NAVY DEPARTMENT
THE DAVID W. TAYLOR MODEL BASIN
WASHINGTON 7, D.C.SUPER-CAVITATING FLOW PAST BODIES WITH
FINITE LEADING EDGE THICKNESS
by
J.. N. Newman
RESEARCH AND 'EVELOPMENT REPORT
SEPTEMBER 1956 Report 1681
Lb
y. Scheepsbouwkúnde
Thhsh Hccdo
Deift
SUPER-CAVITATING FLOW PAST BODIES WITh
FINITE LEADING EDGE THICKIJESS
bT
J0 N0 Newman
TABLE OF CONTENTS..
Page
Introduction 1
T,heory . 1.
Source Distribution Solution .. 2
Deterrninatiòn of' Lift anOEDrag 6
Lift and Drag at Zero Cavitation Number :io
Appilcation'to Foils with Finite-Leading
Edge Thickness 12
Comparison of.Reults With Known Data 13
Discussion of the Resúlts 16
Recommendations . 17
References . 18
SUMMARY
STATEMENT OF THE PROBLEM
The prob.em tO be considered is the effect of increasing the leading edge thickness of a super-caviating foil bèyond the cavity thickness, such that a portion of the back near the leading edge. becomes wetted andcav1tatjon does not commence until some point downstream from the leading edge.
FINDIÑGS .
The investigation has resulted in an approximate. solution for the lift and drag on a foil with a finite
leading edge0 The solution shows that there is a very
large decrease in
the
lift-drag ratio due to the wetted portion of the back, and as such it seems advisable whenever possible to avoid using a foil with a thickleading edge. It was not determined however, at what
angle of attack this type of flow, would occur for a given foil with some arbitrary leading edge. thickness.. RECOMMENDATIONS ...
It is recommended that the above-mentioned type of
flow be avoided by.keeping the thickness of the leading edge less than the cavity thickness0 It. is also
uggested that further research be devoted to determining the
critical angle. ofattack at which the back will be
par-tially wetted, and to determine an exact solution by a
CO1O;màpping.tchn1quee
NOTATION
b., A constant, undefined in tems of physical parámeters CD Drag coefficient =
CL
A
Lift Coefficient
Drag
D1 Drag component, due t& the back
Drag component due to the
face
L Lift
L " Lift coiponent due to thé báck':
L2
Lift
component due to theface
CaVity length measured from the leading
edge
m ' .StÑngth of a source distributionp LOCàI static pressure
CavIty pressure
PO Pres3ure' on the body
Static pressure of the stream at infinity
R. ., ''Radiu5 öf
the
arc of the' 'Circular Arc SectiOn,Re ' Denotes "the real part of"
r ' s Ratio'.. otupper to lower chord lengths:; r
'.2
Chord length
Chord length of' the back (upper
suiface) measured
to, the point of cavitation' inception- Chord length of the face (lower surfáce)
t ' Dummy variable
The x.component of the velocity on the cavity wall
The x-component of the perturbation velocity Tie x-component of the peiturbatiOn velocity
wM.ch is induced by the cavity source dietrbutioñ. u The i-component of' the perturbation velocity which
.18 induced by the body soirce distribution V . The velocity of the fluid at any point in
the flow field
The local perturbation velocÍty... . - u..
y The y-coinponent of the perturbiOñ e1ocity The y-component of the perturbation velocity on, the cavity .
The y-compoùent of.the perturbation velocity.,
.onthebod7
.Aspacé cóordinate parallel toU A dummy variablò' . :.
T. Aspace. coordinate, orthogonal to the x-directIon
:
Y-or.dináte of the.f oil surface !-Ordlnate of the back
t. Y-ordinate of the face.
