1(2012) 1 – 25
1
Vibration attenuation and shape control of surface mounted,
embedded smart beam
Abstract
Active Vibration Control (AVC) using smart structure is
used to reduce the vibration of a system by automatic modification of the system structural response. AVC is widely used,
because of its wide and broad frequency response range, low
additional mass, high adaptability and good efficiency. A lot
of research has been done on Finite Element (FE) models
for AVC based on Euler Bernoulli Beam Theory (EBT). In
the present work Timoshenko Beam Theory (TBT) is used
to model a smart cantilever beam with surface mounted sensors / actuators. A Periodic Output Feedback (POF) Controller has been designed and applied to control the first
three modes of vibration of a flexible smart cantilever beam.
The difficulties encountered in the usage of surface mounted
piezoelectric patches in practical situations can be overcome
by the use of embedded shear sensors / actuators. A mathematical model of a smart cantilever beam with embedded
shear sensors and actuators is developed. A POF Controller
has been designed and applied to control of vibration of a
flexible smart cantilever beam and effect of actuator location on the performance of the controller is investigated.
The mathematical modeling and control of a Multiple Input multiple Output (MIMO) systems with two sensors and
two actuators have also been considered.
Keywords
AVC, FE, EBT, TBT, POF, MIMO, LTI, OLR, CLR
2
3
4
5
6
7
8
9
1
Vivek Rathi and Arshad Hussain
Khan
Mechanical Engineering Department, Aligarh
Muslim University, Aligarh India, 202001
Received 17 May 2012;
In revised form 24 May 2012
∗
Author email: logonrathi@yahoo.co.in, arshad1976@rediffmail.com
INTRODUCTION
Undesired noise and vibrations have always been a major problem in many human activities
and domains. From buildings to atomic force microscopes, all can be disturbed in their normal
functions by vibrations and noise. Recent technological advancements such as the availability
of high–power and low–cost computing, smart materials, and advanced control techniques
have led to a growing use of AVC systems. The implication of active control is that desirable
performance characteristics can be achieved through flexible and clever strategies, whereby
actuators excite the structure based on the structure’s response measured by sensors.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
2
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
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Umapathy and Bandyopadhyay[20] discussed the vibration control aspects of a smart flexible beam for a Single Input Single Output (SISO) case. Hanagud et al. [11] developed a FE
model for an active beam based on EBT and applied optimal output feedback control. Hawang
et al.[13] developed a FE model for vibration control of a laminated plate with piezoelectric
sensors /and actuators. Crawley et al [9] have presented the analytical and experimental
development of piezoelectric actuators as elements of intelligent structures. FE models of a
structure containing distributed piezoelectric sensors / actuators can also be seen in [10, 24].
Detailed survey on various control algorithms used in active vibration control studies has been
presented by Alkhatib and Golnaragi [4]. A detailed comparative studies of different control
algorithms on active vibration control of smart beam has been presented in [14, 21]. Kumar
and Narayanan [15] carried out optimal location studies of sensor-actuator pairs using Linear
Quadratic Regulator (LQR). Li et al. [16] proposed an optimal design methodology for the
placement of piezoelectric actuator/sensor pairs. Molter et al. [17] carried out control design
analysis for flexible manipulators using piezoelectric actuators. In their paper GA technique is
employed for optimization of placement and size of piezoelectric material for optimal vibration
control. Optimal controller design for the location, size and feedback of sensor/actuators have
been carried out in references [12, 23].
Chandrashekhara and Vardarajan [8] have presented a FE model of a piezoelectric composite beam using higher – order shear deformation theory. Aldraihem et al. [3] have developed a
laminated cantilever beam model using EBT and TBT with piezoelectric layers. Abramovich
[1] has presented analytical formulation and closed form solutions of composite beam with
piezoelectric actuators using TBT. Narayan and Balamurugan [18] have presented finite element formulation for the active vibration control study of smart beams, plates and shells
and the controlled response is obtained using classical and optimal control strategies. In the
analyses mentioned above, the controlled response has been obtained based on extension mode
actuation. There have been very few studies based on shear mode actuation and sensing for
the analysis of active structures.
The idea of exploiting the shear mode of creating transverse deflection in beams (sandwiched type) was first suggested by Sun and Zhang [19]. A FE approach was used by Benjeddou et al [6] to model a sandwich beam with shear and extension piezoelectric elements. It
was observed that the shear actuator is more efficient in rejecting vibration than the extension
actuator for the same control effort. Aldraihem and Khdeir [2] proposed analytical models and
exact solutions for beams with shear and extension piezoelectric actuators. The models are
based on TBT and HOBT. Exact solutions are obtained by using the state – space approach.
Azulay and Abramovich [5] studied the effects of actuator location and number of patches
on the actuator’s performance for various configurations of patches and boundary conditions
under mechanical and/or electrical loads.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
47
48
49
50
51
52
53
54
2
POF CONTROL
A standard result in control theory says that the poles of a linear time invariant (LTI) controllable system can be arbitrarily assigned by state feedback. If the original system is time
invariant and the linear combinations are also constrained to the time invariant, the design
problem is to choose an appropriate matrix of feedback gains. The problem of pole assignment
by piecewise constant output feedback with infrequent observation was studied by Chammas
and Leondes [7] for LTI systems.
Consider the system
ẋ = Ax(t) + Bu(t),
y(t) = Cx(t)
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
72
73
74
75
(1)
Where x ∈ Rn , u ∈ Rm , y ∈ Rp , A ∈ Rn×n , B ∈ Rn×m , C ∈ Rp×n
A, B, C are constants matrices and it is assumed that the system (A, B, C) is controllable, observable and stable. Assume that output measurements are available from system at
time instants t = kτ = 0, 1, 2,. . . . Now, construct a discrete LTI system from these output
measurements at rate 1/τ (sampling interval of τ seconds), the system so obtained is called
the τ system and is given by,
x(k + 1)τ = Φτ x(kτ ) + Γτ u(kτ )
y(kτ ) = Cx(kτ )
(2)
u(t) = Kl y(kt),
[kτ + l∆] ≤ t < [kτ + (l + 1)∆], Kl+N = Kl
(3)
Now, design an output injection gain matrix G such that Eigen values of (Φτ + GC) are
inside the unit circle i.e., eig(Φτ + GC)<1.
