Multi-Island Genetic Algorithm and Kriging Model-Based Design of Vehicle Product Comprising Multi-Material

Abstract

In recent years, improvement in multi-material additive manufacturing technology has resulted in technical improvements in multi-material design employed by the automotive industry. Therefore, in this work, an internal trim part (corresponding to an original product composed of polypropylene) of a vehicle was divided into four components using a multi-material design method considering PLA composites. The PLA was reinforced with basalt fibers for realization of the required mechanical properties. The mechanical properties associated with different fiber content (from 0 to 60%) were determined via tensile tests. To reduce the mass of the product, an optimization process combining a Kriging surrogate model with a Multi-Island Genetic Algorithm was used to search for the Pareto solution. The coefficient of determination (R2) and response surface methodology (RSM) surrogate model were used to confirm the validity and accuracy of the Kriging model. The values of R2 were all >0.92 and the low error value of both results demonstrated the effectiveness of the optimization process. Owing to the optimization process, the mass of the PLA composite product was reduced by nearly 9%. Correlation analysis indicated that x2 has the strongest impact on the total mass. Therefore, the optimization process proposed for the multi-material optimal design is feasible and contributes significantly to the attainment of light-weight vehicle parts.