Calvin Lui Department of Mechanical Engineering Stanford University Stanford, CA 94305 March 1998
Acknowledgments
This work is specially dedicated to the graduating class of 1998 and my dear oce-mate, Zhongmin Xiong, here at Stanford University. They encouraged me to compile all these tutorial materials together into one single volume which serves as future references for the ME 131B class. I would like to express my gratitude to them for all their encouragement.
Stanford, California March, 1998
Calvin Lui
ii
Table of Contents
Tutorial One: Thermodynamics Review Solutions to Thermodynamics Review Tutorial Two: Isentropic Flow I Solutions to Isentropic Flow I Tutorial Three: Isentropic Flow II Solutions to Isentropic Flow II Tutorial Four: Normal Shock Solutions to Normal Shock Tutorial Five: Fanno Flow Solutions to Fanno Flow Tutorial Six: Rayleigh Flow Solutions to Rayleigh Flow Tutorial Seven: Angular Momentum Principle Solutions to Angular Momentum Principle Tutorial Eight: Turbomachinery Solutions to Turbomachinery References 1 4 10 14 24 28 40 44 60 66 86 90 105 109 123 127 135
iii
2. What is a simple compressible substance? What does the state principle for a simple compressible substance tell us?
3. What do the following laws of thermodynamics mean to you? Describe them in your own words. (a) Zeroth Law
4. Apart from the above laws of thermodynamics, what other basic principle(s) do we usually apply in analyzing thermodynamic systems?
5. What are the dierent transfer modes for (a) Energy, (b) Entropy?
6. Write down the mathematical form of the First and Second Law for (a) a close system, (b) an open system.
Describe the meaning of each term in the equation. 7. A patent application describes a closed system which at steady-state conditions receives a heat transfer of 500 W at a temperature of 400 K and develops a combined electrical and mechanical power output of 500 W. There are no other energy transfers. Is this claim thermodynamically feasible?
8. What is the ideal gas model? How about perfect gas model? Under what conditions will these models be appropriate in describing real-life phenomena?
9. What is the relation between the specic heats (Cp; Cv ) for an ideal gas?
11. What is an isentropic process? What is its signicance in thermodynamic analysis?
12. What are some common causes for irreversibility in thermodynamic systems?
14. Derive the P T relationships for a perfect gas undergoing an isentropic process.
15. Sketch the following curves on a T s diagram (a) constant pressure, (b) constant density.
Based on the Gibbs equation, explain the dierence in the slope of the above two curves.
state of a system. Extensive property depends on the size or extent of a system. Its value of an overall system is the sum of its individual parts, like entropy and internal energy. Intensive property is independent of the size of a system and is not additive, like temperature and pressure.
2. What is a simple compressible substance? What does the state principle for a simple compressible substance tell us? system Compressible { work mode is associated with volume change R p dV The state principle states that two independent, intensive thermodynamic properties are sucient to fully determine the thermodynamic state of a simple compressible substance, like (T; v); (u; v). (Reminder: Pressure and temperature are not independent of each other in the two-phase region.) For a \non-simple" substance with n independent work modes, we need to know a total of n + 1 independent, intensive thermodynamic properties to completely specify its state. 3. What do the following laws of thermodynamics mean to you? Describe them in your own words. (a) Zeroth Law { Temperature Equality in temperature is a necessary and sucient condition for thermal equilibrium. (b) First Law { Energy Energy is conserved. (c) Second Law { Entropy 4
Simple { there is only one reversible work mode which can alter the energy of the
Entropy can only be produced but not destroyed. (Be careful that it does not
mean that entropy of a system can never decrease. If we have enough heat transfer out of a system, it is possible to have a decrease in the entropy of the system.) It is a powerful tool for us to determine the possible direction of a thermodynamic process.
4. Apart from the above laws of thermodynamics, what other basic principle(s) do we usually apply in analyzing thermodynamic systems?