For l = 0, 1, - - - - - - - - N –1, where an output sampling interval τ is divided in to
N subintervals of width ∆ = τ /N , the hold function being assumed constant. To see the
relationship between the gain sequence {Kl } and closed loop behavior, let {Φ, Γ, C} be a new
system and denote the system sampled at rate 1/∆ as the ∆ system and collect the gain
matrices Kl in to one matrix. If (Φ, Γ) system is controllable and (Φτ , C) is observable, one
can first choose and output injection gain G to place the eigen values of (ΦN + GC) in the
desired locations inside the unit circle and then compute the POF gain sequence {Kl } such
that,
ΓK = G
71
3
(4)
andρ(ΦN + GC) < 1 is satisfied, where ρ is spectral radius.
Werner and Furuta [22] proposed the performance index so that ΓK = G need not be
forced exactly. This constraint is replaced by a penalty function, which makes it possible to
enhance the closed loop performance by allowing slight deviations from the original design and
at the same time improving behavior. The performance index is,
Latin American Journal of Solids and Structures 1(2012) 1 – 25
4
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
J(k) = ∑ [ xTl
∞
l=0
uTl ] [
∞
x
Q 0
T
] [ l ] + ∑ (xkN − x∗kN ) P (xkN − x∗kN )
ul
0 R
k=1
(5)
78
Where, R ∈ Rm×n , Q and P ∈ Rn×n , are positive definite and symmetric weight matrices.
The first term represents ’averaged’ state and control energy whereas the second term penalizes
deviation of G.
79
3
80
3.1
76
77
81
82
83
84
85
86
87
88
FORMULATION
Surface Mounted Sensors and Actuators
The smart cantilever model is developed using a piezoelectric beam element, which includes
sensor and actuator dynamics and a regular beam element based on TBT assumptions. The
piezoelectric beam element is used to model the regions where the piezoelectric patch is bonded
as sensor/actuator, and rest of the structure is modeled by the regular beam element.
The longitudinal axis of the regular beam element (Fig. 1), lies along the X – axis. The
element has constant moment of inertia, modulus of elasticity, mass density and length. The
element is assumed to have two degree of freedom, a transverse shear force and a bending
moment act at each nodal point.
θ
θ
θ
θ
Figure 1
89
A Regular Beam Element
Figure 2
Piezoelectric Beam Elements with
Sensors and Actuators
The displacement relation in the x, y and z direction can be written as,
u (x, y, z, t) = zθ(x, t) = z ( ∂w
− β (x)) ,
∂x
v (x, y, z, t) = 0,
w (x, y, z, t) = w (x, t)
90
91
92
(6)
Where,
w is the time dependent transverse displacement of the centroidal axis, θ is the time
dependent rotation of the cross – section about ‘Y – axis’.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
93
For the static case with no external force acting on the beam, the equation of motion is,
∂ [κGA ( ∂w
+ θ)]
∂x
∂x
∂θ
)
∂ (EI ∂x
94
95
96
5
− κGA (
∂x
The boundary conditions are given as,
At x = 0 w = w1 , θ = −θ1 and At
The mass matrix is given by,
= 0,
∂w
+ θ) = 0
∂x
x=L
w = w2 ,
(7)
θ = −θ2
[Nw ]
[Nw ]
ρA
0
][
] dx
[M ] = ∫ [
] [
[Nθ ]
[Nθ ]
0 ρIyy
(8)
[M ] = [MρA ] + [M ρI ]
(9)
L
T
0
97
98
99
[MρA ] in equation is associated with translational inertia and [MρI ] is associated with
rotary inertia, there expressions are given in the appendix.
The stiffness matrix is given by,
∂
EI
0
[Nθ ]
[Nθ ]
∂x
[K] = ∫ [
]
[
]
×
[
] dx
∂
∂
0 κGA
[Nw ]
[Nθ ] + ∂x [Nw ]
[Nθ ] + ∂x
L
T
∂
∂x
(10)
0
100
Finally we obtain;
⎡
⎢
⎢
EI
⎢
⎢
[K] =
3
(1 + ϕ) L ⎢
⎢
⎢
⎣
101
102
103
104
105
106
12
6L
−12
6L
6L (4 + ϕ) L2 −6L (2 − ϕ) L2
−12
−6L
12
−6L
6L (2 − ϕ) L2 −6L (4 + ϕ) L2
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
(11)
EI
),
Here ϕ is the ratio of the beam bending stiffness to the shear stiffness given by,ϕ = L122 ( κGA
L is the length of beam element. E is the Young’s modulus of the beam material, G is shear
modulus of the beam material, k is shear coefficient which depends on the material definition
and cross – sectional geometry, I is the moment of inertia of the beam element, A is the area
of cross – section of the beam element and ρ is the mass density of the beam material.
The consistent force array is given as,
[Nw ]
q
{F } = ∫ [
] {
} dx.
[Nθ ]
m
L
T
(12)
0
107
108
109
The piezoelectric element is obtained by sandwiching the regular beam element between two
thin piezoelectric layers as shown in figure 2. The element is assumed to have two – structural
degree of freedom at each nodal point and an electric degree of freedom. The piezoelectric
Latin American Journal of Solids and Structures 1(2012) 1 – 25
6
110
111
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
layers are modeled based on EBT as the effect of shear is negligible and the middle steel layer
is modeled based on TBT. The mass and stiffness matrix of piezoelectric layers is given by,
⎡ 156 22lp
54
−13lp ⎤
⎥
⎢
⎢
2
ρp Ap lp ⎢ 22lp
Ep Ip
4lp
13lp
−3lp2 ⎥
⎥ p
p
⎥ [K ] =
⎢
[M ] =
13lp 156 −22lp ⎥
420 ⎢
lp
⎥
⎢ 54
2 ⎥
⎢ −13lp −3l2 −22lp
4l
⎦
⎣
p
p
112
113
114
115
116
117
118
119
120
121
123
124
125
126
128
− 12
l2
4
− l6p
− l6p
2
p
12
2
lp
− l6p
⎤
⎥
⎥
2 ⎥
⎥
⎥
− l6p ⎥
⎥
⎥
4 ⎥
⎦
6
lp
(13)
ta is the thickness of actuator, tb is the thickness of beam, c is the width of beam.
The mass matrix for the piezoelectric beam element is given by,
[M piezo ] = [MρA ] + [MρI ] + [M p ]
(14)
[K piezo ] = [K] + [K p ]
(15)
Stiffness matrix [K piezo ] for the piezoelectric beam element,
3.1.1 Piezoelectric Strain Rate Sensors and Actuators
The linear piezoelectric coupling between the elastic field and the electric field can be expressed
by the direct and converse piezoelectric equations, respectively,
ε = sE σ + dEf
(16)
σ is the stress, ϵ is the strain, Ef is the electric field, e is the permittivity of the medium,
SE is the compliance of the medium, d is the piezoelectric constants.
