See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245429733

Prediction of Delta Wing Leading-Edge Vortex Circulation and Lift-Curve

Slope

Article in Journal of Aircraft · May 1997

DOI: 10.2514/2.2193

CITATIONS READS

29 985

1 author:

Lance W Traub
Embry–Riddle Aeronautical University
132 PUBLICATIONS 1,823 CITATIONS

SEE PROFILE

All content following this page was uploaded by Lance W Traub on 13 October 2017.

The user has requested enhancement of the downloaded file.

450 J. AIRCRAFT, VOL. 34, NO. 3: ENGINEERING NOTES

Acknowledgments a = angle of attack

This paper is, in part, based upon results obtained in a study G = vortex circulation
performed for ARPA, Contract DAAH01-94-C-R022, under g = vorticity per unit length
the direction of M. S. Francis. « = wing apex half-angle
L = leading-edge sweep angle
r = density
References
1
Ericsson, L. E., Mendenhall, M. R., and Perkins, S. C., Jr., ‘‘Re- Subscripts
view of Forebody-Induced Wing Rock,’’ Journal of Aircraft, Vol. 33,
L.E. = leading edge
No. 2, 1996, pp. 253 – 259.
2
Ericsson, L. E., ‘‘Various Sources of Wing Rock,’’ Journal of Air-
v = vortex
craft, Vol. 27, No. 6, 1990, pp. 488 – 494.
3
Ericsson, L. E., and Reding, J. P., ‘‘Fluid Dynamics of Dynamic Introduction
Stall, Part I, Unsteady Flow Concepts,’’ Journal of Fluids and Struc-
tures, Vol. 2, Jan. 1988, pp. 1 – 33.
4
Fratello, D. J., Croom, M. A., Nguyen, L. T., and Domack, C. S.,
T HE leading-edge suction analogy of Polhamus1 provides
an accurate means by which the aerodynamics of slender
sharp wings may be estimated. Slender swept wings typically
‘‘Use of the Updated NASA Langley Radio-Controlled Drop-Model have  owŽ elds dominated by conical leading-edge vortices.
Technique for High-Alpha Studies of the X-29A ConŽ guration,’’
The basic tenet of the suction analogy is that the leading-edge
AIAA Paper 87-2559, Aug. 1987.
5
Ericsson, L. E., ‘‘Moving Wall Effects in Unsteady Flow,’’ Journal
suction force through enforced leading-edge  ow separation,
of Aircraft, Vol. 25, No. 11, 1988, pp. 977 – 990. in combination with sweep, is effectively rotated through 90
6
Ericsson, L. E., and Reding, J. P., ‘‘Asymmetric Flow Separation deg to the plane of the normal force, and manifests as the force
and Vortex Shedding on Bodies of Revolution,’’ Tactical Missile Aero- required to effectively maintain equilibrium of the vortex
dynamics: General Topics, edited by M. J. Hemsch, Vol. 141, Prog- above the wing. The analogy does not, however, yield infor-
ress in Astronautics and Aeronautics, AIAA, Washington, DC, 1992, mation on the leading-edge suction distribution or vortex char-
pp. 391 – 452. acteristics.
7
Ericsson, L. E., ‘‘Unsteady Separation on Slender Bodies at High Use of a panel method allows determination of the leading-
Angles of Attack,’’ Journal of Spacecraft and Rockets, Vol. 30, No.
edge thrust or suction distribution by using the computed at-
6, 1993, pp. 689 – 695.
8
Brandon, J. M., and Nguyen, L. T., ‘‘Experimental Study of Effects
tached  ow sectional lift and vortex drag at each spanwise
of Forebody Geometry on High Angle of Attack Stability,’’ Journal station. The leading-edge suction distribution may also be de-
of Aircraft, Vol. 25, No. 7, 1988, pp. 591 – 597. termined analytically, using an expression derived by Purvis2
9
Bollay, W., and Brown, C. D., ‘‘Some Experimental Results on for arbitrary planforms. To calculate the strength of the lead-
Wing Flutter,’’ Journal of the Aeronautical Sciences, Vol. 8, No. 6, ing-edge vortex, a panel method may be employed, with var-
1941, pp. 313 – 318.
10
ious methodologies being used to estimate the vortex strength,3
Raghunathan, S., Harrison, J. R., and Hawkins, B. D., ‘‘Thick e.g., calculating the velocity above and below the leading-edge
Airfoil at Low Reynolds Numbers and High Incidence,’’ Journal of wake. Euler and Navier – Stokes solvers can also be used to
Aircraft, Vol. 25, No. 7, 1988, pp. 669 – 671. determine the vortex properties, but are computationally ex-
pensive and sensitive to the grid employed.4 None of these
computational methods, however, show explicitly the func-
tional relationship of vortex strength to parameters such as
wing leading-edge sweep, a, and chordwise location.
Prediction of Delta Wing Helmholtz’ vortex theorems ensure that the rate of change
Leading-Edge Vortex Circulation of the spanwise load distribution relates to the rate at which
vorticity is shed from a wing’s trailing edge (g = 2dG/dy).
and Lift-Curve Slope Thus, it may be analogous to assume that the rate of change
of the leading-edge thrust distribution relates to the rate at
which vorticity is shed from the leading edge, and conse-
Lance W. Traub* quently, into the leading-edge vortex. Using this analogy, com-
Texas A&M University, bined with the aforementioned expression of Purvis2 for the
College Station, Texas 77843-3141 leading-edge suction distribution, allows the derivation of an
expression to estimate the leading-edge vortex chordwise cir-
culation distribution. In this Note, an expression is derived to
Nomenclature estimate the strength of the leading-edge vortex of delta wings.
The vorticity shed from the wing leading edge is related to the
b = wing span
rate of change of the wing leading-edge thrust, yielding an
CT = leading-edge thrust coefŽ cient
expression for g, which can be integrated to yield vortex cir-
c = chord
culation. The form of this expression allows derivation of two
cs = sectional leading-edge suction coefŽ cient
equations for the potential constant or attached  ow lift – curve
ct = sectional leading-edge thrust coefŽ cient
slope of delta wings.
kp = potential constant
Hemsch and Luckring5 have derived an expression using a
q = dynamic pressure
Sychev similarity parameter to estimate the strength of the
S = wing area
leading-edge vortex at the wing trailing edge. Using this ex-
t = sectional leading-edge thrust
pression it is possible to estimate the strength of the vortex
U` = freestream velocity
along the wing. This does, however, require the assumption of
w = downwash velocity
conical  ow (i.e., G } x), and two empirical constants.
x = chordwise direction
y = spanwise direction Discussion of Method
Using the Kutta – Joukowski theorem to calculate the lead-
Received Nov. 17, 1996; revision received Feb. 14, 1997; accepted ing-edge thrust associated with the freestream parallel to the
for publication Feb. 14, 1997. Copyright Q 1997 by L. W. Traub.
Published by the American Institute of Aeronautics and Astronautics,
chord gives
Inc., with permission.
*Graduate Student, Aerospace Engineering Department. Associate dG
dt = r (U ` sin a 2 w) dy (1)
Member AIAA. dy
 J. AIRCRAFT, VOL. 34, NO. 3: ENGINEERING NOTES 451

