International Journal of Non-Linear Mechanics 147 (2022) 104232

Contents lists available at ScienceDirect

International Journal of Non-Linear Mechanics

journal homepage: www.elsevier.com/locate/nlm

An equivalent spring-based model to couple the motion of visco-hyperelastic

dielectric elastomer with the confined compressible fluid/air mass
Ajay Kumar a , Aman Khurana a , Atul Kumar Sharma b , M.M. Joglekar a ,∗
a
Department of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee 247667, India
b
Department of Mechanical Engineering, Indian Institute of Technology Jodhpur 342037, India

ARTICLE INFO ABSTRACT

Keywords: Soft materials exhibiting large voltage-induced deformation under external stimuli have gained increasing
Dielectric elastomer bulging actuator attention in the recent past owing to their potential applications in soft transducers. Dielectric elastomer
Air-coupled motion actuators (DEAs) coupled with a fluid/air mass have proved to undergo considerable voltage-induced de-
Nonlinear dynamics
formation. This article presents a framework to couple the motion of an inflated Dielectric elastomer (DE)
Electro-dynamic response
membrane with a confined compressible fluid/air mass and develops a nonlinear dynamic analytical model
Resonance behavior
Variable air column
to predict the polar region deflection of an inflated DE membrane attached to an airtight chamber (namely
Pseudo air spring bulging actuators) stimulated by an electric field. In order to derive the differential equations which determine
the dynamic behavior of the membrane coupled to an airtight chamber, the Euler–Lagrange equation is
employed, which takes into account the effects of membrane pre-stretch, initial inflation pressure, membrane
viscoelastic behavior, and transient electric loading. A hypothesis of pseudo-air-spring is incorporated to
consider the effect of the volume of the confined fluid/air mass. The experimental investigation conducted for
various geometrical and loading conditions in both depressurized and pressurized states of actuation shows
that the developed nonlinear model satisfactorily predicts the influence of fluid/air mass on the dynamic
response of the considered DE membrane undergoing small voltage-induced deformation. Finally, the proposed
framework is incorporated for building insights on several effecting parameters such as inflation pressure,
quality, and quantity of the confined fluid mass, etc., on the electrodynamic behavior of the actuator. The
results reveal that the volume of the air column appreciably alters the nonlinear dynamic behavior of the
bulging actuator. Poincaré plots along with phase diagrams are presented for analyzing the periodicity of the
nonlinear oscillations of the actuator. The underlying theoretical framework and the obtained inferences of the
present article can be implemented effectively in the design of a futuristic class of structures whose motion is
coupled with compressible fluid/ air mass subjected to time-dependent actuation.

1. Introduction successfully to increase the breakdown strength and suppress the

associated snap-through operational instability [24–27].
Dielectric elastomers (DEs) have received considerable attention Edge-clamped circular membrane diaphragm-type actuators are a
in last couple of decades due to large voltage-induced strains [1], type of DE actuator that is widely used in applications like braille
low cost, light weight, high coupling efficiency, low stiffness, high display [7], diaphragm generators [28], fluid pumps [29], acoustic
energy density, etc. Due to these salient characteristics, abundant actuators [30], tunable bionic lens [12] and so on. In this type of
applications have been developed such as artificial muscles [2], en- actuator, a pre-stretched circular DE membrane is deformed axially
ergy harvesters [3], soft robotics [4,5], prosthetic [6], braille display
symmetrically by the combined effect of both mechanical and elec-
[7], resonator [8], actuator [9–11], adaptive optics [12], tunable
trical loading in the form of pressure differential and electric field.
wave-guides [13–15], minimum energy structures [16–18], etc. High
Inflation before electrical loading provides several advantages, includ-
areal strains ranging from 215% to 2200% in DE actuators have
ing increased membrane pre-stretch along with an additional external
been incorporated to improve their actuation performance, which is
typically restricted by the associated electromechanically induced force due to pressure differential, both of which have a considerable
material failures [19–23]. In this regard, pre-stretching the DE impact on the electromechanical actuation [28,31–34]. In context to
membrane is one of the constructive methods that has been used modeling of inflated dielectric elastomer actuators, Zhao and Suo [31]

∗ Corresponding author.
E-mail addresses: [email protected] (A. Kumar), [email protected] (A. Khurana), [email protected] (A.K. Sharma),
[email protected] (M.M. Joglekar).

https://doi.org/10.1016/j.ijnonlinmec.2022.104232
Received 19 March 2022; Received in revised form 30 August 2022; Accepted 1 September 2022
Available online 17 September 2022
0020-7462/© 2022 Elsevier Ltd. All rights reserved.
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

investigated the electromechanical performance of a thin buckling and Kelvin–Voigt Maxwell model [58] have recently been implemented
DE-based actuator using an elastic finite element model. Later, Fox for modeling the viscoelastic effect of the DE membrane.
et al. [32] experimentally studied the static and dynamic response of an The forgoing investigations highlight the role of electromechanical
inflated DE. Based on experimental investigations, they have reported behavior of soft active material on the performance of inflated DE
for the first time that chamber size has a considerable effect on the membrane by considering the effect of fluid pressure. However, to the
electromechanical responses of the DE membrane. Furthermore, Zhu best of the authors’ knowledge, a unified dynamic electromechanical
et al. [33] devised a method for investigating the electromechanical model predicting the response of a DE-based diaphragm-type actuator
response of a pressure-inflated hyperelastic DE membrane without integrated with an airtight chamber considering the combined effect
taking the influence of fluid mass. The developed model differs signifi- of enclosed fluid mass, dynamic loading, and inherent membrane vis-
cantly from experimental results reported by Fox et al. [32] due to the coelasticity, which exists in actual practice, has not been addressed
associated viscoelastic effect of the VHB membrane. Vertechy et al. [28] in the literature so far. To this end, the current study presents a
developed a methodology based on the concept of the reduced model nonlinear dynamic analytical framework of the inflated DE membranes
for investigating the inflated DE, that provides fast design optimization, attached to the airtight chamber considering the effect of enclosed
hardware in the loop simulation, and control. Some of the previously compressible fluid/air mass. An equivalent spring model of the air
reported investigations focused on the anomalous bulging behaviors column and the DE membranes is employed to simulate the coupling
and bulging instability of peripherally clamped DE-based balloon-type behavior of electrically driven DE membranes with the activated air
actuators under electromechanical loading [35–37]. More specifically, column. A pseudo air spring is used to describe the confined air column,
Wang et al. [35] and Liang et al. [36] studied the behavior of instability assuming that the DE membrane deforms homogeneously with small
modes with an anomalous localized bulging and established their ana- transverse deflections. The coupled governing equation is utilized to
lytical model while ignoring the dynamic and the viscoelastic effects. investigate the dynamic motion of DE membranes with the influence
Further, Lv et al. [37] reported a nonlinear viscoelastic model for the of confined fluid/air mass under depressurized and pressurized states
DE membrane for investigating the electromechanical response of DE of actuation. The effects of parameters, such as the bulging pressure,
actuator when driven by a Heaviside electric load. Chen et al. [34] chamber volume, adiabatic constant of the compressible fluid on the
recently proposed a network of multiple interconnected DE balloon-like dynamic electromechanical behavior is analyzed through an extensive
actuators with the benefits of large, continuous controlled actuation. parametric study. Additionally, the thresholds on the dynamic behavior
For achieving required safety in the measure of the DE sensing devices, as a function of the initial inflation pressure and chamber volume are
the interconnected DE membranes via confined fluid are one of the analyzed. Experimental investigations are performed to corroborate the
possible solutions [38–40]. In this regard, Carpi et al. [38] developed a predictive capabilities of the proposed mathematical model.
hydrostatically coupled DE membranes based-actuator, in which one of The rest of the paper is ordered as follows. In Section 2, we devel-
the membranes activated with high voltage and the other separated by oped a nonlinear dynamic governing equation of spherically deformed
an incompressible fluid, behaves like an end-effector. Wang et al. [41] DE-based bulging actuators using the Euler–Lagrange equation. A de-
later provided a theoretical interpretation of the experimental studies tailed fabrication techniques and experimental method of the proposed
conducted by Carpi et al. [38]. Next, Wang et al. [42] proposed a bulging actuators are discussed in Section 3. Section 4 describes the
wave handling system by implementing a series of DE actuators that legitimacy of the proposed analytical framework and a comprehensive
are coupled hydrostatically to each other. In the developed DEAs, each parametric study for revealing the effect of various parameters on
actuator represents an independent unit, proposed to provide a soft the performance of DE-based bulging actuators. Salient inferences and
transfer solution of fragile materials, like food items. Furthermore, few future implications of this investigation are summarized in Section 5.
researchers [43,44] analytically investigated the interconnected DEAs-
based balloon without providing any experimental evidence of their 2. Mathematical formulation and solution method
proposed studies. Most of the previously mentioned studies utilized
an incompressible fluid to interconnect the DE membranes. Later, This section provides an effective analytical framework for pre-
Sasikala et al. [39] developed pneumatically coupled dielectric elas- dicting the electromechanical behavior of the bulging actuator with a
tomer (PCDE) actuators in which two DE membranes, namely active spherically deformed DE membrane coupled with the confined fluid
and passive membranes, are separated by the air column. However, in mass. The formulation is based on the assumption of a thin-walled
their investigation, no theoretical model was provided. Further, Chen membrane undergoing homogeneous deformation. Also, the DE film is
et al. [40] developed a theoretical model of a pneumatic–hydraulic cou- assumed to be incompressible and effects of strain-stiffening, stretch-
pled electromechanical actuator that addresses the electromechanical dependent permittivity are not considered in the present analysis. Fur-
behavior of the DE membrane while ignoring the effects of time- ther, the air-chamber is assumed to be leak-proof. First, we introduce
dependent loading and viscoelasticity. In one of our recent works [45], the kinematics of deformation, followed by a discussion of material
we developed a comprehensive nonlinear dynamic model of the PCDE modeling. Furthermore, pressure-induced deformation is demonstrated
actuator incorporating the effect of time-dependent electromechani- by utilizing potential energy minimization, and the concept of a pseudo
cal actuation and material viscoelasticity. In this work, the idea of air spring is introduced to describe the dynamics of electrically stimu-
pseudo air-spring was introduced for modeling the coupling behavior lated response of an inflated DE membrane coupled with a confined
of the intervening air column. The present work uses a similar mod- fluid mass.
eling approach applied for investigating the dynamics of bulging DE
actuators. 2.1. Kinematics of deformation
Several of the aforementioned DEA applications are expected to un-
dergo dynamic motion during their operation. The literatures A circular dielectric elastomeric membrane having dimensions of
[27,46–53] contain extensive experimental and theoretical studies on radius 𝑅𝑖 and thickness 𝑡0 in the reference state is considered, as
the nonlinear oscillations or dynamic response of DEAs. Apart from dy- depicted in Fig. 1a. The membrane is pre-stretched biaxially by stretch
namic behavior, DE materials also exhibit inherent material viscoelastic 𝜆0 and fixed with the chamber of radius 𝑟, as shown in Fig. 1b. The
effect during electrical stimulation. As a result, studying the material voltage-induced dynamic response can be separated into two phases.
behavior of such membranes is crucial in order to understand the im- The intermediate deflated or inflated condition reached in the first
pact of membrane viscoelasticity on the design and actuation response phase corresponds to the applied pressure difference 𝑃𝑔 as shown in
of inflated DEAs. Several rheological material models, including the Fig. 1c, and the obtained deformation in the second phase relates to
Maxwell model [54], Kelvin–Voigt model [55], Zener model [56,57], the combined influence of pressure differential and electrical loading

2
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 1. A schematic 2-D representation of the DE-based bulging actuator (a) initial state, (b) pre-stretched state attached to the chamber, (c) inflated state due to the applied
pressure difference, (d) final state after the application of electric field, and (e) lumped spring model of DE based bulging actuator.

