a093 7 stevens inst of tech hoboken nj davidson … · november 1978-november 1980 approved for...
Post on 14-Mar-2019
212 Views
Preview:
TRANSCRIPT
A093 7 STEVENS INST OF TECH HOBOKEN NJ DAVIDSON LAB F/G 20/4A PARTIALLY CAVITATING HYDROFOIL IN A GUST.(U)NOV 80 A 5 PETERS T R GOODNAN, I P BRESLIN NOO0 -4-77-C-0122
UNCLASSIFIE ST-DL-O-2118 NL
,TEMIIIIIIIIIIImIIEIIEEEEIIIiIuIuuIini-E.' -
iiiii .,_ ,o "' U-III1 "-° IIIII1111112.2
1I1 10 I 2II 0111IO.25 TES 6 HAT
MICROCOPY RESOLUTION TESTCAt
~'R-2118
DAVIDSONLABORATORY
REPORT SIT-DL-2118
November 1980
A PARTIALLY CAVITATING HYDROFOIL
IN A GUST
.S. Pe ters
T. R. 'oodman
J .. P. Brealin
Final ReportNovember 1978-November 1980
APPROVED FOR PUBLIC RELEASE;DISTRIBUTION UNLIMITED
Prepared forOffice of Naval Research800 N. Quincy StreetArlington, Virginia 22217
2 0
-,
STEVENS INSTITUTE OF TECHNOLOGYDAVIDSON LABORATORYCASTLE POINT STATIONHOBOKEN, NEW JERSEY
NOTICE
IN 1977 A REPORT ENTITLED "THEORY OF A PARTIALLY ANDINTERMITTENTLY CAVITATING HYDROFOIL IN A PERIODICTRAVELING GUST" BY JOHN P. BRESLIN AND THEODORE R.GCCDK'\N !.%S PUBLISHED, AND ONE OR MORE COPIES WERESENT TO YOU.
SUBSEQUENTLY, CERTAIN ERRORS IN THAT REPORT CAME TOLIGHT AND, AS A CONSEQUENCE, THE ENCLOSED REPORT WASPREPARED. THIS REPORT IS INTENDED T PER THEEARLIER ONE. THEREFORE, YOU ARE ADV TO DESTROYALL COPIES OF THE EARLIER REPORT AND SUBSTITUTE THEENCLOSED REPORT IN ITS STEAD.
I
IICIA eIFIESECURITY CLASSIFICATION OF THIS PAGE (Whon Data Entered)
REPORT DOCUMENTATION PAGEREDISRCOS7 BEFORE COMPLETING FORM
SIT-DL-80-2118
A PARTIALLY CAVITTN YRFI NG'T, 5Nv 08-0Nv 8O
f) A.S.,,keters' 12R./Goodmana JP eli )N0IZ:272
Task Order NR-062-568
I. PERFRINGORN ZAT IO AEADADES.PORMEEMENT. ROECT thASKpr
Apprveds InorituteiofTechnlog 61153N; R023-01-01
.OiceIuTO STTEEN (ova tes er ~c eF euid Dyin lcs 20 B frent frov 0ot
800 SPLN TAR Qu n OSettO PAE
ydrodynamc V- Caviatio ii Unsead plpw +Appdrofo)&s
tioNG otENial as thOfndmntliepnent vaia5. IT iLS show that i
th shap DSRBTOSTTN(of ths Iegoraleutosaepeene)hog hc h
lengSTIBth STMNof the t andtac tensted y prBok20tdfesre dir ibuepotiontefi a
IIII.47 SUPLMNTR N ONV5 SSLT ~ YcASPIAINO wS ~l(e lq
beIISRC deteine.n revese e Ofinecssr arIeniyb hloc nonleradr iefrial
tUNCLAS I F I FnL.4_J4ITY CLASSIFICATION OF THIS PAGE(When Dae Entered)
ABSTRACT (Continued)
numerical procedures to solve7 no attempt is made to solve them 0 her.0 Itis shown in an appendix that the earliest possible appearance of a non-vanishing second time derivative of the cross-sectional area of the cavityis to the order of the square ofthe reciprocal of the aspect ratio and thatto this order the pressures at distances large in comparison with the chordof the foil (but not the span) grow logarithmically. It may be inferredfrom this result that the pressures in the far field can be reduced by in-creasing the aspect ratio
I
~i
UNCLASSIFIEDSECURITY CLASSIICATION OF THIS PAGCE~lfb Date KnteRI0
STEVENS INSTITUTE OF TECHNOLOGYIDAVIDSON LABORATORYCASTLE POINT STATIONH4OBOKEN. NEW JERSEY
REPORT SIT-DL-80-21 18
November 1980
A PARTIALLY CA VITA TING HYDROFOIL
* IN A GUST
by
A. S. Peters
T. R. Goodman
J. P. Breslin
Performed Under
Office of Naval Research Contract Nooo14-77-C-0122Task Order NR-062-568
(DL Project 013)
APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED
Approved:.,
P.Se~n, Drco
R-2118
ABSTRACT
An analysis is made of a hydrofoil with partial cavitation in a
traveling gust. The problem is formulated in the time domain with
pressure (acceleration potential) as the fundamental dependent variable.
It is shown that, in the limit, as the aspect ratio of the foil becomes
large, the second time derivative of the cross-sectional area of the cavity
vanishes regardless of the shape of the gust. Integral equations are
presented through which the length of the cavity and the unsteady pressure
distribution on the foil can be determined. These equations are highly
nonlinear and require formidable numerical procedures to solve; no attempt
is made to solve them here. It is shown in an appendix that the earliest
possible appearance of a nonvanishing second time derivative of the
cross-sectional area of the cavity is to the order of the square of the
reciprocal of the aspect ratio and that to this order the pressures at
distances large in comparison with the chord of the foil (but not the
span) grow logarithmically. It may be inferred from this result that the
pressures in the far field can be reduced by increasing the aspect ratio
of the foil.
a
R-21 lP
NOMENCLATURE
A nondimensional cross-sectional area of the cavity
B denotes wetted portion of foil
C denotes cavity
c nondimensional local chord of the cavity
c C/E
F complex function whose real part is P
f in the appendix t/c (the same as h in the text)
h ZIci V/-T
1 unit vector in free stream direction
i unit vector along the chord of the foil
unit vector normal to free stream direction
± unit vector normal to
K a constant0K a constant
nondimensional local chord of foil
n ordinate of cavity surface
P acceleration potential
P nondimensional cavity pressurec
p nondimensional pressure
S semi-span
s nondimensional distance along span
t time
t nondimensional time
u perturbation velocity due to foil
U free stream velocity
u nondimensional perturbation velocity
v vector sum of free stream velocity and gust velocity
v unsteady velocity of fluid
v gust velocitygv nondimensional gust velocitygW denotes wake
x1 distance in free stream direction
x distance along chord of foil
x nondimensional distance along chord of foil
y distance normal to chord of foil
Jk ........ . ... .. Il II....... .. ... I.... II...... -........ ... ...... 1III1 .... . |'Il .... .. . . ...... 3' "
R-21 18
y nondimensional distance normal to chord of foil
z x + iy
a angle of attack
-y x/E
C small quantity inversely proportional to aspect ratio
mapped complex plane = + in
in imaginary coordinate in mapped plane
K 2 unknown functions of time
A dummy spacelike variable
nondimensional difference between ambient pressure and cavity pressure
V CA
nondimensional dummy chordwise coordinate in first section;also in second section real coordinate in mapped plane
p fluid density
o nondimensional dummy spanwise coordinate; also in third section
denotes dummy time
T t/e ; also in third section denotes t-x
* transformed acceleration potential
* nondimensional velocity potential
Wy/
( ) except where specifically defined, denotes first order in c;
similarly for( )2' )'39 ( )4
iii
R-2118
INTRODUCTION
Ship propeller blades are known to cavitate intermittently during
their traverse of regions of large hull wake fractions or low inflow
velocity. Such cavities are responsible for large pressure changes on
the hull with concurrent intense hull and sub-structural vibration. It
is known that the vibratory forces generated by cavitating propellers
are much larger than the forces generated by noncavitating propellers.
The reason for this is that the cavity behaves in the far field like a
source whose strength is proportional to the second time derivative of
the cavity volumel. It is also significant that cavity-induced forces
arising at the first few integer multiples of blade frequency are of the
same order as those at blade frequency. This is in sharp distinction to
the excitations induced by non-cavitating blade loadings which are in-
variably small at all higher multiples of blade frequency, as has been
pointed out by Breslin 1 . Thus, it is clear that the dynamics of unsteady
cavities on propeller blades must be understood and reduced to a calcula-
tion yielding their time-varying geometry before the pressure field (which
induces the hull vibration) can be predicted.
The complete description of intermittent cavitation on propeller
blades requires a three-dimensional representation accounting for the mutual
interaction of cavitating and non-cavitating blade elements as well as the
mutual interaction of the blades. This is clearly a formidable task. As
a first step, we consider the blades to be hydrofoils acting independently
and we examine the limiting case of large aspect ratio. The problem is
thereby reduced to a single two-dimensional hydrofoil subjected to a
periodic gust pattern which is spatially stationary in hull-fixed coordi-
nates and, therefore, travels with the negative speed of the section in
foil-fixed coordinates.
In the past, there was, for an extensive period of time, a continuing
controversy about the possibility of solving unsteady cavity problems in
two dimensions because the existence of time-varying cavities of finite
length implies a time-varying source flow whose potential and non-convective
pressure fields do not vanish at large distance, but grow as log r. Birkhoff2 ,
Leehey 3 and others sought to invoke compressibility as a means for resolving
R-2118
this paradox. Benjamin 4 put the question to rest by pointing out that
such two-dimensional problems are artifacts of attempts to reduce three-
dimensional problems to manageable problems in two dimensions and that
they really represent the inner behavior of three-dimensional flows with
potentials and pressures vanishing as a three-dimensional source, i.e.,
as (r2 + s2)-l/2, where s is the spanwise coordinate. Thus, the logarithmic
character of the mathematical representation of two-dimensional time-varying
cavity flows must be included in order to match the inner expansion of
the outer field. Benjamin went on to assert that the rate of expansion
of the cavity section should be regarded as a parameter of the inner field
and could only be determined by resorting to the three-dimensional outer
field. With these concepts in mind, therefore, the problem will be set up
as a three-dimensional one and the aspect ratio will then be taken to be
large.
