PLASMA CONTROL OF VORTEX FLOW ON DELTA-WING AT HIGH ANGLES OF ATTACK

International Conference on Methods of Aerophysical Research, ICMAR 28 PLASMA CONTROL OF VORTEX FLOW ON DELTA-WING AT HIGH ANGLES OF ATTACK A.D. Budovsky, B.Yu. Zanin, I.D. Zverkov, V.V. Kozlov, A.A. Maslov, B.V. Postnikov, A.A. Sidorenko Khristianovich Institute of Theoretical and Applied Mechanics SB RAS, 639, Novosibirsk, Russia 1. Introduction The main feature of the flow on leeward side of a delta-wing flow is formation of two primary large scale vortices. Each vortex is able to generate the low pressure region on the surface and create additional lift. The contribution of vortex lift in total lift is increased with increasing of angle of attack and sweep angle of the wing. It can reach up to half of total lift at high angles of attack [1]. Therefore, the ability to manipulate and ultimately control of vortex flow over delta-wing is of great practical importance as these vortices mainly determine aerodynamics of such wings. Control strategy is mainly focused on change of vortex breakdown position because this is a reason that leads the vortex to lose the ability to create an additional lift. The vortex breakdown phenomenon is caused by two main factors. The first one is unfavorable pressure gradient that is not so harmful. It always presents at trailing edge of the wing and becomes gradually stronger with increasing of the angle of attack. Therefore influence of this factor on vortex breakdown is predictable enough. The second one is sensitivity of vortex flow to freestream disturbances [2]. The later is more dangerous as it can lead to sudden lift decreasing and moreover to appearance of roll moment. The paper deals with new application of dielectric barrier discharge (DBD) to vortex flow control on delta-wings. It is well known that DBD is able to introduce desired periodic disturbances due to periodic flow acceleration in the boundary layer region [3] as well as acoustic one [4]. These features make it possible to use DBD for vortex breakdown control. 2. Experimental Setup The experiments were conducted at ITAM in low-turbulence subsonic wind tunnel T-324 in velocity range 3 33 m/s. The wind tunnel has square test section 1m 1m 4m. The turbulence intensity Tu =.4% is small enough that facilitates extrapolation of wind-tunnel data to flight conditions and allows to investigate the influence of DBD s disturbances on vortex breakdown phenomenon. The data available in the literature concerning vortex breakdown on delta-wings with sharp leading edge under subsonic flow parameters showed that vortex bursting occur at certain critical angle of attack that depends on aspect ratio of the wing and doesn t depend on Reynolds number. Considering the test section geometry and blockage ratio the experimental model has been prepared for wind tunnel experiments. It is delta-wing with leading edge sweep angle 65, chord length c =.3 m, aspect ratio AR = 1.865 and thickness t =.3 m (Fig. 1). The leading edges of delta-wing are bevelled to 3. This model was designed in view of strain-gauge balance measurements. The model has been made from dielectric material plexiglass. The delta-wing was placed in the test section and secured on strain-gauge balance that Fig. 1. Sketch of the delta-wing model. A.D. Budovsky, B.Yu. Zanin, I.D. Zverkov, V.V. Kozlov, A.A. Maslov, B.V. Postnikov, A.A. Sidorenko

