Multiscale investigation on surface integrity and performance of metal parts by selective laser melting

Abstract

Selective laser melting (SLM) is a popular metal additive manufacturing process which has drawn a great attention in the past decade. It can make functional metal parts directly from the CAD data in a layer upon layer fashion. As-SLMed parts may have near full density comparable to that made by traditional manufacturing methods, SLM has wide application potentials in tool, aerospace, automotive, biomedical, and energy industries. During the SLM process, the interaction between the moving laser beam and the powder leads to very complex physical phenomena. It includes the absorption and reflection of the laser beam; fast heating-cooling cycles; surface tension, wetting, and de-wetting; complex fluid flow and Marangoni convection in the melt pool. Due to the complexity of laser-powder interaction, a comprehensive study of powder-substrate behavior under different process parameters is highly demanded. In addition, high tensile residual stress and part distortion are considered as two major problems for SLMed parts. To understand the formation mechanism of residual stress and part distortion, an efficient prediction tool is highly desired. However, it is very challenging to predict residual stress and distortion of a practical SLMed part as it often requires millions of laser scan which involves the complex multiscale and multi-physics phenomena. Furthermore, poor surface integrity of a SLMed part is detrimental to functional performance such as fatigue. Therefore, a comprehensive understanding of process-surface integrity-fatigue relationship is of great importance. To reveal the underlying mechanisms in SLM, this research focuses on: (1) A critical assessment on the state-of-the-art of SLM process was performed; (2) A thorough characterization of the geometrical, metallurgical, and mechanical properties of single tracks under a broad range of processing conditions in SLM of IN625 alloy; (3) A comprehensive investigation of microstructural evolution of IN625 alloy from as-SLM to different heat treatment conditions; (4) Two efficient multi-physics predictive tools, i.e., stress-thread and temperature-thread based multiscale modeling approaches, to predict residual stress and part distortion of SLMed practical parts at different scanning strategies via thermal loading subroutine and material states variables; (5) A simulation study to investigate the scalability of layer thickness and heat source dimension in fast predicting the distortion and residual stress; and (6) Conclusions and an outlook of the future research.