where w is the downwash associated with the trailing vortex

system. Thus,

dG dt 1
=g= (2)
dy dy r (U ` sin a 2 w)

U ` sin a 2 w should effectively be the upwash distribution at

the wing’s leading edge. However, for the purposes of the
present study, it is not feasible to estimate the local variation
of downwash at the leading edge. The upwash distribution may
be estimated by adapting Purvis’2 approach to the approxi-
mation of U ` sin a 2 w by allowing it to be proportional to
the approximation from Ref. 2, i.e.,

U ` sin a 2 w = K(U ` C T /kp sin a ) (3)

where K is a constant of proportionality.

Noting that the leading-edge thrust is related to the leading- Fig. 1 Correlation of the potential lift-curve slope and wing apex
edge suction by cc t = cc s cos(L), gives angle for simple delta wings, allowing prediction of the potential
constant as a function of AR.
kp sin a cos L rU 2` dcc s
g= (4) would suggest that the strength of the vortex would increase
rKU ` C T 2 dy due to vorticity from the leading edge and then start to de-
crease as the slope of dcct /dy, and hence, the vorticity shed
The leading-edge suction may be estimated as2 changes sign. Euler6 computations of the circulation at the

F Î SD G
2
leading edge and vortex core of a 70-deg delta wing show an
SE0C T 2y 2y 2y essentially linear (or conical) increase in circulation, followed
cc s = 12 1 arcsin (5)
p cos L b b b by a reduction in circulation as the wingtip is approached.
Thus, although the trend of reducing leading-edge G is correct
as the tip is approached, the present method would have G L.E.
Substituting Eq. (5) into Eq. (4) yields tend to zero (with cct) at y = b/2. This is obviously incorrect

