PASSIVE FLOW CONTROL OF SHOCK-WAVE/TURBULENT-BOUNDARY-LAYER-INTERACTIONS USING MICRO VORTEX GENERATORS

Abstract

We evaluate the suitability of micro vortex generators for the passive flow control of shock-wave/turbulent boundary layer interactions. For this purpose, implicit large eddy simulations using the adaptive local deconvolution method are performed. The flow configuration consists of an oblique shock with deflection angle β = 9.5°, impinging on a turbulent boundary layer at Ma∞ = 2.31 and Re = 67.4 ·103. Analysis focuses on the assessment of the relative displacement between the control devices and the shock system, the complex flow structure behind the devices and the low-frequent motions of the separated region. INTRODUCTION Interactions of shock waves with turbulent boundary layers (SWBLI) can be encountered in virtually every high speed application, including engine intakes, turbomachinery or rocket engines. It is well known that shock induced flow separation is followed by severe energy losses and flow distortion degrading overall system performance (Babinsky et al., 2009; Lin, 2002). Additionally, substantial thermal and pressure loads result from the interaction. In conjunction with their highly unsteady nature, SWBLI are of major concern also for the structural integrity and life time of high speed vehicles. In order to address these issues, control devices can be deployed (Délery, 1985). Here, we focus on passive flow control using vortex generators (VGs), which rank among the most promising approaches to boundary layer control (Lin, 2002). Placed upstream of the interaction, these devices induce a pair of counter-rotating, longitudinal vortices that energize the boundary layer flow within their wake. A valuable modification of conventional-type VGs are micro vortex generators (μVGs) possessing a device height of hV G δ0, with δ0 being the 99%-boundary layer thickness. Due to their smaller height, exposure to lower fluid velocities, and reduced surface as well as cross-sectional areas, μVGs result in substantially lower parasitic losses and flow distortion while still efficiently increasing resistance against flow separation (Lin, 2002). However, relative placement to the separated region is crucial as dissipation and de-correlation of the vortex pair is of concern. The considered flat plate turbulent boundary layer (TBL) is characterized by a free-stream Mach number Ma∞=2.31 and Reynolds number Reδ =67.4·103, based on hV G