Towards design automation of topology optimized parts: assessment of shape and parametric optimization-based methods

Abstract In recent years, Additive Manufacturing has experienced significant growth, yet its full potential is constrained by the lack of clear and easily replicable Design for Additive Manufacturing (DfAM) methodologies. Focusing on the product design stage, the current design approach for topology optimized parts involves an iterative re-design process to identify and reduce stress concentrations based on NURBS modelling, which is hardly replicable and heavily influenced by the designer’s experience. This work aims to define two new workflows for DfAM that are easily replicable, highly automated, and based on numerical optimization tools. Leveraging the optimization tools available in 3DExperience integrated CAD platform, after topology optimization, the first workflow involves generating the skeletonization of the resulting geometry and reconstructing it with parametric surfaces, reducing maximum stresses via parametric optimization. The second workflow reconstructs the resulting optimized geometry as a non-parametric B-Rep surface, optimizing maximum stresses through automatic shape optimization. Validation through a case study compares results with the current state-of-the-art approach, evaluating mechanical performance and development lead time, while highlighting designer working time and the pros, cons, and limitations of the proposed methods.