E ects of Canard Position on Wing Surface...

Transaction B: Mechanical EngineeringVol. 17, No. 2, pp. 136{145c Sharif University of Technology, April 2010

E�ects of Canard Positionon Wing Surface Pressure

M.R. Soltani1;�, F. Askari1, A.R. Davari2 and A. Nayebzadeh1

Abstract. A series of wind tunnel tests were performed to study the e�ects of a canard and itsposition on the downstream ow�eld over the wing surface. The wing surface pressure was measuredfor both canard-o� and canard-on con�gurations. In addition, the canard position e�ects on the wingwere investigated at di�erent angles of attack. The canard was installed at three vertical positions andat two di�erent horizontal distances from the wing apex. The results show a remarkable increase in thewing suction peak for the canard-on con�gurations. At low to moderate angles of attack, among thevarious con�gurations examined in the present experiments, the mid-canard con�guration developed ahigher suction on the wing while, at high angles of attack, the upper-canard was found to induce the mostfavorable ow �eld on the wing. In addition, higher suctions were achieved on the wing at moderate tohigh angles of attack, as the wing-canard distance was increased.

Keywords: Canard; Delta wing; Downwash; Leading edge vortex.

INTRODUCTION

Canard-wing con�gurations play an important rolein modern �ghter aircraft design. High performanceaircraft, such as military �ghters, often require high-lift for a wide range of angles of attack to maintaintheir maneuverability. For two-dimensional airfoils, themaximum lift is typically obtained at about 10�-15�angles of attack after which the airfoil stalls. Thus,high performance aircraft require lift enhancementwhen operating under high angle of attack conditions.One technique for obtaining enhanced lift at moderateangles of attack is through the use of delta-shapedplatforms along with strakes. As ow separates alongthe leading edges of a delta wing at moderate angles ofattack, vortical ow results. These vortices producea very low-pressure region and can account for upto 30% of the total lift at moderate to high anglesof attack [1]. A 70� swept delta wing, for example,continues to increase its lift until an angle of attack ofabout 35� [2]. However, there are some limits to the

1. Department of Aerospace Engineering, Sharif University ofTechnology, Tehran, P.O. Box 11155-8639, Iran.

2. Department of Mechanical and Aerospace Engineering, Is-lamic Azad University, Science and Research Campus,Tehran, P.O. Box 14155-4933, Iran.

*. Corresponding author. E-mail: [email protected]

Received 20 June 2009; received in revised form 4 October 2009;accepted 12 December 2009

advantages provided by the delta wing vortices. As theangle of attack is increased, a sudden breakdown inthe vortex structure occurs. This phenomenon knownas vortex bursting, results in a sudden stagnation ofaxial ow and an expansion in the vortex radial size [3-5]. Once the vortex structure is diminished, lift startsto decay. Thus, it is desirable for the vortex burstto be delayed as far as possible to get higher lift.Attempts to control delta wing vortices (under staticconditions) includes blowing [6,7], suction and the useof canards [8]. The canard induces a non uniformdistribution of local angles of attack on the wingsurface, which leads to a non conical vortex formationover the wing, and delays the vortex breakdown tohigher angles of attack. On the other hand, thewing produces an up-wash �eld on the canard, whichincreases its lift. The wing also induces a longitudinalvelocity component on the canard. This inducedvelocity delays the vortex breakdown onset on thecanard and increases its performance at moderate tohigh angles of attack [9,10]. Several investigations havebeen devoted to studying the ow�eld associated withcanard-wing con�gurations. Most of them involve thestudy of canard shape and deection angle as well asthe body angle of attack on the ow�eld. In this paper,a comprehensive survey was performed to underscorethe impact of canard and its position on the structureof the ow�eld over a delta wing platform.

These results can be generalized to investigate

E�ects of Canard Position on the Wing Surface Pressure 137

split canard con�gurations where two canards are usedin a tandem arrangement to increase performance andmaneuverability. Further, in modern conventional highlift systems, interaction of the forward element on themain body is similar to that observed on the canard-wing model considered in the present experiments.