Section II was fixed on the sting of α-variation mechanism so the angle of attack can be changed in the range 45. Design of DBD actuators is similar to one used in most studies devoted to DBD flow control [5, 6]. Electrodes for DBD were made from 5 μm thick adhesive aluminium film. Encapsulated electrode has approximately 7 mm in width; the other was glued 1 mm overlapped and has 7 mm in width too. The barrier was made from three layers of PVC adhesive film of overall thickness 24 μm (Fig. 2). DBD actuators can be placed in various places on the model surface. Exposed electrode W1 Plasma region W3 Dielectric barrier Model body Encapsulated electrode Fig. 2. Scheme of the plasma actuator. DBD power supply (HVG) used in the experiments was the same as used in [7]. HVG is optimized for effective operation (about 5-7% internal loses) in frequency range 5 Hz 5 khz but can be operated in range from 2 Hz to 1 khz with some minor modifications. Powered by HVG, DBD discharges up to 1 m long, 3 W at the gap with 1 min operation time can be initiated. Surface pressure distribution on leeward side of the model was measured using surface pressure probe that was made from medical needle with diameter.8 mm (Fig. 3.). Pressure tap hole of the probe was perpendicular to wing surface and was situated in the near-wall region to measure near-wall pressure. The probe was connected with differential probe pressure sensor Omega PX265-1D5V. The probe was positioned by means of three-component traversing gear. Six-component strain-gauge balance made in ITAM was used to measure aerodynamic loads. As mentioned above the balance was built into the model. Electrical signal from strain-gauge balance was measured and collected using data acquisition system ECKELMANN that consists of 26 independent measuring channels. Each channel is 14 bit 8 khz A/D converter with built-in amplifier and apparatus filter. The measuring of aerodynamic loads by the balance is realized in the coordinate system associated with the balance O 1 X 1 Y 1 Z 1 AOA б flow X 1 M 1 X Xa Ya CM Ц. Т. Fig. 4. Coordinate system for strain-gauge measurements M 1 Y Za. Y 1 O 1 Fig. 3. Surface pressure probe and then the loads measured are converted to coordinate system associated with flow direction OX а Y а Z а (Fig. 4). To get an idea about mean flow and observe vortex evolution with increasing of angle of attack oil-flow visualization as well as smoke flow M 1 Z Z 1 Z visualization was used in the experiments. TiO 1 2 and kerosene suspension was used to realize oilflow visualization method. Smoke flow X 1 visualization was realized by means of laser sheet method presented in Fig. 5. Y 1 2

International Conference on Methods of Aerophysical Research, ICMAR 28 camera smoke generator laser Fig. 5. Smoke flow visualization scheme by means of laser sheet 3. Experimental Results Study of DBD vortex breakdown control on the model of delta-wing was carried out in wide range of chord Reynolds number.6.7 1 6. Series of preliminary tests were performed without DBD actuators. Oil-flow visualization and smoke flow visualization by means of laser sheet were used to get an idea about mean flow over delta-wing and its evolution with increasing of the angle of attack. Oil-flow visualization gives an opportunity to find out such features of the flow as attachment and separation lines. It can be seen from Fig. 6 that on leeward side of the delta-wing at angle of attack 15 tree-vortex separated flow configuration is realized. Analysis of surface streamlines allows us to see primary vortex attachment line (A 1 ) as well as secondary and tertiary vortex separation lines (S 2, S 3 correspondingly) (Fig. 6). Smoke flow visualization assists to observe vortex space location and get an idea on vortex pattern. Comparison of data obtained by means of oil-flow and smoke flow visualization contributes to more detailed understanding of flow structure over the delta-wing as a whole. As shown in Fig. 5 laser sheet illuminates the flow in the plane that was perpendicular to model surface. To raise an amount of data obtained scanning laser sheet method was applied. In other words laser sheet was gradually moved from wing tip to wing (α=15, U =12m/s) trailing edge that allowed us to get information on vortex pattern in various planes along wing chord. Samples of oil/smoke flow visualization prints of the flow over the delta-wing are presented in Fig. 7 for three different angles of attack those correspond to three different flow regimes. The data obtained show that position of vortex breakdown point gradually moves upstream to the wing tip with increasing of the angle of attack. When α is less then 15 vortex breakdown occur somewhere downstream from model trailing edge. When angle of attack reaches 15 right vortex breakdown point overpasses wing trailing edge while left vortex demonstrates absence of breakdown. This tendency was observed in all the tests and connected probably with inaccuracy in model positioning. It also can be noted from Fig. 7a that when angle of attack less or equal 15 secondary vortex separation lines (S 2 see Fig. 6) are slightly curved to centerline of the wing. А 1 Fig. 6. Oil-flow visualization print S 2 S 3 3