𝜙(𝑡), as depicted in Fig. 1d. In the voltage stimulated phase, the motion described, which is based on an incompressible visco-hyperelastic DE
of the DE membrane is coupled with the fluid mass present inside membrane. The neo-Hookean hyperelastic material model is used to
the airtight chamber, pressure intensity, and the applied electric field. model the two nonlinear springs of the DE membrane. The total strain
In this phase, the confined fluid behaves like a pseudo air spring energy for a given state of applied biaxial stretch is represented as
[59–61] and resists the deformation directed by the pressure difference
⎡ 𝜇𝛼 ( 2 ) ⎤
and electric field. The related parameters, 𝑧𝑎0 and 𝑧𝑎 , indicate the ⎢ 𝜆1 + 𝜆22 + 𝜆−2 1 2
𝜆−2 − 3 + ⎥
2
maximum bulging height of the DE membrane before and after the ⎢ ⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟ ⎥
⎢ ⎥
application of electric field, respectively. Similarly, 𝜆𝐵 and 𝜆𝑎 are the ⎢ isotropic ⎥
corresponding longitudinal stretches before and after the application of ⎢ equilibrium ⎥
the electric field. 𝜆01 and 𝜆𝑎1 are the stretches contribution due to the ⎢ ⎥
𝑈membrane = 𝜋𝑅2𝑖 𝑡0 ⎢ ⎥. (2)
applied pressure difference and voltage on the DE membrane. ⎢ 𝜇𝛽 ( 2 −2 ) ⎥
⎢ 𝜆1 𝜆𝑣1 + 𝜆22 𝜆−2 𝑣2
+ 𝜆−2 𝜆−2 𝜆2 𝜆2 − 3
1 2 𝑣1 𝑣2 ⎥
2
⎢ ⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟ ⎥
2.2. Material modeling ⎢ ⎥
⎢ isotropic ⎥
⎢ viscous ⎥
To model the visco-hyperelastic behavior of the DE membrane, we ⎣ ⎦
implement a rheological model comprising two hyperelastic springs
in which 𝜇 𝛼 and 𝜇 𝛽 are the shear moduli of the two nonlinear springs.
and one linear dashpot element, as shown in Fig. 1e. Shear moduli of
the two hyperelastic networks are denoted by 𝜇𝛼 and 𝜇 𝛽 , respectively
2.3. The pressure induced deformation
while the Newtonian viscosity (Pa.s) of the dashpot is denoted by 𝜂.
The deformation gradient 𝐹 = 𝑑𝑖𝑎𝑔 (𝜆1 , 𝜆2 , 𝜆−1
1 2
𝜆−1 ) is used to describe
each parallel unit of the model [62]. Total deformation gradient 𝐹 is In the intermediate state, the membrane is bulged outwards by
multiplicatively decomposed into two parts: elastic deformation gradi- applying a fixed pressure difference 𝑃𝑔 . The corresponding stretches are
ent 𝐹𝑘 (associated with nonlinear springs in the Maxwell element) and calculated using the height (𝑧0 ) derived from the applied initial Pressure
viscous deformation gradient 𝐹𝑣 (associated with the viscous damper) differential under the assumption of spherical and homogeneous defor-
as mation of DE membrane [28]. The longitudinal and latitudinal stretches
resulting from the combined effect of pre-stretch and initial pressure
𝐹 = 𝐹𝑘 𝐹𝑣 = 𝑑𝑖𝑎𝑔(𝜆1 , 𝜆2 , 𝜆−1
1 2
𝜆−1 ) difference are expressed as
= 𝑑𝑖𝑎𝑔(𝜆𝑘1 , 𝜆𝑘2 , 𝜆−1 𝜆−1 )𝑑𝑖𝑎𝑔(𝜆𝑣1 , 𝜆𝑣2 , 𝜆−1
𝑘1 𝑘2
𝜆−1 ),
𝑣1 𝑣2
(1) 𝑧𝑎 20 + 𝑟2
𝜆1 = 𝜆2 = 𝜆𝐵 = 𝜆01 𝜆0 = 𝜆0 . (3)
in which 𝜆𝑘𝑖 represent the 𝑖th principal stretch of the nonlinear spring 𝑟2
in the Maxwell element. The principal stretch 𝜆𝑘𝑖 can be expressed It should be noted that the above equation is arrived at by considering
in terms of internal variable 𝜆𝑣𝑖 as 𝜆𝑖 𝜆−1
𝑣𝑖 . The material model is then the membrane deforming homogeneously by maintaining the spherical

3
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

shape. The peripherally clamped membrane, driven by pressure and Fig. 1e represents the recommended rheological model of the com-
electric potential, is indeed in a state of inhomogeneous deforma- bined system of electrically stimulated DE membrane and pseudo air
tion. The state of equal biaxial stretch exists in the polar region of spring. The suggested model clearly indicates the coupling ability of
the inflated membrane, which changes gradually toward the clamped the electrically stimulated motion of the DE membrane with the fluid
boundary. The inhomogeneity is more prominent in the case of higher mass inside the airtight chamber. The voltage-induced deformation in
pressures, or higher electric fields for a given applied pressure [33], the DE membrane caused by Maxwell stress is directed in an out-of-
for which the shape of the inflated membrane differs from that of a plane motion by applied pressure difference, which is further restricted
spherical cap. In the ensuing analysis, a pressure difference of 300 by confined fluid mass. Based on the principle of thermodynamics, it
Pa with the combination of moderate electric potentials (up to 2.5 is possible to calculate the vertical spring stiffness of the air column
kV) result in an average deviation of 3%–5% between experimental within the chamber [60]. The obtained bulge state after the applied
pressure difference is used to determine the stiffness of the air column.
observations and the model predictions, which justifies the use of
The interior air is assumed to behave like an ideal gas in order to
the theoretical development based on the assumption of homogeneous
model the air spring, and the stiffness of the air spring during voltage-
deformations. The pressure-driven state of inflation is expected to be
induced deformation is assumed to be constant. The pseudo air springs
slow, implying that the isothermal quasi-static condition will prevail.
can be modeled by using Hooke’s spring model for small deflections.
The expression for the work done as a result of applied pressure is
This spring model illustrates the rise in air column strain energy caused
( )
𝑃 𝑉 − 𝑃1 𝑉1 𝑇 by electrical stimulation. To model the moderate and large deflection,
𝑊𝑝 = 2 2 𝛾 𝑖 − 1 − 𝑃𝑎𝑡𝑚 (𝑉2 − 𝑉1 ), (4)
𝛾 −1 𝑇𝑜 we have to include other higher-order terms of Taylor series expansion
in the force–deflection characteristic. The force on the air column is
where 𝑇𝑖 is the initial temperature of the air inside the chamber and
determined by multiplying the gauge pressure by the effective area at
𝑇𝑜 is the temperature of the air blown inside the chamber. Since both
any point on the load–deflection curve, i.e., 𝐹𝑧 = 𝑃𝑔 .𝐴𝑒 . Where 𝑃𝑔 , 𝐹𝑧 ,
temperatures are assumed to be equal, the amount of work done is
and 𝐴𝑒 signify the gauge pressure within the chamber, applied load on
given as
the air column, and effective area corresponding to the applied load,
𝑊𝑝 = 𝑃2 𝑉2 − 𝑃1 𝑉1 − 𝑃𝑎𝑡𝑚 (𝑉2 − 𝑉1 ). (5) respectively. The spring stiffness is now calculated as follows
𝑑𝐹𝑧 𝑑𝐴𝑒 𝑑𝑃
Here, 𝑉1 is the chamber volume (𝑉𝑐 ) of the considered actuators. After 𝐾= = 𝑃𝑔 + 𝐴, (10)
𝑑𝑧 𝑑𝑧 𝑑𝑧 𝑒
applying a pressure difference, the entire volume of the air column 𝑉2
where in the case of small deformation, 𝑧 is assumed to represent the
(including the bulge shape of the DE membrane and chamber) is given
maximum vertical deflection of the air column. An adiabatic process
by
is appropriate for relating the volume change and pressure inside the
𝜋( 2 ) airtight chamber during the voltage-driven dynamic phase. For the
𝑉2 = ± 3𝑟 𝑧𝑎0 + 𝑧3𝑎0 + 𝑉𝑐 , (6)
6 specified location of the applied loading, the effective area of the air
where negative and positive sign represents the depressurized and pres- 𝑑𝑉
spring is articulated as = −𝐴𝑒 [60]. Hence, the stiffness of the
surized state of the DE membrane coupled with the bulging actuation, 𝑑𝑧
air spring for the airtight chamber of the bulging actuators can be
respectively. In this study it is assumed that the relaxation due to represented as
applied preload bulging pressure is completed before the electrical ( )2
stimulation, hence the equilibrium stretch 𝜆𝐵 induced due to the ap- ⎡ 𝛾𝑃 1 𝜋 (𝑟2 + 𝑧2 ) ⎤
⎢ 2 𝑎0 ⎥
plied pressure can be evaluated. At equilibrium condition, the total 𝐾=⎢ ( 2 ) − 𝑃𝑔 𝜋𝑧 𝑎0 ⎥ . (11)
1
⎢ ± 𝜋𝑧𝑎0 3𝑟 + 𝑧 ⎥
strain developed is equal to the viscous strain, i.e., 𝜆𝐵𝑣 ||𝑡=0 = 𝜆𝐵 ||𝑡=0 .
2
+ 𝑉
⎣ 6 𝑎0 𝑐 ⎦
For the stated condition, the elastic energy stored in the hyperelastic
Using the aforementioned assumption of Hook’s spring model, the
element corresponding to the Maxwell element is released [63]. Hence,
induced energy in the air column during voltage-induced deformation
the potential energy of the system, when no electric field is applied is
is written as
written as (
⎡ ⎛ 1 ( 2 2
))2 ⎞ ⎤
𝑈𝑇 = 𝑈membrane ||𝑡=0 − 𝑊𝑃 ||𝑡=0 , (7) ⎢ ⎜ 𝛾𝑃 2 𝜋 𝑟 + 𝑧𝑎0 ⎟ ⎥
⎢ ⎜ ( 2 ) − 𝑃𝑔 𝜋𝑧 𝑎0 ⎟ × ⎥
1
⎢ ⎜ ± 𝜋𝑧𝑎0 3𝑟 + 𝑧2 + 𝑉𝑐 ⎟ ⎥
here, 𝑡 = 0 indicates the starting time of electrical stimulation. Based 1⎢ ⎝ 6 𝑎0 ⎠ ⎥
on the principle of minimization of the potential energy, the following 𝑊𝑎 = ⎢ ⎥. (12)
𝜕𝑈𝑇 2⎢ ⎥
(
⎢ ⎛ ( 𝜆 )2 ( ) 1 2
⎥
expression holds true for = 0 as
) 2 ⎞
𝜕𝜆01 ⎢ ⎜ ⎥
𝑎
𝑟2 + 𝑧2𝑎0 − 𝑟2 − 𝑧𝑎0 ⎟
⎢ ⎜ 𝜆𝐵 ⎟ ⎥
( ) 𝑃𝑔 𝑟3 (( )−1∕2 ( )1∕2 ) ⎣ ⎝ ⎠ ⎦
2𝜇 𝛼 𝑅2𝑖 𝑡0 𝜆20 𝜆01 − 𝜆−4 𝜆−5 −
0 01
𝜆01 − 1 + 𝜆01 − 1 = 0. (8)
4 The stretch generate in membrane due to the applied electric field
Longitudinal stretch 𝜆01 correspond to deflation or inflation state with is written as
the application of initial pressure difference can be obtained by solving
𝜆𝑎 = 𝜆𝐵 𝜆𝑎1 , (13)
algebraic Eq. (8). The central deflection of DE membrane is obtained
𝜆
by utilizing Eq. (3). where 𝜆𝑎1 = 𝑎 is voltage induced stretches in DE membrane and
𝜆𝐵
it can further be expressed by using the initial spherical deformation
2.4. Voltage induced deformation assumption and multiplicative decomposition of stretches from the
stretch state prior to applying initial pressure (Fig. 1b) to that stretch
This subsection introduces a framework for the dynamic response of state subsequent to applying an electric voltage (Fig. 1d) as [28]
DE membrane coupled with electrical stimulation, pressure difference,
𝑟2 + 𝑧2𝑎
and confined fluid mass. The total potential energy of the considered 𝜆𝑎1 = . (14)
system is described in terms of membrane strain energy (𝑈membrane ), 𝑟2 + 𝑧2𝑎0
pressure work (𝑊𝑃 ), electrical work (𝑊𝑒 ), and the energy stored in Therefore, the expressions for a total stretch in DE membrane after
pseudo air spring (𝑊𝑎 ) deformation caused by out-of-plane deflection voltage application is expressed by utilizing the Eqs. (13) and (14) as
of DE membrane as
𝑟2 + 𝑧2𝑎
𝜆𝑎 = 𝜆𝐵 . (15)
𝑈𝑇 = 𝑈membrane − 𝑊𝑃 − 𝑊𝑒 + 𝑊𝑎 . (9) 𝑟2 + 𝑧2𝑎0