Z ..
duy variable
The slope of the back When it is constant (iip to s
The slope of the face when it is cøn5tant
:F1uld density Cavitation number, R0 -P0
=ptJ
A dummy variable VelocitY potential L. dummy variable. .ABSTRACT
An approximate method is develóped for determining the cöefficients of lift and drag.ror a foil whO5e face
1 completely vetted and whose back Is wetted up o 'a
certaIn point 'and 9avltating. downstream of that point,'
at zero àvItat1On flnber. 'The "so].ution obtaInedi$
compared Withexc Qryfor flow past á flat plate,
with exaej theQy
for 1lo
st a non-symmetrica]. wedge,I NTRODUCT1 QN
The effects of cavitating flow past flat plates and thin,foilshve been the subject of recent research by Thlin and Wu. Both developed linearized theories for the flow and solved the problems by conformal mapping. In this way they determined the lift and drag of the cavi-tating hydrofoil as a function of the drag and moment of
the equivalent airfòil,
The limitation that the thickness of such a foil
be less. than the cavity thickness presents practical
difficulties near the leading edge where the cavity approaches zero thickness and leads to vibration and buckling of the foil0 It was, therefore, considered desirable to investigate the effects of increasing the leading edge thickness to some value greater than the
cavity thickness. For such a section the theory based on thin foils will' not hold since the back Is no longer fully cavitating. Consequently, an attempt has been
made to solve the problem using potential flow theory in manner similar to that used by Tulin to
Investi-'
gate the flow past a symmetrical wedge30 THEORY
Two methods have been used to develop approximate
solutions for cavitating flow. The most popular is the
determination o a mathematical model for an equivalent flow. Blrkhoff
and Gilbarg and Rock5 give accoünts of early work along these lines to determine the flow past
fla't plates perpendicular to the stream. More recently
Tulin and Wu developed linearized solutions for flow pst
thin lifting surfaces. The other method is the application
of linearized potential flow theory to determine a source distribution and from this to determine the velocityatd pressure distribution on th surface of the bodya This is the method used by Thilni to determine the cavity shape and drag of a symmetrical wedge.
Theequivalent airfoil was not used In the preset problem bcause of the àdditional singularity at the
leading edge which is not present in the thin foil problem. Çonsequently, linearized potential flow theory was used in
a maìiner similar to that of Tu11n30
1References are listed on Page 18
-1-The presence of a finite leading edge thickness prevents
the inception of back cavitation at the leading edge, Back cavitation does not commence until some point
Sj
and cÒntinues
from' there to the xtremity of the cavity at some point L. For sufficiently low cavitation numbers the cavity length will. be large relativè to the idwer chord length, a condition
necessary for the development of the theory0
It is assumed that no cavitat'ion occurs on the pressure side until the'terrnination of the chord length,. 2, and the cavityextend.s out to (Figure 1) It Is a10 assumed that
the slopes Of. the upper and lower faces,dY ard .dy2
dx
respectively9 and the chord lengths s. and 2 are such that no cavitation occurs.on either the face or the back In the
intervals O x s and O x s,, For this to be true
the Stagnation
po1n must
be at th leading edge0The problem then resolves to that of finding
a soure
distribution which will produce this flow, and to déternzine from it the pressure distribution
along
the wetted pOt1On of the face aridj back. :,The pressure on the backbeyöfld the
point S1
(I.e0
the portiOn of the backlying Insidé the
cavity) is assumed to.be at the cavItypressure,
CônSeuet1
the pressure at any poInt on the surface of the foI1wi1i be IQiown an.d...Dy suitaile lhtegratiorj the lift nd drag caribèdetermined. i.:
SOURCE DISTRIBUTION SOLUTION BOUNDARY CONDITIONS
On. the wetted portions of the face and back
the sreamÏIne slope is equal to the slope of the body:
-()
t)-.«Mfcf
(1) L/...
oc.. J v(Y, y)fj -
í !2LVc
o()-___cjy)r ex.. U..'u.(cjy)U
L Ve.. Li.öhthecavity'wfi the ròsiì.ipand using
BérnouflPs equation.this determines the avitation number as a Í'tthetion of the cavity wall velocity and the stream velocity:
2u.(x3y)
Ö(vYc
(Jo0 Uøo
(&') o(')
where Equation (3) holds over both surfaces. When linearized1thése three equations give:
v(xOt)
o<x<s,
lJ.y(O)
2u(XO+) ¿ 1)002at(iO-)
u00The remaining bounda7 condition is that at the two
points where cavitation commences the flow must-be.. smoöth.
