JOURNAL OF AIRCRAFT

Vol. 32, No. 2, March-April 1995

Finite State Induced Flow Models

Part I: Two-Dimensional Thin Airfoil
David A. Peters*
Washington University, St. Louis, Missouri 63130
Swaminathan Karunamoorthyt
St. Louis University, Cahokia, Illinois 62206
and
Wen-Ming Cao$
Washington University, St. Louis, Missouri 63130

A new finite state aerodynamic theory is presented for incompressible, two-dimensional flow around thin
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

airfoils. The theory is derived directly from potential flow theory with no assumptions on the time history of
airfoil motions. The aerodynamic states are the coefficients of a set of induced-flow expansions. As a result, the
finite state equations are hierarchical in nature and have closed-form coefficients. Therefore, the model can be
taken to as many states as are dictated by the spatial texture and frequency range of interest with no intermediate
numerical analysis. The set of first-order state equations is easily coupled with structure and control equations
and can be exercised in the frequency or Laplace domain as well as in the time domain. Comparisons are given
with Theodorsen theory, Wagner theory, and other methods. Excellent results are found with only a few states.

Nomenclature v = y component of induced velocity, normalized

[A] = matrix of acceleration coefficients, Eq. (35) on V
[B] = eigenvectors of [A] W(t) = Wagner function
b = semichord, m w = total induced downwash, -v
bn = coefficients of expansion, Appendix C Wj = coefficients of w expansion, Eq. (22)
C(k) = Theodorsen function JT, y = Cartesian coordinates normalized on b
c,, = vector of length Af, Eq. (31) Y(Z) = weighting function
£>„„, = matrix, Eq. (30) Z = wake variable, e ~ ^
d,, = vector of length N, Eq. (31) F = total bound vorticity divided by bV
f(x) = generic function /(*) = g(rj) F = normalized vorticity, 172 TT
(^K1?) =
generic function y = vorticity density divided by V
/ = identity matrix yh = bound vorticity density
j — index taken as s, a, n jj = expansion coefficients of vorticity and velocity,
k = reduced frequency Eqs. (11) and (12)
L = lift per unit length divided by 27rpbV2 yw = wake vorticity density
M = pitching moment about midchord over A = elliptical determinant, Appendix A
2irpb2V2 V = Laplace operator
A/, /4 = normalized moment about quarterchord 8(t) = impulse function
m = index 1, . . . , N e = residual error, Eq. (43)
N = number of inflow states A = induced flow due to shed vorticity divided by V
n — index 1, . . . , N A, = expansion terms for A, Eq. (17)
Pf) — pressure differential on airfoil, nondimensional Ay- = complex representation of A,
on pV2 f = streamwise direction
Q[f] = functional, Eq. (18) p = density of air, kg/m3
s = Laplace variable TJ = pressure expansion coefficients, Eqs. (13) and
Tn = eigenvalues of [A] (14)
t = reduced time, normalized on semichord and 4>y = acceleration potentials, Appendix A
freestream ^ = velocity potentials, Appendix A
u = x component of induced velocity, normalized
on V Superscripts
V = freestream velocity, m/s A = from acceleration term
T = transpose
V = from velocity term
= - d/dy
Received Sept. 13, 1993; revision received May 23, 1994; accepted ^ = d/dt
for publication June 17, 1994. Copyright © 1994 by the American
Institute of Aeronautics and Astronautics, Inc. All rights reserved.
= evaluated on upper surface of airfoil
:!:
Professor and Director, Center for Computational Mechanics,
Campus Box 1129. Fellow AIAA. Subscripts
t Associate Professor, Department of Aerospace Engineering, Parks a = asymmetric potential
College. Member AIAA. L = lower surface
^Graduate Research Assistant, Department of Mechanical Engi- 5 = symmetric potential
neering, Campus Box 1185. U = upper surface
313
 314 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL

Introduction of Refs. 13 and 14 attacked the problem for two-dimensional

Background compressible and incompressible flows. Their work is in the
Laplace domain and builds on the work of Sears.5 It utilizes
U NSTEADY aerodynamic models that are useful for
aeroelastic analysis are usually of four general types.
First, there are /c-type aerodynamic theories in which the
numerical methods to locate the poles and to perform the
inverse Laplace transform by contour integration. 13 - 14 Work
motions, pressure, and induced flow undergo only simple har- further developed for the three-dimensional case (finite
monic motion, eikt. These are often used in the V-g method wing), 15 - 16 but that is the subject of Part II, the sequel to this
but, strictly-speaking, they are accurate only at the stability article.
boundary. A second category of model hasp-type (or Laplace- In rotorcraft aeroelasticity, the development of time-do-
domain) aerodynamics in which the aerodynamic variables main (rather than frequency-domain) unsteady aerody-
undergo exponentially growing (or decaying) harmonic mo- namics is particularly crucial due to the existence of periodic
tion. This type of model is utilized in eigenanalysis in which coefficients and nonlinear stall, which preclude superposi-
an iteration on p (i.e., on s) is performed for every eigenvalue tion of Fourier or Laplace solutions. The fundamental fre-
of interest. A third type of aerodynamic model is indicial, in quency-domain result was derived by Loewy. 17 It is an ex-
which a Green's function is utilized with a convolution integral tension of the Theodorsen theory and assumes layers of
to give arbitrary motion response. This is useful in time- vorticity below the airfoil to account for the returning wake.
marching. In principle, these three types of aerodynamics are However, its application is limited to linear problems of
equivalent and can be derived from one another through La- hover and climb. Dinyavari and Friedmann 18 used Fade
place and Fourier transforms. approximants of both Theodorsen and Loewy functions in
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

