This project aims to predict flow curve of AISI 439 steel at different temperature and loading speed (or strain rate). It is known from classical material mechanics that steel will show softening behavior as the temperature elevates and show hardening behavior as the loading speed increases. Experimental results give good agreement on this point. However, it is discovered that at high speed loading, the material temperature also increases. It means that at this condition, the hardening from loading speed and softening from temperature couple each other. This condition is called adiabatic condition.
Normally, one should apply a correction function to decouple the temperature effect out of the high speed testing results. This yields a new flow curve called isothermal flow curve. But this correction function needs human to calibrate its parameters. Therefore, the goal of this project is to develop machine learning approach to predict flow curve of a steel by inputting adiabatic flow curve and predict its isothermal flow curve.
In addition, an unseen data point is included to test the model's robustness. This point does not relate to training or testing set at all.