British Standard

A single copy of this British Standard is licensed to

Tom Magee
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

06 March 2004

This is an uncontrolled copy. Ensure use of the most

current version of this document by searching British
Standards Online at bsonline.techindex.co.uk
 BRITISH STANDARD BS 3923-2:
1972

Methods for

Ultrasonic
Examination
of Welds
Part 2: Automatic examination of fusion
welded butt joints in ferritic steels
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

UDC 621.791.05:620.179.16 + 669.15194.57:621.791.052.4:621.791.5/8:620.179.16

 BS 3923-2:1972

Co-operating Organizations

The Welding Industry Standards Committee, under whose supervision this

British Standard was prepared, consists of representatives from the following
Government departments and scientific and industrial organizations:

Aluminium Federation Institution of Civil Engineers

Associated Offices Technical
Committee*
Association of Consulting
Engineers
British Constructional Steelwork
Association
British Electrical and Allied
Manufacturers Association*
British Railways Board
British Steel Industry*
Crown Agents for Oversea
Governments and
Administrations
Department of Employment
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
Institution of Electrical Engineers
Institution of Mechanical
Engineers
Institution of Production Engineers
Institution of Structural Engineers
Lloyds Register of Shipping*
London Transport Executive
Ministry of Defence, Combined
Ministry of Defence, Navy
Department*
Shipbuilders and Repairers
National Association*
Society of British Aerospace
Companies Limited*

Department of Trade and Industry Welding Institute*

Department of Trade and Industry,
National Engineering Laboratory

The Government department and scientific and industrial organizations

marked with an asterisk in the above list, together with the following, were
directly represented on the committee entrusted with the preparation of this
British Standard:

British National Committee for Non-destructive Testing Society of

This British Standard, having
been approved by the Welding
Industry Standards Committee,
was published under the authority
of the Executive Board on
22nd August, 1972
Non-destructive Testing Great Britain
Electricity Council, the Central Process Plant Association
Electricity Generating Board Society of Motor Manufacturers and
and the Area Boards in Traders
England and Wales Society of Non-destructive
Engineering Equipment Users Examination
Association United Kingdom Atomic Energy
Institute of Physics and the Physical Authority
Society Water-tube Boilermakers
Ministry of Defence, Army Association
National Coal Board Individual firm

BSI 01-1999 Amendments issued since publication

First published August 1965 Amd. No. Date of issue Comments
First revision August, 1972

The following BSI references

relate to the work on this
standard:
Committee reference WEE/34
Draft for comment 70/26486

ISBN 0 580 07436 6

 BS 3923-2:1972

Contents

Page
Co-operating organizations Inside front cover
Foreword ii
1 Scope 1
2 Definitions 1
3 Operators 1
4 Equipment 1
5 Surface condition 1
6 Parent metal examination 1
7 Weld examination 3
8 Evaluation of imperfections 4
9 Test plates 4
10 Presentation of results 4
Appendix A Guidance on the determination of probe characteristics 5
Appendix B Notes for guidance on the D.G.S. diagram 5
Appendix C Method for setting sensitivities where maximum
sensitivity is required 12
Figure 1 Relationship between ratio of diameter to
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

thickness and angle of ultrasonic beam 2

Figure 2 Typical D.G.S. diagram for normal probes 8
Figure 3 Typical D.G.S. diagram for angle probe
of 2 MHz, 20 mm 22 mm crystal size 9
Figure 4 Typical D.G.S. diagram for angle probe
of 4 MHz, 8 mm 9 mm crystal size 10
Figure 5 Measurement of shear wave attenuation 12
Figure 6 Measurement of transfer loss 12
Publications referred to Inside back cover

BSI 01-1999 i
 BS 3923-2:1972

Foreword

This British Standard forms part of a series dealing with methods for the
non-destructive testing of welds. It does not state when this particular type of
testing should be employed not does it give standards of acceptance, as both of
these aspects should be covered in the appropriate application standard or be
agreed between the contracting parties.
This standard was first published in 1965 as a result of ultrasonic examination
procedures being agreed between organizations interested in the manufacture
and use of boilers and pressure vessels. Part 2 has now been revised to cover
automatic examination of butt joints in a wide range of shapes and forms of
ferritic steels.
The use of a D.G.S. (Distance, Gain, Size) diagram for obtaining the appropriate
sensitivity setting only has been introduced although it is not the sole method.
Guidance on this method is given in Appendix B.
The revised standard is not associated with any particular type of fabrication, but
has been prepared to cover a wide range of products and as such lays down the
broad principles of automatic ultrasonic examination. It is emphasized that a
satisfactory technique can only be determined after taking into account all the
relevant factors regarding the equipment to be used and the characteristics of the
weld to be examined.
Prior to ultrasonic examination the weld should be visually inspected and any
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