A fourth type of unsteady aerodynamics is the class of finite order to accommodate some unsteady aerodynamics into
state models. In a precise sense, two-dimensional unsteady periodic-coefficient Floquet stability analysis. Most dy-
aerodynamics has no finite state representation, but is an namic stall models for rotorcraft (e.g., Ref. 19) utilize the
infinite state process [e.g., s /,/(s) terms appear in the transfer Wagner function in a convolution integral to account for
function]. Nevertheless, useful finite state approximations have the time variation in induced flow due to the vorticity shed
been derived. There are several advantages of finite state from stall. One exception is the ONERA dynamic stall
models. First, finite state modeling allows one to cast the model, 20 in which a first-order differential equation pro-
aerodynamics in the same state-space context as the structural vides a single-state approximation to the Theodorsen func-
dynamics and controls. This allows the full complement of tion. (This is in contrast to the normal two-state approxi-
control theory and systems theory to be brought to bear on mations, Refs. 8 and 18.) In Ref. 21, the ONERA model
the problem of aero-servo-elastic control and design. Second, is generalized as a vorticity-based model, but still with a
the existence of explicit states eliminates the necessity to it- one-state Theodorsen model. In Ref. 22, the one-state model
erate on solutions (as in V-g and p-k methods). Instead, the is replaced by a hierarchical finite state inflow model for
entire solution can be obtained in one pass. Third, a state- rotors.
space model is flexible in that it can be exercised in the fre- Present Approach
quency domain, Laplace domain, or the time domain as de- In this article we offer a new type of finite state aerodynamic
sired. model. The model offers finite state equations for the induced
There are several types of finite state models. Vortex-lattice flowfield itself. These equations are derived directly from the
and computational fluid dynamics (CFD) methods can be cast potential flow equations (either velocity or acceleration po-
as finite state models with the number of states being the tential). Thus, no intermediate steps are invoked in which
order of the number of lattice nodes or CFD grid points. This, restrictions are placed on blade motions; and the theory is an
however, is usually such a large number of states that con- arbitrary-motion theory from the outset. In contrast to CFD
ventional control-theory applications are precluded. Instead, and vortex lattice methods, the states represent induced flow
most applications of finite state models to aeroelasticity have expansion fields rather than velocities at discrete nodes. As
utilized a relatively few approximate states. The disadvan- a result, the states are hierarchical, and the equation coeffi-
tages of this are that such states have no direct physical inter- cients are known in closed form. No numerical fitting of fre-
pretation and that they cannot be systematically improved in quency-response or indicial functions is needed.
a hierarchical manner. It would seem, therefore, useful to Furthermore, the induced-flow expansion implies that only
have a more general finite state model. a few states are needed, and the number of states can be
Previous Work chosen a priori based either on the texture required in the
In 1925, Wagner published the indicial function for the lift induced flowfield or on the frequency range of interest. The
response of a two-dimensional, flat-plate airfoil in incom- resultant equations are easily coupled with structural or con-
pressible flow. 1 In 1935, Theodorsen presented the lift fre- trol equations and can be exercised in the frequency-domain,
quency response for the same conditions.2 Garrick then showed Laplace domain, or time domain.
that the two were related (and mutually consistent) by means The above approach can be followed either in two- or three-
of Fourier transform. 3 The use of Laplace transform (/?-ver- dimensional flow. In this article, Part I, we consider two-
sion aerodynamics) was suggested by Jones,4 and applied to dimensional flow about a finite strip (i.e., airfoil) and apply
some problems by Sears.5 Jones4 obtained an approximate the nonpenetration condition to recover thin airfoil theory.
Laplace transform of the Wagner function, 6 but Jones7 was In the sequel, Part II, we will apply the approach to the three-
the first to generalize the Theodorsen function formally for dimensional flow about a disk (i.e., rotor). The disk is taken
p-type motions. At that time, the major mathematical concern as penetrable (an actuator disc) for application to rotorcraft
was whether or not this generalization was applicable for de- rather than circular wings. For flow near the blades, Part II
caying motion (negative real part of s).H This skepticism pre- utilizes the lift equation of Part I, but with the three-dimen-
vailed despite the arguments from analytic continuation.9 Thus, sional induced-flow model. Thus, this article forms a basis for
work on time domain unsteady aerodynamics was at a stand- rotor inflow as well as for two-dimensional aerodynamics.
still. Theoretical Background
Some 25 years later, interest was renewed in time-domain
methods. Hassig10 used rational functions in the Laplace do- Fluid Mechanics
main. Vepa 11 introduced the method of Fade approximants The airfoil lies on the segment y = 0, -1 < x < +1 in
to give a finite state representation of any aerodynamic fre- the nondimensional coordinate system of Appendix A. If we
quency-domain lift function, 11 as did Dowell. 12 The authors define u and v as the nondimensional x and y velocity com-
 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL 315

