https://ntrs.nasa.gov/search.jsp?R=20070004937 2019-04-12T18:49:51+00:00Z High-Lift System for a Supercritical Airfoil: Simplified by Active Flow Control LaTunia Pack Melton∗, Norman W. Schaeffler†, and John C. Lin‡ NASA Langley Research Center, Hampton, Virginia, United States of America ActiveflowcontrolwindtunnelexperimentswereconductedintheNASALangleyLow- Turbulence Pressure Tunnel using a two-dimensional supercritical high-lift airfoil with a 15% chord hinged leading-edge flap and a 25% chord hinged trailing-edge flap. This paper focusesontheapplicationofzero-net-mass-fluxperiodicexcitationneartheairfoiltrailing- edge flap shoulder at a Mach number of 0.1 and chord Reynolds numbers of 1.2×106 to 9×106 with leading- and trailing-edge flap deflections of –25◦ and 30◦, respectively. The purpose of the investigation was to increase the zero-net-mass-flux options for controlling trailing edge flap separation by using a larger model than used on the low Reynolds num- ber version of this model and to investigate the effect of flow control at higher Reynolds numbers. Static and dynamic surface pressures and wake pressures were acquired to de- termine the effects of flow control on airfoil performance. Active flow control was applied both upstream of the trailing edge flap and immediately downstream of the trailing edge flap shoulder and the effects of Reynolds number, excitation frequency and amplitude are presented. The excitations around the trailing edge flap are then combined to control trailing edge flap separation. The combination of two closely spaced actuators around the trailing-edge flap knee was shown to increase the lift produced by an individual actuator. ThephasesensitivitybetweentwocloselyspacedactuatorsseenatlowReynoldsnumberis confirmed at higher Reynolds numbers. The momentum input required to completely con- trol flow separation on the configuration was larger than that available from the actuators used. Nomenclature c = reference airfoil chord C = momentum coefficient, µ (at δ =δ =0◦) ≡ρ hU2/cq LE TE j j C = pressure drag coefficient C = total drag dp D C = lift coefficient C = maximum lift coefficient L L,max C = pressure coefficient, f = excitation frequency, Hz p ≡(P −P )/q s F+ = reduced frequency, h = slot height ≡fx /U sp ∞ P = pressure P = static pressure s q = freestream dynamic Re = chord Reynolds number, c pressure, ≡1/2ρU2 ≡U c/ν ∞ ∞ U = Velocity U = RMS of Jet exit velocity j V = RMS voltage x/c = normalized streamwise loca- RMS tion y/c = normalized cross-stream loca- x = distance from actuator to sp tion trailing edge ∗ResearchScientist,FlowPhysicsandControlBranch,MS170,SeniorMemberAIAA †ResearchScientist,FlowPhysicsandControlBranch,MS170,MemberAIAA ‡ResearchScientist,FlowPhysicsandControlBranch,MS170,AssociateFellowAIAA 1of20 AmericanInstituteofAeronauticsandAstronautics Greek α = angle of attack, degrees δ = TE flap deflection δ = LE flap deflection TE LE ∆ = differential value ν = kinematic viscosity ρ = density Abbreviations 2D = two-dimensional AFC = active flow control AM = amplitude modulation BART = BasicAerodynamicsResearch Tunnel EET = Energy Efficient Transport LTPT = Low Turbulence Pressure Tunnel SHL = simplified high lift ZMF = zero-net-mass-flux LE = leading edge TE = trailing edge VSF = vortex shedding frequency Subscripts m, M = modulation max = maximum value peak = maximum measured value ∞ = freestream conditions I. Introduction ThispaperfocusesontheresultsobtainedforasimplifiedversionoftheNASAEnergyEfficientTransport 12%-thicksupercriticalairfoiloperatinginahigh-liftconfigurationthroughtheuseofactiveflowcontrol. This work extends the Reynolds number range of previous results obtained on a smaller version of the model1,2,3 tested in the NASA Langley Basic Aerodynamics Research Tunnel. The current research effort marks the return of this airfoil section to the NASA Langley Low-Turbulence Pressure Tunnel. The airfoil section was extensively tested in