at a
particular strain level is small enough relative to the
required test precision that it can be ignored, the
strain can be treated as small. If not, the strain is
large, and the nonlinearity must be accounted for to
calibrate with sufcient accuracy.
Since the nonlinearity error at 2000 is normally less
than 0.5%, that level has been taken arbitrarily as the
upper limit of small strain for the purposes of this
Tech Note. The reader should, of course, establish his
or her own small/ large criterion, depending on the
error magnitude compared to the required precision.
The accuracy of the shunt calibration precedure itself
(see Section VII) should be considered when making
such a judgment.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
135
nonlinearity error introduced in the bridge circuit. As a
rule of thumb, the nonlinearity error in this case, expressed
in percent, is about equal to the strain, in percent. Thus, at
low strain levels (below, say, 2000, or 0.2% see inset),
the difference between the simulated and registered strain
magnitudes may not be detectable. For accurate shunt
calibration at higher strain levels, or for precise evaluation
of i nstrument l i nearity, di fferent shunt-cal ibration
relationships may be required. Treatment of nonlinearity
considerations is given in Section IV of this Tech Note.
The procedures described up to this point have referred
only to instrument calibration for compressive strains.
This seems natural enough, si nce shunti ng always
produces a decrease in the arm resistance, corresponding
to compression. There are occasions, however, when
upscale (tension) calibration is more convenient or
otherwise preferable. The easiest and most accurate way to
accomplish this is still by shunt calibration.
Figure 3 illustrates the simple Wheatstone bridge circuit
again, but with the calibration resistor positioned to shunt
the adjacent bridge arm. R
2
(usually referred to as the
dummy in a quarter-bridge circuit). As demonstrated by
Equation (1a), a decrease in the resistance of the adjacent
arm will produce a bridge output opposite in sign to that
obtained by shunting R
1
, causing the instrument to register
a tensile strain. Thus, a simulated compressive strain (
SC2
)
in R
2
, generated by shunting that arm, can be interpreted as
a simulated tensile strain (
ST1
) in R
1
. The special subscript
notation is temporarily introduced here because the two
simulated strains are not exactly equal in magnitude. For
calibration at low strain levels, the difference in magnitude
between
ST1
and
SC2
is small enough that the relationships
given in Equations (5) through (7) are sufciently accurate
for most practical applications. The error in the simulated
tensile strain, in percent, is approximately equal to the
gage factor times the strain, in percent.
The foregoing error arises because shunting R
2
to produce
a simulated compressive strain in that arm, and then
interpreting the instrument output as due to a simulated
tensi le strai n i n R
1
, i nvolves effectively a two-fold
simulation which is twice as sensitive to Wheatstone bridge
nonlinearity. Accounting for the nonlinearity, as shown in
Sections IV and V, permits developing a shunt-calibration
relationship for precisely simulating tensile strains of any
magnitude.
III. Instrument Scaling for Small Strains
Ver y commonl y, when maki ng pract i cal st rai n
measurements under typical test conditions, at least one
active bridge arm is sufciently remote from the instrument
that the leadwire resistance is no longer negligible.
Under these circumstances, the strain gage instrument is
desensitized; and the registered strain will be lower than
the gage strain to an extent depending on the amount of
leadwire resistance. In a three-wire quarter-bridge circuit,
for instance, the signal will be attenuated by the factor
R
G
/(R
G
+ R
L
), where R
L
is the resistance of one leadwire
in series with the gage. The usual way of correcting for
leadwire desensitization is by shunt calibration that is,
by simulating a predetermined strain in the gage, and then
adjusting the gage factor or gain of the instrument until it
registers the same strain.
This section includes a variety of application examples
involving quarter-, half-, and full-bridge strain gage
circuits. In all cases treated here, it is assumed that strain
levels are small enough relative to the users permissible
error limits that Wheatstone bridge nonlinearity can
be neglected. Generalized relationships incorporating
nonlinearity effects are given in subsequent sections.
Quarter-Bridge Circuit
Figure 4 illustrates a representative situation in which
an active gage, in a three-wire circuit, is remote from the
instrument and connected to it by leadwires of resistance
R
L
. If all leadwire resistances are nominally equal, then
R
1
= R
L
+ R
G
and R
2
= R
L
+ R
G
; i.e., the same amount of
leadwire resistance is in series with both the active gage
and the dummy. There is also leadwire resistance in the
bridge output connection to the S instrument terminal.
The latter resistance has no effect, however, since the input
impedance of the instrument applied across the output
terminals of the bridge circuit is taken to be innite. Thus,
no current ows through the instrument leads.
To calibrate in compression, the active gage is shunted
Figure 3 Upscale (tensile) calibration
by shunting adjacent bridge arm.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
136
by a calibration resistor calculated from Equation (7) or
selected from Table 1 for the specied strain magnitude.
After adjusting the sensitivity of the instrument to register
the calibration strain, the effect of the leadwire resistance
is eliminated from all subsequent strain measurements.
Unless additional leadwires are used (as demonstrated in
Figure 6), simulated compressive strain by directly shunting
the remote active gage is usually difcult to implement in
practice. Since the purpose of shunt calibration in this case
is simply to scale the instrument sensitivity as a means
of compensating for leadwire resistance, either upscale
or downscale calibration is equally suitable. Thus, it is
generally more convenient to shunt the adjacent dummy
arm as shown in Figure 4, because this can be done right
at the instrument terminals. It should be apparent from
the gure that the calibration resistor must be connected
directly across the dummy to produce the desired result.
Gage strain cannot be accurately simulated by shunting
from S to P (or from S to P+). After shunting the
dummy with a calibration resistor selected to simulate the
appropriate strain, the instrument sensitivity is adjusted to
register the same strain. At low strain levels, the result is
effectively the same as if the calibration had been performed
by shunting the active gage.
Half-Bridge Circuit
In many stress analysis applications it is necessary (or
at least advantageous) to employ two co-acting gages,
connected at adjacent arms in the bridge circuit, to
produce the required strain signal. A common example
of this occurs when a second gage is installed on an
unstressed specimen of the test material (and maintained
in the same thermal environment as the test object) to
provide temperature compensation for the active gage.
In the special case of a purely uniaxial stress state, with
the principal stress directions known, both gages can
be mounted adjacent to each other, directly on the test
part. One gage is aligned with the applied stress, and
the other is installed in the perpendicular direction to
sense the Poisson strain. This arrangement provides an
augmented bridge output, along with excellent temperature
compensation. Similar opportunities are offered by a beam
in bending. One gage is mounted along the longitudinal
centerline of the convex surface, with a mating gage at the
corresponding point on the concave surface. When the
two gages are connected as adjacent arms in the bridge
circuit, and assuming uniform temperature through the
thickness of the beam, the bridge output is doubled while
maintaining temperature compensation.
All of the foregoing are examples of half-bridge circuits,
since one-half of the Wheatstone bridge is external to
the instrument. Aside from differences in the quality of
the achievable temperature compensation, they differ
principally in their degrees of signal increase. The factor
of signal augmentation is usually expressed in terms of the
number of active gages, N. When the gage in the adjacent
bridge arm senses no applied strain, but serves solely for
temperature compensation, N = 1. With two perpendicular
gages, aligned along the principal axes in a uniaxial stress
eld, N = 1 + , where is the Poissons ratio of the test
material. In the case of the beam, with gages on opposing
surfaces, N = 2, since the gages sense equal and opposite
strains, and the bridge output is doubled.
When N is greater than unity, it is obviously necessary to
adjust the instrument sensitivity by the factor 1/N if the
instrument is to directly register the actual surface strain
sensed by the primary active gage. Furthermore, if the
gage installations are at a distance from the instrument,
additional adjustment of the sensitivity (in the opposite
direction) is required to compensate for the signal loss
due to leadwire resistance. Shunt calibration can correct
for both effects simultaneously, and permit adjusting
instrument sensitivity to register the correct surface strain
at the primary active gage.
Figure 4 Quarter-bridge circuit with active
gage remote from instrument.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
137
Figure 5a illustrates a typical half-bridge circuit, with the
gages located away from the instrument, and with shunt
resistors for downscale (D) and upscale (U) calibration.
Figures 5b and 5c show the physical and circuit arrangements
for N = 1 + and N = 2, respectively. The procedure for
calibration is the same as for the quarter-bridge circuit.
That is, a calibration resistor of the appropriate size is
shunted across the gage, and the instrument sensitivity is
adjusted to register the simulated strain.
Confusion sometimes arises, however, in correlating the
registered strain with the simulated strain (and with the
calibration resistor) when N is greater than unity. The
simplest way of handling this is to generalize Equation (7)
Figure 5 Shunt calibration of
external half-bridge circuits.
5a Basic circuit.
5b Uniform uniaxial stress: N = 1 + 5c Bending beam: N = 2
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
138
so that it includes the number of active gages, Thus,

R
R
F N
R
C
G
G S
G

x 0
x x
6
1
( )

(8)
To calibrate, the strain gage is shunted with a resistor
calculated from Equation (8) and the instrument sensitivity
adjusted to register
s
(). The result is the same, except
for the sign of the instrument output, no matter which of
the two adjacent-arm gages in Figure 5 is shunted. After
calibration, the instrument output will correspond to the
surface strain at the primary active gage. This procedure
accounts for both the signal increase (when N > 1) and the
leadwire desensitization.
When the gage installations are more than a few steps away
from the instrument, it is usually inconvenient to connect
a shunt-calibration resistor directly across the gage as
shown in Figure 5. For such cases, remote shunt calibration
is a common practice. Figure 6 illustrates a half-bridge
circuit with the calibration resistor positioned at the
instrument. In this example, three extra leadwires and a
switch permit connecting the shunt across either arm of
the half bridge. Since shunt resistors are characteristically
in the thousands of ohms, the resis tances of the calibration
leadwires, although shown in the gure, can usually be
neglected in the strain simulation calculations. Equation
(8) is then directly applicable to remote shunt calibration.
If the leadwire resistance is large enough so that 100 x R
L
/
R
C
is greater than about 1/10 of the required calibration
precision (expressed in percent), Equation (8) can be
modied as follows to calculate the correct calibration
resistance:

R
R
F N
R R
C
G
G S
G L

( )
x 0
x x
6
1
2

(9)
In Equation (9), R
L
represents the resistance of one leadwire
between the calibration resistor and gage.
Full-Bridge Circuits
In strain-measurement (stress analysis) applications for
which the half bridge is suitable, the output signal can be
doubled by installing a full bridge, with four active strain
gages on the test object. A representative circuit, including
two supplementary leadwires for remote shunt calibration,
is shown in Figure 7a. In practice, the calibration leadwires
can be connected across any arm of the bridge, and will
always produce the same signal magnitude, but the sign
of the signal depends on which arm is shunted. It will
be noticed, in the case of the full-bridge circuit, that the
leadwire resistance is now in the bridge power and output
leads rather than in the bridge arms. With the assumption
of innite impedance at the bridge output, the resistance
in the latter leads has no effect. However, the resistance in
the power leads reduces the voltage applied to the bridge
proper, and attenuates the output signal accordingly.
Three widely used full-bridge circuit arrangements are
shown in Figures 7b, 7c, and 7d. In the rst two of these, for
6 Remote shunt calibration
of external half-bridge.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
139
7 Remote shunt calibration of
external full-bridge circuit.
7a General circuit
arrangement. Calibration
resistor can be shunted across
any arm. Note that internal half-
bridge is disconnected from
output circuit for full-bridge
operation.
7b Bending beam, with Poisson gage
orientation: N = 2(1 + )
7c Uniform uniaxial stress, with Poisson gage
orientation: N = 2(1 + )
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
140
bending and direct stress, respectively, the number of active
gages is expressed by N = 2(1 + ). This value is substituted
into Equation (8) or Equation (9) when calculating the
calibration resistor to simulate a surface strain of
S
.
The physical arrangement of the gages is the same in
both cases; but, as indicated by the equivalent-circuit
diagrams, the gages are positioned differently in the bridge
circuit to produce the desired signal in each instance. For
applications involving pure bending or torsion, the bridge
output signal can be increased further with the gage and
circuit congurations illustrated in Figure 7d. Since all
four gages are fully active in these examples, N = 4 for
substitution into Equations (8) or (9).
In general, the shunt-calibration relationships appearing
in this section are limited in application to the simulation
of low strain magnitudes, since nonlinearity effects in
the Wheatstone bridge circuit have been ignored. The
equations given here are intended primarily for scaling
the output of an instrument to register the same strain
magnitude that it would if the selected gage were subjected
to an actual strain equal to the simulated strain. This mode
of shunt calibration offers a simple, convenient means for
eliminating the effects of leadwire desensitization and
accounting for more than one active gage (N > 1) in the
bridge circuit.
For calibration at strain levels higher than about 2000,
or for precise evaluation of i nstrument accuracy,
it is ordinarily necessary to incorporate the effects of
Wheatstone bridge nonlinearity in the shunt-calibration
relationships. Nonlinearity considerations are treated in
Section IV, and application examples given in Section V.
IV. Wheatstone Bridge Nonlinearity
As described in TN-507, the common practice with modern
strain gage instruments is to operate the Wheatstone bridge
circuit in a resistively unbalanced mode during strain
measurement. In some instruments, the resulting bridge
output voltage is read directly as a measure of the strain-
induced resistance change(s) in one or more of the bridge
arms. In others, the bridge output signal is unbalanced
(nulled) by injecting an equal and opposite voltage from
a separate circuit which is powered by an equal supply
voltage.
The cause of the nonlinear behavior (when it occurs) can
be demonstrated by reexamining Equation (1a), with
reference to Figure 1. The bridge output voltage under any
initial condition can be expressed as:

e
E
R
R R
R
R R
O
j
(
,
\
,
(

+

+
1
1
1 2
4
4 3

(10)
Considering, for the moment, resistance changes, in R
1

and R
2
(composing the right-handed branch of the bridge
circuit), the output voltage after such changes is:
e
E
R R
R R R R
R
R R
O
j
(
,
\
,
(

+
+ + +

+
2
1 1
1 2 1 2
4
4 3

(11)
7d Bending or torsion: N = 4.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
141
The change in the output signal from the bridge (or the
nulling voltage) is then:
e
E
e
E
e
E
R R
R R
O O O
j
(
,
\
,
(

j
(
,
\
,
(

j
(
,
\
,
(

+
+
2 1
1 1
1 22 1 2
1
1 2
+ +

+ R R
R
R R
(12)
In the usual case, however, R
1
= R
2
= R
G
, the nominal
strain gage resistance. After making this substitution, and
reducing,

e
E
R
R
R
R
R
R
R
R
O G G
G G
j
(
,
\
,
(

+ +
1 2
1 2
4 2 2

(13)
For the quarter-bridge circuit with only a single active
gage (R
1
), R
2
= 0 and:

e
E
R
R
R
R
O G
G
j
(
,
\
,
(

+
1
1
4 2

(14)

Or, introducing the relationship from Equation (4),

e
E
F
F
O G
G
j
(
,
\
,
(

+
4 2

(15)
It is evident from Equations (14) and (15) that in a quarter-
bridge circuit the output is a nonlinear function of the
resistance change and the strain due to the presence
of the second term in the denominator. The nonlinearity
reects the fact that as the gage resistance changes, the
current through R
1
and R
2
also changes, in the opposite
direction of the resistance change. For typical working
strain levels, the quantity 2F
G
in Equation (15) is very
small compared to 4, and the nonlinearity can usually
be ignored. When measuring large strains, or when the
greatest precision is required, the indicated strain must be
corrected for the nonlinearity. With MM Systems 5000,
6000 and 7000, the correction can be made automatically,
and at all strain levels.
Returning to the more general expression for the output
of a half-bridge [Equation (13)], it can be seen that the
nonlinearity terms in the denominator can be eliminated
only by setting R
2
= R
1
. Then, with the resistance
changes in R
1
and R
2
numerically equal, but opposite in
sign, Equation (13) reduces to the linear expression:

e
E
R
R
R
R
O G G
j
(
,
\
,
(

2
4 2
1 1

(16)
Thus, when the separate resistance changes in R
1
and R
2

are such that the total series resistance is unchanged, the
current through R
1
and R
2
remains constant, and the
bridge output is proportional to the resistance change. A
common application of this condition occurs when a beam
in bending is instrumented with a strain gage on the convex
side and another, mounted directly opposite, on the concave
side. The strains sensed by the two gages are then equal in
magnitude and opposite in sign (
2
=
1
). If the two gages
are connected in a half-bridge circuit, as in Figure 5c, the
conditions required for Equation (16) are satised, and the
bridge output, expressed in strain units, is:

e
E
F
O G
j
(
,
\
,
(

2

where || represents the absolute value
of strain in either gage. (17)
This demonstration has dealt with only bridge arms R
1
and
R
2
in Figure 1, but it applies equally to arms R
3
and R
4
. The
principle can be generalized as follows: any combination of
strains and resultant resistance changes in two series bridge
arms (R
1
and R
2
, or R
3
and R
4
) which causes the current in
that branch of the bridge circuit to change will introduce
nonlinearity into the output. Relationships giving the
nonlinearity errors for a variety of commonly used circuit
arrangements are given in TN-507.
With the foregoing principle in mind, we are now in
a position to consider the effect of Wheatstone bridge
nonlinearity on shunt calibration. It should rst be noted
that, in normal practice, only one arm of the bridge is
shunted at a time; and it is never possible, by shunting, to
produce equal and opposite resistance changes in R
1
and
R
2
, or in R
3
and R
4
. Thus, the shunt-calibration procedure
always results in nonlinear, quarter-bridge operation
regardless of how the bridge circuit functions during
actual strain measurement. In Figure 5c, for instance, the
bridge output during strain measurement is proportional
to the surface strain, as indicated by Equation (17). When
either gage is shunted by a calibration resistor, however,
the output is nonlinearly related to the simulated strain
according to Equation (15). As a result, shunting, say, R
G1

in Figure 5c does not exactly simulate, in terms of bridge
output voltage, the behavior of the gage during strain
measurement.
As noted earlier, the effect of the nonlinearity is small
when the strains (actual or simulated) are small. For such
cases, the relationships given in Section III are adequate
to permit calculating the shunt resistor size to simulate a
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
142
given strain magnitude. Because of this, it is preferable to
perform instrument scaling at modest strain levels where
the nonlinearity error is negligible.
When instrument scaling is done at higher strain levels, it
is generally necessary to use special relationships, given
in Section V, to precisely simulate gage behavior by shunt
calibration. There is one notable exception to the latter
statement, however. The bridge output due to shunting
a single gage is indistinguishable from that of a quarter-
bridge circuit with the gage strained in compression. For
this special (but common) case, the simulation is exact at
all compressive strain levels because the nonlinearity due
to shunting is the same as that caused by compressive strain
in the gage. As shown in the Appendix, Equations (5) to (7)
are thus appropriate for compressive scaling of quarter-
bridge circuits at any level of strain.
4
The same is true, of
course, for external half- and full-bridge circuits where
there is only a single active gage, with the remaining bridge
arms used for compensation and/or bridge-completion
purposes.
Both Sections III and V are limited in scope to the subject
of instrument scaling by gage simulation. Shunt calibration
for instrument verication is treated separately in Section
VI. For scaling applications, the si ze of the shunt-
calibration resistor is selected so that the bridge output
voltage is the same for both simulated and actual strains
of the same magnitude. Therefore, when the Wheatstone
bridge arrangement is intrinsically nonlinear (as in Figure
5b, for instance), and the strain level is high, the instrument
indication is accurate only at the simulated strain level.
Subsequent correction may be needed if the instrument is
to accurately register smaller or larger strains.
V. Instrument Scaling for Large Strains
It was demonstrated i n the precedi ng section that
Wheatstone bridge nonl i nearity must general ly be
considered when shunt calibration is used for instrument
scaling at high strain levels. Under such conditions, errors
in gage simulation arise whenever the nonlinearity which
is inherent in shunt calibration differs from that during
actual strain measurement. The relationships given in this
section for quarter-, half-, and full-bridge circuits provide
for precisely simulating the strain gage output at any level of
strain, low or high.
5
Thus, the instrument scaling will also
be precise when the gain or gage factor control is adjusted
to register the simulated strain. It must be kept in mind,
however, that if the strain-measuring circuit arrangement
is nonlinear (as in Figures 4, 5b and 7c), precise scaling is
achieved only at the simulated strain level. At any other
level of strain, some degree of error will be present due to
the nonlinearity.
Quarter-Bridge Circuit
The quarter-bridge circuit, with a single active gage,
is widely used in experimental stress analysis. When
i nstrument scal i ng is done by connecti ng a shunt-
calibration resistor directly across the gage, the simulation
of compressive strain is exact at all strain levels. This is
true because the nonlinearity in shunt calibration is the
same as that during strain measurement. For such cases,
the proper shunt-calibration resistor to simulate a given
strain magnitude can be obtained directly from Table
1, or calculated from Equation (7) of Section II. After
instrument scaling, the indicated strain will be correct at
the magnitude of the calibration strain, but slightly in error
at other strain levels because of the nonlinearity. For most
practical applications, the corrected strain at any different
strain level can be calculated from:

2 %
2 + F
G

S
% ( )

(18)
where: = corrected strain

S
= calibration strain

2 %
2 + F
G

S
% ( )= indicated strain
F
G
= gage factor of strain gage
Since the effect of leadwire resistance on bridge circuit
nonlinearity is normally very small, terms involving R
L
have
been omitted from Equation (18). If the leadwire resistance
is a signicant fraction of the gage resistance, however,
Equation (18) tends to overcorrect for the nonlinearity. In
such cases, the following complete relationship can be used
to obtain more accurate correction:

2 1+
R
L
R
G
j
(
,
\
,
(
%
2 1+
R
L
R
G
j
(
,
\
,
(
+ F
G

S
% ( )

(18a)

Shunting the dummy arm of the bridge (see Figure 4)
produces an upscale signal, and can be used to simulate a
tensile strain in the active gage. For the simulation to be
exact, however, a special shunt-calibration relationship is
required, because the nonlinearity in tension is different
4
Since the nonlinearity due to a resistance increase is different than
for a decrease, precise simulation of a high tensile strain requires a
special relationship, as demonstrated in Section V.
5
The subject relationships are precise or exact with respect to the
specied parameters such as R
G
, R
C
, and F
G
. The effects of toler-
ances on these quantities are discussed in Section VII, Accuracy
Considerations.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
143
from that in compression. If the active gage were subjected
to an actual tensile strain, the resistance of the right-hand
branch of the bridge in Figure 4 would rise, and the current
would decrease correspondingly. However, when the
dummy arm of the bridge is shunted, the resistance of the
branch decreases, and the current rises. This difference can
be accounted for by calculating the calibration resistor so
that the bridge output voltage due to shunting the dummy is
the same as that for a preselected tensile calibration strain in
the active gage. The procedure for doing so is demonstrated
by the following derivation where, for the sake of simplicity,
the effect of leadwire resistance is temporarily ignored
(RL = 0).
Equation (14) in Section IV gives the output voltage from
a resistance change in the active gage (R
1
). The negative of
the same relationship applies to a change in the dummy
arm, R
2
. Thus,

e
E
R
R
R
R
O G
G
j
(
,
\
,
(

+
1
1
1
4 2

(19)

e
E
R
R
R
R
O G
G
j
(
,
\
,
(

+
2
2
2
4 2

(20)
The unit resistance change in the active gage due to a
simulated tensile strain
ST1
is:

R
R
F
G
G ST
1
1

Therefore,

e
E
F
F
O G ST
G ST
j
(
,
\
,
(

+
1
1
1
4 2

(21)
On the other hand, the resistance change in the dummy,
R
2
, is produced by shunting with a calibration resistor,
R
C
. From Equation (3),

R
R
R
R R
G
G
G C
2

+
Substituting into Equation (20),

e
E
R
R R
R
R R
R
R R
O
G
G C
G
G C
G
G C
j
(
,
\
,
(

+
+
2
4 2
2 4

(22)
Equating the two expressions for output voltage,

F
F
R
R R
G ST
G ST
G
G C
1
1
4 2 2 4 +

+
And, solving for R
C
,

R
R
F
R
F
C
G
G ST
G
G ST

( )
1 1
x 10
6

(23)
For the majority of routine applications, any desired
tensile strain in the active gage can be simulated quite
accurately by shunting the dummy gage with the calibration
resistor specied by Equation (23). This relationship can be
compared to Equation (7) to see the difference between
simulating tensile and compressive strains in the active
gage. After scaling for a given simulated tensile strain, the
instrument indication will be correspondingly accurate for
only that tensile strain magnitude. Measurements at other
strain levels (tension or compression) can be corrected, if
necessary, with Equation (18) or (18a).
In extreme cases, when the simulated strain is very large,
and the leadwire resistance is not a negligible fraction of
the gage resistance, slightly greater accuracy in tensile
strain simulation can be achieved by incorporating the
leadwire resistance in the derivation of Equation (23). The
complete expression for the calibration resistor becomes:

R
R
F
R
R
R
R
R
C
G
G ST
G
L
G
L
G

+
j
(
,
,
,
\
,
(
(
(
1
1

(23a)
The second term in Equation (23a) can never be greater
in magnitude than R
G
; and, for typical strain levels, is
negligible compared to the rst term irrespective of
the leadwire resistance. Thus Equation (23) is normally
the appropriate relationship for the shunt resistor used
in upscale calibration. The small error associated with
Figure 8 Percent error from using Equation
(23) instead of Equation (23a). F
G
= 2.0.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
144
Equation (23) is plotted in Figure 8 as a guide to the
very rare circumstances when Equation (23a) might be
necessary.
It is worth noting, for quarter-bridge circuits, that scaling
the instrument by shunting the internal dummy gage (which
is usually a stable precision resistor) can offer distinct
advantages in calibration accuracy. It is common practice,
for instance, to calculate or select the value of the shunt-
calibration resistor on the basis of the nominal gage
resistance. But the resistance of the instal led gage
generally differs from the nominal, due both to its initial
resistance tolerance and to a further change in resistance
during installation. When this occurs, and the active
gage is shunted for compression scaling, the simulated
magnitude is in error accordingly. The extent of the error
can be approximated by the method given in Section VII,
Accuracy Considerations.
One technique for avoiding most of the error due to
deviation in the gage resistance is to temporarily replace
the active gage in the bridge circuit with a precision
resistor equal to the nominal resistance of the gage. The
instrument is then scaled (in compression) by shunting the
xed resistor with a calibration resistor calculated from
Equation (7). After scaling, the active gage is reconnected
to the bridge circuit. It is usually much more convenient,
however, and about equally accurate, to scale in the
tension direction by simply shunting the internal dummy
with a resistor calculated from Equation (23). When the
leadwire resistance is negligible, this procedure is exact,
and independent of the installed gage resistance. Even
with modest leadwire resistance (say, less than R
G
/10),
the error due to a few ohms of gage resistance deviation is
small enough to be ignored. In case of doubt, the installed
gage resistance should be measured. If the resistance is
signicantly beyond the manufacturers tolerance, one of
the two foregoing procedures should always be used for
shunt calibration.
Half-Bridge Circuits
When measuring the maximum principal strain in a
known uniaxial stress state, a simple means for assuring
effective temperature compensation is to mount a second
gage adjacent and perpendicular to the primary gage, and
connect the two gages in a half-bridge circuit as shown in
Figure 5b. Such an arrangement is said to have N = 1 +
active gages, since the bridge output is increased by that
factor. The circuit behavior is nonlinear, however, because
the resistance changes in the two active gages are not equal
and opposite.
It is assumed in the following that R
G1
in Figure 5b
represents the primary gage, and that the object is to scale
the instrument to register the test-surface strain under that
gage. Whether shunting R
G1
to simulate compression in
the primary gage, or shunting R
G2
to simulate tension, the
nonlinearity during scaling is different from that during
actual strain measurement. Thus, two different shunt-
calibration relationships are required for precise strain
simulation, as in the case of the quarter-bridge circuit.
These relationships are developed in the same manner as
before; that is, by enforcing the condition that the bridge
output voltage be identical, whether scaling to a simulated
strain level or measuring the same surface strain with the
primary gage.
To simulate a compressive surface strain,
SC1
, by shunting
R
G1
in Figure 5b, the calibration resistor is calculated
from:

R
R
F
R
R
R
C
G
G SC
G
L
G
+ ( )

+ ( ) +
j
(
,
\
,
(
,
1
1
1
1 1
,,
,
,
,
]
]
]
]
]
]
]

(24)
And, for a simulated tensile strain,
ST1
, generated by
shunting R
G2
.

R
R
F
R
R
R
C
G
G ST
G
L
G
+ ( )

+ ( ) +
j
(
,
\
,
(
,
1
1
1
1
1 1
,,
,
,
,
]
]
]
]
]
]
]

(25)
When substituting into Equations (24) and (25), the
signs of the simulated strains must always be carried,
of course. Also, it is apparent that leadwire resistance,
if present, affects the nonlinear behavior, and must be
known to permit exact simulation. Since the expressions
are relatively insensitive to the quantity R
L
/R
G
, precision
measurement of the leadwire resistance is not ordinarily
required. After scaling the instrument at a simulated
strain,
S
, the registered strain is precisely correct for only
that same magnitude. When necessary, the corrected strain
at any level can be calculated from:

2 1+
R
L
R
G
j
(
,
\
,
(
%
2 1+
R
L
R
G
j
(
,
\
,
(
+ F
G
1 ( )
S
% ( )

(26)
The notation in Equation (26) is the same as in Equation
(18) for the quarter-bridge circuit.
Another half-bridge application of interest is illustrated in
Figure 5c, where directly opposed strain gages are installed
on the convex and concave sides of a rectangular-cross-
section beam in bending. In this instance, the strains in
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
145
R
G1
and R
G2
are always equal and opposite, if only bending
occurs. As a result, the bridge behavior during strain
measurement is linear; and, after scaling at a particular
strain level, remains equally precise at all other strain
magnitudes. Instrument scaling by shunt calibration is a
nonlinear procedure, however, because there is a resistance
change in only a single bridge arm. The simplest approach
is to perform the scaling at a modest strain level where the
error due to calibration nonlinearity is negligible. After
scaling with Equation (8), the instrument will register any
other strain with the same relative precision.
If scaling at a high strain level is necessary, the calibration
resistor can be calculated as follows to provide exact
simulation of a surface strain,
S
in R
G1
(when accompanied
by the strain
S
in R
G2
):

R
R
sF
R
R
R
C
G
G S
G
L
G

+
j
(
,
\
,
(
,
,
,
,
,
,
]
]
]
]
]
]
]
1
1
2 1

(27)
where: |
S
| = absolute value of calibration strain.
Equation (27) is suitable for either upscale or downscale
shunt calibration. If the leadwire resistance is negligible,
the relationship reduces to:

R
R
sF
R
C
G
G S
G

2

(28)
Because the bridge output voltage varies linearly with
strain when actual strains are being measured, no further
correction is required.
Full-Bridge Circuits
When feasible, use of the full-bridge circuit offers several
advantages, including a better signal-to-noise ratio. Typical
applications are: beams in bending, shafts in torsion,
and axially loaded columns and tension links. Although
the simple examples described here do not incorporate
the circuit refinements characteristic of commercial
transducers, it is common practice to infer the magnitudes
of mechanical variables such as bending moment, torque,
and force from the full-bridge strain measurement.
Three representative full-bridge arrangements, illustrated
in Figures 7b, 7c, and 7d, are treated here. The circuits in
Figures 7b and 7d (for bending and torsion) have essentially
the same characteristics, and can be grouped together for
shunt-calibration purposes. In each of these circuits, the
bridge output voltage varies linearly with strain, since
equal and opposite resistance changes occur in arms 1 and
2, and in arms 3 and 4. The nonlinearity of shunt calibration
must be accounted for, however, to achieve exact strain
stimulation at large strains. The proper calibration resistor
to simulate a given surface strain (e.g., the longitudinal
strain, in the case of the beam) can be calculated from the
following:
R
R
NF
R
R
R
R
R
C
G
G S
G
L
G
L
G

+
+
j
(
,
\
,
(
,
,
,
,
,
,
]
]
1 3
2 1 2
]]
]
]
]
]

(29)
Once calibrated according to Equation (29), an accurate
instrument will register the correct strain at any other
strain magnitude. As in the case of the half-bridge circuit,
the leadwire resistance is present in the calibration
relationship, but does not need to be known with high
precision.
The arrangement shown in Figure 7c, for a centrally loaded
column or tension member, is somewhat more complex.
It can be seen from the gure that the bridge currents
change with applied strain, and thus the output voltage is
a nonlinear function of strain even before calibrating with
a shunt resistor. Because the nonlinearity is different in
tension and compression, separate calibration equations
are required as follows:
To simulate compression in R
G1
,
R
R
F
R
R
R
C
G
G SC
G
L
G

+ ( )

+ ( ) +
j
(
,
\
,
(

2 1
3 1 1 2 2
1

44 1 1 2 + ( ) +
j
(
,
\
,
(
,
,
,
,
,
,
]
]
]
]
]
]
]
R
R
L
G

(30)

And, for tension in R
G1
,
R
R
F
R
R
R
C
G
G ST
G
L
G
+ ( )

+ ( ) +
j
(
,
\
,
(

2 1
3 1 1 2 2
4 1
1

++ ( ) +
j
(
,
\
,
(
,
,
,
,
,
,
]
]
]
]
]
]
]
1 2
R
R
L
G

(31)
Equations (30) and (31) provide for correctly simulating
the longitudinal surface strain in compression and tension,
respectively, at any strain magnitude. In common with
the corresponding half bridge (Figure 5b), the instrument
indication at the calibration strain level is precise, but
operation at a different strain level will introduce a small
error due to bridge-circuit nonlinearity. When necessary,
the corrected strain can be calculated from [see Equation
(18) for notation]:
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
146

2 1+ 2
R
L
R
G
j
(
,
\
,
(
%
2 1+ 2
R
L
R
G
j
(
,
\
,
(
+ F
G
1 ( )
S
% ( )