MODEL AND EXPERIMENTALAPPARATUS

The present experiments were performed in a subsonicwind tunnel. The tunnel is of a closed return typeand has a test section with dimensions of 80�80 cm.The maximum attainable speed in the test section is100 m/sec and the Reynolds number varies between5.29�105 and 5.26�106 per meter. Turbulence inten-sity in the test section has been measured to be lessthan %0.1.

The investigations have been performed on acanard-wing-body con�guration. Figure 1, shows themodel installed in the test section. A 60� swept canardis placed in front of the main 60� swept delta wing.The main wing has a constant thickness of 10 mm andis made of aluminum alloy. The wing and canard aspectratios are 1.17 and 1.15, respectively, and both have ahexagonal pro�le.

Seventy eight pressure ports of 0.8 mm innerdiameter were carefully drilled on the upper surfaceof the main wing. Figure 2 shows the pressure tabsarrangement on the wing.

Both the wing and the canard were attached to arectangular cross section fuselage. The model assemblywas such that the canard position on the body couldbe varied in both longitudinal and lateral directions, asshown in Figures 3a through 3f.

The tests were conducted at a constant air speedof 80 m/sec, corresponding to a Reynolds number of1:480 � 106 based on the wing root chord. The modelangle of attack was varied from �10� to 30�. All datawere acquired by an AT-MIO-64E-3 data acquisition

Figure 1. The model installed in the test section.

Figure 2. The pressure tabs arrangement on the wing.

board capable of scanning 64 channels at a rate of500 KHz.

Finally, all data were corrected for the windtunnel sidewalls and the wake blockage e�ects. Ananalytical approach [11] was also used to estimate theerrors involved in the pressure measurements. Both thesingle sample precision and the bias uncertainty in thepressure measurement were estimated. On this basis,the overall uncertainty for the presented data is lessthan �3% [12,13].

Results

The results presented in this section are for variouscases of canard-on and canard-o� con�gurations. Fig-ure 4 shows the wing surface pressure distributionfor both canard-o� and canard-on con�gurations at� = 10�; 20� and 30�. The leading-edge vortexsignature can be observed from the suction peaks forboth con�gurations. At 10 and 30 degree angles ofattack, the wing pressure distribution for the canard-on cases show that the value of the suction peaksfor the front sections, � � 0:5; are higher than thecorresponding locations on the canard-o� ones, whileat the rear sections of the wing, the canard-on surfacepressure is obviously higher (Figure 4). Further, itis observed that the wing vortex core for the canard-on con�gurations lies closer to the wings leading edgethan those of the canard-o� ones. However, for bothcon�gurations, the amount of suction on the wingsurface reduces as it moves streamwise from the wingleading edge to the trailing edge. At � = 30�, thespanwise pressure is nearly constant on the entire wingsurface for the canard-o� con�guration (an indicationof the vortex breakdown) while for the canard-on case,the ow near the leading edge (� = 0:3 and � = 0:4)is seen to be highly energized by the canard vortical�eld and a strong suction is developed in this area(Figure 4c).

The above phenomenon can be observed moreclearly in Figure 5, where a direct comparison ofpressure distributions over the wing is made for the

138 M.R. Soltani, F. Askari, A.R. Davari and A. Nayebzadeh

Figure 3. Various model con�gurations.

canard-on and canard-o� con�gurations at � = 15�and 30� and at four di�erent chordwise sections. Thesuction peak for a 15� angle of attack is nearly sharp,while that for � = 30� is atter and its value is muchlower when compared with the � = 15� case. This isan indication of the vortex burst onset, i.e. at a 30�angle of attack, the vortex on the wing surface bursts;thus, the suction peak has been reduced signi�cantly.

For sections � = 0:5 to � = 0:7, the suction peakindicates a stronger vortex for the canard-on con�gu-rations than that for the canard-o� ones. However, asthe angle of attack is increased to 30�, it seems that theburst position of the canard vortex reaches the wingsurface around � = 0:7 and the canard-on suction inthis case is less than that of the canard-o� (Figure 5).At a farther distance, i.e. � = 0:8, the whole spanwisesection is exposed to the burst vortex of the canard andthe wing.