Section II secondary vortex primary vortex vortex core without breakdown S 2 vortex breakdown a) U = 12 m/s, α = 15 b) U = 12 m/s, α = 17 S 2 kink c) U = 12 m/s, α = 22 Fig. 7. Oil/smoke flow visualization (view from leading edge) Further increasing of the angle of attack lead the vortex breakdown point to move further upstream to wing tip (Fig. 7 a, b). It is interesting to note that lines S 2 begin to curve in leading edge direction. The kink originates at α=17 (Fig. 7b) and moves together with vortex breakdown point further upstream to wing tip with increasing of α too. In Fig. 8 vortex breakdown point position obtained from smoke flow visualization data (U = 14 m/s) and S 2 kink position obtained from oil-flow visualization (U = 12 m/s) along wing chord are presented. As it is seen from Fig. 8 there is certain correlation between vortex breakdown point position and secondary vortex separation line kink position S 2 on the graph. It is known that vortex bursting lead to the abrupt flow stagnation in the vortex core region and its expansion. Thus when vortex bursts the secondary vortex is forced out to wing leading edge and line S 2 is curved in same direction too. At angles of attack less then 21 S 2 kink position is observed downstream from vortex core breakdown point. It is explained by fact that the vortex breakdown phenomenon is non-stationary process and therefore vortex breakdown position is changed. Moreover after disappearing of dark vortex core the process of flow stagnation and vortex expansion need certain period of time and therefore certain distance before achieving of steady burst state. It is also can be noted from Fig. 8 that moving of vortex breakdown position upstream to wing tip with increasing of the angle of attack have approximately linear character. Primary vortex core positions obtained form smoke position of breakdown, x/c 1.2 1.8.6.4 Right vortex Left vortex S 2 break.2 8 12 16 2 24 28 angle of attack, degree Fig. 8. Vortex breakdown and S 2 kink positions vs. angle of attack 4

International Conference on Methods of Aerophysical Research, ICMAR 28 flow visualization for α = 12, 15, 17, 2 and U = 12 m/s are presented in Fig. 9. A12 A15 4 2 Y, mm 2 Y, mm 1-15 -1-5 Z, mm 5 1 15 1 X, mm 2 3-15 -1-5 Z, mm 5 1 15 5 1 25 2 15 X, mm 3 A17 1 A2 4 2 15 Y, mm 2-15 -1-5 Z, mm 5 1 15 1 2 X, mm 3 Y, mm 1 5-2 -1 Z, mm 1 2 1 2 X, mm 3 Fig. 9. Primary vortex core positions Red curves are primary vortex core space locations while blue one are their projections to model leeward surface (Fig. 9). The data are presented for cross-sections and angles of attack when dark vortex cores are still observed. It is clear to see from Fig. 9 A12, A15 that at α = 12 and 15 vortex flow isn t conical. With increasing of x coordinate counted from wing tip along wing chord vortex cores are rose above wing upper side and simultaneously they are shifted closer to the wing leading edge as compared with the case of conical vortex flow. This observation corresponds to oilflow visualization data that demonstrated secondary vortex separation line (S 2 ) kink too. At α = 17 the vortex flow approach to conical state as vortex core lines and their projections become almost straight (Fig. 9 A17). The same is observed for α = 2 before vortex core breakdown point (Fig. 9 A2). Strain-gauge balance measurements of aerodynamic loads on delta-wing were performed in the range of α = 1 4 at freestream velocities U = 12, 21.6, 33.2 m/s. The data obtained are shown in Fig. 1 as C xa (α), C ya (α), C za (α), m xa (α), m za (α) and K, where C xa drag, C ya lift, C za side force, m xa roll moment and m za pitching moment coefficients, K lift-drag ratio, α angle of attack. All of the coefficients are presented in coordinate system associated with flow direction. As it can be seen from C ya (α) graph (Fig. 1) that with increasing of α lift augmentation is observed up to α = 3. The data obtained for velocity range investigated agree well together. At the same time maximum lift is achieved at U = 33.2 m/s. Decreasing of freestream velocity to U = 12 m/s lead the lift value to decrease for all angles of attack. It is also interesting to note that at α = 2 there is a kink in lift curves obtained for all the velocities tested. The kink is connected with the fact that vortex breakdown point overpasses wing trailing edge and affects the aerodynamic characteristics of the wing. 5