Consequently, the slope of the cavity must be coñtinuou with the slope of the body0 This dondition is satisfied by
eliminating solutions for the source distribution which g.o to infinity at these poinis0
SOLUTION
The following solution is based on the simplified assìimp1on that.,1th the s'agna-t1on point at the leading
edge, the
1o'
ô the face i& indépndent of the geoetry of the baák. Conse.que.rtly, theflow on each surface is. assumedtobetbe sainé-tas that oria symmètrical body ofthèsame slope and chord length as the surface in question. The lift is then assumed to bé the difference between the lifts of the two surfaces considéred separately, and the drag is
assumed to be the sum of the two drags. This is a crue assumption since it assumés that the stagnation streamline will be horizontal as it approaches the leading edge, and
this will hot be the case unless the foil Is symmetrical, In which case of course therewIll be ño lift. However, this procedure should give an approximate solütlon and
was followed0
-.3.'-(3)
(la)
Since the
two
surfaces can bé treated sepárately,'. thed.evelopmnt is based on a single surface of slope dy0/dx
and
chord
iength s, and which may be applied to either theface or the back0
4 distribution of sources of s°trength m(x) along the
x-axis
'in the interval O<x gives the following:and thus: if:
g(x,y) J
d(
In w ¿Ji.' 2vr o L. r8 tn6rUx-w) dx'2irI
vfr) y)
.L
. ..Lf
1nfr)y dx'
y
2cr]
oAt points o the
xaxis
the perturbation velôcitiel u(x,O)and v(x,Q will be:L
u(xO) 2ct11 ,-x'
o
o)
tm()
The' boundary
conditions on the x-ax.1 will 'be satste4t (x'x"
j)(-)('
s where :4v(x_xI)2?
ci.o )i.-X'-
-'I2Ù.dx'
iCOr4)= -21h y's)
': '(12)
dxnd thüs
I[2t/'fr)d' 'Sfai(xIM'
¿/ç.
2irJ'
x-X' 2ctJ '2
"o)».
Using the thin airfoil thory inversion formula6 gives m(x) from Equation (il):
I
1Tqxç)(ex)1f
(xix)
=
-2s)(y
fSç/
jd1+]
To satisfy the juncture conditin the term in brackets must vanish at xs: therefore: £
Q=b1rUof
s(J_i))'+
k[woo2
s.
-!fI'
x)fr Lico
EvaluatIng the integrals as shown
2V
J14-K J0 49é I' ---
(x-i)
For.the case of zero cavitation number,
0=
Q,.0 =and,,00
.(13a)
rn(4
Ç13b)2UCfsfo
dtDETERMINATION OF LIFT MID DRAG
DRAG
Assuming the foil to be deeply submerged in an idea].
flow (i.e., neglecting frictional and wavéresitance) the drag on one surface will be given by the integral:
p.c
J
(po r)
X.? dt p[U+ v]Ji - St.(11i-)
u:
dto
dt
o
which may also be written:
- i
t=
r[ut]Í
í()
¼t)
pj
oD
and after linearization this becomes:
O t [u.(-t Y) - u ( c)l '!Z cl 4.
Jo,
-Idtbut
u(t,y0
=u0_0(,t,y0) +
Where u_0(t,y0) Is the x, component of. that part of the: dis
turbánce velocity òn the body
which Is
Induced by the c8vity source distribution and1_0(t,y0)
is the x COmponent of that part of the disturbance velocity on the body whichis .induce4
by the cavity source distribution. . . ..
sInce
..
J0 (yò)°
c.i.to
the equation for drag becomes:
p =
pVc. cr LJ0 o (s)- () yo)
dye
(1+c)
The Integral term ir (11t-c) is evaluated as follows:
using equations (7) and (l3a)
()
r,
2-'-dt
J 2n'
I o[Ç
i.