Sensor Equation: If the poling is done along the thickness direction of the sensors with
the electrodes on the upper and lower surfaces, the electric displacement is given by,
Dz = d31 × Ep εx = e31 εx
127
6
lp
ρp is the mass density of piezoelectric beam element, Ap is the area of piezoelectric patch
= 2 ta c, lp (=L) is the length of the piezoelectric patch. Ep is the modulus elasticity of
piezoelectric material, Ip is the moment of inertia of piezoelectric layer w. r. t. the neutral
axis of the beam
2
(ta + tb )
1 3
]
Ip = cta + cta [
12
2
D = dσ + eT Ef
122
⎡ 122
⎢ lp
⎢ 6
⎢
⎢ lp
⎢ 12
⎢ − l2
⎢ p
⎢ 6
⎢ l
⎣ p
(17)
e 31 is the piezoelectric stress / charge constants, Ep is the Young’s modulus of piezoelectric
material, ϵx is the strain of the testing structure at a point.
The sensor output voltage is,
lp
V (t) = Gc e31 zc ∫ nTl q̇.dx
s
0
Latin American Journal of Solids and Structures 1(2012) 1 – 25
(18)
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
129
130
131
132
133
134
135
7
nTl is the second spatial derivative of the shape function of the flexible beam and as a scalar
vector product as,
V s (t) = pT q̇
(19)
V a (t) = KV s (t)
(20)
εA = d31 Ef
(21)
q̇ is the time derivative of the displacement vector, pT is a constant vector.
The input voltage to an actuator is V a (t) given by,
Actuator Equation: The strain developed by the electric field (Ef ) on the actuator layer
is given by,
The control force applied by the actuator is,
fctrl = Ep d31 cz
a
∫ n2 .dx.V (t).
(22)
lp
136
137
138
139
140
141
z = (ta 2+tb ) , is the distance between the neutral axis of the beam and the piezoelectric layer.
Or as a scalar product as,
fctrl = h.V a (t)
(23)
f t = fext + fctrl
(24)
nT2 is the first spatial derivative of shape function of the flexible beam, hT is a constant
vector.
If any external forces described by the vector fext are acting then, the total force vector
becomes,
142
3.1.2
143
The dynamic equation of motion of the smart structure is finally given by,
144
145
146
Dynamic Equation and State Space Model
M . q̈ + K . q = fext + fctrl
(25)
q = T.g ,
(26)
T is the model matrix containing the eigen vectors representing the desired number of
modes of vibration of the cantilever beam, g is the modal coordinate vector.
Equation (25) is then transformed in to,
∗
∗
M∗ .g̈ + C∗ .ġ + K∗ .g = fext
+ fctrl
(27)
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
8
147
148
149
M ∗ is the generalized mass matrix K∗ is the generalized stiffness matrix, C∗ is the generalized damping matrix, f∗ ext is the generalized external force vectors, f∗ ctrl is the generalized
control force vectors. The structural modal damping matrix is-:
C ∗ = α M ∗ + β K∗ ,
α and β are constants.
The state space model is
150
151
⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
152
153
ẋ1
ẋ2
ẋ3
ẋ4
⎡ x1
⎤
⎢
⎥
⎢ x
⎥
0
I
⎢
⎥
⎥=[
]⎢ 2
∗−1
∗
∗−1
∗
⎢ x3
⎥
−M
K
−M
C
⎢
⎥
⎢ x4
⎥
⎣
⎦
⎤
⎥
⎥
0
0
⎥
⎥+[
] u (t) + [ ∗−1 T ] r(t) (29)
∗−1 T
⎥
M
T
h
M
T f
⎥
⎥
⎦
u (t) is the control input, r (t) is the external input to the system, f is the external force
coefficient vector. The sensor equation for the modal state space form is given by;
y(t) = [ 0 pT
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
(28)
The above system may be represented as,
⎡
⎢
⎢
⎢
T ] ⎢
⎢
⎢
⎢
⎣
x1
x2
x3
x4
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
(30)
ẋ = A .x (t) + B . u (t) + E . r (t)
(31)
y (t) = CT x (t)
(32)
3.1.3 Validation for Surface Mounted Smart Beam
To validate the present formulation and the computer program, a cantilever beam made of steel
which is surface bonded with two PZT layers on both side is considered. The elastic modulus,
poisons ratio and density of steel and PZT are 200GPa, 0.3 and 7500 Kg/m3 and 139 GPa,
0.3 and 7500 Kg/m3 respectively while the strain and stress constants of PZT are 23×10−12
m/V and 0.216 respectively [8]. The length, width and the thickness of the beam are 500 mm,
30 mm and 2 mm respectively while the thickness of each of the PZT layers is 40 µm. The
Voltages at the steel and PZT layers are set to zero. The beam is discretized into 20 elements to
obtain converged results. The beam is excited with 0.2×10−3 Ns impulse load acting on the tip
of the beam. The closed loop response of the tip displacement is obtained using constant gain
negative velocity feedback (CGVF) control with gain Gv =1 and linear quadratic Regulator
(LQR) control Q=106 and R=1 and compared with the response obtained under the same
condition by the Narayanan and Balamurugan [8]. The control is applied after 0.5 seconds.
The present response of the system is very well matched with the published results. Next, the
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
169
170
171
first six open-loop and closed-loop natural frequencies of beam are presented in Table 1 and
compared with reference [18]. These frequencies are in good agreement with the published
results.
Table 1
Open loop natural
frequencies (Hz)
6.809
41.64
115.7
226.2
373.6
557.8
172
173
174
175
176
9
3.2
First six natural frequencies of smart steel cantilever beam
Closed loop natural
frequencies (Hz)
(CGVF control)
6.855
43.19
121.6
237.5
357.5
506.3
Closed loop natural
frequencies (Hz)
(LQR control)
6.834
41.69
115.7
226.3
373.7
557.9
Natural
frequencies (HZ)
[8]
6.89
43.285
121.225
237.72
393.58
589.63
Figure 3
Closed loop response of smart cantilever steel beam (a) Narayanan and Balamurugan [18](reproduced
with permission from Elsevier) and (b) Present obtained with negative CGVF control with Gv =1.
Figure 4
Closed loop response of smart cantilever steel beam (a) Narayanan and Balamurugan [18] (reproduced
with permission from Elsevier) and (b) Present obtained with LQR control with Q = 106 and R=1.
