Multi-material topology optimization considering interface nonlinearity

The work on multi-material topology optimization (MMTO) focusing on plasticity is limited. During linear static analysis, MMTO can outperform single-material topology optimization (SMTO) due to increased design freedom. However, conventional MMTO typically considers only the existence and selection of structural materials, neglecting the bonding between dissimilar materials. In scenarios where joint materials have lower yield stress than structural materials, MMTO solutions may be inferior to SMTO due to potential joint failure, such as yielding. This paper proposed an interface enforcement method (IEM) that improves MMTO solutions by considering the interface locations and encouraging them to move from high-stress to low-stress regions, resulting in better MMTO solutions. This method includes a set-based interface detection algorithm and requires nonlinear analysis to obtain stress and displacement information near interface elements, which is then used to create auxiliary load cases for sensitivity modification. By tuning the interface shifting factor, design sensitivities are modified to favor nonlinear performance. Numerical examples demonstrate that the proposed IEM can effectively reduce nonlinear displacement. Additionally, the proposed IEM avoids expensive nonlinear sensitivity computation and works for finite element analysis (FEA) solvers with black-box characteristics.