Section II 1.9 1.9.8.8.7 Cya.7 Cxa.6.5.6.4.5.4 1 15 2 25 3 35 4.3.2.1 1 15 2 25 3 35 4 3.5 3.2.1 2.5 -.1 K 2 Cza -.2 1.5 -.3 1 -.4.5 1 15 2 25 3 35 4 -.5 1 15 2 25 3 35 4.22.2.5.4.18.3.16.2 mza.14 mxa.1.12.1 -.1.8 1 15 2 25 3 35 4 Fig. 1. Integral characteristics of delta-wing -.2 1 15 2 25 3 35 4 The plot C xa (α) (Fig. 1) demonstrates increasing of drag with increasing of α and drag have its maximum value at α = 35. The curve C xa (α) for U = 12 m/s has higher values at almost all the angles of attack then at higher velocities tested that in turn affect the lift-drag rate K. As it is seen side force and roll moment values are negligible. Data dispersion is increased with decreasing of freestream velocity that can be explained by increasing of measurement relative accuracy with decreasing of measured quantity. It can be seen from pitching moment plot m za (α) (Fig. 1) that center of pressure is shifted to wing tip and increased with growing of the angle of attack. Maximum m za value is achieved at α = 35 that corresponds to global separation on leeward side of the delta-wing. When global separation (separation bubble) is formed the suction peak on leeward side is fully decreased and pressure on the side become the same as on windward side that lead the windward side to play a major role in lift production. This fact results in pressure center shifting and decreasing of pitching moment m za. Further the line turbulators were used on the upper surface of the wing to find out their most 6

International Conference on Methods of Aerophysical Research, ICMAR 28 effective positions for flow control from the point of view of maximum shift of vortex breakdown point. The turbulators were made from synthetic thread with square cross-section 1 1 mm. The various turbulator configurations were tested. It was found that most effective were turbulators placed perpendicularly to the wing leading edge across of vortex flow. The results obtained using the configuration are presented in Fig. 11. vortex breakdown turbulators vortex breakdown a) without turbulators (X break /c =.73) b) line turbulators on the right (X break /c =.58) Fig. 11. Smoke flow visualization (U = 12 m/s, α = 17 ) As it is seen from Fig. 11a that in the case when turbulators are unused vortex breakdown on the right part of the delta-wing occur at 73% of wing chord c from wing tip. Installation of two line turbulators on the right part of the wing results in shift of vortex breakdown point toward wing tip on approximately 15% (Fig. 11b). Surface pressure distribution measurements using surface pressure probe (see Fig. 3) on wing leeward side were performed for the cases presented. The data obtained are shown in Fig. 12 as pressure coefficient -C P. secondary pressure peak primary pressure peak -C P primary pressure peak secondary pressure peak -C P a) without turbulators b) line turbulators on the right Fig. 12. Pressure coefficient distribution (U = 12 m/s, α = 17 ) Fig. 12a shows that when turbulators are unused the vortex flow is symmetrical that is expressed in equal width of pressure peaks generated by primary vortices. Moreover two pressure peak with less intensity that are generated by secondary vortices are well observed on both left and right parts of the wing. Installation of turbulators as described above (see Fig. 11b) leads the right primary pressure peak to become more width due to right primary vortex breakdown (Fig. 12b). Only one secondary pressure peak is observed on the left part of the delta-wing in this case (Fig. 12b). 7