+i/if4ic:t
-7-( s)UsIñg Appendix A the first term of Equatioñ (15) is
evaluated: s
_!1tffd
(J 15d
-iJ ;;
Oy
()
2.wj d
o SAnd using Appendix B the second term of Equation (15) is:
2 d 2 L
i If
¿.\/(-sX
- e) 'J Lotherefore:
Ie
*>)
)it d*.
dt
The lift is calculated in the same manner as wa
drag. Assuming small axtgles o attack, the lift of
face wL11 be r
In Equation (17) and hereaftér, the. following convention lsused* ina.flo from left. to right(F1gureJljft
up ia
6sitivé;slopeu
ndio
d.slópëidö'ntand.:toj ti è.r1ght.is .Lpositie
ôrponding4oùthLuuai
angle of attack0 . .
dt
d.
1a)
And Equation O.c) becomes:
12= p(W
-the drag coefficient, CD
=
+
Linearizing as in Equation (l+) gives:
L =
PUI
rand. u(t,y0) = 0_0(t,yo) +
The integration that follows is consideratll simplified by making the second assumption; namely, that:
) J ¿-w' 6-c'
t
=
opvcos 9t1oed
Equation (17a) thus becomes:
L
Oc.ffrUoQs
fo.o (iyJdJ
(ib
and proceeding as for drag:
r'
f
S.
f
f
(od'
1T
-l.d
-
Yt-
,e-t.J
J7
S8 +
The fi'st
termof Equation
(18) is evaluated fromAnd the second terth becomes:
2u
f 5!th,1'.tí1éY
-JI
t)
/
(
o
The e quátion for Lift is then:
L =
_/]
--.10'3d.
+__
o-
2p
L1C f/4'
(i)
d-r
ir
JV'(#-)(s-t)
AtcL:LIFT AND DRAG AT ZERO CAVITATION NT314BER
The expressions for lift and drag can be considerably simplifiedby assuming zero cavitationnumber.,
Thus the.
cavity length is infinite and the-ve1ocity on the cavity wall equals the stream velocity (UU00 ).
Using Equation(13b) and proceeding as before, th expressions for lift and
drag become: -
L:ff,0coV'f5droíz
-- cit_!'t
dz'e/é(19)
dt
i J,) ir(t-'.)j dr frt)y.
°s
o (1.74) (18)°°L(
4 ..j)(-
yj'J
t-s
where in the last equation the variable has beén changed
from t to t, to conform with the other equations0
Equation (19), likéwise, becomes: 90.
roL.
!IJ_)/vI '4_i)t.0 1t
D=
rj0
(
And. using the
procedure of Appendllx
L0
dt-fi
Dividing Equations (18c) and (19c) by l/2 gives the lift
ánd drag coefficients: S 9(/oo I I
= - ;rJ
dJ0
(-t-À
t(
1 111Jod..
L
f
Sf øo °! I yo I ,Wt ' chL
j_ (°7ff.51
x'-s.
-i)
_)ôI('d.td:
/-s I 1!. JOG1V
cii]
g+j')jC
J0dtY;.
d4
.2. (18b)
(.18c)
Ç19c)
and changirg the order of 'integration,
CL
Cb
(18d)
Equation (18d) can be simplified by .furthêr linearizatiön 'using the seri?s:
Neglecting al]. but the first-power term In
the
series gives:I- irs) dtL'/S_t
b
Final llnearized l±ft and drag coefficients at zero
'cavIttìóñ number are then:
«z' U'J. 2(s-j+fr(:..tj
Fçr the case of a
the assumption is made sums of the, components sepàxatelr on the face
.
