Experimental investigation and genetic algorithm optimisation of stir–ultrasonic cast AZ31–La 2 O 3 nanocomposites

This is the authors' abstract. We don't add key points for this paper.

Optimisation of stir ultrasonic casting parameters is of great importance in enhancing the mechanical properties of Mg-based nanocomposites, especially due to the limitations of AZ31 alloy, which has moderate strength, poor particle wettability, and the development of porosity. La2O3 nanoparticles have been used to enhance AZ31 nanocomposites due to their thermal stability and compatibility with magnesium, thus improving the tensile properties. A three-factor, five-level central composite design was employed to develop a quadratic regression model relating ultimate tensile strength (UTS) to ultrasonic power (UP), ultrasonic time (UT), and squeeze pressure (SP). ANOVA confirmed the statistical significance of the model (F = 184.70, p < 0.0001) with strong predictive capability (R² = 0.964). Sensitivity analysis indicated that UP had the greatest influence on UTS, followed by UT and SP. A genetic algorithm was used to identify the optimal processing parameters of UP = 981.62 W, UT = 69.56 s, and SP = 104.42 MPa, with a predicted maximum UTS of 205.61 MPa. Validation experiments indicated less than 5% variation from predicted results. Microstructural and fractographic studies of the optimised composite material indicated finer equiaxed grains, a uniform distribution of La2O3 nanoparticles, and predominantly ductile fracture characteristics.