On the failure of the maximum principle in coupled thermoelasticity (Q1064102)

The objective of this paper is to point out that the maximum principle does not hold in coupled thermoelasticity. In order to show it, the simple one-dimensional model is considered, described by the following integro-partial-differential equation \[ (1)\quad \partial^ 2\theta /\partial x^ 2(x,t)=(1+a)\partial \theta /\partial t(x,t)- a\int^{1}_{0}\partial \theta /\partial t(y,t)dy. \] Here \(\theta =\theta (x,t)\) is a temperature field and a measures the coupling between thermal and mechanical effects. Equation (1) describes the distribution of temperature in an infinite slab of width 1 made of linear thermoelastic material. For simplicity the material is assumed to be isotropic and homogeneous. The author proves that the solution of the above equation does not satisy the maximum principle no matter how small \(a>0\) is, though for \(a=0\) it does hold. Moreover two further theorems are proved. One of them characterizes the behaviour of associated kernel of the integral operator while the second states that mean temperature is negative for an arbitrary time even for non negative initial temperature. The paper is well written. It will be of interest not only for thermomechanicians but also for mathematicians working in the fields of partial-differential equations and integro-partial-differential equations.