visible flaws recorded. Automatic ultrasonic examination is usually

supplemented by manual examination and both are often used in conjunction
with other testing methods in order to examine completely a welded article or
structure. The use of any non-destructive testing method should always be
considered in relationship to inspection and testing as a whole and the full
benefits of this or any other method often can only be obtained by considering the
results in conjunction with those from other methods.
A British Standard does not purport to include all the necessary provisions of a
contract. Users of British Standards are responsible for their correct application.
Compliance with a British Standard does not of itself confer immunity
from legal obligations.

Summary of pages
This document comprises a front cover, an inside front cover, pages i and ii,
pages 1 to 14, an inside back cover and a back cover.
This standard has been updated (see copyright date) and may have had
amendments incorporated. This will be indicated in the amendment table on
the inside front cover.

ii BSI 01-1999

1 Scope
This Part of this British Standard deals with
methods for the automatic scanning and recording
of imperfections in fusion welded butt joints in
ferritic steels1).
Whereas the procedures apply to all thicknesses of
material, the limiting parameters are:
1) whether the material traversed by the
ultrasonic beam is such that the beam can be
propagated through it sufficiently to achieve the
agreed sensitivity;
2) that with the angles of probe available, the
ultrasonic beam will reflect from the under
surface of the material in such a way as to cover
the full cross section of the weld. For curved parts
the ability of the beam to reflect from the under
surface depends on the ratio of the outside
diameter to the thickness (see Figure 1).
NOTE The titles of the British Standards referred to in this
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
standard are listed on the inside back cover.
2 Definitions
For the purposes of this British Standard the
definitions given in BS 3683-4 apply.
3 Operators
If required, the operator shall demonstrate to the
satisfaction of the contracting parties that he is able
to meet the requirements of any particular
technique adopted.
4 Equipment
4.1 Presentation. The examination shall be
carried out by an agreed automatic or
semi-automatic scan, using a pulse echo technique
with either A or B-scope presentation, and a fully
automatic recording or marking system shall be
used.
4.2 Test frequency. The equipment shall be
capable of working at a test frequency within the
range 1 MHz to 6 MHz.
4.3 Probes2)
4.3.1 The area of each transmitting and/or receiving
crystal shall not exceed 500 mm2. No dimension of
the crystal face shall exceed 25 mm unless
otherwise agreed.
4.3.2 For parent metal examination, zero angle
longitudinal wave probes shall be used.
1)
For the examination of welds used in the manufacture of pipes and tubes on a continuous basis see BS 3889-1A.
2)
Guidance on the determination of probe characteristics is given in Appendix A.
BSI 01-1999
BS 3923-2:1972
4.3.3 For angle probes, the ultrasonic beam shall
have an angle of refraction of between 35 and 70
within the material.
4.4 Coupling medium. The coupling shall be
obtained by either contact, gap or immersion
scanning using a liquid or paste medium suitable for
this purpose.
4.5 Continuity of coupling. There shall be an
indication of adequate acoustic coupling for
example, by recording a compressional wave
transmitted through the coupling or by a device
which indicates any failure of the coupling.
4.6 Recording. The recording or marking system
shall clearly indicate imperfections which require
further investigation by manual scanning.
Any recorded imperfection shall be clearly located
from a positional datum so that it can be accurately
positioned along the weld. It is recommended that
these positional data are located along welds at
intervals of 250 mm.
5 Surface condition
The condition of the surface that will be in contact
with the probe shall be such that a satisfactory
coupling between the probe and the workpiece can
be maintained. Spatter and loose scale shall be
removed.
The surface condition of the weld and adjacent
parent metal on each surface shall be such that it
does not adversely influence the weld examination.
NOTE Depending on the profile and surface condition, dressing
of the weld area may be necessary, even when contact is only to
be made with the parent metal.
6 Parent metal examination
6.1 General. Whether or not the parent metal has
been ultrasonically tested previously, manual
ultrasonic examination shall be made after welding:
1) to locate any flaws, such as laminations, in the
material through which the ultrasonic beam will
pass during examination of the weld,
and
2) to establish the material thickness to enable
the actual beam paths to be determined.
If any flaws are found, their influence on the
inspection of the weld shall be considered and if
possible alternative techniques of scanning shall be
used to ensure complete examination of the weld. If
any section cannot be so tested, this fact shall be
included in the report (see 6.6).
1
 BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

