Investigation of a Reattaching Turbulent Shear Layer: Flow Over a Backward-Facing Step

1980 ◽  
Vol 102 (3) ◽  
pp. 302-308 ◽  
Author(s):  
J. Kim ◽  
S. J. Kline ◽  
J. P. Johnston

Incompressible flow over a backward-facing step is studied in order to investigate the flow characteristics in the separated shear-layer, the reattachment zone, and the redeveloping boundary layer after reattachment. Two different step-heights are used: h/δs = 2.2 and h/δs = 3.3. The boundary layer at separation is turbulent for both cases. Turbulent intensities and shear stress reach maxima in the reattachment zone, followed by rapid decay near the surface after reattachment. Downstream of reattachnent, the flow returns very slowly to the structure of an ordinary turbulent boundary layer. In the reattached layer the conventional normalization of outerlayer eddy viscosity by U∞ δ* does not collapse the data. However, it was found that normalization by U∞ (δ − δ*) does collapse the data to within ± 10% of a single curve as far downstream as x/xR ≈ 2, the last data station. This result illustrates the strong downstream persistence of the energetic turbulence structure created in the separated shear layer.

AIAA Journal ◽  
1986 ◽  
Vol 24 (4) ◽  
pp. 680-682
Author(s):  
Arthur Rizzi ◽  
Charles J. Purcell

2014 ◽  
Vol 136 (12) ◽  
Author(s):  
Juan D'Adamo ◽  
Roberto Sosa ◽  
Guillermo Artana

Active control over a backward facing step flow is studied experimentally by means of plasma based devices. The Reynolds number based on the step height h is 1520. An electrohydrodynamic actuator (EHD), dielectric barrier discharge (DBD) type, is flush mounted to the step wall. The DBD configuration adds momentum locally, normal to the separated shear layer, thus producing strong modifications downstream. The actuation is periodic and its frequency and amplitude are scrutinized to characterize the flow behavior under forcing. Measures of velocity fields for these flows are obtained from particle image velocimetry (PIV). As reported by previous works, the reattachment length shows an important reduction for an optimum forcing frequency. This value closely matches the shear layer flow natural frequency. On the other hand, the flow is less sensitive to the forcing amplitude though the analysis allows us to optimize the actuation in order to save power consumption.


1964 ◽  
Vol 15 (3) ◽  
pp. 247-280 ◽  
Author(s):  
H. McDonald

SummaryAn analysis is presented which enables the boundary-layer thickness parameters of a re-attaching shear layer to be determined when the free-stream flow upstream of the base is supersonic, the base pressure is known, and die initial boundary layer is turbulent. The application of this analysis to some experimental results, on the flow behind blunt-trailing-edge wings and over a back-step where both the base pressure and the initial boundary layer are known, would appear to indicate that the re-attached profile could be specified by one parameter, namely the transformed shape parameter, the transformation used being a turbulent analogue of the well-known Stewartson-Illingworth transformation of the laminar boundary layer and where the shape parameter is defined as the ratio of boundary-layer displacement to momentum thickness. By adopting a value of the shape parameter in advance, it is possible to use the analysis to determine the base pressure by an iterative process and so, on this basis, it is suggested that this analysis is used to replace the existing recompression criterion of the Chapman-Korst model of separated flow which aims to predict base pressures and is known to be capable of improvement.As part of this investigation, an improvement had to be made to the existing compressible turbulent shear layer velocity profiles of Korst and others and this was achieved by means of the compressibility transformation.


1972 ◽  
Vol 52 (1) ◽  
pp. 113-135 ◽  
Author(s):  
P. Bradshaw ◽  
F. Y. F. Wong

Existing experiments on the low-speed flow downstream of steps and fences, and some new measurements downstream of a backward-facing step, are used to demonstrate the complicated nature of the flow in the reattachment region and its effect on the slow non-monotonic return of the shear layer to the ordinary boundary-layer state. A key feature of the flow is found to be the splitting of the shear layer at reattachment, where part of the flow is deflected upstream into the recirculating flow region to supply the entrainment; the part of the flow that continues downstream suffers a pronounced decrease in eddy length scale, evidently because the larger eddies are torn in two. This phenomenon will occur in all cases where a shear layer reattaches after a prolonged region of separation, either at low speed or in supersonic flow. For simplicity, the discussion in the present paper is confined to low-speed flows.


1987 ◽  
Vol 179 ◽  
pp. 439-468 ◽  
Author(s):  
I. P. Castro ◽  
A. Haque

Detailed measurements within the separated shear layer behind a flat plate normal to an airflow are reported. A long, central splitter plate in the wake prevented vortex shedding and led to an extensive region of separated flow with mean reattachment some ten plate heights downstream. The Reynolds number based on plate height was in excess of 2 × 1044.Extensive use of pulsed-wire anemometry allowed measurements of all the Reynolds stresses throughout the flow, along with some velocity autocorrelations and integral timescale data. The latter help to substantiate the results of other workers obtained in separated flows of related geometry, particularly in the identification of a very low-frequency motion with a timescale much longer than that associated with the large eddies in the shear layer. Wall-skin-friction measurements are consistent with the few similar data previously published and indicate that the thin boundary layer developing beneath the separated region has some ‘laminar-like’ features.The Reynolds-stress measurements demonstrate that the turbulence structure of the separated shear layer differs from that of a plane mixing layer between two streams in a number of ways. In particular, the normal stresses all rise monotonically as reattachment is approached, are always considerably higher than the plane layer values and develop in quite different ways. Flow similarity is not a useful concept. A major conclusion is that any effects of stabilizing streamline curvature are weak compared with the effects of the re-entrainment at the low-velocity edge of the shear layer of turbulent fluid returned around reattachment. It is argued that the general features of the flow are likely to be similar to those that occur in a wide range of complex turbulent flows dominated by a shear layer bounding a large-scale recirculating region.


Sign in / Sign up

Export Citation Format

Share Document