Embedded Shear Sensors And Actuators
The piezoelectric element is embedded on discrete locations of the sandwich beam as shown in
Figure (5). The smart cantilever beam model is developed using a piezoelectric sandwich beam
element, which includes sensor and actuator and a regular sandwiched beam element, which
includes foam at the core. A FE model of a piezoelectric sandwich beam is developed using
Latin American Journal of Solids and Structures 1(2012) 1 – 25
10
177
178
179
180
181
182
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
laminate beam theory. It consists of three layers. The assumption made is that the middle
layer is perfectly glued to the carrying structure and the thickness of adhesive can be neglected
and each layer behaves as a Timoshenko beam. The longitudinal axis of the sandwiched beam
element lies along the X – axis. The element has constant moment of inertia, modulus of
elasticity, mass density and length. The element is assumed to have three degree of freedom,
a transverse shear force and a bending moment act at each nodal point.
Figure 5
183
A Sandwiched Beam Element
The displacement relation of the beam u (x, z ) and wu (x, z ) can be written as,
u (x, z) = u0 (x) − zθ (x, t)
184
185
186
187
188
w(x, z) = w0 (x)
(33)
u0 (x) and w0 (x) are the axial displacements of the point at the mid plane,θ (x) is the
bending rotation of the normal to the mid plane.
The beam constitutive equation can be written as,
⎡ ∂u0 ⎤ ⎡ E11 ⎤
⎡ Nx ⎤ ⎡ A11 B11
0 ⎤
⎥⎢
⎢
⎥ ⎢
⎥
⎥ ⎢
dx
⎥⎢
⎢
⎥ ⎢
⎥
⎥ ⎢
∂θ
⎢ Mx ⎥ = ⎢ B11 D11
⎥ + ⎢ F11 ⎥
0 ⎥⎢
(34)
∂x
⎢
⎥ ⎢
⎢
⎥
⎥
⎥⎢
∂w0 ⎥ ⎢
⎢ Qxz ⎥ ⎢ 0
⎥
⎢
⎥
0 A55 ⎦ ⎣ θ + ∂x ⎦ ⎣ G55 ⎦
⎣
⎦ ⎣
A11 , B11 , D11 and A55 are the extensional, bending and shear stiffness coefficients defined
according to the lamination theory,
A11 = c ∑ (Q11 )k (zk − zk−1 ) ,
N
k=1
N
2
),
B11 = c ∑ (Q11 )k (zk2 − zk−1
D11 =
c
3
k=1
N
3
)
∑ (Q11 )k (zk3 − zk−1
k=1
N
A55 = cκ ∑ (Q55 )k (zk − zk−1 ) ,
k=1
Latin American Journal of Solids and Structures 1(2012) 1 – 25
(35)
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
189
190
191
192
193
Zk is the distance of the k th layer from the X – axis, N is the number of layers, k is the shear
correction factor usually taken equal to 5/6.
The boundary conditions are given as,
At x = 0 w = w1 , θ = θ1 ,
u = u1
and At
x=L
w = w2 , θ = θ 2 ,
u = u2
After solving, we get,
⎧
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
[u] = [Nu ] ⎨
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎩
u1
w1
θ1
u2
w2
θ2
⎫
⎪
⎪
⎪
⎪
⎧
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎬ , [w] = [Nw ] ⎨
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎩
⎪
⎪
⎪
⎪
⎭
w1
θ1
w2
θ2
⎧
⎫
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎬ , θ = [Nθ ] ⎨
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎪
⎭
⎩
w1
θ1
w2
θ2
⎫
⎪
⎪
⎪
⎪
⎪
⎪
⎬,
⎪
⎪
⎪
⎪
⎪
⎪
⎭
(36)
[Nu ] = [ N1 N2 N3 N4 N5 N6 ]
[NW ] = [ N7 N8 N9 N10 ]
[Nθ ] = [ N11 N12 N13 N14 ]
195
⎡
⎢
⎢
⎢
⎢
⎢
[M ] = ⎢
⎢
⎢
⎢
⎢
⎢
⎣
197
198
199
M11
M21
M31
M41
M51
M61
M12
M22
M32
M42
M52
M62
M13
M23
M33
M43
M53
M63
M14
M24
M34
M44
M54
M64
M15
M25
M35
M45
M55
M65
M16
M26
M36
M46
M56
M66
⎤
⎡
⎥
⎢
⎥
⎢
⎢
⎥
⎥
⎢
⎥
⎢
⎥ [K] = ⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎢
⎥
⎥
⎢
⎦
⎣
K11
K21
K31
K41
K51
K61
K12
K22
K32
K42
K52
K62
K13
K23
K33
K43
K53
K63
K14
K24
K34
K44
K54
K64
K15
K25
K35
K45
K55
K65
K16
K26
K36
K46
K56
K66
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
(38)
Values of the mass and stiffness matrix coefficients are given in the appendix.
The material constants Q11 , Q22 , Q12 , Q66 , G13 and G23 for foam, steel and piezoelectric
materials are given in table 2. These constants are used to calculate the values of A11 , B11 ,
D11 and A55
Table 2
Material Constants
G12 (MPa)
G13 (MPa)
G23 (MPa)
d31 (m/V)
d15 (m/V)
Q11 (MPa)
Q22 (MPa)
Q12 (MPa)
Q66 (MPa)
Material Properties and Constants
Piezoelectric Material
24800
24800
24800
-16.6×10−9
1.34×10−9
68400
68400
12600
12600
Steel
78700
78700
78700
#
#
215000
215000
2880
78700
Foam
99.9
99.9
99.9
#
#
85.4
85.4
75.8
9.99
Sensor Equation: The charge q(t) accumulated on the piezoelectric electrodes is given
200
201
(37)
Values of N1 to N12 be given in the appendix,
The symmetric mass and stiffness matrices are given by,
194
196
11
by,
Latin American Journal of Solids and Structures 1(2012) 1 – 25
12
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
q (t) = ∬ D3 dA
(39)
A
202
203
204
205
206
207
208
D3 is the electric displacement in the thickness direction, A is the area of electrodes.
The current induced in sensor layer is converted in to the open circuit sensor voltage
V s (t) using a signal-conditioning device with a gain of Gc and applied to the actuator with
the controller gain Kc .
T
6η
2 −l 0 −2 −l ] [q̇] ,
V s (t) = [p] [q̇]
V s (t) = Gc i (t) , V s (t) = e15 c (−12η+l
2) [ 0
[q̇] is the time derivative of the modal coordinate vector, [p] is a constant vector.
The input voltage of the actuator is V a (t), given by,
T
V a (t) = Kc V s (t)
209
210
(40)
Actuator Equation: The strain produced in the piezoelectric layer is directly proportional to the electric potential applied to the layer.
γxz ∝ Ef
211
212
213
γxz is the shear strain in the piezoelectric layer, Ef is the electric potential applied to the
actuator.
