Oyemtau Polombaol alam 7hu dumuwvval

uresge
Crnlonalny dw(c)-
0 At
Moduh4
AD-2

Relu Cn
tulbu po o mrun (nialulungbtdy horuo)O4 dipnoms
Rala
Rruabron
-

umt fo
Umhnsun CaCy,Cz o»d P
Buloi. a emaumbe fw aho am Amknswn
Thwufore t wnknou OL Ca,CyC ad P bena

alna w
wrlk
l k onbmuy qng mennunn gualibro
ludnLng0u olimuafumuon P, u hau tl
Ucbobyryonuli
C
zCom ahee wrllem o

Um mibuTpue hetu ulahe f PCr)

Anwwwe
 SuwulDu

Auuia hahee sulutien

0
d - "a

romn 3 odt 29 (o ato be tuthu a

a a ay1-

heutol

eal geso of 0ound a

Uanal

1-m C
 (

7.4 LINEARIZATION OF THE POTENTIAL EQUATION

The general cquation for compress1ble lows, namcly q (715). Can be

for tlovw past slender or planar bodies Acrofol, slender brdies of

simpliticd
Iypical cxamples ol slendcr bodies Bodics ike wing.
revolution, and sO on are
hodics These
where one dncnsion is smaller than otheIs, are called planar
disturbances. The acrofoil contour becomes the
bodies introduce small
stagnau0N streamline.
with the exception of nose region. the
For the aerofoil shown in Fig. 7.2,
perturbation velocity w is small everywhere.

flow.
Fig. 7.2 Aerofoil in uniform
 Gas Dynamics
256
Small Perturbation Theory
io
arouu
tne acrofoil in
i. 7.2 to be
componChIs
Assume the velocity, , w

(7 17
velocity components av
nd
main
flow u, 0, w
are the aro
where 1, components along the x,
r
y, and
perturbation
(disturbance)
velocity z
directi
pOstulates that the perturbation vol
respectively
The small perturbation theory
main velocity components, i.c. velocities 3Te
to
small compared

1uV UV W V
(7.1%a
Therefore.
V,= V V, <V, V <V. (7.186
Now, consider a flow at a small angle of attack or yaw as shown in Fig. 7.3. Here
V = V. cos a t u, V, = V. sin d + v

Fig. 7.3 Aerofoil at an angle of attack

Since a is small, the above equations reduce to

V = V+ 4, Vy =

Thus, Eq. (7.17) can be used for this case as well.

With Eq. (7.17), linearization of
Eq. (7.15) gives
(1 -M)0.. + 0,,t =0 (7.19)
neglecting all
higher order terms, where M is the local Mach number. Therefore
Eq. (7.16) should be used in solving Eq.
The perturbation velocities
(7.19).
Let o = 0. + p, where
may also be written in potential form, as folows

Therefore, ) may be called the disturbance (perturbation) potential, and ne

the perturbation velocities are given by

W
(7.20)
dy
With the assumptions of small perturbation theory, Eg. (7.16) can be CP
as

4a 1
-(Y- 1)M (7.21)
 PotentialEquation for
Compressible Flow 257

-(Y-1) M
la Binomial theorem, (d./a) can he
expressed

(Y )M.oM
(7.22)
Cuhstituting the above expresSION for (a../a) in the equation

he rclation between M and M.

may be cxpressed as
(neglecting small terms)

M-12 (7.23)
The combination of Eqs. (7.23) and (7.19) gives

(1- M)o, +Oy t+O

V Mo,o. 1+ M (7.24)

Equation (7.24) is a nonlinear equation and is valid for subsonic, transonic, and
supersonic flow under the framework of small perturbations with u < V
U<Vand w VIt is, however, not valid for hypersonic flow even for
slender bodies (since u V.). The equation is called the linearized potential
=

flow equation, though it is not linear.