Several attempts to solve unsteady, partially cavitating hydrofoil
problems have been made, but all have fallen short of the goal set here,
principally because they considered only small disturbances on an estab-
lished steady cavity. In contrast, we must address flow patterns with
cavities whose size can grow and diminish over several orders of magni-
tude during one cycle. In one of the earliest assaults on the non-steadily
cavitating hydrofoil, Steinberg and Karp 6 studied the small disturbance
case, but they imposed a dubious condition, requiring that the velocity be
continuous across the foil-cavity wake while ignoring the cavity closure
condition altogether. Wang and Wu7 studied only small perturbations on an
established steady cavity. Unruh and Bass 8 presume that the cavity leading
edge can be assigned arbitrarily downstream of the foil leading edge, whereas,
many observations in water tunnels and a vacuum tank reveal that, on propel-
ler blades, cavitation almost invariably initiates at the leading edge.
Furthermore, they also imposed the, by now familiar, restriction to per-
turbations of an already existing steady cavity.
The three-dimensional problem posed by intermittent cavitation on
ship propellers has been addressed by Johnsson and S~ntvedt9 , NoordzijlOand,
more recently, by Kaplan, Bentson and Breslin 11. All of these have employed
quasi-steady strip theory utilizing the steady, partially cavitating, sec-
tion theory of Guerst1 with no consideration of unsteady flow mechanics.
An exception to this restriction is presented by Van Houten5 using anumerical procedure.
2
R-21 18
Despite this restriction, surprisingly good correlation with measurements
of hull pressures on models and on ships are reported in Reference 11.
The problem is posed on the pressure as the primary dependent vani-
able. Because the length and cross-sectional area of the cavity (both un-I
known a priori) may undergo large excursions during a cycle, it is not pos-
sible to solve the problem in the frequency domain. Instead, it must be
set up as an initial value problem in the time domain. As will be seen,
this results in a pair of nonlinear integro-differential equations (with
time as independent variable) which must be solved numerically. Inasmuch
as the input is periodic, the output must also be periodic and, even though
an initial value problem is posed, periodicity will be achieved by runningI
through several cycles of input allowing the transients introduced by the
starting conditions to decay. No numerical results are presented here,
however, and the primary result of the analysis is to show that, in two9
dimensions, the second time derivative of the cross-sectional area of the
cavity vanishes identically regardless of the shape of the gust. As a con-
sequence, neither the potential nor the pressure grow logarithmically at
large distance. In an appendix, it is then shown that only to the order9
of the square of the reciprocal of the aspect ratio is there a possibility
that the second time derivative of the cross-sectional area does not vanish.
The solution at this order allows for the possibility that the three-
dimensional flow behaves like a three-dimensional source in the far field.
The practical inference to be drawn from this result is that the vihratory
pressures in the far field will be reduced as the aspect ratio is increased.
Propeller blades of high aspect ratio are known to be more prone to cavitate
than propeller blades of low aspect ratio. So a compromise must be drawn
between low aspect ratio blades which cavitate less and high aspect ratio
blades which induce lower vibratory pressures in the far field even though
they are more prone to cavitate.
The authors are indebted to Program Director Mr. R. Cooper, Fluid
Dynamics Branch of the Office of Naval Research, for support of this work
under Contract N00014-77-C-0122, Task NR-062-568. The analysis has been
conducted under Davidson Laboratory Project 013.
3
R-2118
A 7UTIALLY CAVITATING FOIL
Our present interest is in the linear theory for a partially cavit-
ating planar foil which is immersed in an inviscid, incompressible, fluid
of constant density p. We suppose that the unsteady velocity of the fluid
is given by
V = V + U
where u denotes a perturbation velocity while
v = U i + v (Ut - x- -I g -i
i. =i cosa + sina
= -i sin + j cosa
x = x cosa + y sin
represents a uniform flow upon which is superposed a traveling cross gust
v g. We will take v/U and the angle of attack a to be of the same order
of small magnitude.
Hereafter, all terms will be rendered in dimensionless form by using
U; the semi-span S; and the constant density p as unit quantities. The
linearized velocity vector is then
v = i + [a + v (t-x)]j+ u
where i and i are unit vectors along the x and y axes respectively. We
also suppose that u is source free in the domain of the fluid and that u
is irrotational in the fluid domain minus the wake so that in the latter
domain there exists the velocity potential function 4(x,s,y,t).
Figure I (see following page) shows the area B of the foil and the
cavity surface whose height n is assumed to be small. The trailing edge
of the cavity is assumed to lie always in B; here x = c(s,t) and the foil
leading edge bound the area C under the cavity. The quantity c which ap-
pears in the equation for the trailing edge of the foil is supposed to be
small so that the foil is one of large aspect ratio. This foil with the
straight leading edge suffices for the exposition of the basic ideas.
Its wake W is defined by
L .. .. ... . .. . . .. . . . . .. . . ... . .... . . . m l' I . . .. .. . ." .. .. I 4-
R-21 18
y
AS
(O'1'0)
y~n~x~s~x c2~s~'N
W- y n 0; ,(s ; s 1. W
The velocity potential 4,(x,s,y,t) must satisfy
xx + ss + yy=0
in the domain of the fluid minus the wake and be such that
y (x'sO+,t) = -[ct+v 9(t-x)] cOst)<(s); -l<l
For the same domain, a linearization of the Euler momentum equations pro-
vides the first order acceleration potential
P(x~s~yt) 2* x+ t=-p + f4, +4,t + p
where p is the pressure. The pressure o[x,s,n(x,s,t),t] at the cavity
wall is required to be a function of t only. Since n is small, the linearized
version of this condition is
p(x,s,O+,t) = PC (t)
which leads to the condition
5
R-2118
for (x,s) in c, that is
0 < x < c(s,t)
-l<s< I
The cavity function n(x,s,t) is determined by using the linearized kine-
matic boundary condition for the surface
y = n(x,s,t).
This linearized condition is
nx(x,s,t) + n t =a+ Vg (t-x) + y (X,s,O+,t),O < x < c(s,t); -1 < s < 1
subject to
n(O,s,t) = 0 -1 < s < I
and the closure condition
n[c(s,t) ,s,t] = 0 -1 < s < 1
Let us begin the determination of *(x,s,y,t) with the representation
(x,s,y,t) =- ff y , [f]ddaB+W [(E-X) 2 +(o-S) 2 + y2]3/2
41ff FB+W [(_x) 2 + (a-s) 2 + y2]1/2
1 +121 2
[0] = *t(c,,O+,t) -fl(,I,0-,t)
[y I = y (E,o,O+,t) - *y(E,o,O-,t)
This implies a representation for the acceleration potential P(x,s,y,t).
We have
ax at 1]/ [OirdcB+W [(-x) 2 + (-s) 2 +y2] 3 2
+ Iff ( (02d E2 d4r B+W [((-X)2 +(a-s)2 +y2] 3/2
6
R-2118
At the leading edge 0] = 0; therefore, since y/[ ]3/2 0 as C 00,
an integration by parts yields
a- a 1 -IB+W (-x)
2 + (-s)2 + ]++y23/2
However, we require the acceleration potential
P(x,s,y,t) = ,x + t
to be continuous in the wake and so
ax +PO ) 1 = 4Tf f Y[Pld~daB 2 + (a-S) 2 +y2] 3 / 2
Since we also require the normal velocity y to be continuous in
the wake, we have [cy = 0 there, from which
SI -I dd12ff 1/2 [ydyd,
2 B [(-x)2 + (s)2 +y211/ 2
1 (1a) 1 -. i [ydX daT, f f 17 f ( )
=f • Id ddd-I -1 o (Cx)2y+ (-s)2 + y2]/ 2
1 I (a) [dXddda-1~f [x2+(a-S)2+y2] 1 / 2 ' o [yd o
1 £(a) £ (a)
f f (X) f [4y]dX d~da 6-l o [(&_x)2+(aos)2 +y2] 3 / 2 &
Then
+-L + f ,4 dE daax It) 2=- T 'x ')-l [x2+(a-s)2+y2] I / 2 0 y
W f f T&[(E.x)2 + (-s)2 + y213/2 £" [,y]dA d~da
-1 o (Fx2 (7)2
+ l () (R-x) a ()da- o [(E-x)2 + (a-s)2 +y2 3/2 "t [ Y]dd
S7
R-21 18
and an integration by parts of the second term on the right-hand side of
the last equation shows
1 9.(o)
ax O 2-, t 2 )i=- Y 1/2 f y Jd do- [x 2 +(o-s) 2 +y 2 1 0
I Q.(-x)
+ -1 o [(-x)2 + (a-S)2 +y213/2 {a E y td-[}y
Downstream of the partial cavity and on B, we have [y = 0 by virtue of
the boundary conditions. Therefore
+ = (t 0.daax at 2 I t f 1/2 0dta
ax_ 7r -1 [x2 + (a-s)2 +y 2 ] 0 o
+ 1[ ( y 0 -) /yf - af t 0 1d X -[¢oy d~do .+ f [ (EX)2 + (OS)2 +y21/ at d
This can be expressed in terms of n(x,s,t) if we use the condition
nx (x,s,t) +n t = 4y (X,s,O+,t) +a +v g(t-x) = I which holds on C; and the
closure condition
n[c(s,t),s,t] = 0
with
n(O,s,t) = 0
We find
c(at) [ c c1 [10y dE f n td&= t J" n(Q,ot)dE=At(a,t)
0 0 0
where A is the cross-sectional area of the cavity. Also
C(at)
at & y y
t {nff(X ,at) + nt}dX -n,(E,o,t) -nt
Ik8
R-2118
c c(a, t)
M(,,a,t) = a'2" f nt(X,o,t)dX -2nt(F,o,t) -n (E,a,t)
Our development shows that the acceleration potential can be expressed
as
P(x,s,y,t) = (x,s,y,t) + *t(x,s,y,t)
If *A (at)da
= -I [x2 + (o-s)2 +y2] 1/2 tt
+ l. ft) (C-x) M( ,ot)d~da-f o [(E-x) 2 + (a-s) 2 +y21 3 /2
1 )
+ ,~f fy[P]d~da- o [(E-X)
2 + (a-s)2 +Y2]3/2
which is the same as
P(x,s,y,t) - . Att(s,t) RXn(x 2+y2 )
S I l -sln[ -s l+ V(a-s) 7 + xz + yZ']At (,t)daa lii
-1
1f c(at) (R -x) (s-a) M(ea-t)d~da+r 7- -1 o (&-x)2 +y2 [(s-a) 2 + (E-x)2 +y2] 112
a I(o) (s-a) (P]d~do
+ T,7-- o ( -x)2 +y 2 [(s-a)2+( -x)2+y2l1/ 2
We introduce the quantities:
x = = Cv ; c(a,t) - Ec(a,t)
y = C (a) = ch(a) A Ac2
tI
which implies a high aspect ratio wing and, because of the scaling of the
time, it also implies oscillations whose wave length is of the order of
the chord.