(32)
VI. Instrument Verication
The shunt-calibration procedures described in Sections
III and V are intended specically for instrument scaling
purposes; that is, for adjusting the instrument output to
match a given simulated surface strain. They are not directly
related to the question of verifying the linearity and/or
absolute accuracy of a strain-measuring instrument. It is
implicitly assumed in the preceding sections of this each
Tech Note that the instrument involved is perfectly linear
in its response characteristics and, if direct-indicating, is
perfectly accurate. In practice, however, it is necessary
to periodically verify the accuracy of the instrument by
calibration; and methods for accomplishing this are given
here.
6
By far the most reliably accurate means for instrument
verication is through the use of a laboratory-standard
calibrator such as our Model 1550A. This instrument,
which incorporates true tension and compression strain
si mulation, provides precision calibration of strain
indicators to an accuracy of 0.025 percent. It also eliminates
errors due to the tolerances on the strain gage and shunt
resistances. The calibrator is equipped with three decades
of switches, which permit rapid calibration in small steps
over a very wide strain range (to ~100000).
Whether verication of the strain indicator is to be done
with a precision calibrator or by shunt calibration, it is
important that the procedure be unaffected by leadwire
resistance. When verifying instrument accuracy with
the Model 1550A, for instance, the calibrator should
be connected to the strain indicator with short leads of
generous wire size. Similarly, with shunt calibration, the
leadwire resistance in the shunted bridge arm should be
negligibly small. This can be accomplished, for calibration
purposes, by connecting an installed strain gage or a stable
precision resistor directly across the active gage terminals
of the strain indicator. Either the active or dummy arm
of the bridge circuit can then be shunted to produce,
correspondingly, a downscale or upscale calibration signal.
If the active arm is a strain gage, and is to be shunted, the
installed resistance of the gage must be known accurately.
An alternative approach, which eliminates the effect of
leadwire resistance, is to shunt one arm of the internal
half-bridge commonly found in conventional strain gage
instruments. This procedure requires, of course, that
the resistances of the internal bridge arms be known.
In addition, it requires that the internal half-bridge be
isolated from any balance circuitry which may be present,
or that the effects of such circuitry be incorporated in the
shunt-calibration calculations. In any case, the instruction
manual and circuit diagram for the instrument should be
consulted before attempting to calibrate by shunting the
internal half-bridge.
The calibration relationship for instrument verication
is based on different reasoning than it is for instrument
scaling. In scaling applications (Sections III and V), the
calibration resistor is calculated to develop the same
bridge output voltage that would occur when a strain gage
of specied gage factor is subjected to a given strain. The
instrument gage factor or gain control is then adjusted to
register the simulated strain. The effects of signal loss due to
leadwire resistance, or signal increase form multiple active
gages, are thus compensated for. With this technique, the
nal setting of the gage-factor or gain control is determined
only by the external circuit parameters; and, in the case of
a strain indicator, for example, the resulting gage factor
setting of the instrument would normally be quite different
from that of the strain gage.
In contrast, when calibrating for instrument verication
purposes, the instrument gage factor or gain is ordinarily
preset to some convenient value. The veri f ication
relationship is then written to express the registered strain
(in a perfectly accurate instrument) as a function of the
shunt resistor used to synthesize the strain signal. It will
be seen that the gage factor of the strain gage itself is not
involved in this process. Nor are other external circuit
parameters, except the initial resistance of the shunted
bridge arm, which is usually the nominal resistance of a
strain gage.
In an ideal instrument, the registered strain is related to the
bridge output voltage by:

I
O
C
e
E
x

(33)
where:
1
= strain magnitude indicated or registered
by the ideal instrument.
C = instrument constant at a xed setting
of the gain or gage-factor control.
But the bridge output caused by a unit resistance change
in one arm can be expressed as:

e
E
R
R
R
R
O G
G

+
4 2

(34)
6
As used in this section only, the term "calibration" thus refers
exclusively to the process of instrument verication for linearity
or accuracy.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
147
In Equation (34), the choice of sign depends on which arm
is being shunted. Referring to Figure 1, the sign is positive
for R
1
or R
3
, and negative for R
2
or R
4
. The unit resistance
change in shunt calibration is always negative, however,
and is calculated from Equation (3):

R
R
R
R R
G
G
G C

+
Substituting into Equation (34), and simplifying,

e
O
E
m
R
G
4R
C
+ 2R
G

(35)
Thus, the bridge output is negative when shunting R
1
or
R
3
, and positive for R
2
or R
4
.
Substituting Equation (35) into Equation (33),

I
mC x
R
G
4R
C
+ 2R
G
Or, in general, when shunting any arm of initial resistance
R
G
,

I
G
C G
C
R
R R
+
x
4 2

(36)
And,

R
C R R
G
I
G
C
x
4 2

(37)
where: |
I
| = absolute value of registered strain (ideal).
In the case of a strain indicator with a gage-factor control,
the instrument is designed so that C = 4/F
I
, where F
I
is the
instrument gage factor setting. Then,

R
R
F
R
G
I I
G
C
x

2

(38)
When one arm of the bridge is shunted by a calibration
resistor calculated from Equation (38), the instrument
should indicate the synthesized strain,
I
. Failure to do so
by more than the tolerances on R
G
and R
C
is indicative of
instrument error.
Instead of a strain indicator, the instrumentation may
consist of a signal-conditioning amplier. This type of
instrument is normally equipped with a gain control rather
than a gage-factor control. Its output is simply a voltage
which can then be supplied to an oscillograph or other
device for recording. In the ideal instrument, the voltage
at any gain setting should be strictly proportional to the
bridge output signal. Thus, corresponding to Equation
(33),

V C
e
E
O

(39)
The object of calibration in this instance is to verify the
instrument linearity; that is, to test whether C is, in fact,
constant. Calibration is accomplished by comparing the
measured instrument output voltage to the bridge output
signal at a series of different signal levels. Substituting
Equation (35) into Equation (39),

V mC x
R
G
4R
C
+ 2R
G
Or, in general,

V
R
R R
C
G
C G
4 2 +
j
(
,
\
,
(

(40)
After shunting one arm of the bridge with a calibration
resistor, R
C
, the instrument output voltage is measured,
and the constant, C, calculated from Equation (39). This
operation is repeated at two or more different signal levels
by successively shunting with appropriate calibration
resistors. If the instrument is linear, variations in the
calculated value of the instrument constant should not be
greater than the tolerances on the parameters in Equation
(40).
VII. Accuracy Consideration
As described in the preceding sections, shunt calibration
can be used for either system scaling or instrument
verication purposes. In both cases, the greatest attainable
accuracy with the procedure is limited by errors (deviations
from the nominal or assumed value) in the variables which
enter into the calibration calculations. The error sensitivity
of the method can be demonstrated most easily with a
generalized form of the basic shunt-calibration relationship
[see Equations (3), (4) and (5)].
Let R
1
in Equation (3) represent the actual resistance of the
strain gage, after installation. The factor R
G
in Equation
(4) is replaced by a numerical constant, C, to emphasize the
fact that the nominal resistance of the gage is not changed
by gage installation. Then, the relationships in Equations
(3) and (4) can be reexpressed as:

R
R
R R
C
+
1
2
1

(3a)

R CF
G

(4a)
Combining Equations (3a) and (4a), and solving for the
simulated strain,
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
148

S
G C
R
CF R R
+ ( )
1
2
1

(41)
The total differential of the simulated strain can be
written:

d
R
R
R
R
F
F
S
S S
C
C
S
G
G

1
1
After performing the partial differentiations and dividing
through by
S
= R
1
2
/CF
G
(R
1
+ R
C
),

d R R
R R
dR
R
R
R R
d
S
S
C
C
C
C

+
+
j
(
,
\
,
(

+
j
(
,
\
,
(
1
1
1
1 1
2 RR
R
dF
F
C
C
G
G

(42)
For small deviations, the differentials can be replaced by
nite differences, or:

S
S
C
C
C
C
R R
R R
R
R
R
R R
+
+
j
(
,
\
,
(

+
j
(
,
\
,
(
1
1
1
1 1
2 RR
R
F
F
C
C
G
G

(43)
When multiplied by 100, Equation (43) gives the percent
error in the simulated strain as a function of the errors or
deviations in R
1
(the actual gage resistance), R
C
, and F
G
.
Since R
C
is ordinarily very large compared to R
1
, it can
be seen that the percent error in simulated strain is about
twice that in the gage resistance, and is approximately
equal to that in the calibration resistor and gage factor
(although the signs may differ). In practice, the errors in R
1
,
R
C
, and F
G
vary independently over their respective ranges
of tolerance or uncertainty. Thus, they may tend to be self-
canceling on some occasions; and, at other times, may be
additive. The worst-case errors in simulated strain occur
when R
1
is positive while R
C
and F
G
are negative, and
vice versa. These conditions can be combined into a single
expression by employing the absolute values of the errors:
S
S
C
C
C
C
R R
R R
R
R
R
R R
MAX
+
+
j
(
,
\
,
(
+
+
j
(
,
\
1
1
1
1 1
2
,,
(
+
R
R
F
F
C
C
G
G

(44)
Equation (44) permits calculating the extreme error in
simulated strain from the extreme errors in the other
variables. Practically, however, the extreme errors in R
G
,
R
C
, and F
G
would occur only rarely at the same time, and
with the required combination of signs, to be fully additive.
A better measure of the approximate uncertainty (expected
error range) in S as a function of the uncertainties or
tolerances on the other three quantities can be obtained
by an adaptation from the theory of error propagation.
The latter theory is not strictly applicable in this case
because the individual error distributions are unknown,
are probably different from one another, and may otherwise
violate statistical requirements of the method. However, if
the uncertainties in each variable represent about the same
number of standard deviations, the following expression
should give a more realistic estimate of the uncertainty in
S
than Equation (44):
U
R R
R R
UR
R
R
R R
S
S
C
C
C

+
+
j
(
,
\
,
(
j
(
,
\
,
(
+
+
1
1
2
1
1
2
1
2
CC
C
C
G
G
UR
R
UF
F
j
(
,
\
,
(
j
(
,
\
,
(
+
j
(
,
\
,
(
2 2
(45)
where:
U
X
X
= percent uncertainty in variable X.
As a numerical example, assume that a 350-ohm gage with
a gage factor of 2.0 is to be shunted to simulate a strain of
2000. From Equation (7).

R
C

350 x 10
2.0 x 2000
350 87150ohms
6

This calibration resistor can be found in Table 1; and,
if supplied by Micro-Measurements, has a resistance
tolerance of 0.01%. Assume also that the selected gage
type has tolerances on its gage factor and resistance of
1% and 0.3%, respectively. Since the gage resistance
may have been shifted during installation, however, the
uncertainty in the installed resistance should normally
be taken somewhat larger say, to be conservative, 1%.
Substituting into Equation (45),
U
S
S
( ) ( ) + ( ) ( ) + 1 996 0 996 1
2 2 2 2
. . . x 1.0 x 0.01 00 2 23
2
( ) . %
Equations (44) and (45) will be found helpful guides in
estimating the precision of shunt calibration. They can
also serve in judging whether use of the large-strain
relationships in Section V is warranted under any given
set of circumstances. Thus, if the intrinsic uncertainty in
shunt calibration is many times greater than the renement
obtained by considering large-strain effects, the simpler
relationships in Section III may as well be employed.
When necessary, the overall uncertainty can be reduced
somewhat by accurately measuring the installed gage
resistance and employing this value in the shunt-calibration
equations. Or, alternatively, the effect of resistance deviat-
ion in the gage can be largely eliminated by the methods
described in Section V for quarter-bridge calibration.
T
E
C
H

N
O
T
E
TN-514
Micro-Measurements
149
Tech Note TN-515
MICRO-MEASUREMENTS
Strain Gage Rosettes:
Selection, Application and Data Reduction
T
E
C
H

N
O
T
E
151
Revision 12-Sep-2010
1.0 Introduction
A strain gage rosette is, by denition, an arrangement
of two or more closely positioned gage grids, separately
oriented to measure the normal strains along different
directions in the underlying surface of the test part.
Rosettes are designed to perform a very practical and
important function in experimental stress analysis. It can
be shown that for the not-uncommon case of the general
biaxial stress state, with the principal directions unknown,
three independent strain measurements (in different
directions) are required to determine the principal strains
and stresses. And even when the principal directions are
known in advance, two independent strain measurements
are needed to obtain the principal strains and stresses.
To meet the foregoi ng requi rements, the Mi cro-
Measurements manufactures three basic types of strain
gage rosettes (each in a variety of forms):
Tee: two mutually perpendicular grids.
45-Rectangular: three grids, with the second and
third grids angularly displaced from the rst grid by
45 and 90, respectively.
60-Delta: three grids, with the second and third grids
60 and 120 away, respectively, from the rst grid.
Representative gage patterns for the three rosette types are
reproduced in Figure 1.
In common with single-element strain gages, rosettes
are manufactured from different combinations of grid
alloy and backing material to meet varying application
requirements. They are also offered in a number of gage
lengths, noting that the gage length specied for a rosette
refers to the active length of each individual grid within
the rosette. As illustrated in Figure 2, rectangular and
delta rosettes may appear in any of several geometrically
different, but functionally equivalent, forms. Guidance
in choosing the most suitable rosette for a particular
application is provided in Section 2.0, where selection
considerations are reviewed.
Figure 1 Basic rosette types, classied by grid orientation: (a) tee; (b) 45-rectangular; (c) 60 delta.
5X
(a)
5X
(b)
5X
(c)
Rectangular
Delta
Figure 2 Geometrically different, but functionally
equivalent congurations of rectangular and delta rosettes.
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
152
Strain Gage Rosettes: Selection, Application and Data Reduction
Since biaxial stress states occur very commonly in machine
parts and structural members, it might be presumed that
half or so of the strain gages used in experimental stress
analysis would be rosettes. This does not seem to be the
case, however, and ten percent (or less) rosette usage may be
more nearly representative. To what degree this pattern of
usage reects an inclination for on-site makeup of rosettes
from single-element gages, or simply an undue tendency to
assume uniaxiality of the stress state, is an open question.
At any rate, neither practice can generally be recommended
for the accurate determination of principal strains.
It must be appreciated that while the use of a strain
gage rosette is, in many cases, a necessary condition
for obtaining the principal strains, it is not a sufcient
condition for doing so accurately. Knowledgeability in
the selection and application of rosettes is critical to their
successful use in experimental stress analysis; and the
information contained in this Tech Note is intended to help
the user obtain reliably accurate principal strain data.
2.0 Rosette Selection Considerations
A comprehensive guide for use in selecting Micro-
Measurements strain gages is provided in Reference 1. This
publication should rst be consulted for recommendations
on the strain-sensitive alloy, backing material, self-
temperature-compensation number, gage length, and
other strain gage characteristics suitable to the expected
application. In addition to basic parameters such as the
foregoing, which must be considered in the selection of any
strain gage, two other parameters are important in rosette
selection. These are: (1) the rosette type tee, rectangular,
or delta; and (2) the rosette construction planar (single-
plane) or stacked (layered).
The tee rosette should be used only when the principal strain
directions are known in advance from other considerations.
Cylindrical pressure vessels and shafts in torsion are two
classical examples of the latter condition. However, care
must be exercised in all such cases that extraneous stresses
(bending, axial stress, etc.) are not present, since these
will affect the directions of the principal axes. Attention
must also be given to nearby geometric irregularities,
such as holes, ribs, or shoulders, which can locally alter
the principal directions. The error magnitudes due to
misalignment of a tee rosette from the principal axes are
given in Reference 2. As a rule, if there is uncertainty about
the principal directions, a three-element rectangular or
delta rosette is preferable. When necessary (and, using the
proper data-reduction relationships), the tee rosette can be
installed with its axes at any precisely known angle from the
principal axes; but greatest accuracy will be achieved by
alignment along the principal directions. In the latter case,
except for the readily corrected error due to transverse
sensitivity, the two gage elements in the rosette indicate the
corresponding principal strains directly.
Where the directions of the principal strains are unknown,
a three-element rectangular or delta rosette is always
required; and the rosette can be installed without regard
to orientation. The data-reduction relationships given in
Section 4.0 yield not only the principal strains, but also the
directions for the principal axes relative to the reference
grid (Grid 1) of the rosette. Functionally, there is little
choice between the rectangular and delta rosettes. Because
the gage axes in the delta rosette have the maximum
possible uniform angular separation (effectively 120), this
rosette is presumed to produce the optimum sampling of
the underlying strain distribution. Rectangular rosettes
have historically been the more popular of the two,
primarily because the data-reduction relationships are
somewhat simpler. Currently, however, with the widespread
access to computers and programmable calculators, the
computational advantage of the rectangular rosette is
of little consequence. As a result of the foregoing, the
choice between rectangular and delta rosettes is more
apt to be based on practical application considerations
such as availability from stock, compatibility with the
space available for installation, convenience of solder tab
arrangement, etc.
All three types of rosettes (tee, rectangular, and delta)
are manufactured in both planar and stacked versions.
As indicated (for the rectangular rosette) in Figure 3, the
Figure 3 Rectangular rosettes (of the same gage
length) in planar and stacked construction.
Stacked
Planar
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
153
planar rosette is etched from the strain-sensitive foil as
an entity, with all gage elements lying in a single plane.
The stacked rosette is manufactured by assembling and
laminating two or three properly oriented single-element
gages. When strain gradients in the plane of the test part
surface are not too severe, the normal selection is the
planar rosette. This form of rosette offers the following
advantages in such cases:
Thin and fexible, with greater conformability to
curved surfaces;
Minimal reinforcing effect;
Superior heat dissipation to the test part;
Available in all standard forms of gage construction,
and generally accepts all standard optional features;
Optimal stability;
Maximum freedom in leadwire routing and
attachment.
The principal disadvantages of the planar rosette arise from
the larger surface area covered by the sensitive portion of
the gage. When the space available for gage installation
is small, a stacked rosette may t, although a planar one
will not. More importantly, where a steep strain gradient
exists in the surface plane of the test part, the individual
gage elements in a planar rosette may sense different strain
elds and magnitudes. For a given active gage length, the
stacked rosette occupies the least possible area, and has
the centroids (geometric centers) of all grids lying over
the same point on the test part surface. Thus, the stacked
rosette more nearly approaches measurement of the strains
at a point. Although normally a trivial consideration, it can
also be noted that all gages in a stacked rosette have the
same gage factor and transverse sensitivity, while the grids
in a planar rosette will differ slightly in these properties,
due to their different orientations relative to the rolling
direction of the strain-sensitive foil. The technical data
sheet accompanying the rosettes fully documents the
separate properties of the individual grids.
It should be realized, however, that the stacked rosette
is noticeably stiffer and less conformable than its planar
counterpart. Also, because the heat conduction paths for
the upper grids in a stacked rosette are much longer, the heat
dissipation problem may be more critical when the rosette
is installed on a material with low thermal conductivity.
Taking into account their poorer heat dissipation and their
greater reinforcement effects, stacked rosettes may not be
the best choice for use on plastics and other nonmetallic
materials. A stacked rosette can also give erroneous strain
indications when applied to thin specimen in bending, since
the grid plane of the uppermost gage in a three-gage stack
may be as much as 0.0045 in [0.11 mm] above the specimen
surface. In short, the stacked rosette should ordinarily
be reserved for applications in which the requirement for
minimum surface area dictates its selection.
3.0 Gage Element Numbering
Numbering, as used here, refers to the numeric (or
alphabetic) sequence in which the gage elements in a
rosette are identified during strain measurement, and
for substitution of measured strains into data-reduction
relationships such as those given in Section 4.0. Contrary to a
widely held impression, the subject of gage numbering is not
necessarily a trivial matter. It is, in fact, basic to the proper,
and complete, interpretation of rosette measurement.
3

With any three-element rosette, misinterpretation of the
rotational sequence (CW or CCW), for instance, can lead
to incorrect principal strain directions. In the case of the
rectangular rosette, an improper numbering order will
produce completely erroneous principal strain magnitudes,
as well as directions. These errors occur when the gage
users numbering sequence differs from that employed in
the derivation of the data-reduction relationships.
To obtai n correct results usi ng the data-reduction
relationships provided in Section 4.0 (and in the Appendix),
the grids in three-element rosettes must be numbered in
a particular way. It is always necessary in a rectangular
rosette, for instance, that grid numbers 1 and 3 be assigned to
two mutually perpendicular grids. Any other arrangement
will produce incorrect principal strains. Following are
the general rules for proper rosette numbering. With a
rectangular rosette, the axis of Grid 2 must be 45 away
from that of Grid 1; and Grid 3 must be 90 deg away, in the
same rotational direction. Similarly, with a delta rosette,
the axes of Grids 2 and 3 must be 60 and 120 away,
respectively, in the same direction from Grid 1.
In principle, the preceding rules could be implemented
by numberi ng the grids i n either the clockwise or
counterclockwise direction, as long as the sequence is
correct. Counterclockwise numbering is preferable,
however, because it is consistent with the usual engineering
practice of denoting counterclockwise angular measure-
ment as positive in sign. The gage grids in all Micro-
Measurements general-purpose, three-element planar
rosettes (rectangular and delta) are numerically identied,
and numbered i n the counterclockwise di rection.*
Examples of the grid numbering for several representative
rosette types are illustrated in Figure 4. At rst glimpse, it
might appear that gage patterns (b) and (c) are numbered
clockwise instead of counterclockwise. But when these
patterns are examined more closely, and when the axis
of Grid 2 is transposed across the grid-circle diameter to
satisfy the foregoing numbering rules, it can be seen that
the rosette numbering is counterclockwise in every case.
* Micro-Measurements also supplies special-purpose planar rectangu-
lar rosettes designed exclusively for use with the hole-drilling method
of residual stress analysis. Since these rosettes require different
data-reduction relationships, procedures, and interpretation, they
are numbered clockwise to distinguish them from general-purpose
rosettes.
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
154
Micro-Measurements stacked rosettes are not numbered,
as a matter of manufacturing economics. The user should
number the gages in stacked rosettes according to the rules
given here and illustrated in Figures 2 and 4.
4.0 Principal Strains and Directions
from Rosette Measurements
The equations for calculating principal strains from three
rosette strain measurements are derived from what is
known as a strain-transformation relationship. As
employed here in its simplest form, such a relationship
expresses the normal strain in any direction on a test
surface in terms of the two principal strains and the angle
from the principal axis to the direction of the specied
strain. This situation can be envisioned most readily with
the aid of the well-known Mohrs circle for strain, as shown
in Figure 5**. It can be seen from Figure 5a (noting that the
angles in Mohrs circle are double the physical angles on
the test surface) that the normal strain at any angle from
the major principal axis is simply expressed by:

=
+
+
P Q P Q
2 2
2
cos

(1)
Figure 4 Counterclockwise numbering of grids
in Micro-Measurements general-purpose
strain gage rosettes (see text).
** The Mohrs circle in Figure 5 is constructed with positive shear
strain plotted downward. This is done so that the positive rotational
direction in Mohrs circle is the same (CCW) as for the rosette, while
maintaining the usual sign convention for shear (i.e., positive shear
corresonds to a reduction in the initial right angle at the origin of
the X-Y axes as labeled in Figure 5b).
Figure 5 Strain transformation from the principal strains to the
strain in any direction: (a)
in terms of principal strains

P
, and
Q
, as shown by Mohrs circle for strain; (b) rectangular rosette
installed on a test surface, with Grid 1 at the arbitrary angle
from the major principal axis; (c) axes of the rectangular rosette
superimposed on Mohrs circle for strain.
(a)
(b)
(c)
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
155
Figure 5b represents a small area of the test surface, with
a rectangular rosette installed, and with the reference grid
(#1) oriented at degrees from
p
. Mohrs circle, with the
axes of the rosette superimposed, is shown in Figure 5c.
By successively substituting into Equation (1) the angles for
the three grid directions, the strain sensed by each grid can
be expressed as follows:

1
2
2 2
2
2 2
2
=
+
+
=
+
+ +
P Q P Q
P Q P Q
cos
cos 445
2 2
2 90
3
o
o
( )
=
+
+ +
( )
P Q P Q
cos

(2a)

(2b)

(2c)
When the rosette is installed on a test part subjected to
an arbitrary strain state, the variables on the right-hand
side of Equations (2) are unknown. But the strains
1
,
2

and
3
can be measured. Thus, by solving Equations (2)
simultaneously for the unknown quantities
P
,
Q
, and ,
the principal strains and angle can be expressed in terms of
the three measured strains. Following is the result of this
procedure:

P Q ,

=
+
( ) + ( )
=
1 3
1 2
2
2 3
2
1
2
1
2
1
2
tan
1
22
2 3
1 3

+

(3)

(4)
If the rosette is properly numbered, the principal strains
can be calculated from Equation (3) by substituting the
measured strains for
1
,
2
and
3
. The plus and minus
alternatives i n Equation (3) yield the algebraical ly
maximum and minimum principal strains, respectively.
Unambiguous determination of the principal angle from
Equation (4) requires, however, some interpretation,
as described in the following. To begin with, the angle
represents the acute angle from the principal axis to
the reference grid of the rosette, as indicated in Figure
5. In the practice of experimental stress analysis, it is
somewhat more convenient, and easier to visualize, if this
is reexpressed as the angle from Grid 1 to the principal axis.
To change the sense of the angle requires only reversing the
sign of Equation (4). Thus:

P Q ,

= =

1
2
2
2 1 3
1 3
tan
1

(5)
The physical direction of the acute angle given by either
Equation (4) or Equation (5) is always counterclockwise
if positive, and clockwise if negative. The only difference
is that is measured from the principal axis to Grid 1,
while is measured from Grid 1 to the principal axis.
Unfortunately, since tan 2 tan 2( + 90), the calculated
angle can refer to either principal axis; and hence the
identification in Equation (5) as
P,Q
. This ambiguity
can readily be resolved (for the rectangular rosette) by
application of the following simple rules:
(a) if
1
>
3
, then
P,Q
=
P
(b) if
1
<
3
, then
P,Q
=
Q
(c) if
1
=
3
and
2
<
1
, then
P,Q
=
P
= 45
(d) if
1
=
3
and
2
>
1
, then
P,Q
=
P
= +45
(e) if
1
=
2
=
3
, then
P,Q
is indeterminate (equal
biaxial strain).
The reasoning which underlies the preceding rules becomes
obvious when a sketch is made of the corresponding Mohrs
circle for strain, and the rosette axes are superimposed as
in Figure 5c. A similar technique for assuring correct
interpretation of the angle is given in the form of a step-by-
step algorithm in Reference 3.
The preceding development has been applied to the
rectangular rosette, but the same procedure can be used
to derive corresponding data-reduction equations for the
delta rosette shown in Figure 6. When grid angles , +
60, and + 120 are successively substituted into Equation
(1), the resulting three equations can again be solved
simultaneously for the principal strains and angle. Thus,
for the delta rosette:

P Q ,
=
+ +
( ) + ( ) + ( )
1 2 3
1 2
2
2 3
2
3 1
2
3
2
3

(6)

=
( )
1
2
3
2
3 2
1 2 3
tan
1

(7)
As in the case of Equation (4), the angle calculated from
Equation (7) refers to the angular displacement of Grid 1
from the principal axis. The sense of the angle can again be
changed by reversing its sign to give the angle from Grid 1
to a principal axis:

P Q ,

= =
( )
1
2
3
2
2 3
1 2 3
tan
1

(8)
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
156
In every case [Equations (4), (5), (7), and (8)], the angles
are to be interpreted as counterclockwise if positive, and
clockwise if negative.
Equation (8) embodies the same ambiguity with respect to
the tan 2 and tan 2( + 90) as Equation (5). As before, the
ambiguity can easily be resolved (for the delta rosette) by
considering the relative magnitudes (algebraically) among
the individual strain measurements; namely:
(a)if then
(b)if

1
2 3
1
2
>
+
=
<
,
, P Q P
22 3
1
2 3
2
2
+
=
=
+

,
,
,
then
(c)if and
P Q Q

2 1
1
2
45 ,
,
< = =
=
+
then
(d)if
o
P Q P
33
2 1
2
45 , ,
,
and then
(3)if
o
> = = +
P Q P
then isindeterminate

1 2 3
= =
, P Q
((equalbiaxialstrain)
When the principal angle is calculated automatically by
computer from Equation (5) or Equation (8), it is always
necessary of course, to avoid the condition of division by
zero if
1
=
3
with a rectangular rosette, or
1
= (
2
+
3
)/2
with a delta rosette. For this reason, the computer should
be programmed to perform the foregoing (c) and (d) tests,
in each case, prior to calculating the arc-tangent.
Once the principal strains have been determined from
Equation (3) or Equation (6), the strain state in the surface
of the test part is completely defined. If desired, the
maximum shear strain can be obtained directly from:

MAX
=
P

Q
(9)
Intuitive understanding of the strain state can also be
enhanced by sketching the corresponding Mohrs circle,
approximately to scale. In Equations (3) and (6), the rst
term represents, in each case, the distance from the origin
to the center of the circle, and the second term represents
the radius, or
MAX/2
. With the angle calculated, further
insight can be gained by superimposing the rosette grid
axes on the Mohrs circle, as in Figures 5c and 6b.
5.0 Principal Stresses from Principal Strains
As previously noted, a three-element strain gage rosette
must be employed to determine the principal strains in
a general biaxial stress state when the directions of the
principal axes are unknown. The usual goal of experimental
stress analysis, however, is to arrive at the principal stresses,
for comparison with some criterion of failure. With the
strain state fully established as described in Section 4.0,
the complete state of stress (in the surface of the test part)
can also be obtained quite easily when the test material
meets certain requirements on its mechanical properties.
Since some types of strain gage instrumentation, such as
Figure 6 Delta rosette: (a) installed on a test surface,
with Grid 1 at the angle of from the major principal strain
direction; (b) rosette grid axes superimposed on Mohrs
circle for strain. Note that Grid 2 is to be viewed as +60
(CCW) from Grid 1 in the rosette, and +120 in Mohrs circle.
(a)
(b)
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
157
our System 6000, calculate both the principal strains and
the principal stresses, the following material is provided as
background information.
If the test material is homogeneous in composition, and is
isotropic in its mechanical properties (i.e., the properties
are the same in every direction), and if the stress/strain
relationship is linear, with stress proportional to strain,
then the biaxial form of Hookes law can be used to convert
the principal strains into principal stresses. This procedure
requires, of course, that the elastic modulus (E) and
Poissons ratio () of the material be known. Hookes law
for the biaxial stress state can be expressed as follows:

P P Q
E
= +
( )
1
2

(10a)

Q Q P
E
= +
( )
1
2

(10b)
The numerical values of the principal strains calculated
form Equation (3) or Equation (6) can be substituted into
equations (10), along with the elastic properties, to obtain
the principal stresses. As an alternative, Equation (3)
or Equation (6) depending on the rosette type) can be
substituted algebraically into Equations (10) to express the
principal stresses directly in terms of the three measured
strains and the material properties. The results are as
follows:
Rectangular:

P Q
E
,
=
+
+
( ) + ( )
2 1
2
1
1 3
1 2
2
2 3
2

(11)

Delta:

P Q
E
,

=
+ +
+
( ) + ( ) +
3
1
2
1
1 2 3
1 2
2
2 3
2
3

1
2
( )

