You are on page 1of 90

Chapter 11

Compressible Flow

Introduction
z

Compressible flow variable density, and equation


of state is important
Ideal gas equation of statesimple yet
representative of actual gases at pressures and
temperatures of interest
Energy equation is important, due to the significant
variation of temperature.

11.1 Ideal gas relationship


P = RT
For ideal gas, internal energy u=u(T)
du
u
constant pressure specific heat:
c = ( ) =
du = c dT u2 u1 =

T2

T1

c dT

-For moderate changes in temperature:


u2 u1 = c (T2 T1 )

dT

Enthalpy h=h(T)
h=u+

= u (T ) + RT = h(T )

constant pressure specific heat:


dh = c p dT h2 h1 =

h
dh
cp = ( ) p =
dT
T

T2

T1 c p dT

-For moderate changes in temperature


h2 h1 = c p (T2 T1 )

Since h=u+RT, dh=du+RdT


or
dh du
=
+ R c p c = R
dT dT

k=

cp
c

( 1.4 for air)

Rk
R
cp =
and c =
k 1
k 1

Entropy
1st Tds equation
Tds = du + pd (1/ )

Qh = u +

dh = du + pd (1/ ) +

Tds = dh

dp

dp

---2nd Tds equation

du p
dT
R
ds =
d (1 / )
+ d (1 / ) = c
+
T
T
T
(1 / )
c p dT R
dh (1 / )
=

dp
dp =
T
T
T
p

For constant c p , cv :
s2 s1 = c ln

T2

+ R ln( 1 )
T1
2

T2
p2
= c p ln R ln( )
T1
p1

For adiabatic and frictionless flow of any fluid


ds = 0
or c ln(

or s2 s1 = 0

isentropic flow

T2

T
p
) + R ln( 1 ) = c p ln( 2 ) R ln( 2 ) = 0
2
T1
T1
p1

R
ln( 2 ) = R ln( 2 )
1
k 1 T1

k
T2 k 1
( )

T1

=(

T1

=(

Isentropic
process path,
Pvk = const

2 k
)
1

T
p
T
p
kR
ln( 2 ) = R ln( 2 ) ( 2 ) k 1 = ( 2 )k
k 1 T1
p1
T1
p1
k
T
( 2 ) k 1

Comparison of isentropic and


isothermal compression

Pv = const

2 k
p
p
) = ( 2 ) k = const, for isentropic flow
1
p1

(11.25)

11.2 Mach number and speed of sound


Ma =

V
, V --local flow velocity, c--speed of sound
c

Sound generally consists of weak pressure pulses that


move through air.
Consider 1-D of infinitesimally thin weak pressure pulse
moving at the speed of sound through a fluid at rest.

fluid at rest

Observer moving with control volume

-conservation

of mass

Ac = ( + ) A(c V )
c = c V + c V

V = c
-linear momentum conservation
c cA + (c V )( + )(c V ) A = pA ( p + p ) A

Q ( + )(c V ) A = cA

(continuity)

c cA + ( c V ) cA = pA

V Ac = pA V = p / c
Q V = c

c =

p
c

(continuity)

c2 =

p
p
c=

Isentropic
process path

-Alternative derivation using conservation of energy


instead of momentum equation
V 2
p
+
g z = (loss) --from (5.103)

(loss) 0 for frictionless flow; and (z) 0

p (c V ) 2 c 2
p

= 0 V =

2
2
c
p
= c V

= V = c (from continuity: V = c )

or

c2 =

P
c=

Assume the frictionless flow through the control


volume is adiabatic, then the flow is isentropic.
In the limit p p 0
c =

(11.34)

- For isentropic flow of idea gas

p = c k

p
p
p
k 1
k 1
= k k
= k = RTk c =

= ck

RTk

- More generally, use bulk modulus of elasticity


p
dp
Ev =
=

d /

c =

Ev /

- For incompressible flow

Ev

(11.36)

11.3 Category of compressible flow


- effect of compressibility on CDof a sphere

Why CD increases with Ma?


Can you explain physically?

(from S.R. Turns, Thermal-Fluid Sciences)

- Imagine the emission of weak pressure pulses from a


point source

r = (t t wave )c
where t present time, twave time wave emitted

- stationary point source

- moving point source V< C

- source moving at V=c

- source moving at V>c

- Category of fluid flow


1. Incompressible flow Ma 0.3
unrestricted, linear symmetrical and instantaneous
pressure communication.
2. Compressible subsonic flow 0.3 < Ma < 1.0
unrestricted, but noticeably asymmetrical pressure
communication
3. Compressible supersonic flow Ma 1
formation of Mach wave, pressure communication
restricted to zone of action
4. transonic flow

0.9 Ma 1.2

5. hypersonic flow Ma 5
Example 11.4 Mach cone

(modern aircraft)

(space shuttle)

Mach cone from a rifle bullet


from Gas Dynamics Lab, The Penn. State
University, 2004
from M. Van Dyke, An Album of Fluid Motion

Video: Mach cone of an airplane

(Note the condensed cloud across the shock wave.)