4
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

The required expression of the membrane stored electrostatic energy of applied internal pressure during voltage-induced deformation and
due to the applied electric signal is 𝐸𝑒 = 0.5𝐶𝜙2 . Similarly, 𝑤𝑒 = 𝐶𝜙2 rewritten as
gives the work done on the membrane due to applied electric voltage ( )
⎡ 1 (( )) ⎤
(𝜙). As a result of membrane deformation, the net quantity of electrical ± 𝜋 3𝑟2 𝑧𝑎0 + 𝑧3𝑎0 + 𝑉𝐶 −
⎢ 6 ⎥
energy of the DE membrane is expressed as ⎢ ⎥
⎢ ( (( 2 )) )𝛾 ⎥
⎢ 1 3 ⎥
𝑊𝑒 = 𝐸𝑒 − 𝑤𝑒 = −0.5𝐶𝜙2 . (16) ± 𝜋 3𝑟 𝑧𝑎0 + 𝑧𝑎0 + 𝑉𝐶 ×
⎢ 6 ⎥
𝜀0 𝜀𝑟 𝐴𝑓 𝑃𝑔 ⎢ ⎥
( )1∕2
𝛾 − 1 ⎢⎢ ⎛
The capacitance of the membrane can be defined as 𝐶 = , where 𝑊𝑝 = (1−𝛾) ⎥ .
𝑡𝑓 ⎛⎛ 2
𝜆 𝑎 ( 2 2
) 2 ⎞⎞ ⎞ ⎥
𝐴𝑓 and 𝑡𝑓 are the final surface area and thickness of the DE membrane ⎢ ⎜ ⎜⎜ 3𝑟 𝜆 𝑟 + 𝑧𝑎0 − 𝑟 ⎟⎟ ⎟ ⎥
⎢ ⎜ 1 ⎜⎜ 𝐵 ⎟⎟ ⎟ ⎥
for the small deformation. The final thickness and surface area of the ⎢ ⎜± 6 𝜋 ⎜⎜ ( )
⎟⎟ + 𝑉𝑐 ⎟ ⎥
spherically deformed membrane can also be written as ⎢ ⎜ ⎜⎜ 𝜆𝑎 ( 2 ) 3∕2 ⎟⎟ ⎟ ⎥
( ) ⎢ ⎜ ⎜ ⎜ + 2
𝑟 + 𝑧𝑎0 − 𝑟 2 ⎟⎟ ⎟ ⎥
𝜆𝑎 𝜋 𝑟2 + 𝑧2𝑎0 ⎣ ⎝ ⎝⎝ 𝜆𝐵 ⎠⎠ ⎠ ⎦
−2
𝑡 𝑓 = 𝑡 0 𝜆𝑎 ; 𝐴 𝑓 = . (17)
𝜆𝐵 (25)
Therefore, the expression of the net quantity of electrical energy of the Further, the corresponding expression of the system total potential
DE membrane is given as energy is determined by inserting the Eqs. (2), (12), (18), and (25) into
( 2 2
) 3 2 Eq. (9), accounting for electrical and pressure effects as given in Box I.
𝜀0 𝜀𝑟 𝐴𝑓 2 1 𝜀0 𝜀𝑟 𝜋 𝑟 + 𝑧𝑎0 𝜆𝑎 𝜙 To investigate the dynamic behavior of bulging actuator, the governing
𝑊𝑒 = −0.5 𝜙 =− , (18)
𝑡𝑓 2 𝑡 0 𝜆𝐵 equation is evaluated by substituting Eqs. (22), (26), and (20), into
where 𝜀0 is the vacuum permittivity , i.e., 8.85 × 10−12 𝐹 𝑚−1 and 𝜀𝑟 is the Lagrange equation stated in Eq. (19). The final expression of the
governing equation is written as in Box II. Further, the internal variable,
the relative permittivity.
i.e., viscous stretch 𝜆̇ 𝑎𝑣 is determined using the expressions of 𝐷 and 𝐿
in the generalized form of the Euler–Lagrange’s (E–L) equation as
2.5. Equations governing the dynamic behavior ( )
𝑑 𝜕𝐿 𝜕𝐿 𝜕𝐷
− + = 0. (28)
The non-linear dynamic governing equation of motion of the bulging 𝑑𝑡 𝜕 𝜆̇ 𝑎𝑣 𝜕𝜆𝑎𝑣 𝜕 𝜆̇ 𝑎𝑣
actuator is evaluated using the Lagrange equation based on the princi- Substituting the expression of 𝐿 and 𝐷 in Eq. (28), the differential
ple of least action for non-conservative systems as evolution equation is written as
( )
𝑑 𝜕𝐿 𝜕𝐿 𝜕𝐷 2𝜇 𝛽 ( −4 3 )
− + = 0, (19) 𝜆̇ 𝑎𝑣 + 𝜆𝑎 𝜆𝑎𝑣 − 𝜆2𝑎 𝜆−3
𝑎𝑣 = 0. (29)
𝑑𝑡 𝜕 𝜆̇ 𝑎 𝜕𝜆𝑎 𝜕 𝜆̇ 𝑎 𝜂
where Lagrangian 𝐿 = 𝑇 − 𝑈𝑇 , in which 𝑇 and 𝑈𝑇 indicate the The developed governing and evolution equations stated in Eqs. (27)
kinetic and potential energies of the actuator under consideration, and (29) are nonlinear. The rate-dependent viscous and hyperelastic
respectively. The term 𝐷 represents the Rayleigh dissipation function nature of the DE membrane, as well as the coupled motion of the
which is defined as [16,57,63] membrane and air spring, form the basis of these nonlinearities. In the
( ) present work of dynamic actuation, initially, the DE membrane is under
𝜂 𝜆̇ 2𝑎𝑣 the combined effect of pre-stretch and pressure. Hence the associated
𝐷 = 𝜋𝑅2𝑖 𝑡0 , (20)
2 three initial conditions of the bulging actuators are
𝑑𝜆𝑎 ||
in which 𝜆̇ 𝑎𝑣 denotes the time derivative of viscous stretch in the prin- 𝜆𝑎 ||𝑡=0 = 𝜆𝐵 ; = 0; 𝜆𝑎𝑣 ||𝑡=0 = 1 . (30)
cipal direction. During the electrically induced buckling expansion, the 𝑑𝑡 ||𝑡=0
DE membranes are assumed to have no latitudinal deformation. Thus, Further, the governing equation, evolution equation, and initial con-
the expression for kinetic energy associated with a bulging actuator is dition are solved numerically in Matlab ODE45 solver to interrogate the
expressed as nonlinear electromechanical responses of the DE membrane coupled
with the airtight chamber.
1 ( )
𝑇 = 𝜌 𝜋𝑅2𝑖 𝑡0 𝑧̇ 2𝑎 . (21)
2 3. Experimental method and fabrication techniques of DE based
Here, 𝜌 represents the density of the DE membrane. The time bulging actuators
derivative of 𝑧𝑎 is obtained from Eq. (15). Using these relationships,
the expanded expression for kinetic energy is rewritten as The experimental setup used to characterize the bulging actua-
( 2 )2 ( )−1 ⎤ tors consists of a high voltage amplifier (Matsusada Precision Inc,
⎡ 𝑟 + 𝑧2𝑎0 𝜆𝑎 ( 2 )
1 NF BA4825) in conjunction with a single static head Laser Doppler
𝑇 = ⎢𝜌𝜋𝑅2𝑖 𝑡0 𝑟 + 𝑧2𝑎0 − 𝑟2 𝜆̇ 2𝑎 ⎥ . (22)
8⎢ 𝜆2 𝜆𝐵 ⎥ Vibrometer (LDV, Polytec, OFV534, Waldbronn, Germany) as shown
⎣ 𝐵 ⎦
in Fig. 2a. Based on the ambient pressure, the pressure inside the
In the case of electrical stimulation, the work done by the internal chamber of the bulging actuators is measured using the pressure sensor
pressure is written as (BMP280). The 3-D model of bulging actuators depicted in Fig. 2b is
𝑃𝑔 ( (1−𝛾)
) made up of a cylindrical chamber and flange that were printed using a
𝑊𝑝 = 𝑉2 − 𝑉2𝛾 𝑉2𝑓 , (23) 3D printer (Ultimaker Cura). For investigating the influence of confined
𝛾 −1
fluid mass on the electromechanical response of DE membrane, three
where 𝑉2𝑓 is the volume inscribed by the air column due to voltage- distinct sizes of cylindrical chambers with different volumes (𝑉𝑐 =
induced deformation in the DE membrane and is represented as 100 ml, 𝑉𝑐 = 400 ml, and 𝑉𝑐 = 740 ml) are considered. The model
[ (( ( )1∕2 specifications are listed in Table 1. After pre-stretching to 240%, the
1 𝜆𝑎 ( 2 )
𝑉2𝑓 = 𝑉𝑐 ± 𝜋 3𝑟2 𝑟 + 𝑧2𝑎0 − 𝑟2 DE membrane (3M VHB 4910) is attached to the chamber with the
6 𝜆𝐵
help of two circular rings pressed by a flange. The compliant electrode
( )3∕2 ))]
𝜆𝑎 ( 2 ) like carbon grease (MG Chemical 846–80 G, MG Chemicals Surrey, BC,
+ 𝑟 + 𝑧2𝑎0 − 𝑟2 (24) Canada) is coated on both sides of the DE membrane before attaching
𝜆𝐵
it to the final structure. The electromechanical dynamic responses of
Hence, by invoking the expressions from the Eqs. (6) and (24) into all DEAs were recorded by applying an AC or DC input voltage signal
Eq. (23), we can get the expression for the work done as a result on an initially inflated and deflated bulging states.

