Nonlinear Optimization Algorithms for Adjusting Selective Laser Melting Conditions

Abstract—The possibility and validity of using the mathematical algorithms of nonlinear optimization and machine learning to describe the dependence of the "input" parameters of a setup on the "output" parameters of the synthesized material during selective laser melting of a metal powder of a nickel-based limitedly weldable alloy are considered. The problem is formulated, and the validity of its mathematical solution is substantiated. A mathematical model is developed, and its predictive abilities and the accuracy of predicting the synthesis condition parameters for an "ideal" sample are estimated.