(Can you estimate the airplane speed?)

Ma=0.978

11.4 Isentropic flow of an idea gas


- no heat transfer and frictionless
11.4.1 Effect of variation in flow cross-section area
- conservation of mass

m& = AV = const.
- Conservation of momentum for a inviscid and steady flow
0
1
2
dp + d (V ) + dz = 0
2
dp
dV
=

V
V 2

Since m& = AV = c, ln + ln A + ln V = c
d dA dV
differentiation
+
+
=0
A V

dV d dA
dp

=
+
(=
)
2
V
A

(11.44)

dA
dp d
dp
d V 2
=

=
(1

)
2
2
V
dp
A V
dp
V2
=
(1
)
2
dp / d
V

(11.45)

p
V
=
Since c =
,
and
Ma

dp
dA
2

(1 Ma ) =
2
A
V

(11.47)

dp
dA
1
dV

=
=
2
2
A 1 Ma
V
V

(11.48)

diverging duct

dp
dA
1
dV

=
=
2
2
A 1 Ma
V
V
converging duct

Combining (11.44) and (11.48):


d

dA dA
1
+
=

A
A 1 Ma 2

dA Ma 2
=

A 1 Ma 2

(11.49)

(11.49): For subsonic flow, density and area changes are in the
same direction; for supersonic flow, density and area
changes are in the opposite direction.
dA
A
= (1 Ma 2 )
dV
V
dA
W hen M a = 1
= 0 The area associated with Ma=1 is either
a minimum or a maximum.
dV
From (11.48):

impossible

Therefore the sonic conduction Ma=1 can be obtained in


a converging-diverging duct at the minimum area location.

For subsonic flow converging diverging nozzle


For supersonic flow converging diverging diffuser
throat
subsonic

throat
supersonic

subsonic

Converging-diverging nozzle Converging-diverging diffuser

What is the physical reason that supersonic flow


develops in the diverging duct as long as sonic flow
is reached at the throat?
P-induced flow
P-driven flow
throat

P0

Vavg

Vavg

Pe

Why can V only be accelerated to c in the converging duct,


no matter how low the back pressure Pe is?

11.4.2 ConvergingDiverging Duct Flow


- For an isentropic flow
p0
p
= constant = k
k

0
- streamwise equation of motion for steady, frictionless flow
V2
+ d ( ) = 0, dz neglected

dp

p 10 / k

dp
V2
+ d(
)= 0
1/ k
2
p

p0
1/ k
1/ k
( p0 / 0 )
Q k = k = p
0

k 1
k 1
2
k p01/ k
V
[ p0 k p k ]
=0
2
k 1 0

k p0 p V 2
[ ]
=0
2
k 1 0

-For an idea gas


p0

= RT0 ,

= RT

kR
V2
V2
kR

[T0 T ]
= 0 or c p (T0 T )
= 0 (Q c p =
)
k 1
k 1
2
2

V2
h0 (h + ) = 0
2
kRT0 kRT V 2
=
+
k 1 k 1 2

where h0 is the stagnation enthalpy


k 1 2
kRT0 = kRT +
V
2

T0
k 1 V 2
k 1
Ma 2
= 1+
= 1+
T
2 kRT
2
T
1
=
T0 1 + [(k 1) 2]Ma 2

(kRT = c 2 )

(11.56)

With

= RT

p 0 T
=
p0 T0

p
p0
0
p0 1k
=( )
Q k = k
0

p p 1k
T
( ) =
p0 p0
T0

p k k1 T
( )
=

p0
T0

p
T k k1
=( )
p0
T0

k
p
1
k 1
]
=[
p0 1 + [(k 1) 2]Ma 2

1
]
=[
2
0 1 + [(k 1) 2]Ma

1
k 1

T
1
=
T0 1 + [(k 1) 2]Ma 2

(11.59)
using isentropic relation

(11.60)

(11.56)

k
p
1
] k 1
=[
2
p0 1 + [(k 1) 2]Ma
1

1
] k 1
=[
2
0 1 + [(k 1) 2]Ma

T
1
=
T0 1 + [(k 1) 2]Ma 2

Figure D1 (p. 718)


Isentropic flow of an ideal gas with
k = 1.4.

V2
c pT0 = c pT +
2
Figure 11.7
The (T s) diagram relating
stagnation and static states.

Figure 11.8
The T s diagram for Venturi
meter flow.

