Академический Документы
Профессиональный Документы
Культура Документы
w
1
mLo
2
s
D
L
3
1Y
B
2
s
bD
3
t
6m
e
D
B
2
d
bD
3
t
12m
e
D
_ _
w
xxxx
D
L
2
B
2
s
bD
2
t
pm
0
D
ln
x
1x
_ _ _
_
B
2
d
bD
2
t
2pm
0
D
ln
x
1x
_ __
w
xxx
D
L
2B
2
s
bD
t
m
0
D
B
2
d
bD
t
m
0
D
_
w
xx
DY
L
3
u
xxx
_
D
L
3
B
s
B
d
bD
3
t
3m
e
D
w
xxxx
2B
s
B
d
D
2
t
L
pm
0
D
ln
x
1x
_ _
w
xxx
4B
s
B
d
bD
t
L
2
m
0
D
w
xx
_ _
cosO
1
t
D
L
3
B
2
d
bD
3
t
12m
e
D
w
xxxx
B
2
d
bD
2
t
L
2pm
0
D
_
ln
x
1x
_ _
w
xxx
B
2
d
bD
t
L
2
m
0
D
w
xx
_
cos2O
1
t
D
L
3
P
s
P
d
cosO
2
tw
xx
_
0: 1
B(t) = B
s
+ B
d
cos
1
t
P(t) = P
s
+ P
d
cos
2
t
P(t) = P
s
+ P
d
cos
2
t
N
S
Top skin
Bottom skin
Viscoelastic c Viscoelastic
MRE
L
L
2
L
1
z
x
Fig. 1. Schematic diagram of MRE embedded viscoelastic cored sandwich beam
subjected to periodic magnetic eld and axial load.
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 2
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
and
w
xxx
u
xx
g
n
c
h
t
h
c
_ _
2
u 0: 2
here ( )
,x
represents the differentiation with respect to x.
D(E
t
I
t
E
b
I
b
), D
t
2h
t
. Here g
n
c
g
c
1jZ
c
, g
c
G
c
=2E
t
h
c
=h
t
L=h
t
2
is the shear parameter, subscript c v or e depending
upon viscoelastic or MRE part of the core materials, respectively.
j
1
p
and Z
c
is the core loss factor. Y 31h
c
=h
t
2
. o
s
is the
fundamental frequency of the system. m
ej
(j t,b) and m
0
are the
permeability of the top and bottom faces and magnetic perme-
ability in free spaces.
Solutions of Eqs. (1) and (2) are assumed in the form,
wx,t
r
m 1
q
m
tw
m
x and ux,t
s
k r 1
q
k
tu
k
x 3
where q
m
and q
k
are the time modulation and w
m
x and u
k
x are
the shape functions chosen to satisfy the boundary conditions.
Using single mode approximation in generalized Galerkins
method, Eqs. (1) and (2) can be reduced to the following
non-dimensional temporal equation of motion.
q1iezqea
1
cosO
1
t a
2
cos2O
1
t a
3
cosO
2
tq 0: 4
Due to the presence of three periodically time varying terms in
the temporal equation, which are coefcients of the response q,
this is a parametrically multi-frequency excited equation. The
shape functions used for the numerical calculations are same as
those found in the work of Ray and Kar [21]. To obtain the
solution of the temporal equation of motion (4), here second
order method of multiple scales [30] has been used. In this
method, the displacement q can be represented in terms of
different time scales (T
0
,T
1
,T
2
) and a book keeping parameter e
as follows. The book keeping parameter e is used in order to make
the coefcients z, a
1
, a
2
and a
3
in the same order.
qt; e q
0
T
0
,T
1
,T
2
eq
1
T
0
,T
1
,T
2
e
2
q
2
T
0
,T
1
,T
2
Oe
3
5
Here, T
n
e
n
t, n 0,1,2,3. . .. Taking, D
0
@=@T
0
, D
1
@=@T
1
,
and D
2
@=@T
2
, solving the Eq. (4) and equating the coefcient
of like powers of e, yields the following equations.
Order e
0
:
D
2
0
q
0
q
0
0 6
Order e
1
:
D
2
0
q
1
q
1
2D
0
D
1
q
0
izq
0
a
1
cosO
1
tq
0
a
2
cos2O
1
tq
0
a
3
cosO
2
tq
0
7
Order e
2
:
D
0
2
q
2
q
2
2D
0
D
1
q
1
D
2
1
q
0
2D
0
D
2
q
0
izq
1
a
1
cosO
1
tq
1
a
2
cos2O
1
tq
1
a
3
cosO
2
tq
1
8
General solutions of Eq. (6) can be written as
q
0
AT
1
,T
2
expiT
0
AT,T
2
expiT
0
9
Here, AT
1
,T
2
is the complex conjugate of A(T
1
,T
2
).