ponents, and if we assume the freestream velocity (normalized (with zero initial conditions). It follows that
to unity) is in the f direction, then the continuity and mo-
mentum equations are dPD dyh dyh
(lOa)
dx dx dt
du dv_ _
.., w
dx dy 0 = -^ + -^ (lOb)
dx dt
du
_ du _ ___
i _ — d<&
yH,(*,0 = -f(f - x + 1) (lOc)
dt d£ ~ dX

dv_ dv_ _ -d< Clearly, the velocity u is discontinuous across the vortex sheet;
dt df dy (1) but the normal velocity v is continuous. Thus, <J> V is discon-
tinuous, but d<&v/dx is continuous.
We then define special potential functions <&v and Equations for Bound Vorticity
In order to obtain a set of state-variable equations for the
d<$A -du d$>A -dv induced flow, we can treat the circulation-based equations in
dx dt dy dt the previous section with an expansion. To begin, we consider
the induced flow due only to yh. This induced flow is inde-
- du -dv
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

(2) pendent of the time history of yh. Thus, due to the asymmetry
dx dy about y = 0, we can write $>v (and y/,) as expansions in the
potential functions of Appendix A
The continuity and momentum equations are then fulfilled if
<&A and <&v satisfy

where j takes on the values s, a, 1, 2, 3, 4, . . . , and A implies

evaluation on the upper surface. The velocity field from this
bound vorticity then follows either from the Biot-Savart law, 23
The velocity field may be found from <&v by or from Eq. (4b):

-v = downwash = (12)
—— df (4)
dx dy ^
Thus, the y, represents both bound vorticity coefficients and
The only pressure discontinuity allowed is across the airfoil, downwash coefficients, which are exactly the Glauert velocity
and specification of the time history of this discontinuity de- expansions.
fines the pressures and velocities everywhere. Returning to the dynamic case, we see that these downwash
For the case of edgewise flow (f = x), we have coefficients can be related to the airfoil differential pressure
by Eq. (8a). Thus, if we expand the pressure in a series similar
to that for vorticity
= I
J - dx
df,
-/: -IT* (5)

dt then we obtain from Eq. (8a)

v
d f
- djc (6) (14)
dt J-

It follows that there is a strip of concentrated vorticity y that However, from Appendix A we see that the ^ can be uniquely
exists on y = 0, - 1 < x < o>. On the airfoil ( - ! < * < + ! ) expressed in terms of the <f>y. This allows a balancing of coef-
it is the bound vorticity y/,; and behind the airfoil (1 < x ficients in Eq. (14), since the <f>y are linearly independent. The
< 30) it is the wake vorticity y,v. From the definition of vorticity result of this balancing is a set of ordinary differential equa-
and the integration around a small loop across this strip, one tions that relate the velocity coefficients yy and the pressure
can relate y, <J>K, and PD (the pressure across the vorticity coefficients T,.
sheet):
% = TV, ya = ra (I5a)
y = <S>v - <&V, QL _ 4,^
P/) = (7)
ya ~ ^2 + r, = r, - 2f (isb)
From Eq. (6b) we then have the pressure-vorticity relation (\l2n}(yn^ - ya + l ) + yn = rn - (2ln)T
n = 2, 3, 4, . . . (15c)
= 7,, . f (8a) where f is the normalized total bound vorticity
at J - i

0 = yn. + f + - J yn. d (8b) f ^ I727T = ys + h, (16)

Wake Vorticity
* = j yhdx = - J + | yn. (9) The next step in the derivation is to find equations for the
i
induced flow due to. shed wake vorticity A. From Ref. 23, we
 316 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL

see that this component of downwash can be expanded (on It is important to note in Eq. (23) that w (the total down-
the airfoil) in a similar fashion as w, Eq. (12) wash) is completely determined by the nonpenetration bound-
ary condition. Thus, for small airfoil deformations y(x, t ) ,
A = y (17) -1 < x < +1

where A,, = A () of Ref. 21 and A s is not used since ^.' = 0 on dy dy

— + — (24)
- 1 < x < + 1 . Reference 23 provides formulas for Ay in terms dx dt
of the wake vorticity. For simplicity, we write these as func-
tionals: Therefore, the only unknown in the entire airfoil loading (ry)
is A,,. Of particular interest are the normalized lift L, the
normalized pitching moment about midchord M, and the mo-
~ = -- f ylvg(77) dry ment about the quarterchord M1/4:
77 J 1 77 Jo

f • em M = -k - ir
+ ir,

M1/4 =M - i (25)

r e-"" i r i From Eqs. (23) and the Kutta condition (rv = ra = ys — y,,),
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