Quantitatively, at � = 0:5, the presence of thecanard has enhanced the wing vortex strength by about150% at � = 15� and 25% at � = 30� in comparison

with the canard-o� con�guration. At � = 0:7, thedi�erence between the suction peak values of thewing for the canard-on and canard-o� con�gurationsis about 27% at � = 15� and about 14% at � = 30�.

Regarding further comparisons with the canard-o� cases, the suction peak at all sections on the wing forthe canard-on con�gurations are closer to the leadingedge. At � = 0:8, the presence of the canard hasenhanced the wing vortex strength by about 45% at� = 15�, but at � = 30� the vortex strength for bothcon�gurations has been decreased and the suction peakfor the canard-on con�guration is less than that of thecanard-o� one.

It is well known that the canard induces a down-wash �eld within its span and an upwash �eld outsideits span. The downwash reduces the e�ective angle ofattack considerably in the forward and inner portionof the wing, which leads to a suppression of the owseparation on the wing surface. The upwash �eldincreases the e�ective angle of attack in the outerand rear portion of the wing, which enhances the ow


Figure 4. Surface-pressure distribution for several angles of attack at di�erent sections for the mid-close-canard case.

separation there. This mechanism delays the formationof the wing vortex for the canard-on con�guration.

Due to the nonuniform distribution of the e�ectiveangle of attack along the leading edge of the wing,the wing vortex is fed with vorticity in a di�erentmanner than in the case of the canard-o� con�guration.As a result, the wing for the canard-on con�guration

sees a lower e�ective angle of attack and this leadsto a compensation of an additional lift of the canardthrough a loss of lift of the wing. In the rear portionof the wing, near the trailing edge, the magnitudeof the suction peak is lower than that of the frontpart, at all angles of attack. At high angles of attack,however, the suction peak is removed for the canard-


Figure 5. Direct comparison between canard-o� and canard-on con�guration for the mid-close-canard.


o� con�guration, at all sections, indicating the vortexbreakdown phenomenon.

As observed earlier, for the canard-on con�gura-tion, the suction peak on the wing surface pressure atfront and middle portions are higher than the corre-sponding canard-o� cases while at the rear portion,the amount of the suction peak for the canard-oncon�guration has been reduced. This implies thatthe domain of the favorable inuence of the canard is

mostly restricted to the front and middle portions ofthe wing.

Experimental data for the six di�erent wing-canard con�gurations are shown in Figure 6. Thee�ects of the canard position on the wing surface pres-sure distribution are clearly seen. For this study, thecanard was located at three di�erent vertical positionsand at two di�erent horizontal locations. The modelsare 1-upper close canard, 2-upper-far canard, 3-mid-

Figure 6. E�ect of canard position on the wing pressure distribution.


close canard, 4-mid-far canard, 5-low-close canard and6-low-far canard. Note that all di�erent con�gurationsare shown in Figure 3. The canard-o� data are alsoshown in each �gure for comparison. For the canard-on con�gurations, the wing surface pressure patternremains the same, regardless of the canard position.From Figure 6, the di�erence between the canard-onand canard-o� con�gurations in developing the suctionregion on the wing surface is clearly evident.

The suction peaks lie closer to the wing leadingedge for all canard-on con�guration cases. At � = 10�and for � = 0:6, the presence of the canard has movedthe suction peak toward the wing leading edge, andits value has been increased about 5 to 25 percent,depending on the location of the canard relative to thewing (Figure 6a). At � = 0:8 and for the same angle ofattack, however, the canard has displaced the suctionpeak on the wing towards the leading edge by about 7%of the local chord. At � = 20� and for � = 0:6 and 0.8,the suction peak movement toward the leading edge,due to presence of the canard, was about 3% to 7% ofthe local chord, respectively.

Furthermore, from Figure 6 it is clearly seen thatfor the canard-o� cases, the suction peak is much lessthan for that of the canard-on cases. This phenomenonindicates that the canard not only moves the vortexlocation over the wing surface, but also increases itsstrength.

Further comparison of the experimental datashown in Figures 6e and 6f reveals that at highangles of attack, the wing vortex for the canard-o�con�guration bursts, while for the wing with a canardlocated upstream, the ow�eld has been signi�cantlyimproved, especially for the � = 0:6 case. For � = 0:8,Figure 6f shows that at this angle of attack, only fortwo canard locations, an improvement in the vortical

ow is achieved. For other con�gurations, Figure 6fshows that there still exists a vortex on the wing, dueto the canard, however, this vortex is not as strong asthose observed in Figures 6a through 6e.

