Abstract: The effect of processing parameters on its microstructure after solution treatment is studied by isothermal compression test and metallurgical analysis The predicting model for the relation between equivalent grain size and recrystallized grain volume fraction with strain strain rate and temperature for Ti-15-3 alloy is developed with an artificial neural network method This model is incorporated into rigid-viscoplastic thermo-coupled FEM Code and the microstructure of Ti-15-3 alloy after hot back-extrusion process and solution treatment is simulated. Corresponding experimental research is performed.The simulated results are in good with the measured ones. These studies are significant for determining the processing parameters of Ti-15-3 alloy.

Key words: Artificial neural network FEM, Ti-15-3 alloy Microstructure