dmcv = dt
m _ in
m _ out
X
Z @Z V ~ ( dV ) + ~ V ~ dA ~ = V @t CV CS
~surface + X F ~body F
5. What are the dierent transfer modes for (a) Energy, Mass transfer Heat transfer Work transfer (b) Entropy? Mass transfer Heat transfer There is no entropy transfer associated with work. This is a major dierence between the two energy transfer modes, work and heat. 6. Write down the mathematical form of the First and Second Law for (a) a close system, X X U = Qin Wout X Qin S = T + Ps
(b) an open system. kinetic z}|{ enthalpy potential 2 z}|{ X X V z}|{ dE = X Q _ _ in h + W m _ ( + gz ) out + dt | {z } | {z } {z 2 } | heat trans. non- ow work mass trans. _ in X dS = X Q _s P + ms _ + |{z} dt T {z } | | {z } heat trans. mass trans. production Remarks: Equation can be expressed in an overall or rate form. Examples of non- ow work: { shaftR work, { any PdV type of work such as compression and expansion Enthalpy consists of internal energy and ow work, h = u + P= (Hence, do not _ out again!) double count the ow work in W Internal energy is a measure of microscopic molecular activities while kinetic and potential energies are measures of bulk uid motion. 7. A patent application describes a closed system which at steady-state conditions receives a heat transfer of 500 W at a temperature of 400 K and develops a combined electrical and mechanical power output of 500 W. There are no other energy transfers. Is this claim thermodynamically feasible?
Since it is a closed system, there is no mass
ow. Conservation of mass is trivial. First law (conservation of energy) is satised. Entropy production rate goes negative, hence, second law is violated. As a conclusion, the claim is not thermodynamically feasible.
8. What is the ideal gas model? How about perfect gas model? Under what conditions will these models be appropriate in describing real-life phenomena?
Ideal gas model: { It satises the thermal equation of state: p = R T where R is the gas
constant (dierent for dierent gases.) { u = u(T ) 6
interaction between participating gas molecules.) Perfect gas model: { It is an ideal gas with constant specic heats Cp; Cv . { It is an appropriate model if the temperature variations between states are not too large (together with the conditions for ideal gas behavior.) 9. What is the relation between the specic heats (Cp; Cv ) for an ideal gas?
According to denition
@h Cp = @T p
and
@u Cv = @T v
Adiabatic process is a thermodynamic process in which no heat transfer occurs. It is a good model if { System has good insulation.
7
Isentropic process is reversible and adiabatic. It serves as a limit for real adiabatic process.
12. What are some common causes for irreversibility in thermodynamic systems?
Dissipation like friction, viscous eects Mixing Spontaneous chemical reaction Unrestrained expansion
13. Write down the Gibbs equation.
Tds = du
P 2 d
or
Tds = dh
dP
14. Derive the P T relationships for a perfect gas undergoing an isentropic process.
decreasing pressure s
decreasing density s
Based on the Gibbs equation, explain the dierence in the slope of the above two curves. The slope of any curve on the T s plane is characterized by the derivative dT=ds. From the Gibbs equation, Tds = dh dP Tds = Cp dT dP
Hence, the slope of a constant pressure curve on a T s diagram is equal to T=Cp. Similarly, we can obtain the slope of a constant density curve to be equal to T=Cv . Since Cp = Cv + R, the constant density curve has a steeper slope than the
constant pressure curve at the same temperature.
6. Given a thermodynamic state (T1 ; P1) and its speed in terms of Mach number, M1 . Can you locate its corresponding stagnation state on the T s diagram?
T P1
T1 s
10
Locate the static and stagnation states of both the upstream Station (1) and downstream Station (2) on the same T s diagram.
8. Recall that the incompressible Bernoulli's model gives us 1 V 2: P0 = P + 2 But we obtain, from a compressible analysis, the following result:
P0 = 1 + k 1 M 2 : P 2 These two equations give strikingly dierent expressions for the stagnation pressure. How do you reconcile the dierence between the two? What does this dierence depend on?
k k
11
9. In a wind tunnel experiment, 1.0 kg/sec of air is accelerated through an adiabatic nozzle from an upstream section (P1 = 2.0 bar, T1 = 900 K, A1 = 50 cm2) to a Mach 1.2
ow in the downstream section.
m = 1.0 kg/s
M2 = 1.2 P2 = ? T2 = ? A2 = ?
(b) If the ow is further treated as isentropic, i. What is the downstream cross-sectional area (A2 ), temperature (T2) and pressure (P2 )?
ii. Sketch the variation of pressure, temperature, velocity and Mach number from Station (1) to Station (2).
iii. Trace the process path from Station (1) to Station (2) on a T s diagram.
12
10. Air ows through a passage of unknown shape. The upstream state is characterized as follows:
A2 = 1.0 m2 M2 = ? P2 = ? T2 = ?
Assume that the ow is isentropic. (a) Determine the possible downstream State (2) where A2 = 1:0 m2 (i.e. Find M2 ; P2; T2 .)