F Î SD G
and would suggest that the vortex does not persist downstream.
kp sin a SE0 U ` d 2y 2y
2
2y Note that this problem is not present in Eq. (9) due to the form
g= 12 1 arcsin (6)
K p 2 dy b b b of Eq. (5). A more general formulation for the vortex strength
that encompasses the correct physics is given by Eq. (10):

E
y
which upon differentiating becomes G L.E.
G v( y) = dy (10)

FÎ SD 2 0 y
U ` Skp E 0 sin a 2 2y
g = 1 2
2p K b b This expression may be interpreted as summing the sectional
leading-edge circulation to yield the total vortex circulation.

2
8y 2
b 3 Ï1 2 (2y/b)2
1
2
b Ï1 2 (2y/b)2
G (7)
Equation (10) tends to a constant G v( y) as the leading-edge
thrust, and hence, circulation tends to 0.
Equation (9) shows that for a given spanwise location the
vortex circulation is essentially proportional to the potential
Equation (7) shows that under the assumptions of this anal- constant multiplied by a function of a. Using Hemsch’s5 ex-
ysis, g is proportional to sin(a ), or the velocity component pression for vortex circulation at the wing trailing edge, com-
normal to the wing as well as the potential lift – curve slope bined with Vissers’7 experimental constant, allows determina-
kp. The wing leading-edge sweep does not enter directly into tion of a simple estimate for the attached  ow lift – curve slope
the expression but is effectively included through k p. of delta wings. Hemsch’s expression may be written as
To calculate the strength of the circulation at the leading
edge at a speciŽ c y, Eq. (7) may be integrated using G/U ` c = 4.63 tan0.8« tan1.2a cos a (11)

E Noting that the present analysis shows that (G/U ` c)[1/kp f

y

G L.E.( y) = g dy (8) ( a )] should be a constant for a given spanwise location, and

0 for consistency using the formulation from Ref. 5 for the de-
pendence of G on a, i.e., f (a ) = tan1.2a cos a, gives
yielding, after substituting for S, and E 0 = 1.106p /b
4.63 tan0.8« tan1.2a cos a

F Î SD G
2 = const
1.106cU ` kp sin a 2y 2y 2y kp tan1.2a cos a
G L.E.( y) = 12 1 arcsin
4K b b b
or
(9)
tan0.8«/k p = const (12)
The constant K will be evaluated later using experimental
1
data. For the present method to have more general application, Using Polhamus’ results for kp of delta wings to see if Eq.
i.e., computational prediction of the cct distribution, it is nec- (12) does indeed correlate leading-edge sweep and k p (Fig. 1),
essary to clarify some of its limitations. Panel method com- shows that the data essentially collapse to a constant value of
putation of cc t will show this value tending to zero as the 0.25. This yields two simple expressions for the lift – curve
wingtip is approached [this characteristic is not present in Eq. slope of simple delta wings:
(5)]. Thus, the thrust distribution increases from zero at the
root, to a maximum near the wingtip, and then to zero. This kp = 4 tan0.8« = AR/tan0.2« (13)
452 J. AIRCRAFT, VOL. 34, NO. 3: ENGINEERING NOTES

Concluding Remarks
A method is presented that predicts the leading-edge vortic-
ity and vortex circulation for delta wings. The method relates
the strength of the vorticity shed to the rate of change of lead-
ing-edge thrust. Circulation was calculated by both integrating
the vorticity distribution as well as integrating the sectional
circulation distribution. The study suggests that the leading-
edge vortex strength is proportional to the normal component
of the freestream and the wing lift – curve slope, with sweep
effects effectively being included in the lift – curve slope. The
dependence of vortex strength on the potential constant and
angle of attack also allows determination of two simple ex-
pressions for the attached  ow lift – curve slope of delta wings.

