The Character of Turbulence in the Presence of Smooth and Highly-Irregular Rough Walls
Kenneth T. Christensen
Department of Mechanical Science and Engineering
Most practical wall-bounded turbulent flows of interest occur at high Reynolds numbers (Re) and/or in the presence of non-canonical influences, like roughness, free-stream turbulence, strong pressure gradients, etc. These influences can have a profound impact on the flow and therefore on the overall efficiency and lifetime of a flow system. However, detailed experimental and computational investigations of these complex flows can be extremely difficult. Large-eddy simulation (LES) is a promising computational methodology specifically tailored to high-Re studies wherein the evolution of the larger scales is computed directly while the smaller scales are modeled in some fashion. The first portion of this talk will focus on recent high-resolution PIV measurements in the streamwise–wall-normal plane of a smooth-wall turbulent boundary layer that are used to assess the contributions of underlying vortical structures to the subgrid-scale physics that must be modeled for successful implementation of LES. Substantial evidence exists supporting the presence of hairpin-like vortices in smooth-wall turbulence that serve as the building-blocks for larger-scale vortex packets. The important role that these vortices and vortex packets play in inter-scale energy transfer within the log layer is highlighted using conditional averaging techniques. The second portion of this talk will be devoted to the discussion of on-going PIV measurements of wall turbulence in the presence of a technologically-relevant surface topology–the surface of a damaged turbine blade. The topology considered contains a broad range of spatial scales and its character is quite representative of the surfaces encountered in many practical applications. The statistical and structural features of this rough-wall flow are contrasted with the canonical, smooth-wall case as well as past roughness studies over more classical surface topologies (wire mesh, sand grain, etc.). Of interest is how well classical roughness topologies mimic the impact of this more realistic surface topology, particularly with respect to its influence on the physics of the outer layer.
Christensen received an M.S. in Mechanical Engineering in 1996 from Caltech and
a Ph.D. in Theoretical and Applied Mechanics in 2001 from the