Measurements of the orthogonal blade±vortex interaction using a particle image velocimetry technique R. B. Green, C. J. Doolan, R. M. Cannon Abstract This paper describes the results of application of a particle image velocimetry PIV) technique to an orthogonal blade±vortex interaction experiment. To help resolve the problem of vortex meander during the tests, two PIV systems were used, which produced two velocity vector maps closely separated in time. During the PIV analysis an image-based vector validation scheme was used, which was shown to reduce signi®cantly the number of wild vectors reaching the vector map. Preliminary results from the tests showed that, close to the blade, a signi®cant radial out¯ow was superimposed on the vortex ¯ow ®eld. The radial ¯ow is thought to be due to the dispersion of the vortex axial core ¯ow during vortex cutting, which distorts the vortex ¯ow ®eld and enlarges the vortex. Further away from the blade, no signi®cant radial ¯ow was detected and the vortex remained undis- turbed. 1 Introduction Vortices which form at the tips of helicopter main rotor blades persist for a considerable time after formation. This allows them to interact with other rotor blades or with the airframeitself,producingunwantednoiseandvibration.A particular case of rotorcraft vortex interaction is the or- thogonal blade±vortex interaction. Here, the vortex ap- proaches a downstream tail rotor blade orthogonally and eventually collides with and is severed into two parts by the blade. Recent interest in this vortex interaction has included computational and experimental studies Lee et al. 1998; Marshall and Krishnamoorthy 1997), which have shown that the vortex dynamics were controlled by the impingement of the core axial ¯ow on the blade sur- face. Water ¯ow measurements using PLIF Marshall and Krishnamoorthy 1997) shows vortex core bulging and thinning due to the blockage of the core ¯ow. Surface pressure measurements Doolan et al. 1999a) indicate an increase in pressure where the core ¯ow is directed toward the blade and a decrease when this ¯ow is away from the blade. Flow ®eld measurements during the orthogonal blade± vortex interaction are dif®cult to obtain using conven- tional probe-based techniques due to the fundamental problems of probe interference and vortex wandering. Particle image velocimetry PIV) can provide a global, non-intrusive measurement system, which overcomes the problems of probe-based techniques. The application of PIV is not without its problems, however. Vortex wander or meander can make dif®cult the analysis of the velocity vector maps, since the spatial position of the vortex before the interaction affects the interaction itself Horner et al. 1996). In addition, seeding helicopter rotor-type vortices in wind tunnels has been found to be notoriously dif®cult Horner et al. 1996), and it is often found that the vortex cores are relatively devoid of seeding particles. Velocity information in the vortex core is therefore usually of poor quality. This paper describes the application of PIV to an orthogonal blade±vortex interaction BVI) experiment, in which the above dif®culties are addressed. A vector vali- dation scheme is presented, which can ef®ciently resolve wild vectors even where there is a high density of wild vectors, as may be found in the vortex core. The vortex meander is partially solved using a dual PIV system. The PIV vector maps clearly show a strong radial out¯ow during the orthogonal BVI. 2 Experimental method The experiment consists of two main components, namely the rig for the orthogonal BVI and the PIV system. It is convenient to describe each of these separately. 2.1 Vortex generator and interacting blade The experiments were conducted in the Glasgow Univer- sity 1.15 m ´ 0.85 m closed return, low-speed wind- tunnel. A unique vortex generator, described by Copland 1998) and Doolan et al. 1999b), was placed in the con- traction of the wind-tunnel for the production of convec- ting 3-D vortices in the working section. The vortex generator consists of a rotor rig with a single blade of 0.75 m radius and 0.1 m chord with a NACA 0015 Experiments in Fluids 29 2000) 369±379 Ó Springer-Verlag 2000 Received: 26 April 1999/Accepted: 9 November 1999 R. B. Green &), C. J. Doolan, R. M. Cannon Department of Aerospace Engineering, University of Glasgow, Glasgow, G12 8QQ, UK Financial assistance for setting up the PIV system at Glasgow was obtained from The Dunlop-Chapman bequest of the University of Glasgow, The Royal Society and Spectra-Physics U.K.) Ltd. The experimental work was performed with the assistance of the Engineering and Physical Sciences Research Council. The authors would like to take this opportunity to acknowledge the encouragement and support provided by Prof. R. Galbraith and Dr. F. Coton and the technicians of the Department of Aerospace Engineering, University of Glasgow. 369