this facility as a three-element high-lift configuration.4 It is the goal of the current research effort to achieve similar performance with a single-element simplified high-lift system augmented with active flow control. Prior to describing the present experiment and presenting the results some background information is provided. Experiments at both low5 and high6,7 Reynolds numbers have demonstrated the ability of periodic excitation to effectively and efficiently control separation. These experiments compared steady momentum transfer methods to periodic excitation and showed the momentum requirements were considerably reduced when using periodic excitation. The efficiency of periodic excitation is attributed to exciting the flow at frequenciesthatmatchthenaturalinstabilitiesoftheseparatingshearlayer. AsystemstudybyMcCleanet al.8 indicated that reducing the complexity of high-lift systems, by the application of periodic excitation for controlling separation, would have significant benefits and would represent a high payoff application of this technology. The approach taken was to simplify a conventional three-element high-lift system by replacing the slat and Fowler flap with a simply hinged leading- and trailing-edge flap. The airfoil chosen as a basis for this was the EET airfoil, because of its existing dataset in both cruise and high-lift configurations.4 To achievesimilarperformancetothethree-elementhighliftconfigurationoftheEET,itisexpectedthatlarger trailing-edge flap deflections would be required on the simplified EET high-lift system and the application of active flow control near the leading- and trailing-edge flap shoulder would enable the simplified system to generate lift comparable to the three-element high-lift configuration. Low Reynolds number results of Melton et al.1,2,3 showed that periodic excitation was effective at controlling leading-edge flap shoulder separation, increasing the maximum lift coefficient, C , by 10- L,max 15% at low (less than 5◦ ) trailing-edge flap deflections. Amplitude modulation of the high frequency resonantbasedpiezoelectricactuatorsusedduringtheexperimentstoachievereducedfrequenciescomparable to those used in the experiments of Seifert et al.5,6,7 reduced the momentum required by an additional 50% to 70%, while maintaining the critical peak velocity of the actuator. Multiple excitation locations all controlled by a single actuator were available for controlling separation from the trailing-edge flap shoulder. It was demonstrated that for each slot location there was an optimal trailing-edge flap deflection that was most effective. In the highly curved region downstream of the trailing-edge flap shoulder, small changes 2of20 AmericanInstituteofAeronauticsandAstronautics in excitation slot locations resulted in large changes in the optimal trailing-edge flap deflection angle for lift increment. The momentum requirements for controlling flow separation on the trailing-edge flap were large when compared to other airfoil configurations5,6 and to the momentum required for leading-edge flap shoulder control.1 Additional testing2 indicated that coupling excitation from a slot located just upstream of the trailing-edge flap with excitation from a slot on the trailing-edge flap led to an improvement in airfoil performance. Similar to the leading-edge flap shoulder control, the output from high frequency resonant based actuators was modulated to lower frequencies. The flowfield was shown to be sensitive to the relative phasebetweentwoactuatorswhenactuatorswithdifferentresonantfrequenciesweremodulatedatthesame frequency. For the conditions documented at BART,2 the optimal phase relation was found to be ±30◦. The results from the BART version of the EET SHL model1,2,3 were promising and led to the present study of the model in the LTPT. The organization of this paper is as follows, first results from the current researcheffortarecomparedtothatofLinandDominik4 