Figure 1 Relationship between ratio of diameter to thickness

and angle of ultrasonic beam
During the preliminary examination the operator
shall assess the attenuation characteristics of the
material and the influence that the surface
condition will have on the coupling, in order to
determine the practicability of performing an
effective test (see also 6.5).
6.2 Method. The material shall be examined by the
pulse echo technique using a longitudinal wave
probe normal to the surface of the material.
(See also BS 4336-1A.)
The parent metal shall be ultrasonically examined
over the region covering the entire scanning zone for
the subsequent weld examination.
6.3 Test frequency. The frequency shall be
between one and two times the frequency to be used
for the weld examination so that the attenuation
characteristics assessed under 6.1 are related to the
subsequent shear wave test.
NOTE On thin material it may be desirable to use a double
crystal probe.
6.4 Time base calibration. The time base shall be
calibrated using either the A2 or A3 block as
specified in BS 2704.
6.5 Sensitivity. The method of setting the
sensitivity shall be agreed between the contracting
parties.
Methods of achieving the gain settings are described
in Appendices B and C.
The sensitivity of the flaw detector/probe
combination shall be such that a clear signal will be
obtained from the smallest defect to be detected
throughout the scanning distance.
The size, type and orientation of the smallest flaw to
be detected shall be agreed between the contracting
parties before testing is commenced.

2 BSI 01-1999

6.6 Reporting. A report on the parent metal
examination shall be made describing any local
heterogeneities (e.g. laminations, surface flaws,
areas of high attenuation) which will confuse the
interpretation of the weld examination at such
locations. If no such heterogeneities are found, a
statement that the parent metal examination was
satisfactorily undertaken shall be made.
7 Weld examination
7.1 Method. The examination shall be carried out
with angle probes by an agreed automatic or
semiautomatic scanning system, using the pulse
echo technique with either A or B-scope
presentation. A fully automatic recording or
marking system shall be used.
For imperfections predominantly parallel to the
direction of welding the weld shall be scanned at
right angles to the weld axis.
For imperfections predominantly transverse to the
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
direction of welding the weld shall be scanned along
the weld axis.
NOTE The automatic system may not facilitate the satisfactory
detection of transverse imperfections. In such cases it will be
necessary for an alternative automatic or manual method to be
agreed between the contracting parties before testing is
commenced.
7.2 Test frequency. The test frequency used shall
be compatible with the requirements for sensitivity
in 7.5.
NOTE For guidance, the following frequencies are suitable for
the beam path lengths quoted:
4 MHz to 5 MHz for up to 200 mm
2 MHz to 2.5 MHz for up to 400 mm
1 MHz to 1.25 MHz for over 300 mm
7.3 Scanning. Wherever possible the scanning of
the weld shall be carried out from both sides of the
weld on the same surface. Where the configuration
is such that examination from both sides is not
possible, this fact shall be included in the report
(see Clause 10).
The full cross section of the weld shall be scanned at
the agreed sensitivity. This may be achieved, for
example, by the use of
(1) a variable angle probe;
or (2) a to and fro movement of a fixed angle
probe;
or (3) some special configuration of probes.
The distance between successive scans of the weld
shall not exceed the breadth of the transmitting
crystal.
The speed of movement of the probe(s) shall be such
that the optimum response of the recording
apparatus is obtained.
BSI 01-1999
BS 3923-2:1972
7.4 Equipment calibration
7.4.1 Calibration of time base. For equipment using
A-scope presentation, the time base shall be
calibrated in accordance with the recommendations
of BS 2704 using a type A2 or A3 test block.
Where B-scope presentation is used, direct time
base calibration may not be possible. In this case a
special test block of the same thickness and
curvature as the seam to be examined, with four
slots to represent the corners of the weld, shall be
used to ensure that complete coverage of the weld
area is obtained.
7.4.2 Position of probe index. This shall be
determined in accordance with the
recommendations of BS 2704. If a variable angle
probe is used, the upper and lower limits of the
probe index shall be determined in a similar manner
to that for a fixed angle probe.
7.4.3 Angle of refraction. For a fixed angle probe the
angle of refraction shall be measured in accordance
with the recommendations of BS 2704.
If a variable angle probe is used, the upper and
lower limits of the range of angle shall be
determined in a similar manner to that for a fixed
angle probe.
7.5 Sensitivity. The method of setting the
sensitivity shall be agreed between the contracting
parties.
Methods of achieving the gain settings are described
in Appendices B and C.
The sensitivity of the flaw detector/probe
combination shall be such that a clear signal will be
obtained from the smallest defect to be detected
throughout the scanning distance.
The size, type and orientation of the smallest flaw to
be detected shall be agreed between the contracting
parties before testing is commenced.
With variable angle probes a marked loss in
sensitivity may occur at the higher probe angles and
therefore the sensitivity shall be determined using
that angle which gives the weakest echo.
7.6 Recording. If the recording system is of
the GO/NO-GO type, the echo height shall be
measured at which
1) the recorded signal appears,
and
2) the recorded signal disappears,
and there shall not be a difference of more
than 2 dB.
With a proportional type recording system, the
recording indication shall be compared over its full
range, with the echo height on the flaw detector.
3
 BS 3923-2:1972
8 Evaluation of imperfections
If an imperfection is revealed by the examination
described in Clause 7, a more detailed ultrasonic
examination shall be carried out using an agreed
manual technique which is suitable for the purpose,
such as one of those given in BS 3923-1.
9 Test plates
Due to the restricted width normally required,
welded coupon test plates are not suitable for
automatic scanning, and their examination shall be
carried out by an agreed manual technique.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
4 BSI 01-1999
10 Presentation of results
In the report of the automatic ultrasonic test the
following information shall be included:
1) Job identification reference (including contract
number and part number).
2) Surface condition (as welded, chipped, ground
flush).
3) Date of test.
4) Identity of operator.
5) Instrument used (including serial number).
6) Probes used, including frequency.
7) Scanning techniques.
8) Sensitivity of test. (This shall be recorded at
the end of the test and related instrument
settings shall be compared against and noted
from a suitable test block.)
9) Results obtained.
10) Any other relevant information relating to the
test.

