Unsteady two-Dimensional flow field behind perforated plates on a flat surface

Abstract

The flow over bluff bodies with separation is one of the oldest in fluid mechanics. The separated flows would be considered to have local streamline curvature; such flows are particularly of interest in engineering due to their frequent presence in real-life applications. Pressure fluctuations are of common occurrence in all unsteady fluid flows. A need to measure these has been a long-felt one with reference to aerodynamic applications like jet noise, aircraft cabin noise, etc. Although not recognized widely but of great technical significance, is the fact that pressure fluctuations away from flow boundaries can differ substantially from wall-measured values. Basically, two types of pressure fluctuation measurements in turbulent flows have been attempted in the past, namely surface and within the flow. From two review articles one can understand that pressure fluctuation measurements pose experimental difficulties, particularly for measurements within the flow. In the earlier days, spatial resolution as well as vibration isolation were significant problems. However, presently, the availability of highly sensitive miniature and ultra-miniature pressure transducers has made the task of pressure fluctuation measurements much simpler and more reliable. Studies on two-dimensional separation bubble and associated unsteady pressure fields were conducted for normal plate/splitter combination at zero incidence in a uniform low turbulence region. The normal plate perforation level was varied from 0% to 50% to give different levels of fluid interaction into the otherwise stable bubble that exists for a solid normal plate. The Reynolds number based on step height was varied from 4 × 10³ to 1.2 × 10?. Extensive measurements of mean pressure, surface pressure fluctuations, pressure fluctuations within the flow and longitudinal turbulence intensity were made for all the models. The effect of blockage and free-stream turbulence on the fluctuating pressures on the surface and within the flow are also studied. The facility used for the present experimental study is a suction-type low-speed wind tunnel driven by a 4-bladed fan connected to a 15 HP slip-ring induction motor, having speed control facility. The cross-sectional area of the test section is 610 mm × 610 mm and its length is 2100 mm. The tunnel contraction ratio is 9:1. Several screens and a honeycomb are provided in the upstream settling chamber with a fine mull cloth cover at the bell mouth entry. The transmission of the diffuser vibration was minimized by providing a flexible attachment between the diffuser and the test section. The splitter plate was completely spanning the test section and the perforated normal plates had provision to fit into the splitter plate as two-dimensional bodies. Reattachment length was obtained by flow visualization technique, total pressure probe and shear velocity plots. Of all the three methods adopted, the flow visualization method is more simple and gives better results. The mean pressure measurements were made with the help of pressure tappings provided along the centreline of the splitter plate. The mean static pressure was measured using an alcohol projection manometer via scanivalve. The longitudinal mean velocity and turbulence intensity were measured using a hot-wire anemometer. A similar arrangement was used to measure both the surface and within-the-flow fluctuating pressures. Mean pressures were found to be strongly dependent on perforation level of the normal plate. The shape and size of the bubble vary with different perforated normal plates, that is to say, the bubble gets reduced both in height and length up to 30% perforation level. For higher perforation of the normal plate, the bubble is completely swept out. The study showed that blockage and free-stream turbulence have considerable effect on the fluctuating pressures in the flow field and on the surface. The peak turbulence velocity value observed in the mixing region increases as the reattachment zone is approached and further downstream it decreases. The peak turbulence value occurs around 0.7 Xr to 0.8 Xr, Xr is the reattachment length. The turbulence intensity values are highest for the case of solid normal plate (bleed air is absent) and are lowest for the case of 50% perforation of the normal plate (bleed air is maximum in the present study). The plot of RMS longitudinal velocity fluctuation (u') vs distance measured from the surface of splitter plate (Y) at a section indicates half the maximum value of fluctuation (u'???) designated as (u'?) at a certain distance Y' from the splitter plate. From the analysis of data it was observed that (u'/u'???) is uniquely related with the dimensionless distance Y/Y' for all the perforated normal plates. The maximum value of the shear layer pressure fluctuation is reached upstream of reattachment at an axial location of x equal to about 0.76 Xr. This is in contrast to the observed location of maximum surface pressure fluctuation levels which occur at X/Xr = 0.9. It is interesting to note that for 50% perforation of normal plate, the RMS pressure fluctuation in the flow field gets reduced to around 61.5% as compared to solid normal plate. Analysis of the results shows that the ratio where C'v??? is the maximum fluctuating pressure coefficient, Cpb is the base pressure coefficient and ? is the perforation level in percentage for surface RMS pressure fluctuation levels seems to be constant: C?vmaxCpb(1??)\frac{C'v_{\text{max}}}{C_{pb}(1-\eta)} Cpb?(1??)C?vmax?? For flow field RMS pressure fluctuation levels, it seems to be constant and has a value of about 0.32.