Equation (7.24) may also be written as

( -ME)o.+ Oyt M - M M (1-ME)o.. (7.25)

Further linearization is possible if

M. (7.26)
1- M2 V
With this condition, Eq. (7.25) results in

(7.27)
-ME)0.. +0,, +0=0
nis is the fundamental equation governing most of the compressible TIOW
valid, and Eq. (7.26)
is valid only when Eq. (7.26) is
i (7.27)
Cquation 1s
Vand
different from 1. Hence, Eq. (7.26) can
nly when M. is sufficiently (7.24)
be
fa flows only. For transonic flows, Eq.
d n d Supersonic
used. For M = 1, Eq. (7.24) reduces to

(7.28)
o , o,, +o,, +0::
 k e s the IsO ow piolile
The nonlinearity of lq ( 28)
low problems
ditticult than subsonie or Nupersone
Equation (7.27) is clhpue (.0 l t e t 8 Te oNIIVe) fo udi
for / 1 But in both the
bolie (ie not all tems are positiVe)
is T m e a h i s I he adlvantage of L
gOveming ditlerential cqation
EXAMPLE 7.1 Show that with si periInbton theory Assunplion

Y-1)

will reduce to

1-(y-1) M

Solution By Eq. (7.16).

--1
2

where V V are the main velocity components along X, ' and direetions

By small perturbation theory,

V =
V+ u , V, =
v, V =w

direetions
where u, vand w are the perturbation velocities along along x, y and -

Also, u, v and w are very small compared to

Substituting these in to Eq. (7.16), we have

=1--) + u ++w
a 2

Neglecting and w being small, this reduces to

=1- M 2u . u

M12 2

u therefore,
But IS very small,
mullods
Mulhoof Nonlinear Potential Equation
(Solution) o f No

1ods
Vumerical method The nonl1nearity of Fq24)makes ittedious to
(1) the solutron to the cquation
alve the cquation analytically Howevcr,
numerical methods But a numerical solution not is

can be obtaincd by
configuration in a flow
a ceneral solution, and is valid only for specific geomctry
a

and specified
fieldwith a fixed Mach number
Transformation (Hodograph) method When onevelocity compone
() the resulung curve may
against another velocity component,
is plotted relation may be nonlinear
whereas in the physical plane. the
be linear.
for solving certain transonic flow problems
This method is used
these methods, the boundary conditions need
to

Similarin methods In detail

the equation. (This method is discussed in
be specified for solving
in Chapter 8.)

7.6 BOUNDARY CONDITIONS

shown in Fig. 7.4.

in flow field as
Examine the aerofoil kept
a

f(. , E)

attack
at an angle of
Cambered aerofoil
Fig. 7.4
similar to the
body
is
streamline near
the boundary (SC S)
1n inviscid flow the
conditions
boundary
following
O t O u r . The flow must
satisfy the velocities are
tangential
The
conditnon
condition 1: Kineticflow the velocities are teto
bOundary the body
contour
or tinite
Normal to or
Lne
body contour. veloeities are Zero smalt
perturbation
Atz t o , 7.4, vith
in Fig.
Boundarycondition.2: condition for the
aerofoil
shown
follows:
Body
contour

flow as
lduc be
wrilten

Derti
Prturbation assumptions, may .C.
0.
f)
f= /x, (/ gradl
y, ) V,
Therefore,
on the
surlace,

39)

T V c C l o r
is

(V+u)r v
 Gax Dynamics
264

heretore, Tq ) hecome
but n
(0

0, dl'd 0 Then Fq (7 39 reduc

d

two-dimensonal flows,
For
d

"creters to the body contour and (dz dr) Is tho

wherethe subscript series ol z, we get
u and as power
the body. Expressng

u(N. ) u(T, 0) t
d a) =*
+b»z*-
w(.)=w(r, 0)t bz
the functions ot.r. It the body is sufficiently slend-
where a's and b's are

w(r, 0) d:)
Vtu(r,0) dr
i.e. for sufticiently slender bodies,
it is not necessary to fulfil the boun
if the boundary condition on the t
condition on the contour. It is sufticient
of a body of revolution or the chor
of the body is satisfied, i.e. on the axis
an aerofoil. With u/V <1,
w(r, 0)

For planar bodies: ofidy =

0, and therefore.

