Optimization of Process Parameters in Metal Additive Manufacturing for Defect Reduction Using Machine Learning

Abstract

Metal Additive Manufacturing (MAM) has emerged as one of the most transformative manufacturing technologies in modern engineering industries due to its capability to fabricate complex geometries with reduced material waste and shorter production cycles. Industries such as aerospace, biomedical engineering, automotive manufacturing, and defense have increasingly adopted metal additive manufacturing processes for producing lightweight and high-performance components. Despite these advantages, the widespread industrial implementation of metal additive manufacturing is significantly limited by the occurrence of manufacturing defects such as porosity, residual stresses, balling effects, lack of fusion, thermal distortion, and microcracking. The generation of these defects is strongly influenced by process parameters including laser power, scan speed, hatch spacing, layer thickness, powder characteristics, and thermal gradients. Conventional trial-and-error approaches for parameter selection are inefficient because of the large number of interacting variables involved in additive manufacturing processes. Consequently, there is a growing demand for intelligent optimization methodologies capable of identifying optimal process conditions for defect minimization.