A three-dimensional inverse problem of viscoelasticity

Integro-partial differential equations (45K05) Inverse problems for integral equations (45Q05) Linear constitutive equations for materials with memory (74D05) Wave scattering in solid mechanics (74J20)

Review text

The author considers the viscoelastic problem \[ \rho(x)D^2_t{\mathbf u}(x,t) - L{\mathbf u}(x,t) = {\mathbf f}^0\delta(x-y,t),\quad u\equiv 0\;\text{if } t<0, \] where \({\mathbf f}^0\in \mathbb{R}^3\) is a given vector characterizing the direction of the force at \(y\in\mathbb{R}^3\) and the vector integro-differential operator \(L=(L_1,L_2,L_3)\) is defined by \(L_iu=\sum_{j=1}^{3}\,D_{x_j}\sigma_{i,j}({\mathbf u})\), \(i=1,2,3\), where \[ \sigma_{i,j}({\mathbf u})(x,t)=\lambda_0(x)\delta_{i,j}\text{div}\,{\mathbf u}(x,t)+\mu_0(x)[D_{x_j}u_i(x,t)+D_{x_i}u_j(x,t)] \] \[ +\int_{-\infty}^t \big\{\lambda(x,t-s)\delta_{i,j}\text{div}\,{\mathbf u}(x,s) +\mu(x,t-s)[D_{x_j}u_i(x,s)+D_{x_i}u_j(x,s)]\big\}\,ds. \] Here \(\delta_{i,j}\) is the Kronecker delta, \(\rho\) is the (positive) density of the viscoelastic medium, while \(\lambda_0\) and \(\mu_0\) are the elasticity moduli such that \(\mu_0\) is positive along with \(\lambda_0+2\mu_0\). Assume that the viscosity functions \(\lambda\) and \(\mu\) admit the representations \(\lambda(x,t)=g_1(t)\lambda_1(x)\) and \(\mu(x,t)=g_2(t)\mu_1(x)\), where \(\lambda_1\) and \(\mu_1\) are \textit{unknown}, while functions \(g_1\) and \(g_2\) are \textit{known} and satisfy \(g_1(0)=g_2(0)=1\). \noindent Let \[ c_p(x)=\Big[\frac{\lambda_0(x)+2\mu_0(x)}{\rho(x)}\Big]^{1/2},\quad c_s(x)=\Big[\frac{\mu_0(x)}{\rho(x)}\Big]^{1/2} \] be the velocities of longitudinal and transversal waves and let \(\tau_p(x,y)\) and \(\tau_s(x,y)\) be the geodesic distances corresponding to the Riemannian metrics \(d\tau_p=c_p(x)^{-1}|dx|\) and \(d\tau_s=c_s(x)^{-1}|dx|\), respectively. \noindent Let \(\Omega\subset \mathbb{R}^3\) be a compact domain with a smooth boundary \(\partial \Omega\) such that any pair of points \(x,y\in \Omega\) can be joined by unique geodesics \(\Gamma_p(x,y)\) and \(\Gamma_s(x,y)\). \noindent To determine \(\lambda_1\) and \(\mu_1\) the author assumes to know the two functions \(\alpha^{(p,-1)}:\partial \Omega\times \partial \Omega\to \mathbb{R}^3\) and \(\alpha^{(s,-1)}:\partial \Omega\times \partial \Omega\to \mathbb{R}^3\) characterizing the amplitude and the direction of the displacements of particles of the elastic medium on the fronts of longitudinal and transversal waves. \noindent Taking advantage of known result concerning integral geometry problems, the author can prove the uniqueness of the solution of his inverse problem and its continuous dependence on the pair \((\alpha^{(p,-1)},\alpha^{(s,-1)})\) under the following assumptions: \textit{(i)} \(\rho,\lambda_0,\mu_0\in C^2(\Omega_0)\), where \(\Omega_0\) is a domain containing \({\overline \Omega}\); \textit{(ii)} the family of geodetics \(\Gamma_p(x,y)\) and \(\Gamma_s(x,y)\) are regular inside \({\overline \Omega}\); \textit{(iii)} \(\lambda_1,\mu_1\in C^1(\Omega_0)\).

Recommendations

Cites work

Cited in

(12)

This page was built for publication: A three-dimensional inverse problem of viscoelasticity

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q416881)