Figure 6 also indicates that among the di�erentcanard positions considered here, the up-canard con�g-uration has the strongest and the low-canard has theweakest favorable e�ect on the wing vortex.

The maximum suction value at di�erent anglesof attack for six sections on the wing surface for bothcanard-o� and canard{on cases is shown in Figure 7.According to this �gure, the amount of suction on thewing decreases as moving away from the wing apex. Atthe rear portions of the wing, near the trailing edge,the amount of suction is roughly half that of the frontregion, � = 0:4.

At the front section of the wing, � = 0:4, for lowto moderate angles of attack, the suction peak for thecanard-o� con�guration is higher than the canard-oncase, but, at high angles of attack this phenomenon is

reversed. At the middle section of the wing (� = 0:5to 0.8) the suction for most canard-on con�gurationsis higher than that of the canard-o�, especially forthe mid and up-canard con�gurations. At the rearportion of the wing, �=0.9, the suction peak is reducedsigni�cantly for all con�gurations examined and at allangles of attack, which, as noted earlier, indicates thatthe canard vortex strength at the rear portion of thewing decreases. As shown in Figure 7, for the mid-canard con�guration case, higher suction is achievedat low to moderate angles of attack, and at high anglesof attack, at most sections of the wing, the high-canardcon�guration induces a higher suction peak on the wingsurface.

The wing surface pressure contours at di�erentangles of attack for the canard-o� and canard-on(upper far canard) con�gurations are compared inFigure 8. As shown in this �gure, a low pressure regionover the wing, indicating the presence of the leadingedge vortex, is clearly seen, and it is observed thatfor all cases, the vortex strength near the wing apexis much higher than that in the rear region near thetrailing edge. According to this �gure, the width ofthe low pressure region over the wing surface for thecanard-on case is smaller than the canard-o� one andis aligned closer to the leading edge. The distributionof pressure over the wing at � = 30� indicates that thevortex bursts for the canard-o� case, while the suctionpeak in the pressure distribution for the canard-on caseveri�es that a vortex does exist at the front portion ofthe wing.

For the mid-canard con�guration and at lowangles of attack, the suction peak increases, however,at higher angles of attack, this canard position is notthe best (Figure 8). For high angles of attack, thedata shows that when the canard is located at itsupper position, a remarkable improvement is achievedin the surface pressure pattern. On the other hand,by increasing the longitudinal clearance between thecanard and the wing, the absolute value of the suctionpeak is increased for moderate to high angles of attack.It seems that `upper and far' is a better positionfor the canard from among the wing-canard-bodycon�gurations examined in this investigation.

CONCLUSION

Extensive low speed wind tunnel tests have beencarried out on several wing-canard-body combinations.Surface-pressure results are presented for both canard-o� and canard-on con�gurations. In addition, thee�ect of canard location on wing surface pressure wasinvestigated. The main results of these investigationsare as follows:

1. The downwash of the canard reduces the e�ective


Figure 7. Maximum suction developed at the di�erent sections of the wing.

angle of attack at the inboard section of the wingnear the apex. This inhibits ow separation onthe wing and, therefore, delays the formation ofthe leading-edge vortex. At the wing outboardportion, the upwash of the canard separates the

ow, forming a side-edge vortex.

2. For the canard-on con�gurations, a considerably

high suction region exists in the front portion of thewing and the suction peaks lie closer to the leadingedge when compared with the canard-o� cases.

3. For the low-canard con�guration, minimum suctionis developed on the wing and this con�gurationis shown to have the least performance amongthe cases considered. At low angles of attack,


Figure 8. Pressure distribution over the wing for various angles of attack.


the maximum suction on the wing belongs to themid-canard con�guration, while at higher angles ofattack, the high-canard induces the highest suctionon the wing.

4. By increasing the horizontal distance between thecanard and the wing, higher suction is achieved formoderate to high angles of attack. It seems thatthe upper far position for the canard is a properchoice for the best performance at high angles fromamong the wing-canard-body con�gurations.