Acknowledgment
The author would like to thank Murray Tobak for his helpful
Fig. 2 Comparison of theory and experiment for chordwise var- comments and suggestions.
iation of circulation, L = 75 deg, a = 20 deg.
References
1
Polhamus, E. C., ‘‘A Concept of the Vortex Lift of Sharp-Edge
Delta Wings Based on a Leading-Edge Suction Analogy,’’ NASA TN
D-3767, Oct. 1966.
2
Purvis, J. W., ‘‘Analytical Prediction of Vortex Lift,’’ Journal of
Aircraft, Vol. 18, No. 4, 1981, pp. 225 – 230.
3
Katz, J., and Plotkin, A., Low Speed Aerodynamics, from Wing
Theory to Panel Methods, 1st ed., McGraw – Hill, New York, 1991,
pp. 558 – 562.
4
Kumar, A., ‘‘Accurate Development of Leading Edge Vortex Using
an Embedded Conical Grid,’’ AIAA Journal, Vol. 34, No. 10, 1996,
pp. 2038 – 2046.
5
Hemsch, M. J., and Luckring, J. M., ‘‘Connection Between Lead-
ing Edge Sweep, Vortex Lift, and Vortex Strength for Delta Wings,’’
Journal of Aircraft, Vol. 27, No. 5, 1990, pp. 473 – 475.
6
Longo, J. M. A., ‘‘Compressible Inviscid Vortex Flow of a Sharp
Edge Delta Wing,’’ AIAA Journal, Vol. 33, No. 4, 1995, pp. 680 – 687.
7
Visser, K. D., and Nelson, R. C., ‘‘Measurements of Circulation
and Vorticity in the Leading-Edge Vortex of a Delta Wing,’’ AIAA
Journal, Vol. 31, No. 1, 1993, pp. 104 – 111.

Fig. 3 Comparison of theory and experiment for variation of

circulation with sweep and incidence, x /c = 0.5.

The similarity of the form of Eq. (13) to that for the AR of Impact of Initial Conditions
a delta (AR = 4 tan «) is interesting. Figure 1 also shows
comparisons of k p calculated using Eq. (13) and data from Ref.
on Vortex Breakdown
1; it may be seen that Eq. (13) demonstrates excellent accuracy on Pitching Delta Wings
for AR < 2, as would be expected. Also included in the Ž gure
is the result for the lift – curve slope of slender deltas as per
slender wing theory, i.e., k p = ( p/2)AR. Lars E. Ericsson*
To estimate the constant K for delta wings, comparisons of Mt. View, California 94040
Eqs. (9) and (10) with experimental results of Visser and Nel-
son7 are shown in Figs. 2 and 3. The circulation values pre-
sented are nondimensionalized by the freestream and wing root Introduction
chord. Setting K = 0.5 yields the correlations shown. Figure 2
shows that the present method displays very good correlation
with the experimentally determined chordwise vortex circula-
W HEN comparing the experimental results1 at v̄ = 0.10
for a 52-deg delta wing (Fig. 1) with those for v̄ =
0.107 (Fig. 2) for a 65-deg delta wing,2 the following question
tion distribution. Notice that over the forward part of the wing,
arises. Why can the upstroke characteristics in Fig. 1 hardly
Eqs. (9) and (10) are in good agreement. As the trailing edge
get downstream of the static characteristics, especially consid-
is approached, Eq. (10) tends to predict higher circulation val-
ering that even during the initial portion of the downstroke the
ues, as would be expected noting its formulation. Although not
vortex breakdown in Fig. 2 prefers to stay on the downstream
directly modeling the physics, Eqs. (9) and (10) do predict the
side of the static characteristics? These vortex breakdown char-
correct trends,6 i.e., over the forward part of the wing the lead-
acteristics2 were analyzed in Ref. 3 by Ž rst considering the
ing-edge circulation and vortex circulation are similar; how-
characteristics for purely rampwise changes of the angle of
ever, as the trailing edge is approached the vortex circulation
increases beyond that of the leading-edge circulation. Figure
3 shows that the method closely predicts the trends of sweep Presented as Paper 96-3405 at the AIAA Atmospheric Flight Me-
chanics Conference, San Diego, CA, July 29 – 31, 1996; received Nov.
and a on G/U ` c. The experimental circulation shown is the 3, 1996; revision received Feb. 18, 1997; accepted for publication
maximum measured at the x /c = 0.5 location. The predicted March 5, 1997. Copyright Q 1997 by L. E. Ericsson. Published by
results in Fig. 3 were calculated by evaluating Eq. (9) at b/4 the American Institute of Aeronautics and Astronautics, Inc., with per-
(i.e., x /c = 0.5), and integrating Eq. (10) spanwise from the mission.
wing root to b/4. *Engineering Consultant. Fellow AIAA.

View publication stats