-Any further decrease of the back pressure will not


effect the flow in the converging portion of the duct.
-At Ma=1 the information about pressure can not move
upstream
-Consider the choked flow where at the throat Ma=1,
the state is called critical state

p
p0

Ma =1, critical state (choked flow)


subsoinc

0.528
supersoinc

Critical State:
Set Ma=1 in (11.56), (11.59), (11.60)

p*
2 k k1
=(
)
p0
k +1
For k=1.4

p*
= 0.528

p 0 k =1.4

pk =1.4 = 0.528 patm


*

T*
T*
2
=

= 0.833 or Tk*=1.4 = 0.833T0 = 0.833Tatm
T0 k + 1
T0 k =1.4

*
* p* T0
2 kk1 k + 1
2 k11
) (
)=(
)
= *
=(

= 0.643
k +1
k +1
0 T p0
2
0 k =1.4

Example 11.5
p0 = 101 kPa

0 = 1.23 kg/m3
T0 = 288 K
Find m& = (a)80 kPa, (b)40 kPa.

Critical pressure p* = 0.528p0 =53.3 kPa


(a) pa > p* the throat is not choked
k
p
80
1
] k 1 Ma th = 0.587
=
=[
2
p0 101 1 + [(k 1) 2]Ma
1

1
3
3
k 1
=[
=
kg
m

=
kg
m
]

1.23
/

1.04
/
0
0 1 + [(k 1) 2]Ma 2

T
1
=
T0 1 + [(k 1) 2]Ma 2

T = 269 K

V = Ma * kRT
V = 193 m/s
m& = VA = 0.0201 kg/s

(b) pb=40 kPa < p*=53.3 the flow is choked at the throat Ma = 1
1

1
3
k 1
=[

=
=
]

0.634

0.78
kg/m
0
0 1 + [(k 1) 2]Ma 2

1
T
=
T0 1 + [(k 1) 2]Ma 2

T = 240 K

V = 310 m/s
V = Ma * kRT
m& = VA = 0.0242 kg/s

Figure D1 (p. 718)


Isentropic flow of an ideal gas with
k = 1.4. (Graph provided by Dr.
Bruce A. Reichert.)

Ratio A/A*
AV = * A*V *

A * V *
or
=
*
V
A

with V * = kRT * and V = Ma kRT


*
A
* =

kRT *
1 *
=
kRT Ma Ma 0

T * / T0
T / T0

1
1
+
1
2 k 11
k +1
2
k
1
2
=
(
) [1 +
Ma 2 ] k 1 [
]
Ma k + 1
2
1 1 + [( k 1) 2]Ma 2

k +1

A
1 1 + [(k 1) 2]Ma 2 2( k 1)
*=
[
]
A
Ma 1 + [(k 1) 2]

Figure 11.10 (p. 639)


The variation of area ratio with Mach
number for isentropic flow of an ideal
gas (k = 1.4, linear coordinate scales).

Example 11.8 Air entering subsonically at standard


atmospheric condition, A=0.1+x2
A

1
2

A = r2 r = ( ) = (

0.1 + x 2

1
2

For the flow to be chocked

At throat, x = 0 A* = 0.1
A 0.1 + x 2
p
T
=
, using (11.71) Ma ,
*
A
p0 T0
0.1
d

c
0.98

b
d
0.04

Example 11.9 Air entering supersonically at standard


atmospheric condition, A=0.1+x2
For the flow to be chocked

At throat, x = 0 A* = 0.1
A 0.1 + x 2
p
T
=

,
using
(11.71)
Ma
,
A*
p0 T0
0.1

a
c
0.98

b
d
0.04

Example 11.10 Ma=0.48 at throat, A=0.1+x2


The Ma at throat is 0.48

p T A
Ma = 0.48 ,
, *
p0 T0 A
A 0.1
= * =1.4
*
A A

A* = 0.07
c
b

subsonic-subsonic

subsonic-supersonic(choked)

subsonic-subsonic(choked)

supersonic-supersonic(choked)

Supersonic-subsonic (choked) Supersonic-supersonic (not-choked)

1. For a given (T0, p0), k and converging-diverging duct geometry,


infinite number of isentropic subsonic to subsonic (not choked)
and isentropic supersonic to supersonic (not choked) flow
solutions exist.
2. For choked condition, the flow solutions are each unique.

pext p or pext p ,isentropic flow is possible


p pext p , isentropic flow is not possible

Non-isentropic choked flow

overexpanded
underexpanded
Oblique shock wave (3-D)

V11.5 Rocket engine start-up


V11.6 Supersonic nozzle flow

isentropic

isentropic

Entropy generation

From Zucrow and Hoffman,


Gas Dynamics

11.4.3 Constant Area Duct Flow


z

For constant area isentropic duct flow, the flow velocity,


thus the fluid enthalpy and temperature are constant.

p0

pb

11.5 Nonisentropic flow of an ideal gas


z
z

Fanno flow adiabatic flow with friction.