forthecruiseconfiguration. Then,theAFCresults are presented when control is applied upstream of the trailing-edge flap. The effects of frequency, amplitude and Reynolds number are presented. Similar results follow for the case where excitation is introduced from the trailing-edge flap of the model. Finally, the excitations from the upstream and downstream actuators are combined and the effect documented. II. Wind Tunnel Description The current experimental effort was conducted in the NASA Langley Low-Turbulence Pressure Tunnel, which is a single-return closed-throat pressurized wind tunnel.9 The tunnel can be pressurized from 1 to 10 atmospheres allowing for independent variation of Mach and Reynolds numbers. The test section is rectangular in shape, 0.914 m (3 feet) wide, 2.286 m (7.5 feet) high and 2.286 m (7.5 feet) long. The tunnel has a Mach number range of 0.05 to 0.5, a Reynolds number range of 1.3×106/m (0.4×106/ft) to 49.2×106/m (15×106/ft), and a stagnation pressure range of 101.325 kPa (14.7 psi) to 1013.25 kPa (147 psi). LTPTfeaturesalarge,17.6to1,contractionratioandasetofnineanti-turbulencescreensresultingin extremely low levels of freestream turbulence intensity, less than 0.05% for most cases.10 A schematic of the facility can be seen in Fig. 1(a) and a photograph of the EET-SHL model installed in the LTPT test section can be seen in Fig. 1(b). The tunnel is equipped with a wake rake comprised of twenty five total pressure probes. The wake rake is located approximately 1.34 (reference) airfoil chords, c, downstream of the center of rotation of the tunnel endplates or 0.84c downstream of the model trailing edge. (a)SchematicoftheLTPTcircuit (b) Photograph of the test section with the SHL airfoil modelinstalled,lookingupstream. Figure 1. The NASA Langley Low-Turbulence Pressure Tunnel (LTPT). 3of20 AmericanInstituteofAeronauticsandAstronautics 0.1 LE Actuator TE Actuator Flap Actuators y/c 0 F4F3F2 -0.1 F1 0 0.2 0.4 x/c 0.6 0.8 1 drooped leading edge simple sealed flap 0° to 30° 0°, 20°, 30°, 40°, 60° (a)CoordinatesofthecruiseEETconfiguration. (b)Flapsdeflectedandactuatorlocations. Figure 2. The Simplified High-Lift version of the NASA EET airfoil. III. Model Description The simplified high-lift version of the NASA EET airfoil utilized here is shown in Fig. 2. The model has a cruise (reference)chordlength, c, of 0.762 m (30.0 inches) and features a 15% chord leading-edge flap that canbedeflectedfrom0◦ to–30◦ anda25%chordtrailing-edgeflapthatcanbedeflectedfrom0◦ to60◦.The flowcontrolresultsthatwillbepresentedareforaleading-edgeflapdeflection, δ , of−25◦, andatrailing- LE edge flap deflection, δ , of 30◦ at M=0.1. The two-dimensional model completely spans the test section TE of LTPT and is mounted on two endplates, at the East and West sides (right and left sides, respectively, looking upstream) of the tunnel, which attach the model to its support and an integrated balance. The model is equipped with 270 static pressure taps, where 144 taps are distributed along the centerline of the model and 42 taps on each of two streamwise lines located 6 inches from each endplates. There are three spanwise rows, eachwith14taps, locatedatx/c=0.10, 0.35, and0.94. Additionally, the modelis equipped with several dynamic pressure transducers on the model surface. IV. AFC Actuators OnemotivationfortheLTPTtestwastoincreasetheoptionsforAFCactuatorsthatcouldbeutilizedin theexperiment,duetoaphysicallylargermodel. ForthesmallerBARTmodel,allactuationwasdoneusing internal piezoelectric actuators. Also, on the flap of the BART model, one actuator was used to drive three slotsviaacommoncavity. Thissignificantlylimitedtheflowoutofmultipleslotsandessentiallylimitedthe study to a single slot on the trailing-edge flap. The LTPT model has 6 independently controlled actuators for applying excitation at 6 locations—one on the shoulder of the leading edge, one on the shoulder of the trailing-edge flap, and four on the