Application of passive flow control techniques to attenuate the unsteady near wake of airborne turrets in subsonic flow

Abstract

External airborne installations of cylindrically-shaped objects, such as turrets, can lead to severe degradation of flight performance. This engineering challenge is mainly due to an unsteady and separated vortex flow field that evolves in the wake downstream of the turret. This study focuses on providing an engineering solution to attenuate the wake unsteadiness. Experiments were performed in the Israeli Air Force (IAF) low speed wind tunnel, where we investigated various passive flow control methods on a hemispherical dome attached to a cylindrical body turret with a low aspect ratio (AR) of 1.36 at a diameter-based Reynolds number of $R{e}_D=3.75\times {10}^5$. Three modifications of the baseline turret geometry were examined. The configurations include the addition of a grit roughness strip, a turbulator zigzag trip tape, and a straight edge protuberance (or a bump strip). The velocity fields in the near wake downstream of the various turret configurations were acquired using a particle image velocimetry (PIV) technique. Results show that tripping the flow using a bump strip is the most effective for the turret studied herein. The bump strip prompts symmetrical vortex shedding downstream, which can potentially eliminate the fluctuating side force on the turret, while also contributing the least to any drag increase (or penalty). Implementing such a passive flow technique on cylindrical installations may assist in resolving large fluctuating side forces and meeting operational requirements during flight.

Introduction

External turrets are often installed on military aircraft and serve many purposes, among which are communication, photography, remote sensing, etc. [1]. Such turrets are essentially finite-length bluff bodies under cross-flow conditions, typically comprising a wall-mounted circular cylinder (the base), of height h and diameter D, and a matching hemispherical dome. Studies of airborne turrets have gained much popularity in recent years, in particular for use with unmanned aerial vehicles (UAVs), where both experiments and numerical simulations are employed to gain knowledge on the complex flow field characteristics and the aero-optical effects (e.g., aberrations, distortions, etc.) experienced by such configurations [2], [3], [4], [5], [6], [7].

The flow field around airborne turrets in subsonic conditions is highly three-dimensional, unsteady and dominated by separated vortical structures [4]. The flow is attached on most of the turret front part, where the adverse pressure gradient on the turret forces the flow to separate at some circumferential angle ϕ (from the leading edge of the turret), resulting in a three-dimensional and unsteady flow field. Snyder et al. [1] investigated experimentally turrets with an aspect ratio of $\mathrm{AR}=h/D=1$ and reported that flow separation takes place at the rear of the turret, at about $\varphi ={125}^{\circ }$ from the leading edge for Reynolds numbers between $3\times {10}^5$ to $9\times {10}^5$. Three main flow features characterize the flow field aft of turrets with an $\mathrm{AR}\ge 1$ [4], [8]: i) two counter-rotating trailing (streamwise) vortices separating from the back of the upper hemispherical dome, resulting from the downwash caused by the lower base pressure; ii) a periodic alternate spanwise vortex shedding from the turret sides, which is highly dependent on the turret aspect ratio; and iii) two secondary vortices on both sides of the turret base. All these vortical structures interact with the base horseshoe vortex to create two large streamwise trailing vortices further downstream.

To gain fundamental insights into the flow characteristics generated around turrets with an $\mathrm{AR}\ge 1$, a simplified model of a finite-length wall-mounted circular cylinder can be utilized. As reported in previous studies [4], [7], [9], [10], [11], [12], [13], the flow field around such simplified models, as well as turrets, is dependent on many parameters, among which are the cylinder AR, Reynolds number, thickness of the incoming boundary layer, surface roughness and free stream turbulence. Of these, the two most important parameters that determine the unsteady flow field characteristics downstream are the cylinder AR and its free-end conditions [12], [14], [15]. For a finite-length wall-mounted circular cylinder in a cross-flow, periodic alternate vortex shedding occurs from both sides of the cylinder, forming a Kármán vortex street. The near wake flow characteristics mainly depend on the cylinder AR [16]. When the AR falls below a critical value that ranges from 1 to 7 [14], vortex shedding changes from the anti-symmetrical Kármán type to the symmetric ‘arch’ type. For sufficiently low AR cylindrical geometries, anti-symmetrical Kármán shedding is frequently observed [17]. This highly unsteady environment contributes not only to drag enhancement but also to the development of a highly unsteady side force, which can be larger than the drag force [2]. Such high amplitude unsteady side loads, when developed on the circular cylinder base of airborne turrets [7], may lead to performance degradation of small aircraft such as UAVs.

