Session: Flow Visualization and Regular Poster Session

Paper Number: 158150

158150 - Effects of Leading-Edge Separation Angle on the Characteristics of Separated Flows Around a Rectangular Cylinder 

Abstract: 

Introduction: Separated flows around bluff bodies with sharp-edged rectangular cross-sections are exemplified by the formation of massive recirculation regions as well as elevated turbulence levels and drag compared to those observed around streamlined bodies. In practical applications, the leading-edge geometry of rectangular cylinders is modified to change the characteristics of the separated shear layers and improve their aerodynamic performance. For instance, rounded and chamfered corners as well as extended forebodies of triangular cross-section have been used to control the leading-edge separation angle which, in turn, alters the vortex shedding process and drag characteristics.  The objective of this workis to investigate experimentally the effects of leading-edge separation angle on the drag and turbulent characteristics of separated flows around rectangular cylinders with a chord-to-length ratio of 6. The rectangular cylinders are of thickness t = 20 mm and chord length c = 120 mm.  The investigation is comprised of six leading-edge separation angles, including a baseline rectangular cylinder with sharp edges (θ = 90°), and five triangular cross-sections with half interior angles of θ = 15°, 30°, 45°, 60°, and 75°.

Experimental Method: The experiments were performed in a recirculating open water channel. The test section has dimensions of 6000 mm long, 600 mm wide, and 450 mm high. The water was seeded with silver-coated hollow glass spheres of mean diameter of 10 µm, and specific gravity of 1.4. The water depth was kept at 420 mm for all test cases. The flow was illuminated with a diode pumped dual-cavity high-speed neodymium-doped yttrium lithium fluoride laser with maximum pulse energy of 30 mJ/pulse. Two 4 MP CMOS cameras with resolution 2560 pixels × 1600 pixels were positioned side-by-side to image the flow field simultaneously. The dimensions of the field of view of the cameras were 143.9 mm × 90.0 mm for the first field of view, and 140.9 × 88.0 mm for the second field of view, in the streamwise and vertical direction, respectively.  For each test case, 9000 statistically independent image pairs per field of view were collected at a sampling frequency of 3 Hz.  Data acquisition, image processing, and vector calculations were performed using a commercial software (DaVis version 10). A multi-pass cross-correlation algorithm was used for velocity vector calculations. An initial interrogation area (IA) of 128 pixels × 128 pixels with 50% overlap followed by four final passes of 24 pixels × 24 pixels IA with 75% overlap was used. The vector spacing was 0.33 mm.

Preliminary Results: Contour plots of the streamwise mean velocity are used to visualize the mean flow topology. The separated shear layer from the leading edge reattached on the cylinder regardless of leading-edge angle, and the reattachment length—which was measured as the streamwise distance from the leading edge to the mean reattachment point—decreased linearly from L_r/t = 4.2 to 0.9 as θ decreased from 90° to 30°. No distinct separation bubble was observed for the case of θ =15°. The recirculation length of the wake behind the cylinders also shows distinct dependence on leading edge angle, increasing linearly from L_w/t = 0.83 to 1.5 as θ increases from 15° to 90°. Preliminary results also demonstrate that the mean velocities, Reynolds stresses, and third order moments over the cylinders and in the wake region are strongly dependent on the leading edge separation angle.   

Presenting Author: Bronwyn Rempel University of Manitoba

Presenting Author Biography: Bronwyn Rempel is a M.Sc. student at the University of Manitoba in the department of Mechanical Engineering.