trailing-edge flap itself (see Fig. 2(b)). ThenamingconventionfortheactuatorsfollowsthatshowninFig.2(b)andinMelton,Yao,andSeifert.3 Theactuatorlocatedattheshoulderoftheleading-edgeflapistheleading-edgeactuator,theactuatorlocated immediately upstream of the trailing-edge flap is the trailing-edge actuator and the actuators on the flap itself are referred by number, Flap #1 (F1) is the actuator closest to the trailing edge of the model, Flap #4 (F4) is the slot on the flap closest to the leading edge of the model. The slot heights, h, for all actuators are nominally 0.635 - 0.762 mm (0.025 - 0.030 inches). A summary of the type of actuators used and their performance is provided in Table 1. A. Leading-Edge Actuator The leading edge (LE) actuator assembly is a zero-net-mass-flux (ZMF) actuator, which consists of 4 inde- pendent internally-mounted electromagnetic actuation units. The excitation covered the center 64% span of the model and the AFC slot was located at x/c=0.14. The electromagnetic actuation units were driven in pairs, each pair on a channel of a stereo audio amplifier supplied with a common signal. Typical signal supplied to the actuators was 30 Volts RMS with a peak velocity of around 30 m/s and a frequency range of 80 - 200 Hz (Table 1). 4of20 AmericanInstituteofAeronauticsandAstronautics B. Trailing-Edge Actuator The excitation from the trailing-edge actuator (TE actuator) assembly, located at x/c=0.73, was created by four externally mounted ZMF actuation units. The output from the four actuation units were connected to a common plenum inside the model by sealed pipes that went through the model endplates. Physically, there was a pair of actuation units mounted on the outside of the model endplate on the East side and on the West side of the tunnel. Electrically, one actuation unit on each side of the tunnel formed a pair with an actuation unit on the other side and each of these pairs was connected to a channel of a stereo audio amplifier supplied with a common signal. The typical signal supplied to each pair of actuators was 30 Volts RMS, resulting in a peak velocity of about 100 m/s and a frequency range of 50 - 200 Hz (Table 1). The spanwise uniformity of the TE actuator excitation was assessed prior to the wind tunnel test. Hot- wire measurements acquired at the 3 streamwise pressure tap locations are presented in Fig. 3. During the calibration,datawerenotacquiredbelow100Hzusing30VoltsRMS,whichisthecauseoftheapparentshift in Fig. 3(c). The external actuator arrangement produces the maximum output at the centerline, with the excitation on the West side of the model being the lowest. The ensemble averaged time histories presented in Fig. 4 from the calibration data indicate that the difference in phase between the centerline output and either the East or West output is about 12.5 degrees at 150 Hz operating frequency. The performance of the TE actuator was observed to degrade as the total pressure of the tunnel was increased. ToachievetheRe =1.7×106 andM=0.1condition,thetunnelwasrunatatmosphericpressure c (14.7 psia). At the Re = 3.4×106 and M = 0.1 condition, the tunnel total pressure was raised to 29 psia c andattheM=0.1andRe =6.0×106 and9.0×106 conditions,thetunneltotalpressurewas50psiaand75 c psia, respectively. The output from the TE actuator was characterized by making hot-wire measurements of the exit velocity as a function of drive frequency and amplitude. The RMS of the jet exit velocity, U , j as a function of frequency is shown in Fig. 5. The hot-wire was calibrated at the same elevated pressures using a calibration jet prior to acquiring the characterization data. In addition to the hot-wire, a total pressureprobeutilizingadynamicpressuretransducerwaspositionedattheslotexitatadifferentspanwise location. For the TE actuator, the hot-wire data proved to be unreliable. The data from the total pressure probe indicates that the output velocity of the actuator, for a fixed