Chapter Three - Experimental and numerical investigation of deformation behavior of dual phase steel at elevated temperatures using various constitutive models and ANN

Abstract

The sheet metal undergoes complex deformation paths, and temperature histories during the industrial forming process. Different process parameters like strain, strain rate, and temperature significantly influence the flow stress behavior of the material during forming. It is important for the designers working in sheet metal forming processes to understand the flow behavior of the material in different working conditions. In the present study, uniaxial tensile tests have been conducted on DP steel at two different temperatures (RT and 400C) and three strain rates (0.0001, 0.001, and 0.01 S−1). The microstructure reveals the increase in grain size. In contrast, large size dimples in fractography show the effect of temperature on the flow behavior of the DP steel. The flow behavior of DP 590 steel at the tested temperatures and strain rates has been predicted using different constitutive and artificial neural network (ANN) models. Various statistical parameters such as correlation coefficient (R), mean absolute error (MAE), and standard deviation (SD) have been used to assess the prediction capability of the studied models. The R, MAE, and SD values observed for Johnson-Cook (m-JC) (0.934, 8.68% and 5.69%), modified Zerlli-Armstrong (m-ZA) (0.962, 5.59% and 3.41%), JC-ZA (0.976, 4.38% and 2.89%), and ANN model (0.991, 1.13% and 1.38%). The Johnson-Cook and Zerlli-Armstrong (JC-ZA) model have to be the best, as it predicts the flow behavior based on the physical assumptions that the ANN model lacks, though it has better statistical parameters compared to JC-ZA.