;_ff(id11
i±i'
.2.,i jJ°d*j
:(4ki)
*Ut]
CD,[S'j2
APPLICATÍON TO FOIÏJ WITH INITE LEADIÑG. DGE THICKNESS AT ZERQ CAVITATION NUM5R
foil .sUòh as.is shown inFIgrG 1,
that the tötal lift and drag are the of thé lift and. drag detèrminé4
and the b8.ck0 ls:
L.L1 + L0{J?Y
( (-t)S)
+j5
i'
o F. 18e (18f) Ç22) (23)where the chord length in the denominator of the' coefficients is s By considering s2 to be unity the expressions may be
further simplified without loss of generality0 Denoting'the ratia between the upper and lower chord lengths
= r we have: = ''
U0
(r
I(
IL0
a:;:
+J''
o4"'
+ If
ay d4.U0 d-tPf'
CONPARISON OF RESULTS WITH KNOWN DATA FLAT PLATE
Since several assumptions have been made, it was
con-sidered advisable 'to compare the results obtained from
Equations (2h) and (25) with known data. The 'simplest: comparison is that of a flat plate where r = O and
= 2o Equations (2+) and
(25)
thus beôome: dx p, 1i f' d.
= irThe coefficients
C = 21TÇhc eòsO..2
, i,4o 4 2.ir0
()
62fróm exact theory8" are:
(2+)
=
o')
' I.57°Thus fo th case of a flat pláte, th
approximate
solution for the ift coefficient is
seen to be low by about
12% and the solu on for the drag coefficient Is low by about
3%o Actually the smaller error in the ase of lift Is due
only to the series approximation, Equation '(18e), which
raises the lift coefficient
ratio. 1n3 to 1, or about 10%. Applying Equation (18d) t9
the. flat plate gives:
.-(27a)
Another exact theory which is
known and
is morepertinent to the foils being considered is that fora
zion-synfinetrical wedge (Figure 2). For this case the exact theory7 give.s:*. C 2a3
(_,)2
D LfrSa ¡oc. C-
(-oi1)in-rr(_i---L
(Ls2the constant à is determined from the integral equatIon:
U $2
.For a thin wedge (-,<7r ).Equation
becóe
(Appendix. C)'QÇ...r(
X)-
Cos CA)) 6iX
dZ
where:and:
I
rj.
(39)1 and(31).
.3_fr
(I_acsti(30a)
(31a)*
Note tha.t the sign of has been changed from that in Mì.1ne-ThÖmson to conform with the convention of this paper. Thusas
shown in Figure 2 has a negative value.This solution also is based on the condition that the
stagnation point is at the leading edge, and the condition for this is that the ratio of chord lengths be:
_____\2.
Z - cos cj)S'i.i
dZ
' (
J4co&t),
fWec(cos_cO)JX
- HCDSt
Using EquatIon (32), the approximate solution
(EquatiOns 2+ and
25)
gives the following results:(33)
'w' dt
-cD=
(3+)
For the purpose of comparing the. exact and approximate
solutions, values of CL, ,
and-'
have been plottedvs. r. (Figure )+). As was the case with the flat plate, the
approximate solutions for 'both lift and drag are somewhat
lower than the exact solutions0 Both coeffic1nts' approach
the exact solutions as r approaches unity, or ,a synunetrtcal
wedge.
A third comparison has been made with experimental data
for a circular arc section,9 .'igure
.3.
The experimental data(?igure 5)
is shown in a plot of 11f t,and drag-coefficients.vs. cavitation number for 8 degrees aigle of attack0 At this angle separation did ñot occur at the':.iead'ing edge, but
in-stead at the discontinuity of the uppe' surface0 Thus the
flow corresponds exactly to that of Figure 1. The experimental points are marked as circles on the graph, and the solid lines represent the predicted coefficients assuming ,ull cavitation
from the leading edge according to Wu's theory.
'Applying thé approximate solutions to the circular arc at 8. degrees angle of attack and zero cavitation number gives
the following results (Appendix D): CL
(32)
Separation at Leading Edge O.29
OO+57
Separation at the Bend
Ol65
OO69l
-15-These points have been plotted on theQ O ordinate
'OEf Figüre
5.