(12)
When the test material is isotropic and linear-elastic in
its mechanical properties (as required for the validity of
the preceding strain-to-stress conversion), the principal
stress axes coincide in direction with the principal strains.
Because of this, the angle from Grid 1 of the rosette to
the principal stress direction is given by Equation (5)
for rectangular rosettes, and by Equation (8) for delta
rosettes.
6.0 Errors, Corrections, and Limitations
The obvious aim of experimental stress analysis is to
determine the signicant stresses in a test object as accurately
as necessary to assure product reliability under expected
service conditions. As demonstrated in the preceding
sections of this Tech Note, the process of obtaining the
principal stresses involves three basic, and sequential, steps:
(1) measurement of surface strains with a strain gage rosette;
(2) transformation of measured strains to principal strains;
and (3) conversion of principal strains to principal stresses.
Each step in this procedure has its own characteristic error
sources and limits of applicability; and the stress analyst
must carefully consider these to avoid potentially serious
errors in the resulting principal stresses.
Of frst importance is that the measured strains be as free
as possible of error. Strain measurements with rosettes
are subject, of course, to the same errors (thermal output,
transverse sensitivity, leadwire resistance effects, etc.) as
those with single-element strain gages. Thus, the same
controlling and/or corrective measures are required to
obtain accurate data. For instance, signal attenuation
due to leadwire resistance should be eliminated by shunt
calibration
4
, or by numerically correcting the strain data for
the calculated attenuation, based on the known resistances
of the leadwires and strain gages.
Because at least one of the gage grids in any rosette will in
every case be subjected to a transverse strain which is equal
to or greater than the strain along the grid axis, consideration
should always be given to the transverse-sensitivity error
when performing rosette data reduction. The magnitude of
the error in any particular case depends on the transverse-
sensitivity coefficient (K
t
) of the gage grid, and on the
ratio of the principal strains (
P
/
Q
). In general, when
K
t
1%, the transverse-sensitivity error is small enough to be
ignored. However, at larger values of K
t
, depending on the
required measurement accuracy, correction for transverse
sensitivity may be necessary. Detailed procedures, as well
as correction equations for all cases and all rosette types,
are given in Reference 5.
When strain measurements must be made in a variable
thermal environment, the thermal output of the strain gage
can produce rather large errors, unless the instrumentation
can be zero-balanced at the testing temperature, under
strain-free conditions. In addition, the gage factor of the
strain gage changes slightly with temperature. Reference
6 provides a thorough treatment of errors due to thermal
effects in strain gages, including specic compensation and
correction techniques for minimizing these errors.
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
158
After making certain that strain measurement errors such
as the foregoing have been eliminated or controlled to the
degree feasible, attention can next be given to possible
errors in the strain-transformation procedure for obtaining
the principal strains. A potentially serious source of error
can be created when the user attempts to make up a rosette
on the specimen from three conventional single-element
gages. The error is caused by misalignment of the individual
gages within the rosette. If, for example, the second and
third gages in a rectangular rosette conguration are not
accurately oriented at 45 and 90, respectively, from the
rst gage, the calculated principal strains will be in error.
The magnitude of the error depends, of course, on the
magnitude (and direction) of the misalignment; but it
also depends on the principal strain ratio,
P
/
Q
, and on
the overall orientation of the rosette with respect to the
principal axes. For certain combinations of principal
strain ratio and rosette orientation, 5-degree alignment
errors in gages 2 and 3 relative to Gage 1 can produce an
error of 20 percent or more in one of the principal strains.
Since it is very difcult for most persons to install a small
strain gage with the required precision in alignment, the user
is well-advised to employ commercially available rosettes.
The manufacturing procedures for Micro-Measurements
strain gage rosettes are such that errors due to grid
alignment within the rosette need never be considered. For
those cases in which it is necessary, for whatever reason, to
assemble a rosette from single-element gages, extreme care
should be exercised to obtain accurate gage alignment.
And when the principal strain directions can be predicted
in advance, even approximately, alignment of Gage 1 or
3 in a rectangular rosette, or alignment of any gage in a
delta rosette, with a principal axis, will minimize the error
in that principal strain caused by inter-gage misalignment.
The strain-transformation relationships and data-reduction
equations given in Section 4.0 assume a uniform state of
strain at the site of the rosette installation. Since the rosette
necessarily covers a nite area of the test surface, severe
variations in the strain eld over this area can produce
signicant errors in the principal strains particularly
with planar rosettes.
7
For this type of application, the
stacked rosette is distinctly superior; both because it
covers a much smaller area (for the same gage length), and
because the centroids of all three grids lie over the same
point on the test surface.
The requirements for a homogeneous material and
uniform strain state can be (and are) relaxed under certain
circumstances. A case in point is the use of strain gage
rosettes on ber-reinforced composite materials. If the
distance between inhomogeneities in the material (i.e.,
ber-to-ber spacing) is small compared to the gage length
of the rosette, each grid will indicate the macroscopic or
average strain in the direction of its axis. These measured
strains (after the usual error corrections) can be substituted
into Equation (3) or Equation (6) to obtain the macroscopic
principal strains for use in the stress analysis of test objects
made from composite materials.
8
As noted later in this
section, however, Equations (10)-(12) cannot be used for
this purpose.
There is an additional limitation to the strain-transformat-
ion relationship in Equation (1) which, although not
frequently encountered in routine experimental stress
analysis, should be noted. The subject of the strain
distribution about a point, as universally treated in
handbooks and in mechanics of materials textbooks, is
developed from what is known as innitesimal-strain
theory. That is, in the process of deriving relatively simple
relationships such as Equation (1), the strain magnitudes
are assumed to be small enough so that normal- and
shear-strain approximations of the following types can be
employed without introducing excessive error:
+
2
(13)
sin tan (14)
Although often unrecognized, these approximations
are embodied in the equations used throughout the
contemporary practice of theoretical and experimental
stress analysis (where strain transformation is involved).
This includes the concept of Mohrs circle for strain,
and thus all of the equations in Section 4.0, which are
consistent with the strain circle. Innitesimal-strain theory
has proven highly satisfactory for most stress analysis
applications with conventional structural materials, since
the strains, if not truly infinitesimal, are normally
very small compared to unity. Thus, for a not-untypical
working strain level of 0.002 (2000), the error in ignoring
2
compared to is only about 0.2 percent.
However, strain gage rosettes are sometimes used in the
measurement of much larger strains, as in applications
on plastics and elastomers, and in post-yield studies of
metal behavior. Strain magnitudes greater than about 0.01
(10 000) are commonly referred to as large or nite,
and, for these, the strain-transformation relationship in
Equation (1) may not adequately represent the actual
variation in strain about a point. Depending on the strain
magnitudes involved in a particular application, and on
the required accuracy for the principal strains, it may be
necessary to employ large-strain analysis methods for
rosette data reduction.
9
The nal step in obtaining the principal stresses is the
introduction of Hookes law [Equations (10)] for the biaxial
stress state. To convert principal strains to principal
stresses with Hookes law requires, of course, that the
elastic modulus and Poissons ratio of the test material
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
159
be known. Since the calculated stress is proportion to E,
any error in the elastic modulus (for which a 3 to 5 percent
uncertainty is common) is carried directly through to the
principal stress. An error in Poissons ratio has a much
smaller effect because of the subordinate role of in the
relationship.
It is also necessary for the proper application of Hookes
law that the test material exhibit a linear relationship
between stress and strain (constant E) over the range
of working stresses. There is normally no problem in
satisfying this requirement when dealing with common
structural materials such as the conventional steel and
aluminum alloys. Other materials (e.g., some plastics,
cast iron and magnesium alloys, etc.) may, however, be
distinctly nonlinear in their stress/strain characteristics.
Since the process of transformation from measured strains
to principal strains is independent of material properties,
the correct principal strains in such materials can be
determined from rosette measurements as described in
this Tech Note. However, the principal strains cannot be
converted accurately to principal stresses with the biaxial
Hookes law if the stress/strain relationship is perceptibly
nonlinear.
A further requirement for the valid application of Hookes
law is that the test material be isotropic in its mechanical
properties (i.e., that the elastic modulus and Poissons
ratio be the same in every direction). Although severely
cold-worked metals may not be perfectly isotropic, this
deviation from the ideal is commonly ignored in routine
experimental stress analysis. In contrast, high-performance
composite materials are usually fabricated with directional
fiber reinforcement, and are thus strongly directional
(orthotropic or otherwise anisotropic) in their mechanical
properties. Hookes law as expressed in Equations (10) is
not applicable to these materials; and special constitutive
relationships are required to determine principal stresses
from rosette strain measurements.
8
References
1. Micro-Measurements Tech Note TN-505, Strain Gage
Selection Criteria, Procedures, Recommendations.
2. Micro-Measurements Tech Note TN-511, Errors Due
to Misalignment of Strain Gages.
3. Perry, C.C., Data-Reduction Algorithms for
Strain Gage Rosette Measurements, Experimental
Techniques. May, 1989, pp. 13-18.
4. Micro-Measurements Tech Note TN-514, Shunt
Calibration of Strain Gage Instrumentation.
5. Micro-Measurements Tech Note TN-509, Errors Due
to Transverse Sensitivity in Strain Gages.
6. Micro-Measurements Tech Note TN-504, Strain Gage
Thermal Output and Gage Factor Variation with
Temperature.
7. Troke, .W., Flat versus Stacked Rosettes,
Experimental Mechanics, May, 1967, pp. 24A-28A.
8. Tsai, S.W. and H.T. Hahn, Introduction to Composite
Materials, Technomic Publishing Company, 1980.
9. Meyer, M.L., Interpretation of Surface-Strain
Measurements in terms of Finite Homogeneous
Strains. Experimental Mechanics, December, 1963,
pp. 294-301.
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
160
Appendix
Derivation of Strain-Transformation Relationship
[Equation (1) in text] from Deformation Geometry
Consider a small area of a test surface, as sketched in
Figure A-1. The line O-P, of length L
O
, and at the angle
from the X axis, is scribed on the surface in the unstrained
state. When uniform principal strains
P
and
Q
are applied
in the directions of the X and Y axes, respectively, the point
P moves to P as a result of the displacements X and Y
(greatly exaggerated in the sketch).
It is evident from the gure that:
X =
P
(L
O
cos ) (A-1)
Y =
Q
(L
O
sin ) (A-2)
It can also be seen (Figure A-2), by enlarging the detail in
the vicinity of points P and P, that for small strains:
L X cos + Y sin (A-3)
Substituting from Equations (A-1) and (A-2),
L L
O
(
P
cos
2
+
Q
sin
2
) (A-3)
Or,

=

= +
L
Lo
P Q
cos sin
2 2

(A-5)
But,

cos cos
sin cos
2
2
1
2
1 2
1
2
1 2

= + ( )
= ( )
After substituting the above identities,

=
+
+
P Q P Q
2 2
2
cos

(A-6)
Alternate Data Reduction Equations
In the extensive technical literature dealing with strain
gage rosettes, the user will often encounter data-reduction
relationships which are noticeably different from one
another, and from those in the body of this Tech Note. As
a rule, these published equations yield the same results,
and differ only in algebraic format although proving
so in any given case may be rather time consuming.
Since certain forms of the equations may be preferred for
mnemonic reasons, or for computational convenience,
several alternative expressions are given here. All of the
following are equally correct when the gage elements in the
rosette are numbered as described in this Tech Note.
Figure A-1 Figure A-2
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
161
Rectangular Rosette:
where:
P Q
P Q
,
,
=
+
( ) + + ( )
1 3
1 3
2
2 1 3
2
2
1
2
2
==
+
( ) + ( )
=

1 3
1 2
2
2 3
2
1
2
2 /
, P Q
C C (( ) + ( )
=
+
2
2
2
1 3
2
C
C

(A-7)
(A-8)
(A-9)
(A-9)
Delta Rosette:

(A-10)

(A-11)

(A-12)
where:

P Q ,
=
+ +
+ ( )
+ (
1 2 3 1 2 3
2
2 3
3
2
3
1
3
))
=
+ +
( ) + ( ) + (
2
1 2 3
1 2
2
2 3
2
3 1
3
2

P Q ,
))
= ( ) + ( ) + ( )

2
1
2
2
2
3
2
9
2
/

,

P Q
C C C C
=
+ +
/ 3
3
1 2 3
C

Cartesian Strain Components from Rosette Measurements
It is sometimes desired to obtain the Cartesian components of strain (
X
,
Y
, and
XY
) relative to a specied set of X-Y
coordinate axes. This need can arise, for example, when making strain measurements on orthotropic composite materials.
The Cartesian strain components are also useful when calculating principal strains from rosette data using matrix
transformation methods.*
When the X axis of the coordinate system coincides with the axis of the reference grid (Grid 1) of the rosette, the Cartesian
components of strain are as follows:
Rectangular Rosette:

X
=
1

Y
=
3

XY
= 2
2
(
1
+
3
)
Y
3
2
1
X
* Milner, R.R., A Modern Approach to Principal Stresses and Strains, Strain, November, 1989, pp.135-138.
T
E
C
H

N
O
T
E
TN-515
Micro-Measurements
162
Delta Rosette:

/
/

X
Y
XY
=
= + ( )
= ( )
1
2 3 1
2 3
2 3
2 33

The foregoing assumes in each case that the gage elements in the rosette are numbered counterclockwise as indicated. When
the calculated
XY
is positive in sign, the initial right angle at the origin of the X-Y coordinate system is decreased by the
amount of the shear strain.
Y
3 2
1
X
Tech Note TN-516
MICRO-MEASUREMENTS
Errors Due to Shared Leadwires in
Parallel Strain Gage Circuits
T
E
C
H

N
O
T
E
163
1.0 Introduction
The usual, and preferred, practice with multiple quarter-
bridge strain gage installations used for either static or
combined static/dynamic measurements is to employ
a separate three-wire circuit for each gage. However,
if a number of such gages are connected to a multiple-
channel instrument which simultaneously uses the same
power supply for several channels, the associated bridge
circuits (each of which contains an active and dummy
gage) are effectively in parallel. This arrangement, in
itself, need not cause any problems, provided the power
supply has sufficient capacity to maintain a constant
voltage under varying load. If the two individual current-
carrying P+ and P power supply leadwires in each circuit
have the same resistance and are subjected to the same
temperature, their only contribution to measurement error
is the usual desensitization of the gage factor. But this error
can readily be eliminated by shunt calibration.
Unfortunately, the stress analyst may be motivated under
certain circumstances to use a current-carrying leadwire
that, as shown in Figure 1, is shared by, or common to,
all the active gages. Savings of leadwire can be realized
with this arrangement (sometimes called a chevron)
when the runs between gages and instrumentation are
long. And, savings in manhours of installation time will
be obtained when the number of installations is large.
But, considering the potential problems created by the
use of a common leadwire, the only valid motivations are
those arising from physical and mechanical limitations.
These may include the number of slip rings available for
measurements on rotating equipment; the number of pass-
through conductors possible in a barrier (like the wall of a
pressure vessel) between the instrument and gages; and the
use of multiple-grid gages with an integral bus or solder tab
(common-tab rosettes and certain types of strip gages).
Figure 1. Schematic of parallel
Wheatstone bridge circuits with
common power supply leadwire.
T
E
C
H

N
O
T
E
TN-516
Micro-Measurements
164
Errors Due to Shared Leadwires in Parallel Strain Gage Circuits
A fundamental problem with the use of a common leadwire
is that all data are vulnerable to degradation or even to
complete loss should a single gage (or grid) malfunction.
These malfunctions can range from a short circuit within
a gage, to a low resistance to ground, to an open circuit.
Although a primary reason for avoiding common leadwire
usage, the risk of data loss is not directly relevant to the
following discussion, and will not be treated further.
The more dangerous aspects of common leadwire usage
arise from the often subtle effects that are produced when
the gages function properly. Problems that can result
from these include large initial resistive imbalances of the
Wheatstone bridge circuits, inaccurate shunt calibration,
crosstalk between gage circuits during strain measurement,
and loss of leadwire temperature compensation. These are
the primary subjects of the discussion that follows.
2.0 The Signal from Parallel Circuits
The electrical output, e
o
e
i
, from each of the active
gage circuits shown schematically in Figure 1 depends
upon the power supply voltage, E
P
, and the resistances
of the common leadwire (R
LC
), the active gages (R
G
i
), the
individual return leadwires (R
L
i
), and the dummy gages
(R
D
i
). The resistances of the signal leads are relatively
unimportant because no signicant amount of current
ows through them when a modern instrument with a
high impedance input circuit is used to measure the signal
voltage.
The resistance of the n parallel circuits between points A
and C can be expressed as:

R
R R R
A C
G L D
i
n
i i i

+ +
j
(
,
\
,
(
1
1
1

(1)
provided that the resistance of the leadwire between the
active gages (R
G
i
and R
G
i+1
) is negligible. Because the
common leadwire has some nite resistance, it acts as a
voltage divider to reduce the excitation voltage supplied
to the active and dummy gages. And because it carries
the sum of the currents in all the parallel circuits, the
voltage drop in the common leadwire is n times as great
as for individual return leadwires with the same resistance
(provided all active and dummy gages have nominally
the same resistance). The fraction, H, of the power supply
voltage (E
P
) available to the parallel circuits between
points A and C for any combination of resistances in the
parallel circuits is:

H
E
E
R
R R
A C
P
A C
A C LC
= =

+
( )
(2)
where E
AC
is the actual bridge excitation voltage across
the parallel circuits (assumed to be the same for all).
The significance of this expression is that the current
through the common leadwire and consequently the
bridge excitation voltage at any given moment between
points A and C depends upon not only the resistance
of the common leadwire, the individual leadwires, and
the dummy resistors, but also upon the instantaneous
resistances of all the independently variable active gages
in the parallel network. The effect that this phenomenon
produces in the bridge output will be referred to in the
following discussion as crosstalk.
Applying the voltage division fraction, H, to the active
half-bridge term of the usual expression for output from a
Wheatstone bridge, e
o
e
i
, the signal from any active gage
in the parallel circuit in Figure 1 can be calculated for any
combination of resistance values:

e e E
H R R
R R R
o i P
L D
G L D
i i
i i i

+
( )
+ +
,
,
,
]
]
]
]
1
2
(3)
3.0 Initial Imbalance
Because the common leadwire does not affect the voltage
across the internal half bridge, H is not applied to the 1/2
term in Equation (3). This gives rise to the problem of an
initial imbalance in every circuit, even when the active and
dummy gages are of the same resistance. To illustrate the
magnitude of the initial imbalance, consider the case of n
parallel circuits in which all active and dummy gages are
of the same resistance, R
G
; and all leadwires, including
the common leadwire, have the same resistance, R
L
. If the
instrument gage factor control is set at 2.000, the initial
imbalance, in microstrain units, is:

I
L G
L G
n R R
n R R
( )
+ +
( )

1
2 1
6
10 (4)
Equation (4) is plotted in Figure 2 (on page 165) for
various combinations of the parameters n and R
L
/R
G
.
As demonstrated by the gure, the imbalance can easily
exceed the balance range of commercial strain indicators
and signal-conditioning ampliers.
4.0 Calibration Errors
Shunt calibration of the individual quarter-bridge circuits
to adjust the instrument sensitivity would normally be
done by shunting the dummy in one circuit, under the
condition of zero output from the remaining parallel
circuits. The use of a common leadwire causes no errors
in the actual calibration process itself. However, when
subsequent strain measurements are made, the strain-
T
E
C
H

N
O
T
E
TN-516
Micro-Measurements
165
induced resistance changes in the individual gages
produce changes in the values of R
AC
, H, and ultimately
E
AC
. Consequently, the changes in E
AC
will cause the
bridge output to vary, even when the resistance of the
active gage corresponds to the calibration value. The
calibration factor between resistance change and output
voltage is then no longer correct for the calibrated circuit
and the indicated strains will be in error. Accordingly,
the calibration factor is generally correct only for the
calibration conditions; namely, when the current through
the common leadwire is the same as during calibration.
The calibration error produced when the parallel gages
are strained does not lend itself to generalization, but is
symptomatic of the crosstalk between circuits treated in
the following section.
5.0 Measurement Crosstalk Errors
Crosstalk refers to changes in both sensitivity and output
produced in all parallel circuits by a resistance change in
any one of the circuits. As in the case with the calibration
error, this occurs because the resistance changes in each
of the parallel circuits affect the voltage applied toand
consequently, the output fromall the other circuits.
Output errors can be partially generalized to yield a
cross-talk sensitivity index like that shown in Figure 3.
The approximate incremental output that will occur in
each of the other circuits as the result of a 1000r change
in any one circuit is given by this graph. Note that this
incremental output is not directly related to the magnitude
of any other output that may be present due to strain in the
affected circuits. Rather, these incremental outputs can be
thought of as output shifts that are algebraically added
to each of the indicated strains in the affected circuits.
When the extent of the incremental outputs caused by
crosstalk must be known more precisely, Equation (3)
should be used to calculate the outputs of all circuits
under any set of instantaneous resistances. The errors
produced by crosstalk are relatively small in comparison to
those that can arise from the loss of leadwire temperature
compensation as described in the following section.
6.0 Leadwire Temperature Errors
In a single, isolated, three-wire, quarter-bridge circuit, if
the leadwires are of the same resistance and subjected to
the same temperature change, there is essentially no false
circuit output due to leadwire thermal effects. Since one
of the leadwires is in the active arm of the bridge, and the
other is in the dummy arm, the effects of the resistance
changes cancel. The remaining error, due to a small
change in leadwire desensitization of the gage factor, is
normally small enough to be ignored. But this situation is
entirely different when a common leadwire arrangement is
employed.
The single aspect of common leadwire usage with the
most potential for producing serious errors is the loss of
leadwire temperature compensation. The copper normally
used in leadwires has a rather high thermal coefcient of
100
90
80
70
60
50
40
30
20
10
0
0
.
0
0
5
0
.
0
1
0
0
.
0
1
5
0
.
0
2
0
0
.
0
2
5
0
.
0
3
0
0
.
0
3
5
0
.
0
4
0
0
.
0
4
5
0
.
0
5
0

I
m
b
a
l
a
n
c
e

(
1
0
0
0

)

n = 2
n

=

1
0
n
=
3
n

=
5
Initial R
L
/R
G
Figure 2. Initial imbalance produced by the use of a common
leadwire when all active and dummy gages are of the same
resistance, common and individual leadwires are of the same
resistance, and the instrument gage factor setting is 2.000.
100
90
80
70
60
50
40
30
20
10
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
0
.
0
8
0
.
0
9
0
.
1
0

50
40
30
20
10
0
n = 2
n = 10
I
n
c
r
e
m
e
n
t
a
l

O
u
t
p
u
t

(

)

Initial R
L
/R
G
Figure 3. Range of nominal incremental output produced
in companion parallel circuits by a strain change of 1000r
in any one circuit when the initial resistance of all active
and dummy gages is the same, common and individual
leadwires are all of the same resistance, and the circuit has
been shunt calibrated at 1000r across the active gages,
assuming a package gage factor of 2.000.
T
E
C
H

N
O
T
E
TN-516
Micro-Measurements
166
resistance. When the common leadwire and the individual
leadwires in each of the parallel circuits are subjected
to the same change in temperature, false circuit outputs
will usually result. This phenomenon occurs because
the common leadwire carries the sum of the currents
carried by the individual leadwires. Except for a few
unique combinations of leadwire resistances, thermally
induced resistance changes in the leadwires will produce
a different voltage drop in the common and individual
return leadwires, thus unbalancing the bridge. If, for
example, the active and dummy gages all have the same
initial resistance and the resistances of the common and
individual return leadwires are the same, then the voltage
drop in the common leadwire will be n times as great as
in the individual leadwires. Unlike the case of isolated
quarter-bridge circuits, these voltage changes do not
cancel one another and, as a result, leadwire temperature
compensation is lost.
Figure 4 shows the magnitude of the false output increases
dramatically with both temperature and number of parallel
circuits. These outputs are essentially independent of the
strain levels to which the gages are subjected. As a result,
loss of temperature compensation can lead to very large
percentage errors (or, in extreme cases, even the wrong
sign!) when measuring typical working strains.
7.0 Remedial Measures
By far the best remedy for problems created by a common
power-supply leadwire is to avoid using it altogether.
However, for those cases in which there is no alternative to
the common leadwire, several different techniques can be
I
n
c
r
e
m
e
n
t
a
l

O
u
t
p
u
t

(
1
0
0
0

)

T = 10F (5.6C)
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
0
.
0
8
0
.
0
9
0
.
1
0
Initial R
L
/R
G
n

= 2
5
0
F
4
0
F
3
0
F
20F
Figure 4. Output increment due to leadwire resistance change with temperature when the resistance
of all dummy and active gages is the same, the common and parallel leadwires are all of the same
resistance, and the circuit has been shunt calibrated at 1000r across the active gages, assuming a
gage factor (package data) of 2.000. Note that any thermal output from the strain gages is not included.
I
n
c
r
e
m
e
n
t
a
l

O
u
t
p
u
t

(
1
0
0
0

)

T =
10F
(5.6C
)
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
0
.
0
8
0
.
0
9
0
.
1
0

n

= 3
5
0
F
4
0
F
3
0
F
2
0
F
Initial R
L
/R
G
I
n
c
r
e
m
e
n
t
a
l

O
u
t
p
u
t

(
1
0
0
0

)

T
=
1
0
F
(5
.6
C
)
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
0
.
0
8
0
.
0
9
0
.
1
0

n

= 5
5
0
F
4
0
F
3
0
F
2
0
F
Initial R
L
/R
G
I
n
c
r
e
m
e
n
t
a
l

O
u
t
p
u
t

(
1
0
0
0

)

T

=

1
0
F
(
5
.
6
C
)
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
0
.
0
1
0
.
0
2
0
.
0
3
0
.
0
4
0
.
0
5
0
.
0
6
0
.
0
7
0
.
0
8
0
.
0
9
0
.
1
0
Initial R
L
/R
G
n

= 10
5
0
F
4
0
F
3
0
F
2
0
F
T
E
C
H

N
O
T
E
TN-516
Micro-Measurements
167
employed to minimize the difculties. If, for example, the
strain indicator can be balanced at the test temperature,
and no further temperature changes occur prior to making
strain measurements, there will be, of course, no leadwire
temperature errors. Although this leaves only the relatively
small errors due to crosstalk between parallel circuits, there
may still be a problem if the initial imbalance is beyond
the balance range of the instrument in use. The latter
problem can be overcome by tying the instrument ends of
all dummies together at a common point and inserting a
xed precision resistor in series between points D and C
in Figure 1. When all leadwires, including the common
leadwire, have the same resistance, a series resistor equal
to R
L
(n 1)/n will nullify the initial imbalance (and reduce
crosstalk).
The most effective remedy is one which eliminates the
initial imbalance while maintaining leadwire temperature
compensation, even under varying test temperatures. This
is accomplished by selecting a common power-supply
leadwire with a resistance 1/n times that of the individual
return leadwires. Table 1 gives a simple procedure for
determining the appropriate wire gauge number to achieve
the desired circuit performance. Selection of the common
leadwire resistance in this fashion not only solves the initial
imbalance and leadwire temperature problems, but also
reduces the calibration and crosstalk errors. However,
it should be noted that if the resistance of the common
leadwire is less than 1/n times that of the individual return
leadwires, then all of the problems described here will be
created in reverse.
One simple method of reducing all imbalances and errors
is to reduce the ratio R
L
/R
G
. This can be accomplished by
using higher resistance gages when available.
The many problems of a single common leadwire as
outlined above can be totally eliminated by using Micro-
Measurements Bridge Completion Modules to move
dummy resistors to the gage site, and employing common
leadwires for all parallel circuits to both positive and
negative terminals of the power supply. For this technique
to be effective, both common leadwires must have the
same resistance and be subjected to the same changes in
temperature. The relatively minor problem of leadwire
desensitization remains, but this can be eliminated by shunt
calibrating the active or dummy gages. Unfortunately,
additional wires are required for remote shunting.
8.0 Summary
The use of a common power-supply leadwire can introduce
large errors in strain measurements, along with practical
i nstrumentation problems such as excessive i nitial
imbalance. Unless absolutely necessary, the common
power-supply leadwire should be avoided altogether. When
a common leadwire must be employed, the user needs
to be aware of the error magnitudes, and should always
take steps to minimize these errors through selection of
the proper leadwire size and, whenever practical, through
numerical correction of measured data.

Parallel AWG Number Diameter Ratio
Circuits Reduction (Individual : Common)
2 3 steps 1:
3 5 steps 1:
5 7 steps 1:
10 10 steps 1:
AWG number of the common leadwire should be reduced from
the AWG number of the individual leadwires by the number
of integer steps above to minimize initial imbalance and loss
of leadwire temperature compensation. This assumes the
leadwires are all of the same length and that all dummy and
active gages are of the same nominal resistance.
NOTE: This procedure for circuit compensation by wire
gauge selection is based on single-conductor (solid) wires,
for which there is a well-dened relationship between AWG
number and the conductor cross-sectional area. Since the
wire is normally drawn to a specied diameter, rather than
resistance, precise compensation may require measurement
of the wire resistances. The relationship between wire gauge
and resistance is still more variable with stranded wire, and
the resistance should always be measured to accurately
establish the 1/n resistance ratio between the common and
individual leadwires. In either case, any minor adjustments
necessary to obtain the correct resistance for temperature
compensation can be made by increasing or decreasing the
length of the common leadwire as required.
TABLE 1
GAUGE OF COMMON LEADWIRE
2
3
5
10
2
3
5
10
2
3
5
10
2
3
5
10
Tech Note TN-517
MICRO-MEASUREMENTS
Introduction to Digital Signal Processing
T
E
C
H

N
O
T
E
169
1.0 Introduction
When a signal varies with time, we are usually concerned
not only with its magnitude but also with how it changes.
Oscilloscopes, strip chart recorders and other analog
recording devices enable us to make observations of
the signal by continuously recording and displaying the
measurement data in the time domain. When digital
computers are utilized for this purpose, however, the
magnitude of the signal is sampled only at xed intervals
of time with a complete loss of continuity between. For
data acquired in this form, the mathematics of digital
signal processing can be used to analyze the signal in both
the time and frequency domains. That is, we can know not
only how the magnitude of the signal varied with time, but
also what the amplitudes of any oscillations were over a
spectrum of frequencies.
This Tech Note is intended as a guide to understanding the
necessary considerations for converting an analog signal
into a useful series of discrete digital data, particularly as
related to StrainSmart Data Systems.
2.0 Signal Aliasing
2.1 Fundamentals
Care must be taken when dealing with digital data to
avoid the creation of false, lower-frequency signals by a
phenomenon called aliasing. Do you remember seeing the
spoke wheels of a wagon that appeared to turn backwards
as the wagon rolled across a television or movie screen?
Thats a false visual impression caused by aliasing. When
the wheel is rotating at a slightly slower rate than that at
which the frames are projected, the wheels appear to run
in reverse. Further, if the wheels were rotated at the same
rate as the frames are projected, the wheel would appear to
be static, or not turning at all!
Similar things can happen with an analog signal that
is sampled periodically. Consider a sinusoidal signal
oscillating at, say, 1000 Hz, or 1000 times per second. If
we sample this signal at the same rate as the oscillations
(Figure 1), we might think the signal was static, not varying
with time.
All sampling between 1000 and 2000 times per second
would produce a lower-frequency alias. Shown here is the
333-Hz alias that we would see at a sample rate of 1333
per second (Figure 2).
Sampling at 2000 times per second, the signal would again
appear to be static (Figure 3).
So what is it that we must do to ensure that our instrument
always sees the wagon wheel turning in the right direction?
The answer is to always sample at more than twice the
highest rate of oscillation. When sampling at 4000 times
a second (Figure 4), we can see that it is impossible to
produce either a static or lower-frequency alias from the
measurements.
0 1 2 3
Time, ms
Sampling rate: 1000/s Frequency: 1000 Hz
Figure 1
*
Scans

* * *

0 1.5 3.0
Time, ms
Figure 2
*
Scans
* * * *
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
170
Just how fast should we sample? The mathematics used
in digital signal processing can construct a model of the
analog waveform that produced a particular set of data
by looking at how the magnitudes of the measurements
data vary over time. For frequency domain analysis, the
sampling rate can be quite close to twice the maximum
rate of oscillation. Reconstruction of the actual signal
itself would require a sampling rate of ten or more times
the highest frequency.
2.2 Anti-aliasing Filters
In the general case, we do not know the frequency of any
of the oscillations that might be present in the signal being
measured. But, as was just shown, we do know those
of half or more the sampling rate will produce aliases
during acquisition. Therefore, to ensure that aliases are
prevented, it is necessary to remove all components of
the signal and noise with frequencies of half or more the
sampling rate with a low-pass anti-aliasing lter. This is an
analog circuit through which the input signal must pass on
its way to the analog-to-digital converter. Of course, the
lter not only eliminates any aliasing in the digital data,
but also attenuates any true signals wanted or unwanted
above the stopband of the lter. Thus, when the data is
subsequently analyzed with digital signal processing, we
need not worry about any false lower-frequency signals
being left in the data by the analog-to-digital conversion
process.
3.0 Digital Filters
The sampling rate of the ADC is typically much higher than
that required to extract the necessary information from the
signal within the frequency range of interest. In addition to
allowing unwanted higher frequency components (noise)
to remain in the data, these higher sampling rates will also
increase data storage requirements and analysis time.
In preparation for acquiring data in digital form, the
analog signal being measured is typically passed through
an analog lter to ensure that all components of the signal,
and noise with frequencies corresponding to a half, or
more, of the digital sampling rate of the analog-to-digital
converter (ADC), are removed. As described in Section
2.0, this helps ensure that false lower-frequency signals
called aliases are not introduced into the digital data by
the sampling process itself.
In order to acquire data at a lower rate while avoiding
aliasing errors, it would be necessary to make physical
changes to the analog anti-aliasing lter and slow down
the sampling rate of the ADC. The disadvantage of
this approach is that different analog lter components
are required for each sampling rate. A more practical
solution is to leave the analog lter and ADC sampling rate
unchanged and to mathematically eliminate any unwanted
components from the measured signal by passing the
digital data through a digital lter.
Specifically, low-pass digital filters enable the digital
data coming from the ADC to be decimated. Instead
of the software processing and storing data from every
analog-to-digital conversion, the digital lter allows data
to be sampled at intervals corresponding to every n
th

conversion, effectively reducing the sampling rate without
introducing aliases. And, at the same time, any unwanted
higher frequency components of the measured signal are
eliminated from the digital data as well.
Like an analog lter, the digital lter is selected on the
basis of which frequencies in the signal are to be retained
and which are to be rejected. Low-pass lters, the most
common type, are designed to allow signal components
0 1 2 3
Time, ms
Figure 3
*
Scans
* * * * * *
0 1 2 3
Time, ms
Figure 4
Scans
* * * * * * * * * * * * *
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
171
from 0 Hz (dc) to some nonzero passband frequency, f
o
,
to pass essentially un altered (Figure 5). The lter does
introduce a series of small positive and negative deviations
from the actual signal in the passband. When this ripple
exceeds a certain amount, typically 0.01 dB, it denes the
passband frequency. For frequencies in the transition band
between the passband frequency and higher stopband
frequency, the signal is increasingly attenuated. When the
attenuation reaches a certain level, typically in the vicinity
of 95 dB, it denes the stopband frequency of the digital
lter.
When using digital lters, the user should pay attention to
both the stopband and the transition band. In some cases,
particularly those with lower passband frequencies, the
transition band may be as great as, or even greater than,
the passband itself.
Further, as shown in Figure 5, it should be noted that the
passband and stopband frequencies of digital lters differ
from the cutoff frequency of the commonly used Bessel
and Butterworth analog lters. The cutoff frequency of an
analog lter, typically specied at an attenuation of 3 dB,
usually lies in the transition band between the passband
and stopband frequencies of com parable digital lters.
Digital lters are a combination of mathematical algo-
rithms and fast digital circuits that operate on a series of
digital data acquired over a period of time. The necessity
of using a series of data leads to a delay as the data passes
Passband
Ripple
P
a
s
s
b
a
n
d

F
r
e
q
u
e
n
c
y
Passband Stopband Transition
Band
Digital Filter
Frequency
S
t
o
p
b
a
n
d

