A Novel Time Step Fusion Method with Finite Volume Formulation for Accelerated Thermal Analysis of Laser Additive Manufacturing

Abstract

Laser additive manufacturing has been regarded as a typical green manufacturing process. This paper presents a novel numerical approach termed time step fusion (TSF) along with the finite volume method (FVM), for fast computing the temperature field in a predominant laser additive manufacturing process, namely, selective laser melting. The solution acceleration strategy using TSF is that the entire computational domain is partitioned into multiple subdomains, and in the subdomain distant from the laser source and with the milder thermal gradients, larger time steps are employed. A thin wall is simulated to verify the proposed TSF method and evaluate its effectiveness and efficiency. The results are compared with those of the standard FVM model without TSF. It shows that they are in good agreement in terms of the spatiotemporal thermal profiles. The mean absolute error for the case studies with TSF is below 1 K except for a few spikes of the discrepancy reaching up to 15 K in the molten pool. Meanwhile, up to 28% acceleration in computational speed is obtained with TSF-FVM compared with regular FVM, and 93% time saving is achieved compared with a benchmark FEM model with a commercial solver. An extended case study is also presented to further verify the applicability of the proposed approach for complex geometries. The factors that contribute to the speed are analyzed, and the strategies for potential further improvement are also discussed.