Rayleigh flow constant area duct flow with heat
transfer but without friction

11.5.1 Adiabatic constant area duct flow


with friction (Fanno flow)
z

Consider steady 1-D ideal gas constant area duct flow


with friction

Continuity: m& = VA = const V = const


Energy equation:
V22 V12
m& [ h2 h1 +
+ g ( z 2 z1 )] = Q& + W&
2
V2
h+
= h0 , h h0 = c p (T T0 )
2
V2
T+
= T0 = const
(stagnation
2c p

temp.= const)

( V ) 2
( V ) 2 T 2
T+
= T0 T +
= T0 , where V = const
2
2
2
2c p
2c p p / R

(11.75)

Eq. 11.75 allows us to calculate T for p in the Fanno flow.

Tds equation (2nd law):


1
Tds = dh ( )dp

dT
dp
R
T
p
T
p
s2 s1 = c p ln 2 R ln 2
T1
p1

ds = c p

s s1 = c p ln

T
p
(11.76)
R ln
T1
p1

p1 , T1 , s1 are considered reference values from the entrance.


From (11.75) and (11.76), the Fanno line for variation of p-T-s
can be obtained.

Ex 11.11

Example 11.11 Compressible Flow with


Friction (Fanno Flow)
z

Air (k=1.4) enters [section (1)] an insulated, constant cross-sectional


area duct with the following properties:
T0=284K
T1=286K
p1=99kPa(abs)
For Fanno flow, determine corresponding value of fluid temperature and
entropy change for various values of downstream pressures and plot the
related Fanno line.

Example 11.11
To plot the Fanno line we use Eq. (75) and (76)
T+

(V) 2 T 2
2

2c P ( p / R )

= T0 = constan t

T
p
s s1 = cP ln R ln
T1
p1
k = 1.4

(11.11.2)

R = 286.9J / kg K

From Eq. (14)

(1)+(69)

(11.11.1)

Rk
(11.11.3)
cp =
= ... = 1004J / kg K
k 1
p
p
V =
Ma RTk = 1V1 = 1 Ma1 RT1k (11.11.4)
RT
RT1

Example 11.11
T1 286K
=
= 0.993
To 288K

From Eq. (56) Ma1 = 1 1 / 0.02 = 0.2


0.993

T
1
=
To 1 + k1 Ma2
2

RT1k = ... = 339m / s


(11.11.4)

99103 Pa0.2(339m / s)
V =
= 81.8kg /(m2 s)
(289.6J / kg K )(286K )

For p= 48 kPa
(11.11.1)
(11.11.2)

( V )2 T 2
T+
= ... = 288K T = 278.7 K
2
2
2cP ( p / R )

T
p
s s1 = cP ln R ln = ... = 181.7 J /(kg K )
T1
p1

(56)

Example 11.11
For p=48kPa T=278.7K s-s1=181.7J/(kgK)
For p=41kPa T=275.6K s-s1=215.7J/(kgK)
For p=34kPa T=270.6K s-s1=251.0J/(kgK)

Analysis of Fanno line


Tds equation

dp

dp
Tds = dh
= dh RT

p
For an ideal gas,

dp
p
d dT
)
= cP dT RT (
+

T
V = const, or
Continuity:
Tds = cP dT RT

Q dh = cP dT ;

dp d dT
=
+
p = RT or
p
T

dV
V

dT
dV dT
) = cP dT RT (
+ )
T
V
T

ds cP
1 dV 1

= - R(+ )
dT T
V dT T
Tds = cP dT RT (

Energy eq.:

cP
V2
VdV
dV
= T0 = const dT =

=
T+
2c p
cp
dT
V

cp 1
ds c p
- R( 2 + )

=
dT T
T
V

(11.82)

cp
cp 1
cp
ds
For
=0
= R ( 2 + ) c p R = c = RT 2
dT
T
T
V
V

V = (c p / c ) RTa = kRTa
So, the Mach number at state a is 1.

ds
< 0, V < kRTa subsonic
dT
ds
> 0, V > kRTa supersonic
dT

subsonic

supersonic
Ma=1
supersonic

Since T0 is constant on the Fanno line, the temperature at


point a is the critical temperature T*.

