Characterization of Flow Behaviors by a PSO-BP Integrated Model for a Medium Carbon Alloy Steel

Abstract

In order to characterize the flow behaviors of SAE 5137H steel, isothermal compression tests at the temperatures of 1123 K, 1213 K, 1303 K, 1393 K, and 1483 K, and the strain rates of 0.01 s−1, 0.1 s−1, 1 s−1, and 10 s−1 were performed using a Gleeble 3500 thermo-mechanical simulator. The analysis results of true stress-strain curves show that the flow stress decreases with temperature increasing and strain rate decreasing. In order to accurately and efficiently characterize the complex flow behaviors, the intelligent learning method backpropagation–artificial neural network (BP-ANN) was combined with the particle swarm optimization (PSO), namely, the PSO-BP integrated model. Detailed comparisons of the semi-physical model with improved Arrhenius-Type, BP-ANN, and PSO-BP integrated model for the flow behaviors of SAE 5137H steel in terms of generative ability, predictive ability, and modeling efficiency were presented. The comparison results show that the PSO-BP integrated model has the best comprehensive ability, BP-ANN is the second, and semi-physical model with improved Arrhenius-Type is the lowest. It indicates that the PSO-BP integrated model can accurately describe the flow behaviors of SAE 5137H steel.

Development of Neural Networks to Study Flow Behavior of Medium Carbon Microalloyed Steel during Hot Forming

“…In addition, since the ANN requires that both the input and output data be between 0 and 1 and in order to accelerate convergence in the training phase, we are interested in the data having small values to ensure the settlement of the network into a stable solution. First, for T, the logarithms of the experimental values are used; then, they are normalized between 0 and 1, as similarly performed by G. Quan et al [15]. For T and σ, the following equation is applied: min max min 0.95 ' 1.05 0.95…”

“…Recently, new works that use ANN as a robust tool capable of predicting the flow behavior of medium carbon steels have appeared. Tize Mha et al [13,14], Quan et al [15] Shekh, Kumar and Nath [16] and Ahmadi et al [17], for different medium carbon steels with different alloy levels, concluded that ANN models represent the best alternative to generate flow curves by using experimental data to interpolate or extrapolate points. Pan et al [6] concluded that the Random Committee algorithm could predict flow stress more effectively than ANNs for a Ni-Cr-Mo low-alloy steel.…”