F
r
e
q
u
e
n
c
y
Analog Filter
C
u
t
o
f
f

F
r
e
q
u
e
n
c
y
0
100
0
100
S
t
o
p
b
a
n
d

A
t
t
e
n
u
a
t
i
o
n
A
t
t
e
n
u
a
t
i
o
n
,

d
B
Figure 5
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
172
through the lter. After each new sample is taken, the oldest
data drops off the front of the series, the remaining data is
moved forward in the series, and the data just acquired
is added to the end of the series. Then the algorithm is
applied to the series of data to obtain a calculated value
for the ltered data. The delay, calculated as the time a
particular sample takes to get midway through the series,
is a function of the ADC sampling rate, the number of
terms used in the series, and the passband frequency.
Accordingly, the same digital lter should be selected for
all measurement channels to ensure that all data acquired
at the same time emerges from the digital lters at the
same, but delayed, time.
4.0 Throughput Rates of Digital Systems
The electrical resistance strain gage is an inherently
analog device that utilizes changes in the relative resistance
of the gage to quantify mechanical strains in the surface to
which it is attached. Of course, as readers probably already
know, the strain gage is typically connected to some
form of instrumentation that incorporates a Wheatstone
bridge circuit to provide an analog electrical signal that
varies as the strain changes. Indeed, most sensors
whether they be strain-gage-based transducers, LVDTs,
thermocouples, piezoelectric devices, or a wide variety of
others ultimately produce such a signal.
Unfortunately, the digital computers i ncreasi ngly
incorporated into measurement systems are inherently
i ncompatible with these analog signal s. To store
measurement data in digital form, the analog signal must
be sampled at various points in time and converted to
numbers, i.e., the signal must be digitized. Ideally, the time
between samples should be vanishingly small (approaching
zero). But we know from arithmetic that anything divided
by zero is innitely large. And, of course, the computer can
handle only a nite number of data points. The question
then becomes how infrequently to sample. If the signal is
oscillating on a regular basis, then a minimum of ten data
points per period, for the highest frequency component to
reasonably reconstruct the signal in the time domain, are
typically required. In the frequency domain, any rate of
more than two samples per period will sufce. In order for
these conditions to be met, a digital measurement system
must have a sufcient throughput rate.
In the simplest of terms, the throughput rate is little more
than an indication of how much digital data a specic
combination of hardware and software can acquire per
unit of time. At the instrumentation level, it is primarily
controlled by the number of analog-to-digital converters
(ADCs) being used in the system, and the rate at which the
analog signals being measured can be sampled and digitized.
The useful throughput rate of the overall data acquisition
system, however, is typically much slower because of
such things as (1) the need for oversampling to eliminate
aliasing in dynamic signals, (2) the presence of bottlenecks
in the communications link between instrumentation and
computer hardware, and (3) limitations in the rate at which
software can acquire, reduce, store, and/or present the
digital data.
A calculation of throughput also requires knowledge of
how the instrumentation hardware acquires the data. The
simplest approach is to sequentially sample each data
channel in the system at xed intervals (Figure 6). System
4000, the original Micro-Measure ments data system,
acquires data in this fashion, using a single ADC at a
throughput rate of 25 or 30 samples per second (depending
upon the frequency of the mains power).
A more complicated approach, at the opposite end of the
spectrum, is to simultaneously sample each data channel in
the system (Figure 7). System 6000 does this at rates of up to
10 000 samples per second per channel. Because a separate
ADC is used for each instrumentation card, the theoretical
throughput rate of the system is 10 000 times the number
of channels used in the system. For a 100 channel system,
that would be a million samples per second. However,
because of the limitations in the digital communications
link between instrumentation hardware (where the data
is ac quired) and computer (where the data is stored), the
practical maximum throughput of System 6000 utilizing
Model 6100 Scanners is about 200 000 samples per second
per system for data acquisition and storage. That total can
be from a combination of 20 channels acquiring data at
10 000 samples a second, or of 1000 channels acquiring 200
samples a second.
Multi-
plexer
ADC
Scanner
Communi-
cations
Link
Computer
Signal
Figure 6
ADC*
Scanners
Communi-
cations
Link
Computer
Signals
*Simultaneous sampling
ADC*
ADC*
ADC*
Figure 7
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
173
A substantially higher throughput can be obtained with
System 7000 (or System 6000 using the Model 6200
Scanners), which simultaneously sample and store data
locally on each scanner (Figure 8). With a full complement
of sixteen cards, each System 7000 scanner has a practical
throughput of 256 000 samples per second (128 channels at
2000 samples/second/channel) or 160 000 samples/second
(16 channels at 10 000 samples/second/channel) for the
Model 6200. With this combination, the communications
link bottleneck is virtually eliminated, and the total system
throughput is limited only by the number of scanners used.
A System 7000 with 10 scanners, each lled with sixteen
cards, would have a maximum throughput rate for data
acquisition and storage of 2 560 000 samples per second,
for example.
In addition to sequential and simultaneous sampling
methods, it is also possible for systems to utilize a hybrid
of the two (Figure 9). System 5000 is an example. In this
case, all scanners begin sampling simultaneously, but the
data from each channel in each scanner is sampled and
converted sequentially. Consequently, while the maximum
scanning rate for this hardware is 50 scans per second per
channel and the maximum number of channels is 1200
per system, the maximum throughput rate of the system
is limited by the communications link between scanners
and computer to about 12 500 samples per second. That
could be 250 channels running at 50 samples a second, or
1000 channels collecting and storing data at 10 samples a
second.
As shown here, the useful throughput of a digital system
is a function of many parameters. While it is clear that
the throughput rate can be no greater than the sum of the
analog-to-digital conversions taking place, it is sometimes
less obvious that other processes (digital ltering; transfer
of the data to the computer; data storage, reduction and
display) are often the limiting factors. Indeed, the real
question of throughput is not always how much analog
data can be digitized, but rather how much of the digitized
data can actually be utilized.
5.0 Scanning Rates
As mentioned in Section 4.0, the throughput rate of a digital
data system was dened as the total number of useful data
points that can be acquired, stored, reduced, or displayed
by the system, per unit of time. While many hardware and
software factors contribute to the throughput rate, one
of the most important is the sampling rate, which can be
dened as the total number of useful datum per unit of
time for each signal channel. While the throughput and
scanning rates may be the same for a system containing a
single signal channel, the scanning rate is more commonly
the throughput rate divided by the number of channels.
For static signals, where the measurand does not vary
during the measurement process, the sampling rate is of
little consequence. For dynamic signals, unless the user
has an understanding of how the system acquires data,
how many measurement channels are being sampled, and
how the data are to be used, the term can be particularly
misleading, resulting in data that is somewhat inaccurate,
or completely wrong.
For example, consider a digital measurement system that
acquires data sequentially, with a single analog-to-digital
converter (ADC) running at 100 samples/second, and a
single strain gage on a test part vibrating at 10 Hz. This
sampling rate of 100 samples/second/channel provides
for 10 datum/cycle/channel, and should be adequate to
reconstruct the signal in the time domain. But if the
frequency of the signal increases to 100 Hz, the system can
provide for only 1 datum/cycle/channelclearly insufcient
to reconstruct the signal. While both the throughput and
sampling rates are unchanged, the data becomes meaningless
when the frequency of the signal changes.
Changes affecting the sampling rate can cause similar
cases of undersampling. Consider again the same digital
measurement system acquiring data at 100 samples/second
and a single strain gage on a test part vibrating at 10
Hz. As before, this sampling rate of 100 samples/second/
channel provides for 10 datum/cycle/channel, and should
be adequate to reconstruct the signal in the time domain.
But if the number of channels is increased to 10 and a
multiplexer is used to sample each channel sequentially, the
scanning rate is reduced to only 10 scans/second/channel
Scanners
Communi-
cations
Link
Computer
Signals
*Simultaneous sampling
Local
Storage
Local
Storage
ADC*
ADC*
ADC*
ADC*
Figure 8
Multi-
plexer
ADC
Scanners
Communi-
cations
Link
Computer
Signals
Multi-
plexer
ADC
Simultaneous start/sequential
sampling in each scanner
Figure 9
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
174
and the system can now provide for only 1 datum/cycle/
channel. While the throughput rate and frequency of the
signal were unchanged, the data became meaningless when
the number of channels (and thus the sampling rate) was
changed.
The only cure for undersampl i ng, of course, is to
increase the sampling rate. For systems operating at a
fixed throughput rate (like System 4000, the original
Vishay Micro-Measurements data system), that usually
means decreasing the number of channels being sampled.
More sophisticated systems like the Vishay Micro-
Measurements System 5000, 6000 and 7000 data systems
allow the sampling (scanning) rates to be adjusted until
the maximum scan rate, maximum throughput rate, or
both are reached. System 5000, for example, can scan at 1,
10, 50, and 100 samples/second/channel with a maximum
throughput of 12 500 samples/second/system. System
6000 will scan at 10, 100, 200, 500, 1000, 5000, and 10 000
samples/second/channel with a maximum throughput of
about 200 000 samples/second/system when using Model
6100 Scanners, and virtually unlimited throughput when
using Model 6200 Scanners. System 7000 supports scan
rates of 2048, 2000, 1024, 1000, 512, 500, 256, 200, 128, 100,
and 10 samples/second/channel with virtually unlimited
throughput.
5.1 Time Skewing
Data acquired from various channels are often functionally
related. In stress analysis work, for example, three separate
channels might provide strain data from the three grids
of a strain gage rosette, which are used together to
calculate principal stresses and strains. In this case, when
the measurements were made is important if the three
signals vary with time. When such signals are sampled
sequentially, the resulting data are all taken at different
times, and are said to be skewed. The errors produced by
this skewing depends upon the nature of the signal, the
scanning interval (inverse of the scanning rate) and the
number of intervals between the sampling of any two data
points.
For sinusoidal signals with sequential sampling, the worst-
case errors will occur as the signal is crossing through the
inection points. At these points, the maximum frequency
that can be sampled without any detectable skewing (signal
change of 1 LSB, or less, of the full scale signal over a
single scanning interval) is a function of the sampling rate
and the number of bits into which the ADC digitizes the
full-scale signal. These frequencies are a few hertz at best,
even for relatively high sampling rates. And, of course, the
situation worsens as the number of intervals between data
points increases.
Skewing errors greater than 1 LSB of full scale are detectable
and, as shown in Figure 10, the measurable frequency
range is increased moderately if the accompanying errors
are acceptable.
Of course, all skewing errors can be virtually eliminated by
using a digital data system, like System 6000 and System
7000, that features simultaneous sampling of all signal
channels. In these systems, the maximum time difference
between samples is typically in the nanosecond range, such
that skewing errors for all measurable frequency ranges are
undetectable.
6.0 Uncertainties in Digital
Measurement of Peak Signals
For most engineering parameters that require measure-
ment, the test signal varies continuously over time. A plot
of these signals, made on any analog data recorder as they
vary from one instant of time to the next, produces a line
consisting of an innite number of data points. From a
visual inspection of this line, we can immediately see the
nature of the variable. Does it increase or decrease? Is it
cyclical? What are the maximum and minimum values,
and when did they occur? When measurements are made
digitally, the time between each conversion of an analog
Frequency Response and Sampling Rate
For an analog instrument designed to measure changing input
signals, a key parameter is frequency response, or bandpass,
stated in hertz (Hz). That is a direct indication of how rapidly
the input signal can change, while still being properly amplied
and conditioned by the instrument. For a digital instrument
used in the same application, the key parameter is the sampling
rate, which is the number of samples that can be acquired in a
specied period of time, typically one second. Sampling rate is
sometimes specied in terms of frequency response (Hz), but
while the two are related, they are not the same, and should not
be used interchangeably.
1000
100
10
1
1 10
Sampling Rate, samples/sec/channel
5% Error
2% Error
1% Error
100 1000 10 000
0.1
0.01
0.001
Figure 10. Maximum frequencies that can be
sampled at various skewing error levels.
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
175
mea surement to a digital datum, and the nite data storage
capacity of a computer, limits us to the mea surement
of only a few of the data points on the line of interest.
The question then is how many digital data are needed
to make a good enough reconstruction of the analog
signal for us to obtain meaningful measurement results.
That depends upon the nature of the analog signal, the
digital sampling rate, when the samples are taken, and the
accuracy required.
Suppose, as shown in Figure 11, that a signal varies in a
sinusoidal way with time. Further suppose that digital
samples are taken at ten even intervals throughout
each cycle, beginning at the point where the amplitude
passes through zero. As we can further see when these
measurement values are superimposed on the plot of the
analog signal, it is possible to get a vague notion that
the signal is sinusoidal. But, the largest values actually
measured are only about 95% of the peak value.
If, however, the start of sampling is delayed by a twen tieth
of a cycle, as shown in Figure 12, we can still get the same
notion about the nature of the signal. But, in this case, the
largest measurement values will correspond to the peak
values of the signal.
This problem of timing is present in nearly all digital
measurements because we can seldom ensure that samples
are taken at the peak values. Accordingly, all measurements
of analog signals with digital systems will contain some
amount of uncertainty with regard to capture of the peak
values. In the case of a sinusoidal signal, this uncertainty
is a function of the ratio of the digital sampling rate and
Figure 12
Figure 13
Figure 14
Figure 11
Figure 15
T
E
C
H

N
O
T
E
TN-517
Micro-Measurements
176
the signal frequency, as shown in Figure 13 for a sinusoidal
signal oscillating about zero (with no zero offset).
Of course, many dynamic signals do not vary in a sinusoidal
fashion. Consider the further case of a discontinuous signal
(Figure 14) that varies linearly with time to some maximum
or minimum value before instantly returning to zero (such
as would be experienced by a load-bearing structure
undergoing a uniformly increasing load until it breaks).
Here the worst case scenario is for the signal discontinuity
to occur one sampling interval after the start of the last
measurement. (And, because the event can occur at any time
during a sampling interval, there is always one sampling
interval of uncertainty.) The extent of the error caused by
failure to read the peak value depends not only upon the
rate of sampling, but also upon the rate of change in the
signal and the peak value of the signal. The uncertainty
associated with this error is shown in Figure 15 for various
peak values as a function of the ratio of sampling rate and
signal rate of change.
Uncertainties, unlike errors, cannot be eliminated from
measurements. At best, they can be minimized. And, in
the case of digital measurements of peak values, the only
recourse for minimizing them is to increase the sampling
rate. Par ticular care should be taken if the per-channel
sampling rate of the measurement system decreases with
an increasing number of measurement channels.
Application Note TT-601
MICRO-MEASUREMENTS
Techniques for Bonding Leadwires to Surfaces
Experiencing High Centrifugal Forces
T
E
C
H

N
O
T
E
177
Introduction
Strain gages are often used in the stress analysis of rotating
com ponents such as compressor and turbine rotors, drive
shafts, wheels, gears, etc. On large or high-speed items,
centrifugal forces may reach several hundred thousand
times the force of gravity. As an example, the g-force at a
3-in (75-mm) radius on a component rotating at 50 000
RPM is over 200 000g. The graph below is provided to
estimate g-force.
In strain gage applications subjected to such high centrifugal
forces, it is important not only to properly bond the strain
gage, but also to adequately secure the leadwires to avoid
damage to the gage. The operating temperature range
and other environmental factors also inuence leadwire,
solder, anchoring and protective coating selections.
For long-term applications with operating temperatures
be tween 100 and +400F (75 and +205C), and in
short-term applications with operating temperatures up
to +500F (+260C), M-Bond GA-61 is the recommended
adhesive and coating material. M-Bond AE-10 or AE-15
may be substituted for GA-61 when operating temperatures
will not exceed 100 to +200F (75 to +95C).
The various installation accessories referred to throughout
thi s Appl i cat i on Not e are Mi cro-Measure ment s
Accessories, listed in our Strain Gage Accessories Data
Book and available directly from Micro-Measure ments.
Leadwire Bonding Procedure
Step 1
Chemically clean the specimen surface along which the
lead wires will be routed, following the recommended
techniques used to prepare the surface for bonding strain
gages.
Step 2
Mask the surface with MJG-2 Mylar
Tape to restrict the

adhesive to the desired area.
Step 3
Use a spatula to apply a thin, uniform layer of GA-61
Adhe sive. (The AE-10/15 systems allow brush application
and lower cure temperature.) Remove the Mylar Tape, and
cure the installation for one hour at +350F (+175C).
R
A
D
I
U
S

I
N
C
H
E
S
R
A
D
I
U
S

M
I
L
I
M
E
T
E
R
S
100
2
0
0
R
P
M
5
0
0
R
P
M
1
,0
0
0
R
P
M
2
,0
0
0
R
P
M
3
,0
0
0
R
P
M
4
,0
0
0
R
P
M
5
,0
0
0
R
P
M
6
,0
0
0
R
P
M
7
,0
0
0
R
P
M
8
,0
0
0
R
P
M
9
,0
0
0
R
P
M
1
0
,0
0
0
R
P
M
1
5
,0
0
0
R
P
M
2
0
,0
0
0
R
P
M
2
5
,0
0
0
R
P
M
3
0
,0
0
0
R
P
M
4
0
,0
0
0
R
P
M
5
0
,0
0
0
R
P
M
6
0
,0
0
0
R
P
M
10
1.0
10 100 1,000
g FORCE
10,000 100,000 1M
1000
100
10
T
E
C
H

N
O
T
E
TT-601
Micro-Measurements
178
Caution: Always handle uncured resin systems with care.
To avoid skin contact, use polyethylene gloves. If skin
comes in contact with the resin, wash immediately with
soap and warm water.
Step 4
Lightly sand the surface of the cured coating with SCP-3
400-grit Silicon-Carbide Paper or GC-5 Pumice Powder to
produce a dull matte nish suitable for bonding additional
layers of adhesive.
Step 5
Select an appropriate leadwire. Leadwires with vinyl or
Teon
insulation are not recommended for high g-elds

because high forces can cause the conductor to extrude
from its insulation. Film-insulated leadwire such as 130-
AWN, 130-AWQ, 134-AWN, 134-AWP, 134-AWQ or 136-
AWP are recommended. Bare-copper or tinned-copper
leadwires can also be used when the installation procedures
outlined in this Application Note are followed.
Vinyl- or Tef lon-insulated wires are sometimes used
near the center of a rotati ng component, with the
insulation removed along the length of wire routed to the
high-g radius location. This procedure eliminates a splice
connection from the heavily insulated instrument cable to
the smaller diameter wire used in the high g-eld.
Step 6
Solder the leadwires to the strain gage using +430F
(+221C) or higher melting-point solder. Keep the solder
joints very small and uniform. Remove all uxes according
to recom mended procedures.
Note: Operating temperatures must be limited to 35F
(20C) below solder melting temperature.
Step 7
Clean the leads thoroughly with M-Prep Neutralizer 5A,
and dry completely with a hot-air gun or other appropriate
method.
Step 8
Route leads along the surface of the specimen, securing
the leads at 1-in (25-mm) intervals with hold-down strips of
wire or Mylar tape.
Step 9
Use a spatula to apply a small amount of GA-61 to the
areas between the hold-down points. Work adhesive in
and around the leadwires to prevent air entrapment. Cure
sufciently to solidify resin, typically 12 hour at +300F
(+150C).
Step 10
Remove the anchoring wire or tape and nish-coat the
surface. Cure for two hours at +350F (+175C), or for one
hour at +375F (+190C).
Step 11
When an airfoil surface is required, apply and cure
additional layers of coating, sanding lightly between coats
to produce an optimum bond. When sufcient thickness is
obtained, contour with a portable grinder or by sanding.
A thin nal coat will produce a smooth, glossy nish.
Fiberglass Cloth Method
The following procedure will significantly shorten the
lead wire bonding process. However, the stiffness and
resonant frequency of thin or low-modulus structures may
be affected.
Step 1
Chemically clean the specimen surface as in Leadwire
Bond ing Procedure, Step 1.
Step 2
Select, attach and clean the leadwires according to the pro-
cedures described in Leadwire Bonding Procedure, Steps 5,
6 and 7.
Step 3
Cut FGC-1 Fiberglass Cloth to the desired length.
A second, insulating layer of cloth will be required if bare
leadwires are used, or if lm-coated leads are used above
their recom mended operating temperature.
Step 4
Place the cloth on a clean glass or metal plate. Use a spatula
to work a liberal amount of GA-61 (or AE-10/15) into the
cloth. Turn the cloth over and repeat the application to
completely ll the weave with resin. If desired, trim the
cloth width with scissors (after coating to avoid excessive
fraying of the cloth). Clean scissors immediately with
a suitable solvent, such as GC-6 Isopropyl Alcohol or
CSM-2 Degreaser.
Mylar and Teon are Registered Trademarks of Dupont.

T
E
C
H

N
O
T
E
TT-601
Micro-Measurements
179
Caution: Always handle uncured resin systems with care.
To avoid skin contact, use polyethylene gloves. If skin
comes in contact with the resin, wash immediately with
soap and warm water.
Step 5
Mask the specimen surface as in Leadwire Bonding Pro-
cedure, Step 2, and wet the specimen surface with a lm of
adhesive.
Step 6
Press the coated cloth to the surface. (Omit this step if ade-
quately insulated leads are used.)
Step 7
Position the leadwires on the cloth or surface by pressing
in place with a dental probe or similar tool. Keep leads
slightly separated and wet with adhesive. Careful attention
in this area will avoid air bubble entrapment later.
Step 8
Overlay the leads with a layer of berglass cloth treated as
in Step 4. Press in place.
Step 9
Cover the assembly with a strip of TFE-1 Teon Film,
a layer of SGP-2 Silicone Rubber, and a thin, contoured
pressure plate. Press the assembly firmly against the
surface and tape in position with Mylar tape, applying suf-
cient pressure to maintain the shape while curing.
Step 10
Cure GA-61 for two hours at +350F (+175C), or for one
hour at +375F (+190C). Cure AE-10 at +150F (+65C)
for one hour, or for six hours at room temperature.
Step 11
Remove clamping materials. If additional resin is required,
lightly sand the surface, coat and cure. Surface can be
ground to produce the desired contour. A thin nal coat
will produce a smooth glossy nish.
MICRO-MEASUREMENTS
Silver Soldering Technique for
Attachment of Leads to Strain Gages
T
E
C
H

N
O
T
E
181
Introduction
To avoid excessive reduction in mechanical strength, con-
ventional soft solders are generally limited to maximum
operating temperatures of approximately 50F (30C)
below the solder melting point. The highest melting tem-
perature for easy-to-handle soft solders is in the +570
to +585F (+300 to +310C) range, thus limiting the
operating temperature range to +520 to +535F (+270 to
+280C). Micro-Measurements Solder 570-28R is a soft
solder within this range.
Environmental applications that exceed these operating
temperatures require other lead attachment techniques.
Resis tance soldering, spot welding, and ame welding are
common alternatives. This Application Note is restricted to
the resistance sol dering technique with silver solder paste.
For practical demon stration purposes, the WK-Series
strain gage has been selected.
The soldering unit recommended for this application
is the WRS-1 Resistance Soldering Unit (see Micro-
Measurements Strain Gage Accessories Data Book) which
features a continuously variable power control, tweezer-
type electrodes, and a foot-actuated energizing switch. The
silver solder powder is in a ux suspension, and the melting
temperature is approximately +1240F (+670C).
All accessory items referred to in this Application Note are
described in detail in the accessories data book.
Figure 1 shows a WK-Series strain gage installation,
properly bonded and cured, in preparation for attachment
of leadwires by the silver soldering technique.
The WK-Series strain gage used in the application is one
of few organic-carrier gages capable of operating above
+525F (+275C) for extended periods. Standard leads are
beryllium copper. Nickel-clad copper leads (Option SP30)
are also avail able, and recommended for continuous high-
temperature opera tion.
Bonding
M-Bond 600 or 610 is the strain gage adhesive generally
recommended for high operating temperatures. An
insulating layer, sufcient to prevent contact of any bare
leads with the specimen surface, must be bonded to the
specimen adjacent to the strain gage. This layer is ordinarily
bonded at the same time as the strain gage, using the same
techniques. Bonding instruc tions are detailed in Micro-
Measurements Instruction Bulletin B-130.
Fiberglass cloth (untreated), tissue glass, or a layer of
Kapton
lm are commonly used surface insulators. To

provide an optimum bond, both sides of Kapton lm must
be lightly abrad ed to remove the glossy surface.
Lead Preparation
Step 1
Carefully lift the gage lead ribbons from the surface of
the insulating material by laying the wooden handle of a
cotton-tipped applicator rmly across the gage at the lead
exit point, grasping each lead with tweezers, and raising
the lead at a shallow angle (Figure 2). Avoid introducing
kinks or sharp corners. While maintaining pressure with
the applicator stick, gently stroke the leads from the gage
toward the free ends with a pencil eraser to remove any
adhesive and/or oxidation.
Step 2
For static test measurements, cut three equal lengths of
nickel -clad solid copper leadwire with berglass insulation,
such as 126-GWF [AWG No. 26 (0.016 in, or 0.4 mm,
dia.)].
Dupont Trademark.
Figure 1
T
E
C
H

N
O
T
E
TT-602
Micro-Measurements
182
Step 3
Flame-burnish 1 in (25 mm) of the leadwire ends with a
match to prevent fraying of the berglass.
Step 4
Strip approximately 0.75 in (20 mm) of insulation from
the end of each leadwire being careful not to nick the
conductors. Clean the stripped ends with a pencil eraser to
remove any oxidation.
Step 5
Press these ends rmly against any at surface and form a
C-shaped loop 1 to 2 in (25 to 50 mm) from the stripped end
(Figure 3). This loop technique will prevent any twisting
or rolling of the leadwires during attachment to the gage
leads.
Step 6
Anchor the leadwires to the specimen surface with PDT-1
Drafting Tape as in Figure 4; raise the ends off the surface
for convenient attachment to gage leads.
Step 7
Bend the stripped leadwires at right angles, 0.25 in (6 mm)
from the end of the insulation (Figure 5).
Step 8
Grasp the end of the gage lead with sharp-pointed tweezers
and wrap once around the instrument leadwire at the
formed right angle. Pull with sufcient force to tightly
form the gage lead around the instrument leadwire (Figure
6). To prevent damage to the gage, it is advisable to press
the wooden handle of a cotton-tipped applicator across the
gage at the lead exit point as shown in Figure 2 (Step 1).
Silver Soldering Technique
Step 1
Thoroughly mix the silver-solder paste in the jar in which
it is supplied. The consistency should be that of a thin
mortar. If the paste is too dry, add only a few drops of
water and stir. Be careful not to over-thin the paste. Keep
tightly capped when not in use.
Step 2
Slip a small piece of paper under the leadwires and
over the insulating layer to protect the insulation from
contamination during the silver soldering process.
Step 3
Apply a small amount of paste to each solder joint location
with a dental probe, working the solder into the joints, as
shown in Figure 7. Restrict the paste to the joints only;
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
T
E
C
H

N
O
T
E
TT-602
Micro-Measurements
183
do not allow paste to wet the lead ends. If available, a 3X to
5X magnier is helpful.
Step 4
Connect the tweezer electrode cable and footswitch to the
WRS-1 power control. Inspect the tweezers; if the point
at which the electrodes close together is not clean, lightly
abrade with 400-grit silicon-carbide paper. Always keep
electrodes clean.
Step 5
Adjust the power control to about 90%. Firmly grasp the
instrument leadwire with the electrodes, approximately
0.062-in (1.5 mm) from the end (Figure 8). If available, a 3X
to 5X magnier is helpful.
Step 6
Energize the unit by pressing the footswitch. If the solder
pool does not begin to steam and bubble, reclean the
tweezers. The actual ow of solder will not begin until the
water has boiled away.
When the solder joint has formed, quickly remove the
tweezers and then release the footswitch, in that order,
to min imize the chance of welding the electrodes to the
leadwire.
Flux Removal
Remove the glass-like ux produced by the soldering oper-
ation as follows:
Step 1
Apply a small amount of M-Prep Conditioner A to soften
the ux, and lightly brush the softened material away with
a camel-hair or other soft brush. [Warming the Conditioner
A to +120F (+50C) before applying will accelerate this
process.] Inspect the joint and repeat as necessary until the
solder is shiny in appearance.
Step 2
Flood the area with M-Prep Neutralizer 5A and blot dry
with tissue or a gauze sponge. Remove the protective paper
be tween the leads and insulation.
Cable Anchoring
Step 1
Clip off the excess gage leads and instrument leadwires at
the solder joint, and remove the drafting tape.
Step 2
Carefully form the leadwires at against the surface and
hold in place with Mylar tape or wire. Secure with tape as
shown in Figure 9.
Step 3
Apply an anchor coat of RTV-3145 or M-Bond GA-61 to
the leadwires, avoiding contact with the securing tape or
wire. Cure sufciently to secure the leadwires in posi tion.
Figure 7
Figure 8
Figure 9
Figure 10
T
E
C
H

N
O
T
E
TT-602
Micro-Measurements
184
The same coating selected to anchor the leadwires can be
extended to cover the gage.
Step 4
Remove the securing tape or wire, ll in the uncoated
areas, and complete the curing process.
Figure 10 is representative of a completed, protected, and
cured installation.
Caution: All these materials have a limited life, dependent
upon time, temperature, and exposure to oxidizing atmos-
pheres. For further information, contact our Applications
Engineering Department.
MICRO-MEASUREMENTS
The Proper Use of Bondable Terminals
in Strain Gage Applications
T
E
C
H

N
O
T
E
185
It is common practice to employ bondable printed-circuit
terminals between the main leadwire and the relatively
small and delicate jumper wires to the strain gage. The
primary purpose is to provide an anchor for both sets
of leads, and to prevent forces transmitted along the
main leadwire system from damaging the strain gage or
degrading its performance. In order to obtain maximum
benefit from the use of these terminals, they must be
installed with care and knowledge of their limitations.
Otherwise, the terminal can be a source of performance
degradation for the entire installation.
Micro-Measurements pri nted-ci rcuit termi nals are
produced from 0.0014 in [0.03 mm] electrolytic copper foil
bonded to a carrier of either polyimide lm or berglass-
reinforced epoxy. Micro-Measurements manufactures
both types in a wide variety of congurations as described
in Micro-Measurements Strain Gage Accessories Data
Book. The following instructions will assist greatly in
obtaining proper performance from bondable terminals.
1. Select terminal conguration to match the geometrical
arrangement of the gage/leadwire system, and select
terminal size appropriate to the AWG size of the main
leadwires.
2. Printed-circuit terminal strips are usually bonded
to the specimen with the same adhesive used to
install the gage. After soldering and ux removal, the
terminal strip and jumper wires should be coated with
a protective compound suitable for the environment,
usually the same compound applied over the gage
itself.
3. The polyimide-type terminals (prefix CPF) have
the highest conformability and highest temperature
capability, and are best for general use. However,
the high expansion coef cient of unlled polyimide
may cause loss of bond below 100F [75C], and the
berglass-epoxy terminals (prex CEG) are therefore
preferred for cryogenic applications.
4. Electrolytic copper foil has an inherently poor fatigue
life. Whenever possible, orient the long dimension of
the copper terminal along the axis of minimum strain.
If the strain level exceeds 500, and if the terminal
installation must have a long life under cyclic loading,
use the C conguration terminals cut in half as shown
in Figure 1. This reduces the bonded length of the
terminals, and allows the same solder mass to join
both sets of wires. Failure of the copper between solder
joints is prevented by this technique.
5. Use only rosin-type soldering uxes (such as M-Flux
AR) on printed-circuit terminals, and remove all ux
residue after soldering. Acid and/or chloride type uxes,
particularly the paste type, may cause the copper to
unbond from the backing, and often create serious
electrical leakage problems, especially at elevated
temperature. If the leadwire system is an alloy that
must be soldered with a corrosive ux, pre-tin the wire
with an active acid ux such as M-Flux SS, remove ux
residue, and then solder the wire to the printed-circuit
terminal with rosin ux.
6. Thermal EMF generation and leadwire temperature
differences can create signicant error signals in strain
gage circuits, particularly when high heating or cooling
rates are involved. These problems are minimized
Bondable Terminal Patterns
Type C
Type D
Type L
Type S
T
E
C
H

N
O
T
E
TT-603
Micro-Measurements
186
The Proper Use of Bondable Terminals in Strain Gage Applications
by keeping ther mal masses and jumper wires as
symmetrical as possible. Use neat solder joints on all
terminals with the same amount of solder on each, and
keep all jumper wires the same length from terminal
strip to gage. Figure 2 shows the use of a narrow piece
of tape to mask the center area of a set of terminals.
This prevents nonuniform solder spread, and assists in
obtaining thermal symmetry.
7. When terminal strips are used with high-elongation
strain gages (post-yield measurements), it is preferable
to locate the terminal strip at least
1
16 in [1.5 mm] away
from the end of the gage backing to avoid unbonding
problems due to thick areas of adhesive. If the terminal
backing is placed against the gage backing in an
M-Bond 200 installation, a ramp effect results as
shown in Figure 3, and the bond will generally fail at
the location shown. The half-terminal technique
(paragraph 4) is usually employed if the strain level will
exceed 2 to 3%.
8. The Type S terminal, shown in Figure 4, has a
unique construction; the hole in the center provides
thermal isolation between the solderi ng areas.
This arrangement is popular where soldering and
desoldering of leadwires may be encountered. Type
S terminals are not recommended where high cyclic
endurance is required.
9. High g-fields can create large unbonding forces
between the copper pads and backing material. When
printed-circuit terminals are used in this application,
keep t he sol der
and leadwire mass
t o a mi ni mum.
When practicable,
locate the terminal
st r i p such t hat
centrifugal forces
act either parallel
with the plane of
the termi nal, or
compressive on the
terminal surface;
this assists in keeping the terminal in place. Refer
to Application Note TT-601 relating to techniques
for bonding leadwires to surfaces experiencing high
centrifugal forces.
10. The figures in this Application Note show various
methods of terminal use in gage circuits. Note that in
every case a stress relief loop is used in the jumper
wires between terminal strip and gage to minimize
forces applied to the gage tabs, and to prevent wire
failure at the solder joints.
Please note that reduction of electromagnetic noise
pickup requires special leadwire considerations, and
stress relief loops may be undesirable in these cases.
Tech Note TN-501 provides a detailed discussion of
noise control in strain gage measurements.
Figure 2. A narrow piece of Mylar tape
or drafting tape is used to prevent
excess spread of solder. Remove tape
after soldering, and clean off tape
residue and ux with rosin solvent.
Figure 3. Exaggerated side view of an M-Bond 200 installation
with the terminal strip too close to the gage. Excessive
adhesive builds up at bottom of the gage, and forms a failure
point under high strain conditions.
Terminal
Gage
Adhesive
Bond Failure Point
Figure 4. Use
of Style S
terminals to make
a three-wire
cable connection.
Tin entire copper
pad, leaving neat
dome of solder
Style C terminal
cut in half
lengthwise
Note: When the main leadwire is stranded, it is often
convenient to cut all strands but one to t the size of the
copper pad. The long strand can then be used as the
jumper wire. Soldering is made considerably easier by
this method.
One strand
Twist and tin remaining
strands before cutting
and soldering to terminal
Figure 1. Half-terminal technique for using
bonded terminals in high strain elds.
T
E
C
H

N
O
T
E
TT-603
Micro-Measurements
187
The Proper Use of Bondable Terminals in Strain Gage Applications
No terminals are required with a CEA-
type gage. Two of the leadwires are
twisted to gether; then all three leads
are trimmed and soldered directly to
the copper-coated tabs.
Use of Style L terminal to
bring main lead cable off at
right angles to gage axis.
ADDITIONAL WIRE CONFIGURATIONS Various methods of making three-wire cable connections.
2 Style C
Terminals
2 Style D
Terminals
3 Style C
Terminals
Note: The above methods using Style C terminals
can be combined with the half-terminal technique
shown in Figure 1 on page 186.
Ver ti cal stress-reli ef loop.
Often used when gages have
integral jumper leads.
Side View
MICRO-MEASUREMENTS
Leadwire Attachment Techniques for
Obtaining Maximum Fatigue Life of Strain Gages
T
E
C
H

N
O
T
E
189
Revision 20-Feb-2012
Introduction
Micro-Measurements WA-, WK-, and WD-Series Strain
Gages are manufactured with integral beryllium-copper
lead ribbons. To maximize the inherently good fatigue
life of this gage/lead combination, the gage-to-instrument
leadwire connection must also be highly resistant to
fatigue damage. The techniques described here are
recommended for applications where the cyclic strain
endurance requirements approach the fatigue limits of the
gage in use.
Wi ri ng procedures for gages with i ntegral solder
dots instead of lead ribbons are described in Micro-
Measurements Application Note TT-606, and general gage
soldering techniques are discussed in detail in Application
Note TT-609.
Bonding
M-Bond 600 and M-Bond 610 are the strai n gage
adhesives generally recommended for high-performance
applications. However, any M-Bond adhesive can be used
where temperature conditions and other factors permit.
Although Micro-Measurements strain gage adhesives are
not particularly fatigue sensitive, some potential problem
areas are:
Adhesive failure due to improper surface preparation
or environmental protection.
Voids or air bubbles in the glueline, causing hot
spots in the gage and subsequent premature failure.
Bonding instructions for M-Bond 600 and 610 are detailed
in Micro-Measurements Instruction Bulletin B-130.
Wherever possible, the use of bondable terminals is
recommended to provide an anchor for the instrument
leadwire, preventing damage to the gage when the leadwire
is subjected to any type of force. Micro-Measurements
Type CPF-75C terminals are particularly suitable for this
application within their operating temperature range.
If bonded in a high strain eld, terminals should be cut in
half and/or bonded 90 to the maximum strain direction.
Figure 1 is representative of typical installations.
Lead Preparation and Attachment
Step 1
Carefully lift the gage lead
ribbons from the specimen
surface by laying the wood
extension of a cotton-
tipped applicator firmly
across the gage at the lead
exit point, grasping each
lead end with tweezers
and raising the lead at a
shallow angle (Figure 2).
Avoid introducing kinks
or sharp corners.
Step 2
With tweezers, guide each
lead ribbon upward and
around the tip of a dental
probe to form a smal l
loop above the surface
between the l ead exit
point and the edge of the
gage backing (Figure 3).
Do not twist or kink the
ribbons. Once kinked, a
lead will easily break.
Figure 1
Figure 2
Figure 3
T
E
C
H