We now seek an asymptotic expansion for d(x,s,y,t) such that
9
6
R-2118
O(xOsIy~t) O (CYSCW, :T)= S ~ (C (,W,T )
k=l
whereVk~ ()
L .° k( = 0
Each *k is to be determined by equating terms of like order after the ex-
pansion has been substituted for * in the representation for P. However,
for the time being, let us forego a detailed development of the expansion
and be concerned in this paper with only the lowest order term O1 (x,s,y,t).
In the appendix, however, the higher order terms are examined.
Apart from the term
I4 A t(s,t) in(x 2 +y 2 ) A n(x2 +y 2)),
the lowest order approximation to P can be obtained by neglecting the com-
binations x2 +y 2 and (&-x)2 +y 2 which appear in the radicals in the last
expression for P. With respect to the non-integrated logarithmic term,
we need to consider
IAt In c2 (y 2 +w2 )
and hence
A (s,t) In c = (0 J [ydE .in ctt 21 at0 ly
Withotit loss of generality, we can take vl(g) -1 so that the order of
-- Altt(s,t) In c
is given by in e. Now this order cannot be matched with that of any other
term which appears in the expansion of P. We, therefore, conclude that we
are forced to set
A (s,t) - < s <I tt
With this, it follows that the first order approximation to P is given by
P1 (x,s,y,t) 1 x(x,s,y~t) + * t(x,s,y,t)
10L
R-2118
c 1 c( St) (&-x)MN(l,a,t)d~da
1 0 (- X)2 +
4w as f"- I( T 7 -'X)2 + y2
or
c(st) (C-x)M1(C,s,t)dCP (x,sy,t) = 27r
0 Q _x) 2 + y2
+ (S) y rPlld
o (-x) 2 + y2
We see from the last result that, with s as a parameter, Pl(x,s,y,t) isa two-dimensional harmonic function in the exterior of the cut along the
x-axis from 0 to I(s); and we also see that P1 vanishes as rx---y 4.
Analytic Determination of sLet us turn to the determination of P1 =x ( x s t) +1 and~ (x s'y't).
For typographical convenience, we will drop the subscript and often suppress
the parameter s. With this understanding, the first problem we have to
solve is this: Find P(x,y,t) such that
P(x,y,t) = x(X,y,t) + *t(x,y,t)
is harmonic in the exterior of the slit along the x-axis from 0 to 9(s),
wh i le
P(x,O+,t) = Ox(x,O+,t) + t (x,0+,t)= 1j(t) 0 < x < c(s,t);
P y(x,O+,t) = yx(x,O+,t)+ yt(xO+,t) = 0 c(s,t) < x < E(s);
P y(xO-,t) *yx(xO-,t) + ytx(,O-,t) = 0 0 < x < Z(S)
We assume that, if P(x,y,t) possesses any singularities, then they
exist at only one or both of the points
(0,0) ; (c,o+)
where (0,O)denotes the leading edge of the foil and (c,O+) denotes the
11
gt
R-2118
moving trailing edge of the cavity. Let us suppose that the physically
acceptable solution for P(x,y,t) is the one with the mildest singularities
at (0,0) and (c,O+). With respect to the behavior in the neighborhood
of infinity, we impose the condition dictated by the three-dimensional
analysis, namely
P4*O
as /7 + y-
If z = x + iy, the function defined by
i=r z -c
z ct
8=(t) = V4/-c;
maps the exterior of the slit along the x-axis from 0 to 2Z into the quarter
plane Im 4 > 0, Re > 0 where Im denotes the imaginary part and Re denotes
the real part. For this mapping, the branch /'(-z-)/z is defined as shown:
V(,-z)lZ = jv'(2-z)/zj exp i(61-6)/2
y
z Z/"
z"' (i-z)
e L~ 2
L3 (2.,O)L3
FIGURE 2. PHYSICAL PLANE
12$
R-2118
(The quantity 8 = ,r(-c)/c), for example, is negative.) Then L1 is
mapped
A
M1
i M2 M 3
FIGURE 3. MAPPED PLANE
into M defined by
0 ; ->n>O ;
L2 is mapped into M2 defined by
n - 0 ; 0 < < V'
and L3 is mapped into M3 defined by
n- 0 ; /--< < W
The point at infinity in the z-plane is mapped into
Under this mapping P(x,y,t) is transformed into a function (A,n) which
is harmonic in C > 0; n > 0 with singularities at r = 0 and r = ,. The
function
- t)
vanishes along M1 while
a { -_ " t ) }an
vanishes along M2 and M3 - By using the theory of analytic continuation.
13
7.
R-2118
it follows that ID is the real part of a function of which, except for
poles, is analytic in the 4-plane. (The continuation shows that we cannot
admit branch point singularities.) If we admit only the mildest singular-
ities at 0 and C = ", we have
0 P(t) + cl(t)Re J+ K2 (t)Re C
or in terms of z
P(x,y,t) = (t) +r(t)Re 1 + k2(t)Re z _
z
In order to avoid excessive writing in the sequel, let us express the last
result as
P(x,y,t) = Re F(z,t)
We now have the problem of finding the velocity potential *(x,y,t)
which is harmonic in the xy-plane minus the positive x-axis; exhibits a
wake downstream along the x-axis from x = k to x = o; and which satisfies
x + t = Re F(z,t)
The general solution of this equation can be expressed as
0
O(x,y,t) = f Re F(a+z, a+t)da + O(x-t+a,y,a)a-t
where a is an arbitrary constant. We may suppose that O(x,y,O) 0 for
all (x,y). Then, if we take a = 0, we have
0
O(x,y,t) f Re F(a+z,a+t)da-t
This function is harmonic for 0 < ang z < 2n; and, if we write it in the
formx
O(x,y,t) f Re F(X+iy,X+t-x)dXx-t
it is not difficult to verify that it shows a downstream wake in the sense
that 4, and Oxsuffer a jump, provided t is sufficiently large, as z crosses
the downstream x-axis.
0 14
R-21 18
Determination of the Coefficients
The next step is to enforce the two boundary conditions which 4
must satisfy: the closure condition, and the condition on the regular
part of P at infinity.
The component 4y is
0 ay(x,y,t) f f 7 Re F(a+z,a+t)day -t
- f Re i F(o+z,cy+t)da
ft
-tatxf Re i F(a+z-t,a)da
o•
For 0 < x - Oi < Z, we must have
t-[[I+Vg (t-x)] = (x,O-,t) = f f Re i F(a-t+x-Oi,o)do
0
t-x t
a { f + f Re i F(a-t+x-Oi,o)do}o t-x
However, Re i F(o-t+x-Oi,o)= 0 for 0 < a - t+x - Oi < k or t-x < a < t + Z-x
and, consequently,we must have
at x
t-+v9(tx) f Re i F(a-t+x-Oi,a)d•0
Then, if we set t-x T , we see that for T > 0
9 T 0
or
f [c, +vg ( d)a - f Re i F(o-T,or)da (1)0 0
15S
R-2118
This is the first equation which relates the unknowns Ki, K and 8.2
For c(t) < x + Oi < Z, we must have
a t-[a+V (t-x)] = f f Re i F(o-t+x+Oi,u)du
0
t-x t= { f + f Re i F(o-t+x+Oi ,C)do}
0 t-x
We already know that
at-x-[a+v (t-x)] jxRe i F(a-t+x-Oi,a)da
go0
t-x= x f Re i F(a-t+x+Oi,a)da
0
for t-x > 0 because the space variable a-t+x, in the range
-(t-x) < a - t + x < 0
required in the last integrals lies ahead of the foil and F is continuous
there. Therefore
t0 f Re i F(c-t+x+OiY)do (2)
axt-x
c(t) < x <
is the second equation which relates the unknowns.
Equation (2) can be simplified. A change in the dummy variable of
integration, a-t+x=)=, gives
a ×
0 f f Re i F(+Oi,X+t-x)dX (2a)0
C < x <
Note that
Re i F(X+Oi,X+t-x) = 0
when
c(x+t6x) < x < z16
R-21 18
This inequality defines an upper limit for the integral in (2a). For
simplicity, suppose that
X = c(A+t-x)
has but one root Al for each t-x. Then
X,=C(X 1 +t-x)
and this root is a function of t-x so that we can write
X,= x 1 (t-x)
(The'root XA1 is given by the ordinate of the point of intersection of
and the line
X = a -t+ x
which passes through the point (t,x) with slope equal to 1.) Consequently,
equation (2a) becomes
A (t-x)
-x- f Re i F(X+Oi,A,+t-x)dX = 00
orc (x1+t-x)
a Re i F(X+Oi,X+t-x)dX = 0ax
0
and, if we introduce t -X=T, we have
c(x I +-[)
d f Re i F(X+0i,X+T)dX 0
Integration gives
f Re i F (A+O i, X+t)d X0
A 1(0)
- f Re i F(X+Oi, -)dX = cons't. K
0 17
R-2118
This can be written as
f Re i F[X+O1,X++XI-c(xI+r)]dX = K00
and if we set
t = + +T
we have
c(tC)f Re i F[X+Oi, +t"-c(t*)]dX = K (2b)
00
where t is arbitrary.