Appendix A Guidance on the
determination of probe
characteristics
A.1 Method of measurement of frequency. The
usual method of measuring-probe frequency is to
take the first back wall echo from a steel sample
such as the A2 block of BS 2704 and to take this
high frequency pulse from the end of the amplifier
before rectification and present it on the calibrated
time base of an oscilloscope. The number of full wave
lengths in a one microsecond time interval gives the
probe frequency.
Alternatively, the ultrasonic operator, without any
electronic knowledge, should be quite capable of
measuring the probe frequency provided that the
flaw detector has a fast time base expansion facility,
such as 10 mm steel full screen, and sufficient
resolution on each echo to indicate the high
frequency pulses. In this case, the frequency is
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
derived from the formula:
velocity
Frequency = --------------------------------
wavelength
For boiler plate the velocity in compression is
usually in the order of 5.93 103 m/s and the
velocity in shear is in the order of 3.23 103 m/s.
The velocity is known for the material and the
wavelength is read from the trace as the distance
covered by two consecutive half cycles for a half
wave rectified display. For flaw detectors which
show a full wave rectified display, it would be the
distance covered by four half cycles. From this it
would appear that two wavelengths were being
measured but in fact normal time-base calibration
for flaw detection purposes is always half of the total
distance travelled; the half calibration therefore
cancels out the double wavelength. The half cycles
should be regularly spaced, uneven spacing
indicating a distorted pulse, and probes exhibiting
this feature should not be used with the D.G.S.
system.
BSI 01-1999
BS 3923-2:1972
A.2 Method of measurement of near field
A.2.1 Method 1 (if an immersion tank is available).
Plot the distance/amplitude curve of a small
reflector along the axis of the beam. The near field
length, N, is shown by the maximum amplitude
position before the curve falls away into the far field.
A useful target for this operation is a 2 mm or 3 mm
diameter ball, mounted on the end of a rod whose
diameter is less than the ball at the mounting point.
The near field length for steel can be calculated from
the empirical determination of N for water since the
ratio of the two near fields is inversely proportional
to the ratio of the velocities, i.e.:
N ( steel ) Velocity ( water )
-------------------------- = -------------------------------------------
N ( water ) Velocity ( steel )
A.2.2 Method 2 (if an immersion tank is not
available). In this case, the distance/amplitude
curve of the back echo is plotted from flat parallel
samples of steel of similar velocity whose width
would not restrict the beam. The curve should be
plotted on logarithmic paper and after
approximately three near field lengths, it should
form a straight line. If not, attenuation has to be
allowed for (see B.1.1), and when taken into account
for the back wall echo a straight line should result
in the far field past three near field lengths. If the
straight line is then projected up to the zero decibel
N
line, the intersection point will be -------- or 1.57 N,
2
where N is the near field length.
Appendix B Notes for guidance on the
D.G.S. diagram
B.1 Normal probes. (See Figure 2.)
B.1.1 General. The D.G.S. diagram for single crystal
normal probes was drawn by plotting amplitude in
decibels from a series of disc shaped reflectors with
increasing distance from the probe in water. The
loss due to water attenuation was allowed for in
each case and therefore the graph shows the
reflection conditions for any material assuming no
attenuation.
The distance is given in near fields and is on a log
base. If the near field length N is not known for the
probe in use it may be calculated from the formula
2
D
N = -----------------------------------------
4 wavelength
where D is the crystal diameter.
5
 BS 3923-2:1972
The backwall echo line indicates the maximum
reflection from a large reflector with increasing
distance from the probe, which becomes a straight
line after three near field lengths. This confirms the
radiation law for large reflectors in the far field, i.e.
