Microstructure simulation using self‐consistent clustering analysis

To capture all the individual microstructural effects of complex and heterogeneous materials in structural finite element simulations, a two-scale simulation approach is necessary. Since computational effort such simulations extremely high, different methods exist to overcome this problem. In terms FFT-based microscale simulation, one possibility use reduced set frequencies leading numerical solution Lippmann-Schwinger equation [?]. post-processing step, highly resolved fields may then be reconstructed by using compressed sensing technique stress evaluation method real space therefore not reduced, it most beneficial linear elastic material behavior. Another very recent reduce self-consistent clustering analysis [?,?]. Such split into an offline online stage. Within stage, points high-fidelity representation unit cell are grouped clusters with similar Thereafter, used solve boundary value problem clustered equation. generation based on we propose for these improve efficiency total algorithm. Elastic elasto-plastic composites investigated small strain setting as representative examples.