-2nd law ds>0

subsonic flow
(acceleration)

supersonic flow
(deceleration)

Normal shock

Summary of Fanno flow


behavior

Friction drags acceleration by pressure drop

Friction helps deceleration by pressure rise

-To quantify the Fanno flow behavior, we need to combine


relationship that represents the linear momentum law with the
set of equations already derived.

p1 A1 p2 A2 Rx = m(V2 V1 )
Rx
p1 p2
= V (V2 V1 ), ( Q A1 = A2 = A and m& = AV = C )
A
-Therefore, for the semi-infinitesimal control volumes

dp

w Ddx
A

= VdV

8 w
D2
with f =
, A=
2
V
4

V 2 dx
dp f
= VdV
2D
dp f V 2 dx d (V 2 )
+
+
=0
or
p p 2 D p 2

dp f V 2 dx d (V 2 )
+
+
=0
p p 2 D p 2

(11.88)

After some derivation, (11.88) becomes


2
2
1
d
(
V
)
d
(Ma
) fk
2 dx
(1 + kMa 2 )
Ma

+
=0
2
2
2
2
D
V
Ma

or

(1 Ma 2 )d (Ma 2 )

1 + [(k 1) 2]Ma 2 kMa 4

= f

dx
D

(11.96)

-Integrate to the critical state

Ma* =1

Ma

(1 Ma 2 )d (Ma 2 )
=
2
4
1 + [(k 1) 2]Ma kMa

l*

dx
f
D

2
*
2
1

Ma
f
l
l
1
k + 1 [(k + 1) 2]Ma
+
ln
=
2
2
k Ma
2k
D
1 + [(k 1) 2]Ma

(11.98)

-Note that the critical state does not have to exist,


since for any two section in the Fanno flow,

f l* l
D

) f (l * l ) =
1

f
(l 1 l
D

).

-Other fluid properties in the Fanno flow can also be derived,


as summarized below.

Summary of Property Relations for Fanno Flow


Note: These equations correlate ratio of properties associated
with different positions (a certain position and the choke position),
between which friction loss exists.
2
*
2
l
1

Ma
l
f

1
[(k + 1) 2]Ma
k +1
ln
+
=
(11.98)
2
2
2k
1
+
[(

1)
2]Ma
k Ma
k
D

(k + 1) 2
T
=
,
2
*
1 + [(k 1) 2]Ma
T

(11.101)
1

[(k + 1) 2]Ma
V
=
,
2
V * 1 + [(k 1) 2]Ma
2

V
=
,
*
*

Note: For p/p0, p*/p0*, T/T0


at the same position,
isentropic relations in terms
of Ma or Fig. D1 can be used.

(11.103)
(11.105)

2
T
1
(k + 1) 2
p
=
=
,

2
*
*
*
Ma 1 + [(k 1) 2]Ma
T
p
same position

(11.107)

different positions

p0
p0 p p *
1 2
2
=
=

1 + [(k 1) 2]Ma

p0 * p p * p0 * Ma k + 1

k +1
2( k 1)

(11.109)

Note: These curves


correlate ratio of
properties associated with
different positions (a
certain position and the
choke position), between
which friction loss exists.

Figure D2 (p. 719)


Fanno flow of an ideal gas with
k = 1.4.

p0
p0 *

p
p*

Example 11.12 Choked Fanno flow


Given p0 = 101 kPa, T0 = 288 K

What is the maximum flow rate at the duct?

-For maximum flow rate, the flow must be choked at the exit.

f l* l 1
f (l 2 l 1 ) 0.02 2
=
=
= 0.4
D
D
0.1
p0,1
T1
V1
p1
Fig. D2 Ma1 = 0.63 * = 1.1, * = 0.66,
= 1.7,
= 1.16
*
*
T
V
p
p0
T
p1
1
Fig. D1 with Ma1 = 0.63 1 = 0.93,
= 0.76,
= 0.83
T0
p0,1
0,1
Since T0 =C =288 K
Or, perhaps more logical,
T*
2
T1 = 0.93 T0 = 0.93 288
=
= 0.8333 T * = T0 (0.8333) = 240 T2
T0 k + 1
= 268 K
V * = RT *k = 310m/s ( V2 ) V1 = 0.66 V * = 205m/s V1 = Ma1 kRT1

0,1 = p0,1 / RT0,1 = 1.23 1 = 0.830,1 = 1.02 kg/m3

= 0.63 1.4 386.9 268


= 207m/s

m& = 1V1 A = 1.65 kg/s


*

with friction, use Fig. D2

p p1
1
p2 =
p0,1 =
0.76 101 = 45kPa
p1 p0,1
1.7

p0* = p0,2 = p0,1

1
= 87kPa
1.16

Example 11.13 Effect of duct length on choked Fanno flow

Pd = 45kPa
1m

If the flow is choked.

f l* l 1
D

) = 0.02 1 = 0.2,
0.1

Fig. D1 with Ma1 = 0.70

Fig. D2 Ma1 = 0.7,

p1
V1
=
1.5,
= 0.73,
*
*
p
V

p1
= 0.72, 1 = 0.79
0,1
p0

p * p1
1
p2 = p =
p01 =
0.72 101 = 48.5kPa ( > pd = 45kPa)
p1 p01
1.5
*

1 = 0.79 0,1 , V1 = 0.73V *


m& = 1 AV
1 1 = 1.73 kg/s
-For the same upstream stagnation state and downstream pressure,

l m& or when l is fixed: f m&


Example 11.14 Unchoked Fanno flow

11.5.2 Frictionless constant-area duct flow with heat


transfer (Rayleigh flow)

p1 A1 + mV1 = p2 A2 + mV2 + Rx

Momentum:
or
p0 = p +

V )
(
p+

V 2
2

= const

V )
(
p+

RT

0 (frictionless flow)

= const

const, although there is no friction, why?