(b) Sketch the shape of the associated ow passage found in Part (a).
(c) Sketch the corresponding variation of density, velocity and Mach number from Station (1) to Station (2).
13
3. Stagnation property can be dened for a non-isentropic ow. The stagnation process is a hypothetical process which is not necessarily found in the real physical ow. The stagnation state should be viewed as a reference thermodynamic state with which the ow is compared with. Based on this reason, the stagnation state or properties can be dened for any physical ow regardless it is isentropic or not. In the case of isentropic ow, the stagnation state is constant in the ow. Hence, it serves as a universal reference within the same ow. But the usefulness of the stagnation state as a reference will be highly degraded if the ow is non-isentropic. In this case, P0 changes in the ow. (Depend on the importance of heat transfer in the ow, T0 may also vary in the ow.) 14
4. Under what conditions can we say that the stagnation enthalpy remains constant in a
ow? How about stagnation pressure? Stagnation enthalpy remains constant in a
ow when { the
ow is adiabatic, { the
ow is not subject to any non-
ow work. This can be inferred from the rst law of thermodynamics (conservation of energy). Stagnation pressure remains constant in a
ow when { the
ow is adiabatic, { the
ow is not subject to any non-
ow work, { the
ow is reversible. This can be inferred from the rst and second law of thermodynamics. 5. Is there any limitation on applying the following equations in a
ow analysis? T0 = 1 + k 1 M 2 T 2 ! P0 = 1 + k 1 M 2 P 2 The only limitation is that the
uid should behave as a perfect gas.
k k
6. Given a thermodynamic state (T1 ; P1) and its speed in terms of Mach number, M1 . Can you locate its corresponding stagnation state on the T s diagram? By denition, the stagnation state should have the same specic entropy as the static state. Hence, they should both be on the same vertical line on the T s diagram. By denition, the stagnation temperature can never be lower than the static temperature. Hence, the stagnation state must be located somewhere above the static state. The dierence in temperature between the static and the stagnation states is a measure of the specic kinetic energy carried by the uid: 1V2 h0 = h + 2 1 V2 T0 = T + 2 C
T0
V2 T = 1 2 Cp
15
The higher the
uid velocity, the lower the static temperature compared with its stagnation temperature. This observation points to one important behavior of compressible
ow: interchange between thermal and kinetic energy in an adiabatic
ow. Based on the above conclusions, we can locate the stagnation state relative to its static state on the same T s diagram in the following gure:
T T0,1 T= T1 s V2 2 Cp P0,1 P1
s1 = s0,1
From the T s diagram, we can conrm the fact that the stagnation pressure is
always higher than the static pressure. 7. Consider a
ow through a valve as follows:
Locate the static and stagnation states of both the upstream Station (1) and downstream Station (2) on the same T s diagram. Based on the procedure outline in the previous problem, we can locate the static and stagnation state of upstream Station (1) as our reference. Across the valve, { The ow is adiabatic (no heat transfer). Stagnation temperature remains constant (T0;2 = T0;1). { There are losses due to friction at the valve. Entropy is produced (s2 > s1). Both static and stagnation pressure drop across the valve (P2 < P1; P0;2 < P0;1). 16
crease or decrease across the valve? To reason it out, we need to invoke both conservation of mass and energy. { Assume the same cross-sectional area in both pipes, conservation of mass (COM) gives 1 V1 = 2 V2 { Conservation of energy (COE) gives 2 2 1 = T + V2 T1 + 2VC 2 2 Cp p { Let us examine exhaustively two dierent alternatives: (a) 2 > 1 By COM, a density rise will result in a drop of
ow velocity (V2 < V1). By COE, a drop of
ow velocity will result in a rise in static temperature (T2 > T1 ). But a simultaneous rise in density and temperature cannot possibly produce a pressure drop across the valve. Hence, this case is not feasible. (b) 2 < 1 By COM, a density drop will result in a rise in
ow velocity (V2 > V1). By COE, a rise in
