Strain hardening prediction of materials using genetic algorithm and artificial neural network

It is very important to analyze the characteristics of materials utilized especially in nuclear engineering. Many approximation of strain hardening phenomena or cooling process have been carried out by experiments, but many cases are costly. There are alternative such as modeling and simulation. The purpose of this study is to predict the properties of material due to a particular strain hardening process for molybdenum and austenitic stainless steel. The optimization for some load to get stress and strain was analysed by genetic algorithm. The strain hardening mechanism behavior under stress and strain of material can be modeled using Neural Network with Backpropagation. Levenberg-Marquardt was selected to reach convergence rapidly. The true strain and stress of molybdenum converges to 0.997 and 60489.821 psi. For the austenitic steel are stabilizes to 0.809 and 158255.290 psi. The estimates mean of an exponentially distributed of molybdenum is 4.5667 and austenitic steel is 4.1667.