Continuity:

V = C

Tds eq.: s s1 = c p ln

T
p
Rl n
(11.76)
T1
p1

Eqs. (11.111) and (11.76) can be used to


construct a Rayleigh line with reference
conditions.

(11.111)

Example 11.15 Construction of a Rayleigh line for various


downstream p2 (or T2), with entrance conditions T0, T1, p1

Given T0 , T1 , p1 , 1 , V1 and V = C can be calculated.


-Assume p2 (or T2), then from (11.111) T2 (or p2) can be obtained.
-From (11.76), s2 can be obtained.

( V )
p+

RT

= const

T
p
s s1 = C p ln Rln
T1
p1

(11.111)
(11.76)

Discussion on the Rayleigh line


-At point a on the Rayleigh line,
-After some derivation, we have

1
ds c p V
= +
dT T T [ (T / V ) (V / R) ]
ds
= 0,
dT
ds c p V
1
= +
dT T T [ (T / V ) (V / R )]

For

kR T V
+V = 0
k 1 V R

2
kRT kV 2 ( k 1)V

+
=0
k 1 k 1
k 1
V 2 = kRT

Va = kRTa Ma a = 1

ds
=0
dT
(11.115)

Derivation of ds/dT

dp = VdV
dP

= VdV

Since Tds equation

Tds = dh

dp

= c p dT + VdV
or

ds c p V dV
= +
dT T T dT
cp V
1
= +
T T (T / V V / R )

V = C
d
dV

=
V

p = RT
dp d .dT
=
+

p
T
VdV
dV dT

=
+
RT
V
T
V
1 1 dT

= +
RT
V T dV
T V dT
=
V R dV

At point b, dT / ds = 0
ds c p V
1
= +
dT T T [ (T / V ) (V / R ) ]
dT
1
1
=
=
=0
ds c p V
1
ds
+ [ (T / V ) (V / R) ]
dT
T T
cp V
1
+

T T [ (T / V ) (V / R )]
T V
= V 2 = RT V = RT
V R
V
RT
1
=
Ma b = b =
c
k
kRT

- At point b, the flow is subsonic since k1.

- Now, consider the energy equation,

V 2 2 V1 2
m& h 2 h1 +
+ g ( z 2 z 1 ) = Q n e t + W sh e ftn e t
2

dh + VdV = q,

dh = c p dT =

kR
dT
k 1

c p dT + VdV = q
dT VdV q
+
=
T
c pT c pT
dV
V

V dT
q
V2
+

=
T dV kRT / ( k 1) c pT
1

2
dV q V dT ( k 1) V
q

=
+
=

V
c pT T dV
kRT
c pT

V
T

T V
2
k
1
Ma

V
R

V2
q
2
2
2 1

=
+ ( k 1) Ma =
1 kMa + ( k 1) Ma
1

c pT RT
c pT

1
c pT 1 Ma 2

(11.121)

Derivation for Property Relations for Rayleigh Flow


-linear momentum

p + V 2 = pa + aVa 2

where a is the reference state.

p V 2
or
+
= 1 + a Va 2
pa
pa
pa

2
kV
Va 2 =
Va 2 = a = k , since Va = kRTa
pa
a RTa
kRTa

a 2
p V 2
+
= 1 + Va = 1 + k
pa
pa
pa
p
pa

V 2
1 +
= 1+ k,
p

p
1+ k
=
pa 1 + kMa 2

where V = Ma kRT

(11.123)

a V
T
T
p
T
=
= Ma
=
Ma
Ta
Ta pa
Ta
Va
(1 + k ) Ma
T
p a
T p
=
= Ma =
2
Ta pa
Ta pa
k
1
Ma
+

(1 + k ) Ma
a V
T
=
= Ma
= Ma
2
Ta
k
1
Ma
+
Va

p
1+ k
=
Q
2
pa 1 + kMa

(11.129)

- Due to heat transfer, T0 varies

T0 T0 T Ta
=
T0,a T Ta T0,a
(1 + k ) Ma

1
= 1 + [(k 1) / 2]Ma

2
+
+

k
k
1
Ma
1
(
1)
/
2

2 ( k + 1) Ma 2 1 + [(k 1) / 2]Ma 2
1 + kMa 2

(11.131)