= Q r~T7^ = Q — we then can write

Lsmh(r/)J L "
L = wa + ?Wi - \a + i(vi>,, (26a)
We note that the above formulas imply that a differential
equation for this functional can be obtained based on Eq. M = i(ivfl - \a) - iw2 ~ ^( (26b)
(lOc) and integration by parts
(26c)

dc Historically, w(l + iw, has been called the quasisteady lift,

w
« + ?wi ~ ^« tne circulatory lift, and i(vv rt - ivv2) the
noncirculatory lift. Since M1/4 (and all higher pressure coef-
am = - ficients r;i) are completely determined by airfoil motion, we
concentrate on L and A,, for the remainder of this article. The
power of Eq. (26a) for L in terms of Art is that the AA7 solution
does not need to have the fidelity necessary to meet explicit
boundary conditions. We need only an accurate A,,.
dg/dr? 1
= 2fg(0) (19) Equation (21) gives relationships for the A,,. Unfortunately,
A,, cannot be determined from Eq. (21) alone. One additional
relation is needed that relates A,, to the A,,. From the operator
When Eq. (19) is applied to the A, functional in Eq. (18), notation in Eq. (18) and the shorthand notation Z = e~^,
one must insure that g(0) = /(I) is finite. Therefore, we use we see that if one could approximate unity on 0 < 17 < o° by
A,,., - A / } 4 1 as the left side of Eq. (8), which gives a #(17)
of the form:
bnZ" 0< Z < 1 (27)
g(ri) = [ e ~ ( n 1)TJ
- e- ( / l + 1)7J]/sinh r/ = 2e~fjr> (20)

The resultant differential equations are then one could approximate A w by

A,, - U 2 + A, = 2f i N

A,, (28)
(l/2n)(Xn_l - A,I + I ) + A /f = (2ln)T n = 2, 3, 4 . . . (21)

The similarity with the y-} equations [Eqs. (15)], is striking. Although Eq. (27) can never hold at Z = 0, the approximation
must converge only on the open interval (excluding Z = 0 at
Boundary Conditions 17 —» oo). Thus, unity can be represented as a power series.
It is especially interesting to use Eqs. (15) and (21) to form Closed-form results for the bn are given in Appendix C. The
differential equations for the total induced flow w (downwash result is a set of finite state equations for the A/7 and (more
due to bound plus wake vorticity). If we expand w as importantly) for A,,. The equations are comprised of Eqs. (21)
and (28) with
s
J ^
i = ya 2-7, = (29)
Wj = y, + A,, / ± s (22)
These, along with Eqs. (26), form a finite state theory of
then, by addition of Eq. (15) and (21), we have unsteady aerodynamics in which the states are the Glauert
inflow distributions.

Application
Matrix Form
(\Kn)(*n_i -wn+l) + wn = rn (23) In order to verify the finite state aerodynamic model given
by Eqs. (21), (26a-26c), (28), and (29), we apply it to the
These equations can also be obtained from a direct application cases of the Theodorsen and Wagner functions. To facilitate
of the acceleration potential, Appendix B and Ref. 24. this application, the theory is put into matrix form. First, we
 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL 317

define the following matrices (in addition to the expansion of the eigenvalues and eigenvectors of [A], Tn, and B. (Tn
coefficients £„, Appendix C): are time constants of the flow.)
D = \n (n = m + 1) [B]-*[A][B] = [Tn]
Dnm = -±n (n = m - 1) {A(0} = [B][e-'
DIIIH = 0 n +m ± 1 (30) = (B][e-"T»/Ta][B]-*{c} (41)

4, = i n = 1 The limiting cases of the Theodorsen and Wagner functions

are determined by the bn. An interesting limit is the slope of
dn = 0 AZ ^ 1 the imaginary part of C(k) as k approaches zero. Theoreti-
cally, as k approaches zero the slope of C(k) goes to infinity
cn = 2/n (31) like ///(/:). Interestingly, one can show that the finite state
model gives this slope as
Then we rewrite Eq. (21) as
lim (42)
[D + dbr}{\} + {A} = {c}f (32) d(ik)
where Appendix C shows that this sum is an approximation to
///(O), and approaches — o° as A^ increases.
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

Another interesting limiting case is the limit of C(k) as k

approaches °o, and the limit of W(t) as t approaches zero.
f = (w(l + iw,) - (A,, + U,) Theoretically, these should be equal to each other and equal
L = w(l + 3H>, - A,, + i(n>,, - 3H>2) (33) to 0.5. Although the algebra is a little involved, one can show
that: 1) if 2 bn = I and 2 nbn = 0 (as in the binomial ex-
Alternately, we may write pansion for bn, Appendix C), then C(k) and W(t) will equal
exactly 0.5 in the finite state model; and 2) if 2 bn = 1, then
[A]{\] + {A} = {c}(*tl + iiv,) (34) C(k) and W(t) will approach 0.5 as A^ becomes large. For
example, for W odd
where
lim C(k) =
[/I] - [D + d/?r + c dT + id> r ] (35)

Special Cases £ l\ (43)

= 2, nbn / ( + 1)
The Theodorsen function is defined as the ratio of circu- i, = 1.2,3. / L

latory lift to quasisteady lift (in the complex domain) for

Since the 6,, alternate signs as (-1)" + 1 , W(Q) begins at 0.6
motion of the form eikt. Thus f o r A ^ = l,bi = 1, and quickly approaches 0.5 as Wincreases.
Lastly, one can show that the sum of all the eigenvalues of
C(k) = ^ (36) [A] equals the trace [A]. Thus,
w(l + Aw, T + A,, + ?)