We now turn our attention to the cavity surface, the equation for which is
y = n(x,s,t) 0 < x < c(s,t)
or, if we suppress s,
y = n(x,t) 0 < x < c(t)
and, as we have seen in the three-dimensional analysis, n(x,t) can be deter-
mined by using the linearized kinematic condition
nx(x,t) +nt(x,t) (xO+,t) + [c+v g(t-x)I 0 S x S c(t)
In terms of F this is
tnx(x,t) +n t(x,t) .- f Re i F(u-t+x+Oi,o)da + [a+V (t-x)]
t-x t
- f + Re i F(o-t+x+Oi,a)do} + [a+v (t-x)Io t-x 9
which, by virtue of the first condition (1), reduces tot
nx(x,t) +nt(x't) =.x f Re i F(a-t+x+Oi,a)do 0 : x < c(t)t-x
for t-x>0. Now n(x,t) must vanish at the leading edge, i.e.,
n(O,t) = 0
and the only solution of the differential equation for n(x,t) which satisfies
this condition is
18
R-2118 4
n(x,t) = - [(x-A)Re i F,(A+Oi, ,+t-x)dX0
The formula for the cross-sectional area A(s,t) determined by the
cavity surface and the foil has already been used in the three-dimensional janalysis. Here it is:
c(t)A(t) = f n(x,t)dx
0
C(t)cf [c(t) - X]Re i F[A+Oi ,X+t-c(t)]dX
0
The closure condition to be imposed is
n[c(t),t] = 0 4
This can be enforced by integrating
a tnx(x,t) +nt(x,t) =-x f Re i F(a-t+x+Oi,a)da
axt-x
x= a- f Re i F(X+Oi,X+t-x)dX
0
0 <_ x _ c(t)
Since n(O,t) = 0, we have
x xn(x,t) + f nt(E,t)d& = f Re i F(X+Oi,X+t-x)d1
0 0
or, using n[c(t),t] = 0,
c(t) d c(t) c(t)at n(C,t)d = t- f n(C,t)d&= f Re i F(X+Oi ,X+t-c)dX
o 0 0
which is the same as
c(t)A t(t) = f Re i F[X+Oi,X+t-c(t)]dX (3)
0
19
R-2118
This equation relates the rate of change of the cross-sectional area of
the cavity to the unknown parameters K1, K2 and a. However, Eqs. (2b)
and (3) imply that
A t = K°
from which
Att 0
(as we already know is required) and
A = K t + K0
If'a monotonic growth of the cavity cross-sectional area is precluded,
it follows that we must take
K = 00
It appears then that the first order approximation produces a constant
value for the cross-sectional area of the partial cavity. Equation (3)
then becomes
c(t)f Re i F[X+0i,X+t-c(t)]dX. = 0 (3a)
0
and we now have equations (1) and (3a) for the determination of K1, K 2 '
and a.
The last equation for the determination of the unknowns is
Llzl _*Re F (z,t) = 0
or
Re{ P(t) + + (t)/-} = 0 (4)
The equations (I), (3a) and (4) form a set for the determination of
Ki' K 2 and 8. The equations are linear in KI and K2 , but disconcertingly
nonlinear in 8. Two of them are highly nonlinear integral equations in t
and the ?arameter s. Their solution for given ,vg (t) and p(t) can only
be accomplished by using formidable numerical procedures.
20
R-2118
CONCLUDING REMARKS
The result 4x +4 t =Re F(z,t) presented on page 14, where F(z,t) is g
also given there, can be shown to subsume some classical results. For
example, if the dependence on time is eliminated, and KI and K 2 are eval-
uated from the kinematic condition that the vertical velocity is simply
a, then, together with the (steady) closure condition, the result can be
shown to reduce to that of Guerst 12.
As another example, take K, = 0 and write
+ x +t= '(t) + I/ K2 Re (1 - z)/ 1 / 2
As 6 *, take p - K2 =0 and we have
K t 2~i = -.Re(--/
which is the case of the unsteady, completely wetted, foil.
The analysis presented thus far is in a somewhat unsatisfactory
state inasmuch as it has been shown that Att = 0 so that no two-dimensional
source-like flow at distances large in comparison with the chord of the
foil is possible. It is clear that this result comes about because the
aspect ratio has been taken to be infinite. It is desirable to determine
to what order in aspect ratio a non-zero value of Art will occur. To this
end, we are devising a procedure for the methodical determination of
higher order asymptotic approximations. If c is the reciprocal of the
aspect ratio, and, if the orders of the asymptotic approximations are
characterized by 1, e, E2 PnE, E2...., we find that there are strong
indications that A tt remains equal to zero at least until K corresponds
to e2. Thus, it may be expected that, as the aspect ratio is increased,
the time variations of the cavity cross-sectional area and, hence, the 4
variation of the total volume of the cavity on the three-dimensional
foil) will diminish.
Finally, comparison of the formalism of this report with the two-
dimensional analysis of Van Houten shows that, whereas we have demonstrated
the necessity for taking Att = 0, thereby eliminating the source-like
21 .!
R-2118
behavior of the pressure at large distances, Van Houten has retained
this term in his work. We believe this to be improper, especially in
view of the requirement that the boundary condition at the cavity can
only be a statement that the difference between the cavity pressure and
the ambient pressure is given, whereas, if the pressure grows logarith-
mically at large distances, no ambient pressure can be defined.
The solution of the integral equations which Van Houten analyzes
numerically is equivalent to finding the acceleration potential
P"(x,s,y,t) = +x=1 A*
A- t t ( s ' t ) 9,nlx"-Y-
+ '--ai f-TZ n2la-sl A* (s,t)do47 s-1 l tt
I c(s't) (E - x)M Md2r (Ex) +yL
+I f~ [(P*dE2 7 0 (E-x) L + y,
wherec(s t)A t(s,t) = I! [ y ]dE
M a[C(ysft) w'dX -aty y
and P is subject to the cavity condition
P (x,s,O+,t) = 11(t) 0 < x < c(s,t); -l <s < I;
and the condition
P (xsO+,t) = 0
y
along the wetted surface if we take
(XS* ,-Ot) = - - Vg (t-x)
22
k
R-21 18
there. The representation for P shows that it may possess a logarithmic
singularity at infinity. However, if we admit such a singularity when
P *is determined by the function theory method used above, it turns out
again that A" (s ,t) must be zero in order to satisfy the boundary conditiont t
on the upper side of the foil downstream of the cavity and the closure
condition. Thus, it appears that the first order approximation A =t 0
can be regarded as a consequence of the imposed boundary conditions for
the two-dimensional approximating problem rather than a consequence
of matching in an asymptotic expansion.
4
4
23
R-2118
REFERENCES
1. Breslin, J.P.: "A Theory for the Vibratory Forces on a Flat PlateArising from Intermittent Propeller Blade Cavitation" Symposium onHydrodynamics of Ship and Offshore Propulsion Systems, Paper No. 6,Session 3, Det Norske Veritas, H4vik, 1977
2. Birkhoff, G.: Hydrodynamics, Princeton University Press, p. 34, 1950
3. Leehey, P.: "Boundary Conditions for Unsteady Supercavitating Flows:Fourth Symposium on Naval Hydrodynanics. B.L. Silverstein, Editor.pp. 577-597. Office of Naval Research, ACR-92, 1962
4. Benjamin, T.B.: "Note on the Interpretation of Two-DimensionalTheories of Growing Cavities" J. Fluid Mech. Vol. 19, No. 1, pp. 137-144,1964
5. Van Houten, R.J.: "The Numerical Prediction of Unsteady Cavitationof High Aspect Ratio Hydrofoils" MIT, Department of Ocean Engineering,August 1979
6. Steinberg, H. and Karp, S. : "Unsteady Flow Past Partially CavitatedHydrofoils" Fourth Symposium on Naval Hydrodynamics. B.L. Silverstein,Editor. pp. 551-575. Office of Naval Research, ACR-92,1962
7. Wang, D.P. and Wu, T.Y.: "General Formulations of a PerturbationTheory for Unsteady Cavity Flows" Journal of Basic Engineering ASMETransactions, Series D, Vol. 87, No. 4, pp. 1006-1010, 1965
8. Unruh, J.F. and Bass, R.L.: "Doublet Lattice-Source Method for Cal-culating Unsteady Loads on Cavitating Hydrofoils" J. Hydronautics,Vol. 8, No. 4, pp. 146-153, October 1974
9. Johnsson, C.A. and Sintvedt, T.: "Propeller Excitation and Responseof 230,000 TDW Tankers" Ninth Symposium on Naval Hydrodynamics. R. Brardand A. Casera, Editors, Vol. 1, pp. 5111-G69, 00jffice of Naval Research,ACR-203,1972
10. Noordzij, L.: "Pressure Field Induced by a Cavitating Propeller"International Shipbuilding Progress, Vol. 23, No. 260, pp. 93-105, 1976
11. Kaplan, P., Bentson, J. and Breslin, J.P.: "Theoretical Analysis ofPropeller-Radiated Pressure and Blade Forces Due to Cavitation" PaperNo. 10, Symposium on Propeller-Induced Ship Vibration" Royal Inst.of Naval Architects, London, 10-14 December 1979
12. Guerst, J.A.: "Linearized Theory for Partially Cavitated Hydrofoils"International Shipbuilding Progress, Vol. 6, No. 60, pp. 369-384, 1959
24
R-2118
We proceed to give a method for the theoretical determination of
higher order approximations for the cavitating foil. It is beyond the
scope of the present report to supply all of the details of the develop-
ment. It is expected that a fuller analysis will be produced later. Here,
our object is to find the order of approximation which gives the first
indication that Att #Owhere A is the cross-sectional area of the cavity.
