Comparing predictions from constitutive equations and artificial neural network model of compressive behavior in carbon nanotube–aluminum reinforced ZA27 composites

Abstract

Hybrid carbon nanotube–aluminum reinforced ZA27 composites under hot compressive forces were investigated in the temperature range of 473 – 523 K with strain rates of 0.01 – 10 s−1. From the experimental data, the flow stress curves for increasing strain exhibit typical flow behavior associated with dynamic recrystallization softening. A comparison of predictions from an artificial neural network model and the constitutive equations to describe the hot compressive behavior was performed. Relative errors varied from −4.14 % to 6.75 % for the artificial neural network model and from −15.93 % to 17.29 % using the constitutive equations. The results indicate that the artificial neural network model was more accurate and efficient in predicting hot compressive behavior.

A Review of Dynamic Mechanical Behavior and the Constitutive Models of Aluminum Matrix Composites

“…Liu et al [133] compared the predictive accuracy of the traditional Arrhenius constitutive model and the ANN model for the deformation behavior of (CNT s -Al)/ZA27 composites under high temperature. It is found that because the traditional Arrhenius equation can only describe stable flow processes, such as strain hardening, the DRV and DRX of material, there is a large difference between the prediction results and the experiment when unstable deformations such as micro-cracks and shear bands occur.…”