5
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

⎡ ⎛ ⎞ ⎤
⎢ ⎜ ⎟ ⎥
⎢ ⎜ ⎟ ⎥
⎢ ⎜ 𝜇𝛼 ( ) 𝜇 𝛽 ( )⎟ ⎥
⎢ 𝜋𝑅2𝑖 𝑡0 ⎜ 2𝜆2𝑎 + 𝜆−4 −3 + 2𝜆2𝑎 𝜆−2 + 𝜆−4 𝜆4𝑎𝑣 − 3 ⎟ ⎥
⎢ ⎜ 2 𝑎 2 𝑎𝑣 𝑎 ⎟ ⎥
⎢ ⎜⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟ ⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⎟ ⎥
⎢ ⎜ isotropic isotropic ⎟ ⎥
⎢ ⎜ ⎟ ⎥
⎢ ⎝ equilibrium viscous ⎠ ⎥
⎢ ⎥
⎢ ( )2 ⎥
⎢ ⎛ 𝛾𝑃 1 𝜋 (𝑟2 + 𝑧2 ) ⎞ ⎛(( )2 )1
⎞
2
⎥
⎢ 1⎜ 2 𝑎0 ⎟⎜ 𝜆𝑎 ( 2 ) 2
⎥
+ ⎜ − 𝑃𝑔 𝜋𝑧𝑎0 ⎟ 2
𝑟 + 𝑧𝑎0 − 𝑟 2
− 𝑧𝑎0 ⎟
𝑈𝑇 = ⎢ ( ) ⎥
⎟ ⎜⎝ ⎟
2⎜ 1 𝜆𝐵 (26)
⎢ ± 𝜋𝑧 3𝑟2 + 𝑧2𝑎0 + 𝑉 𝑐 ⎠ ⎥
⎢ ⎝ 6 𝑎0 ⎠ ⎥
⎢ ⎥
( ) ( )𝛾
⎢
⎢ ⎛ ± 1 𝜋 ((3𝑟2 𝑧 + 𝑧3 )) + 𝑉 − ± 1 𝜋 ((3𝑟2 𝑧 + 𝑧3 )) + 𝑉 × ⎞ ⎥
⎥
⎢ ⎜ 6 𝑎0 𝑎0 𝐶
6 𝑎0 𝑎0 𝐶
⎟ ⎥
𝑃
⎢ − 𝑔 ⎜ ⎟ ⎥
⎢ 𝛾 −1 ⎜ ( (( ( )1∕2 ( )3∕2 )) )(1−𝛾) ⎟ ⎥
⎢ ⎜ 1 𝜆𝑎 ( 2 ) 𝜆𝑎 ( 2 ) ⎟ ⎥
⎜ ±6𝜋 3𝑟2 𝑟 + 𝑧2𝑎0 − 𝑟2 + 𝑟 + 𝑧2𝑎0 − 𝑟2 + 𝑉𝐶 ⎟
⎢ ⎝ 𝜆𝐵 𝜆𝐵 ⎠ ⎥
⎢ ⎥
⎢ ( 2 ) 3 2 ⎥
⎢ 2
1 𝜀0 𝜀𝑟 𝜋 𝑟 + 𝑧𝑎0 𝜆𝑎 𝜙 ⎥
⎢ − . ⎥
⎣ 2 𝑡0 𝜆𝐵 ⎦

Box I.

( )2
⎡ ⎛ 𝛾𝑃 𝜋 1 (𝑟2 + 𝑧2 ) ⎞ ⎤
⎢ ( ) ( ) 1 ⎜ 2 𝑎0 ⎟ ⎥
⎢ 2𝜇 𝛼 𝜆𝑎 − 𝜆−5 + 𝜇 𝛽 𝜆𝑎 𝜆−2 −5 4
𝑎𝑣 − 𝜆𝑎 𝜆𝑎𝑣 + ⎜ − 𝑃𝑔 𝑎0 ⎟
𝑧 ⎥
2𝑅2𝑖 𝑡0 𝜆𝐵 ⎜ ± 1 𝜋𝑧 (3𝑟2 + 𝑧2 ) + 𝑉 𝑐
𝑎
⎢ ⎟ ⎥
⎢ ⎝ 6 𝑎0 𝑎0 ⎠ ⎥
⎢ ⎥
⎢ (( )1∕2 )( )−1∕2 ⎥
⎢ 𝜆𝑎 ( 2 ) 𝜆𝑎 ( 2 ) ( 2 ) ⎥
⎢ × 𝑟 + 𝑧2𝑎0 − 𝑟2 − 𝑧𝑎0 𝑟 + 𝑧2𝑎0 − 𝑟2 𝑟 + 𝑧2𝑎0 ⎥
𝜆𝐵 𝜆𝐵
⎢ ⎥
⎢ ⎥
⎢ ( )−1∕2 ⎥
⎢ ⎛ 𝜆𝑎 ( 2 ) ⎞ ⎥
( ) ⎜ ±𝑟 2 𝑟 + 𝑧2𝑎0 − 𝑟2 ⎟
⎢ 𝑃𝑔 𝑟2 + 𝑧2𝑎0 ( 1 (( )𝛾 ⎜ 𝜆𝐵 ⎥
⎢ )) ⎟ ⎥
⎢ − ± 𝜋 3𝑟2 𝑧𝑎0 + 𝑧3𝑎0 + 𝑉𝐶 ⎜ ⎟ ⎥
4𝑅2𝑖 𝑡0 𝜆𝐵 6 ⎜ ( )1∕2 ⎟
⎢ 𝜆𝑎 ( 2 ) ⎥ = 0. (27)
⎢ ⎜ ± 𝑟 + 𝑧2𝑎0 − 𝑟2 ⎟ ⎥
⎢ ⎝ 𝜆𝐵 ⎠ ⎥
⎢ ( (( ( ⎥
⎢ )1∕2 ( )3∕2 )) )−𝛾 ⎥
⎢ 1 𝜆𝑎 ( 2 ) 𝜆𝑎 ( 2 ) ⎥
× ± 𝜋 3𝑟2 𝑟 + 𝑧2𝑎0 − 𝑟2 + 𝑟 + 𝑧2𝑎0 − 𝑟2 + 𝑉𝐶
⎢ 6 𝜆𝐵 𝜆𝐵 ⎥
⎢ ⎥
⎢ ⎥
⎢ ( 2 2
) 2 2 ( 2 )
2 3( )−2 ⎥
⎢ 3 𝜀0 𝜀𝑟 𝑟 + 𝑧𝑎0 𝜆𝑎 𝜙 1 𝑟 + 𝑧𝑎0 𝜆𝑎 ( 2 ) ⎥
⎢ − − 𝜌 𝑟 + 𝑧2𝑎0 − 𝑟2 𝜆̇ 2𝑎 ⎥
2 𝑅2𝑖 𝑡20 𝜆𝐵 8 𝜆3𝐵 𝜆𝐵
⎢ ⎥
⎢ ⎥
⎢ ( 2 )
2 2( )−1 ⎥
⎢ 1 𝑟 + 𝑧𝑎0 𝜆𝑎 ( 2 ) ⎥
⎢ + 𝜌 𝑟 + 𝑧2𝑎0 − 𝑟2 𝜆̈ 𝑎 ⎥
⎣ 4 𝜆2𝐵 𝜆𝐵 ⎦

Box II.

Table 1
Geometrical parameters of bulging actuators.
Geometrical Parameters Sample 1 Sample 2 Sample 3
Final diameter of DE membrane (mm) 50 50 50
Initial pre-stretch 2.4 2.4 2.4
Volume of the air chamber (𝑉𝑐 ) (ml) 100 400 740
Height of the air chamber (mm) 50 55 100

6
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 2. (a) Experimental approach for assessing the actuation response of bulging actuators, and (b) 3-D image of bulging actuator.