We have found that the acceleration potential is
P(x,s,y,t) = x(X,s,y,t) + *t(x,s,y,t)
1 a a 1 c(C;t) y ]d~do
-1 0 [(Ex)2 + (os)2 +y2]1/ 2
+ I (a) y[P]d~do
4-l [(-x) 2 + (-s) 2 +y2]3/ 2
where
[41 = WE'at0) =(',uO+,t) - f(laO-,t)
In order to satisfy the boundary condition for the cavity, we must
satisfy
Ic(a t) [4 ]ddaP(x,s,O+,t)-i [P(x,s,t)] = - 4I f
for 0 < x < c(s,t) ; -1 < s < 1
In order to satisy the boundary condition for the wetted part of
the foil, we must satisfy
xy(X,s,O_,t) + ty(X,S,0±,t)
1 a a xt)y _I a2 1 t(O) /(x-) + (s-o) 2 [Pld~do= 2 ax at y 41T axas 0 (x-)(s-G)
An integration gives
CW
a 0y ( ,s,0,t)dC + at y( 's'O-+t)d
axA-
L A-l
R-2118
I!
2 fx
+ I (o) /,(x- ')2 +(s-a) 2 - I--- Pdd
- (x-) (s-) [P]dd
and if we solve this for 6
f 4,(C,s,O±,t)d& we findx
II
f y s;,0±,,t) dl - f [ y( ,s,t)1dC
xY x
x 1 (G) +________
+ f . f ( . )2 + (sa)2 I .} [P(,a,A+t-x)]d~dadAx-t -1 o s-0
In dimensional terms, the normal velocity along the wetted part of
the foil is required to be
-[ua + v (Ut-x)]
9
In dimensionless terms, this velocity is
v-[a + -Eu (St-sx)]
U
where S is the semi-span. For a foil of large aspect ratio, we can take
KS
where c is small and K is the chord length, so that the normal velocity can
be expressed as
V 6-{a + 9 [T (t-x)]}
U C
or
-[a + V (.-)]
In dimensional terms, the acceleration potential at the cavity wall is
P(t)
A-2w •
R-2118
In dimensionless terms, it is
p P(CU) _(t-P-- U= - = p(-
~I
In dimensional terms, the trailing edge of the foil is
x = Sk(s) = Sef(s)
and in dimensionless terms
x = Ef(s)
In dimensional terms, the trailing edge of the cavity is given by
x = c(s,t) = S C(s ) Sc(s,t)'U
and in dimensionless terms
x c c(s, fr.)= (4
Since the unknowns depend on the parameter- c, let us introduce
x - F-y ; - y C= jj t = ET
and assume the inner asymptotic expansions
O(x,s,y,t) = O(Ey,s,pIETj) = c' (vSit;e) = K cVK(E) 4K(V,S 'u,
')
c(s,-) - c ()c(s,T)E K1I
where
u ° (E)- - 0
K
In terms of the new variables, the boundary condition for the cavity is
(V sO ,T; ) - ['P(y,s,T;C)]
I c( ') (W (w,Tr;E)dwdo
-) f ' /4w Ty + ' - o [(s-a)2 +r2]1 2
A-3
R-21 18 4
where r2 = 2(-2
and 0< Y< c
This can be expanded into
1 ( s T ) kni -
0
+ CI. 1 a,T) 1L0-L Zn 2j0-sJ I Iduida4r as f -s 1
(A- 1)
a+a Ia 2 C 2iw-yi [l-j d0
la-sL.E2(-)2
+ 3 1c(U,-r) a-s 4
-1r 3s f a0-s C2 )2zn2lo-s( j'f? 1dwda
+ o(s2 ;UqsVWqy)
In terms of the new variables. the boundary condition for the wetted
surface is
f 0 (W,S,O±,T;c4dw = ++
Y Y1 a f(a) _____
+i - as }~-)[P(,G,a,+T-Y;E)]dwdadX
where
q c(-)
An expansion of this yields
OD1 c
Y Y4
+ E a f. ( Y) W,G,X+T--Y;c)]dwdadX
Y T - o
A-4I
R-21 18
IYf(a) ['P(W,S,X+T-Y;C)]dwdX
yT 0
+L -J f"f - Zn2Is-ai (X-w)[P(w,sX+T-y;E)]dwdadX
+ 1 f 0a
+4 f _T f o(E2 ;j,s,W) [W(w,a,X+T-y;c)]dwdadXY-T o
where
h(ca-s) ={j >
The expression (A-2) shows that the inner asymptotic expansion of
4)(x,s,y,t) should have the form 4)(x,s,y,t) = 4-)SypsjiE) =cf(ypstiT;E)
where
'(YV~,suTp,) = 4)1(YPsPiPT) + C 2 + C tn + 24+
Then for P(x,s,y,t), we have
P(x,S,Y,t) 4) (xps'y't) + 4)t (x,s~y,t) or
P = 4 (Yts*VT;E) +T
-P I(Yps,j~T) +EP 2
+ £29,nc -P + e2p + .
3 4
The function P(x,s,y,t) is a three-dimension potential function in2 22
the exterior of the foil. If we apply the operator +x + - t h
assumed asymptotic expansion, we find
A- 5
R-21 18
v K(E) [PKyy + PK~lp + 62 P KS
and, hence, each P K(Y's,ii1T) must satisfy
PY (y,5,'JT) + P H (Y'sPT)
in the exterior of the slit defined by
0 < y < f(s) ; '=0
For example, we have
P K + P =PJ 0 K =1,2,3
wh il1eIP +yy P =-p P .
The equations for the I 's can now be found by equating terms of
like order in (A-]) and (A-2).
Consider first In (A-i) there is no match for the term
2r +a f [ ,,]dw0
and, therefore, we must take
(.-+~- f [ 1 d. -- + '- A1 (s,T) =A (S,T) = 0
Here A is the first order approximation to the cavity area.
For the cavity, 4~must then satisfy the integral equation
I a c I (s ,t0P(T) [ P1 (y,s,T)] -L(2 T y T f Znjbw-yj [ , ]dw
0
and, for the wetted surface, it must satisfy
0 c 1 1Tf(s) (P 1 (w,s,X)]dwdX
f (WpSpO±,T)dw± f [0, (W,S,r)]dw + 'Tf f X + ~-Y 2Y 0 0 Y
from which
r A-64
R-2 118
T f(s) [Pl(w;s,X)]dadX
= 21T D f f X+ - T0 0
and1 11 .( Y ' s , O -+ ,r ) + I 1 T ( Y p s , O -± ,T )
f(s) [P (w,s,T) ]dwf(7a + (Y',s , T)]T Y 0
where, on the wetted surface,
= -[a + v (t- )]
As we have seen, these integral equations can be solved by first
finding the analytic function F(z,t) such that
Re F(z,t) = 1ix(x,s,y,t) + t(x'syt)
a, c I f(s) P[IP1dw+ ( y'-) f Zn/(y) 2 +j 2 [qudw + - foir aY o (W-y) 2 + 1a2
2I (ax -aL n -x + I 9" (s) y[P ]d21 x3 nV-.x + P d f (E-x)2 +'0 0
1 c(s,t) ( -x) M1 ( ,s,t)d I P(s) y[P 1 ]df J ( -x)2 + y2 + 27r T -- 70 0
The above representations and the integral equations imply that we must
have
ReF (x+Oi,t) = p(t)
for the cavity and
Re F (z, t) = 0ay
y=O_+
Iy(xs,0±,t) = -[a+v g (t-x)]
along the wetted surface while ReF (z,t) vanishes at infinity.
It has been shown that the first approximation to the cavity surface
y = n (x,s,t) can be found from
A-7
R-2118
nlx(X,s,t) + nit(x,s,t) = [bly(xSO,t)]
x= x" f Re i F(\+Oi, k+t-x)dX
0
xRe i F(x+Oi ,t) + f -i- Re i F(k+Oi ,X+t-x)dX
0
for 0 < x < c. From thisc-c
[1 (C-c,s,O,t)] = Re iF(c-E+Oi,t) - Cf ' Re i F(X+Oi,X+t-x) dX 4y 0 1X=C-
Rei F(c-E+Oi,t) _I aS ReE F(X+Oi X+t-c+c)dXatf e
if c is a sufficiently small positive quantity.
The closure condition
n(c,s,t) = 0
givesc cf nlt( 's't)dC f [bly (E,s,O,t)]d
00
orc
Alt = fRei F( X+Oi,X+t-c)dX
0
orc
A = - SRei FX+Oi,X+t-c)dX
0
but we have found that A tmust vanish. Therefore
L Rei F(c-c+Oi,t)L o qly( - , , t) = .-+o 1 - E:
which is infinite as can be seen from the explicit expression for F. On
the other hance
L [ly (c+cs,O,t] = 0
A-8
R-21 18
Consider second 42 For the determination of *2' we need to
cons ider the term
I* c
We leave c unexpanded and define it to be equal to c1. Indeed,
[,Ivanishes beyond c, so that such terms can make no contribution
to 02. A similar argument will hold for the higher orders. Since no
other term of order F_ Znc appears in (A-i), we must take
A = 4]dw =0
0
We see now that, even for the second order approximation, the second
derivative of the cross-sectional cavity area is zero.
With this, 2 must satisfy
'P2 Y S,O+,T) - -L [P2 (y~st)]= Wc1(s,
for the cavity.