From the constitutive piezoelectric equation, we get,
(41)
γxz = d15 Ef
214
The shear force Qxz is given as
V a (t)
= cGd15
∫ dz
tp
tp
Qxz
(42)
0
215
Or as a scalar product,
fctrl = h . V a (t)
216
217
218
(43)
Where,fctrl = Qxz and hT is a constant vector.
If any external forces described by the vector fctrl are acting, then total force vector becomes,
f t = fext + fctrl
Latin American Journal of Solids and Structures 1(2012) 1 – 25
(44)
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
219
3.2.1
220
The dynamic equation is,
Dynamic Equation and State Space Model:
∗
∗
M∗ .g̈ + C∗ .ġ + K∗ .g = fext
+ fctrl
(45)
C ∗ = α M∗ + β K ∗ ,
(46)
The state space form of the system is obtained as,
221
⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
ẋ1
ẋ2
ẋ3
ẋ4
ẋ5
ẋ6
⎤
⎡ x1
⎥
⎢
⎥
⎢ x
⎥
⎢ 2
⎥
⎢
⎥
⎢ x3
0
I
⎥=[
⎢
⎥
∗−1 ∗
∗−1 ∗ ] ⎢
−M K −M C
⎥
⎢ x4
⎥
⎢
⎥
⎢ x5
⎥
⎢
⎥
⎢ x6
⎦
⎣
(48)
ẋ = A .x (t) + B . u (t) + E . r (t)
(49)
y (t) = CT x (t)
(50)
The above system can be represented as,
223
224
4
225
4.1
228
229
230
231
232
⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
y (t) = [ 0 pT
226
⎤
⎥
⎥
⎥
⎥
⎥
0
0
⎥+[
] u(t) + [
] r(t) (47)
∗−1 T
∗−1 T
⎥
M
T
h
M
T f
⎥
⎥
⎥
⎥
⎥
⎦
The sensor equation for the modal state space form is given by;
222
227
13
T ]
x1
x2
x3
x4
x5
x6
SIMULATION
Simulation of Surface Mounted Sensors and Actuators
The state space representation of the cantilever beam with the surface mounted sensor /
actuator is obtained by using nine regular beam elements and one piezoelectric element as
shown figure (6).
The dimensions and properties of the flexible beam and piezoelectric sensor / actuator used
in the numerical simulation are given in tables 3 and 4 respectively.
Two state space models of the smart cantilever beam have been obtained by keeping the
AR 8 and 15, the length of the beam is kept constant and the thickness of the beam is varied.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
14
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 6
Cantilever Beam with Surface Mounted Sensors and Actuators
Table 3
Physical Properties of Steel Beam
Total length of beam (mm)
Width (mm)
Young’s modulus (Pa)
Density Kg/m3 )
Constants used in C∗
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
l
b
Eb
ρb
α, β
500
24
193x109
8030
10−3 ,10−4
The POF control technique is used to design a controller to suppress vibration of a cantilever
beam. For this purpose three vibration modes are considered. In the first case the AR is
taken equal to 15. Configuration specifications of smart beam are as per table 3 and 4. The
collocated sensor and actuator are placed near the fixed end. The FE model of the surface
mounted cantilever beam is developed in MATLAB using TBT. A sixth order space model
of the system is obtained on retaining the first three modes of vibration of the system. The
first three natural frequencies calculated are 106.03 Hz,658.33Hz, 1826.38Hz respectively. An
impulsive force of 10N is applied for duration of 0.05 sec and the open loop response (OLR) of
the system is obtained as shown in figure 7. A controller based on the POF control algorithm
has been designed to control the first three modes of vibration of the smart cantilever beam.
The sampling interval used is 0.0004 sec. The sampling interval is divided in to 10 subintervals
(N=10). The periodic output gain for the system is obtained by using the algorithm given for
POF controller, the impulse response of the system with POF gain is shown in figure 8.
In the second case the AR is taken equal to 8, and again all other parameters are kept
same as that in the first case for which AR is 15. The first three natural frequencies calculated
are 197.94Hz, 1215.89Hz and 3311.05Hz respectively. The OLR and CLR (with POF gain) of
the system is obtained as shown in figure 9 and 10.
Table 4
Properties of Piezoelectric Sensor /Actuator
Width (mm)
Thickness (mm)
Young’s Modulus (Pa)
Density(Kg/m3 )
Piezoelectric strain Constant (mV−1 )
Piezoelectric stress Constant (VmN−1 )
Latin American Journal of Solids and Structures 1(2012) 1 – 25
b
ta
Ep
ρp
d31
e31
24
0.5
68x109
7700
125x10-12
10.5x10−3
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 7
Figure 9
250
251
252
253
254
255
256
257
258
OLR of Surface Mounted Cantilever Beam with AR = 15
OLR of Surface Mounted Cantilever
Beam with AR = 8
Figure 8
Figure 10
15
CLR of Surface Mounted Cantilever Beam with AR = 15
CLR of Surface Mounted Cantilever
Beam with AR = 8
The variation of control signal with time is shown in figure 11 (for AR = 15) and 12(for
AR = 8).
A POF controller is designed for the Timoshenko beam models. Two cases (AR=15,8 )
have been considered. By comparing the OLR and CLR in the first case for AR=15, we
observed that there was a 89.43% decrease in settling time for the system after applying POF
control, and a change of 89% in settling time was observed for the case with AR=8. Thus, it
can be inferred from the simulation results, that a POF controller applied to a smart cantilever
model based on TBT is able to satisfactorily control higher modes of vibration of the smart
cantilever beam for a wide range of AR.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
16
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 11
Control Signal of Surface Mounted
Cantilever Beam with AR = 15
259
4.2
260
4.2.1 Single Input Single Output (SISO) System
261
262
263
264
265
266
267
268
269
270
271
272
273
274
276
277
278
279
CLR of Surface Mounted Cantilever
Beam with AR = 8
Simulation of Embedded Sensors and Actuators
The FE model of smart cantilever beam based on laminate beam theory is developed. Keeping
the sensor location fixed and varying the position of the actuator, different state space models
of the smart cantilever beam are obtained. A POF controller is designed to control the first
three modes of vibration of the smart cantilever beam. Here an attempt has been made to find
the optimum actuator position for a single input single output (SISO) system. Three cases
have been considered.