ACKNOWLEDGMENT

The work was supported by the AIO and the ADSRC.The authors would like to express their gratitude tothe chairman and personnel for their cooperation andsupport.

NOMENCLATURE

x chordwise coordinate measured from thewing apex,

y spanwise coordinate measured from the root,� local chordwise coordinate along the wing,

nondimensionalized by the root chord,� local spanwise coordinate along the wing,

nondimensionalized by the wing span,� model angle of attack,cp static pressure coe�cient.

REFERENCES

1. Ashraf, A. Omar, Sri, S. \Experimental investigationof incompressible ow around delta wing-slender bodycombinations", International Journal of Applied Engi-neering Research (IJAER), 2(4), pp. 685{698 (2007)

2. Wei-Jen, S., Curtis, W. \Aerodynamic characteriza-tion of a canard guided artillery projectile", AIAA2007-672 (2007).

3. Pandya, S.A. and Aftofmis, M.J. \Computation of ex-ternal aerodynamics for a canard rotor/wing aircraft",AIAA 2001-0997 (2001).

4. Howard, R.M. and O'Leary, J. \Flow �eld study ofa close-couple canard con�guration", AIAA Journal,31(4), pp. 908-914 (1994).

5. Earnshaw, P.B. and Lawford, I.A. \Low-speed windtunnel experiments on a series of sharp-edged deltawings", Aeronautical Research Council, RM 3424(1964).

6. Erickson, G.E. \Water-tunnel studies of leading-edgevortices", Journal of Aircraft, 19(6), pp. 442-448(1982).

7. Seginer, A. and Salomon, M. \Performance augmen-tation of a 60-degree delta aircraft con�guration byspanwise blowing", Journal of Aircraft, 23(2), pp. 801-807 (1986).

8. Wood, N.I. and Roberts, L. \Control of vortical lifton delta wings by tangential leading-edge blowing",Journal of Aircraft, 25(3), pp 236-243 (1988).

9. Lee, M. and Ho, C.M. \Lift force of delta wings",Applied Mechanics Review, 43(9), pp 109-221 (1990).

10. Hummel, H. and Oelker, H.C. \Low-speed character-istics for the wing-canard con�guration of the inter-national vortex ow experiment", Journal of Aircraft,31(4), pp 868-878 (July-Aug. 1994)

11. Rae, H.W., Jr. and Pope, A., Low-Speed Wind TunnelTesting, John Wiley & Sons, Inc. (1984)

12. Thomas, G.B., Roy, D.M. and John, H.L.V., Me-chanical Measurements, Addison-Wesley PublishingCompany (1993)

13. Nayebzade, A. \An experimental survey on the e�ectsof canard position on the ow�eld over the wingin subsonic ow", Msc. Thesis, Sharif University ofTechnology (Jan. 2008)

BIOGRAPHIES

Mohammad Reza Soltani, a professor in theaerospace engineering department of Sharif Universityof Technology, Tehran, has a PhD in aerodynamicsfrom the University of Illinois at Urbana-Champaign,USA. His research interests include applied aerodynam-ics, unsteady aerodynamics wind tunnel testing, windtunnel design, and data processing.

Farshid Askary graduated with a BS degree inaerospace engineering from Sharif University of Tech-nology, from where he also received his MS in aero-dynamics. His research interests include applied aero-dynamics, unsteady aerodynamics wind tunnel testing,and data processing.

A.R. Davari is assistant professor in the departmentof mechanical and aerospace engineering at the scienceand research campus of the Islamic Azad University inIran. He has conducted several experimental studieson unsteady aerodynamics and interference e�ects. Heis also interested in new prediction and optimizationmethods in aerodynamics, such as neural networks andevolutionary algorithms. He graduated with a PhDdegree in 2006 and has published over 11 journal paperssince that time.

Arash Nayebzadeh graduated with a BS degreein aerospace engineering from Sharif University ofTechnology, from where he also received his MS inaerodynamics. His research interests include appliedaerodynamics, unsteady aerodynamics wind tunneltesting, and data processing.

E ects of Canard Position on Wing Surface...

Documents

Transcript of E ects of Canard Position on Wing Surface...