N. Takeuchi and K. Yasuoka: Effect of Extension Length in a Surface Barrier Discharge 1070-9878/09/$25.00 © 2009 IEEE 364 Effect of Extension Length in a Surface Barrier Discharge on the Driving Force of Electrohydrodynamic Gas Flow Nozomi Takeuchi and Koichi Yasuoka Tokyo Institute of Technology Department of Electrical and Electronic Engineering, Meguro-ku, Tokyo 152-8552, Japan ABSTRACT A surface barrier discharge generated between an exposed and a buried electrode induces a unidirectional gas flow, known as an electrohydrodynamic (EHD) gas flow. The discharge behavior and flow characteristics have been investigated. When the exposed electrode is an anode, several streamers propagate above the buried electrode, whereas when the exposed electrode is a cathode, weak light emission is uniformly observed above the buried electrode. The maximum discharge length increases as both the length of the buried electrode and the amplitude of the applied voltage increase. Independent of the buried electrode length, the velocity of the EHD flow increases monotonically as the discharge extends farther, and is uniquely determined by the maximum discharge length. Index Terms — Electrodynamics, surface discharges, gas flow. 1 INTRODUCTION WITH certain types of gas discharges, electrohydrodynamic (EHD) gas flow is generated by the EHD effect, which is an interaction between an electric field and charged particles. The EHD gas flow generated by a dc corona discharge is well known as an ionic wind or corona wind, and has applications in gas pumps [1], electrostatic precipitators [2], and flow control actuators [3, 4]. In a dc corona discharge, positive or negative ions accelerated by the electric field collide with neutral molecules and transfer their momentum to molecules, resulting in gas flow generation. EHD gas flow can also be generated by a surface barrier discharge between an exposed and a buried electrode [5]. This is applicable in active flow control devices for flow re-attachment and drag reduction, among other applications [6, 7]. Although the applied external electric field constantly changes its direction, the surface barrier discharge generates unidirectional gas flow. The mechanism that governs how the discharge induces unidirectional gas flow has been investigated by many researchers. Optical measurements performed using a photomultiplier tube revealed that the structure of the discharge was substantially different in both space and time, with characteristics that depend on the polarity of the exposed electrode [8]. High-speed framing photography revealed that when the exposed electrode was an anode, many positive streamers appeared; however, when the exposed electrode was a cathode, uniform light emission was observed [9]. The streamers which corresponded to discharge current pulses appeared at different locations almost randomly across the span of the dielectric material, and their distribution, averaged over a couple of ac oscillation periods, was approximately uniform [10]. Modeling of the discharge and generated flow has been carried out by several groups. Such asymmetry in the discharge behavior due to the polarity of the exposed electrode was calculated by Font et al by using particle-in-cell and Monte-Carlo methods for pure nitrogen [11]. They indicated that electrons did not significantly affect the momentum transfer to neutrals. When the exposed electrode was the cathode, the discharge induced a weak EHD force due to the momentum transfer from positive ions to neutral molecules in the direction opposite to the generated flow. On the other hand, when the exposed electrode was the anode, the direction of the force was the same as that of the flow, and the amplitude of the force was 10 times larger than that during opposite polarity because the positive ion density was greater. Thus, the EHD flow was generated in the direction of the net force. However, Forte et al experimentally obtained an opposite result with a laser Doppler velocimeter (LDV) that the flow velocity was higher when the exposed electrode was the cathode [12]. The discrepancy may be due to unconsidered oxygen negative ions in the modeling. Boeuf et al performed numerical simulations in pure nitrogen first. They indicated that the EHD force is dominant in a unipolar region [13]. When the exposed electrode was the anode, the calculated current was composed of large, quasi- periodic pulses separated by a low current period [14]. The low current phase was responsible for most of the EHD force because the pulse was caused by the development of a quasi- neutral plasma channel, and the duration of this phase was much longer than that of the current pulse. In their next paper, preliminary results were obtained in air with a simplified model considering only one type of negative ions [15]. They revealed that the downstream-directed EHD force generated Manuscript received on 30 May 2008, in final form 4 October 2008.