In the first case FE model of the smart cantilever beam is obtained by dividing the beam
into 10 elements. The actuator is placed as the 1st element (at the fixed end) and the sensor
is placed as the 8th element as shown in figure 13. The length of beam is 200mm and its cross
– section is 10mm × 20mm. The length of piezoelectric patch is 200mm and its cross – section
is 6mm × 20mm. The material properties used for the generation of FE model are given in
table 2. A ninth order space model of the system is obtained on retaining the first three modes
of vibration of the system. The first three natural frequencies (same for all three models) are
44.9 Hz, 82.4 Hz and 131.5 Hz respectively.
Figure 13
275
Figure 12
Smart Cantilever Beam with Actuator at 1st Position and Sensor at 8th Position
An impulsive force of 10N is applied for duration of 0.05sec and the OLR of the system
is obtained as shown in figure 14 A controller based on the POF control algorithm has been
designed to control the first three modes of vibration of the smart cantilever beam. The
sampling interval used is 0.07 m sec. The sampling interval is divided in to 10 subintervals (N
=10). The impulse response or CLR of the system with POF gain is shown in figure 15.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 14
280
281
OLR of Smart Cantilever Beam with
Actuator at 1st Position
283
285
286
287
Smart Cantilever Beam with Actuator at 5th Position and Sensor at 8th Position
An impulsive force of 10 N is applied for duration of 0.05 sec and the OLR of the system is
obtained as shown in figure 17. The CLR of the system with POF gain is shown in figure 18.
Figure 17
284
CLR of Smart Cantilever Beam with
Actuator at 1st Position
In the second case, the actuator is placed as the 5th element and the sensor is placed as
the 8th element as shown in figure 17. Other parameters are kept same.
Figure 16
282
Figure 15
17
OLR of Smart Cantilever Beam with
Actuator at 5th Position
Figure 18
CLR of Smart Cantilever Beam with
Actuator at 5th Position
In the third case, the actuator is placed as the 10th element (at the free end) and the sensor
is placed as the 8th element as shown in figure 19. Other parameters are kept same as that of
first case.
An impulsive force of 10 N is applied for duration of 0.05 sec and the OLR of the system is
Latin American Journal of Solids and Structures 1(2012) 1 – 25
18
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 19
288
289
obtained as shown in figure 20. The impulse response of the system with POF gain is shown
in figure 21.
Figure 20
293
294
295
Figure 21
CLR of Smart Cantilever Beam
with Actuator at 10th Position
24.
Figure 22
292
OLR of Smart Cantilever Beam
with Actuator at 10th Position
The variation of control signal with time for all the three cases are shown in figure 22 to
290
291
Smart Cantilever Beam with Actuator at 1st Position and Sensor at 8th Position
Control Signal of Smart Cantilever
Beam with Actuator Placed at 1st
Position
Figure 23
Control Signal of Smart Cantilever
Beam with Actuator at 5th Position
Here in the present case, the performance of the controller is evaluated for different actuator
locations while the position of the sensor is kept constant. It can be inferred from the response
characteristics that the actuator locations has negligible effect on the performance of the
controller.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 24
296
297
298
299
300
301
302
303
304
305
4.2.2
307
308
309
310
Control Signal of Smart Cantilever Beam with Actuator at 10th Position
Multi – Input Multi – Output (MIMO) Systems
Active control of vibration of a smart cantilever beam through smart structure concept for a
multivariable system (MIMO) case is considered here. The structure is modeled in space state
from using Finite element method by dividing the beam in to 10 FE and placing the sensors
at the 6th and 10th positions and the actuators at the 4th and 8th position. Thus giving rise
to MIMO with two actuator inputs u 1 and u 2 and two sensors outputs y 1 and y 2 , The POF
control technique is used to design a controller to suppress the first three modes of vibration
of a smart cantilever beam for a multi variable system. The simulations are carried out in
MATLAB. The parameters are kept same as that of the model used for SISO case. The first
three natural frequencies calculated are 45.2 Hz, 83.2 Hz and 136.4 Hz respectively.
Figure 25
306
19
A MIMO Smart Cantilever Beam with Two Inputs and Two Outputs
An impulsive force of 10N is applied for duration of 0.05sec. A controller based on the
POF control algorithm has been designed to control the first three modes vibration of the
smart cantilever beam for the multivariable case. The CLR (sensor outputs y1 and y2 ) with
periodic output feedback gain K for the state space model of the system is shown in figure 27
and 31. Figures 29 and 33 show the variation of the control signal with time.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
20
311
312
313
314
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
Figure 26
CLR of SISO System with Sensor
at 6th Position
Figure 27
Response y1 of MIMO System
with POF
Figure 28
Control Input of SISO System
with Sensor at 6th Position and
Actuator at 4th Position
Figure 29
Control Input u 1 of MIMO System with POF
It can be inferred from the simulation results, that the system’s performance meets the
design requirements. It is also observed that the maximum amplitude of the sensor output
voltage is less for the multivariable case and the response takes lesser time to settle. Controlling
time is considerably reduced with MIMO systems as compared to SISO systems.
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
5
Figure 30
30 CLR of SISO System with Sensor at 10th Position and Actuator
at 8th Position
Figure 31
Response y2 of MIMO System
with POF
Figure 32
Control Input of SISO System
with Sensor at 10th Position and
Actuator at 8th Position
Figure 33
Control Input u 2 of MIMO System with POF
21
CONCLUSION:
An integrated FE model to analyze the vibration suppression capability of a smart cantilever
beam with surface mounted piezoelectric devices based on TBT is developed. In practical
situations a large number of modes of vibrations contribute to the structures response. In this
work a FE model of a smart cantilever beam have been obtained by varying the AR from
8 to 15, the length of the beam is kept constants and the thickness of the beam is varied.
POF control technique is used to design a controller to suppress the vibration of the smart
cantilever beam by considering three modes of vibration. Two different cases have been considered (AR=8,15). The simulation results show that the POF controller based on TBT is
able to satisfactorily control the first three modes of vibration of the smart cantilever beam
for different AR. Surface mounted piezoelectric sensors and actuators are usually placed at
the extreme thickness positions of the structure to achieve most effective sensing and actuation. This subjects the sensors / actuators to high longitudinal stresses that might damage
the piezoceramic material. Furthermore, surface mounted sensors/ actuators are likely to be
damaged by contact with surrounding objects. Embedded shear sensors / actuators can be
used to alleviate these problems. A FE model of a smart cantilever beam with embedded
Latin American Journal of Solids and Structures 1(2012) 1 – 25
22
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
piezoelectric shear sensors / actuators based on laminate theory is developed.
A POF controller is designed to control the vibration of the system. The performance of
the controller is evaluated for different actuator locations while the position of sensor is kept
constant. It was observed from the simulation results that the location of the actuator has
negligible effect on the performance of the controller. A MIMO system with two sensors and
two actuators has also been considered. A POF controller has been designed for the MIMO
smart structure model to control the vibration of the system by considering the three modes
of vibration. The beam with embedded shear sensors / actuators has been divided in to 10 FE
with the sensors placed at the 6th and 10th positions and the actuators placed at the 4th and
8th positions. It can be inferred from the simulation results, that when the system is placed
with the controller, the system’s performance meets the design requirements. It is observed
that the maximum amplitude of the sensor output voltage is less for the multivariable case
and the response takes lesser time to settle.