Substituting Eq. (9) into Eq. (7) leads to
D
0
2
q
1
q
1
2iD
1
AexpiT
0
izAexpiT
0
a
1
2
AexpiO
1
1T
0
a
1
2
AexpiO
1
1T
0
a
2
2
Aexpi2O
1
1T
0
a
2
2
Aexpi2O
1
1T
0
a
3
2
AexpiO
2
1T
0
a
3
2
AexpiO
2
1T
0
cc 10
Here, cc stands for the complex conjugate of the preceding
terms. In this case, one may observe that the solution of Eq. (10)
contains secular or small divisor terms when the non-
dimensional frequency of external axial loading O
2
is nearly
equal to 2 and/or non-dimensional frequency of magnetic eld
strength O
1
is nearly equal to 1. Hence one may have three
different resonance conditions viz. (i) O
2
%2 and O
1
is away from
1, (ii) O
1
%1 and O
2
is away from 2 and (iii) O
2
%2 and O
1
%1
simultaneously. These three conditions are discussed in the
following subsections.
2.1. Principal parametric resonance case (O
2
%2 and
O
1
away from 1)
In this case, one may use detuning parameter s to express the
nearness of frequencies for the resonance conditions, as
O
2
22es, and s O1 11
Substituting Eq. (11) into Eq. (10) and eliminating the secular
or small divisor terms yields
2iD
1
AizA
a
3
2
Aexpi2sT
1
0 12
The particular solution of the Eq. (10) after eliminating the
secular or small divisor terms is
q
1
a
1
2O
1
1
2
1
AexpiO
1
1t
a
1
2O
1
1
2
1
AexpiO
1
1t
a
2
2O
1
1
2
1
Aexpi2O
1
1t
a
1
2O
1
1
2
1
Aexpi2O
1
1t
a
3
2O
2
1
2
1
AexpiO
2
1tcc 13
Substituting Eqs. (12) and (13) into the Eq. (8) and eliminating
the secular or small divisor terms yields
2iD
2
AD
2
1
AGA 0 14
where
G
1
4
a
2
1
O
2
1
2O
1
_ _
1
4
a
2
1
O
2
1
2O
1
_ _
1
4
a
2
2
4O
2
1
4O
1
_ _
1
4
a
2
2
4O
2
1
4O
1
_ _
1
4
a
2
3
O
2
2
2O
2
_ _
From Eq. (12), one may obtain
D
2
1
A
z
2
A
4
a
3
4
_ _
2
A
a
3
4
iz2sAexp2isT
1
15
Substituting Eqs. (15) into Eq. (14), gives
2iD
2
A
z
2
4
a
3
4
_ _
2
G
_ _
A
a
3
4
iz2sAexp2isT
1
0 16
Substituting Eqs. (15) and (16) to equation d=dt D
0
eD
1
e
2
D
2
Oe
2
, one may obtain the following equation:
2i
dA
dt
ieze
2
z
2
4
a
3
4
_ _
2
G
_ _ _ _
A
a
3
2
e
e
2
2
iz2s
_
_ _
Aexp2isT
1
0
17
Putting, A B
r
iB
i
expiest, where B
r
, and B
i
are real and
imaginary parts, respectively, in Eq. (17) and separating the real
and imaginary parts yield the following equations:
2
dB
r
dt
eze
2
a
3
4
z
_ _
B
r
2se
a
3
2
e
2
a
3
2
s
_ _
e
2
z
2
4
a
3
4
_ _
2
G
_ _ _ _
B
i
0
18
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 3
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
2
dB
i
dt
eze
2
a
3
4
z
_ _
B
i
2se
a
3
2
e
2
a
3
2
s
_ _
e
2
z
2
4
a
3
4
_ _
2
G
_ _ _ _
B
r
0
19
Substituting B
r
,B
i
b
r
,b
i
expgt into the Eqs. (18) and (19)
yields the following equations:
2gb
r
eze
2
a
3
4
z
_ _
b
r
2se
a
3
2
e
2
a
3
2
s
_ _
e
2
z
2
4
a
3
4
_ _
2
G
_ _ _ _
b
i
0
20
2gb
i
eze
2
a
3
4
z
_ _
b
i
2se
a
3
2
e
2
a
3
2
s
_ _
e
2
z
2
4
a
3
4
_ _
2
G
_ _ _ _
b
r
0
21
For steady state trivial response, g is equal to zero. One may
obtain the expression for the transition curves by nding the
value of s from the above two equations. Neglecting the terms
O(e
3
) one may write the expression for transition curves of the
second order expansion when O
2
E2 as
O
2
27
e
4
e
2
z
2
a
3
2
_ _
2
4G
_ _
2
16e
2
z
2
a
2
3
4
_ _
_ _
_
e
2
z
2
a
3
2
_ _
2
4G
_ _
_
_
_
_
22
2.2. Simple resonance case (O
1
%1 and O
2
away from 2)
Following the method similar to that described in Section 2.1,
for this simple resonance case O
1
%1 (O
1
1es) and O
2
is away
from 2, the transition curves emanating from O
1
%1 may be
written as
O
1
17
e
8
e
2
z
2
a
2
2
_ _
2
4G
_ _
2
16e
2
z
2
a
2
2
4
_ _
_ _
_
e
2
z
2
a
2
2
_ _
2
4G
_ _
_
_
_
_
23
Here,
G
1
4
a
2
1
O
2
1
2O
1
_ _
1
4
a
2
2
4O
2
1
4O
1
_ _
1
4
a
2
3
O
2
2
2O
2
_ _
1
4
a
2
3
O
2
2
2O
2
_ _
2.3. Simultaneous resonance case (O
1
%1 and O
2
%2)
In this resonance condition when the system is excited simul-
taneously by the external force and the magnetic eld the
frequency of excitation can be given by OO
2
2O
1
22es.
