Prediction and optimization in mask-assisted laser transmission microjoining thermoplastic urethane and polyamide 6 through finite-element analysis, Kriging model, and genetic algorithm integrated method

Abstract

An integrated approach by combining finite-element analysis (FEA), Kriging model, and nondominated sorting genetic algorithm-II (NSGA-II) is utilized to realize modeling and optimization in mask-assisted laser transmission microjoining thermoplastic urethane and polyamide 6 (PA6). First, a three-dimensional FEA model is developed for obtaining the simulation data of the temperature field distribution that can determine the molten pool geometry. Then based on the initial training points generated by the optimal Latin hypercube sampling, the relationships between input parameters (laser power P, scanning speed V, and clamping force F) and weld quality [weld width (WW) and shear strength (SS)] are approximated through the Kriging model. Meanwhile, the main effects and contribution rates of various input parameters on the joint performance are discussed. Finally, the optimal weld quality is characterized as maximum SS and WW with a desired value, the NSGA-II is carried out to solve the multiobjective optimization problem for searching the Pareto-optimal front. The results of validation experiments under the optimal parameters indicate that the corresponding welding joint quality is significantly superior to that under other parameters.