For A,, = \neikt, f= eikl, we have

Tn = 1 k bjn (44)

which also approaches ^ as N becomes large.

Results
{A} = [(D + dbT)ik + I]~l{c}ik We would now like to examine the convergence of the finite
state model for various choices of bn. Figure 1 compares the
A,, - 3^A,,, U, = d r A (37) Re[C(/c)], Im[C(£)], and W(t) with the finite state approxi-
mations for N = 4. In order to see the performance over the
Equation (37) defines the Theodorsen function for any N. It entire range of k or r, the Theodorsen function is plotted vs
is interesting to contrast the simplicity of form of Eq. (37) kl(l + k), and the Wagner function is plotted vs r/(2?r + t).
with the complexity of the Bessel functions in Theodorsen Exact solutions are obtained numerically. One can see that
theory. the augmented least-squares series for bn (see Appendix C
The Wagner function is defined as the circulatory lift w(l + for closed-form) gives a reasonable fit. The other least-squares
IH>, - A,,, for the special case wtl = u ( t ) , step function, and methods (not shown in the figure) give a similar result. The
H>! = 0. From Eq. (26), this reduces to binomial expansion, which is designed to give least error at
large k (and small /), does so, but to the detriment of small
k (or large t) response. Figure 2 shows a similar comparison
W(t) = I - A,,(0 = ' l - (38) for N = 8. The error is greatly reduced, although the binomial
expansion is converging slowly.
Figure 3 presents a quantitative measure of an error norm
where A,,(0 is the solution to Eqs. (34), i.e., defined as the rms error computed on 100 equally spaced
points on the k/(l + k) or t/(27T + r) plot. The error is
[A}{\} + {A} = {c}8(0, {A(0)} = (0) (39) normalized such that the N = 0 case, C(k) = W(t) = 1, has
an error of unity. We see that the binomial expansion con-
or verges at a rate of N~l, whereas the augmented least squares
converges as N~2. However, due to the large factorials in the
[A}{\} + {A} = {0}, {A(0)| = [A]-*{c] (40) augmented method, double precision arithmetic fails for W >
10 for C(k) and N > 8 for W(t). The minimum error for the
Equation (40) can be solved by time-marching or by utilization augmented least squares occurs at N = 8, and is 1% for C(k)
 318 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL

0.8 - -

ReCOOOJ -

0.6 - -

0.4 H———h
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

0.3 -r

-ImC(k)

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
C) T/(T+2jt) C) T/(T+2lC)
Fig. 1 a) Real part of C(k}, N = 4; b) imaginary part of C(k}, N = Fig. 2 a) Real part of C(k), N = 8; b) imaginary part of C(k), N =
4; and c) Wagner function, N = 4. 8; and c) Wagner function, N = 8.

and 1.3% for W(t). The binomial expansion (with smaller states), and with Fade approximants with two and three states,
factorials than the augmented least squares) begins to diverge respectively. Note that the Jones' method, which is based on
after N = 50 for C(k) and TV = 16 for W(t). (Recall that a fit of the Wagner function, has a lower error norm for W(t)
eigenvalue problems are more sensitive than are matrix in- than it does for C(/c). Conversely, the Fade results, based on
versions.) The minimum error is 0.7% for C(k) (N = 50) and a least-squares fit of C(/c), have a lower error norm for C(/c)
5% for W(t) (N = 16). Quadruple precision would extend than they do for W(t). It is at first puzzling to see that one
the accuracy of either method. needs four to six inflow states with the augmented method to
Table 1 provides a summary of these errors and compares obtain the same accuracies as can be obtained with only two
them with those from Jones' approximation (which has two to three states for Jones' and Fade methods. The reason for
 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL 319

Table 1 Error norms or Wagner functions). The method is hierarchical, and the
Methods Theodorsen Wagner states represent the classical Glauert induced flow coeffi-
cients. The resultant, closed-form equations for the inflow
Curve-fit and states give excellent correlation with Theodorsen and
Jones 2.3% 1.8% Wagner functions with four to nine states included. This finite
Fade, N = 2 2.1% 3.8%
Fade, N = 3 1.2% 3.4%
state type of aerodynamic analysis is very useful in aeroelas-
Augmented ticity because it can be used in the frequency domain, Laplace
N =4 3.5% 5.0% domain, or the time domain. Furthermore, the method is
N = 6 1.6% 1.8% more rigorous than other finite state methods in that it is
N = 8 1.0% 1.3% based on first-principle fluid mechanics rather than on curve-
Binomial fitting specialized response functions; and the method is more
N = 4 10.5% 18.3% computationally direct than methods based on extensions in
N = 8 5.6% 10.1% the complex plain that require extensive numerical analysis.
N = 16 2.8% 5.0% The weakness of the method is that convergence is not as
1.3% rapid as is theoretically possible. An appropriate change of
N = 50 0.7%
variable in the inflow expansion could cure this.

