Concurrent Island scanning pattern and large-scale topology optimization method for laser powder bed fusion processed parts

This paper proposes a numerical method for concurrent scanning pattern and topology optimization to design laser powder bed fusion (LPBF) processed large-scale parts. The method developed in this work is motivated by the emerging needs in design-for-LPBF metal additive manufacturing, to reduce the part residual warpage for manufacturability and enable the highresolution structural details for performance. Specifically, the inherent strain-based process simulation solver is developed to predict the part-scale residual warpage induced by LPBF. Then, the concurrent scanning pattern and structural topology optimization algorithm is developed facilitated by the inherent strain solver and accelerated by the PETSc parallel-computing framework. The proposed approach is utilized to solve several 3D benchmark design cases to show its effectiveness in stiffness improvement and residual warpage reduction. The effects of varying the weighting coefficients and scanning patterns, and the importance of considering the concurrent optimization strategy are carefully investigated. The observed numerical phenomena are discussed to summarize the paper.

Automated summary

This paper proposes a numerical method for concurrent scanning pattern and topology optimization in the design of large-scale parts using laser powder bed fusion. The method utilizes an inherent strain-based process simulation solver to predict part-scale residual warpage induced by LPBF. It also develops a concurrent scanning pattern and structural topology optimization algorithm, accelerated by the PETSc parallel-computing framework. The effectiveness of the proposed approach is shown through solving benchmark design cases, improving stiffness and reducing residual warpage.