Machine learning for real-time detection of local heat accumulation in metal additive manufacturing

Abstract

Metal additive manufacturing is associated with thermal cycles of high rates of heating, melting, cooling, and solidification. Some areas within the build experience thermal cycles that depend on the paths of the energy source. Additionally, geometrical features, such as thin walls and overhangs, can lead to heat accumulation, potentially affecting the microstructure, fatigue life, and induced residual stresses that may lead to dimensional distortion and cracking. The identification of significant heat accumulation can be used for part quality monitoring to inform the design process, enhance the quality of printed parts, and optimize the process parameters. This study aims to efficiently identify heat accumulation with affordable in-situ infrared imaging for further characterization and mitigation to enhance the quality of printed parts. A computational framework employing machine learning is developed to identify zones of local heat accumulation in real time. The effectiveness of this approach is demonstrated by experiments conducted on a build with a wide variety of geometrical features. In addition, characterization and detailed analyses of detected local heat accumulation zones are provided.