Following the method similar to that described in Sections 2.1 and
2.2, the expression for the transition curves can be given by
O27
e
4
e
2
z
2
a
2
a
3
2
_ _2
4G
_ _
2
16e
2
z
2
a
2
a
3
2
_ _2
_ _
_
_
e
2
z
2
a
2
a
3
2
_ _2
4G
_ _
_
_
24
Here
G
1
4
a
2
1
O
2
1
2O
1
_ _
1
4
a
2
1
O
2
1
2O
1
_ _
1
4
a
2
2
4O
2
1
4O
1
_ _
1
4
a
2
a
3
O
2
1
2O
1
_ _
1
4
a
2
a
3
4O
2
1
4O
1
_ _
1
4
a
2
3
O
2
2
2O
2
_ _
3. Numerical results and discussions
In this section numerical investigation has been carried out to
study the effects of the key parameters like skin thickness, static
magnetic eld B
s
, static axial force P
s
, amplitude of transverse
periodic magnetic eld B
d
, amplitude of axial periodic load P
d
,
percentage of iron particles and percentage of nano-sized carbon
blacks on the instability regions of the sandwich beam. For this
purpose, the magnetorheological elastomer based on natural
rubber containing different percentage of iron particles and carbon
blacks in nano-size [13,14] are considered for sandwich beams
with simply supported and clamped free end conditions. It has
been observed from the experiment [13] that for same magnetic
eld strength, with increase in percentage of iron particles, the
shear modulus and loss factor increases. Use of higher percentage
of plasticizer also increases the shear modulus and loss factor. For
same percentage of iron particle, with increase in percentage of
carbon black, though the shear modulus increases, the loss factor
decreases [14].
The dimensions and material properties of the sandwich beam are
given as follows. To keep the Euler buckling strength approximately
same for the simply supported and clamped free end conditions, the
spans of the beam (L) is taken as 950 mm for simply supported
and 500 mm for clamped free end conditions. Other physical para-
meters, width, b30 mm; the top and bottom skins thicknesses,
2h
t
2h
b
0.1 mm, the core thickness, 2h
c
2 mm are same for both
the boundary conditions. The top and bottom aluminum skins have
Youngs modulus, E
t
E
b
72 GPa, densities, r
t
r
b
2700 kg m
3
and permeabilities m
e
m
et
m
eb
1.25665e6. Magnetic permeability
in free spaces m
0
4pe7 [24]. The zero eld shear modulus and
Youngs modulus of both MRE and non-MRE parts are same and are
given in Table A1 [13,14].
3.1. Effect of magnetic eld on modal frequencies of sandwich beam
Fig. 2 illustrates the variations in the rst six modal frequen-
cies for a MRE embedded clamped-free sandwich beam without
magnetic eld and with static magnetic eld strength of 0.8 T.
Here, the properties of MRE based on natural rubber containing
80% of iron particles have been taken [13]. It can be seen that the
natural frequencies corresponding to all the modes increase with
the application of magnetic eld. The results show signicant
effect of the magnetic eld on the higher modal frequencies,
while the effect is small on the lower modes. The increase in the
modal frequencies with increasing magnetic eld can be attrib-
uted to increase in the complex shear modulus of the MRE and
thus the increase in the stiffness of the sandwich beam under a
higher magnetic eld. This conrms the potential of the MRE to
control the vibration of the sandwich beam by actively shifting
Fig. 2. Variation of modal frequencies of a clamped free sandwich beam without
and with magnetic eld (B
s
0.8T and B
d
0 T).
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 4
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
the natural frequencies or the resonance conditions by applying
magnetic eld.
Taking the amplitude of dynamic magnetic eld B
d
equal to
zero, the modal frequencies are given in Table 1 which are found
to be in good agreement with those obtained by Zhou and Wang
[24]. The present system reduces to that of Nayak et al. [29] in the
absence of dynamic magnetic eld and hence same frequencies
were found in that work. It may be noted that in Zhou and Wang
[24] a higher order theory has been taken and in the present work
classical sandwich beam theory has been taken. The advantage of
using classical thory over the higher order theory is the less
computational time and memory requirements and simplicity in
analysis procedure.
In the following subsections the instability regions of sand-
wich beams with simply supported and clamped free end condi-
tions have been studied for the three resonance cases. In Section
3.2 the results for the case with non-dimensional frequency of
magnetic eld O
1
%1 and non-dimensional frequency of external
axial loading O
2
away from 2, in Section 3.3 the results for
the case with O
2
%2 and O
1
away from 1 and in Section 3.4 the
results for the case with O
1
%1 and O
2
%2 are presented. The
instability regions are plotted and the effects of many key
parameters like skin thickness, static magnetic eld, B
s
static
axial force, P
s
amplitude of transverse periodic magnetic eld, B
d
amplitude of axial periodic load, P
d
percentage of iron particles
and percentage of nano-carbon blacks on the regions of instability
are investigated. In these plots, the regions bounded by the curves
are unstable and regions outside the curves are stable. The
unstable and stable regions for the sandwich beam with 0.1 mm
skin thickness have been represented by letter U and S, respec-
tively, in Fig. 3a.
