A comparative study on Arrhenius-type constitutive equations and artificial neural network model to predict high-temperature deformation behaviour in 12Cr3WV steel

Abstract

The hot compressive deformation behaviour in 12Cr3WV steel was conducted on a Gleeble-1500 thermo-mechanical simulator at the temperature range of 1223–1373K with the strain rate in the range of 0.01–30s−1 and the height reduction of 60%. Based on the experimental results, strain compensated Arrhenius-type constitutive equations and an artificial neural network (ANN) model with a back-propagation learning algorithm were developed for the characterization and prediction of the high-temperature deformation behaviour in the steel. And then a comparative predictability of the constitutive equations and the trained ANN model were further evaluated in terms of the correlation coefficient (R), the average absolute relative error (AARE) and the relative error. For the constitutive equations, R and AARE were found to be 0.9952% and 3.48% respectively, while for the ANN model, 0.9998 and 0.58% respectively. The relative errors between experimental and predicted flow stress computed from the constitutive equations and ANN model were respectively in the range of −15.46% to 10.46% and −4.12% to 4.08%. Moreover, the relative error within ±1% was observed for more than 85% of the test data sets of ANN model, while only 32% of the test data sets for the constitutive equations. The results indicate that the trained ANN model is more efficient and accurate in predicting the hot compressive behaviour in 12Cr3WV steel than the Arrhenius-type constitutive equations.