Artificial Neural Network Models to predict workability parameters for Dynamic Recrystallisation domain and Instability regimes

For the design of bulk metal working processes like forging, rolling and extrusion, the process parameters namely temperature, strain rate and strain play an important role. In recent years, these parameters are decided on the basis of processing maps, which exhibit safe domains and unsafe regimes for processing. Processing map is an outcome of Dynamic Materials Model (DMM) and the interpretation of the domain of dynamic recrystallization (safe) and regimes of flow instability and cracking (unsafe) in the maps requires considerable expertise as well as detailed microstructural evidences. This paper explains a technique of capturing these features of the processing map using Artificial Neural Network (ANN) models. Details about the ANN architecture, training and testing of these models are explained. The results obtained using the ANN models are compared and validated with those obtained from the processing maps using the DMM