You are on page 1of 17

Chapter 8

Piezoelectric materials: thermal-electrical


analogy

8.1 Smart materials

Smart materials are new generation materials surpassing the conventional structural and
functional materials. These materials possess adaptive capabilities to external stimuli, such as
loads or environment, with inherent intelligence. (Rogers, 1988; Rogers et al., 1988) defined
smart materials as materials, which possess the ability to change their physical properties in a
specific manner in response to specific stimulus input. The stimuli could be pressure,
temperature, electric and magnetic fields, chemicals, hydrostatic pressure or nuclear radiation.
The associated changeable physical properties could be shape, stiffness, viscosity or damping.
Takagi (1990) explained it as intelligent materials that respond to environmental changes at the
most optimum conditions and reveal their own functions according to the environment.
Smartness describes self-adaptability, self-sensing, memory and multiple functionalities of the
materials or structures. These characteristics provide numerous possible applications for these
materials and structures in aerospace, manufacturing, civil infrastructure systems, biomechanics
and environment. There are a number of types of smart material, of which are already common.
Some examples are listed in Figure 8.1.

Figure 8.1: Common smart materials and associated stimulus-response.

There is an increasing awareness of the benefits to be derived from the development and
exploitation of smart materials and structures in applications ranging from hydrospace to
aerospace. With the ability to respond autonomously to changes in their environment, smart
systems can offer a simplified approach to the control of various material and system
characteristics such as noise, shape and vibration, etc., depending on the smart materials used.
With the ability to develop high strains and to act with small or suitably modeled hysteresis, these
materials offer engineers the opportunity to micomanipulate optical devices, small robots, and
other system components.
In aerospace field, smart materials have been developed to suppress vibrations and change shape
in helicopter rotor blades. Shape-memory-alloy devices are also being developed that are capable
of achieving accelerated breakup of vortex waves of submarines and similarly different adaptive
control surfaces are developed for airplane wings. Some materials and structures can be termed
‘sensual’ devices. These are structures that can sense their environment and generate data for use
in health and usage monitoring systems (HUMS) or in particular Structural Health Monitoring
(SHM). To date the most well established application of HUMS and SHM are in the field of
aerospace, in areas such as aircraft checking.
Piezoelectric materials are most popularly used in smart structures and systems, because the same
material is applicable for both sensors and actuators, in principle. Even though transducers, in
general, are devices that convert input energy to a different energy type of output, the
piezoelectric “transducer” is oft en used to denote a device that possesses both sensing and
actuating functions, exemplified by underwater sonar.

8.2 The piezoelectric material

Piezoelectricity is a property exhibited by materials that become electrically charged when


subjected to mechanical stress. The converse effect, in which a mechanical deformation is
induced by an applied electric field, also occurs. Experimental measurement of the piezoelectric
effect was first reported by Pierre and Jacques Curie in 1880. The direct effect may be used in
sensing applications, while the indirect effect may be used in actuation and acoustic transduction.
The piezoelectric effect is caused by an asymmetry in the unit cell and the resulting relationship
between mechanical distortion and electric dipole separation. The effect may be quantified
through the use of suitable piezoelectric coefficients, the measurement of which has been
described in various standards on piezoelectricity.
Since the piezoelectric effect exhibited by natural materials such as quartz, tourmaline, rochelle
salt, etc. is limited in terms of usable power, polycrystalline ferroelectric ceramic materials such
as barium titanate (BaTiO3) and lead zirconate titanate (PZT) with improved properties have been
developed over the last few decades. Likes of PZT ceramic are available in many variations and
are still the most widely used materials for actuator applications today. Prior to polarization
mechanism, PZT crystallites have symmetric cubic unit cells. At temperatures below the Curie
temperature, the lattice structure becomes deformed and asymmetric. The unit cells exhibit
spontaneous polarization i.e. the individual PZT crystallites are piezoelectric.
A traditional piezoelectric ceramic is a mass of perovskite structure (material with the same type
of crystal structure as calcium titanium oxide - CaTiO3), each consisting of a small, tetravalent
metal ion, usually titanium or zirconium, in a lattice of larger divalent metal ions, usually lead or
barium, and O2- ions. Under conditions that confer tetragonal or rhombohedral symmetry on the
crystals, each crystal has a dipole moment, see Figure 8.2.

Figure 8.2: Schematic of the crystal structure of Pb(Zr,Ti)O3 ceramics: above the Curie point, the cell is
cubic (left ); below the Curie point, the cell is tetragonal/rhombohedral.