drive amplitude, is reduced at elevated pressure. ThistrendisconsistentwithotherAFCactuatorcharacterizationsmadeatelevatedtotalpressures by the authors in the same facility. At elevated total pressures, the output from the actuator is reduced and frequency at which the peak velocity is observed shifts slightly towards lower frequencies. Since the total pressure probe was not dynamically calibrated, only the data at the lowest sampled frequency is considered reliable to provide an indication of how the performance is effected by the elevated total pressures. Based on measurements made with the total pressure probe at 50 Hertz drive frequency, the performance of the TE actuator at 29 psia was reduced to 75% of that at atmosphere, reduced to 57% at 50 psia, and reduced to 44% at 75 psia. These values are indicated by the single points presented in Fig. 5. 70 70 70 60 60 60 50 50 50 s 40 s 40 s 40 m/ m/ m/ U,j30 U,j30 U,j30 20 20 20 Centerline Centerline Centerline West West West 10 East 10 East 10 East 0 0 0 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400 Freqeuncy,Hz Freqeuncy,Hz Freqeuncy,Hz (a)VRMS =10. (b)VRMS =20. (c)VRMS =30. Figure 3. TE actuator calibration at three different spanwise locations. 5of20 AmericanInstituteofAeronauticsandAstronautics 120 80 Centerline West 100 East 60 80 ATM U,m/s60 U,m/sj40 29psia 50psia 40 75psia 20 20 0 0 0.002 0.004 0.006 0.008 0 time,s 0 50 100 150 200 250 300 350 400 450 Frequency(Hz) Figure 4. TE actuator ensemble average hot-wire ve- Figure 5. TE actuator frequency sweep at differ- locitymeasurementsatf=150Hzand30VRMS input. ent tunnel total pressures, 30 VRMS. Velocity mea- surementsatatmospherearefromhot-wiremeasure- ments, at elevated pressures total pressure probe. C. Trailing-Edge Flap Actuators The low Reynolds number SHL results on the BART model2,3 indicated that the momentum required to control flow separation on the trailing-edge flap of the model was significantly higher than the momentum requiredforcontrollingleading-edgeflowseparationandhigherthanthemomentumrequiredforcontrolling flow separation on the NACA 0015 at flight Reynolds numbers. The BART data2 also indicated that combining multiple excitations would increase control authority. Furthermore, the BART data3 indicated thatthechordwisespacingoftheexcitationslotsshouldbeclosertogetheraroundthetrailing-edge-flapknee due to the sensitivity to slot location in this highly curved region of the model. Table 1 presents the slot locationsthatwerechosenbasedonthelowReynoldsnumberdatabase. Thetrailing-edgeflapactuators(TE flap actuators) used were a major upgrade from the ones used in BART and the size of the model enabled four actuator assemblies to be placed in the trailing-edge flap of the model. Each flap actuator assembly consisted of a linear array of 19 piezoelectric actuation units. Each array was independent of the others and was divided in half for the purpose of electronic excitation. The division of the array of 19 actuation units was done in a way that the phase between spanwise adjacent actuation units could be varied. For each of the TE flap slots, half of the piezoelectric actuators were connected to one channel of a stereo audio amplifier and the other half were connected to the other channel. The typical signal supplied to each array of actuators was 100 Volts RMS with a peak velocity of approximately 50 m/s at f=1500 Hz (Table 1). Actuator Type x/c Location Frequency Range U peak,max Leading Edge Electromagnetic 0.14 80-200 Hz 30 m/s Trailing Edge Electromagnetic 0.73 50 - 200 Hz 100 m/s Flap #1, F1 Piezoelectric 0.84 1500 Hz† 50 m/s Flap #2, F2 Piezoelectric 0.76 1500 Hz† 50 m/s Flap #3, F3 Piezoelectric 0.74 1500 Hz† 50 m/s Flap #4, F4 Piezoelectric 0.73 1500 Hz† 50 m/s Table 1. Actuator types by location. †Lower frequencies were obtained with the piezoelectric actuator using amplitude modulation. 