Characterization of synthetic jet actuation with application to Ahmed body wake

Abstract

The active flow control by synthetic jet is applied on a road vehicle wake flow. Experiments have been conducted in a wind tunnel. Experimental model is an Ahmed body scaled at 0.7 of the original Ahmed model mounted on aerodynamic balance. Static pressure measurements, wall visualization and PIV as well as hot wire techniques have been used. Synthetic jet actuator has been developed by using electromechanical analogy with the help of the Lumped Element Modeling. It is based on piezoelectric membrane. Its aerodynamic performances have been characterized experimentally. The dynamical response to the membrane power signal (frequency and voltage) is compared to the reduced model (LEM) of the synthetic jet used to scale the actuator. Spatial and temporal evolution of the jet is compared to the existing results and their operating regimes are validated to be used for the control.

The flow downstream from the Ahmed body without control is described. The topology of the longitudinal vortices of the wake and their evolution with the Reynolds number is examined. Spectral analysis is also performed.

For different Reynolds numbers, the aerodynamic efficiency of the drag control is analyzed varying synthetic jet parameters: momentum coefficient, reduced jet frequency and jet position. The study with respect to C_μ variation allows for characterizing the mean topology of the controlled flow. The spectral analysis leads to identify the developed instabilities.

With a rear window tilted at 25°, drag reductions of 8.5% (Re = 1.2 × 10⁶) and of 6.5% (Re = 1.2 × 10⁶) are reached. The control allows to reattach dynamically the rear window separation and to balance the torus vortex structure at the base.

Introduction

New international standards aimed at limiting greenhouse gas emissions are prompting the automobile industry to develop innovative and efficient systems able to reduce fuel consumption on future vehicles. The development of optimal aerodynamic control appears as one of the most promising ways for achieving this objective. Most road vehicles are essentially bluff bodies [1] meaning that their aerodynamic drag is dominated by the pressure drag due to the flow separation at the rear end of the body. Furthermore, it is well established that the flow field in the wake of the vehicle body is highly three-dimensional and unsteady. Instead of using real vehicles, very simplified (generic) vehicle models, reproducing the main features of real vehicles, are used in laboratories. Using this approach, one expects to identify and isolate the more relevant physical parameters of the flow and the impact of the actuation. Ahmed et al. [2], [3], [4], [5], [6] introduced a generic car model whose wake topology depends strongly on the rear slant angle.

Flow separation control is of major interest in fundamental fluid dynamics as well as in various engineering applications [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. Numerous techniques have been explored to control the flow separation either by preventing it or by reducing its effects. These methods range from the use of passive devices to the use of active control devices either steady or unsteady (synthetic jets, acoustic excitation) [19], [20], [21], [22], [23], [24], [25]. Among the various strategies employed in aerodynamic control, conventional passive control techniques, consisting in modifying the shape of the vehicle to reduce the aerodynamic drag, appears as the easiest to implement. Unfortunately, this simplicity is also the main drawback of such devices which are often irrelevant when the flow configuration changes. Indeed, the modification of the shape that produces better aerodynamic properties requires a thorough understanding of turbulent flows around vehicles. Current research efforts are now focusing on active flow control techniques as an alternative to conventional design-modification solutions.

In the case of bluff bodies various actuator technologies have been developed. These actuators can be plasma or fluidic devices. Plasma actuator has been used [26] to control Ahmed body rear window separation. Limited but significant drag reduction has been reached (4%). Continuous [25], pulsed [27] or synthetic [28] jets have been used. Joseph [29] and Joseph et al. [27] used two different actuators (with valves or MEMS) to produce pulsed jets. Drag reduction obtained with valves are higher than with MEMS. The energy balance is however much better with MEMS actuators. Pastoor et al. [28] used ‘zero-net-mass flux actuators’ for a ‘D-shaped cylindrical body’ to control a wake with a closed loop. Because synthetic jets actuator presents an interesting and efficient device for the flow control, it has been studied and characterized both numerically [30], [31] and experimentally [32], [33], [34], [35].

In this paper, the flow around simplified car geometry is controlled by means of a synthetic jet. In areas where the shape of the vehicle naturally generates flow separation, the fluid (air) surrounding the vehicle is periodically sucked in and then blown to form a jet and locally generate vortices in an autonomous manner (no additional air supply). At the actuator orifice exit, the average flowrate during a given cycle is zero, while non-zero net momentum is generated during the blowing and suction phases [9]. The efficiency of this control mechanism has already been demonstrated on continuously curved geometries such as cylinders [16], [17] or wing profiles [12], [36], [37].

The present work addresses the efficiency of a synthetic jet to control the flow on a simple geometry featuring the edges, breaks, low-radius connection curves and the separation regions occurring at the roof, quarter panel, rear window and at the base of an automobile vehicle.

The paper is organized as follows. The experimental setup, the geometry and the actuator are depicted in Section 2. The experimental results are presented and discussed in Section 3. In particular, the aerodynamic drag coefficient (C_d) and the topology of the mean flow in the wake are analyzed as functions of two characteristic parameters of the synthetic jet, namely its momentum coefficient (C_μ) and reduced frequency (F⁺). The experimental results obtained without control are compared to existing experimental and numerical data. The momentum coefficient and reduced frequency are defined as:

$C_{\mu }={\rho }_j{U}_{jm}^2{\Sigma }_j/{\rho }_{\infty }{U}_{\infty }^2{S}_{\infty }$ where U_jm is the amplitude of synthetic jet, Σ_j the slot synthetic jet area, S_∞ the vertical maximum surface of the body and $F_w^{+}=f_j/f_w$ f_j is the jet frequency and f_w is the natural flow frequency.