Summary of Property Relations for Rayleigh Flow


p
1+ k
=
pa 1 + kMa 2

(11.123)

T (1 + k ) Ma
=
2
Ta 1 + kMa
(1 + k ) Ma
a V
=
= Ma
2
1
k
Ma
Va
+

(11.128)
(11.129)

1 + k ) 2
(
p0
p0 p pa
2
k
1
[(
1)
/
2]Ma
=
=
+

p0,a
p pa p0,a
1 + kMa 2 k + 1

T0 T0 T Ta
=
=
T0,a T Ta T0,a

2 ( k + 1) Ma 2 1 + [(k 1) / 2]Ma 2
1 + kMa
2

k 1

(11.133)

)
(11.131)

a V
,
Va

Figure D3 (p. 720)


Rayleigh flow of an idea gas with
k = 1.4. (Graph provided by Dr.
Bruce A. Reichert.)

Effect of Ma and Heating/Cooling for Rayleigh Flow


(Example 11.16)

Note: p0 is not constant, although frictionless

11.5.3 Normal Shock Waves


-Normal shock waves involves:
deceleration from supersonic to subsonic
a pressure rise
an increase of entropy

(from Cengel and Cimbala,


Fluid Mechanics, 2006)

FIGURE 1230

V11.6 Supersonic nozzle


flow
V11.7 Blast waves

Schlieren image of a normal shock


in a Laval nozzle. The Mach number
in the nozzle just upstream (to the
left) ofthe shock wave is about 1.3.
Boundary layers distort the shape of
the normal shock near the walls and
lead to flow separation beneath the
shock.

Derivation for Normal Shock Waves


z infinitesimal thin control volume surrounding the
shock wave: friction and heat transfer negligible, A=C
V = const
continuity:
linear momentumfriction negligible:same as Rayleigh line

p + V = const or
2

energy:

V )
(
p+

RT

= const

with z = 0, q = 0

V2
h+
= h0 = const
2
For an ideal gas,

( V ) T 2
2

T+

= T0 = const same as Fanno line


2c p p / R
T
p
Tds relationship: s s1 = c p ln Rln
T1
p1
Q: The irreversibility in shock wave is not from friction or heat transfer.
What is it from?

Since shock wave flows have the same energy eq. for Fanno flows
and same momentum eq. for Rayleigh flows, thus for a given V,
gas (R, k), and conditions at the inlet of the normal shock (Tx, px, sx),
the conditions downstream of the shock (state y) will be on both a
Fanno line and a Rayleigh line that pass through the inlet state
Mom., mass,
(state x).
Total energy,
- For Rayleigh line,
mass, Tds eqs.
p y p y pa
=
px pa px
py
px
1+ k
1+ k
,
Q
=
=
pa 1 + kMa y 2 pa 1 + kMa x 2

1 + kMa 2x

=
p x 1 + kMa 2y
py

Tds eqs.

(11.140)

- For Fanno line


1
*
py py p
2
1
(k + 1) / 2
p
= *
and * =
Ma 1 + [(k 1) / 2]Ma 2
px p px
p
1/ 2

p y Ma x 1 + [(k 1) / 2]Ma 2x

px Ma y 1 + [(k 1) / 2]Ma 2y

(11.148)

Total energy, mom.,


mass, Tds eqs.

Combining (11.140) and (11.148)


1/ 2

p y 1 + [(k 1) / 2]Ma 2x

=
2
px 1 + [(k 1) / 2]Ma y

Ma x 1 + kMa 2x
=
Ma y 1 + kMa 2y

2
Ma
x + [2 /( k 1)]
Ma 2y =
[2k /(k 1)]Ma 2x 1

1 + kMa 2x
2k
k 1
2
=
=

(11.149) into (11.140)


Ma
x
px 1 + kMa 2y k + 1
k +1

(11.149)

py

(11.150)

- For Fanno line, Eq. 11.101

Tx
(k + 1) / 2
(k + 1) / 2
=
=
and
T * 1 + [(k 1) / 2]Ma 2y
T * 1 + [(k 1) / 2]Ma 2x
Ty

Ty T * 1 + [(k 1) / 2]Ma 2x
= *
=
Tx T Tx 1 + [(k 1) / 2]Ma 2y
Ty

(11.144)

[1 + [(k 1) / 2]Ma 2x ]{[2k /(k 1)]Ma 2x 1}


(11.149) into (11.144)
=
(11.151)
{(k + 1) 2 /[2(k 1)]}Ma 2x
Tx
Ty

Summary of Property Relations across Shock Wave


If Max is known, property ratios across the shock can be known:
2
Ma
x + [2 /( k 1)]
Ma 2y =
[2k /(k 1)]Ma 2x 1