the amplitude is inversely proportional to the
distance, so that if the distance is doubled the
amplitude is halved: a 6 dB reduction. This law
provides a simple method of measuring attenuation.
If a difference of more than 6 dB is measured, then
this will be due to attenuation in the material,
leaving a simple calculation for the attenuation
factor.
Example 1
Plate thickness 30 mm
Probe frequency 4 MHz
Probe diameter 10 mm
Near field length 17 mm
The second back echo at 60 mm distance is greater
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
than 3N (51 mm), therefore, the second and fourth
echoes are on the 6 dB slope.
Suppose the difference in echo amplitude measured
between second and fourth echo = 10 dB. Total beam
path between second and fourth echo = 120 mm.
Therefore, the attenuation of the plate material
at 4 MHz is
10 6 1 echo set to full screen height.
---------------- = ------ dB/mm or 33 dB/m
120 30
Alternatively, back echoes in the near field may be
used. Any difference between their amplitude in
excess of what is shown in Figure 2 indicates the
amount of attenuation over the total distance
travelled between the two echoes.
NOTE The attenuation factor in the example 1/30 dB/mm
means that the sound is attenuated by 1/30 dB for every
millimetre of sound path travelled, bearing in mind that the
sound in a reflection technique has to travel there and back.
Since the flaw detector calibration is always for the
distance in front of the probe and not the total
distance travelled, it may be more convenient to
express the factor in the same way, i.e. 1/15 dB/mm
or 66 dB/m. It is important to indicate in which form
the factor is being used.
The lines below the back echo line show the
amplitude of reflectors which are smaller than the
beam width. These values are shown in relation to
the crystal diameter and, for small reflectors within
the near field, the nodes and anti-nodes of
sensitivity are clearly indicated. They do not arise
with a large reflector since all the incident energy is
totally reflected.
Well into the far field all these lines indicate that
reflections from small reflectors follow the inverse
square law, so that if the distance is doubled then
the amplitude is reduced by a factor of 4 (12 dB).
6 BSI 01-1999
The term small reflectors is related to the beam
width and not to the crystal diameter, since
reflectors greater than the probe diameter behave
as small reflectors when far enough away to be
encompassed within the spreading beam width.
This is shown in Figure 2 by the lines with S values
of 1 and 2 times the crystal diameter.
B.1.2 Setting the sensitivity. In order to use Figure 2
to determine the sensitivity setting, the operator
first needs to know the smallest disc shaped flaw or
equivalent reflector size that has to be detected and,
secondly, at what amplitude he wishes to record it.
The following plate testing example illustrates the
method:
Example 2
Plate thickness 100 mm
Probe frequency 4 MHz
Probe diameter 25 mm
Near field length 104 mm
Smallest flaw to be detected equivalent to
flaw 5 mm diameter (0.2S). Defect echo height 2
screen divisions.
The first back echo will be at a distance of
approximately 1 near field. It can be seen from
Figure 2 that close to the surface an 0.2S flaw will
give an echo approximately 26 dB below the back
Near the bottom of the specimen such a flaw will
only be 17 dB below but at the worst position, 0.5N
distance (52 mm), it will be 29 dB below.
In order to find the flaw at 2 screen divisions at the
worst point (29 dB below), the back echo has first to
be set to 2 screen divisions and then increased
by 29 dB. This calibration ensures that all flaws
that have to be detected will give a signal of not less
than 2 screen divisions height, whether or not there
is attenuation present.
It is obvious when the minimum record level is
drawn on the diagram that reflections from certain
depths could give echoes greater than that level and
yet still be acceptable. To determine therefore
whether or not the echo is greater or less than the
minimum size or in fact to determine its actual
equivalent flaw size, the gain should be reduced by
the decibel difference between the 0.2S line and the
record level at the defect depth. The 2 screen
divisions is then the minimum record level at the
distance.