For the wetted surface, 2 must satisfy
f (w,s,O±,T)dw T- f [0 (w, s,-)]dw
T If (a) ____________crd T..L f(S) [P2 (w,s, X))d(d X
from which
T f(s) [P (w,s,X)Idw~dX(YpspO±VT) 2 ~~( ,) +--
2p+2 0YS'r] 27r Dy 02
and hence
2p (YSOT)+ 2 T(YpSQ±,r)
7Yf 7s) [P (W,S,T)]dw
A- 9
R-21 18
These integral equations can be solved by first finding the potential
function
Re F2 (zt) = 2x(X,syt) + 2 t(x,s,y,t)
c v f(s) iJ[P 2 ]dw
2-r ""ay+ T n (-y) + 2 T[]d+" (W-y)z + P z0 0
c 9 (S) ysP 2[]dP
(x+ f kn/(-x)2 +y2 [2yd+0 0
Sc (U-x)M2 dE I 9,(s) y[P2] dE
=21T 27 (0x' y / ( -x) 2 + y2_
The function Re F2 (z,t) must vanish at infinity and the integral
equations show that it must be such that
Re F2 (x+Oi,t) = P2 (x,s,O+,t)
for 0 < x < c, while along the wetted surface
Re F2 (z,t) y=0 + = (2xy(X,s,O,t) + 2ty(X,s,O,t)
where d 2y(x,s,O,t) is an induced velocity on the wetted surface and
P2 (x,s,O+,t) is an induced acceleration potential at the cavity. The
latter quantities are to be determined by matching inner and outer expan-
sions.
Consider next 3: The component 3 is to be determined by equating terms
of order 62 in E in (A-l) and (A-2); but first we need to consider the terms
c i"+ - ) f E2 n [1 e Idw
Because the term of order e2 (zn0)2 cannot be matched in (A-1), it
follows that we must have
c
A = =0
0
A-IO- 4I,
R-2118
We can now see that 43 must satisfy
P (Y'S o+PT) 1 [P3(Y,S,-r)]3 2 3p3(y,s,O+,t') - -P 3 ys,)
c1
f Zn~w-yj [ 3 ] dw0
Cl
0c1
1 32 )2- 8 ' s' f (W-Y)2 [¢ I dw
0
for the cavity, i.e., 0 < y < pl; p = 0+.
For the remainder of the foil 03 must satisfy
f 4 31 (W,s,O±,T)dwY
c s T f(s)
-+ 203 f f [P3 (ws,
'W)]dwdd
Y 0 0
1 1 f()-4r- f .23 f f h(o-s) (X+y-T-w) [P1 (w,a,X)]dwdad
o -1 o
from which
3 f(s) [P3(),s,X)]dwd€311 ' '0+ T =+ 31,1 ' 2ny" 0 X,+Y-T-(W
I T a 3 1 f (c)
+- T 3 f f h (oy- s)[ P I(w ,7;)d d,7 d X
0 1 0
and
3,4Y 5,0± 4) +031T (YVSO±pT)
+A + ( ,,T)
A-11
R-2118
a -s) [P3( ,sX)]dw
2Tray Y-W
1 a3 1 f (o)+* - f f h(o-s) [P1(w,a,X)]dwdo
These integral equations, which express the boundary conditions,
can be solved by first finding the potential function
P3(ys,T) =Re F3(zt)c1
=- a- + f nV(w-y)2 +Z [,I d
+ L f(s) i[P3]dw 1 + c 1+ L I -( _ ) + -- 2 a F [t ]d w
_iT(aT a a2 1d
8ff ay + -) ' I {(w-y) 2 + 21 [0 Idw
0
1 a2 f(s)
0
where z = Y + i. From the integral equations, we find that Re F3 (Z,T)
must satisfy
Re F3(y+Oi,T) = P3 (Y,s,O+,r) for the cavity and
Re F3 (z ,-) =0 3 y (Y'sO 'T) + 0 3p,(y,s,OaT)
for the remainder of the foil; while F3(z,T) must have a pole at infinity.
If
c2, a f [01dw
is not zero, 04 remains to be determined from the equations for order c2
and there is a coupling between the approximations of third and fourth
order. In the solution for ReF 3, the induced quantities 0 3p(ys,O,T)
A-12ij. .5
and 3(y~,O+,) ar at irstR-21I 18 IandP 3YSO+,) ae a fistunknown. Ultimately they are to be found
by matching inner and outer expansions.Finll, lt s cnsder4): Along the cavity interval 4~must
satisfy
P (Y,s,cj+,t) - P4 2 [ 4 (YIsIT)]
r Clf X j- j (41(, ,)d
0
+ "s' n cl-l C dd12.2 as f-si f []d4 da-l o
+ay 1 2cl
1 330- s 1 33 1 1 S (w-y)2 [l + in l-i da
Along the wetted surface, 4 must satisfy
*4 (Y,S,O±,T)
= ± I[~()]- ~ tf~s [P4 (w ,s,X)]ddX
0 0
+ I 93 f~a 3 Is-al + II!k Lzn2s--0 ol~sa 2 n25 P1 (w,a,X)]dwdadX
fro wic + h(a-s) {nj)X+y-T-wj+l)
0) (YvsO±,T) "44jT(y'S,O±,T)
= ,* 33 1 3f(S) [P (w,S,T)]dw
0
1 33 1 f (a) 3I-l+~L. n2sa+ -~~f f 3 s-al+ Ia 2 1s-a lF7 S 0 )- [P 1(W,a,T)]dwdaL+h(a-s) (inly-wi +11
A- 13
R-21 18
In accordance with these integral equations, the function
Re F (zT ) (--+ a-) f YnV'(w-y) 2 + - [I+]dw f(s)[
27r 3y a 4 7 f (wy) z +r0 0
= ~*£ny fi2 a 1 f 7 ) p(P 4 ]dw
n i r y a(_t f [ ]dw+ T f y) y
27r 77YTZ+=) 7Tf[1 4 [ 414I0 W 4
must. satisfy
Re F4(y+Oi,) =
P (y,s,O+,T)4 1c 1
- ..± .2L n21o-sl f [41 ]dcoda
-.i as --1 0-s o~
c
+ .f1 (w-y) 2 ZnjW-YI [€1 1dw0
as I1 O C-S ( (_y)12 +Zn 2fa-sj} [4lQdwda
for the cavity, and
a Re F (zT)j
=4P (YS,O±,T) + *4 pT(y,s,O±,T)
J' f 1jT -L :0 +4I2. Zn 2js-a1S -1 o S [P Idwda
h(a-s) (Znly-wj+ 1} Jfor the wetted part of the foil.
As we have seen, the function P4 . 04y + 0 T must satisfy the Poisson
A-14
..
R-21 18
equation
P + =-P4yy 4PjIJ Iss
where
c
P Y,,P1)f 9,n Ay-w~) + jj t4 dw(y,s,p,) ( -_ a+T) n ( -) [lji
I fs) [IIP1 ]dw+ r 0 +P2
It can be shown that if S4 is an appropriate particular solution of the
Poisson equation, then
P4 = S 4 + Re F4 (z,T)
The potential function Re F4 possesses a logarithmic singularity at
infinity, namely,
ci
f [4]dw . kn/Yl7+1JP2
provided
c1
A4 = ] [(,]dw
does not vanish. So far there is no apparent reason for supposing that
A =0.4,TT
We conclude from our analysis that, if A(s,t) is the cross-sectional
cavity area and if Att does not vanish, then it must be of order equal to
or higher than c2.