Table 2 to zero. Further, the resulting form of Eq. (31) is employed in the
Material parameter of VHB4910. governing Eq. (27) to obtain the DC dynamic response as represented in
Material parameter VHB4910 the plot of Fig. 3. The depressurized and pressurized mode of actuation
𝜀 (F/m) 4.1 × 10−11 is considered under the influence of varying air column capacity with
𝜇𝛼 (kPa) 20 the pressure difference of 𝑃𝑔 = 300 Pa. The variability of the air
𝜇𝛽 (kPa) 40
𝜌 (kg∕m3 ) 1200
column capacity is taken into account by considering the geometrical
configuration in accordance with samples 1, 2, and 3 as indexed in
Table 1. The DC voltage is taken in the range 2000 to 5000 V with
the increment step of 200 V. The voltage-induced central deflections
4. Results and discussions for each incremental step of DC voltage are recorded after the settling
time of dynamic oscillation while ignoring the associated viscoelastic
This section discusses the reliability of the proposed model by creep. For an air chamber capacity of 𝑉𝑐 = 100 ml, the experimentally
corroborating the analytical model with experimental results in the observed DC response of the central displacement follows well with the
instance of bulging actuators for various model specifications as men- obtained analytical results in both pressurized and depressurized states
tioned in Section 3 and listed in Table 1. Further, the present section (Figs. 3a–b). However, it is found that the analytically predicted values
highlights the various analytical and experimental characterization of of the central deflection are less as compared with the experimental
inflated or deflated DE-based bulging actuators with varying chamber value for samples 2 and 3 (for 𝑉𝑐 = 400 and 740 ml) under the influence
volumes. The DE membrane (VHB 4910) required material parameters of both pressurized and depressurized states of actuation. This can be
for further electromechanical analysis of air-coupled behaviors of in- due to the fact explained as: the inflated membrane whose periphery
flated DE membrane are indexed in Table 2 [64–66]. For evaluating is fixed leads to inhomogeneous deformation whereas the governing
the influence of membrane viscoelasticity, the viscosity parameter (𝜂) equations derived in Section 2 is taking the assumption of homogeneous
are considered to be 100 Pa-s [64]. deformation. The homogeneous deformation of the inflated membrane
is well suitable for small out-of-plane deformation which arises at the
4.1. Comparison of electromechanical response obtained from the analyti- relatively lower voltage and chamber volume (𝑉𝑐 = 100 ml) and further
cal model and experimental observations goes into the inhomogeneous deformation regime at higher voltage and
chamber volume (𝑉𝑐 = 400 and 740 ml). This is the reason why the
This subsection provides the comparison of electromechanical–air experimental result deviates from the analytical solution at a higher
coupled responses of the developed model for the different sets of voltage with the combination of relatively larger chamber volume (𝑉𝑐
chamber volume by implementing the analytical formulation discussed = 400 and 740 ml) but matches well at 𝑉𝑐 = 100 ml.
in Section 2 and experimental observations. The capability of the
model is investigated by using both DC and AC electrical loading in 4.1.2. Response of AC electrical loading
depressurized and pressurized states of actuation. The following is a This subsection examines the dynamic behavior of time dependent
general representation of a time dependent electrical signal electrical loading on the inflated DE membrane coupled with an airtight
chamber of bulging actuator for pressurized and depressurized state.
𝜙(𝑡) = 𝜙𝑑𝑐 + 𝜙𝑎𝑐 (𝑡), (31)
The electromechanical responses of the proposed model are corrobo-
where 𝜙𝑑𝑐 is the offset DC voltage and 𝜙𝑎𝑐 (𝑡) is the time-dependent AC rated using experimental data obtained from all three samples listed
voltage waveform. The offset DC voltage 𝜙𝑑𝑐 shifts the initial equilib- in Table 1 to examine the influence of variable air volume. The DE
rium position, and 𝜙𝑎𝑐 (𝑡) imparts a dynamic component of actuation membrane of the considered bulging actuators are subjected to the
about the equilibrium position. combined effect of both constant pressure (𝑃𝑔 = 300 Pa) and time-
varying electrical loading. This subsection presents the experimental
4.1.1. Response of DC electrical loading and analytical studies using a sine voltage waveform as a time-varying
The influence of DC loading on the dynamic behavior of the inflated electrical loading with an input frequency apart from the resonance
membrane coupled with the confined air column is presented in this frequency. To investigate the nonlinear dynamics of the associated os-
subsection. Therefore, in Eq. (31), we employed an electrical signal in cillation, we considered both with and without offset loading condition
the form of the Heaviside step function with a constant DC voltage of the harmonic signal. The amplitude of AC voltage 𝜙𝑎𝑐 (𝑡) was fixed
(𝜙𝑑𝑐 ) and the harmonic component of the electrical signal (𝜙𝑎𝑐 ) set at 2.5 kV for the DC offset of 𝜙𝑑𝑐 = 0 and 2.5 kV. The associated

7
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 3. Comparison of analytical and experimental DC dynamic response in (a) a pressurized, and (b) a depressurized state of actuation of a DE membrane coupled with the
bulging actuator.

Fig. 4. Comparison of analytical and experimental time history response of DE membrane coupled with bulging actuator with varying chamber volume obtained from a pressurized
state of actuation, for offset DC voltage (a) 𝜙𝑑𝑐 = 0 kV, and (b) 𝜙𝑑𝑐 = 2.5 kV.

Fig. 5. Comparison of analytical and experimental time history response of DE membrane coupled with bulging actuator with varying chamber volume obtained from a depressurized
state of actuation, for offset DC voltage (a) 𝜙𝑑𝑐 = 0 kV, and (b) 𝜙𝑑𝑐 = 2.5 kV.

input voltage harmonic frequency is kept at 1 Hz. The result shows oscillation is nearly doubled in the analytical and experimental scale.
that the electromechanical responses obtained from experimental and This is evident in the output responses with zero offsets in Figs. 4a and
analytical simulation in pressurized and depressurized mode of ac- 5a for pressurized and depressurized mode of actuation, respectively.
tuation are opposite in sign, representing the inward and outward However, the frequency of oscillation decreases and becomes almost
motion, respectively as shown in Figs. 4 and 5. The experimental equal to the input signal when the membrane is stimulated with an
values for samples 1, 2, and 3 are quite near to the predicted values, offset closer to the AC component as indicated in Figs. 4b and 5b. This
as per the developed governing equations. Hence, it can be evident can be due to the fact that as the frequency of oscillations decreases
that the developed equivalent spring-based model effectively captures with increase in the offset DC voltage due to electrostatic softening and
the coupling behavior of electromechanical dynamic response with the the developed model successfully capture this effect.
confined fluid/air mass. If the actuators are stimulated with a harmonic The dynamic responses obtained from AC loading in both depressur-
signal without an offset DC voltage, the frequency of output dynamic ized and pressurized states for the given load-voltage range enhanced

8
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 6. The percentage increment of voltage-induced deflection of DE membrane coupled with a bulging actuator having chamber size 𝑉𝑐 = 400 and 740 ml compared with the
chamber size of 𝑉𝑐 = 100 ml for a given AC voltage loading of (𝜙𝑑𝑐 = 0 and 2.5 kV with 𝜙𝑎𝑐 = 2.5 kV) in (a) pressurized, and (b) depressurized state.

Fig. 7. Comparison of analytical and experimental resonance behavior of DE membrane coupled with a bulging actuators of chamber volume (a) 𝑉𝑐 = 100 ml, (b) 𝑉𝑐 = 400 ml,
and (c) 𝑉𝑐 = 740 ml.

with the increase in the chamber volume, as evidenced by the exper- of actuation. For this analysis, an offset DC voltage of 𝜙𝑑𝑐 = 2.5
imental and analytical results, with sample 1 shows the lowest and kV is applied. Figs. 7 (a–c) depicts the analytical and experimental
sample 3 shows the highest value of the central deflection. When the frequency response for the given samples with chamber volumes of 𝑉𝑐
chamber volume increases from 100 ml to 740 ml, the percentage = 100, 400, and 740 ml, and their corresponding resonance peak are
increment of the central deflection increases in both pressurized and listed in Table 3. The deflection indicated in the frequency response
depressurized states as shown in Figs. 6(a–b). This is because as the curve (Fig. 7) obtained from the analytical simulation and experimental
chamber volume increases the stiffness of the air column decreases, and observation is normalized with the peak value. The first resonance
the restriction offered by the confined air column during out-of-plane peak revealed by the analytical simulation falls inside the interval
electromechanical deformation decreases. The maximum percentage of the experimental observation for a given chamber volume of 𝑉𝑐
increment is found in the depressurized state of actuation for the = 100 ml. For the chamber volume of 𝑉𝑐 = 400 and 740 ml, a
chamber size of 𝑉𝑐 = 740 ml. It is nearly 316% and 347%, respectively small discrepancy is seen in the first resonance peak with the average
obtained from the experimental and analytical analysis at the given error of about 24% and 26%, respectively. This is probably due to
offset of 𝜙𝑑𝑐 = 2.5 kV. the several influencing parameters like electrode mass, environmental
conditions such as temperature and humidity effect, strain-stiffening
4.1.3. Resonance behavior effect, and inhomogeneous deformation that arises in case of higher
Generally, the dynamic analysis at various frequencies is needed chamber capacity, etc., which is not incorporated while investigating
to examine how frequency affects the amplitude of the displacement natural frequencies using the developed theoretical model. As a result,
response. Therefore, the dynamic analysis was performed using a burst predicting resonating frequency with the developed model is very much
chirp signal. Since the dielectric constant is nearly remains constant suitable at smaller chamber volume.
below 1 kHz [67], the input frequency is limited to 1 kHz in this study.
The dynamic response of the considered actuators using a burst chirp 4.1.4. Pressure differential loss of confined fluid mass in electrical loading
signal can reveal the effect of the entire frequency on displacement The loss of pressure differential in the airtight chamber of the given
in a single experiment. In general, the analysis of frequency response actuators during voltage-induced deflection determined the amplitude
with the burst chirp, the amplitude of the AC voltage should be kept of the dynamic oscillation in the DE membrane. Pressure differential
at 10%–15% of the DC voltage to avoid breakdown due to resonance. loss during electrical loading can be affected by various factors. The
The dynamic frequency analysis is represented by the pressurized mode experimental investigation predict that the applied initial inflation

9
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Table 3
Resonance peaks of DE membranes coupled with bulging actuators with different chamber volume.
Volume of the chamber Experimental Analytical
F1 (Hz) F2 (Hz) F3 (Hz) F1 (Hz) F2 (Hz) F3 (Hz)
𝑉𝑐 = 100 ml 300–325 495–505 600–610 316 706 728
𝑉𝑐 = 400 ml 240–260 370–380 565–575 190 – –
𝑉𝑐 = 740 ml 210–230 360–370 570–590 162 – –

Fig. 8. Effect of inflation pressure on the pressure differential loss during electrical loading for a given chamber volume (a) 𝑉𝑐 = 100 ml, (b) 𝑉𝑐 = 400 ml, and (c) 𝑉𝑐 = 740 ml.

pressure difference is one of the important factors. A set of experi- 4.2. Electromechanical responses under DC and AC loading evolution with
ments were performed to evaluate the loss of pressure during electrical variable air column
loading owing to the influence of initial pressure difference. In the
experimental front, the DC voltage signal with a magnitude of 5 kV This subsection investigate the variable air column effect on the
is applied on the inflated DE membrane in the pressurized state, and electromechanical response of bulging actuators. Effect of quantity
the pressure drop caused by the applied voltage is measured. For and quality of the air present inside the chamber are investigated by
chamber volume of 𝑉𝑐 = 100 ml, 400 ml, and 740 ml, eight such data developed analytical model.
are recorded, with an increment of pressure difference 𝑃𝑔 = 200 Pa
ranging from 100 Pa to 1500 Pa. A similar result is obtained from the 4.2.1. Effect of quantitative variation of the fluid column
analytical model in which loss of pressure is calculated by utilizing It is vital to analyze how the amplitude of deflection changes as
adiabatically change of air column volume during electrical loading. chamber volume changes in order to explore the dynamic behavior of
The change of the air volume inscribed by actuators during electrical inflated bulging actuators under a variety of electric fields and chamber
loading is evaluated by using maximum central deflection of spherically volumes. The deflection-volume characteristic of DE membrane cou-
deformed inflated DE membrane obtained from the governing equation pled with the airtight chamber obtained from the analytical simulation
stated in Eqs. (27) and (29). The loss of pressure differential due at a given harmonic sine voltage waveform (𝜙𝑑𝑐 = 2500 V and 𝜙𝑎𝑐
to the applied voltage is plotted against the initial inflation pressure = 2500 V) and varied initial inflation pressures (𝑃𝑔 ) as represented
difference and shown in Figs. 8(a–c) for the chamber volume of 𝑉𝑐 in Fig. 9a. Similarly, the deflection-volume characteristic for various
= 100 ml, 400 ml, and 740 ml. The experimental results show that applied DC voltages at a given inflation pressure difference of 𝑃𝑔 =
the pressure differential loss observed in the given actuation field of 300 Pa is illustrated in Fig. 9b. The voltage-induced central deflection
the considered actuators varied almost linearly with initial bulging first increases rapidly as the chamber volume increases but eventually,
pressure. The analytical model predicts the similar trends as observed it reaches a saturation state at a certain value in the sufficiently larger
in the experimental investigation. volume, as shown in Figs. 9a and b. Consequently, the enhancement