One may use the instability regions developed in this work for
suppressing the vibration of the system. For example for a
sandwich beam with 0.1 mm skin thickness the system is
unstable at B
d
4 with a frequency of O
1
1 (Fig. 3a), and hence
the system will vibrate. Now to suppress this vibration one may
change only the frequency of the AC component of the magnetic
eld (e.g., O
1
0:9 or O
1
1:08). Similarly for a sandwich beam
with 0.5 mm thickness (B
d
4, O
1
1) to suppress this vibration
one may either decrease the amplitude of AC component of the
magnetic eld (e.g., B
d
2:4) or change the frequency of the AC
component of the magnetic eld (e.g., O
1
0:9 or O
1
1:08). In
the similar way one may interpret the instability region devel-
oped in this work for attenuation or control of vibration by
suitably changing the system parameters.
3.2. Simple resonance due to magnetic eld (O
1
%1 and
O
2
is away from 2)
In this case, the system is subjected to magnetic eld with a
frequency nearly equal to the natural frequency of the system and
the frequency of external axial loading is away from the principal
parametric instability zone. So the instability regions are inde-
pendent of P
d
.
Initially a simply supported sandwich beam has been consid-
ered for the analysis. Fig. 3a and b show the effect of skin
thickness of a simply supported sandwich beam on the transition
curves with MRE containing 80% of iron particles and MRE
containing 33% of iron particles with 7% of carbon black, respec-
tively. With increase in skin thickness the instability region
decreases as the skins strengthen the bulk exural rigidities of
the sandwich beam. From Fig. 3a it is noted that with decrease in
B
d
the unstable region decreases and the system can operate for a
wide frequency range for lower value of B
d
. Further, there exists a
critical value of B
d
B
dcr
as shown in Fig. 3a below which the
system has no unstable region and hence can operate at any
frequency without vibration. By using less percentage of iron
particles with carbon black in MRE, the critical B
d
value below
which the system becomes stable for frequency O
1
, marginally
decreases which is clearly illustrated by the comparison of B
dcr
values for different MRE patch as given in Table 2. This is due to
the fact that, though the shear modulus of MRE increases with
carbon black, the loss factor decreases. From Fig. 3 it may also be
observed that the system with higher skin thickness remain
stable for all frequency range for higher value of B
d
.
Table 1
Comparison of the modal frequencies obtained in the present analysis with those
obtained by Zhou and Wang [24].
Magnetic eld Modal frequencies (Hz)
Mode Present analysis
with B
d
0
Zhou and Wang
([21], Fig. 4a)
1 T
1 67.338 68.816
2 144.923 152.441
3 222.145 225.08
Fig. 3. Effect of skin thickness on the instability regions of the sandwich beam with (a) MRE, containing 80% of iron particles and (b) MRE containing 33% of iron particles
and 7% of carbon black, subjected to dynamic magnetic eld P
s
1N and B
s
0.
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 5
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
The results presented in Fig. 4 clearly illustrate the inuence of
static axial force P
s
and static magnetic eld B
s
on the transition
curves. Comparing Fig. 4a and b, with increase in static axial load
P
s
from 5N to 10N the instability regions decrease and the critical
value of B
d
increases. This is due to the fact that with increase in
P
s
, the compressive force in the skins makes the system stiffer.
One may also observe that with increase in iron particles from
60% to 80% in MRE the instability regions decreases but with
increase in carbon black from 4% to 7% in MRE containing 33% of
iron particles the instability regions increases. This can be
attributed to the relatively increase in the loss factor with
increase in percentage of iron particles and decrease in the loss
factor of the MRE with increase in carbon blacks.
Comparing Fig. 4a and c it is observed that with increase in
static transverse magnetic eld density B
s
from 0.1 T to 0.2 T, the
width of the instability regions decreases and the critical value of
B
d
increases. Here the critical values of B
d
are more in comparison
to that in case of change of static load which is given in Table 3.
This can be related to the fact that the relative variation in the
beam stiffness could be greater than that of the beam subjected to
variation of the static load. It should be noted that due to the
increase of eld dependent shear modulus of the MRE patches
with magnetic eld, the stiffness of the beam changes.
Now considering a clamped free MRE embedded sandwich,
since similar trends in the instability regions are observed as in
case of simply supported boundary condition for the variation the
Table 2
Comparison of B
dcr
for different skin thicknesses with different MRE patch.
Skin thickness (mm)
B
dcr
Core with MRE patch
containing 80% of
iron particles
Core with MRE patch
containing 33% of iron
particles and 7% of
carbon black
0.1 0 0
0.3 1.152 0.829
0.5 2.549 2.264
Fig. 4. Inuence of static axial load P
s
and static magnetic eld B
s
on instability regions. (a) P
s
5 N, B
s
0.1 T; (b) P
s
10 N, B
s
0.1 T and (c) P
s
5 N, B
s
0.2 T.
Table 3
Values of critical dynamic magnetic eld, B
dcr
of the simply supported sandwich
beam with different MRE patch.