To prepare a piezoelectric ceramic, fine powders of the component metal oxides are mixed in
specific proportions then heated to form a uniform powder. The powder is mixed with an organic
binder and is formed into structural elements having the desired shape (discs, rods, plates, etc.),
for this research, a 120µm and 250µm diameter fibres were used. The elements are fired
according to a specific time and temperature, during which the powder particles sinter and the
material attains a dense crystalline structure. The elements are cooled, then shaped or trimmed to
specifications, and electrodes are applied to the appropriate surfaces. Above a critical
temperature, the Curie point, each perovskite crystal in the fired ceramic element exhibits a
simple cubic symmetry with no dipole moment (Figure 8.2 left). At temperatures below the Curie
point, however, each crystal has tetragonal or rhombohedral symmetry and a dipole moment
(Figure 8.2 right).
To minimize the internal energy of the material, ferroelectric domains form in the crystallites of
the ceramic. Within these volumes, the orientations of the spontaneous polarization are the same.
The different orientations of bordering domains are separated by domain walls. The direction of
polarization among neighbouring domains is random, however, so the ceramic element has no
overall polarization (Figure 8.3a). A ferroelectric polarization process is required to make the
ceramic macroscopically piezoelectric as well. For this purpose, a strong electric field of several
kV/mm is applied to create an asymmetry in the previously unorganized ceramic compound,
usually at a temperature slightly below the Curie point (Figure 8.3b). The electric field causes a
reorientation of the spontaneous polarization. At the same time, domains with a favorable
orientation to the polarity field direction grow and those with an unfavorable orientation shrink.
The domain walls are shifted in the crystal lattice. The polarisation is associated with a change
in length, Δ𝑆. After polarization, most of the reorientations are preserved even without the
application of an electric field. However, a small number of the domain walls are shifted back to
their original position, e.g., due to internal mechanical stresses (Figure 8.3c). The element now
has a permanent polarization, the remnant polarization Pr , and is permanently elongated of Δ𝑆# .

Figure 8.3: Polarizing of a piezoelectric ceramic, left, random orientation of a polar domains prior to
polarization, centre, polarization in DC electric field and right, remnant polarization after electric field is
removed.

8.3 The piezoelectric effect

Mechanical compression or tension on a poled piezoelectric ceramic element changes the dipole
moment, creating a voltage. Compression along the direction of polarization, or tension
perpendicular to the direction of polarization, generates voltage of the same polarity as the poling
voltage (Figure 8.4b). Tension along the direction of polarization, or compression perpendicular
to the direction of polarization, generates a voltage with polarity opposite that of the poling
voltage (Figure 8.4c). These actions are the direct effect (i.e. for actuation), where by the ceramic
element converts the mechanical energy of compression or tension into electrical energy.

Figure 8.4: Direct and indirect effect of piezoelectric material showing voltage generation when
compressed and shape change when voltage is applied.

If a voltage of the same polarity as the poling voltage is applied to a ceramic element, in the
direction of the poling voltage, the element will lengthen and its diameter will become smaller
(Figure 8.4d). If a voltage of polarity opposite that of the poling voltage is applied, the element
will become shorter and broader (Figure 8.4e). If an alternating voltage is applied, the element
will lengthen and shorten cyclically, at the frequency of the applied voltage. This is the converse
effect (i.e. for sensing), where by electrical energy is converted into mechanical energy.
It is now necessary introduce a 3D system of reference in the representative unit cell, for which
the axis x3 has been chosen aligned with the polarization direction, as illustrated in Figure 8.5.
Figure 8.5: undeformed and deformed configuration of a PbTiO3 cell under the action of a stress 𝜎%
applied along the direction of polarisation x3

If we apply a mechanical stress 𝜎% directed along x3 and applied to the surface of the cell of
normal x3 the cell will strain along x3 with the result of increasing the distance between the atom
of titanium and the geometrical center of the unit cell. This can be viewed as an increase of the
polarization of the material from the remnant polarization 𝑃# to 𝑃# + ∆𝑃 measured along the x3
axis. For small values of the applied stress the ∆𝑃% is proportional to 𝜎% .
∆𝑃% = 𝑑%% 𝜎% (8.1)
with 𝑑%% a positive constant with two pedices indicating the direction of polarisation (x3) and the
one of the applied stress (𝜎% ). On the contrary if the mechanical tensile stress is produced along
a direction (x1) normal to polarisation the effect on the displacement of the titanium atom with
respect to the centre of the cell will be the opposite, resulting in a negative 𝑑%+ (Figure 8.6)
∆𝑃% = 𝑑%+ 𝜎+ (8.2)

Figure 8.6: undeformed and deformed configuration of a PbTiO3 cell under the action of a stress 𝜎+
applied along a direction x1 normal to the direction of polarisation.