10
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 9. Influence of chamber volume variability on the central deflection of DE-based bulging actuators (a) for a given harmonic sine voltage waveform (𝜙𝑑𝑐 = 2500 V and 𝜙𝑎𝑐
= 2500 V) with varying level of initial inflation pressure 𝑃𝑔 = 300, 600 and 900 Pa; and (b) for a given inflation pressure of 𝑃𝑔 = 300 Pa with varying potential difference 𝜙𝑑𝑐
= 2000, 4000 and 6000 V.

of the deflection due to rise in the volume is ineffective at larger

chamber volume. Fox et al. [32] reported a similar trend based on
experimental investigation. The reason for this can be; at a sufficiently
larger chamber capacity, the resistance offered by the air column
during electrical stimulation of the DE membrane is negligible, and
hence the influence of the air column can be neglected. This allows the
free movement of the DE membrane. As a result, it attains the highest
possible deformation, determined by the combined influence of applied
pressure and electrical loading. However, at low chamber volume, the
stiffness of the air spring is very high, and therefore the voltage-induced
deflection is relatively low.
Further, the voltage-induced (𝜙𝑑𝑐 = 2500 V, and 𝜙𝑎𝑐 = 2500 V)
time history response of bulging actuators obtained by utilizing the
derived theoretical model, in both situations, i.e., with (𝑉𝑐 = 100 ml)
and without the influence of the air column at 𝑃𝑔 = 300 Pa, is shown
in Fig. 10. The situation, without air column effect, is obtained by
considering the total potential energy (𝑈𝑇 stated in Eq. (9)) without
including the energy of the air spring (𝑊𝑎 ). The electromechanical
Fig. 10. AC Dynamic response of DE-based bulging actuator with and without air
response of the DE membrane without considering the air column effect column effect.
is quite significant (amplitude of central displacement = 333 μm) when
compared to the response obtained with the influence of the air column
for a given chamber capacity (amplitude of central displacement = pressure difference 𝑃𝑔 = 300 Pa and chamber volume 𝑉𝑐 = 100 ml,
42 μm) as illustrated in Fig. 10. Without the air column effect, there is the electromechanical response of a bulging actuator filled with various
no restriction owing to air spring during voltage-induced deformation types of gases under AC and DC electrical loading, as determined by an
of an inflated DE membrane, and it is possible to achieve the maximum analytical model, is illustrated in Figs. 11(a–b) and 12(a–b). The time
voltage-induced deformation of combined loading of pressure differ- history response (Figs. 11a and 12a) of with and without pre-stretching
ence and electric field. This scenario is analogous to the condition when shows that voltage induced deflection of the DE membrane under AC
the chamber volume of the investigated actuator is kept sufficiently loading (𝜙𝑑𝑐 = 2.5 kV and 𝜙𝑎𝑐 = 2.5 kV) is in increasing order when
large (𝑉𝑐 = ∞), as mentioned earlier. The voltage-induced central dis- the chamber is filled with monoatomic (𝛾 = 1.67), polyatomic linear
placement obtained from the deflection-volume characteristic (Fig. 9a) (𝛾 = 1.4), and polyatomic nonlinear gases (𝛾=1.33). It is also inferred
at the saturated state is around 333 μm, which is exactly the same from Figs. 11a and 12a that the amplitude of oscillation of the central
as the deflection obtained without the effect of the air column for deflection increases due to the effect of pre-stretching. This is because
the identical electric field (𝜙𝑑𝑐 = 2500 V, and 𝜙𝑎𝑐 = 2500 V) and as the pre-stretching increases, the thickness of the DE membrane
pressure difference (𝑃𝑔 = 300), as shown in Fig. 10. Hence, the response decreases. Therefore, the electric field rises for a given voltage which
of the DE membrane coupled with the airtight chamber with either turns to increases the Maxwell stress [68,69]. The increase of Maxwell
a sufficiently high chamber volume or without the influence of air stress enhanced the electromechanical response of the DE membrane,
and the reported model successfully captured this effect. On the other
column are interchangeable, and both arise the limiting condition of
hand, the voltage-deflections characteristic obtained by DC loading gets
deflection for the given electric field and pressure difference.
steeper as the adiabatic constant decreases, as shown in Figs. 11b and
12b. Thus, the results demonstrate that as the adiabatic gas constant of
4.2.2. Effect of qualitative variation of the fluid column and membrane the filled gas increases, the tendency of electromechanical deformation
pre-stretching of the DE membrane decreases, and vice versa, i.e., it is maximum in
The DE membrane electromechanical actuation response is also the case of polyatomic nonlinear gases and minimum in monoatomic
altered by the gas quality filled inside the chamber of the bulging ac- gases. We also found from the DC response (as represented in Figs. 11b
tuator when the size of the chamber is in permissible range to activate and 12b.) that the mechanically pre-stretched membrane required less
the air spring. According to Eq. (11), it is clear that the air column electric field to achieve the desired level of electromechanical actuation
stiffness is a direct function of the adiabatic gas constant. For a given for a given quality of the confined fluid.

11
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 11. Effect of enclosed fluid quality on electromechanical actuation of the bulging actuator without membrane pre-stretching under (a) AC, and (b) DC loading.

Fig. 12. Effect of enclosed fluid quality on electromechanical actuation of the bulging actuator with membrane pre-stretching of 𝜆0 = 2.4 under (a) AC, and (b) DC loading.

Fig. 13. The cumulative effect of pressure and voltage on the central deflection of DE-based bulging actuator (a) with, and (b) without considering the enclosed air column effect.

4.2.3. Cumulative effect of pressure and voltage will changes the curvature of the membranes. The cumulative behavior
The cumulative influence of pressure differential and input voltage of both scenarios is almost similar, but the response obtained without
on the actuation of the DE membrane coupled with the confined considering the effect of the air column is magnified in nature. From
fluid/air mass is presented in this subsection. The analysis is restricted the response curve, it is also observed that as the pressure difference
to two extreme states, such as with and without considering the air increases, the output response of the membrane varies almost linearly
column effect, and represented in Figs. 13a and b, respectively. When
with a positive slope in both the considered cases. However, when the
the effect of the air column is taken into account, the chamber volume
input voltage is relatively high, the electromechanical response enters
is considered as 𝑉𝑐 = 100 ml. The absence of the air column effect
the nonlinear region, and a drastic increase in displacement is noted
simply represents the large size of the chamber where the influence
of air spring becomes negligible, as mentioned in Section 4.2.1. The as the pressure difference rises. The variability of the voltage induced
variability of the center deflection is represented by linearly increas- deformation associated with both the membranes is investigated col-
ing voltage signals at a rate of 5000 V/s. The changing nature of lectively with the nature of considered DE films at a given pressure
the curve for both cases, i.e., with and without the effect of the air range of 200 to 600 Pa, and a voltage range of 0 to 5000 V. The
column demonstrates how variations in pressure and applied voltage appropriate electromechanical region of the specified actuators can be

12
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 14. Voltage-stretch response of DE-based bulging actuators with varying level of inflation pressure for a given chamber volume (a) 𝑉𝑐 = 100 ml, (b) 𝑉𝑐 = 400 ml, (c) 𝑉𝑐 =
800 ml, and (d) 𝑉𝑐 = ∞.

800 ml, and 𝑉𝑐 = ∞) with varying inflation pressure (𝑃𝑔 ), and rep-
resented in Fig. 14. The case with 𝑉𝑐 = ∞ represents the condition
without considering the effect of the air column. It implies that the
stretch of the DE membrane rises with increasing voltage. In the case
of lower inflation pressure, the slope of the voltage-stretch response for
the given chamber volume is steeper as compared with higher inflation
pressure. Although, it is also inversely related to the variability of the
chamber volume for given inflation pressure. Lesser slope in the higher
inflation pressure for a given chamber volume indicates a favorable
condition for electromechanical deflection. Whereas lesser slope in
larger chamber volume for a given pressure indicates less restriction
by confined air inside the chamber, as a result, more electromechan-
ical deflection. Generally, large deflection leads to instability in the
elastomeric film [70,71]. The sharp bending of the voltage-stretch
curve represents the instability condition as clearly seen in Fig. 14.
The voltage required for the sharp bending decreases as chamber
volume increases. The electromechanical instability also depends on the
inflation pressure. The required breakdown voltage corresponding to
inflation pressure for a different set of the chamber size is represented
Fig. 15. Effect of inflation pressure on the breakdown voltage of DE membrane coupled in Fig. 15. The curve indicates that as the pressure increases the
with variable chamber capacity of the bulging actuators.
breakdown voltage of the given actuator decreases. Similarly, for a
given inflation pressure, the operating range of the voltage loading is
higher for the less chamber volume, and this is minimum when the
simply interpreted from the resulting curve for the given size of the chamber volume of the actuators is very large. This is because as the
chamber, depending on the required application. chamber volume decreases, the response of the DE membrane decreases
due to the large restriction offered by confined air. As a result, for the
given loading condition, the deflection of a particular membrane is less
4.2.4. Dynamic stability of DC and AC loading than its capabilities, thereby increasing the voltage required to achieve
To better understand the impact of the air column effect and in- sufficient deformation for breakdown.
flation pressure on the electromechanical response of DE membrane Next, dynamic stability and periodicity are important factors that
coupled with a bulging actuator, the voltage-stretch response is ob- determines the performance of the considered actuators. Fig. 16 rep-
tained analytically for different sizes of the chamber (𝑉𝑐 = 100, 400, resents the phase portraits of DC loading (𝜙𝑑𝑐 = 5000 kV) obtained

13
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 16. Phase portraits of DE membrane coupled with bulging actuator under DC loading with varying level of inflation pressure 𝑃𝑔 = 300, 600 and 900 Pa with (a–c) chamber
volume 𝑉𝑐 = 100 ml, and (d–f) chamber volume 𝑉𝑐 = ∞.