A-15
JM
R-2118 1
DISTRIBUTION LIST(Contract NOOOl4-77-C-0122, Task NR 062-568)
12 DEFENSE DOCUMENTATION CENTER I NASA Scientific and Technical
Cameron Station Information Facility
Alexandria, VA 22314 P.O. Box 8757Baltimore/Washington International
Technical Library AirportDAVID W. TAYLOR NAVAL SHIP Maryland 21240
RESEARCH & DEVELOPMENT CENTERAnnapolis Laboratory 1 LibrarianAnnapolis, MD 21402 Dept. of Naval Architecture
UNIVERSITY OF CALIFORNIAProfessor Bruce Johnson Berkeley, CA 94720
Engineering DepartmentNAVAL ACADEMY 1 Professor P. Lieber
Annapolis, MD 21402 Dept. of Mechanical Engineering
UNIVERSITY OF CALIFORNIALibrary Berkeley, CA 94720
NAVAL ACADEMYAnnapolis, MD 21402 1 Professor W.C. Webster
Dept. of Naval ArchitectureProfessor T. Francis Ogilvie UNIVERSITY OF CALIFORNIADept. of Naval Architecture & Berkeley, CA 94720
Marine EngineeringUNIVERSITY OF MICHIGAN I Professor J.V. WehausenAnn Arbor, MI 48105 Dept. of Naval Architecture
UNIVERSITY OF CALIFORNIAProfessor W.S. Vorus Berkeley, CA 94720Dept. of Naval Architecture &
Mechanical Engineering I Professor J.A. SchetzUNIVERSITY OF MICHIGAN Dept. of Aerospace EngineeringAnn Arbor, MI 48105 VIRGINIA POLYTECHNIC INSTITUTE
Blacksburg, VA 24061Professor C.S." YihDept. o'- Engineering Mechanics 1 DirectorUNIVERSITY OF MICHIGAN OFFICE OF NAVAL RESEARCH BRANCHAnn Arbor, MI 48105 OFFICE
495 Summer StreetDr. Coda Pan Boston, MA 02210SHAKER RESEARCH CORPORATIONNorthway 10 Executive Park I CommanderBallston Lake, NY 12019 PUGET SOUND NAVAL SHIPYARD
Bremerton, WA 98314Professor S. CorrsinDept. of Mechanics and Dr. Alfred Ritter
Materials Sciences CALSPAN CORPORATIONTHE JOHNS HOPKINS UNIVERSITY P.O. Box 235Charles and 34th Street Buffalo, NY 14221Baltimore, MD 21218
Professor G. Birkhoff I LibraryDept. of Mathematics NAVAL WEAPONS CENTERHARVARD UNIVERSITY China Lake, CA 93555Cambridge, MA 02138
1 Professor E. ReshotkoI Professor G.F. Carrier Division of Chemical
Division of Engineering and Engineering ScienceApplied Physics CASE WESTERN RESERVE UNIVERSITY
Pierce Hall Cleveland, OH 44106HARVARD UNIVERSITY
Cambridge, MA 02138 1 Commander
CHARLESTON NAVAL SHIPYARDCommanding officer Naval BaseNROTC Naval Administrative Unit Charleston, SC 29408MASSACHUSETTS INSTITUTE OFTECHNOLOGY 1 Professor J.M. Burgers
Cambridge, MA 02139 Institute of Fluid Dynamics andApplied Mathematics
Professor J. Kerwin UNIVERSITY OF MARYLAND
Dept. of Ocean Engineering College Park, MD 20742
MASSACHUSETTS INSTITUTE OFTECHNOLOGY I Professor Pai
Cambridge, MA 02139 Institute for Fluid Dynamics andApplied Mathematics
Professor P. Leehey UNIVERSITY OF MARYLANDDept. of Ocean Engineering College Park, MD 20742
MASSACHUSETTS INSTITUTE OFTECHNOLOGY 1 Technical Library
Cambridge, MA 02139 Naval Weapons Surface CenterDAHLGREN LABORATORY
I Professor C.C. Mei Dahlgren, VA 22418
Dept. of Civil EngineeringMASSACHUSETTS INSTITUTE OF I Computation & Analyses Laboratory
TECHNOLOGY NAVAL WEAPONS SURFACE CENTERCambridge, MA 02139 DAHLGREN LABORATORY
I Dahlgren, VA 224181 Professor J. Nicholas Newman
Dept. of Ocean Engineering 1 Dr. C.S. Wells, Jr.Room 5-324A ADVANCED TECHNOLOGY CENTER, INC.MASSACHUSETTS INSTITUTE OF P.O. Box 6144
TECHNOLOGY Dallas, TX 75222CAMBRIDGE, MA 02139
1 Professor A. McKillopProfessor A.R. Kuhlthau Dept. of Mechanical EngineeringDirector, Research Laboratories UNIVERSITY OF CALIFORNIA
for Engineering Sciences Davis, CA 95616Thornton HallUNIVERSITY OF VIRGINIA 1 Dr. J.A. YoungCharlottesville, VA 22903 JAYCOR
1401 Camino Del MarDirector Del Mar, CA 92014
OFFICE OF NAVAL RESEARCH BRANCHOFFICE
536 South Clark Street
Chicago, IL 60605
R-2118
Professor Ralph Burton 1 Dr. J.P. Craven
Dept. of Mechanical Engineering UNIVERSITY OF HAWAII
& Astronautical Sciences 1801 University Avenue t
NORTHWESTERN UNIVERSITY Honolulu, HI 96822
Evanston, IL 602101 Professor L. Landweber
Dr. Martin H. Bloom Institute of Hydraulic Research
Director of Gas Dynamics Research UNIVERSITY OF IOWA
POLYTECHNIC INSTITUTE OF NEW YORK Iowa City, IA 52242 4
Long Island CenterFarmingdale, NY 11735 1 Professor E.L. Resler
Graduate School of Aerospace
Technical Documents Center EngineeringBuilding 315 CORNELL UNIVERSITYARMY MOBILITY EQUIPMENT RESEARCH Ithaca, NY 14851
CENTERFort Belvoir, VA 22060 1 Dr. D.E. Ordway
SAGE ACTION, INC.Professor E.R. Lindgren P.O. Box 416Dept. of Engineering Sciences Ithaca, NY 14850UNIVERSITY OF FLORIDA231 Aerospace Engineering Building I Professor V.W. Goldschmidt
Gainesville, FL 32611 School of Mechanical EngineeringPURDUE UNIVERSITY
Technical Library Lafayette, IN 47907WEBB INSTITUTE OF NAVAL ARCHITECTUREGlen Cove, NY 11542 1 Professor J.W. Miles
Institute of Geophysics & Planetary
Dr. M. Poreh Physics, A-025TECHNION-ISRAEL INSTITUTE OF UNIVERSITY OF CALIFORNIA
TECHNOLOGY San DiegoDept. or iil Eniretr'ng La Jolla, CA 920 3Haifa, Israel
1 FLOW RESEARCH, INC.Dr. -,.P. Breslin Los Angeles DivisionDavidson Laboratory 9841 Airport Blvd., Suite 1004STEVENS INSTITUTE OF TECHNOLOGY Los Angeles, CA 90045Castle Point StationHoboken, NJ 07030 1 Professor A.T. Ellis
Dept. of Applied Mathematics and
Dr. T.R. Goodman Engineering SciencesDavidson Laboratory UNIVERSITY OF CALIFORNIASTEVENS INSTITUTE OF TECHNOLOGY San DiegoCastle Point Station La Jolla, CA 92037Hoboken, NJ 07030
1 Professor W. LindbergDr. D. Savitsky Dept. of Mechanical Engineering
Davidson Laboratory UNIVERSITY OF WYOMING
STEVENS INSTITUTE OF TECHNOLOGY Laramie, WY 82070Castle Point StationHoboken, NJ 07030
R-2118
Mr. Otto Decker I Professor J. WuMECHANICAL TECHNOLOGY, INC. College of Marine Studies968 Albany-Shaker Road UNIVERSITY OF DELAWARELatham, NY 12110 Newark, DE 19711
Mr. Virgil Johnson, President 1 Technical LibraryHYDRONAUTICS, INC. NAVAL UNDERWATER SYSTEMS CENTER7210 Pindell School Road Newport, RI 02840Laurel, MD 20810
1 OFFICE OF NAVAL RESEARCHMr. M.P. Tulin New York Area OfficeHYDRONAUTICS, INC. 715 Broadway - Fifth Floor7210 Pindell School Road New York, NY 10003Laurel, MD 20810
I Professor J.J. StokerCommander Courant Institute of MathematicalLONG BEACH NAVAL SHIPYARD SciencesLong Beach, CA 90801 NEW YORK UNIVERSITY
251 Mercer StreetProfessor John Laufer New York, NY 10003Dept. of Aerospace EngineeringUNIVERSITY OF SOUTHERN CALIFORNIA I Professor D.D. FullerUniversity Park Dept. of Mechanical EngineeringLos Angeles, CA 90007 COLUMBIA UNIVERSITY
New York, NY 10027Lorenz G. Straub LibrarySt. Anthony Falls Hydraulic I Professor V. Castelli
Laboratory Dept. of Mechanical EngineeringUNIVERSITY OF MINNESOTA COLUMBIA UNIVERSITYMinneapolis, MN 55414 New York, NY 10027
Professor J. Ripkin I Professor H.G. ElrodSt. Anthony Falls Hydraulic Dept. of Mechanical Engineering
Laboratory COLUMBIA UNIVERSITYUNIVERSITY OF MINNESOTA New York, NY 10027Minneapolis, MN 55414
I ENGINEERING SOCIETIES LIBRARYDr. E. Silberman 345 East 47th StreetSt. Anthony Falls Hydraulic New York, NY 10017
LaboratoryUNIVERSITY OF MINNESOTA I SOCIETY OF NAVAL ARCHITECTS ANDMinneapolis, MN 55414 MARINE ENGINEERS
74 Trinity PlaceProfessor Paul F. Pucci New York, NY 10006Mechanical Engineering DepartmentNAVAL POSTGRADUATE SCHOOL I Technical LibraryMonterey, CA 93940 NAVAL COASTAL SYSTEM LABORATORY
Panama City, FL 32401Library
i.AVAL POSTGRADUATE SCHOOL I Professor A.J. AcostaMonterey, CA 93940 Dept. of Mechanical Engineering
CALIFORNIA INSTITUTE OF TECHNOLOGYPasadena, CA 91109
R-2118
Professor H.W. Liepmann I CommanderGraduate Aeronautical Laboratories NORFOLK NAVAL SHIPYARDCALIFORNIA INSTITUTE OF TECHNOLOGY Portsmouth, VA 23709Pasadena, CA 91109
1 Mr. C. duP Donaldson AssociatesProfessor M.S. Plesset AERONAUTICAL RESEARCH OF PRINCETON,Engineering Science Department INC.CALIFORNIA INSTITUTE OF TECHNOLOGY 50 Washington RoadPasadena, CA 91109 Princeton, NJ 08540
