Estimating the absorption efficiency in a laser welding process using a nonlinear inverse problem method

Abstract

A numerical-experimental methodology is presented in this study to estimate the absorption efficiency in a laser welding process by estimating the rate of energy transferred to a metal plate. The iterative Function Specification Method was modified to account for moving temperature sensor thermal sensitivity as a function of time and position relative to the welding bead. Thus, highly nonlinear problems can be solved by using a high-temperature gradient in the measurement sensor region. Three experiments on an AISI 1020 steel sheet were carried out using a 3 kW fiber laser and a 3 m/min welding speed. A thermo-fluid model was used with solid–liquid phase changes, buoyancy forces, and the Marangoni effect in the welding pool to model the physical phenomena. A code in Matlab was developed to solve the inverse problem. The direct problem was solved using COMSOL Multiphysics through the Livelink for Matlab feature. The average absorption efficiency was 79.5% for the welding process. A comparison was made between the geometry of the welding bead obtained in experiments with the numerically calculated welding bead to validate the model. The results obtained in this article are intended to assist simulations in laser welding processes and are in agreement with the literature data.