Session: 03-08-02: Micromechanics and Multiscale Modeling II

Paper Number: 107355

107355 - Mechanistic Machine Learning Method for Multiscale Analysis of Injection-Molded Composites 

Although CAE has built a strong reputation as a verification, troubleshooting and analysis tool for industrial applications, there is an increasing need for multi-scale multi-physics analysis including new materials and processes, for which current software packages are still missing. Short-fiber-reinforced composites (SFRC) are high-performance engineering materials for lightweight structural applications. Typically, SFRC structures are manufactured by injection molding, which induces heterogeneous microstructures, and the resulting nonlinear anisotropic behaviors are challenging to predict by conventional numerical models. In this work, we will introduce a machine learning-based multiscale analysis method by integrating injection molding-induced microstructures, composite material homogenization, and Deep Material Network (DMN) [1] in the finite element simulation software LS-DYNA. 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 at a computational speed orders-of-magnitude faster than the high-fidelity direct numerical simulation. To model industrial-scale SFRC products, transfer learning [2] is utilized to generate a unified DMN database, which effectively captures the effects of different fiber orientations and volume fractions on the composite mechanical properties. Numerical examples including electronics drop test simulations will be presented to demonstrate the promising performance of this machine learning-based multiscale method released in LS-DYNA R14 [3, 4].

References

[1]      Liu, Z. and Wu, C. T. Exploring the 3D architectures of deep material network in data-driven multiscale mechanics. Journal of the Mechanics and Physics of Solids. 127, 20-46. (2019). https://doi.org/10.1016/j.jmps.2019.03.004

[2]      Liu, Z., Wei, H., Huang, T. and Wu, C. T. Intelligent multiscale simulation based on process-guided composite database. 16^th LS-DYNA International Conference. (2020). https://arxiv.org/abs/2003.09491

[3]      Wei, H., Wu, C.T., Lyu, D., Hu, W., Rouet, F.H., Zhang, K., Ho, P., Oura, H., Nishi, M., Naito, T. and Shen, L. Multiscale simulation of short-fiber-reinforced composites: from computational homogenization to mechanistic machine learning in LS-DYNA. 13^th European LS-DYNA Conference. (2021). https://www.dynalook.com/conferences/13th-european-ls-dyna-conference-2021/composites/wei_ansys_lst.pdf

[4]      Wei, H., Wu, C. T., Hu, W., Su, T. H., Oura H., Nishi, M., Naito T., Chung S., Shen L. LS-DYNA machine learning-based multiscale method for nonlinear modeling of short-fiber-reinforced composites. Journal of Engineering Mechanics. (2022). accepted.

Presenting Author: Haoyan Wei Ansys Inc.

Presenting Author Biography: Dr. Haoyan Wei is an R&D engineer for LS-DYNA development at ANSYS Inc. He earned a Ph.D. from the University of California San Diego in 2018, and his research focuses on accelerated finite element/meshfree methods and reduced-order modeling technologies for multiscale computer-aided engineering analysis. Currently, Dr. Wei is developing a machine learning-based finite element simulation framework in LS-DYNA for nonlinear multiscale analysis of composite materials and structures.

Authors:

Haoyan Wei Ansys Inc.
C. T. Wu Ansys Inc.
Wei Hu Ansys Inc.
Tung-Huan Su Ansys Inc.