Many physical systems that are important in engineering, fluid mechanics, and astrophysics are heterogeneous — for example, when a solid interacts with a liquid, or when an interdependence develops between the physics of an electromagnetic field and that of a plasma flow. Although numerical methods based on partial differential equations exist for simulating individual components of such heterogeneous systems, combining them into a more comprehensive multiphysics model that both represents the interaction and preserves the structure of the individual systems remains a complex problem for both physics and for the field of high-performance computing (HPC).
The project SNuBIC is investigating methods for the modeling and simulation of coupled systems described by partial differential equations. In Stuttgart we are focused on developing improved methods for multiphysics simulations at interfaces between physical systems, particularly from the perspective of making them computationally efficient when using high-performance computing.
Once completed, our multiphysics coupling framework could be used by scientists and engineers to improve simulations of heterogeneous systems. One major application area is astrophysics, where scientists are interested in understanding the interactions between magnetic fields and physical objects, which are described using different numerical methods. Our methods preserve numerical structures to ensure that the magnetic field remains approximately divergence free, while offering coupling capabilities for simultaneously simulating diverse regimes in an efficient way.
01. November 2022 - 31. October 2026
www.snubic.io
Exascale Computing
HPC for Industry
Optimization & Scalability
Training and Scalable Algorithms
DFG
See all projects
Research Scientist