Volume 1, Issue 3 e1028
RESEARCH ARTICLE
Full Access

Asymptotic analysis of quasistatic electro-viscoelastic problem with Tresca's friction Law

Mohamed Dilmi

Mohamed Dilmi

Laboratory of Pure and Applied Mathematics, University of M'sila, M'sila, Algeria

Search for more papers by this author
Mourad Dilmi

Corresponding Author

Mourad Dilmi

Applied Mathematics Laboratory, Department of Mathematics, Setif I-University, Setif, Algeria

Mourad Dilmi, Applied Mathematics Laboratory, Department of Mathematics, Setif I-University, 19000 Setif, Algeria.

Email: [email protected]

Search for more papers by this author
Hamid Benseridi

Hamid Benseridi

Applied Mathematics Laboratory, Department of Mathematics, Setif I-University, Setif, Algeria

Search for more papers by this author
First published: 10 April 2019
Citations: 2

Abstract

This paper is devoted to the study of the asymptotic analysis of a mathematical model involving a frictionless contact between an quasistatic electro-viscoelastic and a conductive foundation in a three-dimensional thin domain Ωε. The contact is modeled with Tresca friction law. We derive a variational formulation of the problem and prove its unique weak solution. Then, we study the asymptotic analysis when the small parameter of the domain tends to zero. Finally, the specific Reynolds limit equation and the limit of Tresca free boundary conditions are obtained.

1 INTRODUCTION AND STATEMENT OF THE PROBLEM

Scientific research in mechanics are articulated around two main components: one devoted to the laws of behavior and the other on boundary conditions imposed on the body. For the constitutive law, we consider an quasistatic electro-viscoelastic body with Tresca free boundary friction conditions in the dynamic regime occupying a bounded homogeneous domain of 3D. The piezoelectric is characterized by the coupling between the mechanical and the electrical properties of the material: it was observed that the appearance of electric charges on some crystals was due to the action of body forces and surface tractions and, conversely, the action of the electric field generated strain or stress in the body (see the work of Batra and Yang1 and the references therein). This coupling leads to the appearance of electric potential when mechanical stress is present and, conversely, mechanical stress is generated when electric potential is applied. Piezoelectric materials for which the mechanical properties are elastic are called electro-elastic materials and those for which the mechanical properties are viscoelastic are called electro-viscoelastic materials. Antiplane shear deformations are one of the simplest examples of deformations that solids can undergo. In the work of Sofonea and Dalah,2 the authors studied antiplane frictional contact problem of electro-viscoelastic cylinders. Chougui and Drabla3 studied a frictional contact problem with adhesion between an elastic piezoelectric body and a deformable obstacle. Contact problems for electro-viscoelastic materials were considered in the works of Sofonea et al.4, 5 In most cases, the authors are interested in studying the weak formulation of the solution, then prove the existence and uniqueness of the solution.

Here, we continue this line of research and study the asymptotic analysis of the electro-viscoelastic quasistatic problem with friction modeled by the law of Tresca, in a thin domain urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0001, where 0 < ε < 1 is a small parameter that will tend to zero. The boundary of Ωε will be noted urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0002, such that urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0003 is the upper boundary of equation x3 = εh(x1,x2), urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0004 is the lateral boundary and ω is a fixed bounded domain of urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0005 of equation x3 = 0, which is bottom of the domain Ωε.

In mathematical literature, there are many research papers regarding the asymptotic analysis of the problems that arise in a thin domain of which we mention; Benseridi and Dilmi6 studied the asymptotic analysis of a dynamical problem of isothermal elasticity with nonlinear friction of Tresca type. The asymptotic behavior of a dynamical problem of nonisothermal elasticity with nonlinear friction of Tresca type was studied in the work of Saadallah.7 Bayada and Lhalouani8 investigated the asymptotic and numerical analysis for a unilateral contact problem with Coulomb's friction between an elastic body and a thin elastic soft layer. Benseghir et al9 studied the theoretical analysis of a frictionless contact between two general elastic bodies in a stationary regime in a three-dimensional thin domain with Tresca friction law. The asymptotic analysis of some problem in mechanics of the fluids in a thin domain in the stationary case is found in the works of Bayada and Boukrouche10 and Dilmi et al.11

Our work is structured as follows. In Section 1, we present the model of the electro-viscoelastic, and we give some notation on the data of the problem. In Section 2, we derive the variational formulation for the problem and state our result of existence and uniqueness of the solution. In Section 3, we employ the change of variable z = x3/ε to transform the initial problem posed in the domain Ωε at a new problem posed on a fixed domain Ω independent of parameter ε. We establish some estimates independent of parameter ε, for the displacement urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0006 and the electric potential urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0007. Finally, we give the convergence results, the limit problem, and its uniqueness in Section 4.