MRE patch with
B
dcr
Static load (5 N)
and static
magnetic eld
(0.1 T)
Static load (10 N)
and static
magnetic eld
(0.1 T)
Static load
(5 N) and static
magnetic eld
(0.2 T)
60% IP 1.158 1.16 2.001
80% IP 2.854 3.123 3.512
33% IP and 4% CB 2.266 2.536 2.836
33% IP and 7% CB 0.46 0.462 1.162
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 6
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
skin thickness, the instability regions are not plotted but the
critical values of B
dcr
are given in Table 4. Like simply supported
case, here also the instability regions decreases and the critical
amplitude of dynamic magnetic eld B
dcr
increase with increase
in skin thickness.
Fig. 5 shows the instability regions of a clamped-free sandwich
beam subjected to static axial force P
s
and static magnetic eld B
s
.
The values of B
dcr
with the variation of P
s
and B
s
, are presented in
the Table 5. Since, B
d
depends on the length of the beam, in this
case the maximum value of B
d
decreases in comparison to that of
the simply supported case. From Tables 3 and 5 it is observed that
the B
dcr
values in case of simply supported sandwich beam are
more than that of the clamped free end condition. Hence, to
obtain the instability regions one should analyze the system for
specic boundary condition.
Fig. 6 shows the time response of a clamped-free sandwich
beam considering two different cores with MRE patch containing
60% and 80% of iron particles. This corresponds to the point
marked A in Fig. 5. As the point is outside the instability region,
the system is supposed to be stable which is validated from the
time response shown in Fig. 6. One may clearly observe that by
increasing the percentage of iron particles in MRE patch, the
system response can be reduced for the same magnetic eld
and frequency corresponding to the point A. The settling time is
reduced from t 100 to t 60.
Now taking the same amplitude of magnetic eld strength, if one
increases the frequency of the applied magnetic eld O
1
, the system
may be unstable. For example, taking frequency O
1
corresponding
to point B as marked in Fig. 5, Fig. 7a shows the time response for
the transverse displacement of the clamped-free sandwich beam
with MRE embedded core. In Fig. 5, point B is clearly in the
unstable region, which is in good agreement with the response
curve shown in Fig. 7a. For the same amplitude of magnetic eld
strength, Fig. 7b shows the time response for transverse displace-
ment of the system corresponding to the points C, as marked in
Fig. 5. Clearly this point is in stable region and is in good agreement
with the response curve shown in Fig. 7b. The time response is
obtained by numerically solving the temporal Eq. (4).
3.3. Principal parametric resonance (O
2
%2 and O
1
is away from 1)
In this subsection, effect of percentage of iron particles,
percentage of carbon black, static axial force and static magnetic
eld has been investigated for the simply supported and clamped
free boundary conditions when the system is excited at a
frequency O
2
%2 (i.e. nearly equal to twice the fundamental
frequency) and O
1
is away from 1.
Table 4
Comparison of B
dcr
for different skin thicknesses with different MRE patch for a
clamped-free sandwich beam.
Skin thickness
(mm)
B
dcr
Core with MRE patch
containing 80% of iron
particles
Core with MRE patch
containing 33% of iron
particles and 7% of
carbon black
0.1 0 0
0.3 0.953 0.687
0.5 2.115 1.878
Fig. 5. Instability regions of the sandwich beam subjected to static axial force
P
s
5 N and static magnetic eld B
s
0.1 T.
Table 5
Values of critical dynamic magnetic eld, B
dcr
of the clamped free sandwich beam
with different MRE patch.
MRE patch with
B
dcr
Static load (5 N)
and static
magnetic eld
(0.1 T)
Static load (10 N)
and static
magnetic eld
(0.1 T)
Static load
(5 N) and static
magnetic eld
(0.2 T)
60% IP 0.993 1.063 1.63
80% IP 2.447 2.897 3.272
33% IP and 4% CB 1.942 2.324 2.941
33% IP and 7% CB 0.394 0.424 1.069
Fig. 6. Time response at point A for MRE patches containing 60% and 80% of iron
particles; key as in Fig. 5.
Fig. 7. Time responses for (a) point B and (b) point C; key as in Fig. 5.
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 7
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
The inuence of skin thickness on the transition curves of a
simply supported sandwich beam with MRE containing 80% of iron
particles and MRE containing 33% of iron particles with 7% of carbon
black respectively are shown in Fig. 8a and b. With increase in skin
thickness the instability region decreases. From Fig. 8a it is observed
that there exists a critical value of P
d
P
dcr
below which the system
has no unstable region and hence the system can operate at any
frequency without vibration. Also it may be noted that, by using less
percentage of iron particle with carbon black in MRE, the critical
value of dynamic load P
d
below which the system becomes stable
for the frequency O
2
, marginally decreases. This is due to the fact
that, though the shear modulus of MRE increases with carbon black,
the loss factor decreases. The critical values of dynamic load P
dcr
are
given in Table 6. To compare the results of the present case with that
of [29] where dynamic magnetic eld (AC component) is not
considered, Fig. 9 is plotted for various combination of B
s
and B
d
.