from the analytical simulation with varying level of operating pressure Further, we study the effect of the air column and inflation pres-
(𝑃𝑔 = 300, 600, and 900 Pa). In which, Figs. 16(a–c) indicate phase sure on the nonlinear dynamic stability under AC loading. The same
portraits obtained for a chamber volume of 𝑉𝑐 = 100 ml, whereas aforementioned condition is used to investigate the stability transition.
Figs. 16(d–f) indicate phase portraits obtained for a sufficiently large Poincaré maps and the phase portraits are drawn for a particular
chamber volume (𝑉𝑐 = ∞), where the influence of the air column excitation frequency of 100 Hz and AC loading of 𝜙𝑎𝑐 = 2500 kV with
can be neglected and allows the free movement of the DE membrane. offset DC voltage of 𝜙𝑑𝑐 = 2500 kV. By segmenting the phase-plane
It is observed from Fig. 16, as pressure 𝑃𝑔 increases, DE membranes diagram at each exciting load period, the Poincaré map was developed.
required a relatively less time to attain an equilibrium state in both the The envelope of the phase portraits which indicates the magnitude of
considered actuators. Similarly, the chamber volume has a significant a minimum and maximum stretch value of the DE membrane increases
impact on the time required to reach equilibrium. Lower chamber as the chamber volume increases for the given inflation pressure. Simi-
volume (𝑉𝑐 = 100 ml) required more time to reach equilibrium than larly, the size of the envelope increases as inflation pressure increases.
higher chamber volume (𝑉𝑐 = ∞) for a given pressure as shown in The phase path is a nearly closed-loop for both chamber sizes, indicate
Figs. 16(a–c) and Figs. 16(d–f). It is also inferred that the chamber the stable vibration of he system. From the phase portrait shown in
volume influences the degree of nonlinearity as illustrated Fig. 16. A Fig. 17, it is evident that the actuators with 𝑉𝑐 = 100 ml take longer to
higher value of 𝑉𝑐 implies lower nonlinearity compared to a smaller settle to the equilibrium position than those with 𝑉𝑐 = ∞. The Poincaré
value of 𝑉𝑐 (100 ml). maps which appear to be distorted for the lower value of inflation

14
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

Fig. 17. Poincaré maps along with phase portraits of DE membrane coupled with bulging actuator under AC loading with varying level of inflation pressure 𝑃𝑔 = 300, 600 and
900 Pa with (a–c) chamber volume 𝑉𝑐 = 100 ml, and (d–f) chamber volume 𝑉𝑐 = ∞.

pressure and chamber volume indicate that the inflated air coupled coupling, the membrane viscoelastic effect, and transient electric load-
dielectric-based actuators exhibit aperiodic oscillation, i.e., the mem- ing. The coupling behavior of voltage-induced deformation with the air
brane experiencing chaotic actuation [72,73]. Whereas Poincaré map of column is obtained by incorporating the concept of pseudo air spring in
the DE membranes coupled with larger chamber volume (𝑉𝑐 = ∞) form conjunction with the Euler–Lagrange equation leading to the governing
the close loop at the higher pressure 𝑃𝑔 = 600 and 900 Pa as shown equations of dynamic equilibrium. Finally, considerable experimental
in Figs. 17(e–f), revealing that the system possesses quasi-periodic work was conducted to explore the capabilities of the established
oscillations [48,74,75], when the influence of air is negligible. analytical model. The proposed nonlinear dynamic model satisfac-
torily predicts the voltage-induced central deflection of the inflated
5. Conclusion
DE membrane coupled with the air-tight chamber and successfully
This paper presents an experimentally validated nonlinear dynamic encompasses the influence of the variability of chamber capacity in
analytical framework to predict the motion of the inflated DE mem- both depressurized and pressurized states of actuation with membrane
brane coupled with a compressible confined fluid/air mass and stimu- undergoing small deformation. From the AC dynamic response with
lated by an electric field under the influence of pressure difference. The the chamber volume of 100, 400, and 740 ml, it is inferred that the
developed framework is based on the assumption of homogeneous de- percentage increment of the central deflection increases as the chamber
formations, and takes into account the impacts of electromechanical–air volume increases in both depressurized and pressurized states. The

15
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

percentage increment of central deflection is found to be maximum [9] A.K. Sharma, A. Khurana, M.M. Joglekar, A finite element model for investigating
in the depressurized state as 316% and 347%, respectively obtained the thermo-electro-mechanical response of inhomogeneously deforming dielectric
elastomer actuators, Eur. J. Comput. Mech. (2021) 387–408.
from the experimental and analytical analysis at the given offset of
[10] A.K. Sharma, P. Kumar, A. Singh, D.M. Joglekar, M.M. Joglekar, Electromechan-
𝜙𝑑𝑐 = 2.5 and chamber size of 𝑉𝑐 = 740 ml. The prediction of the ical instability of dielectric elastomer actuators with active and Inactive Electric
resonance frequency is also assessed by implementing the proposed Regions, J. Appl. Mech. 86 (6) (2019) 061008.
framework for a smaller chamber volume. There is a slight discrepancy [11] A.K. Sharma, Design of a command-shaping scheme for mitigating residual
found in the first resonance peak for the chamber volume of 𝑉𝑐 = vibrations in dielectric elastomer actuators, J. Appl. Mech. 87 (2) (2020).
400 and 740 ml with the average error of about 24% and 26%, this [12] F. Carpi, G. Frediani, S. Turco, D. De Rossi, Bioinspired tunable lens
with muscle-like electroactive elastomers, Adv. Funct. Mater. 21 (21) (2011)
is due to the influence of electrode mass, temperature and humidity
4152–4158.
effect, strain-stiffening effect, etc. It is also found that the pressure
[13] Z. Alam, A.K. Sharma, Functionally graded soft dielectric elastomer phononic
differential loss inside the chamber during electromechanical loading crystals: Finite deformation, electro-elastic longitudinal waves, and band gaps
increases as initial bulging pressure increases. It is also concluded that tunability via electro-mechanical loading, Int. J. Appl. Mech. (2022).
the effect of restriction offered by the air column is negligible when [14] A.K. Sharma, M.M. Joglekar, D.M. Joglekar, Z. Alam, Topology optimization
the DE membrane is coupled with a large airtight chamber volume and of soft compressible phononic laminates for widening the mechanically tunable
band gaps, Compos. Struct. 289 (2022) 115389.
achieved maximum possible deflection due to negligible air stiffness
[15] A.K. Sharma, M. Kosta, G. Shmuel, O. Amir, Gradient-based topology optimiza-
for the given voltage. Therefore, it can also be modeled without con- tion of soft dielectrics as tunable phononic crystals, Compos. Struct. 280 (2022)
sidering the effect of the air column at a large chamber volume. The 114846.
breakdown voltage is inversely related to the chamber volume due to [16] A. Khurana, A. Kumar, S.K. Raut, A.K. Sharma, M.M. Joglekar, Effect of vis-
the free movement of the inflated DE membrane in the case of large coelasticity on the nonlinear dynamic behavior of dielectric elastomer minimum
energy structures, Int. J. Solids Struct. 208 (2021) 141–153.
chamber volume. The key finding of this work can effectively predict
[17] A. Khurana, A.K. Patra, M.M. Joglekar, An energy-based model of dielectric
behavior of the homogeneously deformed electromechanically coupled
elastomer minimum energy structures with stiffeners: Equilibrium configuration
DE-based devices incorporating a compressible intervening medium and the electromechanical response, Mech. Adv. Mater. Struct. (2022) 1–19.
that has a significant impact on their electromechanical performance, [18] A. Khurana, A. Kumar, A.K. Sharma, M.M. Joglekar, Dynamic modeling of dielec-
and the investigation can be further incorporated to the inhomogeneous tric elastomer-based minimum energy structures with membrane entanglements
deformation of the DE-based actuators coupled with compressible fluid and finite extensibility, Sådhanå 47 (2022) 152.
mass. [19] R. Pelrine, R. Kornbluh, Q. Pei, J. Joseph, High-speed electrically actuated
elastomers with strain greater than 100%, Science 287 (5454) (2000) 836–839.
[20] A.K. Sharma, N. Arora, M.M. Joglekar, DC dynamic pull-in instability of a
CRediT authorship contribution statement dielectric elastomer balloon: An energy-based approach, Proc. Royal Soc. A 474
(2211) (2018) 20170900.
Ajay Kumar: Conceptualization, Methodology, Software, Investiga- [21] C. Zeng, X. Gao, Stability of an anisotropic dielectric elastomer plate, Int. J.
tion, Data curation, Writing – original draft. Aman Khurana: Con- Non-Linear Mech. 124 (2020) 103510.
[22] A. Khurana, D. Kumar, A.K. Sharma, M.M. Joglekar, Static and dynamic insta-
ceptualization, Methodology, Software, Investigation, Data curation,
bility modeling of electro-magneto-active polymers with various entanglements
Writing – original draft. Atul Kumar Sharma: Writing – review & and crosslinks, Int. J. Non-Linear Mech. 139 (2022) 103865.
editing, Validation, Supervision. M.M. Joglekar: Writing – review & [23] A. Khurana, M.M. Joglekar, G. Zurlo, Electromechanical stability of wrinkled
editing, Validation, Supervision. dielectric elastomers, Int. J. Solids Struct. 246 (2022) 111613.
[24] G. Kofod, P. Sommer-Larsen, R. Kornbluh, R. Pelrine, Actuation response of
Declaration of competing interest polyacrylate dielectric elastomers, J. Intell. Mater. Syst. Struct. 14 (12) (2003)
787–793.
[25] J. Huang, T. Li, C. Chiang Foo, J. Zhu, D.R. Clarke, Z. Suo, Giant, voltage-
The authors declare that they have no known competing finan- actuated deformation of a dielectric elastomer under dead load, Appl. Phys. Lett.
cial interests or personal relationships that could have appeared to 100 (4) (2012) 041911, 4.
influence the work reported in this paper. [26] A.K. Sharma, N. Sheshkar, A. Gupta, Static and dynamic stability of dielectric
elastomer fiber composites, Mater. Today: Proc. 44 (2021) 2043–2047.
Acknowledgment [27] A.K. Sharma, M.M. Joglekar, Effect of anisotropy on the dynamic electrome-
chanical instability of a dielectric elastomer actuator, Smart Mater. Struct. 28
(1) (2018) 015006.
This research is supported financially by the Science and Engi- [28] R. Vertechy, G.P. Papini Rosati, M. Fontana, Reduced model and application
neering Research Board (SERB), India through Grant No. EMR/2017/ of inflating circular diaphragm dielectric elastomer generators for wave energy
003289. harvesting, J. Vib. Acoust. 137 (1) (2015).
[29] N.C. Goulbourne, M.I. Frecker, E. Mockensturm, Electro-elastic modeling of a
dielectric elastomer diaphragm for a prosthetic blood pump, in: Smart Structures
References
and Materials 2004: Electroactive Polymer Actuators and Devices, Vol. 5385,
Eapad, International Society for Optics and Photonics, 2004, pp. 122–133.
[1] C. Ma, Y. Xia, T. Lu, Large actuation in an electromechanical actuator using gel, [30] K. Hochradel, S. Rupitsch, A. Sutor, R. Lerch, D. Vu, P. Steinmann, Dynamic
elastomer, and oil, Int. J. Non-Linear Mech. 124 (2020) 103499. performance of dielectric elastomers utilized as acoustic actuators, Appl. Phys.
[2] P. Brochu, Q. Pei, Dielectric elastomers for actuators and artificial muscles, in: A 107 (3) (2012) 531–538.
Electroactivity in Polymeric Materials, Springer, 2012, pp. 1–56.
[31] X. Zhao, Z. Suo, Method to analyze programmable deformation of dielectric
[3] D. Yang, Y. Xu, M. Ruan, Z. Xiao, W. Guo, H. Wang, L. Zhang, Improved
elastomer layers, Appl. Phys. Lett. 93 (25) (2008) 251902, 3.
electric energy density and conversion efficiency of natural rubber composites
[32] J. Fox, N. Goulbourne, Electric field-induced surface transformations and experi-
as dielectric elastomer generators, AIP Adv. 9 (2) (2019) 025035.
mental dynamic characteristics of dielectric elastomer membranes, J. Mech. Phys.
[4] S. Gour, D. Kumar, A. Khurana, Constitutive modeling for the tear fracture of
Solids 57 (8) (2009) 1417–1435.
artificial tissues in human-like soft robots, Eur. J. Mech. A Solids (2022) 104672.
[5] M. Franke, A. Ehrenhofer, S. Lahiri, E.-F. Henke, T. Wallmersperger, A. Richter, [33] J. Zhu, S. Cai, Z. Suo, Resonant behavior of a membrane of a dielectric elastomer,
Dielectric elastomer actuator driven soft robotic structures with bioinspired Int. J. Solids Struct. 47 (24) (2010) 3254–3262.
skeletal and muscular reinforcement, Front. Robot. AI 7 (2020) 510757. [34] F. Chen, J. Cao, H. Zhang, M.Y. Wang, J. Zhu, Y. Zhang, Programmable
[6] E. Biddiss, T. Chau, Dielectric elastomers as actuators for upper limb prosthetics: deformations of networked inflated dielectric elastomer actuators, IEEE/ASME
Challenges and opportunities, Med. Eng. Phys. 30 (4) (2008) 403–418. Trans. Mechatronics 24 (1) (2018) 45–55.
[7] P. Chakraborti, H.K. Toprakci, P. Yang, N. Di Spigna, P. Franzon, T. Ghosh, [35] F. Wang, C. Yuan, T. Lu, T. Wang, Anomalous bulging behaviors of a dielectric
A compact dielectric elastomer tubular actuator for refreshable Braille displays, elastomer balloon under internal pressure and electric actuation, J. Mech. Phys.
Sensors Actuators A 179 (2012) 151–157. Solids 102 (2017) 1–16.
[8] T. Li, S. Qu, W. Yang, Electromechanical and dynamic analyses of tunable [36] X. Liang, S. Cai, New electromechanical instability modes in dielectric elastomer
dielectric elastomer resonator, Int. J. Solids Struct. 49 (26) (2012) 3754–3761. balloons, Int. J. Solids Struct. 132 (2018) 96–104.