The complex flow characteristics around wall-mounted circular cylinders and turrets have also been studied numerically. Studies that utilized the Reynolds Averaged Navier-Stokes (RANS), unsteady RANS (URANS) and Partially Averaged Navier-Stokes (PANS) modeling approaches have reported only limited agreement between the numerical results and available experimental data [7], [18], [19], [20], [21], [22]. Due to the unsteady nature of the problem, many studies have utilized higher-fidelity Large Eddy Simulation (LES) and hybrid RANS/LES methodologies, reporting an improved accuracy of the numerical results [3], [10], [23], [24], [25], [26], [27]. These studies demonstrated the importance of using high-fidelity models to accurately resolve the flow field around wall-mounted circular cylinders and turrets. Moreover, they underlined the importance of combined numerical and experimental investigations for a given problem of this kind. It is noteworthy, however, that at realistic Reynolds numbers relevant for aircraft in flight (on the order of millions), a properly resolved LES for a full three-dimensional problem is impractical, due to the large mesh requirements and computational resources. For this reason, LESs have only been used to model low Reynolds number flow problems [10], [28], [29]. For numerical solutions of realistic wall-mounted circular cylinders and turrets, hybrid RANS/LES methods are preferred [3], [23], [24], [25], [26], [30].

As most airborne turrets are sensitive to the external flow environment, aerodynamic shielding with fairings is frequently preferred, as previously stated in the literature [1], [19], [31], [32]. These solutions, however, may not be practical for most UAVs in which weight balance is crucial for vehicle performance. Several studies have reported that active flow control devices, such as suction actuators and synthetic jets, can be used for delaying flow separation on turrets [3], [33]. Similar improvement of the flow field around turrets was also achieved by using passive flow control devices, such as vortex generators, porous surfaces and forward partition plates [3], [21], [34]. Other devices to passively control the flow around the circular cylinder base include geometric modifications, such as helical wires and strakes wound around the circumference, dimple surfaces, spanwise splits, circumferential grooves, spanwise grooves, and roughness strips [35], [36], [37], [38], [39] – all of which promote transition and presumably lock the separation line over the cylinder surface. Passive flow control, if effective, is a logical choice for turrets installed on UAVs: it does not involve any moving parts to complicate the turret assembly, does not require additional energy to control the flow, and tends to be effective over a range of flow regimes, as opposed to other active flow control techniques.

In this study, we investigate experimentally a turret model comprising a hemispherical dome on a cylindrical body ($h/D=1.36$), which is similar to typical airborne turrets mounted on UAVs. The flow conditions in the wind tunnel we used (such as Reynolds number and turbulent intensity) are similar to those that a full-scale turret would experience during flight. Given the relatively high Reynolds number [40], [41], [42], a supercritical flow regime [35] is expected around the turret model, with anti-symmetrical vortex shedding, dominating the near wake downstream of the cylindrical base, due to its low AR [23], [43]. Herein, we study how local surface modifications on the cylinder base can mitigate the turret's wake flow unsteadiness, thus, eliminating the large unsteady side loads exerted on it. The choice of these passive flow control devices and their location on the turret model is based on preliminary hybrid RANS/LES CFD simulations, performed to map the flow around the baseline turret and identify the location where flow separation is initiated (see further details in Appendix A). The passive flow control devices investigated were chosen because of their easy installation on real-world turrets. A particle image velocimetry (PIV) technique was utilized to measure the near wake turbulence downstream of the turret model at the Israeli Air Force's (IAF) wind tunnel facility. The near wake flow dynamics were characterized using the mean and turbulent flow properties as functions of the various flow control devices. In addition, the flow patterns formed at the wake were analyzed using proper orthogonal decomposition (POD). Finally, the ‘drag penalty’ associated with the passive flow control devices was estimated from the momentum deficit and Reynolds normal stresses measured in the near wake region.