(11.149)

2k
k 1
2
Ma x
=
px k + 1
k +1

(11.150)

[1 + [(k 1) / 2]Ma 2x ]{[2k /(k 1)]Ma 2x 1}


=
{(k + 1) 2 /[2(k 1)]}Ma 2x
Tx

(11.151)

py

Ty

y Vx
(k + 1)Ma 2x
=
=
x Vy (k 1)Ma 2x + 2

k +1
k 1
2 k 1
[
Ma x ] [1 +
Ma 2x ] k 1
2
= 2
1
2k
k

1
[
Ma 2x
] k 1
k +1
k +1
k

p0, y
p0, x

(11.154)
k

(11.156)

Figure D4 (p. 721)


Normal shock flow of an idea
gas with k = 1.4. (Graph
provided by Dr. Bruce A.
Reichert.)

Example 11.17 Stagnation pressure drop across a normal shock


For Ma x
p0, y

py
px

and

p0, y

p0, x

py

px

p0, x
0.5

10

0.06
1

29

2.5

Ma x

1
1

Ma y

Across a normal shock, adverse pressure gradient occurs which


can cause flow separation. Therefore, shock-boundary layer
interactions are of great concern to designers of high speed flow
device.

Example 11.18 Supersonic flow pitot tube


Given: p0, y = 4114 kPa, T0 = 555K, px = 82 kPa
k +1
2 k 1
[
Ma
x]
p0, y p0, y p0, x
2
=
=
Rayleigh Pitot tube formula
1
px
p0, x px
k 1 k 1
2k
[
Ma 2x
]
k +1
k +1
414kPa
=
=5
82kPa
k

Fig. D1 Ma x = 1.9
Vx = Ma x cx = Ma x kRTx
T0, x = T0, y ,

Tx
= 0.59 Tx = 327K
T0, x

Vx = 678 m/s
Note: Incompressible calculation for pitot tube would give the wrong result.

Example 11.19 Normal shock in a converging-diverging duct


pIII
p
(a)
=? for shock at x = 0.3 m
= ? for normal shock at exit, (b)
p0
p0

x=-0.5

x=0.5

0.98
2.8

0.04

-shock at x=0.5m
From Ex 11.8: Ma x = 2.8 and
Fig. D4 Ma x = 2.8(at exit):

py
p0, x
p0, y
p0, x

px
= 0.04 at x = 0.5m
p0, x
py
px

= 9,

p0, y
p0, x

= 0.38

p y px
p
=
= 9 0.04 = 0.36 = III
px p0, x
p0, x
= 0.38 considerable energy loss

-shock at x=0.3m
px
From Ex 11.8: Ma x = 2.14,
= 0.1
p0, x
Fig. D4 with Ma x = 2.14 across the shock
py
px

= 5.2, Ma y = 0.56,

p0, y
p0, x

= 0.66

(Conditions downstream the shock at x =0.3m)

Note: for isentropic A*=0.1 (a minimum)

Consider the isentropic decelerating flow downstream the shock


Ay
= 1.24 (Here, A * is used as dummy)
Also, Fig. D4
A*
A2 0.1 + (0.5) 2
=
= 1.842
2
Ay 0.1 + (0.3)
Ay A2
A2
=
= 1.24 1.842 = 2.28
A * A * Ay
A2
A* =
= 0.15
2.28
A2
= 2.28, Fig. D1 (isentropic) Ma 2 = 0.26,
A*

p2
= 0.95
p0, y

p2
p2 p0, y

=
= 0.95 0.66 = 0.63
p0, x p0, y p0, x
Note:

p0, y
p0, x

= 0.66, shock at x = 0.3m >

p0, y
p0, x

= 0.36, shock at exit x = 0.5m

Figure D1 (p. 718)


Isentropic flow of an ideal gas with
k = 1.4. (Graph provided by Dr.
Bruce A. Reichert.)

11.7 Two-Dimensional Compressible Flow


Supersonic
- Flow acceleration across a Mach wave

Vt1 = Vt2
Vn2 > Vn1
V2 > V1

- Supersonic flows decelerate across compression Mach wave

from M. Van Dyke,


An Album of Fluid
Motion

Small wedge angle-attached shock

Large wedge angle-detached shock

from M. Van Dyke, An Album of Fluid Motion

(from Cengel and Cimbala, Fluid Mechanics, 2006)

Figure 11.4 (p. 591)


The schlieren visualization of flow (supersonic to subsonic) through a row of
compressor airfoils. (Photograph provided by Dr. Hans Starken, Germany.)

3-D shock wave around a model space shuttle

from Gas Dynamics Lab, The Penn. State University, 2004