0)
w. , -dE
and
Ie. the condition is satistied in the plane of the body. In Eqs. (7.4-)
the elevation above the v-axis is neglected.
 FLOW: THE
PRANDTL
TW0-DIMENSIONAL
8.2 FLOW
SUBSONIC
GLAUERT RULE FOR

The Prandtl-Glauert Transformations of

solution
to relate the
possible
Frandtl and Glauert have shown that it is flow
solution.

incompressible
l o w about a body to the

271
272 Is yna
hiller

tuanslommao
o t
wodme
Ihe 01
os
aplee
manner

ndIs noIal the low Jhee

where
s along
the low
duce0
lesCriptioi 0 the equaur
roblem, ne te
prohlen

complete
are
not the loo
however, condiions

boundauy
ADsloncd ilo o
specily
Ihe can be Ia
to and (8.)
ns (8)
following transfomalon

K2
Zinc)
O(r. 2)K2 Pme ner
arc for incomoreee
varables
with subscript "inc
the flow. Combinin
In Eq. (8.3). are
subscript
compressible
lor
without
the vaniables
and (8.3).
We get
Oinc =0
(1-M:)K 2
dX inc

1.e.
0 inc
K (1-MPm+K2C9ne 0
dMinc 0zinc
(8.2) if
the incompressible potential equation
This is identical to
K-1-M
determined from the boundary conditions
Now, K, is to be
theory (see
Section 7.61. we
by small perturbation
For slender bodies,

W dz
V+u dx
poent
be expressed in tems ot
K1. Equation (8.5)
can
Since
ul
function as

.N
0

d
Wine
ne
Vn
Also, by Eq. (8.3),

(00 - K20 de
0
 r l a r t Pule 273

With this relation and T:qs. (8.6), we get

KK:d
dx
(8.7a)

d
(ds
dK-Md. (8.76)

it is scen that K2 can be determined from the boundary

From Eq. (8.7b),
Conditions.
that the slope of the profile in the
Equation (8.76) simply
means

of the
compressible flow pattern 1Is
(R2-M.) times the slope
corresponding profile in the related incompressible flow pattern.
For further treatment of similarity law, let us consider the three specific
in which the
versions of the problems, namely, the direct problem (Version I),
the indirect problem (Version ll), which is
body profile is treated as invariant,
distribution around the body in
the case of equal potentials (the pressure
flow are taken as the same), and the
incompressible flow and compressible
streamline analogy (Version III), which is also called Gothert's rule.

The Direct Problem-Version

Consider an invariant profile. In this case. there is no transformation of

we have the
at all. For the profile to be invariant. from Eq. (8.7b),
geometry
condition

K (8.8)
yI-M
Therefore, Eq. (8.7b) reduces to
ddinc (3.9)
dx dxine However,
contradicts the original transformation equations (8.3).
Cquation (8.9) since the Prandtl-Glauert
C erTor involved in this contradiction is not large
transformation is valid only for small perturbations.
By Eq. (8.3), we have
(8. 10)
inc

V-M ue

from the body. Sinee

8.3) is valid only for streamlines away
theory.
the error

Prandtl-Glauert transformation is based on small to

perturbation

this, sonc
erTor
co
Cases thickness of the bod'In addition
ncreasing
IS
introduced by the above contradiction (see Eq. (8.9) ody in
compressible

a
flow by
quation (8. 10) shows that the streamlines in
i n c o m p r e s s i b l e

low are more those the body

separated than around
anamoun given by AM In otier words, by the exstence of body n

tlowfield, the streamlines arC more diSDlnced

h
rewble fllow than i
i comprcvsible
ncompressible Mow, as shown in ig. R.I, 1.C, the disturbIcestots.oduced
object are larger in compresSible loV than n 1nCompressible ( and
y t
humber, Ihis 18 so because in Comnra
nerease with the rise in Mach sible fi
as the low pISSCS duc to accodl
OvCr the body
there is density decrcase
whereas in incompressible lOw there iS no change in density at all Th
and hence by streamtL
body there is a drop density,
in
across the
say, aTea
4.4), the strcamtube area increases as the len it
velocity relationships (Section
decreases. At M= 1, this disturbanee becomes infinitcly larue an
an
treatment is no longer valiId.

(a) Incompressible flow (b) Compressible flow

Fig. 8.1 Aerofoil in uniform flow.

