arXiv subjects

Tung-Huan Su

Publications and source records attributed to Tung-Huan Su.

3 recordsLinked to original sources

arXiv2026

A Parametric Multiscale Surrogate Framework Based on Texture-Generalizable Deep Material Networks for Polycrystal Modeling

This work presents a computational framework for parametric multiscale surrogate modeling of polycrystalline materials. The framework integrates a physics-based Deep Material Network (DMN), specifically an Orientation-aware interaction-based Deep Material Network (ODMN), with two data-driven components: a Texture-Adaptive Clustering and Sampling (TACS) scheme that provides a reduced yet statistically consistent representation of crystallographic texture, and a Graph Neural Network (GNN) that infers the micromechanical equilibrium parameters of the ODMN from grain-level interaction graphs. Referred to as the TACS-GNN-ODMN framework, this combination introduces a microstructure-to-parameter mapping that constructs fully parameterized surrogate models for previously unseen microstructures without retraining. The resulting surrogate model preserves the micromechanical structure of the underlying formulation while substantially reducing computational cost. Numerical results show that the framework accurately predicts nonlinear mechanical responses and crystallographic texture evolution under several loading conditions, in close agreement with full-field direct numerical simulations (DNS). The method further achieves more than two orders of magnitude speed-up over fast Fourier transform (FFT)-based simulations. The framework thus provides an efficient and physically consistent strategy for multiscale modeling of polycrystalline materials and is well suited to large-scale simulations and real-time applications.

cs.CE↗

arXiv2025

Orientation-aware interaction-based deep material network in polycrystalline materials modeling

Multiscale simulations are indispensable for connecting microstructural features to the macroscopic behavior of polycrystalline materials, but their high computational demands limit their practicality. Deep material networks (DMNs) have been proposed as efficient surrogate models, yet they fall short of capturing texture evolution. To address this limitation, we propose the orientation-aware interaction-based deep material network (ODMN), which incorporates an orientation-aware mechanism and an interaction mechanism grounded in the Hill-Mandel principle. The orientation-aware mechanism learns the crystallographic textures, while the interaction mechanism captures stress-equilibrium directions among representative volume element (RVE) subregions, offering insight into internal microstructural mechanics. Notably, ODMN requires only linear elastic data for training yet generalizes effectively to complex nonlinear and anisotropic responses. Our results show that ODMN accurately predicts both mechanical responses and texture evolution under complex plastic deformation, thus expanding the applicability of DMNs to polycrystalline materials. By balancing computational efficiency with predictive fidelity, ODMN provides a robust framework for multiscale simulations of polycrystalline materials.

cs.CE↗

arXiv2023

LS-DYNA Machine Learning-based Multiscale Method for Nonlinear Modeling of Short Fiber-Reinforced Composites

Short-fiber-reinforced composites (SFRC) are high-performance engineering materials for lightweight structural applications in the automotive and electronics industries. Typically, SFRC structures are manufactured by injection molding, which induces heterogeneous microstructures, and the resulting nonlinear anisotropic behaviors are challenging to predict by conventional micromechanical analyses. In this work, we present a machine learning-based multiscale method by integrating injection molding-induced microstructures, material homogenization, and Deep Material Network (DMN) in the finite element simulation software LS-DYNA for structural analysis of SFRC. DMN is a physics-embedded machine learning model that learns the microscale material morphologies hidden in representative volume elements of composites through offline training. By coupling DMN with finite elements, we have developed a highly accurate and efficient data-driven approach, which predicts nonlinear behaviors of composite materials and structures at a computational speed orders-of-magnitude faster than the high-fidelity direct numerical simulation. To model industrial-scale SFRC products, transfer learning is utilized to generate a unified DMN database, which effectively captures the effects of injection molding-induced fiber orientations and volume fractions on the overall composite properties. Numerical examples are presented to demonstrate the promising performance of this LS-DYNA machine learning-based multiscale method for SFRC modeling.

cs.CE↗