Again the'lift coefficient is low,by about13% when separation occur's at the leading edge'0 Unfor
tunately9experimenta]. datais not available belowr= 0.2,
but the general tì'end of the points seems to be close tó that given by CL +
C0
The approximate solution gives a drag coeff1cet whih is about 10% high for the case of cavitation from the leading edge and the general trend ofthe experimental points is in the direction of C,.
+ CD
Lu"
'2DISCUSSION OF,- TIE. RESULTS
In the developent of the, solution it has been assumed
that the stagnation streamline was horizontal9 thus' allowing
the face and back to be treated separately, and that the
lift and drag forces.4ue to the velocities induced by the body
source d1s,tr1butipn'were zero0 As a result this. solition is,
only an approximate one, the accuracy of which can only..be determined from comparison with iown data, such as is dOnò In the preceding section0 The greatest divergence from
kilown' data is in the case o' the flat plate, where, the
approximate solution 'for drag' is 23% 'less than the exact
solution. Strictly speaking. this comparison should not' be
made since.the stagnation point is not at the leading
ege9
"unless 2Q, in which case the comparison becomes trivial.The difference between the two solutions should not be takèn lightly since it is not own to what extent the' position of
the Stagnation point will 'influence the f maI accuracy. If. nO other conclusion can, be da from this, it can at least
be seen. that it is possible' for .a solution based n completely
different conditions to give an answer which is approximately
correct.
Comparisons with data for foils whe±e the stagnation 'point Is ..at, the leading edge. are more' encouraging. For the
case of 'the non-symmetrical wedge, or equivalently the foil with, flat surfaces', the lift and drag coéfficients aré both
1ow, "but for. . r 0, the maximum difference is only abont
7% for 'both, and as,r increases the difference becomes 'smaller.
There is ;o 'assurance,of coursé, that the same accuracyri11.
exist for foil with curved surfaces arid this fact' is borne 'out by the
comparis9n
with experimental data, although this1n any event, the accùracy of the results is Sufficient for the conclusion to be drawn that the performance of a foil with a partially wetted back is greatly Inferior to a fully
cavitating foil. This. Is shown for the wedge in Figure 6 where Lc is plotted vs0 r In this case when the back is
D
wetted over % of the chord length, there will be a loss In lift-drag ratio of 3O as compared to a fully cavitating
flat plate0 The obvious conclusion to be drawn therefore, is that a partially wetted back.should.be avoided if at all
possible.
BEC OMMENDATI ONS
There are two prificipal 'sho±tcomings in the solution as it now stands0 The first is that assumptions have been
made which diminish the accuracy. and thus the authority of the results0 The second is that the assumption is made that the stagnátion 'poInt is at the leading edgewithout
stating the conditions undrzMchhiiLbe.o0
The approximating assumptions can best be overcome by
solving the problem by conformal mapping. This has been attempted without success but further efforts woul.be
worthwhile.
Determining the conditioñs at which the stagnation point will be at the leading edge, or particúlarly, the angle of
attack f or any given foil, wOuld also contribute greatly to
.*e
value of the solution0 Mime-Thomson solvedthls problem for the..non-synixnetricaLwedge (EquatiQn32),
but no solution has been obtálned fòr curredurfaoes0Ì Ç
Further rsearcb could also be directed towai,'d extending
the soltion'toc6vèr finite cavitation númbersG It is quite
possible that the effect of the partially-wetted back is not a great under such condltions;and this infórmátion would be. most valuablè .f or désign urposes. . .
ACKNÓWLEDGEÌITS
The author is indebted to A. J. Tachmindji for his
constant help and encouragément throughout the investigation.
REFERENCES
1. Thun, N. P., "Supercavitating Flow Past Foils and :StrUts,1 Proceedings.o.f a .Sympostum on Cavitation
in Hydrodynamics, National Physical Laboratory,
September
1955,
H. M. Stationery Office2. Wu, T. Yao-tsu, "A Free Streamline Theory for Two-Dimensional FullyCavitted Hydró*ils," Report No. 21.47, Hydrodynamics Laboratory, CIT, Pasadena, Califox'nia; July
1955.