ow velocity will result in a drop of static temperature (T2 < T1 ). A simultaneous drop in density and temperature can produce a pressure drop across the valve. We can conclude that this is indeed the case. Based on the above conclusions, we can locate the static and stagnation state of downstream Station (2) with respect to those of upstream Station (1) on the same T s diagram as follows:
T T0,1 = T0,2 P0,1 P0,2 P1 P2
{ However, the non-trivial part is on the change in density. Does density in-
T1 T2 s1 s2 s
17
8. Recall that the incompressible Bernoulli's model gives us 2 P0 = P + 1 2 V : But we obtain, from a compressible analysis, the following result:
P0 = 1 + k 1 M 2 : P 2 These two equations give strikingly dierent expressions for the stagnation pressure. How do you reconcile the dierence between the two? What does this dierence depend on? Assume a perfect gas model, we can write the incompressible model as: P0 = 1 + 1 V 2 P 2 RT 2 1 = 1 + 2 k kV RT 1 = 1 + k M2 2 Results of the incompressible model are compared with those of the compressible model in the following gure (for the k = 1:4 case):
k k
2.0
Incompressible Compressible
1.8
1.6
P0 / P
1.4 1.2 1.0 0 0.2 0.4 0.6 0.8 1.0
number range (M < 0:3). Dierence between the two models becomes apparent for M > 0:4 (compressibility is no longer negligible). The value of P0=P predicted by the two dierent models and the percentage dierence between them are tabulated in the following table for further reference: 18
We can observe that the two models give nearly identical results in the low Mach
Incompressible Compressible % Dierence M1 = 0:10 1.0070 1.0070 0.0017283 M1 = 0:30 1.0630 1.0644 0.13438 M1 = 0:70 1.3430 1.3871 3.1794 M1 = 1:0 1.7000 1.8929 10.192 M1 = 2:0 3.8000 7.8244 51.434 M1 = 3:0 7.3000 36.733 80.127 To nd out the dierence between the two models, we can invoke the binomial expansion technique as follows:
P0 = 1 + k 1 M 2 P 2 k) M 6 + : : : 2 + k M 4 + k (2 = 1+k M 2 48 {z } | 8 correction terms The leading correction term of the compressible model to its incompressible counterpart scales with the fourth power of the
ow Mach number. This explains why the dierence between the incompressible and compressible models increases so dramatically with compressibility (Mach number).
k k
9. In a wind tunnel experiment, 1.0 kg/sec of air is accelerated through an adiabatic nozzle from an upstream section (P1 = 2.0 bar, T1 = 900 K, A1 = 50 cm2) to a Mach 1.2
ow in the downstream section.
m = 1.0 kg/s
M2 = 1.2 P2 = ? T2 = ? A2 = ?
(a) Sketch the general shape of the nozzle section. For air, R = 286:9 J / kg K. 19
We can use the thermal equation of state P = RT to compute the ow density, which gives 1 = 0:775 kg/m3 Assume one-dimensional ow, m _ = 1 A1 V1 We obtain V1 = 258 m/sec The speed of sound at Station (1) is given by q c1 = k R T1 = 601 m/sec This gives a Mach number of M1 = 0:429 at Station (1). Going from subsonic ow (M1 = 0:429) to supersonic ow (M2 = 1:2), we need to pass through a converging-diverging nozzle. (b) If the ow is further treated as isentropic, i. What is the downstream cross-sectional area (A2 ), temperature (T2) and pressure (P2 )? For an isentropic ow, the following quantities are constant in the entire ow: A; P0; T0 ; 0 They are invariant in the ow. Cross-sectional area (A2), static temperature (T2) and pressure (P2) can all be computed in a similar manner as follows: A2 = A2 =A 2 = 1:03044 = 0:6866 ) A = 34:3 cm2 2 A1 A1 =A 1:50072 1 T2 = T2=T0;2 = 0:77640 = 0:8051 ) T = 725 K 2 T1 T1=T0;1 0:96434 P2 = P2 =P0;2 = 0:41238 = 0:4683 ) P = 93:7 kPa 2 P1 P1 =P0;1 0:88065 ii. Sketch the variation of pressure, temperature, velocity and Mach number from Station (1) to Station (2). As the ow passes through the converging-diverging nozzle, velocity rises (nozzle is a ow accelerator), pressure drops (as a result of velocity rise, from momentum equation), temperature drops (as a result of velocity rise, from COE), Mach number rises (as a result of velocity rise and temperature drop). 20
Given P1 and T1 , the thermodynamic state at Station (1) is fully specied.
iii. Trace the process path from Station (1) to Station (2) on a T s diagram.
T T0,1 = T0,2 T1 P2 P0,1 = P0,2 P1
T2 s1 = s2 s
10. Air ows through a passage of unknown shape. The upstream state is characterized as follows:
A2 = 1.0 m2 M2 = ? P2 = ? T2 = ?