An identical behaviour is observed along third axis x2 with


∆𝑃% = 𝑑%, 𝜎, (8.3)
with 𝑑%, = 𝑑%+
In case of a shear stress 𝜎- the Ti4+ is displaced on the right, aligned with x1, resulting in a positive
coefficient 𝑑+- (Figure 8.7)
∆𝑃+ = 𝑑+- 𝜎- (8.4)

Figure 8.7: deformed configuration of a PbTiO3 cell under the action of a shear stress 𝜎- .

Also for shear stresses an identical behaviour can be observed in case of a shear stress 𝜎. for
which we have
∆𝑃, = 𝑑,. 𝜎. (8.5)
with 𝑑,. = 𝑑+-
We also recognize that no variation in the polarization of the crystal structure can be obtained if
applying a 𝜎/ stress, that is a shear stress contained in the x1x2 plane. In conclusion by ordering
the 𝑑01 coefficients in a 3x6 matrix and recognizing all the cases where no effect of variation of
polarization in x3 or of polarization along x1 and x2 is observed, we obtain the direct piezoelectric
coupling matrix 𝑑.

0 0 0 0 𝑑+- 0
𝑑 = 0     0     0    𝑑+-     0    0 (8.6)
𝑑%+ 𝑑%+ 𝑑%% 0 0 0
with
∆𝑃 = 𝑑  𝜎
(8.7)
being ∆𝑃 = ∆𝑃+   , ∆𝑃,   , ∆𝑃% and 𝜎 = 𝜎+   , 𝜎,   , 𝜎%   , 𝜎.  , 𝜎-   , 𝜎/ .

In an analogous way the physical reason for the converse piezoelectric effect could be
investigated.

Figure 8.8: deformation of a PbTiO3 unit cell under the effect of an electric field E3 applied in the
direction of polarisation x3.
The effect of an application of an electric field along the x3 direction will be the moving in the
positive direction of x3 the positive charges and in the opposite direction the negative ones (Figure
8.8). This will result in a stretching of the cell along x3 and a squeezing along x1 or x2. In terms
of strain components 𝜀+ , 𝜀, and 𝜀% we will then obtain
𝜀% = 𝑑%% 𝐸%
𝜀+ = 𝑑+% 𝐸% (8.8)
𝜀, = 𝑑,% 𝐸%
with 𝑑+% = 𝑑,% for reasons that are analogous to what discussed for direct piezoelectric case.
Similarly, by applying an electric field E1 along the x1 direction, due to the laws that regulate the
action of an electric field on a charged particle, the center of positive charges (in our
simplification the Ti atom) will move along x1 while the centre of negative charges will move the
opposite direction. Due to the positions of Ti along x3 this will result in a positive shear strain of
the unit cell as represented in the Figure 8.9.

Figure 8.9: effect of the application of an electric field E1 in the direction x1 normal to the polarisation
direction.

In this case we will obtain

𝜀- = 𝑑-+ 𝐸+ (8.9)
and in analogous way also the remaining coefficient would be obtained

𝜀. = 𝑑., 𝐸, (8.10)
with 𝑑., = 𝑑-+
In this case, by putting in an ordered matrix all the d constants, we obtain a 6x3 matrix which
happens to be the transpose of the matrix obtained for the direct piezoelectric coupling
0 0 𝑑+% 0 0 𝑑%+
0 0 𝑑+% 0 0 𝑑%+
𝑑 7 = 0       0      𝑑%% = 0       0      𝑑%% (8.11)
0 𝑑., 0 0 𝑑+- 0
𝑑-+ 0 0 𝑑+- 0 0
0 0 0 0 0 0
being 𝑑+% = 𝑑%+ and 𝑑+- = 𝑑-+ , with
𝜀 = 𝑑 7  𝐸
(8.12)
where 𝐸 = 𝐸+   , 𝐸,   , 𝐸% is the electric vector and 𝜀 is the vector that lists the second order
tensorial components of the strain 𝜀 = 𝜀+  , 𝜀,  , 𝜀%  , 𝜀.  , 𝜀-  , 𝜀/ . For the strain tensorial
components it is worth noting that 𝜀+ = 𝜀++  , 𝜀, = 𝜀,,  ,  𝜀% = 𝜀%%  , 𝛾. = 2𝜀,% = 2𝜀%,  , 𝛾- =
2𝜀+% = 2𝜀%+  , 𝛾/ = 2𝜀+, = 2𝜀,+  for the well known relation between the engineering shear
component 𝛾01 and the tensorial component 𝜀01  of the strain term for which 𝛾01 = 2𝜀01 with i, j =
1, 2, 3. The circumstance that the piezoelectric metrics obtained for the direct effect is the
transpose of the one obtained for the converse effect makes symmetric the overall constitutive
metrics. This evidence is related to the conservative nature of both elastic and electric forces.
The polarization Pi (C m−2) that is induced in an insulating, polarizable material (a dielectric) by
an applied electric field vector Ei (V m−1) is given by:

𝑃0 = 𝜒01 𝐸1 (8.13)
where 𝜒01 (F m−1) is the tensor known as the dielectric susceptibility of the material. Relation
(8.13) is valid only for linear materials or in a linear limit for nonlinear materials and, in general,
Pi depends on higher-order terms of the field. The total surface charge density that is induced in
the material by the applied field is given by the dielectric displacement vector, Di (C m−2):

𝐷0 = 𝜖= 𝐸0 + 𝑃0 (8.14)
Scalar 𝜖= = 8.854 ∙ 10D+,  F/m is known as the dielectric permittivity of a vacuum. It follows
from (8.13) and (8.14) that

𝐷0 = 𝜖= 𝐸0 + 𝜒01 𝐸1 = 𝜖= 𝛿01 𝐸1 + 𝜒01 𝐸1 = 𝜖= 𝛿01 + 𝜒01 𝐸1 = 𝜖01 𝐸1 (8.15)


where 𝜖01 = 𝜖= 𝛿01 + 𝜒01 is the dielectric permittivity of the material and 𝛿01 is Kronecker’s
symbol. For most ferroelectric materials 𝜖= 𝛿01 ≪ 𝜒01 and 𝜖01 ≅ 𝜒01 . In practice, the relative
dielectric permittivity, 𝜅01 = 𝜖01 𝜖= , also known as the dielectric constant of the material, is more
often used than the dielectric permittivity. Using free energy arguments it can easily be shown
that 𝜒01 (as well as 𝜖01 and 𝜅01 ) must be symmetrical tensor 𝜒01 = 𝜒10 with only six independent
components. Symmetry of a material may further reduce the number of independent components
of the permittivity tensor.
Another coefficient g can be defined that relates the mechanical stress applied to the electrical
field E produced:
𝑑01 𝑑01
𝐸0 = 𝑔01 𝜎1 = 𝜎1         ⟹           𝑔01 = (8.14)
𝜖01 𝜖01

The coefficients d and g are very significant for the performance of the piezoelectric material
considered as a smart material capable of actuation and sensing functionalities. In fact, once
assigned the level of the electric field available, a material with a high d is best suited for
performing actuation while a high g is interesting for maximizing the performances of sensors of
the state of stress 𝜎.
Another parameter is the piezoelectric coupling factor k. It is a measurement of the overall
strength of the electromechanical effect. It is often defined as the square root of the ratio of
electrical energy output to the total mechanical energy input (in the case of the direct effect) or
the mechanical energy available to the total electrical energy (in the case of the converse effect).
The value of k is of course always less than unity because energy conversion is always
incomplete.

8.4 Constitutive equations for piezoelectricity

The presence of an additional polarization ∆𝑃, due to the direct piezoelectric effect, can be viewed
also as an additional term to the electric displacement vector D. This means that the final form
for the expression of the electric displacement can be written as
N
𝐷=𝜖  𝐸 + 𝑑  𝜎
(8.15)
Where 𝜖 is the dielectricity (permittivity) matrix obtained for zero or constant stress 𝜎
(superscript (𝜎)  ), that can be written as
𝜖++ 0 0
𝜖= 0 𝜖,, 0 (8.16)
0 0 𝜖%%
for orthotropic material.
In an analogous way the expression of strain obtained from piezoelectric effect can be extended
for including the strain due to stress assuming a linear elastic behaviour of the piezoelectric
material:

𝜀 = 𝐹 (R)  𝜎 + 𝑑 7  𝐸
(8.17)
where F(E) is the compliance matrix obtained for a generalized Hooke law obtained with zero or
constant electric field E (superscript (E) ). For orthotropic material:
𝐹++ 𝐹+, 𝐹+%    0        0        0    
𝐹+, 𝐹,, 𝐹,% 0 0 0
𝐹 𝐹,% 𝐹%% 0 0 0
𝐹   = +%   (8.18)
   0        0        0     𝐹..    0        0    
0 0 0    0     𝐹--    0    
0 0 0    0        0     𝐹//
The ferroelectric ceramics generally used by the piezoelectric material manufacturers present a
symmetry in their hexagonal crystallographic structure and can be considered transversely
isotropic materials, being the plane normal to the direction 3 of polarization the isotropy plane.
In that case 𝐹,, = 𝐹++ ,  𝐹,% =   𝐹+% , 𝐹-- = 𝐹.. , 𝐹// = 2 𝐹++ − 𝐹+, and 𝜖++ = 𝜖,, .
Compliance coefficients 𝐹01 are related to the engineering material properties, Young’s moduli
𝑌0 , shear moduli 𝐺01 and Poisson’s ratios 𝜈01 by
1 1 1 𝜈,+ 𝜈+, 𝜈%+ 𝜈+%
𝐹++ = , 𝐹,, =  , 𝐹%% =  , 𝐹+, = − =−  , 𝐹+% = − =−
𝑌+ 𝑌, 𝑌% 𝑌, 𝑌+ 𝑌% 𝑌+

𝜈%, 𝜈,% 1 1 1
𝐹,% = − =− , 𝐹.. =  , 𝐹-- =  , 𝐹// =
𝑌% 𝑌, 𝐺,% 𝐺%+ 𝐺+,
We can write:
𝜀+ 𝐹++ 𝐹+, 𝐹+%    0        0                          0                       𝜎+ 0 0 𝑑%+
𝜀, 𝐹+, 𝐹++ 𝐹+% 0 0 0 𝜎, 0 0 𝑑%+ 𝐸
𝐹+% 𝐹+% 𝐹%% +
𝜀 0 0 0 𝜎%
𝛾. =    0       𝜎 + 0       0      𝑑%% 𝐸, (8.19)
   0        0     𝐹..    0        0     . 0 𝑑+- 0
𝛾- 𝜎- 𝐸%
0 0 0    0     𝐹..    0     𝑑+- 0 0
𝛾/ 0 0 0    0        0     2 𝐹++ − 𝐹+, 𝜎/ 0 0 0

𝜎+
𝜎,
𝐷+ 𝜖++ 0 0 𝐸+ 0 0 0 0 𝑑+- 0
𝜎
𝐷, = 0 𝜖++ 0 𝐸, + 0     0     0    𝑑+-     0    0 𝜎% (8.20)
.
𝐷% 0 0 𝜖%% 𝐸% 𝑑%+ 𝑑%+ 𝑑%% 0 0 0 𝜎
-
𝜎/

8.5 Correspondence between piezoelectric strain and thermal strain

The generalized Hooke’s law taking into account the thermal effect can be written:

𝜎 = 𝐶 (R)  𝜀 − 𝐶 (R) 𝛼  ∆𝑇


(8.21)
where 𝛼 is a 6 x 1 thermal expansion coefficient vector and ∆𝑇 = 𝑇 − 𝑇= is a temperature
difference, relative to a reference temperature 𝑇= .
Eq. 8.17 can be written as:
D+ D+ 7
𝜎 = 𝐹 (R)    𝜀 − 𝐹 (R) 𝑑  𝐸 (8.22)
or, introducing the stiffness matrix 𝐶 (R) obtained for zero or constant electric field:

𝜎 = 𝐶 (R)    𝜀 − 𝐶 (R)  𝑑 7  𝐸
(8.23)
Combining Eq. 8.21 and 8.22 , piezoelectric strains and thermal strains relationship is obtained:
𝑑 7  𝐸 = 𝛼  ∆𝑇
(8.22)
As only d31, d32, d33, d24 and d15 are non-null, Eq. 8.22 can be reduced to:
∆𝑉%
𝑑%0 = 𝛼0 ∆𝑇            with      𝑖 = 1,2,3
𝑡
∆𝑉,
𝑑,. = 𝛼. ∆𝑇 (8.23)
𝑡
∆𝑉+
𝑑+- = 𝛼- ∆𝑇
𝑡
where ∆𝑉% is the voltage difference between the electrodes and t is the thickness of the
piezoceramic actuator while ∆𝑉+ and ∆𝑉, are the in-plane voltage differences. If both out-of-
plane and in-plane electric fields are applied, the Eq. (8.23) should be used for the application of
the thermal analogy. In practice, i.e. in one-dimensional actuation cases, the in-plane electric
∆bc
fields E1 and E2 are negligible. Forcing E1 and E2 to zero and setting 𝐸% = , an exact analogy
d
between piezoelectric strain and thermal strain is obtained through the following relation:
∆𝑉%
𝑑%0 = 𝛼0 ∆𝑇            with      𝑖 = 1,2,3 (8.24)
𝑡
Eq. (8.24) indicates that the voltage actuation of a piezoceramic can be simulated exactly using
conventional three-dimensional elastic elements with the thermal actuation. One has to impose
∆bc
the thermal expansion coefficient 𝛼0 as 𝑑%0 and to impose a temperature change ∆𝑇 equal to .
d
It is important here to emphasize the limitations of thermal analogy. First, the variation of
electrical potential has to be assumed: since the piezoelectric element are used as actuators, the
tension between electrodes is imposed, and the variation of potential is likely to be linear in the
thickness (constant electric field). The variations in the in-plane directions parallel to the
electrodes is hard to postulate. Since the related effect is expected to be small compared to the
effect of out-of-plane tension, the in-plane electric potential is assumed constant (null electric
field: E1 = E2 = 0). Second, since the thermal effect is not reciprocal (temperature produces strains
but strains do not produce temperature), only the converse effect of linear piezoelectricity can be
modelled. It means that piezoelectric element is acting as actuator and not as sensor.

8.6 Bending actuation - The bimetallic strip analogy

Let us now consider how to induce a bending deformation by means of an axial induced strain
actuation. In the model representing the action of an axial actuator the active material is working
against an axial stiffness. In this way axial displacement is produced with the relevant axial force,
both depending on the ratio of the stiffness of the actuator and the one of the passive structure. It
is now interesting to understand how the effect of an expansion of an axial actuator could produce
a bending effect on a structure. This could be done by letting the actuator be patched on the
surface of the passive structure, as in the well known case of a bimetallic strip, represented in
Figure 8.18.
Figure 8.10: the bimetallic strip analogy for induced bending

We consider two bars of metals of different characteristics, namely with two different thermal
expansion coefficients, perfectly bonded along a middle surface and we heat the structure by
increasing the temperature of an increment ∆𝑇. The effect will be a bending of the overall beam.
In fact, by assuming 𝛼e < 𝛼g , the part of the bar made of metal A will expand more than the part
B. Since the two bars are adherent, the metal at the tip will constrain the displacement of the metal
at the bottom. This will produce an axial expansion and a bending as represented in Figure 8.11.
For sake of simplicity let us assume that both metals behave as linear elastic materials and have
the same Young modulus.

Figure 8.11: mechanism of bending actuation in a bimetallic strip

By establishing the coordinate system in Figure 8.11, it is easy to write the governing equations
for the bending behaviour of the bimetallic strip, considered as a bending beam.
Constitutive equations (different for part A and part B):
1
𝜀g = 𝜎 + 𝛼g ∆𝑇
𝐸 g
1 (8.25)
𝜀e = 𝜎e + 𝛼e ∆𝑇
𝐸
Kinematic equations:

𝜀h 𝑦 = −𝜒𝑦 (8.26)
Equilibrium equations:
𝑑𝑇 𝑑𝑀
= 0                   = 0                  𝑀 𝐿 = 0                𝑇 𝐿 = 0 (8.27)
𝑑𝑥 𝑑𝑥
Being M the bending moment, T the shear force and the bending curvature.
Since 𝑀 0 = 0 one can immediately obtain, assuming a unitary width of the beam:
d = d
𝑀= 𝜎  𝑦  𝑑𝑦 = 𝐸 𝜀 − 𝛼g ∆𝑇  𝑦  𝑑𝑦 + 𝐸   𝜀 − 𝛼e ∆𝑇  𝑦  𝑑𝑦 =
Dd Dd =
d = d (8.28)
,
= 𝐸𝜒 𝑦  𝑑𝑦 − 𝐸𝛼g ∆𝑇 𝑦  𝑑𝑦 − 𝐸𝛼e ∆𝑇 𝑦  𝑑𝑦 = 0
Dd Dd =

that leads to
1 % ∆𝑇
𝑡 𝜒=− −𝛼g 𝑡 , + 𝛼e 𝑡 , (8.29)
12 2
in conclusion
∆𝑇 ,
𝑡 𝛼e − 𝛼g 6∆𝑇
𝜒= 2 =− 𝛼e − 𝛼g (8.30)
1 % 𝑡
𝑡
12
and by substitution in 8.26:
6∆𝑇
𝜀h 𝑦 = 𝛼e − 𝛼g 𝑦 (8.31)
𝑡
Let us now imagine that in this two layer structure, a thermal deformation is induced only in the
part B and no thermal deformation is present in the part A. This can be viewed as an induced
strain Λ produced in the active layer B in the presence of a passive layer, where no induced strain
is present. From the viewpoint of the induced strain this case could be represented by previous
equations by setting Λ = 𝛼e ∆𝑇 and considering the passive layer as a material with 𝛼g = 0.
By substituting in eq. 8.30 we obtain the bending curvature as
6
𝜒=− Λ (8.32)
𝑡