...
3.
ThUn, M. P., "Steady Tto-D1ninsional Cavity Flows About Slender Bodies," TMB Report83I., May 1953
1+0 Blrkhoff, G0, "Hydrodynamics," Princeton University Press for Universityof Cincinnati,
1950.
5.
Gilbarg, D. and Rock, D. H., "On-Two Theories of Plane Potential Flows with Finit Cavities," Naval O±dnance Lab.. MemOrandum8718,
August 19+6heng, H. L and Rott, N., "Generalization of the
Inversion Formula of TI
iñMrfóil Théory
:of Rational Mechanics and Analysis, Vol. 3,No.3,19514. MUfle-Thomson, L. M., '!Theoretical Hydrodynamics," The
NeMillan Company, New York,
1938, 1950.
: Iamb0 R., "Hydrodynamics," Sixth Edition, Dover
Pub1i-cations, New York,
195
Pardn, B. R., "Experiments on Circular Arc and Flat Plate HydrOfoils in Noncavitating and in Ful]. Cavity F1ow,'! eport No. +7-6, CIT, Hydrodynamics Laboratory,
t
SOLUTION OF ,FREQUENTLY VSED INTEGRALS
Let I(m, n,
S) =
Substituting z
sin2O. gives:
a.
Ib Çb1
d6
I(m)v)
=
J
-cos'e+I- n,',,,I.(m,n, (1
)= 2
for the solution of.
i---1
I
f±'o'X'
J
s
V )Lg x'-x
APPENDIX A
Re
md for the solution of
(
-Is
-l9
.-I='_2.j/'!&
ùc'(j
.tt9)+2e
syt
>»s
f' ('-x)fr-)
4-c)
,es) I(
£S
!(2!)
-'---çP) e- O) e-.j
=t)r J/*.
.x'..s'f'o
EVALUATION OF THE SECOND INTEGRAL OF EQUATION (1g)
.it IViJ;r;
,s.
i. dyet
dv0 L(5.Z!
/' {
fr
(e-s, c-s ¿-s, o) Jit)0
therefore: f
4f7_i_j
$4,
Were:
s
r..!.
¿'
.t-.)e')dv y-r-s X1,/
¿ir'
-jrI.
APPENDIX B dc oi(;
-tç.S
J. I
o -
-
d-\
-t-s< 4tA
I s,4)
nbut:
ÍkC
'.
=É e (
(t)o
_i :;i0
c1(
aJt
dt
X
J is determined as follows:
/((\t.(s(
OJOdt
-c-..j
I d-t .ç .2J J
d-6 Oj'T
o -o Sddt
T
f L
o o(5
I
(-c-Lb-s)
-
(--)(c-s) i
j
2
d-t ck
th! 4
cJ-c
dd
-21
SOLUTION OF EQUATIONS (30), (31), and (32) since.:
=(
SI.yivc')j
\ fv/a(X#4)
and:. Thus:. :°° .'then:
f(cosx-o.c
)siIid
=
_2f 1s14)Va(%'t(A))J1
Assuming a thin wedge, 1, and the integral may be
approximated by:
-
f
.-cq&u», zdx
/-eos
f
and by the same method:
dZ
(j- Cotw)'
/,,I+ cos '(A)
I I
i-
coS (A). APPENDIX C 2 -22-I +co.s ¿A))fra1
2()
r-x)1APPENDIX D
CALCULATION OF LIFT AND DRAG COEFFICIENTS FOR THE
CIRÇULAR ARC SECTION
At.8°
ángle
of attack, the fáce is tangent to the x-axis fat the leading edge:Therefore,. sine the face isá circular arc of radius R:
-lx
= tanQ where O = sin.