Thepotential function for compressible flow given by Eq. (8.3)is

P ine
(8.11a
1-M
By Eqs. (7.20) and (7.48), we have

Cdx C-2
Using Eq. (8.1la), the perturbation velocity and the pressure coefficient ay
expressed as follows:

1
Ox
1 -M OXne
Therefore

ine
VI-M
(N17
C
-M
Since the lift coC and the
moment cocefticient
Cv
pitching
n g moment
intcgrals of C. they can be expressed followine lEq. (8.1 b) as coene

C (S.I1

-M
 milarity Rule 275

(dC d)
d
IM (1d)
a t pate m Compressible low
For a
dC 2/
d ( c)
-M
CNM
CM
(%1
I-M
Cimilarly, we can express the circulation in
compressible flow in terms of
circulation in incompressible flow as

inc
VI-M
(8.11g)
From the discussions I of the Prandtl-Glauert
on version
transformation.
the following two statements can be made:

1. Streamlines for compressible flow are farther apart from each other by
1/1-M than in incompressible flow.
2. The ratio between aerodynamic characteristics in compressible and

incompressibleflows is also 1/1- M

From Eqs. (8.11c) and (8.11 f), infer that the locations
we
centre and centre of pressure do not change with M., as they are ratios between
of aerodynamic
CM and C
he theoretical lift-curve slope and the drag coefficient from the Prandtl
ulauert rule and the measured C and Cp versus Mach number for symmetrical
NACA-profiles of different thickness are shown in Fig. 8.2. From this tigure it
Seen that the thinner the aerofoil, the better is the accuracy of the P-G rule
Or 6% aerofoil, there is good agreement up to M. = 0.8; for 12% aerotoil aiso
grcement is good up to M = 0.8; thus 12% may be taken as the limit of
of the P-G rule. For 15% acrofoil, there is good agreement up
Pcability
oM 0.6 But above this, there is no more any agreenment. However. for
a to

oniCaircraft, the profiles uscd are very thin; from practical pont ot so a

theP-G rule is very good even with the contradicting assumptions

involved.
Mach number, there is decrease in lift. This explainned by can be
Fig. SOme
8.2. There is a sudden increase n drag when the local speed
inereases

beyond the sonic

c speed.
sp This is because at sonic point on the profile,
there S

-shock which giv rise to separation of boundary layer. as shown

Cgives s o n e w here
Mach number which gives sonic velocity
on rcam
the boundary deereases with
Mach number M. M
the eritical
increasingthi called
CSS rati0 of profile, The P Grule is valid oinly up to
about
PRANDTL-GLAUERT RULE FOR SUPERSONIC
FLOW: VERSIONSIAND I

t i o n 8 . wehale seen lhe snmilatily ules ln abson l Nty let

mlanty rules for supeisonme lows We ea

N hiseussioms on sImulanty rule lor suibsene coupresauhle flow, thal th

i t o t i n the translomaton I'q. (8.) should have tlie olluwiny rclatioti

on the tloW regime:

dependng
for subsonic low

K=M. -| for supersonic flow

in general, we can write

Therefore.

(8.16)
K--M
difference between the treatment of supersonic
However, there is one important
cannot find any incompressible flow in
the
low and subsonic flow, i.e. we
supersonic flow regime.

Subsonic Flow

We know that for subsonic flow the transformation

relations are given by
Eq. (8.3) as
XineX, inc K z, = K2 ine
The transformed equation is
+
Kil0--)hnel =0
K2[(1 -I (.)ne
be identical to
be satisfied by this equation in order
to
the condition to
and
Eq. (8.2) is
K, - 1 -M
inis case the above transformed cquation becomes Liaplace s cquation.

Supersonic Flow
i e Iransformation relations for supersonic llow are

iun
The l l l
the trnslomcd vanables.
Dies wilh "prime*" are
, ake the M:iclh unber
al.

ncse transformations is to

cqualion (8.1) to vanish.

With the above lransloralon rclations, the gv

K1(1 M:, K^i

 so the
above cquation becoes
low, Ar T, and
For supersonic

we can sCe
that the Mach number M
cquatio.
By inspection of this
eliminated from the
above cquation witlh

The cquation becomes

- =F0 (8.17a)
Now we nmust find out as to which supersonic Mach number this flow belongs
The original form of the governing differential equation for this kind of
low. given by Eq. (8.1), is

(M-1)0x =
0 (8.17b)
For Eqs. (8.17a) and (8.17b) to be identical, it is necessary that

M. 2
By following the arguments of P-G rule for subsonic compressible flow, we can
show the following results for versions I and Il of the Prandtl-Glauert rule for
supersonic flow.