6of20 AmericanInstituteofAeronauticsandAstronautics V. Cruise and High-Lift Baseline Airfoil Performance The cruise configuration (δ =δ =0◦) of the model was tested prior to deflecting the leading- and LE TE trailing- edge flaps and applying active flow control as was done with the BART model.1 The SHL BART dataset of Melton et al.1 was acquired at Reynolds numbers much lower than the data of Lin.4 However, a comparisonatthe same Re canbemadebasedondatafromthecurrentresearcheffort. Figure6compares c the lift values at Re =2.5×106 and Re =9×106. The SHL has a much larger slope change with Reynolds c c number when compared to the data of Lin. The increase in the stall angle with Reynolds number for the SHL model is due to the sidewall venting utilized for the SHL data. The data of Lin were acquired without sidewall venting and with the sidewall venting holes for boundary-layer control taped. The SHL model was observed to stall sooner (between 15◦ and 16◦) at Re = 9×106 when the endplates were taped and no c sidewall venting used. The large Reynolds number sensitivity seen with the SHL EET model is believed to be primarily due to blockage effects that are larger for this 0.762 m chord model when compared to the 0.55 m chord model of Lin. When wall corrections11 are applied to the data there is a closer agreement between the lift data for the two models (Figure 6). The high-lift configuration for the model combines a leading-edge flap deflection and a trailing-edge flap deflection (δ =−25◦, δ =30◦). Figure 7 compares the pressure distribution on the BART SHL model LE TE to those on the LTPT SHL model. The C distributions presented, at α = 8◦, indicate the presence of a p laminar separation bubble on the low Reynolds number BART model at the LE flap shoulder (x/c = 0.15). The lower surface C values are almost identical although there are differences near the hinge locations, p x/c=0.25 and x/c=0.75, for the two datasets. Wind tunnel wall interference is the reason for larger C L generated by the BART model. The solid blockage for the BART model is about 70% higher than that of the LTPT model. The baseline C distributions also indicate that the TE flap is completely separated for p all cases studied. 2.5 -5 Re=9.0x106 x 2 RRRRReeeeecccccc=====92292.....00555xxxxx111110000066666,,,,LCLCiioonnrrrreecctteedd --43 xxxxxxx xxxxx x RRRRReeeeeccccc=====01369.....774005xxxx1111x00001666606,BART xxxxxxxx x x CL1.5 Cp -2 x xx -1 x xxxxxxxxxxxxxx x xxx xxx xxxxxxxxxxxxxxxxxxxxxxxxx 1 x 0 x x 0.50 5 10 15 20 1 0xxxxxxxxxxxxxxxxxxxx0x.x2xxxxx x0.4xxxxxxx0.x6xxx xxxx0x.8xxx x 1 AngleofAttack(Deg) x/c Figure 6. Lift coefficients for the cruise configuration Figure 7. Baseline Cp distribution comparing BART of the EET model. data to LTPT data. M=0.1, α=8◦. VI. TE Actuator Control Results A. Effect of TE Actuator Excitation Frequency The effects of changing the frequency of the TE actuator on C are presented in Fig. 8(a) at two angles L of attack (α=0◦ and 8◦). The data, acquired at Re = 1.7×106, indicate that F+=0.44 and 0.89 are the c best frequencies in the range presented for increasing C . Comparable lift increments are obtained with L F+ = 0.44 and F+ = 0.89 at α = 0◦. At α = 8◦, F+=0.89 is not as effective as F+=0.44 at increasing lift. The form drag data shown in Fig. 8(b) indicate that C is increased at α = 0◦ with F+=0.89 and at dp α = 8◦ with F+ = 0.44. The sensitivity to F+ at these two angles of attack provides the basis for the F+ 7of20 AmericanInstituteofAeronauticsandAstronautics 1.4 2.19 0.11 0.15 α=0o α=0o α=8o α=8o 1.38 2.16 1.36 2.13 0.1 0.13 o0 o8 o0 o8 = = = = α1.34 2.1α α α C,L C,L C,dp C,dp 1.32 2.07 0.09 0.11 1.3 2.04 1.28 2.01 0.08 0.09 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2 F+ F+ (a)CL (b)Cdp Figure 8. TE actuator frequency scan. Rec=1.7x106;δLE=−25◦;δTE=30◦;M=0.1. values used throughout the current research effort. These results are consistent with the results of Melton, Yao, and Seifert2 when using excitation from this location on the model. B. Effect