N
O
T
E
TT-604
Micro-Measurements
Revision: 20-Feb-2012
190
Leadwire Attachment Techniques for
Obtaining Maximum Fatigue Life of Strain Gages
Step 3
Insulated, three-conductor, stranded tinned-copper
leadwire cable is normally selected for static or static/
dynamic measurements. Refer to Micro-Measurements
Strain Gage Accessories Data Book for the selection of a
suitable leadwire. Typically recommended cable types are
Micro-Measurements 326-DFV (fat, and most convenient
for routine applications) or 326-DSV (twisted, and shielded
for use when electrical noise elds are present). Both cables
are color-coded red, white, and black. The 326-DFV is
used to illustrate the following step-by-step procedures.
Cut the leadwire cable to the appropriate length. At the
gage end of the leadwire, separate the three color-coded
conductors, and strip 0.5 in (13 mm) of insulation from
each. A thermal wire stripper is recommended to help
prevent nicking the wires.
Step 4
Separate a single wire strand from both the red conductor
and the black conductor.
Twist the remaining strands of the red conductor together,
leaving the single strand separated. Twist the remaining
st rands of t he bl ack
conductor together with
all strands of the white
c ondu c t or , l e av i ng
the si ngle strand from
t he bl ack conduct or
separated.
Ti n t he t wo t wi st ed
bundles of stranded wire
with solder. With diagonal cutters, trim each bundle to
approximately 0.1 in (2.5 mm) from the insulation, being
careful not to trim the two single strands of wire extending
from the red conductor and the black/white conductor
(Figure 4).
Step 5
Tin the bonded terminals
with sol der. Pl ace the
tinned, trimmed sections
of the leadwires on the
terminals, and anchor the
cable in this position with
drafting tape. Solder the
leads in place (Figure 5),
avoiding large masses of
solder.
Step 6
With tweezers, form the single wire strands and insert
through the gage lead loops as shown in Figure 6. Tighten
the gage lead loops by pulling gently with a pair of tweezers
(Figure 7). To avoid damage to the gage joint, lightly press
the wood extension of a cotton-tipped applicator over the
area as indicated in the gure.
Solder the joints and cut off excess lead ribbons and wire
strands.
Step 7
Remove solder fux with rosin solvent, gently blotting dry
with a gauze sponge. If any wire strands come in contact
with the specimen, gently lift with a dental probe.
Protective Coating Application
Refer to Micro-Measurements Strain Gage Accessories
Data Book for the selection and application of a suitable
protective coating.
Note: WD-Series Strain Gages are supplied with single
beryllium-copper leads. The WA- and WK-Series gages, as
shown throughout this Application Note, are supplied with
dual leads. The WD gages should be handled in the same
manner as the dual lead connections.
Figure 4
Figure 5
Figure 6 Figure 7
MICRO-MEASUREMENTS
High-Elongation Strain Measurements
T
E
C
H

N
O
T
E
191
Introduction
Experimental stress analysis of structural materials
sometimes requires testing to complete failure. In such
cases, particularly with ductile materials, failure is often
preceded by large local strains, the magnitudes of which
are of interest to the test engineer.
High-elongation strain measurements place severe demands
on the gage installation, and necessitate special gage and
adhesive selection and surface preparation procedures.
This Application Note outlines recommendations for gage
and adhesive selection, surface preparation, gage bonding
and wiring, and protective coating selection for high-
elongation strain measurements.
High elongations are usually expressed as percentages: 1%
elongation is equal to 10 000 microstrain; 5% elongation
is equal to 50 000 microstrain. The maximum rated
elongation of any bonded strain gage is based on a single
tensile measurement. Cyclic strains of high levels must
be limited to a value much lower than maximum due to
fatigue limitations on the strain gage foil. Consult Micro-
Measurements Tech Note TN-508, Fatigue Characteristics
of Micro-Measurements Strain Gages, for details on strain
gage performance under repetitive loading conditions.
Gage Selection
Selection of proper strain gages for use in high-elongation
testing is based on both anticipated strain levels and test
temperature. Polyimide (E) backing is normally selected
for this type of service because it has superior elongation
capabilities and an operating temperature range suitable
for most high-elongation testing. The E-backing is available
with constantan (A) foil or fully annealed constantan
(P) foil.
EA-Series
Properly installed and wired EA-Series (polyimide-backed
constantan foil) strain gages are capable of measuring
maximum elongations in the range of 3% 5%. While gage
lengths of 1/8 in (3 mm) or longer will normally achieve
5%, shorter gage lengths may be limited to 3%. Since most
structural materials (e.g., metals) yield well below this
limit, the EA-Series is a popular choice for use in obtaining
yield point information on these materials.
EP-Series
EP-Series gages are recommended when measurement
requirements are beyond the 3% 5% elongation capability
of the EA-Series. P-alloy is a fully annealed constantan foil
processed for very high ductility. A properly bonded
and wired EP-Series strain gage is capable of strain
measurements to 20% (200 000 microstrain) or greater.
As in the EA-Series, smaller gages will exhibit a lower
maximum elongation capability.
Options
Any leadwire attachment option, such as Option W
(encapsulated gages with integral terminals) or Option
LE (encapsulated gages with leadwires), will lower the
maximum elon gation capability of EP-Series gages. But
because the elongation limits of EA-Series strain gages
are more moderate (3% 5%), and because options will
normally withstand at least the lower end of this range,
they are not generally a limitation in the selection of
EA-Series gages.
To obtain maximum possible elongation, select EP-Series
gages without options.
Adhesive Selection
High-elongation strai n measurements place severe
demands on the adhesive system. The adhesive must be
rigid enough to prevent gage relaxation (creep), yet exible
enough to permit large deformations without cracking.
Recommended adhesives are listed in the following
selection chart.
Surface Preparation
The selection and implementation of proper surface
preparation procedures is equally as important as using
proper gages and adhesives. High-elongation measurements
demand closer attention to recommended procedures than
do normal elastic strain measurements.
Micro-Measurements Instruction Bulletin B-129, Surface
Preparation for Strain Gage Bonding, outlines steps for
surface preparation on a variety of materials. These
procedures produce a smooth, clean surface, usually
havi ng been abraded i n a si ngle di rection duri ng
T
E
C
H

N
O
T
E
TT-605
Micro-Measurements
192
preparation. Because of higher bondline forces involved
in high-elongation measurements, the surface should be
altered further for greater bond strength, as indicated in
the following procedures:
1. Completely prepare the surface of the test specimen as
described in Bulletin B-129.
2. Abrade the specimen surface in a direction 45 to the
intended axis of strain measurement. On soft materials,
such as aluminum, use 320-grit silicon-carbide paper;
on harder materials, such as steel, use 60-grit.
3. Lightly abrade, with the appropriate grit paper as
indicated in Step 2, in a direction 90 to the rst abrasion.
This will produce a cross-hatched abraded surface.
Typical surface roughness desired approximates 250in
(6.4 m) rms.
4. Repeat the degreasing steps outlined in Bulletin B-129.
5. Condition the surface with Micro-Measurements
M-Prep Conditioner A (if appropriate for the specimen
materials as detailed in B-129), scrubbing with cotton-
tipped applicators. Wipe dry with a gauze sponge.
6. Neutralize the surface with Micro-Measurements
M-Prep Neutral i zer 5A (i f appropriate for the
speci men mat er i al ), s cr ubbi ng wi t h cot t on-
tipped applicators. Wipe dry with a gauze sponge.
Many materials oxidize readily in air. If allowed to
form on the surface, oxidation will greatly reduce
bond strength and elongation capabilities. It is strongly
recommended that the gage bonding operation be
completed as soon as possible after the surface is
prepared.
Gage Bonding
Follow the standard bonding procedures outlined in the
respective instruction bulletin for the adhesive selected.
Gage Wiring
Relatively high displacements occur with high-elongation
strain measurements, suggesting the use of an external
ter-minal strip. The strain gage should be wired to the
terminal strip with a small-diameter, single-conductor
wire no larger than 30 AWG (0.25 mm diameter) as shown
in Figure 1. Drafting tape (Micro-Measurements PDT-1)
applied over the gage during soldering is recommended to
restrict the ow of solder to the tab ends.
Adequate strain relief loops, as shown in the following
illustration, are of particular importance in high-elongation
measurements. Note also that the leads attached to the
strain gage tabs approach in a direction 90 from the
maximum principal strain axis.
ADHESIVE
TYPE
GENERAL DESCRIPTION
PREFERRED
CURE
ELONGATION CAPABILITIES
320F
(195C)
+75F
(+24C)
+200F
(+95C)
M-BOND
200
Special pretested grade of cyanoacrylate,
certied for strain gage use. Fast room-
temperature cure.
3 min at
+75F (+24C)
Not
applicable
6%
Up to 15% for
installations less
than 30 minutes old
Not
applicable
M-BOND
AE-10
Two-component, 100%-solids epoxy
system. Capable of room-temperature cure.
Transparent, medium-viscosity adhesive.
6 hr at
+75F (+24C)
1% 6 10% 10 15%
M-BOND
AE-15
Two-component, 100%-solids epoxy
system. Moderately elevated temperature cure.
Transparent, medium-viscosity adhesive.
2 hr at
+150F (+65C)
2% 15% 15%
M-BOND
GA-2
Two-component, partially lled, 100%-solids
epoxy system. May be cured at room
temperature. Higher viscosity than AE system.
6 min at
+75F (+24C)
followed by
1 hr at +125F (+50C)
4% 10 15% 15 20%
M-BOND
A-12
Two-component, lled, 100%-solids epoxy
system. Paste-like consistency. Specically
recommended for elongations above 10% on
steel.
2 hr at
+165F (+75C)
Not
applicable
15 20% 15 20%
For additional specications, consult Micro-Measurements Strain Gage Accessories Data Book
and specic adhesive instruction bulletin.
T
E
C
H

N
O
T
E
TT-605
Micro-Measurements
193
After soldering, all tape and soldering f lux should be
removed with several brush applications of rosin solvent
(Micro-Measurements RSK-1).
Gage Linearity
A discussion of high-elongation strain gage measurements
is not complete without including the subject of gage
linearity under plastic straining conditions.
Although constantan alloy is one of the most linear of all
strain gage alloys, it does exhibit small changes in gage
factor during high elongation.
Without treating this subject in detail, it can be shown in
theory that the gage factor of the strain gage approaches a
value of 2+ when the gage alloy is strained beyond its elas-
tic range. The elastic range for constantan is approximately
0.5%.
Using the 2+ guideline, the gage factor at a strain level of
10% would be close to 2.1 in tension and 1.9 in compression.
At a 20% strain level, the gage factor approaches 2.2 in
tension and 1.8 in compression.
While these numbers are correct in theory, they have
not, to our knowledge, been substantiated by actual test
results.
Instrumentation
Wheatstone bridge circuits are commonly used in strain
gage measurements. These circuits are inherently nonlinear
when large gage resistance changes are encountered.
Consult Micro-Measurements Tech Note TN-507, Errors
Due to Wheatstone Bridge Nonlinearity, for correction
procedures.
Common Installation Problems
Unbonding of the Gage
Assuming the proper adhesive has been selected (see
Chart), loss of bond before maximum gage elongation
is reached can usually be traced to a contaminated or
improperly prepared surface. A visual examination of the
bond area will usually detect this. Cured adhesive on the
back of the strain gage but not on the specimen generally
indicates improper or incomplete surface preparation.
It is important to note that some materials, like polyethylene,
are difcult to bond; and it will not be possible to reach
maximum gage elongation before bond failure.
Unbonding of the Gage Tabs
Gage tab unbonding is often due to an excessive amount
of solder, which reinforces the gage tabs, and causes them
Tape Coverage
Strain relief loop
Style C terminal
cut in half lengthwise
Single Strand
Twist and tin remaining strands before
cutting and soldering to the terminal
Tin entire copper
terminal, leaving neat
dome of solder
HALF-TERMINAL TECHNIQUE FOR USING BONDED TERMINALS IN HIGH STRAIN FIELDS
Note: When the main leadwire is stranded, it is
often convenient to cut all strands but one to t
the size of the copper terminal. The long strand
can then be used as the jumper wire. Soldering
is made considerably easier by this method.
Figure 1
Tape Coverage
Strain relief loop
Style C terminal
cut in half lengthwise
Single Strand
Twist and tin remaining strands before
cutting and soldering to the terminal
Tin entire copper
terminal, leaving neat
dome of solder
T
E
C
H

N
O
T
E
TT-605
Micro-Measurements
194
to unbond as a result of their inexibility. Drafting tape
is recommended for restricting the amount of solder (see
Gage Wiring).
Excessive soldering temperatures, or terminals placed too
close to gage tabs, can also contribute to this problem.
Gage Grid Failure (Open Circuit)
Assuming that the strain gage readings were within the
strain range capability of the gage, and that the backing
remains bonded, premature grid failures are often the
result of high local strains within the area covered by the
gage. Since the strain gage is an averaging device, it will
indicate average strain along its grid length. Steep strain
gradients can cause localized excessive strain damage
while the gage may have been indicating a strain that was
well within its elongation limits.
Grid failures may also result from strain concentrations
caused by inclusions in the adhesive layer (unmixed
adhesive particles, dirt, etc.) or by uneven gluelines, often
caused by irregular or pitted specimen surfaces or uneven
clamping pressures.
Elongations Exceeding 20%
Strain measurements, particularly on materials such as
plastic or rubber, sometimes surpass the capability of
the EP-Series strain gage. Measurements in this range
can be obtained through simple exure devices designed
to reduce the strain level on the gages. These devices,
commonly referred to as clip gages, are shown in Figures
2A and 2B. Strain gages are installed on the top and
bottom of the clip, and the clip mounted to the specimen
by bonding or spot welding, depending on the specimen
material. The displacement of the feet under strain causes
corresponding resistance changes in the strain gages. With
the gages wired into either a half or full Wheatstone bridge,
their resistance changes are additive, thereby increasing
the amount of available signal. Wherever space permits,
full bridges are recommended because the effective bridge
output is twice the half-bridge signal for the same foot
displacement.
Since the clip gage is normally a nonlinear device, the
gage output is calibrated against known displacements
before installation. The gage zero reading is monitored
during the calibration, since a permanent zero offset after
calibration can usually be traced to localized yielding of
the metal strip. If this is indicated, the calibration should
be repeated several times to check for reproducibility.
A clip gage is particularly useful in measuring very large
specimen strains or actual displacements occurring
between two bodies, as with expansion joints or crack
openings.
Figure 2A Clip Gage (Half Bridge)
Figure 2B Clip Gage (Full Bridge)
THIN METAL STRIP HARDENED
TO RETAIN SHAPE
FOOT
STRAIN GAGES
TOP AND BOTTOM
COMPRESSION
TENSION
BRIDGE
EXCITATION
STRAIN GAGES
TOP AND BOTTOM
BRIDGE
EXCITATION
MICRO-MEASUREMENTS
Soldering Techniques for Lead Attachment to
Strain Gages with Solder Dots
T
E
C
H

N
O
T
E
195
Introduction
Many gages are now purchased with integral solder dots.
These offer advantages of accurately controlled solder
joint size, convenience when working in conned areas,
and of having pretinned hard-to-solder alloys such as
isoelastic and Karma. Controlled joint size, in turn, helps
to increase gage fatigue life, and minimizes thermocouple
effects.
If proper installation techniques are used, there should
be no difculty experienced in obtaining well-soldered
connections; but care must be exercised. Open-faced gages
with Option S (solder dots only) should be coated with
adhesive at the time of installation. Such encapsulation
effectively prevents dot enlargement and undesirable
spreading during subsequent soldering steps. To expose
the dots, lightly abrade with 400-grit silicon-carbide paper
or simply solder through this over coat. Encapsulated gages
do not require additional coating with adhesive.
The following are important guidelines:
Soldering Temperature
T backed gages with Option S, and S Series gages,
have dots composed of +570F (+300C) lead-tin-silver
solder. Micro-Measurements wire solders have melt ing
points ranging from +361 to +570F (+183 to +300C);
the most common are the +361 and +570F (+183 and
+300C) solder wire with rosin-core ux. Solid wires are
also available, including a popular +430F (+220C) type.
A temperature-controlled soldering station, such as the
Mark V or Mark VIII Soldering Units, is recommended.
Tip temperature should be high enough to assure good
wetting of the solder but not so high as to remove the
dots, vaporize ux, or hinder keeping the iron properly
tinned and clean. Unfortunately, most uncontrolled irons
are quite capable of tip temperatures in excess of +900F
(+480C), which is much too high for general strain gage
soldering.
Soldering Tip Design
Never solder with a sharply pointed tip (Figure 1). A hot
point applied to a solder dot usually draws out the solder.
Use a clean, at 1/16-in (1.5-mm) wide chisel or screwdriver
type tip held at against the work, as in Figure 2.
Figure 1
Correct
Shape
Incorrect
Shape
Figure 2
Iron Direction
Flat Surface
Iron Direction
Correct
Incorrect
T
E
C
H

N
O
T
E
TT-606
Micro-Measurements
196
Fluxing
Proper uxing is essential. Enough M-Flux AR, either
from melting of ux-core solder or adding M-Flux AR
separately, must be introduced to assure complete wetting
action. This is parti c ular ly important with +570F (+300C)
solder as there is an increased tendency for the ux to be
boiled away at higher temperatures.
If spikes are produced instead of smooth beads, it is a
sign of inadequate uxing, dwelling too long with the iron,
and/or an improper iron temperature.
Rosin ux should always be used in making connections
to Option S gages. All ux residue should be thoroughly
re moved with RSK Rosin Solvent.
Bead Formation
It is important to rst build up a well-formed bead on
each dot as in Figure 3. This is done by laying rosin-core
solder wire across each dot, rmly applying the iron for
one second, then simultaneously lifting both wire and iron.
See Figure 4. If the rst attempt fails, simply repeat this
procedure, making certain to use enough ux. Feeding a
cored solder into the joint area during heat application will
increase the available ux.
Hold-down
Each connecting wire must be carefully and firmly
an chored in place with spring loading against the dot
bead, as illustrated in Figure 5, before soldering the
connection. When this is properly done, a one-second
touch of a hot iron and fresh solder with f lux should
complete the process.
Solder Compatibilities
All Micro-Measurements solders are compatible with each
other. Of course, if one type is incorporated in the dot and
another added from wire, a mixture is produced. This new
alloy cannot be expected to have melting and strength
properties any better than those of the lower temperature
component. In special cases of mixing +430 and +570F
(+220 and +300C) solders, a nal alloy similar to +361F
(+183C) solder can be encountered.
Figure 6 illustrates a typical installation. Note that solder
dot (Option S or SE) gages are always connected to bonded
ter minals for strength. Refer to Application Note TT-603,
The Proper Use of Bondable Terminals in Strain Gage
Applica tions, for other arrangements.
Figure 3
Before
After
Figure 4
Iron Direction
Solder
Flat Surface
Figure 5
After Soldering
Figure 6
Before Soldering
T
E
C
H

N
O
T
E
TT-606
Micro-Measurements
197
thi s Appl i cat i on Not e are Mi cro-Measurement s
Book and available directly from Micro-Measurements.
Application Note TT-609, Strain Gage Soldering Techniques,
provides a detailed discussion on general strain gage
soldering techniques.
V I S H AY MI C R O - ME A S U R E ME N T S
STRAI N GAGE I NSTALLATI ON AND
PROTECTI ON I N FI ELD ENVI RONMENTS
Appl i cat i on Not e TT-607
A
P
P
L
I
C
A
T
I
O
N

N
O
T
E
VMR-TC0607-0501
Introduction
Field installation of strain gages presents the gage installer
with several unique problems, particularly when long-
term measurements are required, and massive structures or
inclement weather are encountered.
For example, in some eld conditions, proper surface
preparation may be difcult or impossible, and required
adhesive curing temperatures may not be attainable.
This Application Note outlines recommendations for gage
and leadwire selection, installation, and protective coat-
ings to maximize strain gage performance under these
conditions. Wherever possible, preparation of the test-part
surface should be in accordance with procedures outlined
in Application Note B-129. Surface preparation of an
area extending at least 2 in [50 mm] beyond each edge of
the strain gage will provide an optimum bonding surface
for both the adhesive and the protective coating. In some
instances these procedures may necessarily be adapted to
eld conditions, bearing in mind that where some steps are
shortened or eliminated, gage performance and service life
may be reduced.
Gage Selection
Bondable Strain Gages
Adhesively bonded strain gages are preferred over weldable
strain gages when highest accuracy is desired.
When an adhesive cure is possible, select Vishay Micro-
Measurements gages from the CEA-, L2A, and C2A-Series;
the EA-, WA-, or WK-Series with Option W; the CEA-
Series with Option P2; or the EA-Series with Option P.
Pressure application to the installation during adhesive cure
may require special clamping xtures. Methods for applica-
tion of clamping pressure include deadweights, spring xtures
secured with bolts or magnets, rubber bands, spring clamps,
and vacuum fixtures. Refer to Application Note
TT-610 for strain gage clamping techniques. Adhesive curing
temperatures are often attainable with heat lamps, hot air blow-
ers, heat guns, strip heating pads, or heating tape. Cyanoacrylate
adhesives, such as M-Bond 200, are popular because of their
short-term curing requirements. However, the installation
surface must rst be warmed to at least +65F [+20C] prior
to bonding. Because of its sensitivity to moisture absorption,
M-Bond 200 is not recommended for field installations
exceeding six months.
Refer to Catalog 500 for complete strain gage descriptions.
this Application Note are listed in Catalog A-110 (VSD-
DB0065) and available directly from Vishay Micro-
Measurements.
Weldable Strain Gages
Where structure size or weather conditions will not permit
adhesive curing, weldable strain gages are an alternate
approach for gage installation.
The Weldable Strain Gage consists of a specially manufac-
tured strain gage prebonded to a metal carrier for spot weld-
ing to the test surface. Although weldable gages are more
costly than bondable gages, the overall installation cost is
reduced signicantly because of the shorter installation time
and elimination of adhesive-curing requirements.
Refer to Catalog 500 for complete weldable gage descrip-
tions, and Application Note B-131 for installation details.
With either bondable or weldable strain gages, a fully encap-
sulated gage type will ensure maximum stability, and a gage
resistance of at least 350 ohms will minimize errors associat-
ed with leadwire desensitization. When maximum accuracy
is required below 50F [45C] or over +150F [+65C],
modied-Karma (K) alloy is recommended.
Leadwires
Selection
It is important to select the proper wire type and size to
help ensure stability of measured data. Moisture absorption
through the leadwires can result in apparent gage instability
and drift; therefore, the leadwire insulation should be tested
for its waterproof properties before installation.

Vishay Micro-Measurements
Strain Gage Installation and Protection in Field Environments
Revision 20-Jan-05 2
Where long lengths of leadwire are required, select a wire
gauge large enough to maintain minimal leadwire resis-
tance, preferably less than 1% of the strain gage resistance.
Quarter-bridge gage circuits utilizing a three-wire connec-
tion to the instrument are preferred in order to minimize
temperature effects and leadwire desensitization.
When lengths of leadwire are spliced together, the splice
joints can be protected and waterproofed with a heat-shrink-
able wire splice sealant such as HST-1. HST-1 can be used
to seal splice connections of stranded, insulated leadwires
ranging from 18 to 34 AWG.
For further protection of long-term installations, encasing
the leadwire in waterproof conduit is recommended.
Preparation
A common source of gage failure is an inadequate bond
between the protective coating and the leadwire insulation
(or conduit) in the lead exit area. Without proper seal, mois-
ture can be drawn into the gage area along the leadwires by
capillary action. This junction is important to successful
long-term strain gage installations.
Vinyl-insulated leadwires should be coated with a layer of
M-Coat B, an air-drying nitrile-rubber compound.
Untreated Teon
-insulated leadwires require etching with

a Teon etchant, such as TEC-1 Tetra-Etch
Compound.
The conductors of at-cable leadwire should be separated
prior to treatment to assure an even application around
each conductor.
Thermally stripping and tinning the leads after treatment
prevents contact of the priming materials with bare wires.
Attachment
Field installations often fail because of improper solder selec-
tion or inadequate removal of soldering uxes. Fluxes are
generally hygroscopic and, when not completely removed,
contribute to gage instability and drift. Rosin uxes are rec-
ommended for strain gage use, and are removed with several
applications of RSK Rosin Solvent.
Some industrial fluxes, particularly zinc-chloride paste,
are not suitable for strain gage soldering. Vishay Micro-
Measurements solders and uxes have been selected speci-
cally for this purpose.
Refer to Catalog A-110 for detailed information on lead-
wire, solders, and accessories.
Protective Coatings
For maximum stability of the strain gage, the installation
requires sealing with a protective coating that is immediately
stable chemically, and highly resistant to water-vapor trans-
mission. Mechanical protection is also often required to
shield the installation from the effects of earth lls, concrete
pours, tool handling, structural assembly, etc.
Moisture is the most common cause of eld installation
failures. Its presence usually results in low electrical resis-
tance to ground and causes circulating currents, electrical
noise, and desensitization of the measurement. Grid corro-
sion and intragrid conductive paths can also result, causing
negative or positive drift in output, depending upon which
cause is predominant.
A good protective coating not only seals the gage instal-
lation from moisture, but will also seal in any moisture in
the area at the time of coating application. For maximum
stability in high-moisture environments, it is therefore
important to warm the gage installation until any moisture
is removed prior to applying the protective coating. The
protective coating should be applied immediately after gage
and leadwire installation to prevent moisture from collect-
ing in the gage area.
The M-Coat F protective coating system is an excellent choice
for immediate and thorough protection in eld environments.
Refer to Instruction Bulletin B-134 for specications and
application instructions for the M-Coat F system. Catalog
A-110 describes additional protective coatings that are
available.
Data Acquisition
A properly installed and protected strain gage installation
does not alone ensure stable, drift-free measurements in hostile
eld environments. Instrumentation and temperature changes
at the gage location can inuence strain measurements.
Instrumentation
Portable instruments are ideal for eld measurements, but
they require special consideration to maintain accuracy and
repeatability. Battery-powered models should employ bat-
tery-check circuits to assure proper sensitivity and proper
calibration. In addition, an instrument which powers the
gage only during strain measurement assists in gage sta-
bility. Refer to Bulletin 600, Strain Gage Instrumentation
Selection Guide, for a selection of high-quality portable
instrumentation.
Teon is a Registered Trademark of DuPont.
Tetra-Etch is a Registered Trademark of W. L. Gore.
Strain Gage Installation and Protection in Field Environments
Vishay Micro-Measurements
Revision 20-Jan-05 3
To check stability over long periods, an instrument zero-
reference channel is recommended. Zero-reference channels
are generally single- or full-bridge precision resistors that
can be connected to the instrument before and after each
measurement sequence to check the zero stability.
Refer to Catalog A-110 for specications on precision resis-
tors for strain gage circuits.
If leadwires will be disconnected from the instrument
between measurements, sealing the tinned ends when not
in use will prevent moisture absorption. Sealing can be
accomplished by dipping the lead ends in melted wax, or
encasing in a waterproof junction box. When solderless
connectors are selected, they should have low and repeat-
able contact resistance.
Temperature Effects
Self-temperature-compensated strain gages are recom-
mended for single-gage measurements. A properly installed
and wired gage of this type will typically result in a thermal
output of less than 50 over a range of +40 to +125F
[+5 to +50C]. Where expected temperature ranges will
be wider, or where higher accuracy is desired, a surface
temperature measurement taken at the gage location with
each strain measurement can be used with the gage-package
data to correct the strain gage readings for effects of tem-
perature. Where feasible, a temperature-compensating half-
bridge gage installation can be used as an alternate method
of temperature compensation, and may be preferred in
some applications.
Refer to Catalog 500 for temperature sensor description.

MICRO-MEASUREMENTS
Techniques for Attaching Leadwires to
Unbonded Strain Gages
T
E
C
H

N
O
T
E
203
Introduction
Strain gages must sometimes be installed in areas where it
is impossible, or at least very difcult, to attach leadwires
after gage bonding. Examples include:
Deep holes or recesses where gage tabs cannot be
reached with a soldering iron.
Poor-heat-conducting materials such as plastics, where
soldering heat may damage the gage or specimen.
Explosive or i nf lammable envi ronments where
appliances such as soldering irons are not permitted.
Under these and similar conditions, attaching leadwires
to the gage before bonding will greatly simplify gage
installation. This Application Note outlines recommended
procedures. The various installation accessories referred to
throughout this Application Note are Micro-Measurements
Book and available directly from Micro-Measurements.
The use of gage types with preattached terminals is
preferred in order to minimize chances of heat damage
to unbonded gage backings during soldering. In addition,
f lexible backings are preferred to reduce chances of
breakage during handling and cleaning steps.
Recommended Micro-Measurements types (in order of
preferred selection):
1. E-backed gages with Option W
2. CEA-Series gages
3. W-backed gages with Option W
Consult Micro-Measurements Precision Strain Gages
Data Book and Tech Note TN-505 for detailed gage
descriptions.
Solder Selection
Any of the soft solders listed in our Strain Gage Accessories
Data Book can be used for preattaching leadwires. In
general, solders having the lowest melting point which
satisfy test temperature and adhesive cure temperature
requirements should be selected. This is because an
unbonded strain gage backing has a much lower tolerance
for high soldering temperatures.
Leadwire Selection
Ideally, the smallest practicable diameter leadwire should
be soldered to the gage to reduce solder connection height
and its effect on uniform clamping pressure. This is not
without limitation since the smaller the leadwire the
greater its resistance and corresponding effect on circuit
stability and sensitivity.
As a general guideline, no larger than stranded 30
AWG wire, 0.031 in (0.8 mm) overall diameter including
insulation, should be used with the 100%-solids epoxy
adhesives. These include M-Bond AE-10, AE-15, GA-2,
and GA-61 Adhesives.
With solvent-thinned adhesives such as M-Bond 43-B, 600,
and 610, attainment of uniform clamping pressures is more
critical. The maximum recommended leadwire diameter
for these adhesives is 0.01 in (0.25 mm) including insulation.
Leadwire types with flm insulation, such as 130-AWN or
134-AWP are preferred.
Soldering Preparation
1. Using gauze sponges and CSM-2 Degreaser, thoroughly
degrease the surface of a smooth, clean, steel or
aluminum plate at least 1/8 in (3 mm) thick.
2. Using blunt-nose tweezers which have been cleaned
with Neutralizer 5A, remove the gage from its mylar
folder and place on the cleaned plate. Tabs on the gage
should face up.
3. Place a short piece, 1 to 2 in (25 to 50 mm), of PDT-1
drafting tape over the grid area of the gage, leaving one-
half of the soldering tabs exposed.
4. Press tape over gage and onto work surface (Figure 1).
If the gage does not have good thermal contact with the
surface, a weight can be applied to hold gage in place.
Leadwire Preparation
1. Select wire type and size to be used according to
recommendations outlined in Leadwire Selection.
2. Thermally strip approximately
1
4 in (6 mm) of insulation
from end of wire to be attached to gage. (Thermal
stripping prevents broken or nicked wires often resulting
from mechanical stripping methods.)
T
E
C
H

N
O
T
E
TT-608
Micro-Measurements
204
Techniques for Attaching Leadwires to Unbonded Strain Gages
3. Tin exposed wire with solder selected.
4. Trim tinned conductor so that approximately 0.1 in
(2.5-mm) extends from end of insulation. Longer tinned
conductor length is not advisable since it may extend
beyond gage backing and contact specimen during gage
bonding and clamping.
Note: Whenever possible, a three-leadwire system should
be employed on single gage (quarter-bridge) installations.
The use of two-wire systems should be limited to dynamic-
only strain measurements or in static measurements where
the two-lead system is not more than a few inches long and
can be connected to a three-lead cable for routing to the
measuring instrument.
Soldering General
The use of a temperature-regulated soldering iron is
always recommended for strain gage soldering. This
is particularly critical when soldering is to be done on
unbonded gages. Unbonded gages have less capacity to
dissipate soldering heat.
Soldering stations having variable heat control should
be adjusted to a tip-temperature high enough to allow
quick melting of the solder without vaporizing the ux or
hindering proper cleaning and tinning of the iron tip.
Note: On Mark V and Mark VIII Soldering Units, this
adjustment is automatically accomplished by turning
heat selection on the front panel to the range of the solder
being used.
When using soldering irons without heat controls, they
should be powered through a variable transformer. A
setting of 70 to 75% of line voltage is a good starting
position for adjusting tip temperature.
Soldering Procedures
The object of these procedures is to attach the leadwire to
the gage tab with the least amount of solder practicable.
1. Adjust tip temperature as described in Soldering
General.
2. Clean and tin the soldering tip, adding enough solder
to form a small pool. A screwdriver-shaped tip is
preferred.
3. Place one drop of M-Flux AR on gage tab or terminal
to be wired.
4. Hold tinned wire in place on tab. Drafting tape is often
helpful to keep wire from moving during soldering.
5. Without adding more solder to iron, frmly press fat
surface of iron tip onto wire and tab and hold for one
second.
6. Repeat steps 3 to 5 on each tab to be wired, adding
small amounts of solder to iron when necessary before
each tab is soldered.
7. Inspect connections. Incomplete or peaked solder
joints should be resoldered with additional ux on gage
tab and limited amounts of solder reapplied to iron tip.
Gage Removal And Cleaning
1. After soldering has been completed, thoroughly ood
the gage and tape area with several brush applications
of RSK Rosin Solvent.
2. Work the solvent brush around solder areas and under
tape. Do not try to lift gage/leadwire assembly until
rosin solvent has loosened tape from gage face.
3. Lift gage assembly by the leadwires and dip into a
container of clean rosin solvent. Scrub the solder
connections and gage face with brush provided with the
rosin solvent.
4. Remove assembly from rosin solvent and blot dry
between layers of gauze sponges. Use caution not to
snag gauze on gage backing or solder connections.
5. Place gage on a clean work surface with the bonding
surface exposed. Wipe with a cotton applicator slightly
moistened with Neutralizer 5A. Rolling the cotton
applicator as you wipe will improve cleansing. Allow to
air dry 5 minutes.
6. Slip cleaned gage into acetate folder and tape shut. This
will minimize contamination and provide lead support
until ready for installation.
Tape Over Gage Face
Leaving Half of
Tab Exposed
Metal Plate
Figure 1
T
E
C
H