Appendix A: Potential Functions

with Elliptical Coordinates
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

Coordinates
We define an x, y coordinate system with positive x down-
stream and positive y in the direction of positive lift. All
0.1 -- lengths are nondimensional on 6, such that the airfoil is on
the line segment2 — l < ^ < + l , _ y = 0. Pressures are nor-
malized on pV , and time is normalized on the semichord
ERR over velocity. (It follows that velocities are normalized on V.)
The transformation from elliptical coordinates to Cartesian
coordinates is given by
0.01 --

x = cosh 77 cos </>, _y = sinh 77 sin </>

0 < </> < 277, 1<77<00 (Al)

0.001
Thus, upstream </> = TT, x = -cosh 17, y = 0; downstream
<p = 0, x = cosh 77, y = 0; and on the airfoil 77 = 0, ;c =
a)
cos </>, y = 0. The transformation of derivatives is given by

dx J_ sinh 77 cos </> —cosh 77 sin cf>

(A2)
A ^cosh 77 sin (/> sinh 77 cos <
dy

where
A = sinh 2 77 + sin 2 </> = cosh277 - cos2^
ERR
= sinh2?] cos2(/> + cosh2r/ sin2(/>
0.01 - - Laplace's Equation
Laplace's equation is

^ + ^ = ^—— +
T7l)=0 < A3 >
0.001

Through separation of variables, we obtain the fundamental

b) __________ solutions
Fig. 3 Error norm for a) V(Re C)2 + (Im C)2 and b)
1, (/>, 77, 77(/>

this phenomenon is found in a study of the eigenvalues of the e'lri

system. The accuracy of the finite state model is obtained
almost entirely from the first two to three eigenvalues (which
are real), even for N = 50. Thus, only a few states are re-
quired; but they are a combination of many of the Glauert When these solutions are used to represent pressure, the pres-
states. One might suspect, therefore, that a change of variable sure must vanish at infinity and have discontinuities only on
could improve the convergence of the finite state results. the airfoil. Thus, acceptable pressure functions are
Conclusions flri
<£„ = e sin(ft</>) n = 1, 2, 3, . . . (A4)
A finite state induced flow model has been developed from
first principles (as opposed to numerical fits of Theodorsen Due to the A2 in the denominator of Eq. (A3), we must also
 320 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL

consider derived potential functions that have singularities at If we expand this concept, we can develop differential equa-
A = 0 (leading and trailing edge of airfoil). We define sym- tions for Wj based on Eq. (3a):
metric and antisymmetric functions

2 =
= X ^// tsin 0 cosn
n= 1,3,5

(series converges 17 =t= 0) It follows from Eqs. (4) and (5) that wy = r/, and that

= 2j (~^)(*>n =
- [sin 0 cos (B2)
/ / = 2,4,C>

(series converges 77 =1= 0) (A5) The nature of the acceleration potential method is that the
pressure (or velocity) expansions should balance on the air-
We also define <£, = 4>t - 4 <£v, which yields zero net pressure foil. The <J>', ^' relationships in Appendix A show that Eq.
across the airfoil. (B2) is balanced when

Velocity Potentials
We also define a set of velocity potentials based on ,, = TH « = 2, 3, . . . (B3)
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

This is identical to the result from our thin-airfoil derivation,

(A6) Eq. (23). However, the above equations are not complete
since they omit the distributions ws and wa. The singularities
associated with the derived functions 4> v , <!>„, Ws imply that
TT — <f> = X 2^,, (series converges everywhere) the theory must be extended to include these singularities in
/i=l,3,5
order to obtain ws, w(l equations. For the special case of no-
shed vorticity (F = yv + iyi = 0-» wx + ±w{ = 0, Ay = 0),
- e" 7 7 sin </> = ~ 2¥,, the W's terms disappear; and wtl = r(n vvs = r, become the
/i = 2, 4, 6
two extra equations. These cause the velocity balance to be
(series converges everywhere) fulfilled everywhere in the flowfield. These, then, agree with
thin-airfoil theory (y, = vvy), Eq. (15a).
For the general case, the equation cannot be balanced
everywhere; and the far field must become a "natural" bound-
ary condition. The balance of l/sinh 3 ry leading-edge singular-
(n = 2, 3, 4, . ities gives
2 \n + 1 n - 1
(TV — w> v ) = —(TU — (B4)
(A7)
In other words, the classical thin-airfoil pressure distribu-
On the airfoil (TJ = 0) tion gives the classical inflow. (This is also equivalent to
A,, = -A s , which implies no leading-edge singularity due to
shed vorticity.) The balance of 1/sinh 17 singularities gives a
0) (A8) second equation

2Z
Gradients of Potentials
Let ( )' = — d( )/dy. This gives <$>,', as the normal acceler-
ation and Wn the normal velocity: - (w, + wrt) = 0 (B5)

<E>S' = sinh 17 cosh ^[(cosh2?/ + cos2</>)sin2c/> where Z ^ ^~ T J . At 17 = 0, Z = 1, this gives

-f (sin2<£ — sinh2r/)cos2(/)]/A3
<&', = sinh r\ cos </>[sinh2i7 cos2^ - 3 cosh2r/ sin2</)]/A3 wa) = 0 (B6)