Assume that the ow is isentropic. (a) Determine the possible downstream State (2) where A2 = 1:0 m2 (i.e. Find M2 ; P2; T2 .) This problem is very similar to the last one except that the Mach number at the downstream Station (2) is unknown. The only information we have about Station (2) is its cross-sectional area. Before we tackle this problem, let us examine the variation of A=A with Mach number (refer to Section 13-3 in Fox & McDonald).
A = 1 A M
21
1 + k 2 1 M2 1 + k21
!2(
k+1 k 1
2.5
2.0
A / A*
1.5
1.0
From the above gure, we observe that there are two possible solutions to
this question, namely a subsonic solution and a supersonic solution. Both solutions have the same A=A ratio: A2 = A2 A1 A 1 = 1:0 (1:09437) (1) = 2:18874 A A1 A 0:50 2 1 A2 From the isentropic
ow table, { subsonic solution is M2 = 0:277, { supersonic solution is M2 = 2:298. To nd the pressure at Station (2), we can use the following procedure: P2 = P2 P0;2 P0;1 = P2 (1) 1 = 1:3871 P2 P1 P0;2 P0;1 P1 P0;2 0:72093 P0;2 Similarly, the temperature at Station (2) can be found in a similar manner: T2 = T2 T0;2 T0;1 = T2 (1) 1 = 1:0980 T2 T1 T0;2 T0;1 T1 T0;2 0:91075 T0;2 For the subsonic solution (M2 = 0:277), P2 = 0:9481 ) P = 6:58 bar 2 P0;2 T2 = 0:9849 ) T = 292 K 2 T
0;2
= 0:4864 ) T2 = 144 K 22
(b) Sketch the shape of the associated ow passage found in Part (a). For the subsonic solution, the ow passage will be a diverging one. It serves as a diuser. For the supersonic solution, the ow needs to go from subsonic to supersonic. Hence, a converging-diverging passage will be necessary. It serves as a nozzle. (c) Sketch the corresponding variation of density, velocity and Mach number from Station (1) to Station (2). For the subsonic solution, { velocity drops ( ow through a diuser), { pressure rises (as a result of velocity drop, from momentum equation), { temperature rises (as a result of velocity drop, from COE), { density rises (as a result of pressure rise and isentropic ow), { Mach number drops (as a result of velocity drop and temperature rise). For the supersonic solution, { velocity rises ( ow through a nozzle), { pressure drops (as a result of velocity rise, from momentum equation), { temperature drops (as a result of velocity rise, from COE), { density drops (as a result of pressure drop and isentropic ow), { Mach number rises (as a result of velocity rise and temperature drop).
23
2. For a steady, quasi-one-dimensional, adiabatic ow without wall friction, what do the following principles simplify to: (a) Conservation of Mass: (b) Momentum Equation: (c) Conservation of Energy: (d) Second Law of Thermodynamics:
3. We have discussed how to locate the stagnation state of a given
ow state (T1; P1; M1 ) last week. How about its sonic () state? Can you locate it on the T s diagram?
T P1
T1 s
24
4. Complete the following table with increases, decreases, remains constant for an isentropic ow:
P T V c M P0 0 T0 A P T P T V c M P0 0 T0 A P T
25
5. Choose the best answer in the following questions which concern the sonic state in an adiabatic, non-isentropic
ow: (a) T decreases/increases/remains constant in the
ow. (b) P decreases/increases/remains constant in the
ow. (c) A decreases/increases/remains constant in the
ow. 6. A large supply chamber containing air at 6.0 atm and 300 K is connected to a converging nozzle on the left side and a converging-diverging (C-D) nozzle on the right side. Both nozzles share the same minimum passage area of 100 cm2. The C-D nozzle has an exit-to-throat area ratio of 1.2.