APPENDIX
1. Translational mass matrix:
⎡ [70φ2 + 147φ + 78]
⎢
⎢
⎢ [35φ2 + 77φ + 44] L
ρI
⎢
4
[MρA ] =
⎢
2
[35φ2 + 63φ + 27]
210(1 + φ) ⎢
⎢
⎢
⎢ − [35φ2 + 63φ + 26] L
⎣
4
L
4
L2
+ 14φ + 8] 4
[35φ2 + 63φ + 26] L
4
2
− [7φ2 + 14φ + 6] L4
[35φ2 + 77φ + 44]
[7φ2
[35φ2 + 63φ + 27]
+ 63φ + 26]
[70φ2 + 147φ + 78]
− [35φ2 + 77φ + 44]
L
4
L2
+ 14φ + 6] 4
− [35φ2 + 77φ + 44] L
4
2
[7φ2 + 14φ + 8] L4
− [35φ2 + 63φ + 26]
L
4
[35φ2
L
4
− [7φ2
346
347
2. Rotational Mass matrix
⎡
36
− (15ϕ − 3) L
−36
− (15ϕ − 3) L
⎢
⎢
2
2
(10ϕ + 5ϕ + 4) L (15ϕ − 3) L (5ϕ2 − 5ϕ − 1) L2
− (15ϕ − 3) L
⎢
[MρI ] = ⎢
2
2
⎢ (10ϕ + 5ϕ + 4) L
(15ϕ − 3) L
36
(15ϕ − 3) L
⎢
2
2
⎢
(5ϕ − 5ϕ − 1) L
− (15ϕ − 3) L
(15ϕ − 3) L (10ϕ2 + 5ϕ + 4) L2
⎣
348
3. The mass matrix for the sandwich beam element ,
⎡
⎢
⎢
⎢
⎢
⎢
[M ] = ⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
349
M11
M21
M31
M41
M51
M61
M12
M22
M32
M42
M52
M62
M13
M23
M33
M43
M53
M63
M14
M24
M34
M44
M54
M64
M15
M25
M35
M45
M55
M65
M16
M26
M36
M46
M56
M66
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
Where,
1 γ L 3 I1
1
1
1 γ L 2 I1
,
M
=
M
=
−
, M14 = M41 = LI1 ,
M11 = LI1 , M12 = M21 =
13
31
2
2
3
2 (12η−L )
4 (12η−L )
6
Latin American Journal of Solids and Structures 1(2012) 1 – 25
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
23
350
γ L I1
1 γ L I1
M15 = M51 = − 12 (12η−L
2 ) , M16 = M61 = − 4 (12η−L2 ) , M24 = M42 =
2
2
1 γ L I1
,
2 (12η−L2 )
M22 =
,
⎛
L⎜
1 ⎝
35
3
⎞
⎟
⎠
−294I3 ηL2 + 35I2 L3 + 1680I2 η 2 + 13I3 L4 − 420I2 ηL
+ 42I1 L2 + 42γ 2 I1 L2
(12η−L2 )
⎛ 11I3 L5 − 10080I2 η 2 + 1260I3 η 2 L − 231I3 L3 η + 126γ 2 I1 L3 ⎞
⎟
L⎜
⎠
⎝
+ 1260LI1 η + 840I2 ηL2 + 21I1 L3
1
,
2
2
210
(12η−L )
4
2
2
2
2
2
3 L(3I3 L −28γ I1 L −28I1 L −84I3 ηL +560I3 η )
,
2
70
(12η−L2 )
5
2
2
⎛ 13I3 L + 10080I2 η + 2520I3 η L − 378I3 L3 η + 252γ 2 I1 L3 ⎞
⎟
L⎜
⎠
⎝
− 2520LI1 η − 840I2 ηL2 − 42I1 L3
1
2
420
(12η−L2 )
2
M23 = M32 =
M25 = M52 =
M26 = M62 =
351
M55 =
⎛
L⎜
1 ⎝
35
13I3 L4 − 35I2 L3 + 42γ 2 I1 L2 + 42I1 L2 − 294I3 ηL2
+ 420I2 ηL + 1680I3 η 2
352
⎛
L⎜
⎝
1
210
⎛
L⎜
⎝
1
210
11I3 L
252I3 η
5
,
(12η−L2 )
+ 10080I2 η 2 + 1260I3 η 2 L − 231I3 L3 η + 126γ 2 I1 L3 ⎞
⎟
⎠
+ 1260LI1 η − 840I2 ηL2 + 21I1 L3
2
M56 = M65 =
M66 =
⎞
⎟
⎠
(12η−L2 )
L − 42I3 L η + 10080I1 η 2 + 63γ 2 I1 L4 + 28I1 L4 ⎞
⎟
− 420I1 ηL2 + 2I3 L6 + 2520I2 η 2 L − 210I2 ηL3 ⎠
2
2
2
,
4
(12η−L2 )
2
.
