Cheng-Hau Yang
Postdoc Appointee
Argonne National Laboratory
I am a postdoctoral researcher at Argonne National Laboratory, where I focus on developing scalable numerical frameworks for solving multiphysics problems. I work under the supervision of Dr. Gary Hu and Dr. Mark Messner.
I have extensive experience developing multiphysics finite‑element frameworks for solving partial differential equations (PDEs), including work in MOOSE (ANL), deal.II (Google X), and Dendro (ISU).
I earned my Ph.D. in Mechanical Engineering from Iowa State University, advised by Dr. Baskar Ganapathysubramanian and Dr. Adarsh Krishnamurthy. My doctoral research focused on advanced numerical methods—including the Shifted Boundary Method (SBM), in collaboration with Dr. Guglielmo Scovazzi, and the Finite Cell Method (FCM)—to solve partial differential equations on irregular domains. This work enhanced the accuracy, robustness, and scalability of simulations across various engineering applications.
During my Ph.D., I interned at Google X, where I worked on parallel parasitic extraction simulations using deal.II, under the supervision of Dr. Dino Ruić. There, I developed a remeshing-free circuit geometry optimization technique that reduced runtime by 50% per optimization iteration. I also improved software usability by introducing configuration file support and integrating GoogleTest for automated testing.
Beyond numerical methods, I have actively contributed to the development of high-performance computing (HPC) software, the creation of benchmark datasets for scientific machine learning (SciML), and the integration of data-driven models into physics-based simulations. My recent work includes multiphase flow simulations, machine learning-assisted constitutive modeling, the moving immersed (shifted) boundary method, and fluid-structure interaction simulations.
I am passionate about bridging scientific computing and machine learning, advancing simulation capabilities for complex systems, and developing open-source tools to support and empower the research community.
I earned my Ph.D. in Mechanical Engineering from Iowa State University, advised by Dr. Baskar Ganapathysubramanian and Dr. Adarsh Krishnamurthy. My doctoral research focused on advanced numerical methods—including the Shifted Boundary Method (SBM), in collaboration with Dr. Guglielmo Scovazzi, and the Finite Cell Method (FCM)—to solve partial differential equations on irregular domains. This work enhanced the accuracy, robustness, and scalability of simulations across various engineering applications.
During my Ph.D., I interned at Google X, where I worked on parallel parasitic extraction simulations using deal.II, under the supervision of Dr. Dino Ruić. There, I developed a remeshing-free circuit geometry optimization technique that reduced runtime by 50% per optimization iteration. I also improved software usability by introducing configuration file support and integrating GoogleTest for automated testing.
Beyond numerical methods, I have actively contributed to the development of high-performance computing (HPC) software, the creation of benchmark datasets for scientific machine learning (SciML), and the integration of data-driven models into physics-based simulations. My recent work includes multiphase flow simulations, machine learning-assisted constitutive modeling, the moving immersed (shifted) boundary method, and fluid-structure interaction simulations.
I am passionate about bridging scientific computing and machine learning, advancing simulation capabilities for complex systems, and developing open-source tools to support and empower the research community.