Section II Absence of secondary pressure peak on the right have the same explanation as was given earlier in the case of oil-flow visualization. When primary vortex is broken down its core diameter is increased thus forcing out the secondary vortex closer to wing leading edge. As DBD electrodes have simple construction the experiments on vortex flow control by means of DBD surface plasma discharge was carried out using various DBD-activator configurations. The first configuration used is presented in Fig. 13. In this case DBD-activators were placed along wing leading edge on leeward side of the wing and upper electrode was 7 mm away from wing leading edge. upper electrodes plasma on vortex cores without breakdown Fig. 13. Smoke flow visualization (DBD-activator on leading edge, U = 12 m/s, α = 17 ) Here it was supposed that DBD-activator will be able to influence on shear layer shedding from the wing leading edge due to introduction of mass flow as well as probably acoustic disturbances. But data obtained showed that such configuration was ineffective. As modification of the configuration the second DBD-activator was installed on windward side of the model in the same manner. But it wasn t crowned with success too. In this case DBD-activator was situated in the secondary vortex region i.e. in the separation zone as analysis of oil-flow visualization data showed and disturbances generated by DBD were not sufficient for desired flow disturbing. DBD-activators installed in the same manner as turbulators proved to be most effective for flow control. In this case DBD-electrodes are perpendicular to wing leading edge. The distances between upper electrodes were 3 mm. DBD-activator generates flow acceleration toward wing tip. The experiments using the configuration were performed in the freestream velocity range U = 1 3 m/s at α = 1 25. Fig. 14 shows an example of vortex flow pattern in the laser sheet section using smoke flow visualization with and without DBD for the configuration. DBD actuators plasma on vortex cores with no breakdown breakdown а) DBD off b) DBD on f = 1 khz Fig. 14. Smoke flow visualization (view from trailing edge), U = 12 m/s, α = 15 It is seen from Fig. 14a, that there are two vortices with well defined cores on the left as well as right part of the wing in the case DBD off. If DBD was activated on the right part of the wing (Fig. 14b) only left vortex had the core while right one is destroyed. 8

International Conference on Methods of Aerophysical Research, ICMAR 28 4. Conclusions Research of vortex flow control on the model of delta-wing was carried out at subsonic speed with chord Reynolds number.6.7 1 6. Dielectric barrier discharge plasma on the surface of the model was used for vortex breakdown control. Effects of the angle of attack of the model, discharge frequency, location and geometry of discharge actuators were studied in experiments. It was found that DBD is able to influence on vortex breakdown position and can be applied for vortex flow control on delta-wing at high angles of attack. The experimental results showed that variations of DBD location, power and frequency have significant effect on the vortex breakdown control efficiency. The most effective was DBD placed perpendicularly to the leading edge across of vortex flow. REFERENCES 1. William H. Wentz and David L. Kohlman., Wind tunnel investigations of vortex breakdown on slender sharpedged wings, Technical Report FRL 68-13, University of Kansas Center for Research, Inc., 1968. 2. Mabey D.G., Similitude relations for buffet and wing rock on delta wings, Progress in Aerospace Sciences, Volume 33, Number 7, 1997, pp. 481-511(31) 3. J. Reece, Roth and Xin, Dai, Optimization of the Aerodynamic Plasma Actuator as an Electrohydrodynamic (EHD) Electrical Device, AIAA Paper No. 26-123, 26. 4. Corrie Baird, C. L. Enloe, Thomas E. McLaughlin and James W. Baughn, Acoustic testing of the dielectric barrier discharge (DBD) plasma actuator, AIAA Paper No. 25-565, 25. 5. Post, M., and Corke, T., Separation Control on High Angle of Attack Airfoil Using Plasma Actuators, AIAA Paper No. 23-124, 23. 6. Post, M., and Corke, T., Separation Control Using Plasma Actuators - Stationary and Oscillating Airfoils, AIAA Paper No. 24-841, 24. 7. Anatoly Maslov, Andrey A. Sidorenko, Boris Yu. Zanin, Boris V. Postnikov, Alexey D. Budovsky and Norman D. Malmuth, Plasma Control of Flow Separation on Swept Wing at High Angles of Attack, AIAA Paper No. 28-54, 28. 9