Estimation of Grain Size Distribution of Friction Stir Welded Joint by using Machine Learning Approach

Abstract

Machine learning has widely spread in the areas of pattern recognition, prediction or forecasting, cognitive game theory and in bioinformatics. In recent days, machine learning is being introduced into manufacturing and material industries for the development of new materials and simulating the manufacturing of the required products. In the recent paper, machine learning algorithm is developed by using Python programming for the determination of grain size distribution in the microstructure of stir zone seam of Friction Stir Welded magnesium AZ31B alloy plate The grain size parameters such as an equivalent diameter, perimeter, area, orientation etc. were determined. The results showed that the developed algorithm is able to determine various grain size parameters accurately.