Prediction of flow stress in isothermal compression of Ti–6Al–4V alloy using fuzzy neural network

Abstract

Isothermal compression of Ti–6Al–4V alloy at the deformation temperatures ranging from 1093 K to 1303 K with an interval 20 K, the strain rates ranging from 0.001 s −1 to 10.0 s −1 and the height reductions ranging from 20% to 60% with an interval 10% were carried out on a Thermecmaster-Z simulator. Based on the experimental results, a model for the flow stress in isothermal compression of Ti–6Al–4V alloy was established in terms of the fuzzy neural network (FNN) with a back-propagation learning algorithm using strain, strain rate and deformation temperature as inputs. The maximum difference and the average difference between the predicted and the experimental flow stress are 18.7% and 4.76%, respectively. The comparison between the predicted results based on the FNN model for flow stress and those using the regression method has illustrated that the FNN model is more efficient in predicting the flow stress of Ti–6Al–4V alloy.