O,', = He~ lll7 [cosh 17 sin c/> sin(i7(/>)

- sinh r/ cos </> cos(i7^>)]/A a divergent (although functionally correct) series. To manip-
(A9)
ulate this into a better-behaved series, we divide by Z and
^' = sinh 17 cos $/A integrate from 0 to 1. The result is

_ i I
1 + Z

-o (B7)

Appendix B: Acceleration Potential where ///(()) is approximated by — 2 b,Jn (Appendix C). This
In Eq. (11), <&v is expanded in a <f>y series for the static series still fails to converge at Z = 1, (1 = 2 - 2 + 2-2
case, in order to find the induced flow due to bound vorticity. + 2 - 2 . . .). However, when the diverging series of 2 is
 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL 321

replaced by a series of bn (Appendix C), we have a well- gularity at 77 = 0. Thus, for the least-square error from Eq.
behaved model: (C8) to be finite

(T,, - wa) = \^b» \(rn - nO - - K + *„) (B8) error =

2 „ \_ n j

As a special case, consider no inflow due to bound vorticity: one must have 2 bn = 1 in order to cancel the singularity.
Ta = 0, T,, = (2//i)T ¥ , Eqs. (15). The term ws + wa = TS + Thus, we also consider modifications of Eq. (C7) to meet this
ra = rv cancels every bnrn term yielding criterion. In particular, we define three possibilities: 1) mod-
ified least squares, bNreplaced by bN + (- 1)N; 2) augmented
least squares, bn taken from formula for N - 1 terms and
bnwn (B9) bN = (- l)N+1; and 3) least squares fit (q = - 1) with formal
constraint on E bn = 1.
This is the same as from the vorticity equations, Eq. (28).
Also note that the ws + wa coefficient, 2 bn/n, approaches Acknowledgments
30, -///(O), as N becomes large. This work was sponsored by the U.S. Army Aeroflight-
dynamics Directorate (ATCOM), NASA Grant NAG-2-728,
Appendix C: Shed-Wake Expansion Robert A. Ormiston, Technical Monitor.
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

Consider the approximate expansion

References
^z"' o< z < i (ci) 'Wagner, H., "Uber die Entstehung des dynamischen Auftriebs
von Tragflugeln," ZAMM, Bd. 5, Heft 1, Feb. 1925, pp. 17-35.
2
Theodorsen, T., "General Theory of Aerodynamic Instability and
If we divide by Z and integrate from 0 to 1, we have the Mechanism of Flutter, 11 NACA Rept. 496, May 1934, pp. 413-
433.
"Garrick, I. E., k t On Some Reciprocal Relations in the Theory of
(C2) Nonstationary Flows,11 NACA Rept. 629, 1938.
4
Jones, R. T., "Operational Treatment of the Nonuniform Lift
Theory to Airplane Dynamics, 11 NACA TN 667, March 1938, pp.
Evaluation at Z — 1 yields the following: 347-350.
5
Sears, W. R., "Operational Methods in the Theory of Airfoils in
Non-Uniform Motion, 11 Journal of the Franklin Institute, Vol. 230,
(C3) July 1940, pp. 95-111.
Mones, R. T., "The Unsteady Lift of a Wing of Finite Aspect
Ratio, 11 NACA Rept. 681, June 1939, pp. 31-38.
Thus, 2 bjn will approach oo as A^ is increased. 7
Jones, W. P., "Aerodynamic Forces on Wings in Non-Uniform
To apply a weighted least-squares calculation of bn, we Motion, 11 British Aeronautical Research Council, R & M 2117, Aug.
minimize /, where 1945.
<s
Bisplinghoff, R. L., Ashley, H., and Halfman, R. L., Aeroelas-
ticity, Addison-Wesley, Reading, MA, 1955.
J= dz (C4) ''Edwards, J. W., Breakwell, J. V., and Bryson, A. E., Jr., "Reply
by Authors to Vepa,11 Journal of Guidance and Control, Vol. 2, No.
5, 1989, pp. 447, 448.
Y(Z) is a weighting function. Different weights will result in '"Hassig, H. J., "An Approximate True Damping Solution of the
different inflow theories. For Y(Z) = Z* (Z = 1 is the most Flutter Equations by Determinant Iteration, 11 Journal of Aircraft,
important point), we obtain an expansion Vol. 8, No. 11, 1971, pp. 825-889.
"Vepa, R., "On the Use of Fade Approximants to Represent
Unsteady Aerodynamic Loads for Arbitrarily Small Motions of Wings,11
- 1 - (1 - Z) A (C5) AIAA Paper 76-17, Jan. 1976.
l2
Dowell, E. H., "A Simple Method for Converting Frequency
this further implies Domain Aerodynamics to the Time Domain, 11 NASA TM-81844,
Oct. 1980.
"Edwards, J. W., Ashley, H., and Breakwell, J. V., "Unsteady
bn = (-1)"- N\(N - n)\ ' nbn = 0 (C6) Aerodynamic Modeling for Arbitrary Motions,11 AIAA Paper 77-
451, March 1977.
l4
Edwards, J. W., Breakwell, J. V., and Bryson, A. E., Jr., "Ac-
These are the binomial expansion coefficients of (1 - Z)N. tive Flutter Control Using Generalized Unsteady Aerodynamic The-
For Y(Z) = Z(/, q = integer, we have a family of bn. The ory, 11 Journal of Guidance and Control, Vol. 1, Jan.-Feb. 1978, pp.
case q = - 1 (smallest possible integer value) gives a closed- 32-40.
15
form expression Matsushita, H., "Aerodynamic Transfer Functions of a Finite
Wing in Incompressible Flow,11 20th Aircraft Symposium, Fukuoka,
Japan, Nov. 1982.
b = (-p.-
; (C7) l6
Miyazawa, Y., and Washizu, K., "A Finite State Aerodynamic
^ Model for a Lifting Surface in Incompressible Flow,11 AIAA Journal,
Vol. 21, No. 2, 1983, pp. 163-171.
In this formulation, the 1, bn = 2 for TV odd, and ^ bn = 0 l7
Loewy, R. G., "ATwo-Dimensional Approximation to Unsteady
for W even. Aerodynamics in Rotary Wings,11 Journal of the Aeronautical Sci-
From Eq. (18) we see that the functional form of the A,,, ences, Vol. 24, No. 2, 1957, pp. 81-92.
ls
A,, relationship is Dinyavari, M. A. H., and Friedmann, P. P., "Unsteady Aero-
dynamics in Time and Frequency Domains for Finite Time Arbitrary
1 Motion of Rotary Wings in Hover and Forward Flight,11 AIAA Paper
(C8) 84-0988, May 1984.
sinh r] /I = i sinh 17 1M
Beddoes, T. S., "Representation of Airfoil Behavior, 11 Vertica,
Vol. 7, No. 2, 1983, pp. 183-197.
2()
This shows that the error in the r\ functional involves a sin- Tran, C. T., and Petot, D., "Semi-Empirical Model for Dynamic
 322 PETERS, KARUNAMOORTHY, AND CAO: TWO-DIMENSIONAL THIN AIRFOIL

