Overview
"Scientists study the world as it is; engineers, create the world that has never been." -- Theodore von Karman
Welcome to the Intelligent Computational Mechanics (InCOME) Group led by Dr. Qizhi He. Our group specializes in the convergence of computational mechanics, multiscale and multiphysics material modeling, and scientific machine learning, with main focuses on developing innovative mathematical and computational approaches that facilitate a profound understanding of natural and engineered systems. Our current research interests encompass a range of topics, including advanced numerical methods (FEM, meshfree methods, etc.) for materials modeling, data-driven solid mechanics, damage and fracture mechanics, reduced-order modeling, material inverse design and structural optimization, and physics-informed deep learning for geophysical inverse problems. We are interested in advancing the frontiers of integrating scientific machine learning with computational mechanics.
Our overarching goals are:
1) Advancing the fundamental understanding and predictive capacities of natural and engineered materials, e.g., porous media, composites, biological tissues, and energy storage materials
2) Providing high-performance computational methods and software tools for earth/geo-science in support of environmental health, cleaner energy, and national security
3) Improving the sustainability of civil structures and infrastructure under extreme conditions due to climate change and anthropogenic activities
For more information about our active research, see Research and Publication.
Recent News
- Paper (11/2023): New paper on arXiv: Neural-Integrated Meshfree (NIM) Method: A differentiable programming-based hybrid solver for computational mechanics
- Talk (10/2023) Honored to be invited to present our recent work "Neural Meshfree Method: A Hybrid Solver for Computational Mechanics" at Advances in Computational Mechanics (ACM 2023) — a conference celebrating the 80th birthday of Thomas J.R. Hughes, in Austin, Texas, from October 22-25, 2023.
- Event (10/2023) We hosted a mini-symposium titled "Data-driven Computational Solid Mechanics" during the 2023 SES Annual Technical Meeting. Gratitude to all speakers for their valuable contributions and to our fellow chairs for their generous assistance.
- Member (09/2023) Welcome to new CEGE graduate students Binyao and Zihan to join the Group!
- Paper (08/2023): Our paper "A Hybrid Deep Neural Operator/Finite Element Method for Ice-Sheet Modeling" was accepted for publication in Journal of Computational Physics.
- Short course (07/2023): Dr. He joined the instructor team to offer a Short Course "Machine Learning for Solid Mechanics" at 17th U. S. National Congress on Computational Mechanics. The slides of my lab session can be found here.
- Paper (04/2023): Congratulations to Honghui for publishing a paper titled "Modeling density-driven flow in porous media by physics-informed neural networks for CO2 sequestration" in Computers and Geotechnics. [More]
- Paper (02/2023): Our paper "Physics-informed neural networks of the Saint-Venant equations for downscaling a large-scale river model" was accepted for publication in Water Resources Research.
- Paper (01/2023): Our preprint A Hybrid Deep Neural Operator/Finite Element Method for Ice-Sheet Modeling is released.
- Talk (December, 2022) Dr. He gave a seminar talk about Reduced Order Modeling and Physics-Constrained Deep Surrogate Model at University of Illinois Urbana-Champaign.
- Talk (November, 2022) PhD student Honghui presented his recent studies on "efficient meshfree method and CO2 density-driven flow modeling by using physics-informed neural networks" in the Structures Seminar at the Department of Civil, Environmental, ang Geo- Engineering.
- Event: We are organizing two symposiums on Data-Driven Computational Solid Mechanics and Data-Driven Additive Manufacturing in the 17th U.S. National Congress on Computational Mechanics (USNCCM) at Albuquerque, New Mexico, July 23-27, 2023 (https://17.usnccm.org/). Welcome to submit an abstract to MS 413 Data-Driven Computational Solid and Geological Mechanics or MS 605 Physics-Based and Data-Driven Solutions for Additive Manufacturing before Jan 15th, 2023.
- Member (August 2022) Welcome to new CEGE graduate student Honghui Du to join He Group!
- Grant (June 2022) Our group received a U of M Informatics Institute (UMII) seed grant award (Medium) to develop a novel knowledge-augmented machine learning tool for a fast and reliable defect prediction in 3D printing. Congratulations!
- Member (June 2022) Welcome new group members Ziyang Gao, Amarachi Nzeukwu, and Yuxiang Wan!
- Talk (May 2022): Dr. He will co-chair the session DS04.04: Accelerating Materials Discovery in Symposium DS04 of the 2022 MRS Spring Meeting on May 8-13, Honolulu. He will also present a recent work on the application of physics-constrained deep neural network for redox flow battery modeling on May 10.
- Talk (April 2022): Dr. He gave a contributed talk titled "Enhanced Physics-Constrained Deep Neural Networks for the Redox Flow Battery Modeling,” at AIRES 3: Machine Learning for Robust Digital Twins.
- Event: The USACM Thematic Conference on Meshfree and Finite Element Methods with Applications will be held in Berkeley, USA, September 25-27, 2022. Abstracts can be submitted on the conference website [link] and the deadline for abstract submission is May 1, 2022.
- Paper (Mar 2022) Our paper “Physics-constrained deep neural network method for estimating parameters in a redox flow battery” is accepted for publication in Journal of Power Sources. [More]