Investigation on multi-objective optimization for in-situ laser-assisted machining of glass-ceramic

As a typical optical hard and brittle material, the efficient machining methods of glass-ceramic have always been a research hotspot. Based on previous research, this article conducted surface damage analysis on glass-ceramic using in-situ laser-assisted machining (LAM) orthogonal experimental results, and summarized that the surface damage of glass-ceramic mainly includes scratches, pits, and brittle fracture. Surface analysis confirmed that in-situ LAM can effectively reduce cutting forces and improve surface quality compared to conventional cutting. The artificial neural network (ANN) and genetic algorithm (GA) were used to fit and train and conduct multi-objective optimization for the data from in-situ LAM orthogonal experiments with resultant cutting force and surface roughness as eigenvalues. The Pareto optimal front curve with multiple groups of optimal solutions was obtained through multi-objective optimization using GA. The actual in-situ LAM experimental values were compared with the predicted values in the Pareto front, the relative error of the resultant force and the relative error of the surface roughness are both very small. In-situ LAM experiments based on response surface methodology (RSM) with surface roughness as the characteristic value were conducted. The optimal machining parameters for RSM optimization, as well as the minimum values for resultant force and surface roughness were obtained. Through comparative analysis, it was found that RSM has better multi-objective optimization performance than GA. Research content of this article provides reference and guidance for the multi-objective optimization analysis method of hard and brittle materials such as glass-ceramic after LAM.