Prediction of mechanical properties of cast Mg-Al-Zn alloys

Purpose: The aim of this work was to predict the mechanical properties of the cast magnesium alloys of the MCMgAl12Zn1, MCMgAl9Zn, MCMgAl6Zn and MCMgAl3Zn type, by appliance of artificial neural networks. Design/methodology/approach: In this work there were performed investigations concerning the influence of chemical composition of the Mg-Al-Zn alloys onto the structure and mechanical properties, as well there was carried out the prediction of these properties compared to the mass elements concentration. Findings: It was also determined the type of phases, which crystallise during the alloy crystallisation process - including the well-known Mg17Al12 phase. It was also investigated which elements have the highest influence on the prediction of properties as a result of changed concentration of particular elements. Practical implications: The investigations have been performed for the reason to found application possibilities in the foundry industry, as well to make possible an optimisation of the chemical composition by prediction of mechanical properties based on changeable chemical composition of the alloy for the reason to achieve material with desirable structure and mechanical properties. Originality/value: Chemical composition influences the phases and eutectics crystallised during the solidification process of the alloys, and that fore determine also material structure and properties, as well as allow it to predict these parameters applying artificial neural networks.