Prediction on hot deformation behavior of spray formed ultra-high strength aluminum alloy-A comparative study using constitutive models

The hot deformation behavior of spray formed ultra-high strength aluminum alloy was investigated by isothermal compression tests at 613-723 K and 0.001-1.0 s(-1). The flow stresses were predicted by using phenomenological model (strain-compensated Arrhenius equation), physical-based model (diffusion model) and artificial neuron network. The predictive ability has been weighed through statistical analysis including relative errors, correlation coefficients and box charts. It has shown the artificial neural network contains the best capability to predict the flow stress over the applied phenomenological and physical-based model. The strain-compensated Arrhenius can track the flow stress well in all processing conditions. Moreover, the diffusion model can well describe the metallurgical mechanisms during deformation by the microstructure observation. (C) 2017 Elsevier B.V. All rights reserved.