This example assumes no attenuation in the
material at the frequency in use, although by taking
the back echo as a reference any attenuation
present would only make the sensitivity too high
rather than too low. If the attenuation has been
measured it may turn out to be insignificant
1 - dB/mm (5 dB/m) in view of the thickness
e.g. ---------
200
involved, If the attenuation was significant,
however, as shown in Example 1 (1/30 dB/mm) then
this should be taken into account as follows:
Draw in the attenuation curve as shown. In this
case, since the back echo has been taken as the
reference reflector, the allowance for attenuation
is greater the nearer the reflector is to the
surface. The difference at the surface in this case
is 6.7 dB i.e. (100 mm 2 1/30).
Assuming now that a reflector in excess of the record
level was found at a depth of 0.3N, two corrections
should be made to the gain: 4 dB for the difference
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
between the 0.2S line and the record level, and 5 dB
for the record level to attenuation line, i.e. the total
correction to be made is a reduction in echo height
of 9 dB after which the decision as to whether or not
the flaw is rejectable is directly related to the record
level of 2 screen divisions.
If it is necessary to determine the actual equivalent
flaw size for the flaw echo at 0.3N, its amplitude in
decibels above the 2 screen divisions should be
measured and reduced by the decibel value, record
level to attenuation line. For example:
Flaw depth 62 mm = 0.6N
Echo height above 2 screen division = 19 dB
Record level to attenuation line = 3 dB
Therefore, final position above record
level = 16 dB (on 0.4S line)
From Figure 2, S crystal diameter = equivalent
flat-bottomed hole diameter
Therefore: 0.4 25 = 10 mm diameter
NOTE 1 In the example used the back echo happened to be
at 0.96N but it could have been 0.4N or 5.5N or any other value,
depending on the material thickness and the probe near field
length. Furthermore, the S values and the distance values in
near fields could both be given directly in millimetres once the
probe crystal diameter and near field length are known.
NOTE 2 The D.G.S. diagram shown in Figure 2 is based on the
established sound radiation laws. It is essential, therefore, that
the amplifier and probe characteristics in use do not give
conflicting results from those shown on the diagram. In other
words, the flaw detector and probe combination should be
entirely compatible with the D.G.S. diagram in use.
BSI 01-1999
BS 3923-2:1972
B.2 Angle probes. (See Figure 3 and Figure 4.)
B.2.1 General. The D.G.S. diagram for angle probes
is essentially the same as for normal probes, except
that distance is measured from the probe index and
therefore most of the near field is fortunately lost in
the probe. The horizontal axis as before gives the
distance along the beam path; corresponding
equivalent surface distances are shown for the
different angle probes. Note that in the case of the
two D.G.S. diagrams shown in Figure 3 and
Figure 4, the probe size is stated and therefore the
equivalent flat-bottomed hole sizes and the
horizontal axis have been calibrated directly in
millimetres.
In order to use these diagrams to determine the
sensitivity setting the operator first needs to know
the smallest disc-shaped flaw or equivalent reflector
size that has to be detected and secondly, at what
amplitude he wishes to record it.
To set the sensitivity for weld testing with an angle
probe however, is more complex than using a simple
normal probe. Because there is no backwall echo to
use as a standard reference, a separate reference
echo has to be used, such as the 100 mm radius of
the A2 calibration block (see BS 2704). This means
that an allowance has to be made for the different
surface conditions (transfer loss) between the test
block and the plate, the different attenuation factors
between the two materials and also the difference in
distance between the 100 mm radius reflector and
the farthest defect distance in the material.
The procedure to arrive at the correct gain setting to
ensure that the smallest defect to be detected
throughout the scanning distance will at least give
an echo at the appropriate screen height, taking into
account the factors mentioned above, is as follows:
1) Measure attenuation of material (see B.2.2).
2) Measure transfer loss between A2 test block
and plate (see B.2.3).
3) Read difference in decibels between back echo
from 100 mm radius on A2 block to minimum
defect size at worst position on D.G.S. diagram
which should be available for the probe in use.
4) Set back echo from 100 mm radius to 2
divisions, plus difference in decibels from (3) plus
transfer loss from (2) plus difference in
attenuation from 100 mm radius on A2 block to
maximum testing distance in plate.
7
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
BSI 01-1999 8