of TE Actuator Excitation Amplitude In Fig. 9 the effect of changing the excitation amplitude for the Re = 1.7×106 case is presented for c two frequencies; F+=0.44 and F+=0.89. The results indicate that the F+=0.44 case is more effective at increasing the total aerodynamic force, the lift increases above the baseline as does the drag. For the F+=0.89case,boththe∆C andthe∆C areinitiallynegative. The∆C reachesaminimumearlyonand L dp L then begins to increase, while the ∆C continues to decrease, which is an indication of less flow separation dp onthetrailing-edgeflap. FortheF+=0.89excitationcase,∆C beginstoincreasewhenC exceeds0.36%, dp µ at which point ∆C becomes positive, although the C levels continue to remain less than the baseline L dp case (C = 0%). The C distributions at α = 8◦ corresponding to C =0.36% for the two frequencies are µ p µ presentedinFig.10. ThedataindicatethattheC valuesusedarenotsufficienttocompletelyeliminateflow µ separation on the trailing-edge flap. The data also indicate that the two frequencies have different effects on the flowfield. On the trailing-edge flap the higher F+ excitation increases the model trailing-edge pressures while the F+=0.44 case decrease the pressures on the trailing-edge flap. Additionally, the F+=0.44 case increases the upper surface suction, thus increasing the overall circulation for the airfoil. In contrast, the F+=0.89 case locally improves the TE pressure recovery indicating a degree of separation control on the flap. C. Effect of Reynolds Number The data presented in Fig. 11 indicate the effect of varying the excitation amplitude (F+ = 0.89) of the actuator at Re = 1.7×106 and Re = 3.4×106. The excitation has the same effect on the flowfield for c c these two Reynolds numbers. At higher Reynolds numbers, Re = 6×106 and Re = 9×106, this was no c c longerthecase. ThedifferencesathigherRe arebelievedtobeduetoactuatoroutputdecreasingastunnel c total pressure increases. The actuator was characterized by making velocity measurements via a hot-wire anemometer. These characterizations were performed at total pressures of 50 psi and 75 psi, corresponding to Re = 6×106 and Re = 9×106 at M=0.1, and indicated a decrease in actuator output with increasing c c total pressure (Fig. 5). The pressure distributions acquired at Re = 1.7×106 (Fig. 12) and Re = 3.4×106 (Fig. 13) provide c c a possible explanation for the results obtained at Re = 3.4×106 that seem to contradict the hot-wire c calibration data. Note, identical dimensional frequencies are used for the two Reynolds numbers of Figs. 12 8of20 AmericanInstituteofAeronauticsandAstronautics 0.3 0.015 F+=0.44 F+=0.44 F+=0.89 F+=0.89 0.2 0.01 0.005 CL0.1 dp ∆ C ∆ 0 0 -0.005 -0.1 -0.01 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 C ,% C ,% µ µ (a)∆CL (b)∆Cdp Figure 9. Effect of TE actuator excitation amplitude on ∆CL and ∆Cdp. Rec=1.7x106;δLE=−25◦;δTE=30◦;M=0.1,α=8◦. -5 Baseline -4 F+=0.44 F+=0.89 TESlot -3 p -2 C -1 0 1 0 0.2 0.4 0.6 0.8 1 x/c Figure 10. Cp distribution with TE actuator control comparing F+=0.44 and F+=0.89. Rec=1.7×106 ; M=0.1; α=8◦. 9of20 AmericanInstituteofAeronauticsandAstronautics -3 0.25 Baseline 0.2 RReecc==13..74xx110066 FF++==00..8437 -2 F+=1.03 0.15 CL0.1 Cp -1 ∆ 0.05 0 0 -0.050 10 20 30 40 1 0 0.2 0.4 0.6 0.8 1 VRMS x/c Figure11. ∆CL vs. VRMSforTEactuatorcomparing Figure 12. Cp distribution for TE actuator excitation. two Reynolds numbers. α=0◦; F+=0.89. Rec=1.7×106; α=0◦. Only every fourth data point marked with symbols. -3 1.6 Baseline F+=0.47 1.55 VRMS=15 VRMS=22 F+=0.94 VRMS=30 -2 F+=1.09 1.5 VRMS=30,Rec=3.4x106 1.45 p -1 C CL1.4 1.35 0 1.3 1.25 1 0 0.2 0.4 0.6 0.8 1 1.2 x/c 0 0.2 0.4 0.6F+ 0.8 1 1.2 Figure 13. Cp distribution for TE actuator excitation. Figure 14. Frequency scan with varied amplitude for Rec=3.4×106; α=0◦. Only every fourth data point TE actuator. α=0◦; Rec=1.7×106 . marked with symbols. 10of20 AmericanInstituteofAeronauticsandAstronautics