Microstructural prediction through artificial neural network (ANN) for development of transformation induced plasticity (TRIP) aided steel

The prediction of the amount of retained austenite as a function of chemical composition and heat treatment is important for achieving the desired properties in TRIP (Transformation Induced Plasticity) aided steel. In the present work, three experimental steels (CMnSiAlP, CMnSiAlNb and CMnSiNb) made in vacuum induction furnace were suitably heat treated in hot dip processing simulator (HDPS) to produce multiphase TRIP microstructure. The process parameters were determined with the aid of multilayered perception (MLP) based artificial neural network (ANN) models in combination with the results of the study of the transformation behaviour. Amount of retained austenite in microstructure measured by optical microscopy and X-ray diffraction technique had shown a good agreement with that predicted through the afore mentioned model. All three alloys were found to have an excellent strength–ductility balance and significantly good strain hardening exponent (n) value. Among the three grades, CMnSiAlNb grade was observed to have a better combination of properties in terms of high strength and ductility.