Application of improved GRNN model to predict interlamellar spacing and mechanical properties of hypereutectoid steel

In this study, mechanical properties of hypereutectoid steels after thermomechanical processing have been evaluated using standard mechanical tests and a theoretical model has been developed using a novel artificial neural network approach. The K-folder cross validation (K-CV) was used to improve the generalized regression neural network (GRNN) to model and predict the lamellar spacing and mechanical properties of hypereutectoid steels with a small sample data. The independent variables in the model were the alloying elements. The dependent parameters were the interlamellar spacing, tensile strength, yield strength, section shrinkage and hardness. A comparison between the predicted values by improved GRNN with the experimental data indicates that the well trained model can provide accurate results. The effects of alloying elements can be evaluated by the developed model and help to achieve the desired lamellar spacing and mechanical properties. This would help the materials engineer suitably design the alloying compositions to obtain the desired combination of plasticity and strength properties.