Comparison of turbulent flows over surfaces of rigid and flexible roughness

The work aims at examining the influence of flexible surface roughness on wall-bounded turbulent flows. The experiments are conducted in a refractive-index-matched turbulent channel (using NaI solution) with a test section of 1.2 m in length and 50 mm square cross section. The rough samples consist of tapered cylinders of 0.35 mm in base diameter with 3 mm in height and 4 mm spacing in a 25 cm stretch. Two types of transparent polydimethylsiloxane (Sylgard 184 and Solaris) are used to generate roughness with about an order of magnitude difference in compliance (i.e. bulk elastic modulus \quad of \quad 1.8 Mpa vs. 0.2 Mpa). The dimension and the elastic modulus of roughness elements are designed such that the rough surface with higher modulus shows no deformation (namely rigid roughness) while the one with lower value deforms appreciably under the present flow conditions. Flow measurements are conducted using digital inline holographic PTV to obtain 3D velocity fields in a volume of 10 mm (streamwise) x 50 mm (wall-normal) x 10 mm (spanwise) with a temporal resolution of 0.16 ms and a spatial resolution of 1.1 mm/vector, above both rigid and deformable rough surfaces under two Reynolds numbers. A selection of instantaneous samples, mean velocity profile, turbulent fluctuation, and energy spectra as well as conditionally-sampled flow fields are analyzed to quantify the effect of roughness compliance on general turbulent statistics and coherent flow structures.