What we study

Turbulent flows with non-uniformities in space or time. We conduct DNS/LES to understand flow physics and evaluate predictive models of turbulence in practical applications where computational constraints prevent resolving the smallest scales. Focus areas include boundary layers experiencing strong pressure gradients (such as airfoil flows and atmospheric boundary layers) and the examination of quasi-laminarization and boundary layer separation processes under strong pressure gradients.

Research addresses roughness in applications including sea-going vessels, airplane wing icing, turbine blade erosion, and, at larger scale, plant or ocean canopies. Key contributions:

• Built a roughness-resolved simulation database spanning many surface geometries, in equilibrium and non-equilibrium flows.
• Found invariants in the roughness sublayer that inform separation prediction.
• Developed RANS roughness models, including machine-learning closures, that preserve the near-wall stress balance.
• Extended these methods to roughness in more complex flows, such as compressible and rotating wall turbulence.

The group conducts pore-resolved simulations of water and solute exchange in aquatic environments, characterizing exchanges between streams and sediment. Key findings:

• Identified roughness pumping effects (grain roughness induces additional roughness-scale interfacial pressure variation) at grain scale.
• Discovered that grain roughness induces significant subsurface flow paths and long transit times previously ignored in hyporheic exchange models.
• Roughness modulates bedform-induced exchange differently.
• Long-term goal focuses on transport of conservative and reactive solutes across multiscale topographies.

Developing fast-prediction tools for fan acoustic and efficiency characteristics. The group participates in the Consortium III for the Development of Ultra-High Efficiency Quiet Fans, formed by 4 universities and 4 industrial partners. Accomplishments:

• Demonstrated that minimal-span simulations effectively approximate fan blade flows.
• Collected extensive DNS, LES, and measurement databases.
• Developing a generalized wall-pressure-spectrum closure for equilibrium and non-equilibrium boundary layers applicable to acoustic prediction in fan applications.

Developing reduced-order models of undulatory fish swimming validated against high-fidelity simulation and experiment. We combine Lagrangian dynamics, direct numerical simulation, and water-tunnel measurements.

• Modeled the effect of fin in a body-fin model using nonholonomic constraints (fin as a frictionless keel), validated for steady swim in a wide parameter range.
• Built a 3D-printed “land fish” prototype for mechanical validation and a refractive-index-matched silicone water-tunnel model for future PIV measurements of deformation and surrounding flow.