Combination of peridynamics and genetic algorithm based topology optimization methods for additive manufacturing-friendly designs

Topology optimization (TO) is a practical tool to generate light-weighted engineering structures for various manufacturing industries. However, manufacturing constraints and surface smoothing are still considerable challenges for TO algorithms. Existing TOframeworks utilize mechanical analysis approaches that discretize the whole domain with elements or particles. Therefore, obtained geometries from TO have been criticized for their complex shapes. In this study, we propose a coupled framework to generate additive manufacturing (AM)-friendly designs which result in less support structure and higher surface quality. For this purpose, the generative topology optimization method (GTO), which uses genetic algorithm to search for the best alternative set of geometry within all the possible topology results, is coupled with the peridynamics topology optimization (PD-TO) method to evolve the PD-TO results into AM-friendly shapes. The PD-TO discretizes the problem domain using equally spaced particles during the TO process. Hence, PD-TO generates a point cloud file with relevant artificial material density values in the final state. Then, the GTO method utilizes the point cloud and material densities as an input file to achieve better final geometry. AM-friendly designs achieved from GTO are compared with the initial results obtained from PD-TO to demonstrate the efficiency and capability of the proposed method