Let h(.) be a function to class C1 defined on ω such that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0008
we note urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0009, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0010. The domain Ωε is given by
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0011
with urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0012 as its boundary.

Let uε(x,t) be the displacement field and φε(x,t) be the electric potential in Ωε.

The electro-viscoelastic constitutive law is given by
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0013
where d(.) the strain tensor, which is given by
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0014
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0015 is the piezoelectric tensor, where urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0016 is the piezoelectric coefficient, (Eε)T is the transposed of the tensor Eε, με, λε are the coefficients of Lamé, I3 is the identity, θε is the viscosity term, and βε is the electric coefficient of permitting.
Let n = (n1,n2,n3) be the unit outward normal to Γε, and
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0017
are, respectively, the normal and the tangential of the displacement uε, and the normal and tangential components of stress tensor field σε.

The classical model for quasistatic electro-viscoelastic problem is as follows.

Problem ( PE.V). Find the displacement field urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0018, and the electric potential urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0019 such that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0020()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0021()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0022()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0023()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0024()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0025()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0026()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0027()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0028()
where 1 and 2 represent the equilibrium equations, 3-5 represent the boundary condition for the displacement field and normal velocity, 6 represent the boundary condition for the electric potential, and 7 represent the boundary condition for the electric charges, 8 represent the law of friction of Tresca, where |.| indicate the Euclidean norm, and kε is a given function,  f  ε represent the densities of the forces, qε represent the densities of the electric charges, and 9 represent the initial condition of displacement.
We suppose that the coefficients of Lamé με, λε and the viscosity term θε satisfies
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0029
we also suppose that the electric coefficient of permitting and the piezoelectric parameter satisfies
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0030

2 WEAK FORMULATION

To get the weak formulation, we recall some standard spaces: L2ε) is the space of Lebesgue for the norm urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0031, H1ε)3 is the space of Sobolev, which is given by
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0032
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0033 is the closing of Dε)3 in H1ε)3, and H−1ε)3 is the dual space of urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0034.

We suppose that the problem of electro-viscoelastic (PE.V) admits a solution (uε,φε) regular, and let (v,ψ)∈(H1ε)3,H1ε)). We multiply Equation 1 by the element urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0035, in the same way for Equation 2 by the element ψ ∈ H1ε), then we integrate on Ωε and, using the Green's formula, we obtain the following variational problem.

Problem  urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0036. Find the displacement field urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0037, and the electric potential urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0038 such that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0039()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0040()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0041()
where
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0042
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0043()
and
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0044()

Theorem 1.Under the assumptions

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0045()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0046()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0047()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0048()

The problem 10-12 admits a unique solution (uε,φε) ∈ Kε × Qε, such that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0049()

Proof.The proof is based on the regularization method, which is based on an approximation of nondifferentaible term jε(.) by a family of differentaible onse urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0050, where

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0051
and we build a problem approximate
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0052
Using Galerkin's method, we show that there exists a unique solution urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0053 of this last approximate problem (see the works of Duvaut and Lions12, 13). Then, the limit of urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0054 when ζ tends to zero is a solution of 10-12.

3 CHANGE OF THE DOMAIN AND A PRIORI ESTIMATES

To study the asymptotic analysis of the problem, we will use the technique of change of the variable z = x3/ε, thus, ∀(x,x3) ∈ Ωε, we find (x,z) in Ω such that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0055
with urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0056, its boundary. We define now on Ω the following functions:
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0057
For the data of problem 1-8, it is supposed that they depend of ε in the following manner:
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0058
with urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0059, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0060, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0061, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0062, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0063, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0064, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0065, and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0066 do not depend on ε.

Now, we define the function spaces on Ω.

Let
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0067
and
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0068
Vz is the Banach space with the norm
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0069
Then, the problem 10-11 is equivalent to the following problem.
Problem  urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0070. Find the displacement field urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0071, and the electric potential urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0072 such that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0073()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0074()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0075
where
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0076
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0077
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0078
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0079
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0080
and
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0081
Now, we establish the a priori estimates for the displacement urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0082 and the electric potential urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0083.

