Simultaneous numerical approximation of microstructures and relaxed minimizers (Q1374751)

The problem of minimizing an integral functional with nonquasiconvex integrand in a bounded compact open set of admissible functions with Lipschitz continuous boundary is considered. A numerical method, which is based on the construction of a minimizing problem alternative to the relaxed problem is introduced. It approximates simultaneously the highly oscillating minimizing sequences, which are called microstructures, and the minimizers of the corresponding relaxed problem. The alternative minimizing problem provides the base for the discrete problems and the numerical method. Existence and convergence of the discrete solutions are proved. An error estimate is given. A numerical example is supplied. The method is a general one in the sense that the problem is solved completely numerically. It means that it avoids the requirement of other methods on the analytical description of the quasiconvex envelope of the energy density and its numerical approximation during the computation.