A Modified Simulated Annealing-Based Method for Hybrid Lattice Support Structure Design in LPBF Additive Manufacturing

Abstract

When designed effectively, support structures play a critical role in quickly dissipating heat and mitigate part distortion without driving up excessive costs within the additive manufacturing metals technique of Laser Powder Bed Fusion (LPBF). Lattices, composed of individual unit cells strategically arranged to achieve a desired function, are a promising solution as a support structure. Prior research utilizing gradient-based optimizers to design lattice support structures for heat dissipation poses challenges regarding limited design domain exploration and non-differentiable objective functions. Non-gradient-based optimizers are an alternative solution but existing optimizers, such as traditional simulated annealing (SA), are known to be more computationally expensive compared to gradient-based optimizers, rendering it challenging to optimize the heat dissipation of lattice support structures. This paper introduces a modified SA-based method to design lattice structures for LPBF by efficiently optimizing the distribution of a library composed of various types of unit cells, thereby creating hybrid lattice support structures (hLSS). A stage-dependent annealing swapping strategy is created and integrated into the method for efficient design domain exploration. Homogenization approximation and equivalent static loading are also performed in each iteration step to make the design optimization process computationally tractable. Two case studies validate the method by designing hLSS for a cantilever beam and a bracket. The results of these case studies show the method's ability to achieve material cost savings of up to 61% and post-processing cost savings of up to 62% when compared to a solid support domain while satisfying manufacturing constraints.