Buoyancy modulation and scalar transport in turbulent flows over smooth and rough walls

Edward Silberman Award Ceremony and Seminar

Silberman Award Recipient: Michael Heisel, PhD Candidate in Civil, Environmental, and Geo- Engineering

Keynote Speaker: Elie Bou-Zeid, Associate Professor of Civil and Environmental Engineering; Director, Program in Environmental Engineering and Water Resources, Civil and Environmental Engineering, Princeton University

The workings of how buoyancy modifies the statistics and structure of turbulence in wall-bounded flows, such as the atmospheric boundary layer (ABL), continues to be a topic with various open fundamental questions and multiple pressing applied needs. In the first part of this talk, a model that connects the budgets of the three velocity variance components and captures how the energy redistribution terms vary with the flux Richardson number is proposed. The results of this model are first tested against large eddy and direct numerical simulations. The model is then used to inquire about how the turbulence transitions between different regimes as the Richardson number varies. It is shown that the asymptotic limits of this transition are reasonably reproduced and are consistent with a large corpus of experiments under both unstable and stable conditions.

The complexity added to the problem by wall roughness is then considered. Since heat (or buoyancy) is advected with the flow, broad similarity is expected between the widely-studied momentum transfer problem and its scalar counterpart. However, unlike momentum that is dominated by form drag over very rough walls, scalar transport must occur through the viscous exchanges at the solid-fluid interface, which might result in transport dissimilarity. To examine underlying physics of the problem, a suite of large-eddy simulations (LES) is conducted over large three-dimensional roughness elements. The results point to (i) the importance of matching the Reynolds number of the simulations to real-world values, and (ii) the significance and dissimilarity (momentum versus scalars) of the dispersive fluxes inside and right above near the canopy. Implications for parameterizations in urban canopy models are then discussed.