Use of Artificial Neural Networks for Prediction of Phase Transformation Temperatures of Shape Memory Alloys

The knowledge of phase transformation behavior of shape memory alloys has always great importance. In this study, we used artificial neural network (ANN) model to understand the complex relationship between composition and transformation temperatures of Ti-Ni-Pd shape memory alloys. The developed model was able to predict the phase transformation temperatures namely austenite start (As), austenite finish (Af), martensite start (Ms) and martensite finish (Mf) temperatures with good accuracy. The effect Ti, Ni and Pd concentration (at. %) on phase transition temperatures was studied systematically. The influence of Ti-poor and Ti-rich alloy compositions on phase transformation temperature was explained with the help of the index of relative importance. The current model is very useful to guide the actual experiments for designing and optimizing the applications of the shape memory alloys.