1. The stiffness matrix for the sandwich beam element is
⎡
⎢
⎢
⎢
⎢
⎢
[K] = ⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
353
K11 =
AA11
,
L
K12
K22
K32
K42
K52
K62
K15
K25
K35
K45
K55
K65
K16
K26
K36
K46
K56
K66
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
11
= K51 = − AB
,
L
4
2
2
2
2
2
6 A (D11 L +10A55 L +10γ A11 L −20η D11 L +120 D11 η )
,
2
2
5L
(12η−L )
(D11 L4 +10A55 L2 +10γ 2 A11 L2 −20η D11 L2 +120 D11 η 2 )
(12η−L2 )
1 AL(−D11 L −10B11 γ L −60A55 L−60γ A11 L+120γ B11 η)
,
K62 = 10
2
(12η−L2 )
⎛ 2L6 D11 − 30L4 D11 η + 180L2 D11 η 2 − 15L5 γ B11 + 2160A55 η 2 ⎞
⎟
A⎜
⎝
⎠
+ 60A55 L4 − 360A55 ηL2 + 180γL3 B11 η + 45γ 2 L4 A11
,
15L(12η−L2 )2
3
K33 =
K14
K24
K34
K44
K54
K64
AB11
11
, K13 = K31 = 0, K14 = K41 = − AA
, K15
L
L
3
2
2
AL(D
L
−10B
γ
L
+60A
L+60η
A
L+120γ
B
11
11
55
11
11 η)
1
,
− 10
2
(12η−L2 )
11
, K25 = K52 = − 65 A
K24 = K42 = − AB
L
L
K26 =
K13
K23
K33
K43
K53
K63
K12 = K21 =
K16 = K61 = 0, K22 =
K23 = K32 =
K11
K21
K31
K41
K51
K61
2
2
,
2
K34 = K43 = 0,
Latin American Journal of Solids and Structures 1(2012) 1 – 25
24
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
354
3
2
2
1 AL(D11 L −10B11 γ L +60A55 L+60γ A11 L+120γ B11 η)
11
, K44 = AA
,
2
2
10
L
(12η−L )
11
, K46 = K64 = 0,
K54 = AB
L
⎛ L6 D11 − 60L4 A55 − 90L4 A11 γ 2 − 60L4 η D11 + 360D11 L2 η ⎞
⎟
−A⎜
⎠
⎝
+ 4320A55 η 2 − 720A55 ηL2
K63 =
,
30L(12η−L2 )2
4
2
2
2
2
2
)
(D
L
+10A
L
+10γ
A
L
−20η
D
L
+120
D
η
11
55
11
11
11
6A
,
2
5L
(12η−L2 )
K35 = K53 =
K45 =
K36 =
K55 =
1
K56 = K65 = − 10
K66 =
355
AL(−D11 L3 −10B11 γ L2 −60A55 L−60γ 2 A11 L+120γ B11 η)
,
(12η−L2 )
6
5
⎛ 2L D11 + 15L B11 γ − 30L4 D11 η + 2160A55 η 2 + 180L2 D11 η 2 ⎞
⎟
A⎜
⎝
⎠
+ 60A55 L4 − 360A55 ηL2 − 180γL3 B11 η + 45γ 2 L4 A11
.
15L(12η−L2 )2
2
References
356
357
[1] H. Abramovich. Deflection control of laminated composite beams with piezoceramic layers - close form solution.
Composite Structures, 43(3):2217–231, 1998.
358
359
[2] O. J. Aldraihem and A. A. Khdeir. Smart beams with extension and thickness-shear piezoelectric actuators. Journal
Of Smart Materials And Structures, 9(1):1–9, 2000.
360
361
[3] O.J. Aldraihem, R.C. Wetherhold, and T. Singh. Distributed control of laminated beams: Timoshenko vs. euler bernoulli theory. Journal of Intelligent Materials System and Structures, 8:149–157, 1997.
362
363
[4] R. Alkhatib and M.F. Golnaraghi. Active structural vibration control: a review. Shock and Vibration Digest,
35(5):367–383, 2003.
364
365
[5] L.E. Azulay and H. Abramovich. piezoelectric actuation and sensing mechanisms - closed form solution. Composite
Structures, 64(3):443–453.
366
367
[6] A. Benjeddou, M.A. Trindade, and R. Ohayon. New shear actuated smart structure beam finite element. AIAA
Journal, 37:378–383, 1999.
368
369
[7] A.B. Chammas and C.T. Leondes. Pole placement by piecewise constants output feedback. International Journal
Of Control, 29(1):31–38, 1979.
370
371
[8] K. Chanrashekhara and S. Vardarajan. Adaptive shape control of composite beams with piezoelectric actuators.
Journal of Intelligent Material Systems and Structures, 8:112–124, 1997.
372
373
[9] E.F. Crawley and J. De Luis. Use of piezoelectric actuators as elements of intelligent structures. AIAA Journal,
25:1373–1385, 1987.
374
375
[10] C. Doschner and M. Enzmann. On model based controller design for smart structure. Smart Mechanical Systems
Adaptronics International, pages 157–166.
376
377
[11] S. Hanagud, M.W. Obal, and A.J. Callise. Optimal vibration control by the use of piezoelectric sensors and actuators.
Journal Of Guidance, Control and Dynamics, 15(5).
378
379
[12] Q. Hu and G. Ma. Variable structure maneuvering control and vibration suppression for flexible spacecraft subject
to input nonlinearities. Smart Materials and Structures, 15(6):1899–1911, 2006.
380
381
[13] W. Hwang and H.C. Park. Finite element modeling of piezoelectric sensors and actuators. AIAA Journal, 31(5):930–
937, 1993.
382
383
[14] S. Kapuria and M.Y. Yasin. Active vibration control of piezoelectric laminated beams with electroded actuators and
sensors using an efficient finite element involving an electric node. Smart Materials and Structures, 19, 2010. 045019.
384
385
[15] K.R. Kumar and S. Narayanan. Active vibration control of beams with optimal placement of piezoelectric sensor/actuator pairs. Smart Mater. Struct, 17, 2008. 055008 (15pp).
Latin American Journal of Solids and Structures 1(2012) 1 – 25
Vivek Rathi et al / Vibration attenuation and shape control of surface mounted, embedded smart beam
25
386
387
[16] Y. Li, J. Onoda, and K. Minesugi. Simultaneous optimization of piezoelectric actuator placement and feedback for
vibration suppression. Acta Astronautica, 50.
388
389
390
[17] A. Molter, A. Otavio Alves da S., S. Jun Ono Fonseca, and V. Bottega. Simultaneous piezoelectric actuator and
sensor optimization and control design of manipulators with flexible links using sdre method. Mathematical Problems
in Engineering, 2010:23. Article ID 362437.
391
392
[18] S. Narayanana and V. Balamurugan. Finite element modelling of piezolaminated smart structures for active vibration
control with distributed sensors and actuators. Journal of Sound and Vibration, 262:529–562, 2003.
393
394
[19] C.T. Sun and X.D. Zhang. Use of thickness shear mode in adaptive sandwich structures. Smart Materials and
Structures, 4(3):205–207, 1995.
395
396
[20] M. Umapathy and B. Bandyopadhyay. Control of flexible beam through smart structure concept using periodic
output. System Science, 26.
397
398
[21] C.M.A. Vasques and J.D. Rodrigues. Active vibration control of smart piezoelectric beams: Comparison of classical
and optimal feedback control strategies. Computers and Structures, 84:1459–1470, 2006.
399
[22] H. Werner and K. Furuta. Simultaneous stabilization based on output measurements. Kybernrtca, 31(2):395–411.
400
401
[23] S. X. Xu and T.S. Koko. Finite element analysis and design of actively controlled piezoelectric smart structures.
Finite Element in Analysis and Design, 40(3):241–262, 2004.
402
403
[24] M.Y. Yasin, N. Ahmad, and N. Alam. Finite element analysis of actively controlled smart plate with patched
actuators and sensors. Latin American Journal of Solids and Structures, 7:227–247, 2010.
Latin American Journal of Solids and Structures 1(2012) 1 – 25