An enhanced layer lumping method for accelerating simulation of metal components produced by laser powder bed fusion

Abstract

Mechanical simulations are usually implemented via inherent strain method in a sequential layer-by-layer manner to predict residual deformation of large metal builds given the bottom-up fabrication nature of the laser powder bed fusion (L-PBF) process. However, it is very time-consuming since too many layers need to be simulated in a large component. In this paper, an enhanced layer lumping method (ELLM) is developed to accelerate the layer-wise simulation while maintaining its accuracy. Based on meso-scale modeling, material property parameters including yield stress and inherent strain values are adjusted for the lumped layers to avoid overestimation errors in the residual stress and deformation introduced by layer lumping. The material property adjustment incorporated in layer lumping is the key feature of the proposed ELLM. Computational time can be greatly reduced (e.g., 70% decrease for the cantilever beam case) by the ELLM. Good accuracy for residual deformation prediction by the ELLM has been demonstrated by comparing a benchmark layer-wise simulation with experiment for a large Inconel 718 cantilever beam. Moreover, the scalability and robustness of the proposed method is also fully verified through a complex canonical part.