We simulate turbulence at high fidelity, and build the models that predict it.
At the Turbulence Simulation and Modeling (TSM) Lab at Michigan State University, we understand the fundamentals of turbulence through large-scale, high-fidelity simulations and develop next-gen physics- or data-based predictive methods. Applications address engineering and environmental problems challenging for traditional approaches.

High-fidelity simulations quantify where standard models fail, in unsteady or spatially developing wall-bounded flows under strong pressure gradients.

Roughness-resolved simulation database and physics-based modeling to improve RANS closure for arbitrary surfaces in complex flows.

Pore-resolved simulations of stream–sediment exchange — revealing ‘roughness pumping’ and long subsurface transit times.

Fast-prediction tools for fan acoustics and efficiency — part of the Consortium for Ultra-High Efficiency Quiet Fans.

A nonholonomic-constraint reduced-order model of fish swimming (the fin acts as a frictionless keel), validated against simulations and a physical land-fish prototype.
Prof. Yuan developed large-scale, high-fidelity numerical simulation methods for complex wall-bounded turbulent flows. Her research goal is to push the boundaries of physical understandings of complex, realistic turbulence, and to develop physic-based and empirical closures for a wide range of applications.