Figure 8.12: mechanism of bending actuation in the presence of an active and a passive layer of a beam
In Figure 8.12 the bending deformation of a two layer structure (part B active; part A passive) is
illustrated. Of course this result can be obtained directly but the example reinforces the concept
that an induced strain Λ can be considered as the source for actuation, no matter what is the cause
(in the present case the temperature) that is generating it.
Coming back to the bimetallic strip, an axial deformation is also produced in the two layer system.
By expressing the axial force N as done for the bending moment we obtain:
d = d
𝑁= 𝜎  𝑑𝑦 = 𝐸 𝜀 − 𝛼g ∆𝑇  𝑑𝑦 + 𝐸   𝜀 − 𝛼e ∆𝑇  𝑑𝑦 =
Dd Dd =
(8.33)
= 2  𝐸  𝑡  𝜀 − 𝐸  ∆𝑇  𝑡  𝛼g + 𝐸  ∆𝑇  𝑡  𝛼e
from which we obtain:
∆𝑇
𝜀= 𝛼 + 𝛼e (8.34)
2 g
Since 𝛼g > 𝛼e we will have always a tensile elongation 𝜀 for the overall beam.
For the stress in the layer A we obtain:
𝛼g 𝛼e 𝐸  ∆𝑇
𝜎g = 𝐸  ∆𝑇 − + =− 𝛼g − 𝛼e (8.35)
2 2 2
For the layer B:
∆𝑇
𝜎e = 𝐸 𝛼 + 𝛼e − 𝛼e ∆𝑇 (8.35)
2 g
that is:
𝛼g 𝛼e 𝐸  ∆𝑇
𝜎e = 𝐸  ∆𝑇 + − = 𝛼g − 𝛼e (8.36)
2 2 2
the value for stress will correspond to tension in the layer B and to compression in the layer A as
clear by intuition. In fact the part A, willing to expand more than part B, is constrained by B not
to expand, that is subjected to compression. The opposite is true for part B. In analogy to what
previously discussed for the bending part of the response of the bimetallic strip, it is easy to
understand that a similar effect could be obtained in presence of an actuating layer with an
induced strain Λ, say the B layer, and a passive layer, say the A layer. In that case the solution can
be obtained by assuming Λ = 𝛼e ∆𝑇 in B and Λ = 𝛼g ∆𝑇 = 0 in A, so that:
Λ
𝜀= (8.37)
2

8.7 Pure bending induced by patched actuators

Let us now consider the pure bending induced by a pair of patched actuators on a passive
substrate. In a 2D space let us consider a cantilever beam of unitary width, length L and thickness
2tp covered at the top and the bottom surfaces by two active layers each one of thickness ta. The
passive part is made of a linear elastic material being the Young modulus equal to EP . Let u(x)
and w(x) be the displacement components in x and y directions. Let us consider the case for which
the top active layer is activated with an induced active strain Λ = −λ and the bottom layer is
actuated in phase opposition with Λ = λ. In the hypothesis of a Euler Bernoulli behaviour of the
cantilever beam, the governing equation of the problem can be written as follows:
- Constitutive equations:
1
Active  layers            𝜀 = 𝜎 + Λ        or        𝜎 = 𝐸g 𝜀 − Λ
𝐸g
1
Passive  layer            𝜀 = 𝜎        or        𝜎 = 𝐸} 𝜀
𝐸}
- Equilibrium equations:
𝑑𝑇 𝑥
Translation  along  𝑦             =0
𝑑𝑥
𝑑𝑀 𝑥
Rotation  around  𝑧             = −𝑇 𝑥
𝑑𝑥
- Kinematic equation:
𝑑, 𝑣
Curvature            𝜒 =
𝑑𝑥 ,
And, with the hypothesis of plane section remaining plane and normal to the axis of the beam,
the displacement field reads:
𝑢 𝑥, 𝑦 = 𝑢 𝑥, 0 − 𝑦𝜃 𝑥
(8.38)
𝑣 𝑥, 𝑦 = 𝑣 𝑥, 0
‰Š h
with 𝜃 𝑥 = , so that in conclusion:
‰h