Stall of Airfoils in View of the Application to the Calculation of for Use in Rotor Aeroelasticity and Control Analysis, 1 ' Mathematical
Response of a Helicopter Blade in Forward Flight," Sixth European and Computer Modelling, Vol. 19, Nos. 3/4, 1994, pp. 109-123.
23
Rotorcraft and Powered Lift Forum, Bristol, England, UK, Sept. Dowell, E. H., Curtiss, H. C., Jr., Scanlan, R. H., and Sisto,
1980. F., A Modern Course in Aeroelasticity, Sijthoff and Noordhoff, Al-
2l
Peters, D. A., 'Toward a Unified Lift Model for Use in Rotor phen aan den Rijn, The Netherlands, 1980.
Blade Stability Analyses,' 1 Journal of the American Helicopter So- 24
Cicala, P., "Le Azioni Aerodinamiche sui Profili di ala Oscillanti
ciety, Vol. 30, No. 3, 1985, pp. 32-42. in Prezenza di Corrente Uniforme, 11 Memoirs of the Royal Academy
-Peters, D. A., and Su, A., "An Integrated Airloads-Inflow Model of Science, Torino, Serie 2, Tomo 68, Parte I, July 7, 1935.
Downloaded by TEXAS A & M UNIVERSITY on March 21, 2013 | http://arc.aiaa.org | DOI: 10.2514/3.46718

Rotary Wing
Structural Dynamics and Aeroelasticity
Richard L Bielawa
This new text presents a comprehensive account of the controlling vibrations, dynamic test procedures, stabil-
fundamental concepts of structural dynamics and ity analysis, mechanical and aeromechanical instabili-
aeroelasticity for conventional rotary wing aircraft as well ties of rotors and rotor-pylon assemblies, unsteady
as for the newly emerging tilt-rotor and tilt-wing concepts. aerodynamics and flutter of rotors, and model testing.
Intended for use in graduate level courses and by prac- The text is further enhanced by the inclusion of prob-
ticing engineers, the volume covers all of the important lems in each chapter.
topics needed for the complete understanding of rotorcraft AIAA Education Series
structural dynamics and aeroelasticity, including: basic 1992, 584 pp, illus, ISBN 1-56347-031-4
analysis tools, rotating beams, gyroscopic phenomena, AIAA Members $54.95 Nonmembers $75.95
drive system dynamics, fuselage vibrations, methods for Order*: 31-4(830)

Place your order today! Call 1-800/682-AIAA

&ALAA Sales Tax: CA residents, 8.25%; DC, 6%. For shipping and handling add $4.75 for 1 -4 books (call
for rates for higher quantities). Orders under $100.00 must be prepaid. Foreign orders must be
American Institute of Aeronautics and Astronautics
prepaid and include a $20.00 postal surcharge. Please allow 4 weeks for delivery. Prices are
Publications Customer Service, 9 Jay Could Ct., P.O. Box 753, Waldorf, MD 20604 subject to change without notice. Returns will be accepted within 30 days. l\lon-U.S. residents
FAX 301/843-0159 Phone 1 -800/682-2422 8 a.m. - 5 p.m. Eastern are responsible for payment of any taxes required by their government.