Active Flow Control of Separated Turbulent Flow Over a Hump Using RANS, DES, and LES

Abstract

Most practical flows in engineering applications are turbulent, and exhibit separation. Losses due to separation are undesirable because they generally have adverse effects on performance and efficiency. Therefore, control of turbulent separated flows has been a topic of significant interest as it can reduce separation losses. It is of utmost importance to understand the complex flow dynamics that leads to flow separation and come up with methods of flow control. In the past, passive flow-control was mostly implemented that does not require any additional energy source to reduce separation losses but it leads to increasing viscous losses at higher Reynolds number. More recent work has been focused primarily on active flow-control techniques that can be turned on and off depending on the requirement of flow-control. The present work is focused on implementing flow control using steady suction in the region of flow separation. The present work is Case 3 of the 2004 CFD Validation on Synthetic Jets and Turbulent Separation Control Workshop, http://cfdval2004.larc.nasa.gov/case3.html, conducted by NASA for the flow over a wall-mounted hump. The flow over a hump is an example of a turbulent separated flow. This flow is characterized by a simple geometry, but, nevertheless, is rich in many complex flow phenomena such as shear layer instability, separation, reattachment, and vortex interactions. The baseline case has been successfully simulated by Gan et al., 2007. The flow is simulated at a Reynolds number of 371,600, based on the hump chord length, C, and Mach number of 0.04. The flow control is being achieved via a slot at approximately 65% C by using steady suction. Solutions are presented for the three-dimensional RANS SST, steady and unsteady, turbulence model and DES and LES turbulence modeling approaches. Multiple turbulence modeling approaches help to ascertain what techniques are most appropriate for capturing the physics of this complex separated flow. Second-order accurate time derivatives are used for all implicit unsteady simulation cases. Mean-velocity contours and turbulent kinetic energy contours are examined at different streamwise locations. Detailed comparisons are made of mean and turbulence statistics such as the pressure coefficient, skinfriction coefficient, and Reynolds stress profiles, with experimental results. The location of the reattachment behind the hump is compared with experimental results. The successful control of this turbulent separated flow causes a reduction in the reattachment length, compared with the uncontrolled case. The effects of steady suction on flow separation and reattachment are discussed.