First-principles Design And Grain Size Prediction Model Of Magnesium-containing Lightweight High Entropy Alloys

High entropy alloys(HEA)have great application prospects in the field of engineering because of excellent strength,hardness,toughness and thermodynamic stability.The traditional high-entropy alloys are mostly transition group elements,which makes the overall density of the alloy larger and limits further application.At the same time,the research on the formation mechanism of high-entropy alloys is still immature,so reducing the density of high-entropy alloys and deepening the understanding of the formation mechanism of high-entropy alloys have become an important research direction in the field.In order to achieve the lightweight and diversified development of high-entropy alloys,a stable structure with magnesium light weight and high entropy alloys is obtained by the first principle calculation,and its mechanical properties are studied.A method of predicting grain size by machine learning is proposed to solve the grain refinement problem of light alloys.The main research contents are as follows:Firstly,the Mg AlAgNdSn with the highest structural stability was selected by calculating the forming enthalpy of the Mg light and high entropy alloy with the highest structural stability.The effect of Si on the mechanical properties of magnesium containing lightweight alloys was further investigated.The results show that the plasticity of magnesium alloy can be improved by increasing the content of the Si element.Secondly,the criteria for considering the comprehensive properties of magnesium based light entropy alloys were established and the MgAlAgErSi,MgAlAgGdSi and MgAlAgNdSi structures with good comprehensive properties were selected.Finally,grain size models suitable for light alloy nucleation and growth are summarized,as well as the application range,advantages and disadvantages of various models.According to the limitations of traditional models,a method based on machine learning is proposed to predict grain size.The results showed that the mean value of R2 in the training set was 0.964 and the mean value of R2 in the test set was 0.809,which was better than that predicted by traditional models.In the paper,the tentative research on MgAlAgXY lightweight high entropy containing magnesium can provide some useful guidance for the design and development of high entropy alloys.The prediction of grain size by machine learning methods will deepen the theoretical knowledge of metal solidification and nucleation,which is of great significance to realize industrial production and process optimization of magnesium-containing lightweight alloy.