Temperature-Aware Recurrent Neural Operator for Temperature-Dependent Anisotropic Plasticity in HCP Materials

TL;DR

This paper introduces TRNO, a temperature-aware neural network model that efficiently predicts temperature-dependent anisotropic plasticity in HCP materials, outperforming traditional models in accuracy and speed.

Contribution

The paper presents TRNO, a novel time-resolution-independent neural architecture for modeling complex temperature-dependent plasticity in HCP materials, with improved efficiency and generalization.

Findings

TRNO achieves high predictive accuracy across diverse conditions.

TRNO outperforms GRU and LSTM models in training and prediction.

Multiscale simulations with TRNO are at least 1000 times faster.

Abstract

Neural network surrogate models for constitutive laws in computational mechanics have been in use for some time. In plasticity, these models often rely on gated recurrent units (GRUs) or long short-term memory (LSTM) cells, which excel at capturing path-dependent phenomena. However, they suffer from long training times and time-resolution-dependent predictions that extrapolate poorly. Moreover, most existing surrogates for macro- or mesoscopic plasticity handle only relatively simple material behavior. To overcome these limitations, we introduce the Temperature-Aware Recurrent Neural Operator (TRNO), a time-resolution-independent neural architecture. We apply the TRNO to model the temperature-dependent plastic response of polycrystalline magnesium, which shows strong plastic anisotropy and thermal sensitivity. The TRNO achieves high predictive accuracy and generalizes effectively across diverse loading cases, temperatures, and time resolutions. It also outperforms conventional GRU and LSTM models in training efficiency and predictive performance. Finally, we demonstrate multiscale simulations with the TRNO, yielding a speedup of at least three orders of magnitude over traditional constitutive models.