It may be noted that in [29] instability regions were found by using a
different method. The instability regions with B
d
0 are found to be
in good agreement with [29]. FromFig. 9 one may observe that while
application of static (B
d
0) or dynamic magnetic eld improves the
stability of the system by decreasing the instability region, applica-
tion of static magnetic eld (DC component) has more stabilizing
effect than the dynamic (AC component) magnetic eld.
Fig. 10 shows the instability regions for different static load
and magnetic elds for four different types of MRE patch and
critical values of dynamic load P
dcr
for these cases are given in
Table 7. It has been observed that with increase in percentage of
iron particles the stability of the system improves and the P
dcr
value increases. But for the same percentage of iron particle with
increase in percentage of carbon black, P
dcr
decreases. Also it may
be noted that doubling the static load has less effect in increasing
P
dcr
values than doubling the magnetic eld. Hence one can
effectively increase the magnetic eld for active vibration reduc-
tion as it improves the stability of the system signicantly.
The effect of skin thickness on the instability regions for the
clamped free sandwich beam are found to be similar to simply
supported case and the instability regions in this case also decrease
with increase in skin thickness as the critical dynamic load P
dcr
increases. Table 8 presents the variation of critical dynamic load P
dcr
for MRE with 80% iron particles and 33% iron particle with 7% carbon
black. It is observed that, in the former case, one can have higher
P
dcr
. This may be due to the fact that in the later case, though the
shear modulus is more, the loss factor is less.
The inuence of P
s
and B
s
on the instability regions for the
clamped free boundary conditions are found to be similar to the
that of a simply supported beam and the values of critical
dynamic load P
dcr
for different MRE patches are presented in
Table 9. As with increase in static magnetic eld the shear
modulus of the MRE increases, which causes the increase in
overall stiffness of the system, the critical value of dynamic load
P
d
increases. The increase of P
dcr
makes the system more stable
and hence the system can operate at a wide range of frequency
Fig. 8. Effect of skin thickness on the instability regions of the sandwich beam with (a) MRE contains 80% of iron particles and (b) MRE containing 33% of iron particles and
7% of carbon black P
s
1 N and B
s
0 T, subjected to dynamic magnetic eld.
Table 6
Comparison of P
dcr
for different skin thicknesses with different MRE patch.
Skin thickness
(mm)
P
dcr
Core with MRE patch
containing 80% of iron
particles
Core with MRE patch
containing 33% of iron particles
and 7% of carbon black
0.1 0 0
0.3 0.9 0.7
0.5 2.8 2.3
Fig. 9. Transition curves of a simply supported sandwich beam with MRE
containing 80% of iron particles for different combinations of B
s
and B
d
values
2h
t
0.1 mm, P
s
1 N.
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 8
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
without any vibration. Here, with increase in iron particles from
60% to 80% in MRE the instability regions decreases but with
increase in carbon black from 4% to 7% in MRE containing 33% of
iron particles the instability regions increases. This is due to the
relative increase of loss factor with increase in percentage of iron
particles and decrease of the loss factor of MRE with increase in
percentage of blacks.
3.4. Simultaneous principal parametric and simple resonance
(O
2
%2 and O
1
%1)
In this case the system is subjected to dynamic magnetic eld
with a frequency nearly equal to the natural frequency of the
system and an external axial loading with a frequency nearly
Fig. 10. Inuence of static axial load, P
s
and static magnetic eld B
s
on instability regions. (a) P
s
5 N, B
s
0.1 T; (b) P
s
10 N, B
s
0.1 T and (c) P
s
5 N, B
s
0.2 T.
Table 7
Values of critical dynamic axial force, P
dcr
of the simply supported sandwich beam
with different MRE patch.
MRE patch with
P
dcr
Static load (5 N)
and static
magnetic eld
(0.1 T)
Static load (10 N)
and static
magnetic eld
(0.1 T)
Static load
(5 N) and static
magnetic eld
(0.2 T)
60% IP 1.5 1.7 2.2
80% IP 3.1 3.4 4.1
33% IP and 4% CB 2.6 2.8 3.1
33% IP and 7% CB 1.0 1.1 1.7
Table 8
Comparison of P
dcr
for different skin thicknesses with different MRE patch.
Skin thickness
(mm)
P
dcr
Core with MRE
patch containing
80% of iron particles
Core with MRE patch
containing 33% of iron
particles and 7% of
carbon black
0.1 0 0
0.3 0.9 0.69
0.5 2.7 2.25
Table 9
Values of critical dynamic axial force, P
dcr
of the clamped free sandwich beam with
different MRE patch.
MRE patch with
P
dcr
Static load (5 N)
and static
magnetic eld
(0.1 T)
Static load (10 N)
and static
magnetic eld
(0.1 T)
Static load
(5 N) and static
magnetic eld
(0.2 T)
60% IP 1.483 1.692 2.1
80% IP 3.076 3.573 4.096
33% IP and 4% CB 2.567 2.785 3.098
33% IP and 7% CB 0.987 1.098 1.654
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 9
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
equal to twice the fundamental frequency. Hence the system is
subjected to simultaneous principal parametric resonance due to
external axial loading and simply resonance due to the magnetic
eld. The instability regions are plotted for simply supported and
clamped free boundary conditions.
