A machine learning guided investigation of quality repeatability in metal laser powder bed fusion additive manufacturing

Abstract

Additive manufacturing has entered the phase of industrial adoption, for which its quality repeatability is of vital importance to industries where functional parts with consistent mechanical properties are desired. This concern will manifest with large scale implementation of such technology, affecting not only the reliability of products but the reputation and profitability of a business. The root cause to this problem is obscure demanding a systematic approach to identify potential influencing parameters for better process control. In this article, the quality repeatability of laser powder bed fusion (L-PBF) technology, in terms of static mechanical properties of printed parts, was quantified using relative standard deviation, and a machine learning approach for root cause analysis was demonstrated. While most of the prior work focused on the effect of laser-related process parameters to part properties, this research emphasises on the downstream production parameters while keeping laser-related parameters fixed. It was found that the combinational effect of part location and post-chamber pressure drop heavily influences the quality of printed parts. A follow-up experiment with the new process control was able to produce parts with improved quality repeatability. This proves the effectiveness of the proposed approach for process control of L-PBF at large scale implementation.