𝑑𝑣 𝑥, 0
𝑢 𝑥, 𝑦 = 𝑢 𝑥, 0 − 𝑦
𝑑𝑥
(8.39)
𝑣 𝑥, 𝑦 = 𝑣 𝑥, 0
Considering now the expression of the strain component along x we obtain, by labelling 𝑢 𝑥, 0
as 𝑢 𝑥 and 𝑣 𝑥, 0 as 𝑣 𝑥 :
𝑑𝑢 𝑥, 𝑦 𝑑𝑢 𝑥 𝑑, 𝑣 𝑥
𝜀= = − 𝑦 (8.40)
𝑑𝑥 𝑑𝑥 𝑑𝑥 ,
Since it is immediate to recognize that no axial deformation will be induced for the anti-
symmetric nature of the actuation, we can now restrict the analysis to the purely bending
component,
𝑑, 𝑣 𝑥
𝜀h = − 𝑦            or  else            𝜀h = −𝜒𝑦 (8.41)
𝑑𝑥 ,
Now we can add the boundary conditions for the bending problem
- for the clamped edge of the beam
𝑑𝑣 0
𝑣 0 = 0        and         =0
𝑑𝑥
- for the free edge of the beam
𝑀 𝐿 = 0        and        𝑇 𝐿 = 0
From the equilibrium equations it is easy to obtain that the shear force and the bending moment
are always zero along the x axis. In order to obtain the value for the curvature 𝜒 induced by the
actuation we may set the expression of the bending moment equal to zero.
d‹ Œd•
𝑀= d•
𝜎  𝑦  𝑑𝑦 =
Dd‹ Œ
,
d• (8.42)
Dd• d• Œd
, ‹
=− 𝐸Ž 𝜀 + 𝜆  𝑦  𝑑𝑦 − d•
𝐸•  𝜀  𝑦  𝑑𝑦 − d•
𝐸Ž   𝜀 − 𝜆 𝑦  𝑑𝑦
Dd‹ D
, ,

But since 𝜀h = −𝜒𝑦:


Dd• Dd• d•
𝑀=− 𝐸Ž −𝜒𝑦  𝑦  𝑑𝑦 − 𝐸Ž 𝜆  𝑦  𝑑𝑦 − 𝐸• −𝜒𝑦  𝑦  𝑑𝑦 +
Dd‹ Dd• Dd‹ Dd• Dd•
d• Œd‹ d• Œd‹
− 𝐸Ž −𝜒𝑦  𝑦  𝑑𝑦 − 𝐸Ž −𝜆  𝑦  𝑑𝑦 =
d• d•
Dd• Dd• d• (8.43)
, ,
= 𝐸Ž 𝜒  𝑦  𝑑𝑦 − 𝐸Ž 𝜆  𝑦  𝑑𝑦 − 𝐸• 𝜒  𝑦  𝑑𝑦 +
Dd‹ Dd• Dd‹ Dd• Dd•
d• Œd‹ d• Œd‹
+ 𝐸Ž 𝜒  𝑦 ,  𝑑𝑦 − 𝐸Ž 𝜆  𝑦  𝑑𝑦
d• d•

By calculating the integrals we obtain:


2 % 2
𝑀 = 𝐸Ž 𝜒 𝑡Ž + 3𝑡Ž, 𝑡• + 3𝑡Ž 𝑡•, − 𝐸Ž 𝜆 −𝑡Ž, − 2𝑡• 𝑡Ž − 𝐸• 𝜒 𝑡•% (8.44)
3 3
,
And since M = 0 and by dividing each term by 𝐸• 𝑡•% :
%
% , ,
𝐸Ž 𝑡Ž 𝑡Ž 𝑡Ž 𝐸Ž 3𝜆 𝑡Ž 𝑡Ž
𝜒 +3 +3 +1 = − −2 (8.45)
𝐸• 𝑡• 𝑡• 𝑡• 𝐸• 2𝑡• 𝑡• 𝑡•
d‹
so that setting 𝛼 = we finally obtain:
d•

𝐸Ž 3𝜆
−𝛼 , − 2𝛼
𝐸• 2𝑡•
𝜒= (8.46)
𝐸
1 + Ž 𝛼 % + 3𝛼 , + 3𝛼
𝐸•
For 𝐸Ž = 𝐸• :

and for 𝛼 very small (but not zero):


6𝛼 𝑡Ž
𝜒=− = −3𝜆 , (8.47)
2𝑡• 𝑡•

You might also like