Fig. 11a and b shows the inuence of dynamic magnetic eld
on the instability regions for simply supported sandwich beam. It
is observed that with increase in magnetic eld the stability of the
system improves. The critical value of dynamic load, P
dcr
increases
and the instability regions moves upward. The values of P
dcr
for
simply supported sandwich beam are given in Table 10. From this
table it is observed that the value of P
dcr
increases by 96.7% with
the variation of dynamic magnetic eld from 0.2 T to 0.6 T for the
beam having the MRE patch containing 80% of iron particles. This
characteristic indicates that the applied magnetic eld increases the
exural rigidity of the sandwich beam due to eld dependent shear
modulus of MRE. Similar to previous resonance conditions, here also
the instability regions decrease with increase in percentage of iron
particles and increases with increase in carbon black.
Also for the same boundary condition, the inuence of static
axial load on the instability regions is shown in Fig. 11c and d and
the critical dynamic magnetic eld, B
dcr
obtained in this resonance
condition are given in Table 10. It is noticed that the value of B
dcr
increases by 9.25% at a dynamic load, P
d
1N and with variation of
static load from 5N to 10N for simply supported sandwich beam
having the MRE patch containing 80% of iron particles.
Similar to simply supported case, for the clamped free sandwich
beam Fig. 12ad shows the instability regions due to dynamic
magnetic eld and axial load. The critical values are given in
Table 11. It is observed that the value of P
dcr
increases by 53.95%
with the variation of dynamic magnetic eld from 0.2 T to 0.6 T for
the beamhaving the MRE patch containing 80% of iron particles. Also
for sandwich beam with MRE patch containing 80% of iron particles
it is noticed that the value of B
dcr
increases by 46.23% at a dynamic
load of P
d
1 N and varying the static load from 5 N to 10 N.
4. Conclusions
In this work an attempt has been made to study the para-
metric instability regions for simply supported and clamped free
Fig. 11. Inuence of dynamic magnetic eld B
d
and static axial load P
s
on the transition curves for simply supported sandwich beam. (a) P
s
1 N, B
d
0.2 T; (b) P
s
1 N,
B
d
0.6 T; (c) P
d
1 N, P
s
5 N and (d) P
d
1 N, P
s
10 N.
Table 10
Values of critical dynamic axial force, P
dcr
and critical dynamic magnetic eld, B
dcr
of a simply supported sandwich beam with different MRE patches for combination
resonance case.
MRE patch with
P
dcr
(P
s
1 N) B
dcr
(P
d
1 N)
B
d
0.2 T B
d
0.6 T P
s
5 N P
s
10 N
60% IP 1.1 4.2 2.006 2.836
80% IP 3.00 5.9 3.935 4.299
33% IP and 4% CB 2.00 5.0 2.854 3.512
33% IP and 7% CB 0.6 3.7 1.344 2.506
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 10
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
sandwich beams with MRE embedded viscoelastic core subjected
to time varying transverse magnetic eld and axial periodic load
using perturbation method. The temporal equation of motion,
which contains two frequency parametric excitation terms, is
solved using second-order method of multiple scales. Instability
regions are found for three different resonance conditions viz.,
simple resonance, principal parametric resonance and simulta-
neous principal parametric and simple resonance. With the
increase in skin thickness the bulk exural rigidity of the
sandwich beam increases and the instability regions decreases.
Due to the change of shear modulus of MRE and magnetoelastic
loads on the conductive skins under the application of the
magnetic eld, the dynamic property of the sandwich beam
changes. Also the dynamic property changes due to the
application of static and dynamic load. The instability regions
decrease with the application of static magnetic eld and static
axial load. With the increase in percentage of iron particles in case
of MRE patch containing iron particles the instability regions
decreases. However with the increase in percentage of carbon
black in case of MRE patch containing less percentage of iron
particles and carbon blacks, the instability regions increases.
Though in all these resonance conditions above mentioned
similar observations have been realized, the actual instability
regions and response amplitude depends on the particular reso-
nance condition and the concerned boundary condition. Hence
based on the operating range of the sandwich beam, these points
should be considered while designing a sandwich structure. To
have a vibration free sandwich beam the control parameters (viz.,
amplitude and frequencies of magnetic eld and axial load)
should be in the stable regions of the instability plots. These
plots can easily be developed using the expressions developed for
transition curves in this work. Hence, this work will nd
application for passive and active vibration suppression of
sandwich beams.
Appendix
Following the assumptions and strain displacement relations
in [29] where the expressions for strain energy (U), kinetic energy
(T) and non-conservative work done (W
nc
) due to axial load and
Fig. 12. Inuence of dynamic magnetic eld B
d
and static axial load P
s
on the transition curves for clamped free sandwich beam. (a) P
s
1 N, B
d
0.2T; (b) P
s
1 N,
B
d
0.6 T; (c) P
d
1 N, P
s
5 N and (d) P
d
1 N, P
s
10 N.
Table 11
Values of critical dynamic axial force, P
dcr
and critical dynamic magnetic eld, B
dcr
of a simply supported sandwich beam with different MRE patches for combination
resonance case.