Theorem 2.Under the same assumptions as in Theorem 1, there exists a constant c independent of ε such that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0084()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0085()

Proof.Let (uε,φε) the solution of the problem 10-12, then we have

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0086()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0087()

We derive 25 for t, then choose ψ = φε and we use Korn's inequality urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0088, we find

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0089
this implies that
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0090
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0091()

By integration of 26, we have

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0092
using Poincaré's inequality
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0093
we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0094()

By applying the Hölder and Young inequalities, we find

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0095()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0096()
and
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0097()

By Poincaré's inequality

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0098
and Young's inequality, we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0099()

From 27-31, we find

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0100()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0101()

It is necessary to estimate ∇φε(0). From 11, we deduce that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0102
thus, ∀ψ ∈ Qε,
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0103
we are multiplying this inequality by urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0104, we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0105
where
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0106
On the other hand, the sum of 32 with 33 is given as
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0107()

As

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0108
we multiply the inequality 34 by ε, we get
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0109()
where
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0110
is a constant independent of ε.

Using Gronwall's lemma in inequality 35, we get

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0111
where c is a positive constant independent of ε. Consequently, we find
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0112

4 CONVERGENCE RESULTS AND LIMIT PROBLEM

Theorem 3.Under the assumptions of Theorem (2), there exists urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0113, i = 1,2 such that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0114()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0115()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0116()

Proof.According to Theorem 2, there is constant c independent of ε such that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0117
Using these estimates with the Poincaré's inequality in the domain Ω, we obtain 36, for 37-38, according to 22-23, and 36.

Theorem 4.With the same assumptions of Theorem 3, (u,φ) satisfies the following relations:

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0118()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0119()
and the limit problem
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0120()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0121

Proof.As J(.) is convex and lower semicontinuous

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0122
thus, by using the convergence results of the Theorem 3, we get
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0123()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0124()

We choose now in the variational equation 42

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0125
we find
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0126
By using Green's formula and while choosing w2 = 0 and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0127, then w1 = 0, and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0128, we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0129
On ω, we have urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0130, thus we get
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0131()

As urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0132, urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0133, then 44 is valid in L2(Ω).

Theorem 5.Let

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0134
Under the same assumptions of Theorem 4, the traces of the displacements τ(x,t), s(x,t) and electric potential l(x,t), satisfy the following inequality:
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0135()
and the following limit form of the Tresca boundary conditions:
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0136()

Moreover, (u,φ) satisfies the following weak form of the Reynolds equation:

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0137()
where
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0138

Proof.For 45 and 46, it is enough to follow the same techniques in the work of Bayada and Boukrouche.10 To prove 47, we integrate 41 between 0 and z, it is seen that

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0139
Integrating for the second time between 0 and z, we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0140()
if z = h(x) in 48, we find
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0141()

By integrating 48 from 0 to h(x), we get

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0142()

According to 49 and 50, we deduce 47.

Theorem 6.The solution (u,φ) of limit problem 41 is unique in L(0,T,Vz) × L(0,T,Vz).

Proof.Let us suppose that there are two solution (u∗1,φ∗1) and (u∗2,φ∗2) of variational equations 39 and 40, we have

urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0143()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0144()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0145()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0146()
Take urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0147 in 51 and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0148 in 52 (respectively), and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0149 in 53 and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0150 in 54 (respectively). By summing the two inequalities and two equations, we obtain
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0151()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0152()
We put urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0153 and urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0154, according to 55 and 56, we find
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0155()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0156()
By the integration of 57 and 58, we get
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0157()
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0158()
the two inequalities 59 and 60, given
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0159
Now, by Gronwall's lemma, we find
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0160
using Poincaré's inequality, we deduce
urn:x-wiley:cmm4:media:cmm41028:cmm41028-math-0161

Biographies

  • biography image

    Mohamed Dilmi is a PhD in applied mathematics from the University of M'sila. His research area includes the asymptotic analysis of some boundary value problems in a thin film.

  • biography image

    Mourad Dilmi is a full professor in the Department of Mathematics at the Sétif 1 University (Algeria). He holds degrees in Magister (2001) and a PhD in Applied Mathematics (2009). He is the author of more than thirty research papers in topics such as existence and uniqueness results for boundary value problems, singularity of the solutions, transmission problem, and asymptotic analysis in a thin domain.

  • biography image

    Hamid Benseridi is currently a professor in the Department of Mathematics, Faculty of Science, University of Sétif 1. His research area includes applied mathematics (modeling in continuous porous media) and asymptotic analysis in a thin domain.

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.