16
A. Kumar, A. Khurana, A.K. Sharma et al. International Journal of Non-Linear Mechanics 147 (2022) 104232

[37] X. Lv, L. Liu, J. Leng, Y. Liu, S. Cai, Delayed electromechanical instability of a [56] A.K. Sharma, M.M. Joglekar, A computationally efficient locking free numer-
viscoelastic dielectric elastomer balloon, Proc. R. Soc. Lond. Ser. A Math. Phys. ical framework for modeling visco-hyperelastic dielectric elastomers, Comput.
Eng. Sci. 475 (2229) (2019) 20190316. Methods Appl. Mech. Engrg. 352 (2019) 625–653.
[38] F. Carpi, G. Frediani, D. De Rossi, Hydrostatically coupled dielectric elastomer [57] A. Khurana, D. Kumar, A.K. Sharma, M.M. Joglekar, Nonlinear oscillations of
actuators, IEEE/ASME Trans. Mechatronics 15 (2) (2009) 308–315. particle-reinforced electro-magneto-viscoelastomer actuators, J. Appl. Mech. 88
[39] S. Sasikala, K. Madhavan, G. Ramesh, S. Saggar, P. Predeep, P. Chowdhury, (12) (2021).
Electro-mechanical response to the harmonic actuation of the pneumatically [58] J. Zhang, J. Ru, H. Chen, D. Li, J. Lu, Viscoelastic creep and relaxation of
coupled dielectric elastomer based actuators with and without load, Int. J. Solids dielectric elastomers characterized by a Kelvin-Voigt-Maxwell model, Appl. Phys.
Struct. 110 (2017) 58–66. Lett. 110 (4) (2017) 044104.
[40] L. Chen, W. Chen, Y. Xue, M. Zhang, X. Chen, X. Cao, Z. Zhang, G. Li, T. [59] F. Chang, Z. Lu, Dynamic model of an air spring and integration into a vehicle
Li, Investigation of the state transition and moving boundary in a pneumatic– dynamics model, Proc. Inst. Mech. Eng. D 222 (10) (2008) 1813–1825.
hydraulic coupled dielectric elastomer actuator, J. Appl. Mech. 86 (3) (2019) [60] M. Presthus, Derivation of Air Spring Model Parameters for Train Simulation,
031004, 12. 2002.
[41] H. Wang, S. Cai, F. Carpi, Z. Suo, Computational model of hydrostatically coupled [61] F.J. De Melo, A.B. Pereira, A.B. Morais, The simulation of an automotive air
dielectric elastomer actuators, J. Appl. Mech. 79 (3) (2012) 031008, 8. spring suspension using a pseudo-dynamic procedure, Appl. Sci. 8 (7) (2018)
[42] T. Wang, J. Zhang, J. Hong, M.Y. Wang, Dielectric elastomer actuators for soft 1049.
wave-handling systems, Soft Robot. 4 (1) (2017) 61–69. [62] M. Otténio, M. Destrade, R.W. Ogden, Acoustic waves at the interface of a pre-
[43] W. Sun, H. Wang, J. Zhou, Actuation and instability of interconnected dielectric stressed incompressible elastic solid and a viscous fluid, Int. J. Non-Linear Mech.
elastomer balloons, Appl. Phys. A 119 (2) (2015) 443–449. 42 (2) (2007) 310–320.
[44] F. Chen, M.Y. Wang, Simulation of networked dielectric elastomer balloon [63] D. Eder-Goy, Y. Zhao, B.-X. Xu, Dynamic pull-in instability of a prestretched
actuators, IEEE Robot. Autom. Lett. 1 (1) (2016) 221–226. viscous dielectric elastomer under electric loading, Acta Mech. 228 (12) (2017)
[45] A. Kumar, A. Khurana, A.K. Sharma, M.M. Joglekar, Dynamics of pneumatically 4293–4307.
coupled visco-hyperelastic dielectric elastomer actuators: Theoretical modeling [64] Y. Li, I. Oh, J. Chen, H. Zhang, Y. Hu, Nonlinear dynamic analysis and active
and experimental investigation, Eur. J. Mech. A Solids 95 (2022) 104636. control of visco-hyperelastic dielectric elastomer membrane, Int. J. Solids Struct.
[46] A. Kumar, A. Khurana, A.K. Sharma, M.M. Joglekar, Dynamic analysis of 152 (2018) 28–38.
anisotropic dielectric viscoelastomers incorporating humidity effect, J. Braz. Soc. [65] J. Zhang, H. Chen, B. Li, D. McCoul, Q. Pei, Coupled nonlinear oscillation and
Mech. Sci. Eng. 44 (8) (2022) 1–14. stability evolution of viscoelastic dielectric elastomers, Soft Matter 11 (38) (2015)
[47] J. Fox, N. Goulbourne, On the dynamic electromechanical loading of dielectric 7483–7493.
elastomer membranes, J. Mech. Phys. Solids 56 (8) (2008) 2669–2686. [66] M. Kollosche, G. Kofod, Z. Suo, J. Zhu, Temporal evolution and instability in a
[48] B.-X. Xu, R. Mueller, A. Theis, M. Klassen, D. Gross, Dynamic analysis of viscoelastic dielectric elastomer, J. Mech. Phys. Solids 76 (2015) 47–64.
dielectric elastomer actuators, Appl. Phys. Lett. 100 (11) (2012) 112903. [67] X. Liu, J. Zhang, H. Chen, Ambient humidity altering electromechanical actuation
[49] J. Sheng, H. Chen, B. Li, Y. Wang, Nonlinear dynamic characteristics of a of dielectric elastomers, Appl. Phys. Lett. 115 (18) (2019) 184101.
[68] B. Li, H. Chen, J. Qiang, S. Hu, Z. Zhu, Y. Wang, Effect of mechanical pre-stretch
dielectric elastomer membrane undergoing in-plane deformation, Smart Mater.
on the stabilization of dielectric elastomer actuation, J. Phys. D: Appl. Phys. 44
Struct. 23 (4) (2014) 045010.
(15) (2011) 155301.
[50] A.K. Sharma, S. Bajpayee, D.M. Joglekar, M.M. Joglekar, Dynamic instability
[69] G. Kofod, The static actuation of dielectric elastomer actuators: How does
of dielectric elastomer actuators subjected to unequal biaxial prestress, Smart
pre-stretch improve actuation? J. Phys. D: Appl. Phys. 41 (21) (2008) 215405.
Mater. Struct. 26 (11) (2017) 115019.
[70] A. Goriely, M. Destrade, M. Ben Amar, Instabilities in elastomers and in soft
[51] A. Khurana, A. Kumar, A.K. Sharma, M.M. Joglekar, Effect of polymer chains
tissues, Quart. J. Mech. Appl. Math. 59 (4) (2006) 615–630.
entanglements, crosslinks and finite extensibility on the nonlinear dynamic
[71] Y. Su, W. Chen, M. Destrade, Tuning the pull-in instability of soft dielectric
oscillations of dielectric viscoelastomer actuators, Nonlinear Dynam. 104 (2)
elastomers through loading protocols, Int. J. Non-Linear Mech. 113 (2019)
(2021) 1227–1251.
62–66.
[52] D. Kumar, S. Sarangi, Dynamic modeling of a dielectric elastomeric spherical
[72] H. Vatanjou, Y. Hojjat, M. Karafi, Nonlinear dynamic analysis of dielectric
actuator: An energy-based approach, Soft Mater. 19 (2) (2021) 129–138.
elastomer minimum energy structures, Appl. Phys. A 125 (9) (2019) 583.
[53] R. Mangan, M. Destrade, Gent models for the inflation of spherical balloons, Int.
[73] A. Khurana, A.K. Sharma, M.M. Joglekar, Nonlinear oscillations of electrically
J. Non-Linear Mech. 68 (2015) 52–58.
driven aniso-visco-hyperelastic dielectric elastomer minimum energy structures,
[54] K.A. Khan, H. Wafai, T. El Sayed, A variational constitutive framework for the
Nonlinear Dynam. 104 (3) (2021) 1991–2013.
nonlinear viscoelastic response of a dielectric elastomer, Comput. Mech. 52 (2)
[74] E. Bortot, Nonlinear dynamic response of soft thick-walled electro-active tubes,
(2013) 345–360.
Smart Mater. Struct. 27 (10) (2018) 105025.
[55] R. Sarban, B. Lassen, M. Willatzen, Dynamic electromechanical modeling
[75] K. Kashyap, A.K. Sharma, M.M. Joglekar, Nonlinear dynamic analysis of aniso-
of dielectric elastomer actuators with metallic electrodes, IEEE/ASME Trans.
visco-hyperelastic dielectric elastomer actuators, Smart Mater. Struct. 29 (5)
Mechatronics 17 (5) (2011) 960–967.
(2020) 055014.

17