P = 6.0 atm Pamb D T = 300 K B C A
Converging Nozzle
CD Nozzle
(a) Let us consider the converging nozzle on the left. i. Compare the pressure level at Point A, B, C and D. ii. If the ambient pressure is reduced to 5.0 atm, what is the mass ow rate in the nozzle? iii. How much do we need to lower the ambient pressure (relative to the chamber pressure) to reach the choking point of this converging nozzle? iv. What is the corresponding mass ow rate at the choking condition? (b) Let us consider the C-D nozzle on the right. i. If the ambient pressure is set at 5.0 atm, do you expect the mass ow rate in the C-D nozzle to be the same as that in the converging nozzle computed before? 26
ii. How much do we need to lower the ambient pressure for the nozzle to operate at its rst critical point? iii. What is the corresponding mass ow rate at the rst critical point? iv. At the design point (third critical), A. what is the ambient pressure? B. determine the density and velocity at the exit plane. (c) Look back to your calculations, i. How do you compare the ambient pressure which is required to choke the converging and C-D nozzle? Which one is higher? Can you explain it? ii. How do you compare the mass ow rate between the two nozzles: A. before choking? B. after choking? 7. You are asked to build a supersonic wind tunnel with operating Mach number of 2.0 in the test section. The plenum conditions are constantly kept at 300 K and 10.0 bars. Due to cost factor, air ow is delivered at a rate of 1 kg/sec. (a) If the ow is treated as isentropic, what is the downstream cross-sectional area? (b) If the entropy change between the plenum and the test section is 40 J/kg-K, what will be the cross-sectional area in the test section? Compare the result with Part (a) and label the two states on the same T s diagram.
27
In the absence of heat transfer and non-
ow work, the stagnation enthalpy of
the
ow is constant . 2. For a steady, quasi-one-dimensional, adiabatic
ow without wall friction, what do the following principles simplify to: (a) Conservation of Mass: V A = constant Remarks: The mass
ow rate (product of density, velocity and area) is required to be a constant to conserve mass. In subsonic
ow, the change in density is not so drastic. Velocity varies in a way which is consistent with our everyday experience. But the situation is so much dierent in the supersonic regime. Density change becomes very appreciable. Take the case of a supersonic
ow in a converging passage: the density increase outweighs the area decrease and forces velocity to go down in order to conserve mass
ow. Similar behavior is found in a supersonic
ow in a diverging passage but opposite eects are observed. In summary, the \strange" behavior of supersonic
ow is caused by the appreciable density change. It seems counter-intuitive because the world we encounter with on a daily basis operates mostly in the incompressible regime. (b) Momentum Equation:
dP = V dV
Remarks: Pressure and velocity change in opposite direction to each other in both subsonic and supersonic regimes. (c) Conservation of Energy:
2 h0 = h + V2 = constant 28
Remarks: When the ow speeds up, the uid cools down and vice versa. This interchange between static enthalpy and kinetic energy is fundamental in understanding an adiabatic ow. (d) Second Law of Thermodynamics:
s = constant
3. We have discussed how to locate the stagnation state of a given
ow state (T1; P1; M1 ) last week. How about its sonic () state? Can you locate it on the T s diagram? By denition, the sonic () state should have the same specic entropy as its static state. Hence, they should both be on the same vertical line on the T s diagram. By denition, the Mach number of the sonic state is unity. Hence, its location relative to that of its static state depends on the
ow Mach number. If the
ow is subsonic (M < 1), the sonic state will be below its static state. If the
ow is supersonic (M > 1), the sonic state will be above its static state. Based on the above conclusions, we can locate the sonic state relative to its static state on a T s diagram in the following gures: Subsonic Case
T T1
* T1
* T1
Supersonic Case
* P1
P1
P* 1 P1
T1
s1
s1
29
4. Complete the following table with increases, decreases, remains constant for an isentropic ow:
P T V c M P0 0 T0 A P T
Subsonic Flow: Converging Channel Diverging Channel decreases increases decreases increases decreases increases increases decreases decreases increases increases decreases remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant Supersonic Flow: Converging Channel Diverging Channel increases decreases increases decreases increases decreases decreases increases increases decreases decreases increases remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant remains constant
P T V c M P0 0 T0 A P T
Remarks: Both the stagnation state and the sonic state are constant in an isentropic ow. They serve as convenient reference states for the ow. 30
Trace each process path on a T s and a P diagram. Subsonic Case Supersonic Case
T Diverging T1 Converging
* T1
T
* T1
Converging T1 Diverging
s1
s1
P k = constant
P P* 1
P k = constant
P1 Diverging P* 1 Converging
Diverging P1 Converging
* 1
* 1
5. Choose the best answer in the following questions which concern the sonic state in an adiabatic, non-isentropic
ow: (a) T remains constant in the
ow. For any specic gas, the ratio of stagnation temperature to sonic temperature is a constant: T0 = 1 + k 1 T 2 Since the stagnation temperature of an adiabatic
ow is constant, so is the sonic temperature. (b) P decreases in the
ow. For any specic gas, the ratio of stagnation pressure to sonic pressure is a constant: ! P0 = 1 + k 1 P 2
k k
31
in the
ow direction, so does the sonic pressure. (c) A increases in the
ow. To see this point clearly, we can evaluate the mass
ow rate at the sonic point: where c = k R T : We know from previous results that T ; c remain constant but P decreases in the
ow. This leads us to conclude that decreases in the
ow also, from the ideal gas equation. Hence, A has to increase to conserve the same mass
ow rate. 6. A large supply chamber containing air at 6.0 atm and 300 K is connected to a converging nozzle on the left side and a converging-diverging (C-D) nozzle on the right side. Both nozzles share the same minimum passage area of 100 cm2. The C-D nozzle has an exit-to-throat area ratio of 1.2.
