Peridynamics topology optimization of three-dimensional structures with surface cracks for additive manufacturing

Official URL: https://dx.doi.org/10.1016/j.cma.2022.115665

Additive manufacturing (AM) is an effective approach to fabricating intricate shapes obtained from topology optimization (TO). However, it may cause undesired manufacturing-induced defects/cracks due to high thermal residual stresses. This study proposes a PeriDynamics-enabled three-dimensional Topology Optimization method (PD-TO) for designing structures by considering surface cracks for the AM processes. The PD-TO approach employs a bi-directional evolutionary structural optimization method and uses particle discretization of geometry for mechanical analysis. Crack surfaces are generated by breaking three-dimensional nonlocal interactions of the particles, and thus, during the optimization process, complex multiple structural discontinuities can be diligently modeled. First, the proposed approach is validated by solving benchmark problems without cracks. For each benchmark geometry, the PD-TO analysis is then performed by considering different positions and numbers (single/multiple) of cracks. These analyses extensively investigate and demonstrate the effects of a priori knowledge of residual stress-induced damages/cracks on the optimum topology for additive manufacturing. Besides, the smoothing operation is applied to the optimum designs to transform voxel shapes into AM-friendly smooth surfaces. These geometries are manufactured by an extrusion-based AM process to demonstrate the practical engineering application of the proposed method. Finally, the comparison of numerical results is also supported by the experimental tests conducted on the optimized topologies. Overall, it is confirmed that the PD-TO approach is a viable and accurate optimization tool for additive manufacturing considering possible process-induced damages.