Professor A. Roshko I Chief, Document SectionGraduate Aeronautical Laboratories Redstone Scientific InformationCALIFORNIA INSTITUTE OF TECHNOLOGY CenterPasadena, CA 91109 ARMY MISSILE COMMAND
Redstone Arsenal, AL 35809Professor T.Y. WuEngineering Science Department 1 ARMY RESEARCH OFFICECALIFORNIA INSTITUTE OF TECHNOLOGY P.O. Box 12211Pasadena, CA 91109 Research Triangle Park, NC 27709
Director I Dr. H. Norman AbramsonOFFICE OF NAVAL RESEARCH BRANCH SOUTHWEST RESEARCH INSTITUTE
OFFICE 8500 Culebra Road1030 E. Green Street San Antonio, TX 78228Pasadena, CA 91101
1 ONR Scientific Liaison GroupMr. R. Wade AMERICAN EMBASSY - Room A-407TETRA TECH., INC APO San Francisco 96503Marine & Environmental Engineering
Division I Editor630 N. Rosemead Boulevard Applied Mechanics ReviewPasadena, CA 91107 SOUTHWEST RESEARCH INSTITUTE
8500 Culebra RoadTechnical Library San Antonio, TX 78228NAVAL SHIP ENGINEERING CENTERPhiladelphia Division 1 Technical LibraryPhiladelphia, PA 19112 NAVAL UNDERSEA CENTER
San Diego, CA 92132Technical LibraryPHILADELPHIA NAVAL SHIPYARD 1 Dr. Andrew FabulaPhiladelphia, PA 19112 Code 4007
NAVAL UNDERSEA CENTERDr. Paul Kaplan San Diego, CA 92132OCEANICS, INC.Technical Industrial Park 1 OFFICE OF NAVAL RESEARCHPlainview, NY 11803 San Francisco Area Office
760 Market Street, Room 447Technical Library San Francisco, CA 94102NAVAL MISSILE CENTERPoint Mugu, CA 93041 1 Library
PEARL HARBOR NAVAL SHIPYARDCommander Box 400PORTSMOUTH NAVAL SHIPYARD FPO San Francisco 96610Portsmouth, NH 03801
R-2118
Technical Library 1 Dr. J.M. RobertsonHUNTERS POINT NAVAL SHIPYARD Dept. of Theoretical and AppliedSan Francisco, CA 94135 Mechanics
UNIVERSITY OF ILLINOISProfessor Bruce H. Adee Urbana, IL 61803Dept. of Mechanical EngineeringUNIVERSITY OF WASHINGTON I Techn~cal LibrarySeattle, WA 98195 MARE ISLAND NAVAL SHIPYARD
Valleno, CA 94592
Dr. Steven Crow, PresidentPOSEIDON RESEARCH 3 OFFICE OF NAVAL RESEARCH11777 San Vicente Blvd., Suite 641 Code 438Los Angeles, CA 90049 800 N. Quincy Street
Arlington, VA 22217Librarian
NAVAL SURFACE WEAPONS CENTER I OFFICE OF NAVAL RESEARCHWhite Oak Laboratory Code 200Silver Spring, MD 20910 800 N. Quincy Street
ArlTrngon, VA 22217Mr. J. RogersNAVAL SURFACE WEAPONS CENTER I OFFICE OF NAVAL RESEARCHWhite Oak Laboratory Code 210Silver Spring, MD 20910 800 N. Quincy Street
Arlington, VA 22217
Fenton Kennedy Document LibraryTHE JOHNS HOPKINS UNIVERSITY 1 OFFICE OF NAVAL RESEARCHApplied Physics Laboratory Code 211Johns Hopkins Road 800 N. Quincy StreetLaurel, MD 20810 Arlington, VA 22217
Professor E.Y. Hsu I OFFICE OF NAVAL RESEARCHDept. of Civil Engineering Code 212STANFORD UNIVERSITY 800 N. Quincy StreetStanford, CA 94305 Arlington, VA 22217
Dr. R. L. Street I OFFICE OF NAVAL RESEARCHDept. of Civil Engineering Code 221STANFORD UNIVERSITY 800 N. Quincy StreetStanford, CA 94305 Arlington, VA 22217
Professor Milton van Dyke I OFFICE OF NAVAL RESEARCHDept. of Aeronautical Engineering Code 473STANFORD UNIVERSITY 800 N. Quincy StreetStanford, CA 914305 Arlington, VA 22217
Professor R. DiPrima 1 OFFICE OF NAVAL RESEARCHDept. of Mathematics Code 480RENSSELAER POLYTECHNIC INSTITUTE 800 N. Quincy StreetTroy, NY 12181 Arlington, VA 22217
Professor J.W. Holl 1 OFFICE OF NAVAL RESEARCHDept. of Aerospace Engineering Code 481THE PENNSYLVANIA STATE UNIVERSITY 800 N. Quincy StreetUniversity park, PA 16801 Arlington, VA 22217
R-2l 18
6 OFFICE OF NAVAL RESEARCH I Code 6034+Code 1021P (ONRL) NAVAL SHIP ENGINEERING CENTER800 N. Quincy Street Center BuildingArlington, VA 22217 Prince George's Center
Hyattsville, MD 207826 Code 2627
NAVAL RESEARCH LABORATORY I Code 6101EWashington, DC 20375 NAVAL SHIP ENGINEERING CENTER
Center BuildingICode 4+000 Prince George's CenterNAVAL RESEARCH LABORATORY Hyattsville, ND 20782Washington, DC 20375
1 Code 6110IDr. D.L. Hunston NAVAL SHIP ENGINEERING CENTERCode 6170 Center BuildingNAVAL RESEARCH LABORATORY Prince George's CenterWas-hington, DC 20375 Hyattsville, MD 20782
77PS ~1LNAVAL RESEARCH LABORATORY NAVAL SHIP ENGINEERING CENTERWashington, DC 20375 Center Building
Prince George's CenterMir. R.J. Hansen Hyattsville, MD 20782Code 84+41NAVAL RESEARCH LABORATORY I Code 6136Washington, DC 20375 NAVAL SHIP ENGINEERING CENTER
Center BuildingMr. L. Benen Prince George's CenterCode 0322 Hyattsville, MD 20782NAVAL SEA SYSTEMS COMMANDWashington, DC 20362 1 Code 614+0
NAVAL SHIP ENGINEERING CENTERIMr. J. Schuler Center BuildingCode 032 Prince George's CenterNAVAL SEA SYSTEMS COMMAND Hyattsville, MD 20782Washington, DC 20362
I Dr. A. Powell (Code 01)ICode 03B DAVID W. TAYLOR NAVAL SHIPNAVAL SEA SYSTEMS COMMAND RESEARCH & DEVELOPMENT CENTER .Washington, DC 20362 Bethesda, MD 20084
1Dr. T. Peirce I Mr. W.M. Ellsworth (Code 11)Code 03512 DAVID W. TAYLOR-NAVAL SHIPNAVAL SEA SYSTEMS COMMAND RESEARCH & DEVELOPMENT CENTERWashington, DC 20362 Bethesda, MD 20084 4
1Library I Dr. W.E. Curmmins (Code 15)Code 09GS DAVID W. TAYLOR NAVAL SHIPNAVAL SEA SYSTEMS COMMAND RESEARCH & DEVELOPMENT CENTERWashington, DC 20362 Bethesda, MD 20084
Dr. W.B. Morgan (Code 154) 1 Dr. A.L. SlafkoskyDAVID W. TAYLOR NAVAL SHIP Scientific Advisor
RESEARCH &. DEVELOPMENlT CENTER COMM'ANIDANT OF THE MARINE CORPSBethesda, M~D 20084 Code AX
Washington, DC 20380Mr. J.H. McCarthy, Jr. (Code 1552)DAVID W. TAYLOR NAVAL SHIP I Librarian Station 5-2
RESEARCH & DEVELOPMENT CENTER COAST GUARD HEADQUARTERSBethesda, MD 20084 NASSIF Building
400 Seventh Street, SWMrs. Joanna Schot (Code 1843) Washington, DC 20591DAVID W. TAYLOR NAVAL SHIP
RESEARCH & DEVELOPMENT CENTER 1 Mr. M. PitkinBethesda, MD 20084 Assistant Administrator for Commercial
DevelopmentILibrary (Code 56141) Maritime AdministrationDAVID W. TAYLOR NAVAL SHIP 14th & Constitution Avenue, N.W.
RESEARCH & DEVELOPMENT CENTER Washington, D.C. 20230Bethesda, MD 200814
1 Mr. R. KissICode 03 Director,Office of Ship ConstructionNAVAL AIR SYSTEMS COMMAND Maritime AdministrationWashington, DC 20361 14th & Constitution Ave., N.W.
Washington, D.C. 202301Code 03B I Mr. P. ThurstonNAVAL AIR SYSTEMS COMMAND AIR FORCE OFFICE OF SCIENTIFICWashington, DC 20361' RESEARCH/A
Building 410ICode 310 Boiling AFBNAVAL AIR SYSTEMS COMMAND Washington, DC 20332Washington, DC 20361
1 Chief of Research & DevelopmentICode 5301 OFFICE OF CHIEF OF STAFFNAVAL AIR SYSTEMS COMMAND Department of the ArmyWashington, DC 20361 Washington, DC 20310
ISTRATEGIC SYSTEMS PROJECTS OFFICE I Dr. G. KuhinDepartment of the Navy Fluid Mechanics SectionWashington, DC 20376 NATIONAL BUREAU OF STANDARDS
Washington, DC 20234IMr. Norman Nilsen (SP 2022)STRATEGIC SYSTEMS PROJECTS OFFICE 1 NATIONAL SCIENCE FOUNDATIONDepartment of the Navy Engineering DivisionWashington, DC 20376 1800 G Street, NW
Washington, DC 20550IOCEANOGRAPHER OF THE NAVY200 Stovall Street 1 Science and Technology DivisionAlexandria, VA 22332 LIBRARY OF CONGRESS
Washington, DC 2051401Commander
NAVAL OCEANOGRAPHIC OFFICEWashington, DC 20373
R-2118
Professor H.P. PaoDepartment of Aerospace and
Atmospheric SciencesTHE CATHOLIC UNIVERSITY OF AMERICAWashington, DC 20017
Professor J.V. FoaSchool of Engineering and Applied
ScienceGEORGE WASHINGTON UNIVERSITYWashington, DC 20006
Defense Research & Development AttacheAUSTRALIAN EMBASSY1601 Massachusetts Avenue, NWWashington, DC 20036
Dr. A Thiruvengadam
DAEDALEAN ASSOCIATES, INC.Springlake Research Center15110 Frederick RoadWoodbine, MD 21797
Dr. A.S. Peters
12 Argyle Way RRIHarwich, MA 02645 i
Professor T.V. DaviesDept. of Mathematics
THE UNIVERSITYLeicester LEI 7RHEngland
")
IJ 3v (w,s,O±,tr)dw
C1 Tf(s)
= 1 f( 31(W'S'T)Idw +-L f f [P3 (w,sX)JdwdXY 0 0
1T ;3 I f (a)- -- s7 f f h(a-s) -(A+Y-T-Wu) [P,(w,a,A)jdwdadA
0 o
om which
( y ~ , O ± r ) = ± 4. [ ~ ~ y, , )] 3~ f (s ) [P 3 (w ,s aX) J d w d XLv 3 i 21r 3y 0 0 +T-
4+ .-- -- s f h(a-s) [Pl (w,7;) 3fd-d~d0 - o
*314 $SO±,T) +-t-P (Y ,s,O±,T)
-1 (- +
A-I11
DAT
top related