Favorable pressure gradient turbulent flow over straight and inclined ribs on both channel walls

Abstract

This paper reports on experimental study of turbulent flows over straight and inclined transverse ribs of square and triangular cross sections attached to the bottom and top walls of an asymmetric converging channel. The pitch-to-height ratio of the ribs was 10. A particle image velocimetry technique was used to conduct extensive velocity measurements at channel midspan and in planes close to the leading and trailing edges of the inclined ribs. From these measurements, spatial averaged profiles of the mean velocity and higher order statistics were obtained to study the effects of rib geometry, pressure gradient, spanwise plane, and rib inclination on the flow characteristics. The results show that rib geometry has no significant effects on the mean flow and turbulent quantities. The roughness effects produced by the straight ribs outweighed pressure gradient effects in the inner region of the flow. As a result, the skin friction coefficient is nearly independent of pressure gradient. The Reynolds shear stress and turbulent transport of the shear stress are also independent of pressure gradient. On the contrary, favorable pressure gradient decreased the Reynolds normal stresses in the outer region and increased the magnitudes of the triple velocity correlations and transport of turbulent kinetic energy (TKE). The three-dimensional secondary motion produced by the inclined ribs distorted the mean flow pattern and substantially diminished the ribs’ effectiveness to augment skin friction and turbulence. For example, the skin friction over the inclined ribs is only 50%–70% of the value measured over the straight ribs. Furthermore, the size of equivalent sand grain required to produce the same amount of drag is one-tenth to one-third of the rib height for the inclined ribs compared to two- to fourfold for the straight ribs. The inclined ribs also reduced the level of the Reynolds stresses, triple velocity correlations, and transport of both the turbulent kinetic energy and Reynolds shear stress compared to the straight ribs. The skewness factors showed that important structural differences exist between flows over straight and inclined ribs.

Numerical simulation of a spatially developing accelerating boundary layer over roughness

“…FPG leads to an increase of C f and a decrease of bulk Reynolds number. In the outer layer, competing effects of FPG and roughness were found on the mean flow and the Reynolds stresses; however, close to the rough wall, both roughness and acceleration intensify the fluctuations significantly, with the Reynolds-stress anisotropy unaffected by FPG; this was different from the observations of both Coleman et al (1977) and Tachie & Shah (2008). Similar observations on the opposite effects of FPG and roughness and the higher C f were also made from large-eddy simulations (LES) of spatially developing flows over roughness by Piomelli & Yuan (2013).…”

“…Similar observations in a spatially developing flow by Cal et al (2009) support this theory. Tachie & Shah (2008) also reported the increase of drag. In a sink flow, on the other hand, U RS decreases (relative to ν/k) with K due to the decrease of flow rate Q, as K = ν/Q; thus k + decreases, and the flow tends towards the hydraulically smooth regime and laminarization (Yuan & Piomelli 2014b).…”

“…The turbulent kinetic energy decreased when normalized using U ∞ , especially its wall-normal and spanwise components, while the streamwise component was unchanged, leading to higher anisotropy. Tachie and co-workers (Tachie, Agelin-Chaab & Shah 2007;Tachie & Shah 2008) conducted experiments in flows over bar roughness under strong FPG achieved with converging side walls or an inclined bottom wall in open-channel flows. Contrary to the observation of Coleman et al (1977), their data showed that C f , the mean-velocity defect, and the Reynolds stresses were largely independent of the pressure gradient.…”

Extension of roughness noise to bluff bodies using the boundary element method

“…The joint effects of a favourable pressure gradient and roughness on a turbulent boundary layer in the fully rough regime are of interest here. There are a number of studies in the literature focusing on the effect of mild and moderate pressure gradients on the mean and Reynolds stress profiles [24,25,26,27]. Cal et al [24] showed that the effect of roughness on the Reynolds stresses could be scaled with the friction velocity for a zero-pressure gradient and favourable pressure gradient, separately.…”

“…They also observed an increase of the roughness Reynolds number along the accelerated rough wall boundary layer, discarding the appearance of transitional effects. Tachie and Shah [25] studied a turbulent channel with two-dimensional roughness. They concluded that the effect of the pressure gradient on the mean and Reynolds stress profiles was much weaker than the effects of roughness, so that the pressure gradient could be ignored.…”