The subject of flow control is a diverse one, ranging from active to passive control techniques, with a multitude of technologies and levels of complexity. Perhaps the most widespread and simplest of these are passive vane-type vortex generators; small winglets protruding normal to the surface generating high intensity streamwise vortices which promote entrainment of higher streamwise-momentum flow. After the early demonstration of their usability by Taylor in 1947, 1 these types of VGs have received considerable attention throughout the past decades, spurred on in recent years by the advent of more advanced nonintrusive flow measurement techniques. For extended reviews, the interested reader is referred to Lin 2 and Ashill et al. 3 Traditionally, research focussed on device optimisation for particular flow regimes by studying the sensitivity of backflow and skin friction modulation, studying influence of boundary layer state and pressure gradients on the growth and effectiveness of this passive flow control technique. This approach, however, left gaps in the knowledge of the underlying flow physics and the manner in which VG-induced vortices interacted with the encompassing boundary layer. In the late 1980s and early 90s, works by Westphal also delved into the dynamics and structure of the vortices, with a view towards vortex wandering and its effect on the mean flow properties. 4 Vortex meandering was apparently also responsible for observed high shear stresses in the vortex core region. Angele and Muhammad-Klingmann, 5 as well as Cathalifaud et al. 6 again investigated vortex dynamics using particle image velocimetry (PIV), and found that spanwise wandering was larger than wall-normal fluctuations. Vortex stretching was also hypothesized, further reaffirming the non-stationary nature of the vortices. Recent works by Lögdberg et al. 7, 8 investigated the evolution of passively generated streamwise vortices up to a relatively long downstream distance of 450 device heights (or 95δ) and found that control was quite robust to variations in VG placement, relative to mean separation line.

Previous research has focussed on practical vane shapes, in configurations which are of course representative of general applications. However, test cases and inflow boundary conditions (whether from an experimental, or computational point of view) are generally considered static and ideal; that is, the influence of fluctuating or non-ideal inflow conditions on the flow induced by vortex generators has received little attention. It is the subject of this paper to address the latter situation, and describe the flow induced behind VGs, subject to skewed inflow conditions.