Multi-objective optimization of engineering properties for laser-sintered durable thermoplastic/polyamide specimens by applying a virus-evolutionary genetic algorithm

Abstract In this work an enhanced virus-evolutionary genetic algorithm has been developed and applied to optimize a three-objective optimization problem related to an important additive manufacturing/rapid prototyping operation known as selective laser sintering (SLS). Selective laser sintering is a manufacturing process where laser is used as the power source to sinter powdered materials such as metals, superalloys, or even nylon and polyamides. The laser source is collectively applied to points in space determined by a solid model, and further binds the material to fabricate solid components. A response surface experiment was established to study the effect of crucial SLS process parameters on the optimization objective of density, hardness, and tensile strength. With reference to the experimental results, a statistical analysis was conducted to further obtain regression models with high coefficient of determination so that the objectives can be reliably predicted. The models were considered as the objective functions for simultaneously maximizing all three objectives and study the trade-off among them. To compare the results and show that the proposed virus-evolutionary genetic algorithm is prominent, two other population-based heuristics were applied to the same problem, namely multi-objective Greywolf algorithm (MOGWO) and multi-verse optimization algorithm (MOMVO). To evaluate the algorithms and judge superiority with reference to the non-dominated solution sets obtained, the hypervolume indicator as well as statistical results were considered for examination. It was shown that the proposed virus-evolutionary genetic algorithm can advantageously maximize all three objectives simultaneously and constitute a beneficial optimization module with strong potentials to optimize not only the SLS problem formulated in this work, but general engineering optimization problems with quite promising and practically viable outcomes.