N
O
T
E
TT-608
Micro-Measurements
205
Techniques for Attaching Leadwires to Unbonded Strain Gages
Additional Suggestions
1. Before bonding, make a resistance and continuity check
of the gage/leadwire assembly. Test instruments should
not apply more than 1 to 2 volts during measurement.
2. Follow instructions for gage bonding according to the
adhesive selected. When positioning gage for adhesive
application and alignment on specimen, handle gage
by the leadwires to minimize force applied to the solder
connections. Tape leadwires to surface after gages are
in position.
MICRO-MEASUREMENTS
Strain Gage Soldering Techniques
T
E
C
H

N
O
T
E
207
Introduction
The most common method of maki ng el ectri cal
connections in strain gage circuits is by means of soft
solders, in wire form. Other methods, such as spot
welding, brazing, compression bonding, paste solders,
and conductive epoxies, are also available, but nd only
limited application. Solders have many advantages for
strain gage use they are low in cost, readily available in
various alloy compositions to provide a range of melting
temperatures, and are easily obtained in the form of either
solid wire or wire with a core of ux. They are convenient
to use, and offer an excellent combination of electrical and
mechanical properties.
Although soldering is basically a simple procedure,
it must be done with appropriate tools, supplies, and
techniques to assure accurate strain measurement. This is
particularly true when test requirements are severe in the
sense of approaching the limits of the strain gage circuit
capabilities; e.g., long-term stability, high-elongation
measurements, fatigue endurance, etc. Use of improper
materials or techniques can signicantly degrade strain
gage performance.
The purpose of this Application Note is to outline
recommended procedures and materials for attaching
leadwi res to strai n gage solder tabs or to bonded
printed-circuit terminals. These reliable, experience-
proven methods are based on the use of a professional
quality soldering station, in conjunction with Micro-
Measurements solders and installation accessories.
Soldering Station and Pencil
For precision soldering of strain gages, it is always necessary
to use a temperature- or power-controlled soldering station
that provides low voltage and adjustable temperature to
the soldering iron tip. An unregulated soldering iron,
connected directly to the power line, is not ordinarily
suitable for strain gage use because the tip temperature is
apt to be far too high. This tends to oxidize the tip, and to
instantly vaporize the ux, making soldering much more
difcult. In addition, the unnecessarily high temperature
may damage the strain gage, the bonding adhesive, or even
the test specimen. For these reasons, the soldering station
should incorporate provision for adjusting the soldering
temperature to suit varying installation conditions and
requirements. The temperature must be adjusted, of
course, to accommodate the melting points of the different
solders commonly used for strain gage connections, but
also to allow for environmental conditions such as drafts
or outdoor soldering in cold weather. Moreover, the
temperature controller should be carefully designed to
ensure that it does not generate electrical noise that could
adversely affect nearby measuring instruments when both
are in use.
Design of the soldering pencil also requires special
consideration. It should be light in weight, with a very
exible power cord, and with the gripping area thermally
insulated from the heating element. These characteristics
contribute to the comfort, ease, and precision of soldering,
and minimize operator fatigue during long periods of
use. The soldering tip itself should be of the at, chisel, or
screwdriver type. Pointed tips should not be used, because
they tend to draw solder away from the work area, and thus
make it more difcult to achieve a proper joint. In contrast,
at tips act to conne the solder, while offering greater
surface area for better heat transfer and more effective
soldering, generally.
Micro-Measurements soldering units incorporate all of the
above features and a number of others, designed to help the
user easily make consistent, reliable solder joints. These
soldering units are widely used by professional strain gage
installers everywhere, in both stress analysis laboratories
and in transducer manufacture.
Solder Selection
The Micro-Measurements Division stocks a broad range
of solder types to meet various installation and test
require ments. While solders are sometimes selected to
provide specic electrical or mechanical properties, the
most common basis for selection is simply the melting-
temperature range. Low-melti ng-poi nt solders, for
example, are generally used for strain gage installations
on nonmetallic test parts to avoid damaging the gage,
bonding adhesive, or test material due to overheating. In
contrast, high-temperature solders are normally selected
only when necessary to satisfy elevated-tem perature
testing requirements. These solders are somewhat more
difcult to handle because the higher working tem perature
T
E
C
H

N
O
T
E
TT-609
Micro-Measurements
208
rapidly vaporizes the f lux, and oxidizes the soldering
tip, both of which tend to impede the soldering process.
Specially designed soldering tips are recommended for
high-temperature use.
For routine applications, where test conditions do not
dictate the use of either a low- or high-temperature solder,
an alloy with an intermediate melting temperature is the
normal selection. The 63/37 tin-lead alloy ( Type 361A-
20R) is an excellent choice for general-purpose strain
gage soldering. As an eutectic alloy, it has a sharply
defined melting tem perature a characteristic that
largely eliminates cold solder joints. The addition of a
trace of antimony provides superior performance when
the soldered connections will be exposed to very low
(cryogenic) temperatures for long periods of time.
The general-purpose solders are supplied with a core of
activated rosin f lux. This makes soldering much more
con venient, and is particularly useful in eld applications
where accessory liquid rosin ux (M-Flux AR) may not be
available. Solid-wire solder, with externally applied acid
ux (M-Flux SS), is recommended for making soldered
connections to Micro-Measurements K- and D-alloy
(modied Karma and isoelastic) strain gages. Rosin-core
solders should not be used in conjunction with acid ux.
Silver solder ( Type 1240-FPA) is available for applica-
tions where leadwire connections will be exposed to
temperatures above about +550F (+290C). This solder,
in paste form, is not suitable for attaching wires directly
to strain gage solder tabs or to bondable terminals, but is
intended for connecting instrument leads to preattached
strain gage leads, as with WK-Series gages using a special
resistance soldering unit. Techniques for making leadwire
connections with silver solder are described in Micro-
Measurements Application Note TT-602, Silver Soldering
Technique for Attachment of Leads to Strain Gages.
Soldering Flux
The function of a soldering ux is to remove oxidation
from the members being joined (solder tabs, terminals,
leadwires), and to prevent further oxidation during
soldering. For making leadwire splices, or soldering
directly to constantan foil or copper terminals, the ux
contained in a rosin-core solder is usually sufficient.
With higher temperature solders, however, it may be
necessary to supply additional ux. A liquid activated-
rosin ux such as M-Flux AR is recommended for this
purpose.
Acid uxes should never be used on constantan strain
gages or copper terminals, or for splicing copper leadwires;
and paste f luxes, containing chlorides, should not be
used under any circumstances for strain gage soldering.
When tinning bare (without soldering options) solder
tabs of Micro-Measurements K- and D-alloy strain
gages, a liquid acid ux (M-Flux SS) is recommended.
After the tinning operation, the residual f lux must be
completely neutral i zed within one to two minutes;
and then the leadwire joint can be completed using the
same solder and M-Flux AR rosin ux or a rosin-cored
solder.
Preparation of the Soldering Tip
New soldering tips should always be tinned with solder prior
to initial use. This is easily accomplished by wrapping one
to two in (25 to 50 mm) of solder wire around the working
portion of the tip while the soldering iron is cold, before
applying power to the soldering station. If rosin-core
solder is used, no external ux is required. With solid-wire
solder, however, the wrapped tip should be dipped into
liquid rosin ux (M-Flux AR) to provide sufcient ux
for initial tinning. Set the control on the soldering station
to the appropriate temperature range for the solder, and
apply power to the unit. Allow the soldering pencil to heat
until the solder wrapped around the tip melts completely.
Remove excess melted solder from the tip with a dry gauze
sponge. Never knock the heated soldering pencil against
any object to remove excess solder, since this may result in
personal injury or damage to the soldering pencil.
Note: Cross-alloying of solders can change the electrical,
chemical, thermal and mechanical properties of the solder
being used. To prevent cross-alloying, it is recommended
that only one type of solder be used with each soldering
tip. Of course, if one type of solder is incorporated in a
gage with solder dots and another type is added, a mixture
is produced. This mixture cannot be expected to have
melting and strength properties any better than those of
the lower temperature component.
Oxidation of the soldering tip seriously hinders the
soldering operation. The tendency for oxidation can be
minimized by ensuring that excess melted solder remains
on the tip at all times when it is not actually in use.
Negligent maintenance practices, or wiping the hot tip
with materials that char on the surface, will produce a
buildup of oxide that prevents proper soldering. If the tip
does become oxidized, the following procedure is effective
for cleaning and re-tinning:
1. Set the soldering station to the appropriate temperature
range for the solder in use.
2. Place several drops of M-Flux SS on a glass plate.
Re-tin the soldering surface by holding the heated tip in
the SS ux while feeding solder onto the tip. A generous
amount of solder is essential for proper tinning.
T
E
C
H

N
O
T
E
TT-609
Micro-Measurements
209
3. Wipe the excess solder from the tinned tip with a dry
gauze sponge. For severely oxidized tips, it may be
necessary to repeat this operation several times to
obtain a properly tinned surface. The soldering tip
should never be led or sanded, since this may remove
the plating on the tip, accelerating the oxidation and
leading to the early deterioration of the tip. After the
cleaning operation, remove excess solder, re-tin and
clean the tip several times, using rosin-core solder, or
solid-wire solder withM-Flux AR.
Tinning Solder Tabs and
Bondable Terminals
All strain gage solder tabs, terminals, and leadwires must
be properly tinned before making soldered connections.
This helps ensure active surface wetting and good heat
transfer during the soldering operation. Tinning stranded
leadwires to produce a formable solid conductor will also
greatly simplify the leadwire attachment procedure.
Before tinning the solder tabs on open-face (unencapsulat-
ed) strain gages, the measuring grid should be protected
with PDT-1 drafting tape. The drafting tape is positioned
to cover the entire grid and the upper portion of the solder
tabs, as shown in Figure 1. This not only shields the grid
from soldering flux and inadvertent solder splash, but
also restricts the ow of solder on the tabs. The tinned
area on the solder tabs should be only large enough to
easily accommodate the leadwire size in use. The latter
consideration is particularly important when making
installations for dynamic applications or large-strain
measurement.
The tinning procedure for strain gage tabs and terminals
consists of rst cleaning and reapplying a small amount of
solder to the hot soldering iron tip. Next, apply a drop of
M-Flux AR to the tab or terminal (this step can be omitted
if a rosin-core solder is used). When soldering directly to
bare Karma or isoelastic foil, use M-Flux SS on the gage
tabs only. Hold the soldering pencil in a nearly horizontal
position (<30), with the at surface of the tip parallel to
the solder tab or terminal. Place the solder wire at on the
gage tab, and press rmly with the tinned hot soldering tip
for about one to two seconds, while adding approximately
1/8 in (3 mm) of fresh solder at the edge of the tip. This
procedure assures that there is sufcient solder and ux
for effective tinning. Simultaneously lift both the soldering
pencil and solder wire from the tab area.
NOTE: Lifting the soldering iron before lifting the solder
may result in the end of the solder wire becoming attached
to the tab; lifting them in the reverse order can leave
a jagged (spike) solder deposit on the tab. When the
operation is performed properly, it will produce a small,
smoothly tinned area on the tab or terminal.
If M-Flux AR or a rosin-core solder is used in the tinning,
it is not necessary to remove the residual soldering ux
at this time. However, when M-Flux SS is employed to
tin the bare solder tabs of K- or D-alloy gages, the acidic
ux residue must be removed immediately following the
tinning operation. To remove the residue, apply M-Prep
Conditioner A liberally, and wash the area with a soft
brush; then blot dry with a clean gauze sponge. Next, wash
again with freely applied M-Prep Neu tralizer 5A, and blot
dry with a clean gauze sponge.
NOTE: Special procedures for tinning and wiring strain
gages supplied with preattached solder dots are described
i n Micro-Measurements Appl ication Note TT-606,
Soldering Techniques for Lead Attachment to Strain Gages
with Solder Dots.
Tinning and Attaching Leadwires
Of course, leadwire ends must be stripped of insulation
before tinning, and this should be done with a thermal
wire stripper to avoid the damage to the wire that often
occurs when mechanical wire strippers are used. After
the wires are stripped, the ends of stranded conductors
should be twisted tightly together before tinning. The bare
leadwire ends can then be tinned easily with the following
procedure:
1. Remove excess solder from the soldering tip, using a
dry gauze sponge. Then melt fresh solder on the hot tip
to form a hemisphere of molten solder about twice the
diameter of the wire to be tinned.
2. If rosin-core solder is used, slowly draw the bare wire
through the molten solder while continuously adding
fresh solder to the interface of the wire and soldering
tip. With solid-wire solder, apply M-Flux AR to the
Figure 1 Gage grid and upper portion of
solder tabs masked with drafting tape.
T
E
C
H

N
O
T
E
TT-609
Micro-Measurements
210
wire end before starting to tin, and proceed in the same
manner. This will produce a smooth, shiny coating of
solder over the bare wire.
For applications employing bondable terminal strips and
stranded instrumentation wire, it may be convenient to use
a single strand of the wire as a jumper between the terminal
and the strain gage solder tab. In such cases, the single
wire strand should be separated out before twisting and
tinning the remaining strands (see Micro-Measurements
Application Note TT-603, The Proper Use of Bondable
Terminals in Strain Gage Applications).
Leadwires should be formed and routed to the strain
gage or terminal strip, then firmly anchored to the
test-part surface with drafting tape before making the
soldered connection. Attempting to route the leadwires
after completi ng the solder joi nt wi l l often result
in damage to the gage or terminals. Routing into the
connection area should be along a minimum strain
direction (such as the Poisson direction in a uniaxial
stress eld) particularly for high elongation or dynamic
tests. The tinned leadwire end should be trimmed short
enough so that it will not protrude through the connection
area, and cannot inadvertently make electrical contact
with the test-part surface or adjacent solder connections.
Figure 2 illustrates this stage in the procedure. In the nal
preparatory step, bend the leadwire end slightly to form
a spring-like loop, and tape the wire rmly in place over
the connection area, using PDT-1 drafting tape. The tape
should be within about 1/8 in (3 mm) of the connection
area, as shown in Figure 3.
Clean and re-tin the soldering iron tip with fresh solder.
The temperature of the iron should be adjusted so that the
solder is easily melted, without rapidly vaporizing the ux.
If the iron temperature is either too low or too high, it may
cause poor solder connections, or it may damage the strain
gage, terminal, or bonding adhesive. Apply a small amount
of M-Flux AR to the joint area and, holding the soldering
pencil nearly horizontal, rmly press the at surface of the
tip on the junction for about one second; then lift the tip
from the soldered joint. If needed, additional ux can be
provided during the joining operation by feeding a little
fresh solder into the joint from a spool of rosin-core solder.
This procedure should result in a smooth, hemispherical
solder joint, without any peaks or jagged areas. If the
solder joints are not smooth and uniform in size, repeat the
soldering procedure, using additional ux and/or solder as
necessary.
Cleanup and Inspection of
Soldered Joints
After completing the soldering operation, it is imperative
that all traces of residual f lux be completely removed
with RSK Rosin Solvent. The same solvent is used to
soften the mastic of the drafting tape, permitting its easy
removal. Do not try to pull away the tape with tweezers
or other tools, because this may result in damage to the
soldered connections or the strain gage grid. Thoroughly
clean the entire installation area with generously applied
rosin solvent and a soft-bristled brush. Clean the solder
connection area until no visible signs of residual f lux
remain, and blot the area dry with a clean gauze sponge.
Any traces of residual f lux can cause gage instability
and drift, and will inhibit bonding of the installations
protective coating. Incompletely removed soldering ux is
the most common cause of degraded performance in strain
gage installations. Residual ux mixed with a protective
coating application can completely destroy the coating
objective.
Visually inspect the soldered joints for any gritty or jagged
joint surfaces, and for traces of ux. Solder connections
Figure 2 Trimming leadwire ends before taping in place. Figure 3 Leadwire end taped to surface
in preparation for soldering.
T
E
C
H

N
O
T
E
TT-609
Micro-Measurements
211
should be smooth, shiny, and uniform in appearance.
Any soldered joints that look questionable should be
re-soldered, and ux removed. Check the resistance-to-
ground of the completed gage installation, using the Model
1300 Gage Installation Tester. Low or marginal resistance
readings suggest a leakage path between the soldered
connections and the test-part surface. This condition
usually results from residual soldering ux, or from bare
leadwire conductors partially shorting the gage tabs or
terminals to the test part. Soldered joints should not be
tested by pulling on the leadwire, or by probing at the joint
area. These practices frequently cause lifting or tearing of
the solder tab from the gage backing material.
Summary
The ability to make consistently good soldered joints is
essential for precision strain gage measurements. The
techniques described here are straightforward and easily
mastered, but they are most effective when used with
professional soldering equipment which is specially
designed for making soldered connections in strain gage
circuits. The soldering pencil should be lightweight, with
a at chisel or screwdriver tip, and it should be connected
to the soldering station with a very exible power cord.
Requirements for the soldering station include low-voltage
operation of the soldering pencil, and provision for
temperature adjustment to suit the type of solder and the
application conditions. The equipment should not generate
electrical interference that could affect sensitive measuring
instrumentation. Solder selection is based primarily on the
expected operating temperature range of the strain gage
installation; and all solder tabs, bondable terminals, and
leadwire ends should be tinned before soldering the joints.
Soldered joints should always be smooth and shiny, with
no jagged or irregular edges, and all traces of residual ux
must be thoroughly removed prior to the application of
protective coating. Use of the recommended materials and
techniques, with careful attention to detail, will result in
consistently proper and reliable soldered connections.
MICRO-MEASUREMENTS
Strain Gage Clamping Techniques
T
E
C
H

N
O
T
E
213
Introduction
All organic strain gage adhesives require that some form of
clamping pressure be applied and maintained throughout
the curing stage. In the case of the fast-curing M-Bond
200 ad hesive, clamping pressure is applied with the thumb
or nger, and maintained for a minimum of one minute.
How ever, M-Bond 200 is not recommended for elevated-
tempera ture or long-term strain measurements. For these
applications, or when gages must be installed in conned
areas such as inside tubing or bored holes, epoxy or epoxy-
phenolic ad he sives are commonly selected.
With epoxy-based adhesives it is always necessary to
maintain a specied uniform clamping pressure while the
adhesive is curing. The instruction bulletins accompanying
all Micro-Measure ments strain gage adhesives include, in
each case, the recommended clamping pressure and curing
cycle. Since the curing process usually takes several hours
and may involve elevated temperatures, it is important that
the clamping device be physically stable and capable of
holding the specied force for the required time.
This Application Note describes several popular, and some
unusual, gage clamping methods and hardware.
Clamping Hardware
The mechanical arrangement used to obtain a steady, non-
shifting clamping force obviously depends on the nature
of the test object and the gage installation. Sometimes a
simple spring clamp will sufce. In other cases, it may be
necessary to devise a more elaborate clamping xture. For
any type of clamp, the assembly generally consists of four
components:
1. A release lm is used between the gage/adhesive and
the pressure pad. It prevents the pressure pad from
adhering to the adhesive layer. For low-temperature-
curing epoxies [<+150F (+66C)] such as AE-10/15
and GA-2, this lm is usually gage installation tape
(PCT-2M). For higher-temperature-curing epoxies,
such as 600, 610, 43-B and GA-61, it is usually a sheet of
Teon lm (TFE-1) or Mylar tape (MJG-2).
2. A resilient rubber pressure pad is used to provide a
uniform clamping pressure over the gage area. It should
be soft enough to conform to slightly irregular surfaces
but not so soft as to extrude from under the clamping
plate. Re c om mended hardness is durometer A40-60.
Silicone gum or silicone rubber are preferred because of
their high-tempera ture capability.
3. A metal clamping plate serves to distribute the clamping
force over the entire area of the pressure pad. It should
be formed to match the contour of the pressure pad or
test part and should have sufcient thickness to prevent
deformation under normal clamping force.
4. A force application device is used to apply steady
force on the metal clamping plate. It could range from
simple dead weights to sophisticated vacuum pads with
integral heaters.
Calculating Clamping Pressure
All Micro-Measurements instruction bulletins for epoxy
adhesives specify a recommended clamping pressure.
Specied in pounds per square inch (psi) or kilonewtons
per square meter (kN/m
2
), clamping pressure is calculated
by dividing the applied force by the surface area of the
rubber pad in contact with the strain gage and specimen
surface. Normally the desired clamping pressure is known
from the instruction bulletin, and the proposed pressure
pad area can be measured. Therefore, the necessary
quantity to be determined is clamping force.
Registered trademark of DuPont.

Figure 1
T
E
C
H

N
O
T
E
TT-610
Micro-Measurements
214
As an example, the gage shown in Figure 1 is to be bonded
with M-Bond 610 adhesive at a recommended clamping
pressure of 45 psi (312 kN/m
2
). The gage pattern selected
can be con ven iently covered with one of the clamping
pads from the GT-14 pad kit. These pads measure
0.5 x 1.25 in (13 x 32 mm). The necessary clamping force
can be computed by multiplying the required clamping
pressure by the pad area:
clamping force = clamping pressure x pad area
= 45 psi x 0.625 in
2
(312 kN/m
2
x 0.0004 m
2
)
= 28 lbs (125 N)
Thus a 28-lb (125-N) clamping force for the example in
Figure 1 would yield the recommended 45 psi (312 kN/m
2
)
clamping pres sure.
Clamping Techniques
A selection of gage clamping techniques is shown in the
sketches that follow. Whether one of these or another
clamping scheme is used, there are two precautions that
need to be observed in every case. The rst is to always
apply the clamp ing force through some form of spring
(preferably low-rate). Rigid clamps (C-clamps, hose clamps,
etc.), such as those shown in Figures 4 and 5, should not be
used without a spring in series with the clamping force.
The spring is neces sary to ensure that the force does not
vary signicantly with small dimensional changes which
may occur as the adhesive f lows and cures, or as the
temperature changes. The second is to always provide a
means for determining the actual clamping force, thus
making certain that it is within the range specied by the
supplier. This can be done by precalibrating the spring
for force versus deection, or by direct measurement on
the clamping assembly with an inexpensive spring scale
(e.g., a sh scale). The latter technique is particularly
appropriate for arrangements such as those shown in
Figures 2 and 7.
The selection of additional clamping methods is limited
only by the imagination of the gage installer. Following the
guide lines provided in this Application Note, there should
be very few ap plica tions where adequate gage clamping
pressure cannot be achieved.
For further assistance in applying the clamping techniques
illustrated, contact our Applications Engineering Depart-
ment.
Figure 2 Common spring clamp available in various sizes.
Clamping force exerted should be checked at different
jaw openings, as force may vary.
Figure 3 Spring clamp used in a reverse (pushing)
mode. Force can be checked by pressing handles
down on platform or bathroom scale.
Figure 4 Hose clamp with spring in series to
provide constant and measurable force.
Figure 5 C-clamp with compression spring that can be
removed and checked for force versus deection.
T
E
C
H

N
O
T
E
TT-610
Micro-Measurements
215
Figure 6 Magnetic or metal block that can be attached
to structure with dental cement or M-Bond 200.
Spring can be calibrated separately
Figure 7 Rubber-band clamp (usually surgical tubing) wrapped
tightly around structure can be force-calibrated with spring scale.
Figure 8 Silicone rubber plug expands to provide
clamping force to gage inside hole. Difcult to calibrate.
Figure 9 Silicone rubber tubing has both ends plugged
and tted with air hose. Under pressure, tube expands
to provide clamping force for gage inside pipe. Gage can
also be prewired and temporarily installed upside down
on tubing before insertion in pipe. Adhesive is then applied
to gage bonding surface and assembly slid into pipe
to the proper position and inated.
Figure 10 Commercially available vacuum pads adhere to
surface through suction. Internally heated models are available.
Figure 11 Homemade vacuum clamping xture.
Putty is formed into a shallow box shape with vacuum
hose molded into side. After gage placement, a plastic
lm (usually mylar) is placed over putty frames, and the
vacuum is ap plied. Normally heat lamps are used to
cure elevated-temperature-curing epoxies.
MICRO-MEASUREMENTS
Strain Gage Installations for Concrete Structures
T
E
C
H

N
O
T
E
217
Introduction
The installation of strain gages for concrete structures
presents several unique challenges to the installer, whether
measurements are made on the concrete surface or within
the concrete, or on reinforcement bars within the structure.
For example, special preparation is required to ensure that
strains on the irregular surface of the concrete are fully
transmitted to the strain gage; and when gages are bonded
to reinforcement bars, provisions are necessary to protect
the installation from mechanical damage during fabrication
and from the hostile environment of the concrete itself.
This Application Note outlines recommendations for gage,
leadwire and protective coating selections and installations
under these conditions. The surface preparation materials
and installation accessories referenced throughout are
described in detail in the Micro-Measurements Strain
Gage Accessories Data Book.
Installing Strain Gages on Concrete
and Other Irregular Surfaces
Strain gages can be satisfactorily bonded to almost any
solid material including concrete if the surface is
properly prepared. For smooth surfaces on nonporous
materials, only the basic operations of solvent degreasing,
abradi ng, appl ication of layout l i nes, conditioni ng
and neutralizing are required. For concrete and other
materials with an uneven, rough and porous surface, an
extra operation must be added to ll the voids and seal
the surface with a suitable precoating before the gage is
bonded.
Degreasing
Use a stiff-bristled brush and a mild detergent (Figure 1)
to remove any loose soil or plant growth. Rinse with clean
water. A degreaser such as CSM-2 may be needed if oils
and greases are present. Remove surface irregularities with
a wire brush, disc sander, or grit blaster. Blow or brush all
loose dust from the surface.
Conditioning
Generously apply M-Prep Conditioner A, a mildly acidic
solution, to the surface in and around the gaging area. Scrub
with a stiff-bristled brush. Blot contaminated Conditioner
A with gauze sponges. Rinse the area thoroughly with
clean water. Reduce the surface acidity by scrubbing with
M-Prep Neutralizer 5A. Blot with gauze sponges and
rinse with water. Dry the surface thoroughly. Warming
the surface gently with a propane torch or heat gun will
hasten evaporation.
Filling
Application of a 100%-solids adhesive to the gaging area
(Figure 2) will provide a suitable gage-bonding surface.
For test temperatures up to +200F (+95C), M-Bond
AE-10 is normally used. At higher temperatures, M-Bond
GA-61 is recommended. In applying the adhesive as a
sealer to the surface, work the adhesive into any voids, and
level to form a smooth surface. After the adhesive is cured,
it should be abraded with 320-grit abrasive paper until the
base material is exposed. (If a thin adhesive, like M-Bond
200, will be used to bond the gage, the base material should
not be exposed.)
Layout Lines
Using a ballpoint pen or round-pointed metal rod,
burnish layout lines. Scrub them with Conditioner A, apply
Neutralizer 5A, and dry as before. Supplemental layout
lines may be drawn with ink on the concrete outside the
gaging area.
Figure 1 Conditioning the surface for gage installation.
T
E
C
H

N
O
T
E
TT-611
Micro-Measurements
218
Gage Bonding
Normal procedures should be followed for bonding the
gage to the prepared gaging surface. Special notice should
be paid to several points, however. First, the gage length
of strain gages used on concrete should be at least 5 times
the diameter of the largest aggregate in the concrete. This
often results in the use of patterns with gage lengths of
1 in (25 mm) or more. N2A-Series or encapsulated
EA-Series gages, which tend to lie atter during handling,
are highly recommended for their ease of installation
under these circumstances. Further, bonding with a quick-
curing adhesive, like M-Bond 200, is not recommended,
even when test conditions may warrant its use. Accurate
gage alignment and an even application of pressure as the
adhesive is cured are more difcult when bonding longer
gages. A slower curing adhesive, like M-Bond AE-10
shown in Figures 3 and 4, will allow time for realigning the
gage, if necessary. It will also enable the use of a suitable
pressure pad and clamping xture as outlined in Micro-
Measurements Application Note TT-610.
Soldering
Concrete and adhesive fillers are relatively poor heat
conductors. Accordingly, care should be taken when
soldering leads directly to the strain gage. Excessive
heating of the tabs can be eliminated by using gages with
Option W (integral printed circuit terminal), or Option P
(preattached leadwires), which is shown in Figure 5.
Attention to these procedures will help ensure successful
installations of strain gages on the surface of concrete
and other similar solids. If you have any questions about
your particular applications, contact our Applications
Engineering Department for recommendations.
Strain Measurement
Within Concrete Structures
Micro-Measurements EGP-Series Embedment Strain
Gages (Figure 6) are specially designed for measurement of
mechanical strains within concrete structures. The sensing
grid has an active gage length of 4 in (100 mm) to average
strains in aggregate materials, and is fully encapsulated in a
polymer concrete material to closely match the mechanical
properties of typical structural concrete, guard against
mechanical damage, and to protect against moisture and
corrosive attack. EGP-Series Gages incorporate a 10-ft
(3-m), jacketed, three-conductor cable for ease of use in
eld installations, and are compatible with conventional
strain measurement instrumentation.
Gage Installation
No preparation of the gage itself is required; however, as
Figure 2 Filling of the surface.
Figure 3 Adhesive application.
Figure 4 Application.
Figure 5 Finished installation with N2A-06-40CBY-120 gage
with Option P (preattached leadwires).
T
E
C
H

N
O
T
E
TT-611
Micro-Measurements
219
with bonded or welded strain gages, EGP-Series Gages
must be accurately aligned along the intended strain
measurement direction during the installation process.
Care should be taken to secure the gage in the desired
location and orientation, and to tie the leadwire cable to
any available support, before the concrete is poured. While
the Embedment Gage must be completely encapsulated
in concrete to ensure complete strain transfer from the
structure, normal pouring techniques are usually all that
are required.
Cable Splices
EGP-Series Gages are provided with a 10-ft (3-m) cable
to allow for making cable splices outside the concrete
structure. When splices are required, all connections
should be soldered and then protected from moisture and
other contamination with a suitable cable splice sealant.
Strain Gage Installation
On Concrete Reinforcement
Strain gage installation on reinforcing rods follows the same
general procedure recommended for most steel specimens.
These rods are, however, subjected to mechanical abrasion
and a moist, corrosive environment. Accordingly, the
following special attention is required:
Surface Preparation
1. Degrease with a degreaser (CSM-2) over at least a
6-in (150-mm) length of the bar at the proposed gage
location.
2. Descale and smooth the rebar around its circumference
with a grinder wheel. (Aluminum oxide or silicon carbide
abrasive of approximately 50 mesh is preferred.) A 3-in
(75-mm) length general ly provides a suff iciently
large descaled area for gage and protective coating
installations. Surface nish after this operation should
be about 180 microinches (5 m) rms.
3. Wet abrade with Conditioner A and 220-grit silicon
carbide wet-or-dry paper (SCP-1). Use sufficient
Conditioner A to prevent material from drying on the
rebar surface while abrading.
4. Wipe dry with a clean gauze sponge (GSP-1), then
repeat Step 3 (with 320-grit paper) and dry again.
5. Surface nish should be 63 to 125 microinch (1.6 to
3.2 m) rms at the completion of the second wet-
abrading operation.
6. Lay out the gage locations.
7. Scrub the installation area with Conditioner A and
a cotton applicator (CSP-1). Wipe dry with gauze
sponges.
8. Scrub the area thoroughly with Neutralizer 5A and a
cotton applicator and wipe dry with a gauze sponge
as previously noted. This step must be accomplished
thoroughly to neutralize all traces of Conditioner A
used in Steps 3 through 7.
9. Mask an area with PCT-2M gage installation tape (or
MJG-2 Mylar
) tape at the gage location to minimize

ow-out of adhesive for subsequent protective coating
application.
Adhesive Selection
M-Bond AE-10 adhesive is a good selection when a room-
temperature cure of a eld application is required. This
adhesive will cure in 6 hours at +75F (+24C). Other
adhesives that may be used, depending upon the test
environment, are M-Bond AE-15, M-Bond 600/610, or
GA-61 adhesive. Application of the adhesive should follow
the specic instructions accompanying it.
Gage Selection
CEA-Series gages are the most popular choice when the
cross section of the bar is 1/8 in (3mm) or larger in diameter.
Where very stable installations are required (e.g., for tests
Figure 6 EGP-Series Embedment Gage.
Figure 7 CEA-Series Weldable Gage.
T
E
C
H

N
O
T
E
TT-611
Micro-Measurements
220
in excess of one year) on 1/4 in (6 mm) or larger diameter
rebar rod, WK-Series gages are recommended. When
conditions are not favorable for bonding gages, CEA- and
LWK-Series Weldable Gages (Figure 7) may be used.
Leadwire Considerations
When utilizing one active strain gage (quarter-bridge
conguration), it is good practice to use a three-leadwire
system. Micro-Measurements EA- and CEA-Series strain
gages can be supplied with a preattached three-leadwire
cable (Options P and P2, respectively) to eliminate the
need for attaching leadwires at the job site, and to reduce
installation time.
Alternately, leadwires may be soldered to the strain gage
tabs after gage bonding. If a parallel or twisted cable is
used, separate the individual (leadwire) conductors for
a distance of about 1 in (25 mm) from the cable end and,
if Tef lon
-insulated cable is used, etch the insulation

with Tetra-Etch compound; if vinyl-insulated cable is
used, prime the insulation with thinned M-Coat B. These
materials should not be allowed to f low onto the bare
strands of the conductors.
After allowing the M-Coat B to air dry for at least two
hours at room temperature [about +75 (+24)], thermally
strip the leadwire ends and tin and solder the wires to the
strain gage tabs. For most rebar installations, 361A-20R
solder will give excellent results. Carefully remove all rosin
flux from the soldered connections using rosin solvent
(RSK-1) before applying the protective coating.
Environmental Protection
Apply M-Coat J to the gage i nstal lation careful ly
following the procedures outlined in Micro-Measurements
Instruction Bulletin B-147. The coating should be built up
to provide approximately
1
4 in (6 mm) thickness completely
surrounding the rebar (Figure 8) at the gage location, and
should be carried back far enough to cover the leadwire
area previously primed with M-Coat B. Allow this coating
to cure 24 hours at +75F (+24C), or 4 hours at +125F
(+50C).
As a nal step, the instrumentation leads extending from
the gage, out through the concrete, should be placed in
conduit to prevent mechanical damage to the leadwire
system. Of course, the complete installations should be
thoroughly checked with a Model 1300 Gage Installation
Tester before and after the concrete is poured. A properly
installed and protected strain gage is capable of many
years of service on embedded reinforcing bars, providing
data about load effects throughout the life of a concrete
structure from initial construction forces to unexpected
severe loading conditions.
Figure 8 Cut-away view of installation.
MICRO-MEASUREMENTS
The Three-Wire Quarter-Bridge Circuit
T
E
C
H