Figure 2 Typical D.G.S. diagram for normal probes

BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
9 BSI 01-1999

BS 3923-2:1972
 10

BS 3923-2:1972
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
BSI 01-1999

Figure 4 Typical D.G.S. diagram for angle probe of 4 MHz, 8 mm 9 mm crystal size

Example 3
Probe of 4 MHz, 8 mm 9 mm crystal size
Smallest defect to be detected equivalent
to 2 mm diameter disc-shaped flaw
Maximum defect distance 150 mm
Transfer loss 4 dB
Attenuation of A2 block 40 dB per metre beam
path in front of probe
Attenuation of plate 80 dB per metre beam path
in front of probe
The setting required to bring a 2 mm diameter
equivalent flaw to the record level at the worst
distance (150 mm) would be an echo from
the A2 block 100 mm radius to 2 divisions,
plus 30 dB from (3) plus 4 dB from (2), plus 8 dB
difference in attenuation from 100 mm radius on A2
block to maximum testing distance in plate,
i.e. an A2 back echo to 2 divisions plus 42 dB.
With this sensitivity, the minimum equivalent flaw
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
size to be recorded will at least reach 2 screen
divisions, wherever it appears throughout the
scanning distance. It is possible as with the normal
probe test that echoes in excess of 2 screen divisions
could be acceptable. To determine whether or not
this is so, and at the same time to evaluate the
equivalent flaw size for any such echo, proceed as
follows:
a) Draw on the diagram the attenuation curve as
shown.
b) Locate the defect depth on the diagram in
distance and amplitude above the record level.
c) Reduce the defect position by the difference in
decibels between the record level and the
attenuation line.
Example 4
Defect echo 50 mm beam path from screen
Echo height 34 dB greater than record level
of 2 screen divisions
Record level to attenuation line 8 dB
Therefore, new position on
diagram = 34 8 = 26 dB above record level and
hence equivalent flaw size is 4 mm diameter.
B.2.2 Measurement of shear wave attenuation. The
shear wave attenuation of the plate can be
measured as follows:
1) Use one probe of the same angle and frequency
as will be used for the test and calibrate the
screen for beam path.
2) Use two probes, separate transmitter and
receiver, and locate as shown in Figure 5 so that
signal A. B1 appears at 50 mm and signal A. B2
at 100 mm on the screen.
BSI 01-1999
BS 3923-2:1972
3) Measure the difference between through
transmission signal height A. B1 and A. B2.
4) Subtract the difference in decibels indicated on
the D.G.S. diagram, the remainder being
attenuation over half skip distance on the screen
or 100 mm of beam path.
Example 5
Probes of 4 MHz, 8 mm 9 mm crystal size
1
-- skip distance 50 mm
2
Signal height A. B1 44 dB
Signal height A. B2 54 dB
------
10 dB
Difference according to D.G.S.
diagram, 5 dB (50 mm to 100 mm)
Therefore,
10 dB 5 dB
attenuation = ----------------------------------- = 1 10 dB mm of
50 mm
beam path on screen.
B.2.3 Measurement of transfer loss. The transfer
loss between the A2 block and the plate under test
can be measured as follows:
1) Use one probe of the same angle and frequency
as will be used for the test and calibrate the
screen for beam path.
2) Use two probes, separate transmitter and
receiver, and locate as shown in Figure 6.
3) Measure difference between through
transmission signal height of A2 block and plate.
4) Subtract the difference indicated on the D.G.S.
diagram, the remainder being transfer loss.
Example 6
Probes of 4 MHz, 8 mm 9 mm crystal size
Through transmission signal
( 1-- skip on screen), A2 block 56 dB
2
Through transmission signal
( 1-- skip on screen), plate 68 dB
------
2
12 dB
Difference according to D.G.S.
diagram, 3 dB (50 mm to 80 mm)
Attenuation difference according to attenuation
curve for A2 block at 50 mm and attenuation
curve for plate at 80 mm, 5 dB.
Therefore, transfer loss = 12 3 5 = 4 dB.
11
 BS 3923-2:1972
Appendix C Method for setting
sensitivities where maximum
sensitivity is required
After following the requirements of the standard for
a particular probe selection and time-base
calibration the sensitivity settings should be made
as follows:
1) The probe is applied to the test surface and the
gain controls are adjusted so that the grass
level of the material grain structure is
raised 2 mm high above the base-line at the
appropriate testing distance. Any reduction or
increase in the sensitivity to facilitate the
exploration of defects is to be noted in decibels.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI
Figure 5 Measurement of shear wave attenuation
Figure 6 Measurement of transfer loss
12
2) The probe is next placed on the A2 calibration
block (see BS 2704) and a record made of the
difference in the attenuator reading required to
bring the echo from the 100 mm radius for angle
probes, or the 100 mm parallel section for normal
probes, down to full screen height.
3) The attenuator difference between the 100 mm
radius for angle probes, or the 100 mm parallel
section for normal probes, of the A2 block and
that required to raise an echo from a 1.5 mm
drilled hole at an appropriate beam path
distance, is recorded.
It should be stated whether this drilled hole is
parallel to or at right angles to the test surface.
BSI 01-1999
 BS 3923-2:1972