Analogy version I For this case of invariant profile in supersonic flow,

K2
M-1
Compute the flow about the given body at = 2 . For any othcr
supersonic Mach number, the aerodynamic coefficients are given dy

CM = (8.18a)
C'M M-1
Analogy version I Here the requirement is to find a transformation
i
profile, by which we can obtain a body, for wlhich the governing
cqu
Eq. (8.17a) with exactly the same pressure distribution as the actual or
which the goveming equation is
Eq. (8.17b). For this,

The derivation of the above two

results are left to the reader as an c
From the above results, we see that in supersonic low M.. = 2 plays tn Sannc

role as M =
0 in subsonic low
For version 1I, we can write

C (8. 18b)
C
 'rlar t/ Pule 281

Gothert rule For any given body, at yiven Mach

l:
Analogy , version
versi
the rule
Analogy ranstormcd shape by using
find the
her M..

M- (9)

and 7 are h e cambcr

and thickncss of the g ven
the angle ol atlack, / and the
here (d 1s
a r c lor the
transformed body
The primcd quantilies
hody, respectively.
lor the actual body
unprimed o n e s are
characteristics of thc transformed body for
acrodynamic
Compute the
characteristics of the giVen body at the gi ven Mach
2. The acrodynamiC
M
number M.. follow from
CM (6.20)
C C
C C'M M-1
a
floWs by usSing
subsonic and supersonic
We can state the Gothert rule for

modulus:|l -M
subsonic and
compressible
Om the discusSIons on similarity rules for
ready-made
subsonic flow therc 1s
a
it is clear that in
upCTSonic flows, cases we can
use the Prandu
Hence for such
ncari/cd solution for M. 0. =

rulc. But for supersonic flow the lincar theory

cquations are Very simple
uT rule.
ncrClore, we can convenicntly use the Gothern
 KARMAN RULE
FOR TRANSONIC
8.4 THE vON
FLOW
case of two-dimensional transonic
(7.24), for the present
The potential equation
tlow. reduces to

(1 M-)0 +
: M Mo. (8.21

form due to Sprieter (see also Liepmann and Roshko

Equation (8.21) results in
a

1963)) for M.. 1, as

=

2/3

C
(C
l(y+1)M TC ) (8.21a

where
1-M (8.21b)
2/3

l(7+) M
and C, 1s the similarity pressure coefficient. It follows from Eq. (8.21a) that u

lift and drag coefficients are given by

23

CL
( C()
(S.21e

(y+1) M3
5/3
( (S.21d
CD D()
lY+ ) M
Equations (8.21a) and (8.21c) and (8.21d) are valid for local as well as t
tie, the fineness

values. Sometimes, instead ot the thickness-to-chord ratio tc. tne

ratio" defined as in Fig. 8.5 is used.

 20

lo
ol an anglo ol atlack
Fig. 8.5 Wedgo

shown in Fig. 8.5,

For the vedge
= 2 tanO
tanG. C
2 c
fineness ratio (at angle of
attack = 0).
is called the
tc
The ratio
tan
(8.22)
tan (0 t a) =

and the "minus" sign is for the

surface
is for the upper must use the
where the "plus" sign and Cp. we
For finding the local values
of Cp, CL
lower surface.
defined by these equations.
fineness ratio

Karman Rule
Use of
other affinely
we can find solutions for
know the solution for
one profile, 8410.
If we
NACA profiles designated by 8405,
related profiles. For example, the only the absolute
same nose radius etc.;
distribution,
8415 all have the s a m e extended to transonic
floW range
This rule c a n be
is different.
magnitude of t/c transonic range,
the aerodynamic
8.6, it is s e e n that in the values to
as well. From Fig.
that the proper
with Mach number, so
characterS change very quickly
and Cp
are X, CL. Cp
and C; instead they
be considered are not M, C, Cp
w e c a n make
the following remarks
From the discussions made so far,
equation
flows, the goverming
1. For subsonic and supersoniC
trom
that the results
0 is independent of
so
+ =

-M tlow, tne
be applicd to any gas; but for transonie
Similarity rules can
Therefore, the results have
equations are not independent of y y for
polential suitable correction
to be
properly applied to different gases, with steam.
transonic range c a n
be calibrated for
C8a probe used for air in
2 For transonic flow,

CpC :-(
For subsonic floW,

C-