N
O
T
E
221
Introduction
Since the invention of the electrical resistance strain gage
more than a half century ago, the Wheatstone bridge has
become the sensing circuit of choice in most commercially
available strain gage instrumentation. This is due in large
measure to its inherent ability to:
1) detect the small resistance changes produced in the
strain gage as it follows even minute dimensional
changes on the surface of a test part under load,
2) produce a zero output voltage when the test part is at
rest, and
3) provide for compensation of temperature-induced
resistance changes in the strain gage circuit.
To varying degrees, each of these factors is essential for
accurate strain gage measurements.
In the majority of strain gage appl ications for the
determination of the state of stress on a test-part surface,
individual strain gage elements, whether from uniaxial
or rosette strain gage configurations, are connected
independently to the Wheatstone bridge in a quarter-
bridge arrangement. As discussed in the following sections,
the wiring scheme chosen to connect the strain gage to the
bridge circuit has a signicant effect on the accuracy of
measured strain data.
The Wheatstone Bridge
The Wheatstone bridge circuit in its simplest form (Figure
1) consists of four resistive elements, or bridge arms (R
1
,
R
2
, R
3
, R
4
), connected in a series-parallel arrangement,
with an excitation voltage source (E). The connection
points formed by (adjacent) pairs of bridge arms and the
leadwires from the excitation voltage source are input
corners of the bridge; and those formed by pairs of bridge
arms and the signal (e
o
) measurement leads are output
corners. It is worth noting for this discussion that each
input corner is adjacent to each output corner, and each
bridge arm is connected between two adjacent corners.
Also, if the bridge circuit is resistively symmetrical about
an imaginary line drawn through both output corners,
the output voltage e
o
will be exactly zero, regardless of the
excitation voltage level, and the bridge will be balanced.
Two-Wire Circuit
For an initially balanced bridge, if one of the resistors
in Figure 1 is replaced with a strain gage of precisely the
same resistance value and connected with two leadwires
having negligible resistance, the bridge circuit remains
at balance. But in practice the leadwires will have some
measurable resistance R
L
as shown in Figure 2 on page
222. And because both leadwires are in series with the
strain gage between adjacent corners of the bridge circuit,
the bridge arm resistance becomes R
G
+R
L1
+R
L2
, causing a
signicant lack of symmetry and an unbalanced condition
in the bridge, resulting in a non-zero output voltage at e
o
.
If the initial imbalance is modest, it may be mathematically
subtracted from subsequent measured strain readings; but
large imbalances may cause a more serious problem. As
an example, a 20-ft [6 m] length of two-conductor AWG26
[0.4 mm dia.] copper leadwire has a room-temperature
resistance value of about 1.7 ohms. Wired in a two-wire
connection with a 120-ohm strain gage, and connected to a
measuring instrument with a gage factor setting of 2.0, this
would produce an initial bridge imbalance corresponding
to about 7000. This offset may signicantly limit the
available measurement range of the instrumentation, and
also should be considered when correcting for Wheatstone
bridge nonlinearity (see Tech Note TN-507).
e
o
E
R
4
R
2
R
3
+
+

R
1
Figure 1. Basic Wheatstone Bridge Circuit.
T
E
C
H

N
O
T
E
TT-612
Micro-Measurements
222
Further, the leadwires are a parasitic resistance in the gage
arm of the bridge, and effectively reduce or desensitize
the gage factor of the strain gage, resulting in a reduced
signal output. For modest values of leadwire resistance,
the desensitization is approximately equal to the ratio
of leadwire resistance to strain gage resistance. In the
example given here, this results in about a 1.5% loss in
sensitivity.
And f i nal ly, a more serious problem may result i f
the temperature of the leadwires changes during the
measurement process. Copper leadwi res change i n
resistance approximately 22% of their room-temperature
resistance value for a 100F [55C] temperature change.
For the 120-ohm gage circuit in Figure 2, this would result
in an error equivalent to approximately 156 microstrain,
or about 4700 psi [0.03 GPa] in steel, for a 10F [5.5C]
temperature change in the leadwire system.
All three of these effects increase in severity with increased
leadwire resistance. It is worth noting that use of a 350-ohm
strain gage circuit will reduce each of these effects, but
cannot eliminate completely the associated measurement
errors. But the three-wire circuit described in the following
section will reduce the loss in sensitivity, and essentially
eliminate the initial imbalance problem and the error that
results from temperature changes in the leadwire system.
Three-Wire Circuit
With the three-wire circuit shown in Figure 3, the negative
output bridge corner is electrically moved from the top
of R
4
, to the bottom strain gage tab at the end of R
L3
.
In this conguration, leadwire R
L1
and strain gage R
G

comprise one arm of the bridge, and R
L2
with resistor R
4

the adjacent arm. For an equal-arm bridge, if leadwires
R
L1
and R
L2
are initially the same type and length, their
resistances will be equal, and the two respective bridge
arms will therefore be equal in resistance. The bridge is
resistively symmetrical about a line through the bridge
output corners, and the bridge is balanced. And regardless
e
o
E
R
4
R
2
R
3
R
G
R
L1
R
L2
+
+
e
o
E
R
4
R
2
R
3
R
G +
+
R
L1
R
L3
R
L2
Figure 2. Two-wire quarter-bridge circuit.
Figure 3. Three-wire quarter-bridge circuit.
T
E
C
H

N
O
T
E
TT-612
Micro-Measurements
223
of leadwire temperature changes, so long as the two
leadwires are at the same respective temperature, the
bridge remains in balance. Additionally, because only
one leadwire is in series with the strain gage, leadwire
desensitization is reduced about 50% compared to the two-
wire conguration. The third wire (R
L3
) in Figure 3 is a
voltage-sensing wire only, it is not in series with any of the
bridge arms, therefore it has no effect on bridge balance or
temperature stability.
While the three-wire circuit offers several advantages over
the two-wire circuit, in some special applications involving,
for example, slip rings or feed-through connectors, not
enough connections may be available for a continuous
three-wire system from the gage site to the instrument
terminals. In these cases, use of a two-wire lead system
between the strain gage and the connector, and a three-
wire circuit between the connector and the measuring
instrument, is recommended to minimize the total length
of the two-wire system.
The foregoing discussion applies primarily to measurement
of static strains using a measuring instrument with
dc-coupling between the bridge circuit and the amplier
input terminals. For measurement of purely dynamic
strains when only the peak-to-peak amplitude of a time-
varying signal is of interest, the two-wire system may
sometimes be used effectively by selecting a signal-
conditioning amplier that provides for ac-coupling of the
input signal, and blocks the effects of temperature-induced
changes in leadwire resistance.
In summary, benefits of the three-wire circuit include
intrinsic bridge balance, automatic compensation for the
effects of leadwire temperature changes on bridge balance,
and increased measurement sensitivity compared to the
two-wire configuration. The three-wire hookup is the
recommended conguration for quarter-bridge strain gage
circuits for static strain measurement. The two-wire circuit
can sometimes be used effectively for special situations
such as dynamic-only measurements with ac-coupled
instrumentation, or in static strain applications where the
length of the two-wire system can be kept very short.
An Alternate Degreasing Solvent
Application Note VMM-2
Micro-Measurements

Document Number: 11165 For technical questions, contact: micro-measurements@vishaypg.com www.micro-measurements.com
Revision: 01-Apr-10 1
When preparing the surface of a test specimen for the
installation of strain gages, the preferred general-purpose
degreasing agent for most applications is CSM-2 Degreaser.
Instructions for performing this important part of the gage
installation process when using CSM-2 are given in Vishay
Micro-Measurements Application Note B-129, Surface Pre-
paration for Strain Gage Bonding.
In some cases, the specimen may be adversely affected by
this powerful solvent. In others, company or governmental
regulations may preclude its use. When faced with these
restrictions, GC-6 Isopropyl Alcohol often may be utilized as
a suitable alternate degreasing agent.

CLEANING PROCEDURES
GC-6 is supplied in specially selected containers to avoid
contamination of the isopropyl alcohol from the container
itself. To use this solvent, pour it directly from the container
into a clean gauze sponge (GSP-1) or apply with a
compatible dispensing container, available from laboratory
supply houses. Please note, however, that GC-6 has a
relatively high flammability, and any dispenser used should
not be of the type that atomizes or produces a mist or fine
spray.
Any degreasing agent must be kept free of contaminants to
work effectively. Avoid dipping the sponge into the liquid or
holding the sponge against the mouth of the container.
The saturated sponge, folded in quarters, is used to
vigorously scrub the gaging area.
As the sponge becomes soiled, refold it to expose a clean
portion. When completely soiled, discard the sponge and
continue with a fresh sponge moistened with GC-6.
When the surface is thoroughly degreased, no soil will be
visible on the sponge after scrubbing. When this condition is
achieved, allow the solvent to evaporate from the specimen
surface for about 30 seconds. The drying time may be
reduced by immediately blotting the degreased area with a
clean, dry sponge to remove excess solvent.
Following completion of the degreasing step, continue with
surface preparation procedures as outlined in Application
Note B-129, Surface Preparation for Strain Gage Bonding.
External Bridge Completion For Strain Gage Circuits
Micro-Measurements


The Wheatstone bridge used in most strain gage measure-
ment circuits usually consists of (a) the gages for actively
measuring the strains and (b) precision resistors
incorporated in the measuring instrument for completing the
circuit, as shown below.
Typical three-wire quarter-bridge strain-gage circuit requires
three bridge completion resistors.
Sometimes, however, the circuit must be completed external
to the instrument. The most common reason, of course, is
when bridge completion resistors are not built into the
instrument; but accuracy considerations can also be a factor.
In certain test conditions, external bridge completion can
reduce the effects of noise pick-up in the leadwire system, or
eliminate the effects of contact resistance. Suitable elements
for completion of the bridge outside the instrument in these
cases include:
Bonded strain gages
Precision resistors
Bridge completion modules (BCM's)
Functional differences, along with differences in product
configuration and installation techniques, provide both
advantages and disadvantages for each type of external
bridge completion.
BONDED STRAIN GAGES
Bonding strain gages to the test part (individually or with
multiple grids on the same backing) is a good choice when
large changes in test temperature are expected, since this is
the only completion method that will provide temperature
compensation. Installation is relatively easy and
inexpensive. Both active and completion gages must be of
the same alloy, with the same S-T-C number. For highest
accuracy, all should be from the same production batch of
foil. Of course, the temperature of gages in adjacent arms of
the bridge must always be identical for effective temperature
compensation. Further, care must be taken to install the
completion gages in a region that (a) is free from applied
strains, and (b) can undergo thermal expansion without any
mechanical restraint.
PRECISION RESISTORS
Temperature-stable precision resistors are an alternative to
strain gages for bridge completion. They are particularly
useful when nonstandard bridge resistances are required,
and they are also beneficial when higher power levels must
be dissipated within the bridge circuit. Precision resistors are
designed to be very insensitive to temperature and,
therefore, cannot be used to provide compensation for any
thermal output from the active gages in the circuit. Unlike
strain gages, precision resistors are usually not attached
directly to the test part. Rather, a terminal strip is typically
bonded to the part for anchoring the resistor leads. As with
strain gages, care must be taken to protect both the leads
and solder joints from the test environment, making
installation somewhat demanding.
BRIDGE COMPLETION MODULES
Micro-Measurements MR-Series Bridge Completion
Modules (BCM's) were introduced in 1989. Since then, they
have become the preferred method of bridge completion at
the gage site when temperature compensation within the
bridge is not required. Even more insensitive to temperature
than precision resistors, they are certainly the easiest to
install. These networks of matched resistive grids come
ready for use with quarter and/or half bridges of active gages.
Their integral solder tabs minimize the number of solder
connections required to complete the circuit, while a foam
R
4
R
2
R
3
R
1
V out V in
External Bridge Completion For Strain Gage Circuits
Micro-Measurements

www.micro-measurements.com For technical questions, contact: micro-measurements@vishaypg.com Document Number: 11168
2 Revision: 01-Apr-10
tape attaches each BCM to the test part. This foam
decouples the BCM from any strains present in the
underlying surface.
While the subject of external bridge completion is too broad
to adequately cover here, much more detailed information is
available in the technical literature. The Tech Notes on
thermal output in strain gages (TN-504, Strain Gage Thermal
Output and Gage Factor Variation with Temperature) and
noise in strain gage circuits (TN-501, Noise Control in Strain
Gage Measurements) are particularly useful in this regard.
And, as always, our Applications Engineering Department
will be pleased to help you select the right compensating
strain gage, precision resistor, or bridge completion module
for your particular strain measurement application.
How to Install a Strain Gage in a Small, Deep Hole
Micro-Measurements

It is sometimes necessary to install a strain gage deep in a
hole which is too small to permit using conventional gage
placement and bonding methods. When this situation arises,
the technique described here (modified as appropriate to suit
specific needs) can often be used very effectively in solving
the problem.
SURFACE PREPARATION
As in all strain gage installations, the first operation is to
prepare the surface of the hole for gage bonding. The
procedure for doing so generally parallels that for bonding a
gage to any other surface (see Micro-Measurements
Application Note B-129, Surface Preparation for Strain Gage
Bonding):
1. Thoroughly clean the hole of all metal chips and loose
dirt; then degrease with CSM-2 Degreaser or other
degreasing agent compatible with the test material.
2. If the hole surface is rough from machining, and cannot
be reamed, it should be sanded to the proper finish with
abrasive paper. This can be done by taping the abrasive
paper to the end of a long, slender rod, and working the
rod back and forth in the hole. After abrading, step 1 must
be repeated.
Note: In the following steps it is preferable to position the test
object with the open end of the hole facing downward so that
excess liquids tend to drain out instead of collecting at the
bottom of the hole.
3. Wrap a clean gauze sponge around the end of the rod,
dip it in M-Prep Conditioner A, and scrub the hole
thoroughly.
4. Repeat step 3 with a clean, dry gauze sponge to wipe the
hole dry.
5. Repeat steps 3 and 4 until a clean sponge remains clean
after the operation.
6. Using a clean gauze sponge dipped in M-Prep
Neutralizer 5A (instead of Conditioner A), repeat step 3.
7. Repeat step 4.
GAGE PREPARATION
It is next necessary to select the strain gage and mount it on
a temporary carrier for accurate placement and alignment in
the hole. The preferred type of gage for this application is one
of Micro-Measurements CEA-Series, because of its
combination of ruggedness, flexibility, and ease of lead
attachment. Proceed as follows:
1. Select a strain gage with a grid size and pattern suitable
for the strain-measurement task.
2. Preattach leadwires to the gage as instructed in
Application Note TT-608, Techniques for Attaching
Leadwires to Unbonded Strain Gages.
3. Cut a rectangular piece of Mylar to the dimensions shown
below. This will serve as the temporary carrier for gage
placement.
4. Adhere the gage, top-side down, to the carrier with Duco
or rubber cement.
GAGE BONDING
After the cement has dried, and the gage is attached to the
temporary carrier, bonding inside the hole is done as follows:
1. Apply a thin layer of M-Bond AE-10 (room-temperature-
curing epoxy) adhesive to the exposed back of the strain
gage, and to the bonding area in the hole.
2. Insert the rolled-up carrier into the hole to a pre-marked
depth (distance A minus B in illustration below), and
secure it in this position with adhesive tape.
Mylar carrier
Gage adhered to carrier
grid side down,
bonding side up
t
A
x (D - 2t)
D
B
A
Air Pressure
Surgical Tubing
Mylar carrier
How to Install a Strain Gage in a Small, Deep Hole
Micro-Measurements

3. Carefully route the leadwires along the inserted sheet,
and tape them to the test piece surface where they come
out of the hole.
4. Insert a length of surgical tubing (sealed at the end) into
the hole until it is beyond the gaging area by at least 2
hole diameters. The outside diameter of the tubing
should be about 0.040 in (1mm) smaller than the hole.
5. Apply air pressure to the tube at about 10psi (70kPa), and
maintain the pressure for at least 6 hours, and preferably
24 hours.
6. Remove the air pressure and then remove the surgical
tubing.
7. Slowly and gently peel the carrier away from the gage.
Alternatively, trim off the protruding portion of the carrier
flush with the hole edge, and leave the remainder in the
hole.
In Search of the Perfect Zero
Micro-Measurements

The Wheatstone bridge, developed more than a century ago
to accurately measure the resistance of electrical
components, is the sensing circuit in which electrical
resistance strain gages are most commonly used. The
output signal from the bridge depends upon the values of the
resistive components in each of the four arms of the bridge.
The bridge is said to be zero-balanced when the resistance
of the components in the arms yield zero bridge output. This
is an ideal condition seldom achieved in practice.
Nearly all strain-gage instrumentation and transducer
read-outs incorporate the Wheatstone bridge circuit.
Typically they contain a power supply for providing a voltage
to the power corners of the bridge, an amplifier for increasing
the output from the bridge to useful levels, and, in many
cases, a read-out device for indicating the strain or other
units of measurement. The resistive components in the arms
of the bridge may be either bridge completion resistors, or
one or more active strain gages in a transducer or installed
on a test part.
Most strain-gage-based transducers and bridge completion
modules are designed to present a combination of resistive
components to the Wheatstone bridge circuit that will
produce zero output when no loads or load-induced strains
are present. Few, if any, completely achieve this goal
because of tolerances in the resistive values of the
components. Some small - and normally insignificant - output
will usually be present. However, incorrect gage installations
or damage to components in the bridge circuit can lead to
large imbalances requiring attention before accurate
measurements can be made.
The extent of any initial zero imbalance can be readily
determined, of course, from the values indicated by
strain-gage instrumentation or the transducer read-out. For
this to be an accurate indication of bridge balance, however,
the output of the instrument itself must be zero when the
output from the signal corners of the bridge is zero. The star
bridge shown here is commonly used for achieving that
condition. The resistors between S+ and S- simulate the
impedance of the Wheatstone bridge circuit while providing
a zero signal to the amplifier input. The pair of resistors
between P+ and P- places a test load on the power supply
and provides a reference to circuit common for output circuits
requiring one.
The star bridge is constructed with four resistors. Each
should have a resistance value of approximately half the
impedance of the Wheatstone bridge circuit. Therefore,
175-ohm resistors should be used in the star bridge if
350-ohm gages are present in the Wheatstone bridge.
Bridge completion resistors with values of 60, 175 and 500
ohms are available in the Micro-Measurements Strain Gage
Accessories databook. (For instrument that incorporates
resistive balancing circuits, the values of the star-bridge
resistors between the power corners must be as nearly equal
in value as possible for maximum accuracy.)
To use the star bridge, disconnect the sensor circuit from
your instrument and connect the star bridge in its place. After
turning your instrument ON, the balance controls can then be
adjusted to produce zero output. (If you are uncertain of how
this is done, consult the instruction manual supplied with
your instrument for the proper method.) When the star bridge
is disconnected and the active components are reconnected,
any output indicated will be the result of imbalance in the
Wheatstone bridge circuit itself.
The causes of any large imbalances in the Wheatstone
bridge itself should be isolated and corrected. Smaller
imbalances can be virtually eliminated by simply readjusting
the balance of the instrument, or by subtracting the initial
imbalance from subsequent measurements.
R
4
R
2
R
3
R
1
P+
S+
P
S
Wheatstone Bridge
Excitation Signal
R
4
R
2
R
3
R
1
P+
S+
P
S
Star Bridge
Excitation
Signal
Installation Verification
Micro-Measurements


Zero-shift and zero-drift in the resistance of strain gages can
arise from a number of sources, including the gage
installation itself. This series describes the necessary steps
for the verification of strain gage installations to ensure
stability.
AFTER BONDING
A recommended checklist follows for inspection of
installations prior to both leadwire attachment and
application of protective coatings. These, by necessity, are
limited to visual inspections. Magnification is helpful when
dealing with small gages.
Adhesive Uncured
Use a DPR-1 Dental Probe to check the exposed adhesive
on the specimen surface beside the gage backing after
bonding. If adhesive is soft or rubber-like, either complete the
cure (if possible) or replace the installation.
Gage Misaligned
Visually inspect to ensure that the alignment marks on the
gage are in the desired orientation. Resulting errors are
discussed in Micro-Measurements Tech Note TN-511,
Errors Due to Misalignment of Strain Gages.
Misaligned Aligned
Backing Unbonded
If visibly obvious, gage should be replaced. If in doubt,
recheck electrically for zero-return and stability under load
after leadwire connections have been made.
Grid Wrinkled or Creased
Installer must judge whether caused by normal undulation of
adhesive or by mishandling during bonding. If obviously
wrinkled or creased, the installation should be replaced.
Bubbles
Usually a concern only when directly under gage grid. If
bubble area is small in comparison to grid, check for
zero-return and stability under load when electrical
connections have been made.
Unacceptable Bubble-free
Bumps
Usually caused by entrapped foreign matter, undissolved
crystals of adhesive resin, or filler. May reduce fatigue
endurance. Usually a concern only when directly under gage
grid. If particle is small in comparison to grid, check for
zero-return when electrically connected.
Unacceptable Bump-free
Residual Tape
Adhesive tapes (and their mastics) used during installation
will interfere with environmental protection if not
removed.Rosin Solvent may be helpful in removing residual
pieces from the gage and specimen.
Micro-Measurements

Adhesive Layer Uneven
Usually caused by uneven clamping pressure, uneven
"gluelines" have a varying color density when viewed
through the gage backing. May affect strain transmission,
thermal output, and magnitude of errors due to the location
of the grid above the specimen surface during bending.
Regaging is recommended when pronounced.
Adhesive flow
The adhesive must be present under the entire grid but
should not flow between the top surface of the gage and the
transfer tape and/or pressure pad. Check for adhesive on
solder tabs and terminals that will interfere with soldering.
Discoloration
Unusual darkening of backing or solder tabs may indicate
excessive curing temperatures. If soldering is difficult, clean
tabs with a soft (usually "pink") pencil eraser. (After
protecting the gage grid with drafting tape, stroke the tabs
away from the tape. Avoid side-to-side or back-and-forth
motions.) Check thermal output and zero-return after
electrical connections are made.
Gage Identification
Good practice requires that each gage be identified for
reference to the engineering data supplied with the gage.
Serialize at the gage site, or prepare a diagram showing
gage locations and identification.
AFTER LEADWIRE ATTACHMENT
The following checklist gives the recommended inspections
immediately after the completion of leadwire attachment, but
prior to the application of a protective coating.
Cold Solder Joints
Caused by failure of the entire solder joint to reach the proper
soldering temperature, leadwire movement during soldering,
or insufficient flux. Solder may not "wet" the leadwire and
solder tabs on the strain gage or terminal strip. Cold joints
are characterized by an uneven, "flaky" appearance and
poor solder flow.
Cold Solder Joint
Solder Peaks
Sharp peaks of solder are usually the result of insufficient flux
(use of solder without flux, loss of flux due to excessive
soldering temperature, or failure to remove rosin-core solder
wire and soldering iron tip from the solder joint
simultaneously). Peaks may interfere with flux removal and
environmental protection. Large solder masses of any shape
are undesirable in tests involving high acceleration or
deceleration.
Solder Peak
Solder bridges
Solder tabs are often closely spaced. Care must be taken to
ensure that solder joints or excessively long leadwire strands
do not form an electrical connection between tabs. A visual
inspection is usually adequate. However, an electrical
resistance check is recommended if any doubt exists.
Solder Bridge
Wiring Errors
Be certain that all leadwires are connected to their intended
circuit locations. Wire markers and color coding are helpful.
Electrical verification of connections is recommended for
installations with long runs or large bundles of leads.
Installed Resistance
A properly installed strain gage will usually retain a nominal
grid resistance within the tolerance shown on the
Engineering Data Sheet supplied with it. (Gages installed on
a small radius or with a heat-curing adhesive may exceed
these limits.) The Model 1300 Gage Installation Tester is an
ideal instrument for verifying the deviation of installed
resistance (sum of gage grid and leadwire resistances) from
the nominal value.
Residual Flux
Visually inspect for flux residue after cleaning with rosin
solvent. When cleaning solvents have evaporated, the
insulation resistance (leakage) between the gage grid and
specimen (if electrically conductive) should be at least
Micro-Measurements

10,000 megohms. Caution: Always use a megohm meter,
like that incorporated into the Model 1300, that applies a test
voltage of less than 100Vdc.
Residual flux
Residual Moisture
Condensation and other forms of moisture may also affect
insulation resistance. Dry the installation with warm air and
recheck resistance as previously described just prior to the
application of any environmental protection.
Zero Return
Connect the installation to a strain indicator and
zero-balance. If possible, load the specimen to produce a
strain of about the same magnitude expected in the test and
unload immediately. The reading should return to within 5
microstrain of zero. If loading is impractical, protect the gage
grid with drafting tape and gently press the grid with a soft
(usually pink) rubber eraser. Poor zero return may indicate
bubbles, inclusions, or unbonded areas in the adhesive layer
under the grid.
Stability
A strain gage subjected to a static strain over a period of time
should yield an indicated strain that is stable within a few
microstrain. Unexplained changes in indicated strain while
the test specimen is under a static load may also be due to
bubbles, inclusions, or unbonded areas.
AFTER ENVIRONMENTAL PROTECTION
Circuit Resistance
A repeat check of the resistance of the gage grid and
leadwires will ensure that no short circuit, open circuit, or
unintentional grounding has developed during installation of
the protection system.
Insulation Resistance
Any moisture or residual flux trapped under the
environmental protection may affect the insulation resistance
(leakage) between the gage grid and specimen (if electrically
conductive). Inappropriate or incorrectly installed coatings
may also produce similar undesirable effects. A megohm
meter should be used to ensure that the leakage resistance
is at least 10 000 megohms (and preferably 20 000
megohms or more). For this test, always use a megohm
meter, like that incorporated into the Model 1300 Gage
Installation Tester, that applies a test voltage of less than 100
Vdc to avoid damage to the installed gage.
If the insulation resistance is low, the protective coating
should be removed and the installation carefully cleaned and
dried. Check the resistance again before replacing the
coating. If the coating system requires a supplementary
insulation over unencapsulated grids and leadwire
connections, be sure it is in place. Some coatings contain
solvents which must evaporate before an acceptable
insulation resistance is obtained.
Visual Checks
Use STW-1 Tweezers or a DPR-1 Dental Probe to locate any
unbonded areas between the specimen surface and the
environmental protection. Make certain that any holes or
voids in the coating are not deep enough to compromise the
environmental protection system. Look for uncured areas
which may indicate incomplete mixing or curing. Remove
coating and apply again if any doubt exists regarding the
integrity of the protection system.
Unbonded Coating Coating Void
Leadwire Anchoring
Leadwires should be held firmly in place for those coatings
which harden or cure. Any motion during these processes
may produce "tunnels" between the coating and leadwire
through which moisture and contaminates can invade the
gage installation. (This is particularly true for M-Coat W-1
microcrystalline wax.) Visual inspection of the leadwire entry
into the protective coating will reveal the presence of any
separation between them.
Unanchored Cable
Anchored Cable
Micro-Measurements

Reinforcing Effects
The extent of reinforcement produced by the protective
coating depends upon the dimensions and mechanical
properties of the specimen, as well as the protection system
itself. While the stiffening effect is often negligible,
specimens having thin sections or those made of
low-modulus materials warrant special attention. The relative
magnitude of the effect on measurement data often can be
determined by applying a small preload to the specimen both
before and after installation of the environmental protection.
If preloads are not permitted, a similarly sized specimen of
the same material can be used to make an estimation of the
extent of reduction in indicated strain under similar loading
conditions.
Tunnel No Tunnel
Viscoelastic Effects
Most coatings behave in a viscoelastic manner under load
and produce reinforcing effects that, in time, approach zero.
Conversely, when long-standing loads are removed,
reinforcements that are opposite in sign but equal in
magnitude to the initial effects are produced. These, too, will
diminish with time. If either of these coating-related,
viscoelastic effects on the indicated strains are unacceptable
in a particular application, an alternate method of
environmental protection must be sought.
With the exception of encapsulated gages used in
immediate, short-term indoor tests, every strain gage
installation should be protected by an appropriate
environmental protection system. Properly selected and
applied, Micro-Measurements protective coating systems
will enhance the quality of your strain measurements by
protecting gage installations against chemical attack,
mechanical abuse, and electrical malfunctions.
Most gage installations will exhibit none of these faults if the
installation techniques outlined in Micro-Measure- ments
instructional materials and training programs are followed.
However, field applications made in harsh environments, or
even laboratory installations with difficult gage locations on
the test specimen, warrant special attention. If you have any
questions concerning your particular applications, our
Applications Engineering Department will be pleased to
assist you in establishing a program for verifying the integrity
of strain gage installations.
Installing Gages with Option P
Micro-Measurements

Gages supplied with Option P can be installed in a wide
variety of applications. General gage installation procedures
provided in Micro-Measurements technical literature and
training programs provide a sound foundation; however, the
need for leadwire attachment after gage bonding is
eliminated by Option P. Although application conditions may
dictate certain installation procedures, the following
guidelines are recommended for maximum performance.
They supplement any standard procedures when installing
gages with preattached leadwires.
GAGE HANDLING AND CABLE SUPPORT
The jumper wires, while sturdy, should never be used to
support the weight of the cable. After removing the gage and
its protective window from the pouch in the plastic box, hold
the gage/cable assembly by the cable bundle. Release a
comfortable working length of cable and place the gage on a
chemically clean surface. Temporarily secure the remainder
of the bundle before removing the gage from its protective
window with a pair of blunt-nosed tweezers.
Separate the jumper wires by approximately 1/8 in (3mm)
and apply a strip of PCT-2M gage installation tape to the
gage face on which the jumpers are soldered. This transfer
tape can be applied in the direction transverse to the
leadwires on shorter gages. If applied longitudinally over the
gage, jumper wires, and end of the cable, however, the tape
will provide additional strengthening during gage handling
and bonding.
The jumper wires can become permanently attached to the
specimen surface by adhesive flash during gage bonding. In
critical applications, this should be avoided by covering the
jumpers with a strip of TFE-1 Teflon film. Cut a strip wide
enough to extend about 1/8 in (3mm) on either side of the
jumpers. Carefully lift the tape from the surface at a shallow
angle. Using the mastic on the transfer tape to hold the film
in place, leave about 1/16 in (1.5mm) of space between the
edges of the tape and gage. With the bonding face of the
gage down, reposition the gage on the clean surface and
gently press the tape and film together around the jumpers
with a clean CSP-1 cotton swab.
When ready to bond the gage, carefully lift the tape from the
surface at a shallow angle. Unsecure the cable bundle and
hold the gage/cable/tape assembly by the bundle when
moving it to the installation site.
SURFACE PREPARATION
Surface preparation for bonding is essentially identical to that
described in Application Note B-129, Surface Preparation for
Strain Gage Bonding, for gages with preattached leadwires.
Because of the relatively long jumper wires, a larger-than-
normal clean area may be required for cable anchoring and
application of a protective coating.
Temporarily secure the cable bundle to the specimen, and,
using the transfer tape, position the gage for bonding.
GAGE BONDING
To preclude damage to the cable insulation, the adhesive
selected should be curable at a temperature of no more than
+180F (+80C).
Lift the gage end of the gage/tape assembly at a shallow
angle to the specimen surface (less than 45 degrees) until
the gage is no longer in contact with the surface. Continue
lifting the tape until it is free from the surface about 1/2 in
(13mm) beyond the gage. Tuck the loose end of the tape
under itself and apply the adhesive according to the
instructions provided. Whenever possible, lift the jumper
wires from the surface before curing the adhesive.
After the adhesive has cured, remove the transfer tape by
pulling it back directly over itself, peeling it slowly and
steadily off the surface. An application of RSK-1 Rosin
Solvent will quickly soften the mastic and ease tape removal.
Then, gently lift the cable and jumper wires and remove the
Teflon film.
Installing Gages with Option P
Micro-Measurements

PROTECTIVE COATINGS
Option P provides electrical insulation of the gage grids,
solder joints, and conductors on the gaged end of the cable.
Nevertheless, an additional protective coating is
recommended when the gage installation is exposed to
moisture, chemical attack, or potential for mechanical
damage. When applying coatings, pay particular attention to
the junction between the jumper and cable leadwires. The
vinyl insulation covering this connection does not provide a
waterproof barrier. Apply an appropriate coating in this area
before the cable is permanently anchored to the specimen.
STRAIN RELIEF
Before unsecuring the cable bundle, it is recommended that
strain relief loops be provided in both the cable and jumper
wires. Form these loops over the handle of a dental probe or
rod of similar diameter. The jumper-wire loop, which should
lie in the plane of the specimen, is usually held in place by
the protective coating. The cable loop, which should remain
upright, is usually located outside the coated area. 3145 RTV
silicone rubber can be used as a cable anchor.
After the cable has been properly anchored, the remainder of
the bundled cable can be unsecured and routed to the strain
gage instrumentation. If a cable extension is added,
remember that it should be attached by soldering. Alligator
clips, twist caps, and most other mechanical connections
should be avoided when making electrical connections within
strain gage circuits.
Gages supplied with Option P can be installed readily with
these techniques in most applications. Should you have any
questions about your particular application, our Applications
Engineering staff here at Micro-Measurements are always
ready to assist you.
Installing Gages with Option P2
Micro-Measurements

CEA-Series gages supplied with Option P2 can be installed
in a wide variety of applications. General gage installation
procedures provided in Micro-Measurements technical
literature and training programs provide a sound foundation;
however, the need for leadwire attachment after gage
bonding is eliminated by Option P2. Although application
conditions may dictate certain installation procedures, the
following guidelines are recommended for maximum
performance. They supplement any standard procedures
when installing gages with preattached leadwires.

