Prof. Wing Kam Liu
Department of Mechanical Engineering
Northwestern University

AI-Empowered CAE: Real-Time Engineering Analysis for Rapid Prototyping and High-Fidelity Simulation

ABSTRACT: Traditional computer-aided engineering (CAE) workflows are often hindered by scalability issues, computer-aided design (CAD)-to-mesh bottlenecks, and slow convergence. This talk introduces two agentic artificial intelligence (AI) systems, Immersed Tensor Decomposition (ITD) and Convolution Finite Element Method (C-FEM), designed to overcome these classical limitations. ITD targets the early prototyping stage, using generative AI and tensor-based reduced-order modeling to provide rapid performance predictions directly from sketches, reducing turnaround times from months to minutes. C-FEM focuses on high-fidelity analysis during later development stages. By utilizing machine learning-guided high-order interpolation rather than intensive human-guided remeshing, C-FEM enables automated, near real-time simulation optimized for multi-GPU environments.

BIOGRAPHY: Wing Kam Liu is the Walter P. Murphy Professor of Mechanical Engineering at Northwestern University and a leading expert in computational mechanics, multiscale modeling, and mechanistic data science. A former president of the International Association for Computational Mechanics (IACM), he has pioneered the development of Hierarchical Deep-learning Neural Networks (HiDeNN). Liu is the cofounder of HiDeNN-AI, where he leads the development of HiDeNN-SIM. This platform integrates physics-based modeling with data-driven AI to enable efficient exploration of high-dimensional design spaces. His research has been applied across the aerospace, automotive, additive manufacturing, and semiconductor industries to enhance design efficiency and predictive accuracy.