The nonlinear optimization problem in heat conduction

From MaRDI portal
Publication:2388367





A body \(D\) in a domain \(\Omega\) with a given temperature \(\varphi\) on its surface \(\partial D\) and zero on \(\partial\Omega\) is considered and the domain \(\Omega\) itself is to be optimized to minimize the functional \(\int_{\partial D}\Gamma(x,u_\mu(x))d\sigma\), where \(u_\mu\) seems to be a normal flux of the solution \(u\) to the heat equation \(\Delta u=0\) on the isolating layer \(\Omega\setminus D\) whose volume is prescribed (equal to 1). This equality constraint is then penalized by a suitable nonsmooth penalty function as if it were only a unilateral constraint. For this, \(\Gamma(x,\cdot)\) is assumed convex, increasing, and coercive. This penalization is proved to be exact if the penalty parameter is small enough, and a lot of fine regularity properties of the solutions to the penalized (and thus also the original) problems is proved, in particular analyticity of the free boundary \(\partial\Omega\).




Cited in
(26)








This page was built for publication: The nonlinear optimization problem in heat conduction

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2388367)