P = 6.0 atm Pamb D T = 300 K B C A
m _ = A c
Converging Nozzle
CD Nozzle
(a) Let us consider the converging nozzle on the left. i. Compare the pressure level at Point A, B, C and D. The main point of this part is to visualize the pressure variation and uid acceleration within the supply chamber. When the uid \senses" the pressure dierential between the inner chamber, P0, and the surrounding ambient, Pamb , it accelerates from negligible velocity at chamber pressure to some nite velocity closed to the nozzle inlet. Associate with this ow acceleration, there is a corresponding pressure drop. We can treat the ow going through an \imaginary" converging passage from the inner chamber to the nozzle inlet. 32
Between the inlet and exit, the
ow continues to accelerate and pressure
continues to drop. We can conclude that
PA > PC > PD The cause of pressure dierence between Point B and C is apparent after we draw the streamlines around the inlet of the converging nozzle. Since the streamlines curve around the corner, there is a positive pressure gradient developed in the normal (to the streamline) direction. Hence, PC > PB The comparison between pressure level at Point B and D depends on the exact nozzle geometry and requires further quantitative analysis. ii. If the ambient pressure is reduced to 5.0 atm, what is the mass
ow rate in the nozzle? In this type of problem, we always need to check if the converging nozzle is choked at Pamb = 5:0 atm. For a converging nozzle, we learned that the ambient pressure has to be lower than 52.8 % of the chamber pressure before choking occurs. In this case, Pamb = 5:0 = 0:833 > 0:528 P0 6:0 Hence, the nozzle is not choked. Furthermore, we can conclude that the pressure at the exit plane is the same as the ambient value. For the given pressure ratio Pexit = 0:833 P0 We can nd out from the isentropic
ow table that Mexit = 0:517 And the temperature ratio is Texit = 0:94924 T0 which gives an exit temperature of Texit = 284:8 K. Using the thermal equation of state for an ideal gas P = RT we obtain an exit density of exit = 6:201 kg / m3.
33
iii. How much do we need to lower the ambient pressure (relative to the chamber pressure) to reach the choking point of this converging nozzle? For a converging nozzle, the ambient pressure has to be lower than 52.8% of the chamber pressure to choke the converging nozzle. This corresponds to an ambient pressure of
If Pamb is lower than 3.17 atm, the exit plane pressure will not be the
exit = 4:478 kg/m3 m _ = 14:2 kg/m3 (b) Let us consider the C-D nozzle on the right. i. If the ambient pressure is set at 5.0 atm, do you expect the mass
ow rate in the C-D nozzle to be the same as that in the converging nozzle computed before?
