Optimized mechanical properties of magnesium matrix composites using RSM and ANN

Abstract

This study investigated the effect of stirring speed, stirring time, and particle weight fraction on the mechanical properties of magnesium matrix composites (Mg-MMCs) synthesized by the stir casting process. In addition, response surface methodology (RSM) and artificial neural network (ANN) models were used to optimize process parameters and create predictive models for evaluating the mechanical properties of Mg-MMCs. According to the results, the optimal parameter conditions for maximum mechanical properties based on the desirability function methodology were achieved at a stirring speed of 312.8 rpm, a stirring time of 11.9 min, and a weight fraction of particles of 9.9 wt%. In conjunction with the ANN and RSM models, the expected findings in this study will present beneficial recommendations in selecting main process parameters, contributing to the development of a good database for magnesium composites in manufacturing and mechanical performance evaluation.