Design of Molten Pool Geometry in Additive Manufacturing by Analysis of Inverse Problems

Abstract

Abstract Metal additive manufacturi ng is expected to achieve a manufacturing revolution by merging the advance technologies such as information technology and image-sensing technology. The concept on the design of process parameters in metal additive manufacturing by means of machine learning is discussed in the present work. In order to realize the discussed concept, the process parameter in laser additive manufacturing is tried to be predicted from the images of molten pool geometry by the analysis of inverse problems with machine learning. It is found that the prediction of pitch width, which is one of the process parameters, can be achieved by probabilistic classification.