Development of Predictive Models for Formability Study of Ti-6Al-4V alloy at Elevated Temperatures

Abstract

Deep drawing experiments have been performed in order to study formability of Ti-6Al-4V alloy sheet at temperature ranging from room temperature to 400°C. It is found that below 150°C, formability of the material is very poor and above 150°C till 400°C, Limiting Draw Ratio (LDR) is found to be 1.86 which is considerably lesser than other structural alloys. Moreover, one of the important qualitative aspects in formability is to study the effect of temperature on thickness distribution of properly formed cups. In this paper, thickness distribution at various temperature and blank diameter is evaluated from experiments and is modelled using intelligent techniques such as Artificial Neural Networks (ANN) and least squares Support Vector Machines (SVM). These techniques give an accurate prediction of thickness strain which is helpful for prediction of failures in the punch-corner region and wall region at any unknown punch diameter and temperature. The predicted result using intelligent techniques shows good agreement with experimental data.