SURFACE PREPARATION
Surface preparation for bonding is described in Application
Note B-129, Surface Preparation for Strain Gage Bonding,
for gages with preattached leadwires.
GAGE HANDLING AND CABLE SUPPORT
After removing the gage and its protective window from the
pouch in the plastic box, hold the gage/cable assembly by
the cable bundle. Release a comfortable working length of
cable and place the gage on a chemically clean surface.
Temporarily secure the remainder of the bundle before
removing the gage from its protectivle window with a pair of
BTW-1 Blunt-nosed Tweezers.
Apply a strip of PCT-2M Cellophane Tape to the gage face.
This transfer tape should be applied in the direction
transverse to the leadwires.
When ready to bond the gage, carefully lift the tape from the
surface at a shallow angle. Unsecure the cable bundle and
hold the gage/tape assembly by the bundle when moving it
to the installation site.
GAGE BONDING
To preclude damage to the cable insulation, the adhesive
selected should be curable at a temperature of no more than
+180F (+80C).
Lift the gage end of the gage/tape assembly at a shallow
angle to the specimen surface (less than 45 degrees) until
the gage is no longer in contact with the surface. Continue
lifting the tape until it is free from the surface about 1/2 in
(13mm) beyond the gage. Tuck the loose end of the tape
under itself and apply the adhesive according to the
instructions provided. Since Option P2 leadwires are
attached directly to the gage, a cushion may be used over
the grid area to promote uniform pressure over the entire
gage area. With M-Bond 200 , a folded gauze sponge can be
used as a cushion. However, the thumb by itself is sufficient
when placed transverse to the leadwires. When M-Bond
AE-10 / 15 is used, the silicone gum pad should be placed
over the gage grids and butted to the ends of the leadwire on
the solder tabs. M-Bond 600 and 610 should not be used on
gages with Option P2 because their elevated curing
temperatures are excessive for the vinyl cable insulation.
After the adhesive has cured, remove the transfer tape by
pulling it back directly over itself, peeling it slowly and
steadily off the surface. An application of M-LINE Rosin
Solvent will quickly soften the mastic and ease tape removal.
Installing Gages with Option P2
Micro-Measurements

PROTECTIVE COATINGS
Option P2 provides electrical insulation of the gage grids, but
not the solder joints. Therefore, an additional protective
coating is recommended when the gage installation is
exposed to moisture, chemical attack, or potential for
mechanical damage. When applying coatings, pay particular
attention to the junction between the solder tabs and cable
leadwires. Apply an appropriate coating in this area before
the cable is permanently anchored to the specimen.
STRAIN RELIEF
Before loosening the cable bundle, it is recommended that
strain relief loops be provided in the cable. Form these loops
over the handle of a M-LINE DPR-1 Dental Probe or rod of
similar diameter. The jumper-wire loop, which should lie in
the plane of the specimen, is usually held in place by the
protective coating. The cable loop, which should remain
upright, is usually located outside the coated area. M-LINE
3145 RTV silicone rubber can be used as a cable anchor.
After the cable has been properly anchored, the remainder of
the bundled cable can be loosened and routed to the strain
gage instrumentation. If a cable extension is added,
remember that it should be attached by soldering. Alligator
clips, twist caps, and most other mechanical connections
should be avoided when making electrical connections within
strain gage circuits.
Gages supplied with Option P2 can be installed readily with
these techniques in most applications. Should you have any
questions about your particular application, our Applications
Engineering staff is always ready to assist you. Don't hesitate
to give them a call.
Leadwire Selection
Micro-Measurements

Leadwires, properly selected and installed, should play a
passive, yet critical, role in strain measurement. Their
primary function is to deliver power to those portions of the
strain gage circuit remote from the measurement instrument
and to retrieve the strain-induced measurement signal from
the gage installations. Ideally, leadwires take nothing from
the measurement signal and add nothing to it.
While all leadwires do influence the signal to some extent in
actual practice, Vishay Micro-Measurements offers a
complete line of wires and cables which can be used to
minimize leadwire errors. Following is a brief outline of the
more important selection considerations:
Temperature Capabilities
Number of Conductors
Diameter of Conductor
Solid vs. Stranded Wire
Twisted/Shielded Cables
Adhesion to Environmental Protection
TEMPERATURE CAPABILITIES
Copper conductors are used extensively in strain gage
applications because they represent a good compromise
between electrical conductivity and cost. Their "use
temperature" range is influenced by both the electrical
insulation and conductor plating, if any, provided with them.
Multi-stranded wires with tin coatings and PVC insulation are
the most economical choice and are widely used in strain
gage applications. They are, however, restricted to an upper
temperature limit of about +180F (+80C). Because PVC
cracks at low temperatures, these leadwires should only be
used above -60F (-50C). Silver plating extends the use
temperature of multi-stranded wire. With Teflon coatings, the
range is -452F to +500F (-269C to +260C), and with
polyimide, -452F to +600F (-269C to +315C). Teflon is
often used for room-temperature applications when fire
codes specify that the insulation must not support
combustion.
Polyurethane enamels on unplated single conductor wires
are usable between -100F and +300F (-75C to +50C),
while polyimide coatings extend the limit from -452F to
+600F (-269C to +315C). Nickel plating and fiberglass
braid insulation on single conductor wires provide the widest
use temperature range, -452F to +900F (-269C to
+480C).
NUMBER OF CONDUCTORS
Vishay Micro-Measurements offers 1-, 3-, and 4-conductor
wires and cables. Single-conductor wires are used almost
exclusively for making intrabridge connections between
gages, or between gage and terminal strip. Three-conductor
cables are used for quarter-bridge circuits and half-bridge
applications. Four-wire cables, of course, are for connecting
full bridges to the strain gage instrumentation.
DIAMETER OF CONDUCTOR
From the standpoint of error reduction, the largest possible
wire diameter should be used to minimize leadwire
desensitization of gage factor and leadwire-related thermal
output. From a practical standpoint, however, smaller wires
are often necessary because of their influence on
measurements (reinforcement, mass effects, etc.), or
physical size of wire (ability to solder to gage tabs). As a
good rule of thumb, the longer the leadwire, the larger the
wire diameter should be.
SOLID VS. STRANDED WIRE
In general, stranded wire is more flexible and less
susceptible to fatigue-failure. Therefore, solid wires are
recommended for use primarily in making intrabridge
connections where they can be readily formed to the shape
of the test part and bonded to the surface to minimize fatigue.
An exception is nickel-plated wires, used for elevated
temperatures and available only as solid conductors.
TWISTED/SHIELDED CABLES
Special twisted and/or shielded cables are generally needed
when leadwires pass through electric and/or magnetic fields
which can superimpose electrical noise on the measurement
signal. A detailed discussion of these noise sources, and the
recommended cable selections to minimize their effects, is
given in Tech Note TN-501, Noise Control in Strain Gage
Measurements.
ADHESION TO ENVIRONMENTAL PROTECTION
A leadwire is the only portion of a gage installation which
should extend beyond the bounds of the protective coating.
As such, it always represents a possible avenue for moisture
invasion into the installation. A good moisture-proof bond
between leadwire coating and protective coating is essential
for long-term stability. To ensure the best possible bonds,
cables with PVC insulation should be precoated with thinned
M-Coat B before environmental protections are applied.
Teflon cable coatings should be etched with TEC-1
Tetra-Etch Compound to improve the bond.
Leadwire Selection
Micro-Measurements

The Vishay Micro-Measurements Strain Gage Accessories
databook, presents an extensive selection of wires, cables,
and accessories to handle a broad range of applications.
Each has been thoroughly field tested and found to give
excellent sensor performance when properly used in the
specified environments. If you have any questions about the
right leadwire for your specific application, our Applications
Engineering Department will be pleased to make specific
recommendations.
Mechanical Connectors in Strain Gage Circuits
Micro-Measurements


A question that often arises, especially for short-term strain
gage measurements, is: "Can I use mechanical connectors
(screw-type, etc.) to simplify connecting and disconnecting
strain gages in my instrument circuit?".
As for many other simple questions, there is no simple
answer to this one. In general, however, it is preferable to
use as few mechanical connections as possible in any strain
gage circuit. This is because mechanical connections are
potentially less stable in joint resistance than soldered
connections. When such an answer is given, the next
question is apt to be: "But what about my strain gage
instruments? All of them are equipped with binding posts or
other types of mechanical connectors." True enough. The
connectors are there for the convenience of the instrument
user, but these are carefully chosen for the intended
purpose, and are often gold-plated or gold-flashed to
minimize and stabilize the contact resistance.
It is always necessary to keep in mind the relatively small
resistance changes involved in strain gage measurements.
For a uniaxial working stress of, say, 30 000psi (207MPa) in
steel, the strain level is 1000 microstrain. With a 120-ohm
gage (G.F. ~ 2.0), the strain-induced resistance change is
only about 250 milliohms. Under these conditions a
5-milliohm change in contact resistance will cause an
indicated zero-shift of some 20 microstrain -- enough to
alarm a careful practitioner. Yet contact resistance changes
of this magnitude can be caused by an oxidized connector
contact surface or by a poorly terminated leadwire. Even
changing the insertion depth of the stripped end of the
leadwire into the connector can produce a measurable
zero-shift. With AWG #26 wire (0.4mm dia.), for instance,
changing the insertion depth by as little as 0.3 in (~8mm)
when reconnecting the wire will offset the zero by close to 5
microstrain.
The lesson to be learned from the above is that if mechanical
connectors are to be employed within the bridge circuit, a
number of precautions should be taken to help ensure
accurate strain measurements. One obvious measure is to
use high-quality connectors, gold-plated if possible. In
addition, however, leadwire terminations should be carefully
made. If the wire is stranded, the strands should be wound
snugly together, and then uniformly and smoothly tinned.
When connecting to a binding post, the wire should be
inserted to about the same depth each time, and the binding
post should be tightened for a firm but non-crushing grip on
the wire. After the strain indicator is turned on and balanced,
there should be no perceptible strain indication due to
shaking the free portion of the wire. With conventional
screw-down binding posts, the preferred leadwire
termination is a "spade" terminal, which should be attached
to the leadwire by soldering, not by crimping. The spade
terminal allows repeatable insertion depth into the connector,
and it is not subject to crushing if the binding post is over
tightened.
Even the foregoing steps do not guarantee complete
freedom from contact resistance variations. The quality of a
mechanical connection generally degrades with time; and
the rate of degradation depends upon the environment in
which the system operates. Thus, monitoring and
maintenance of the connection may be necessary to
preserve its initial performance. One way to monitor the
stability of the connection resistance is to use a highly stable
resistor (such as the Vishay Micro-Measurements S-Type) in
a verification channel. If the resistor is connected to identical
terminals and exposed to the same environment as the
active circuits, zero-shift in the verification channel may be
indicative of contact resistance variations in the connectors.
When contact resistance variations are identified, cleaning of
the contact areas may be necessary. Leadwire terminations,
if not plated, can be polished; and binding posts or other
gold-plated connectors can be washed with GC-6 isopropyl
alcohol (but never abraded).
Mechanical Connectors in Strain Gage Circuits
Select only, high-quality connectors -- gold-plated when
practical
Use solid, robust leadwire terminations
Firmly clamp leadwires in connectors, always at the same
point on the lead wire
"Exercise" connectors periodically by moving and re-
tightening wire clamps, and by several insertion/ removal
cycles of plug-and-socket connectors
Monitor and maintain the connections for resistance
stability and repeatability
In performing the previous step, be certain to distinguish
between contact resistance changes and thermoelectric
potentials
Pressure Pads for Bonding Gages
to Contoured Surfaces
Micro-Measurements

A number of Micro-Measurements adhesives, including the
popular M-Bond AE-10/15 epoxy-based system, require the
uniform application of pressure during curing. For flat
surfaces, silicone rubber pads, such as Vishay
Micro-Measurements SGP-1 or -2, can be used in
conjunction with a metal back-up plate and some vehicle for
load application (dead weight, spring clamps, etc.). With a
little more effort, the same approach can sometimes be used
on simple curved surfaces (cylinders, pipes, etc.) if the radius
of curvature is sufficiently large.
But when surfaces have steep or compound curvatures (like
the pipe elbow shown above), custom-shaped conformal
pads are usually required to distribute the applied load in a
uniform fashion.
PRESSURE PADS
Procedures
Form a back-up plate, slightly larger than the gage matrix, to
the approximate shape of the surface to be gaged.
Squeeze out RTV-3145 into a small metal or plastic cup. The
quantity should be sufficient to cover the gaging area to a
thickness of about 0.2 in (5mm). Add several drops of water
to the RTV silicone rubber and mix thoroughly. Apply the
mixture to the uncleaned surface (to make removal easier
after cure).
Press the backup plate in place, leaving a minimum
thickness of at least 0.1 in (2.5mm).
Cure will take about 24 hours. Remove the contour-pad/plate
assembly from the surface and trim any excess rubber from
the plate to complete the fabrication process.
Use the pad while following standard gage installation
procedures. These pads can withstand adhesive cure
schedules of at least 350F (175C).
Pressure Pads for Bonding Gages to Contoured Surfaces
Micro-Measurements

Water should NEVER be added to RTV silicone rubbers
used as protective coatings! In addition to the creation of an
excess number of voids during mixing, the quality of the bond
between substrate and coating is substantially reduced when
water is added. While that effect is desirable when
fabricating a conformal pressure pad, just the opposite is
needed in a protective coating.
Micro-Measurements
Remote Sense

Remote sense or, more correctly, remote sensing of
excitation voltage, is commonly recommended for use with
precision, commercial transducers to prevent leadwire
resistance changes (due to changes in either temperature or
length) from affecting transducer span, or sensitivity. This
feature is less frequently used with home-made
transducers, and is seldom used in strain gage test
applications.
TRANSDUCER APPLICATIONS
Leadwire attenuation presents a potentially significant error
source in transducers utilizing a Wheatstone bridge circuit.
The leadwires represent a parasitic resistance, and a portion
of the excitation voltage intended for the bridge circuit is
dropped in the leadwire system, reducing the voltage
actually present at the transducer, and effectively reducing
the transducer sensitivity. If the temperature remains
constant during the setup and measurement process, then
correctly applying shunt calibration, or a physical calibration
of the transducer, is all that is necessary to correct for the
effects of leadwire attenuation.
However, if the temperature of the leadwire system changes
during the measurement process, the resistance will change
correspondingly, changing the amount of attenuation, and
again resulting in a calibration error. The resulting error is
typically a fraction of one percent, but with commercial
transducers, which demand the highest possible accuracy,
or when attempting to obtain the best possible performance
from a home-made transducer, even that small error may not
be acceptable. In those instances, if the measuring
instrument is provided with remote sense capability, the error
can be effectively eliminated
For remote-sense operation, in addition to the pair of
leadwires supplying excitation voltage to the transducer,
another pair of wires would be connected between the
excitation terminals on the transducer and the remote sense
terminals at the instrumentation. The remote sensing circuit
in the instrumentation is designed to continuously monitor
the excitation voltage actually present at the transducer and,
depending upon the instrument design, to respond in one of
two ways if that voltage attempts to change.
Change the instrument sensitivity inversely with the voltage
monitored at the transducer. In other words, if the voltage at
the transducer were to drop due to an increase in leadwire
resistance, reducing the output signal from the transducer,
the instrument gain would effectively be increased by a
corresponding percentage; and if the voltage increased,
sensitivity would be reduced.
Increase or decrease the excitation supply voltage to main-
tain a constant voltage at the transducer, regardless of
changes in leadwire resistance. Micro-Measurements Model
3800 Wide-Range Strain Indicator is represenative of this
design.
STRESS ANALYSIS APPLICATIONS
Leadwire attenuation is directly dependent upon the ratio of
wire resistance to strain gage bridge resistance, and by
keeping the wire resistance to no more than a few percent of
the bridge resistance, the error will be on the same order of
magnitude. For example, for a three-wire quarter-bridge cir-
cuit, if the leadwire resistance is one percent of the gage re-
sistance, the uncorrected error due to attenuation will be
approximately one percent and can easily be eliminated by
proper application of shunt calibration. In a three-wire quar-
ter-bridge circuit, this is generally accomplished by applying
the shunt calibration resistor across the dummy arm of the
Wheatstone bridge, and adjusting the instrument gage factor
or gain control to display the proper synthesized strain value.
For a half- or full-bridge circuit, remote shunt calibration can
be employed, and the instrument adjusted as for the quar-
ter-bridge circuit.
As noted earlier, remote sense is generally not used in strain
gage measurement applications. The leadwire temperature
effects can greatly affect bridge balance (and that is why the
three-wire hook-up is strongly recommended for quar-
ter-bridge circuits), but the effect of leadwire temperature
change on sensitivity is generally very small, and normally
can be ignored in stress analysis applications.
In those cases where the leadwire resistance is a significant
percentage of the gage resistance, or if the wire temperature
will vary more than perhaps 200F (110C), remote sense
can be employed to advantage in half- or full-bridge circuits.
For a half bridge, the remote sense wires would be connec-
R
G
R
G
R
G
R
G
R
L
R
L
P+
S+
P
S
I
n
s
t
r
u
m
e
n
t
T
r
a
n
s
d
u
c
e
r
R
G
R
G
R
G
R
G
R
L
R
L
P+
S+
P
S
I
n
s
t
r
u
m
e
n
t
T
r
a
n
s
d
u
c
e
r
RS
RS
Remote Sense
Micro-Measurements

ted at the remote half bridge; and for a full bridge, the
connections would be made as for a transducer. Remote
sense is not practicable with a quarter-bridge circuit.
SPECIAL APPLICATIONS
While the subject of constant-current excitation is beyond the
scope of this article, it is worth noting that in some special
applications, constant-current (rather than constant-voltage
and remote sensing) can be used effectively to reduce or
eliminate temperature-dependent leadwire effects. It may
sometimes be used with commercial or home-made
transducers having long leadwires with appreciable
resistance; but it should be borne in mind that the transducer
must be designed for constant-current excitation. For
example, if a constant-voltage transducer is used with
constant-current excitation, it may defeat any built-in
compensation for changes in elastic modulus (of the
transducer) with temperature. And, constant-current
excitation can offer an advantage in high-temperature strain
measurement testing where copper wire (with its inherent
low resistance) cannot be used. In these applications, a
uniaxial gage would typically be employed with a four-wire
constant current hook-up, eliminating the Wheatstone bridge
circuit, and the need for remote sensing.
Micro-Measurements Model 2200 Signal Conditioning
Amplifier System offers both constant-voltage excitation with
remote sensing and constant-current excitation.
Quality Control of Strain Gage Installations
Micro-Measurements

An indispensable preliminary to the start-up of any
experimental stress analysis test is verification that all
components in the measuring system are functioning
properly to produce accurate, reliable data. In the case of
strain gage tests, the measurement system consists of three
principal elements: the gages installed on the test part, the
leadwires, and the strain indicating or recording
instrumentation. With no intention of underestimating the
importance of the latter two topics, this article will
concentrate on verification of the strain gage installation.
If not previously detected during system check-out, an
improperly installed or otherwise malfunctioning strain gage
will often (but not always) make itself evident when strain
measurements are being taken. But once the test has been
started (or, worse yet, completed), it may be very costly to
discover missing, inaccurate, or unreliable data. Fortunately,
there are a few simple tests that can be performed on
installed gages (preferably before the application of
protective coating) which will uncover most sources of
potential failure or poor operation. The tests are easily
applied, and should be routinely employed for every gage
installation, but are imperative when long-term accuracy and
stability are important.
QUALITATIVE TESTS
The gage installation should first be given a close visual
inspection. If any discolored or jagged solder joints are
observed, these should be re-soldered to achieve smooth
solder deposits. All traces of soldering flux must be
completely removed with an appropriate solvent such as
M-Line Rosin Solvent. The gage itself should have a uniform
color, since discolored areas visible through the backing may
indicate voids in the adhesive bond line. If voids exist, the
gage output is apt to be unstable when excitation is applied.
A further check on bond quality can be made with the gage
connected to a strain indicator. Pressing gently on the grid
with the eraser end of a lead pencil should produce a small
output indication which returns to zero with removal of the
pressure. (Note: in performing this test on open-face gages,
the pressure should be applied through a teflon strip to
protect the grid.) Erratic meter indication or failure of
zero-return suggests an incomplete bond, and the gage
should be replaced.
RESISTANCE TO GROUND TEST
When a strain gage is installed on a metal test part, one of
the most basic tests for the quality of the installation is the
resistance between the gage grid and the part. The
insulation resistance can be measured very easily, and with
no risk of damage to the gage, using the
Micro-Measurements Model 1300 Gage Installation Tester. If
the gage is properly installed and tested under laboratory
conditions, resistance to ground should be in excess of
20 000 megohms. A resistance of 10 000 megohms is
considered about the minimum for reliably accurate strain
measurement. Readings below this value generally indicate
some kind of quality defect in the gage installation. Low
readings may be caused, for example, by trapped foreign
matter, moisture, residual soldering flux, or backing damage
from soldering. They may also result from moisture migrating
to the installation after passing through pinholes in the
leadwire insulation. The Model 1300 can be used very
effectively to test the integrity of the leadwire insulation, too.
For this purpose, the insulated leadwire is placed in a
water-filled metal container, and the resistance between the
wire conductor(s) and the container measured. The reading
should again be at least 20 000 megohms.
INSTALLED GAGE RESISTANCE TEST
Among its other capabilities, the Model 1300 provides for
convenient, precise measurement of the gage resistance
after installation. The instrument indication is displayed as
the percent deviation from the nominal gage resistance, with
full-scale ranges of either 5% or 1%. Since Vishay Micro-
Measurements strain gages are manufactured to tightly
controlled tolerances on gage resistance, the magnitude of
the resistance shift due to installation provides a very
sensitive index to certain aspects of gage installation quality.
Shifts in gage resistance during installation should not
ordinarily exceed 0.5% when using room-temperature-curing
adhesives. With elevated-temperature-curing adhesives, on
the other hand, the resistance shift may be greater due to the
thermal stress in the gage after return to room temperature
from the adhesive curing temperature. The magnitude of the
shift depends on the curing temperature, the materials
involved, and other variables, but should never exceed 2%,
and should be uniform among identical gage installations to
within about 0.5%.
Excessive resistance shifts are usually indicative of one or
more flaws in the gage installation - poor installation
technique, questionable solder connections, damage to the
gage grid, leadwire damage, etc. A unique feature of the
Model 1300 is that the completely wired strain gage can be
connected to the instrument in a 3-wire circuit, just as it would
be to a strain indicator. Thus, when the Model 1300 is used
in its deviation mode, the effect of the normal resistance in
matched leadwires is automatically cancelled, and the
instrument will therefore detect resistive faults not only in the
gage, but also in the leadwires and solder joints.
After applying a protective coating to the gage, and prior to
placing it in service, retesting for the resistance to ground
and installed resistance is highly recommended. This will
ensure that the gage has not been damaged in the process
and that the presence of the coating has not significantly
lowered the leakage resistance. It should be noted, however,
that solvent-thinned coating compounds will exhibit lower
leakage resistance until the coating is fully cured and the
Quality Control of Strain Gage Installations
Micro-Measurements

solvents evaporated. If the resistance to ground remains low
after the coating is completely cured, the coating compound
may be contaminated or otherwise deficient in dielectric
properties.
It should be obvious that while passing all of the foregoing
tests is a necessary condition for satisfactory gage
performance, it is not in itself sufficient to guarantee that the
gage will perform properly under actual measurement
conditions. At the same time, any installation that falls short
of meeting the indicated standards should be viewed with
suspicion, and not relied on for critical applications or those
requiring accurate data.
Remote Shunt Calibration
Micro-Measurements

It is common knowledge - or at least it should be - that shunt
calibrating a remote bridge arm at the instrument binding
posts will introduce a calibration error several times greater
than the error anticipated from leadwire desensitization
considerations alone. What probably is not common
knowledge is that the indicated calibration magnitude will be
increased, not decreased, as expected from desensitization
considerations. The significant point is that if a remote bridge
arm is shunt calibrated at the instrument terminals and
attempts are made to mathematically or experimentally
adjust gage factor or instrument gain to correct the error, the
sensitivity will be adjusted in the wrong direction and
increase the error. Worse, perhaps, is that this is an "unsafe"
error, in that output from the transducer being used is larger
than the instrument indicates.
To illustrate the point, two examples are shown for strain
gages: a quarter-bridge and a full-bridge (which could be any
type of strain gaged transducer). In both cases, the gages (R
G) are 120 ohms and have a gage factor (K) of 2. The shunt
value (R SH) of 59 880 ohms is selected because it produces
an indicator reading of 1000 microinch/inch when used with
120 ohm gages, and there is no lead resistance. For each
example, the indicated calibration strain is listed for:
(a) calibration at the instrument terminals
(b) calibration at the bridge arm by use of dedicated
calibration leads
(c) calibration directly across the bridge arm
Note that for high shunt values and reasonable lead
resistances, (b) and (c) are equal. Further, for (c) - and (b), if
the values are approximately the same - the error can be
removed by adjusting the gage factor setting (i.e., gain) of the
instrument.
QUARTER BRIDGE
SHUNT POSITION INDICATED STRAIN
A 1051 microstrain
B 985 microstrain
C 985 microstrain
FULL BRIDGE
SHUNT POSITION INDICATED STRAIN
A 1034 micRostrain
B 969 microstrain
C 969 microstrain
There really is no legitimate reason for shunting a remote
bridge arm at the instrument terminals in quarter- and
half-bridge operation, although it may seem convenient for
indicating compressive strains. However, the practice is
common with external full-bridges or strain gage
transducers. The examples shown should serve to illustrate
how great the errors can be if remote calibration is not used.
Essentially all Micro-Measurements instruments provide
remote calibration capability to eliminate the associated
error. The leadwire resistances shown (R L) are not
exaggerated. Fifty feet of #26 AWG wire has a resistance of
about 2 ohms.
In summary, terminal shunting of external bridge arms can
cause significant errors that cannot be corrected simply by
adjusting the instrument gain to show the expected value. It
is a practice that should be avoided, and it is for this reason
that Strain Smart Data Systems do not provide shunt
calibration for an external full bridge.
RTV Silicone Rubber Coatings
Micro-Measurements

Many materials can be used to protect strain gage
installations. Perhaps none is more versatile for short-term
applications than room-temperature vulcanizing (RTV)
silicone rubber. The list of this material's capabilities is
indeed impressive:
Available as an easy-to-apply single-component coating
with uncured consistencies ranging from a low-viscosity
brush-on material for thin coats, to a medium viscosity
self-levelling form for use on level surfaces, to a
high-viscosity no-run paste for vertical and overhead
applications.
Cures at room temperature, yet is usable over a
temperature range of -75 to +550F (-60 to +290C).
Has a low modulus of elasticity that is ideal for thin or
flexible structures for which coating reinforcement effects
may become significant.
Provides the installation with good protection from
mechanical abuse. (Also works well for anchoring
leadwires to the surface of the specimen, for most
applications.)
Provides good short-term protection from water; resists
many chemicals; and can be used in radiation and vacuum
environments.
In order to cure, most single-component silicone-rubber
materials must absorb a small amount of moisture from the
surrounding atmosphere. This moisture reacts with the
material during polymerization to produce condensation
by-products, commonly acetic acid or methyl alcohol. Those
producing the acetic acid by-products smell like vinegar and
SHOULD NOT BE USED AS A PROTECTIVE COATING for
strain gages.
Acetic acid is a relatively weak etchant. However, trace
amounts trapped within a silicone rubber coating will
immediately attack both open-faced gages and copper
leadwires. (Any residual soldering flux will accelerate the
process.) This chemical corrosion of the gage grid, solder
joints, and leadwires will cause the resistance of the
installation to increase, resulting in a zero-drift with time.
Additionally, the fatigue life of the gage may be affected.
To ensure these problems are not encountered,
Micro-Measurements offers three qualified RTV silicone
rubber coatings which produce a safer methyl alcohol
by-product. M-Coat C is a naptha-thinned brush-on material
for thin, conformal coatings; 3140 RTV is a translucent,
self-levelling coating; and 3145 RTV is an opaque, thick
paste. These coatings will protect gages for several days in
high humidity or water-splash environments. Even with these
qualified materials, however, zero-drift will occur over time at
high relative humidities - (typically in 8 to 10 days with 3140
RTV and 3145 RTV, and somewhat sooner for the thinner
M-Coat C). When this happens, the resistance of the
installation may initially decrease (due to internal shunting
between gage grids or tabs). Within a few more days,
galvanic corrosion effects can cause a net increase in
resistance
Because these silicone rubber coatings require no mixing
and can be brushed (M-Coat C), poured (3140 RTV), or
spread (3145 RTV) on the surface, they are easily applied.
For maximizing their protective qualities, the specimen
surface to which the coatings are applied must be chemically
clean.
Quality of the bond often can be improved by applying an
appropriate primer to chemically activate the specimen
surface. (Micro-Measurements offers RTV Primer No. 1,
which can be used on a wide variety of materials.) However,
if the surface is properly cleaned for gage bonding (and is not
re-contaminated during gage installation), the application of
a primer is often unnecessary.
Special care should be taken when applying silicone rubber
coatings to composite materials. The low surface energy of
many polymeric matrix materials, migration of plasticizers to
the surface, and incomplete evaporation of surface-cleaning
solvents can adversely affect the coating quality.
Easy to apply. Effective as protection against mechanical
abuse and chemical attack. The versatile RTV silicone
rubber coatings may be the answer to your short-term gage
coating requirements. If you haven't used them before or
need more details, don't hesitate to call our Applications
Engineering Department concerning your specific
application.
Surface Preparation of Composites
Micro-Measurements

Micro-Measurements Application Note B-129, Surface
Preparation for Strain Gage Bonding, provides the general
procedures for properly preparing surfaces of test specimens
for strain gage bonding. Many of the methods described
there are also appropriate for composite materials.
Following are some supplementary techniques for producing
a chemically clean, dull matte finish needed for making good
strain gage installations on these materials.
SOLVENT CLEANING
The number of resin systems, fibers, and fillers used in
composites is ever increasing. No single cleaning solvent is
ideal for them all. However, isopropyl alcohol (GC-6) is
satisfactory for many resin systems. If in doubt about
compatibility with a solvent, consult the resin system
manufacturer. And don't forget to verify compatibility with any
exposed fibers and fillers.
Composites fabricated with silicone releasing agents, or
structures coming into contact with silicone greases or oils
prior to gaging, require special cleaning. Strain gages cannot
be successfully bonded to surfaces contaminated with these
materials. Light concentrations can often be removed by
several washing/drying cycles performed with M-Prep
Conditioner A warmed to approximately +95 deg F (+35 deg
C). Procedures recommended by the silicone manufacturer
must be used on heavier contaminations, or when
water-based cleaners are disallowed.
SURFACE ABRASION
Matrix-rich surfaces can usually be abraded with 320-grit
silicon carbide paper (SCP-2) to produce a satisfactory matte
finish. However, unless their surfaces have been machined
flat, test specimens of matrix-poor composites or those with
textured surfaces require alternate techniques. For these,
the grit on the silicon carbide paper will not reach into the
"valleys" to produce the desired finish. For small areas (a
square inch or so), try a fiberglass brush as shown here.
For larger areas or a large number of smaller areas, an
air-abrasive unit may be used to advantage.
Care must be taken, however, to ensure that neither the air
nor the abrasive contaminate the surface. Accordingly, either
bottled dry air or shop air highly filtered at the unit inlet is
recommended. Clean alumina in the 27-50 micron size is the
ideal abrasive. Prior to abrasion, mask the gage area with
drafting tape (PDT-1). Handling the air nozzle as if spray
painting, start and finish each pass on the drafting tape to
avoid excess abrasion. Dust off loose particles (with a brush
if necessary) and wipe the area with surface cleaning solvent
and gauze sponges (GSP-1). Repeat as necessary until the
surface has a satisfactory matte finish. Dry the surface with
a hair dryer for a few seconds prior to gage installation to
ensure complete solvent removal.
CONDITIONING AND NEUTRALIZING
A final cleaning with Micro-Measurements M-Prep
Conditioner A and Neutralizer 5A is recommended following
solvent cleaning and abrasion, provided water-based
surface preparations are permitted. To minimize absorption,
however, these materials should be applied and removed as
quickly as possible. Further, in most cases, the surface
should be warm-air dried following their application.
The surface preparation of composite materials for strain
gage installation is relatively straightforward, provided these
special requirements are taken into account. Our
Applications Engineering Department will be pleased to
provide you with additional details about these techniques
and installation accessories.
Three Leadwire Attachment
Micro-Measurements

Application Note TT-612, The Three-Wire Quarter-Bridge
Circuit, describes the advantages of the three-wire quarter-
bridge circuit, and explains why it is generally recommended
for static strain applications. Because the three-wire circuit
has become the standard in contemporary strain gage
technology, gage users need to be familiar with techniques
for connecting three leadwires to the solder tabs of a gage.
Briefly described here are procedures for attaching leadwires
to several widely used types of strain gages.
CEA-SERIES GAGES
This popular line of general-purpose gages features a fully
encapsulated grid, and large, copper-coated solder tabs.
Because of the special tab construction, the leads from a
three-conductor cable such as Micro-Measurements
326-DFV can be attached directly to the solder tabs as
shown below in Diagram (1). Following is the lead
attachment procedure, assuming that the gage is already
bonded (or welded) in place.
Diagram (1)
1. Strip about an inch (25mm) of the vinyl insulation from the
end of the cable with a thermal wire stripper.
2. The cable conductors are distinguished by their
insulation colors - red, white, and black. Separate the
strands belonging to the red conductor from the others,
and twist the strands tightly together.
3. Twist the combined strands from the white and black
conductors tightly together to form a single leadwire.
4. Tin both resulting leadwires (the red, and the
white/black).
5. Dress the leads to project axially from the cable end, and
trim off all but the last 3/32 inch or so (~2mm) of the bare
wire.
6. Tin the solder tabs of the gage.
7. After forming a small hump in the cable, near the end,
tape the cable in place so that the tinned leads press
down directly on the solder tabs.
8. Solder the leads to the tabs to complete the attachment.
9. Clean the gage installation with Rosin Solvent, and apply
an appropriate protective coating.
These and other detailed instructions for making soldered
connections to strain gages are given in Micro-
Measurements Application Note TT-609, Strain Gage
Soldering Techniques.
GAGE INSTALLATIONS WITH BONDABLE TERMINALS
The usual practice for standard strain gages (not equipped
with copper-coated solder tabs) is to install a set of bondable
solder terminals adjacent to the tab end of the gage, as
shown in diagrams (2) to (5). These terminals are commonly
bonded in place simultaneously with the gage. The ends of
the cable conductors are then soldered to the terminals; and
short, single-strand strain-relief jumpers complete the
connections to the gage solder tabs.
Attachment of the cable conductors to the terminals is done
with essentially the same procedure as for the CEA-Series
gage described previously. In fact, when only two terminals
are used, as in (2), the procedure is identical.
Diagram (2)

Tutorial Galgas Montaje

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tutorial Galgas Montaje

Uploaded by

Copyright:

Available Formats

Tech Note TN-501-2

, for instance) that have been

Allegheny Ludlum Steel Co.

-based computer program

50/50 Tin-Indium solder should be used. This

Registered Trademark of Indium Company of America.

= the error as a percentage of the actual

Equations (6a) and (7a) can be rewritten:

As an alternative to the preceding methods, a quick graphical

Following is a numerical example utilizing rst Equations

Correction of Principal Strains

0.75 1000 0.335 1 0.285

= Strain along axis of intended measurement

vanishes. In order to better illustrate the order of

and substitute the value of n; including

in terms of principal strains

Tape to restrict the

insulation are not recommended for high g-elds

Mylar and Teon are Registered Trademarks of Dupont.

lm are commonly used surface insulators. To

-insulated leadwires require etching with

Registered trademark of DuPont.

) tape at the gage location to minimize

-insulated cable is used, etch the insulation

You might also like