-dx
.. .=.
tan(sin)
.And:
= =
tan(8°tan1
O.1i00)
dx
ll62
Also for equations (21i-) and (25) to hold,
2
must
be unity which determines R:R
=1:
=and: r
= ____
= _____
= O)+88Equations (2+) and (25.). become:
CD..= i frr
d'/12
1Uo
YrxJ
L0
Ir'
j0 3fjz_x 2:-x)
vrr
The first integral in each equation is easily evaluatê4.to give CL1 and.. CD]: : CL1 = P
-O, 133
= .23 -=Oli9'1.
(2+a) (2 5a)But thé evaluation of the integrals. for and
considerably more difficu]t. Consequently, they were
siinp]4fied by assuming constant.. Since the
maximum value is R =
359
and the minimumvalue Is ij' =the exror invOlved is small.
he aver&ge value was taken as 3 2.
huss 4
-(I
wdx__
T J0 2fr-r)4 ,ri-;:, 'j2
O298 -. o
= + OO23)+OO69l
xs,
yco A
X
Figúre i -. Öiagrani of the Foil In Super-Cavitating Flow
400
8.550 R
FIgure 2 - The Foil as a Figure
3 -
The CircularNon-Synunetrical Wedge Arc Section
.25...
2
2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 r
Figure 4 - Lift and Drag Coefficients for Thin Non-Symmetrical Wedges
26
- - Approximate Mime-Thomson
Solution
-1.0
0.9
0.6
0.5
0. 4
0.3
0.0
KEYo c)
FRI EXPEHIMENFAL DATA WITHSEPARATION FIIOEI THE DISCONTINUITY
V CD) IN THE UPPER SURFA
- CL PREDIC11]J FR! YIU' S THEÇY FUI
CD SEPARATION FRCTvI TIE LEADING EDGE
ODEFFICIErSifS BASED ON
FINITE-LEADING EDGE THICKNESS SOLUTION
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
CAVITATION NUMBER, o
Figure
5 -
Comparison of Approximate Solution with ExperimentalData for Circular Arc Section at 8 Angle of Attack
27 /
t
/
o/
/
o/
/
/
/ /
/
o/
/ /
'I
L2 Ö0.2
0.1
1.1
1.0
0.9
0.8
cl4
0.7
0
0.6
0.4
0.3 0.20.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Figure 6 - Lift-Drag Ratios For Thin Non-Symmetrical Wedges
28 I "
\'
\'
-Approximate Milne-ThomsOn Solution
-\
N
INITiAL DISTRIBUTION
Copies
8 Chief Bureau of Ships, Technical Library (Code 312),
for distribution:
5 -
Technical Libraryi - Civilian Consultant to the Chief (Code 106) i - Preliminary Design (Codé +2l)
i - Propeller and Shafting (Code 55+)
2 Chief of Naval Research, Fluid Mechanics Branch,
Code +38, Attn: M. L. Tulin
Director,- Ordnance. Research Lab., Penn State
University, UniversityPark, Pa.
1 Commander, Naval Ordnance Test Station, Pasadena Annex, Pasadena, California
1 Experimental Towing Tank, Stevens Institute of Technology, 711 Hudson Street, Hoböken, N.J. 1 Head, Department of Naval Architecture aùd
Marine Engineering, MIT, Cambridge
39,
Nass.,&ttn: Prof. L. Troost
Director, National Advisory Committee for Aeronautics, LangleyFleld, Va., Attn: Mr0 Parkinson
1 Director, National Advisory Committee for
Aeronautics, Washington, D. C.
i Gibbs & Cox, Inc., 21 West
St0,
New York 6, N. Y., Attn: Dr0 S. F. berner1 Director, Iowa Institute of Hydraulic Research, State University of Iowa, Iowa City, Iowa
Director, St0 Anthony Falls Hydraulic Lab., University of Minnesota, Minneapolis 11f, Minn. 1 Director, Hydrodynamic Lab,', California Institute
of Technology, Pasadena +, Calife.
9 BrItish Joint Services Mission (Navy taff), P. 0. Box 165, Benjamin Franklin Station9 Washington, D0 C.
3 Canadian Joint Staff, 1700 Mass. Ave., N.W.
Wasblngton.6, D. C.