Example of maximum sensitivity record

Calibration: Screen calibrated to 250 mm full

path length
Probe selection: 2 MHz, 45, single crystal
Testing To give grass level up
sensitivity: to 250 mm path
A2 calibration (100 mm radius) echo
block: requires 28 dB inserted to bring
it down to full screen height
Test block with Hole was drilled at 125 mm path
1.5 mm hole: length parallel to the test
surface; requires 16 dB inserted
from testing sensitivity level to
give full screen height from this
hole
Allowance can then be made for the transfer loss
and the average attenuation of the test piece as
described in Appendix B.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

BSI 01-1999 13
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

14
blank
 BS 3923-2:1972

BSI publications referred to in this standard:

This standard makes reference to the following British Standards:

BS 2704, Calibration blocks and recommendations for their use in ultrasonic flaw detection.
BS 3683, Glossary of terms used in non-destructive testing. Part 4. Ultrasonic flaw detection.
BS 3889, Methods for non-destructive testing of pipes and tubes. Part 1A. Ultrasonic testing of ferrous pipes
(excluding cast).
BS 3923, Methods for ultrasonic examination of welds. Part 1. Manual examination of fusion welded butt
joints in ferritic steels.
BS 4336, Methods for non-destructive testing of plate material. Part 1A. Ultrasonic detection of laminar
imperfections in ferrous wrought plate.
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

BSI 01-1999
 |
|
|
|
|
|
|
|
|
BSI British Standards Institution
|
|
|
|
|
|
| BSI is the independent national body responsible for preparing British Standards. It
|
| presents the UK view on standards in Europe and at the international level. It is
| incorporated by Royal Charter.
|
|
| Revisions
|
|
| British Standards are updated by amendment or revision. Users of British Standards
|
| should make sure that they possess the latest amendments or editions.
|
|
| It is the constant aim of BSI to improve the quality of our products and services. We
|
| would be grateful if anyone finding an inaccuracy or ambiguity while using this
| British Standard would inform the Secretary of the technical committee responsible,
|
| the identity of which can be found on the inside front cover. Tel: 020 8996 9000.
|
| Fax: 020 8996 7400.
|
|
| BSI offers members an individual updating service called PLUS which ensures that
|
| subscribers automatically receive the latest editions of standards.
|
|
| Buying standards
|
| Orders for all BSI, international and foreign standards publications should be
|
| addressed to Customer Services. Tel: 020 8996 9001. Fax: 020 8996 7001.
|
|
| In response to orders for international standards, it is BSI policy to supply the BSI
|
Licensed Copy: Tom Magee, Howden Power, 06 March 2004, Uncontrolled Copy, (c) BSI

| implementation of those that have been published as British Standards, unless

|
| otherwise requested.
|
|
| Information on standards
|
| BSI provides a wide range of information on national, European and international
|
| standards through its Library and its Technical Help to Exporters Service. Various
|
| BSI electronic information services are also available which give details on all its
|
| products and services. Contact the Information Centre. Tel: 020 8996 7111.
|
| Fax: 020 8996 7048.
|
|
| Subscribing members of BSI are kept up to date with standards developments and
| receive substantial discounts on the purchase price of standards. For details of
|
| these and other benefits contact Membership Administration. Tel: 020 8996 7002.
|
| Fax: 020 8996 7001.
|
|
| Copyright
|
|
| Copyright subsists in all BSI publications. BSI also holds the copyright, in the UK, of
|
| the publications of the international standardization bodies. Except as permitted
| under the Copyright, Designs and Patents Act 1988 no extract may be reproduced,
|
| stored in a retrieval system or transmitted in any form or by any means electronic,
|
| photocopying, recording or otherwise without prior written permission from BSI.
|
|
| This does not preclude the free use, in the course of implementing the standard, of
|
| necessary details such as symbols, and size, type or grade designations. If these
|
| details are to be used for any other purpose than implementation then the prior
| written permission of BSI must be obtained.
|
|
| If permission is granted, the terms may include royalty payments or a licensing
|
| agreement. Details and advice can be obtained from the Copyright Manager.
|
| Tel: 020 8996 7070.
|
|
|
|
|
|
|
|
|
BSI |
|
389 Chiswick High Road |
|
London |
|
W4 4AL |
|
|
|
|
|
|