MRE patch with
P
dcr
(P
s
1 N) B
dcr
(P
d
1 N)
B
d
0.2 T B
d
0.6 T P
s
5 N P
s
10 N
60% IP 1.098 3.3 1.431 2.604
80% IP 2.988 4.6 2.894 4.232
33% IP and 4% CB 1.997 4.2 2.323 3.501
33% IP and 7% CB 0.598 2.6 1.086 2.331
B. Nayak et al. / International Journal of Non-Linear Mechanics ] (]]]]) ]]]]]] 11
Please cite this article as: B. Nayak, et al., Multi-frequency excitation of magnetorheological elastomer-based sandwich beam with
conductive skins, Int. J. Non-Linear Mech. (2011), doi:10.1016/j.ijnonlinmec.2011.08.007
magnetoelastic loads are given by the following equations:
U
1
2
_
L
0
E
t
I
t
w
2
,xx
dx
1
2
_
L
0
E
t
A
t
u
2
t,x
dx
1
2
_
L
0
E
b
I
b
w
2
,xx
dx
1
2
_
L
0
E
b
A
b
u
2
b,x
dx
1
2
_
L
0
G
n
c
A
c
g
2
c
dx A1
T
1
2
_
L
0
mw
2
,t
dx: A2
W
nc
1
2
_
L
0
Ptw
2
,x
n
t
w
2
,x
n
b
w
2
,x
dx
_
L
0
m
m
t
dw
,x
m
m
b
dw
,x
dx
A3
here ( )
,x
represents the differentiation with respect to x. A
t
, A
c
and
A
b
are the cross sectional area of top, core and bottom layers,
respectively. g
c
denotes the shear strain in the core. u
t
and u
b
denote axial displacements of the center-line of the top and
bottom skins, respectively.
As the skins are conductive, the magnetoelastic load applied to
the conductive skins is equivalent to the horizontal force n
j
and
the distributed moment m
j
, which are expressed in terms of the
longitudinal displacement (u
j
) and transverse displacement (w
j
)
as in [24] given below.
n
j
B
2
bh
j
m
ej
u
j,xx
, A4
and
m
m
j
B
2
bh
j
m
0
p
2ln
x
Lx
u
j,x
h
j
2p
w
j,xx
ln
x
Lx
w
j,x
_ _
B
2
bh
j
3
12m
ej
w
j,xxx
:
A5
Here, subscript j t and b for top and bottom skins, respec-
tively, m
0
and m
ej
are respectively the permeability of the free
space and the jth skin materials.
Following non-dimensional parameters are used in this ana-
lysis:
t o
s
t, O
1
O
1
=o
s
; O
2
O
2
o
s
; x x=L, u u=L,
wmL
3
=D w and ww=L A6
here t is the non-dimensional time.
The expressions for the coefcients of Eq. (4) are given below
a
1
B
r
d
e
D
2
t
m
0
3m
e
L
2
h
22
h
1
2D
t
pL
h
21
h
1
4
h
11
h
1
_ _
,
a
2
B
d
d
e
D
2
t
m
0
12m
e
L
2
h
22
h
1
D
t
2pL
h
21
h
1
h
11
h
1
_ _
and
a
3
P
d
d
e
h
11
h
1
Here
o
2
L
D
mL
4
, d
o
2
L
o
2
s
, B
s
B
2
s
bD
t
L
2
m
0
D
, B
d
B
2
d
bD
t
L
2
m
0
D
,
B
r
B
s
B
d
bD
t
L
2
m
0
D
, P
s
P
s
L
2
D
, P
d
P
d
L
2
D
and the book keeping parameter e.
o
2
s
o
2
L
h
22
h
1
_ _
1Y 1
h
31
h
13
h
22
h
33
g
n
c
h
c
=h
t
2
h
44
_ _ _
B
s
D
2
t
m
0
6m
e
L
2
D
t
pL
h
21
h
22
2
h
11
h
22
_ _
B
d
D
2
t
m
0
12m
e
L
2
D
t
2pL
h
21
h
22
h
11
h
22
_ _
P
s
h
11
h
22
_
h
1
_
1
0
w
m
x
2
dx, h
11
_
1
0
dw
m
x
dx
_ _
2
dx,
h
21
_
1
0
ln
x
1x
_ _
d
2
w
m
x
dx
2
dw
m
x
dx
_ _
dx,
h
22
_
1
0
d
2
w
m
x
dx
2
_ _2
dx,
h
33
_
1
0
du
k
x
dx
_ _
2
dx, h
31
_
1
0
d
2
w
m
x
dx
2
du
k
x
dx
_ _
dx,
h
13
_
1
0
d
2
u
k
x
dx
2
dw
m
x
dx
_ _
dx, and
h
44
_
1
0
du
k
x
dx
_ _
2
H
1
du
k
x
dx
_ _
2
H
2
_ _
dx,
The expressions of H
1
and H
2
for the MRE and non-MRE parts
are
H
1
1HxL
1
HxL
2
, H
2
HxL
1
HxL
2