An experimental investigation of an incompressible turbulent boundary layer in the vicinity of separation

Abstract

The following is an investigation into the effects of small changes in the static pressure distribution on the development of an axisymmetric, incompressible, turbulent boundary layer with incipient separation. The pressure distribution was closely controlled to study three cases, in which the skin friction was either approximately zero, slightly negative, or slightly positive along a fixed length. Mean flow and turbulence structure in air were measured using pulsed-wire and hot-wire anemometry.These measurements show characteristic properties of steady turbulent boundary layers both on the verge of separation and with a long, shallow separation bubble. There is an asymptotic velocity defect law near separation. A linear relationship between χW, the wall value of the reverse-flow parameter, and the form parameter H12 suggests the importance of χW in characterizing the boundary layer. The occurrence of the first reverse-flow events coincides with the vanishing of the logarithmic law, the asymptotic mean velocity profile, and a sudden drop in the values of the skewness SW and the flatness FW of the skin friction. This implies that the presence of instantaneous reverse flow is associated with a complete change in the nature of the near-wall flow, well upstream of mean separation. As the three cases were investigated in a single test section under closely controlled conditions with the same experimental techniques, this data set is well suited to a sensitivity study. It is possible to show the effect of small changes in the upstream pressure gradient on the separation region and to distinguish the effect of mean reverse flow from that of the adverse pressure gradient. This effect of the reverse flow is displayed most clearly in a plateau in $\overline{u^{\prime 2}}$ near the wall and in unusual behaviour of the skewness and the flatness profiles over the inner half of the boundary layer.

Direct numerical simulation of a separated turbulent boundary layer

“…The experiments on separation of a flat plate turbulent boundary layer include the works of Perry & Fairlie (1975), Simpson et al (1977Simpson et al ( , 1981a, Dengel & Fernholz (1990), Driver (1991) and Alving & Fernholz (1996, 1995. Some of these investigations have also tried to develop different scalings of the velocity profile in both outer and inner variables.…”

“…They compared the Durbin-Belcher and Perry-Schofield scalings, with the conclusion that the latter works better than the former. However, they did not actually use the velocity scale proposed by Perry & Schofield (1973), but rather determined their velocity scale so that the velocity profiles close to separation collapse with the profile given by Dengel & Fernholz (1990).…”

2D planar jet control for manipulating the separation bubble behind the vertical fence

“…9 shows the reverse-flow factors, which means the percentage of negative streamwise velocities from the total measurement data. If the value of the reverse-flow factor is 50%, that location approximately indicates the dividing streamline or reattachment length at Y/H=0 position (Dengel and Fernholz [20]). Siller and Fernholz [9] reported that the reverseflow region was only reduced in the time mean data.…”

Evaluation of a Vortex Generator Model in Adverse Pressure Gradient Boundary Layers

“…The separation bubble in experiments is defined as the region where at least 50% back flow at the wall is developed, i.e., the wall back flow coefficient wall 0:5. According to Dengel and Fernholz [18], was extrapolated to the wall from the data points in the region y 1:5 to 10 mm in order to estimate wall . The resulting geometrical properties such as the separation location x sep , the reattachment location x att , the length l sep as well as the height of the separation bubble h sep [defined as Uh sep 0] are given in Table 1.…”