34
For this C-D nozzle case, we also need to check if the nozzle is choked at
Pamb = 5:0 atm. The main dierence between the C-D nozzle and the converging nozzle is that the choking pressure ratio is dependent on the exit-to-throat area ratio (not a universal constant anymore). With an area ratio of 1.20, we nd from the isentropic ow table that the subsonic solution gives a pressure ratio P = 0:78997 < 5:0 P0 6:0 Hence, we conclude that { the ambient pressure is high enough that the ow is not choked { the ow remains subsonic within the C-D nozzle { exit ; Mexit; Texit are the same as those in the converging nozzle case Since the exit area is 1.20 times as large as that of the converging nozzle, we expect a 20 % increase in the mass ow rate. Hence, m _ = 13:0 kg/m3
ii. How much do we need to lower the ambient pressure for the nozzle to operate at its rst critical point? The rst critical point corresponds to an isentropic, subsonic solution with Mach 1.0 ow at the throat. We obtain from the isentropic ow table that Pamb = 0:78997 ) Pamb = 4:74 atm P0 iii. What is the corresponding mass ow rate at the rst critical point? Once this converging-diverging nozzle is choked at its rst critical point, we know that Mach 1.0 is achieved at its minimum ow area, i.e. at the throat. Furthermore, Pthroat ; Tthroat ; Mthroat are the same as those of the converging nozzle choked case. Hence, we expect the same mass ow rate as that of the converging nozzle choked case m _ = 14:2 kg/m3 iv. At the design point (third critical), A. what is the ambient pressure? The third critical point corresponds to an isentropic, supersonic solution in the C-D nozzle. 35
(c) Look back to your calculations, i. How do you compare the ambient pressure which is required to choke the converging and C-D nozzle? Which one is higher? Can you explain it? For the converging nozzle, Pchoke = 3:17 atm. For the C-D nozzle, Pchoke = 4:74 atm. We conclude that the C-D nozzle is choked at a higher ambient pressure than the converging nozzle. This conclusion can be explained by the following pressure plot for C-D nozzle operation:
Pamb / P0
(subsonic ow), the C-D nozzle is choked at a higher back-to-plenum pressure ratio. The exact value of the this pressure ratio depends only on the exit-to-throat area ratio. 36
A. Pamb =P0 > 0:790, both nozzles are not choked. B. 0:790 > Pamb =P0 > 0:528, only the C-D nozzle is choked. C. Pamb =P0 < 0:528, both nozzles are choked. ii. How do you compare the mass
ow rate between the two nozzles: A. before choking? Before any choking occurs, the C-D nozzle has a higher mass
ow rate (20 % higher) than the converging nozzle simply because the exit area of the C-D nozzle is 20 % larger than that of the converging nozzle. As the ambient pressure is reduced, the C-D nozzle gets choked rst. Once it is choked, its mass
ow rate is not aected by the ambient pressure anymore. Meanwhile the mass
ow rate of the converging nozzle keeps increasing as the ambient pressure is reduced. B. after choking? The mass
ow rate is the same in both nozzles after they are both choked. Graphically, the mass
ow rate of the two nozzles can be compared as follows:
m 14.2 kg/sec CD Nozzle Converging Nozzle
0.528
0.790
Pamb / P0
7. You are asked to build a supersonic wind tunnel with operating Mach number of 2.0 in the test section. The plenum conditions are constantly kept at 300 K and 10.0 bars. Due to cost factor, air ow is delivered at a rate of 1 kg/sec. (a) If the ow is treated as isentropic, what is the downstream cross-sectional area? To achieve a supersonic ow in the test section, we need a C-D nozzle connecting the plenum and the test section, with a Mach 1.0 ow (sonic state) right at the minimum throat area. 37
T = 0:83333 ) T = 250 K T0 The ideal gas equation further gives us the density at the throat: = 7:365 kg/m3 The size of the throat can be found from the mass
ow rate equation A m _ = V throat p = k R T Athroat ) Athroat = 4:285 cm2 To nd out the cross-sectional area at the test section, we need to relate the Mach 2.0
ow in the test section with the sonic state at the throat. From the isentropic
ow table, Msection = 2:0 gives Asection = 1:68750 ) Asection = 7:230 cm2 Athroat (b) If the entropy change between the plenum and the test section is 40 J/kg-K, what will be the cross-sectional area in the test section? Compare the result with Part (a) and label the two states on the same T s diagram. The
ow remains to be adiabatic. Hence, T0 ; T remains constant even in this non-isentropic
ow. However, the entropy increase in the nozzle causes a drop in the stagnation pressure. (Take State 1 to be the plenum state and State 2 to be the test section state in the following analysis.) ! ! T P 0;2 0;2 s2 s1 = Cp log T R log P 0;1 0;1 P s s 0;2 2 1 ) P0;1 = exp R = 0:870 ) P0;2 = 8:70 bar From the isentropic
ow table, M2 = 2:0 gives P2 = 0:12780 ) P2 = 1:11 bar P0;2 T2 = 0:55556 ) T2 = 167 K T
0;2
P = 5:283 bar
38
The ideal gas equation further gives us the density in the test section:
2 = 2:32 kg/m3 The cross-sectional area in the test section can be found from the mass
ow rate equation m _ = 2 V2 A 2 q = 2 M2 k R T2 A2 ) A2 = 8:31 cm2 Comparing this result with that of Part (a), we conclude that a larger test section area is necessary when irreversible eects are taken into account. It is also interesting to point out that the ratio between the area obtained in Part (a) and Part (b): Asection;a = 7:230 cm2 = 0:870 Asection;b 8:310 cm2 is the same ratio as the stagnation pressure loss. Hence, we conclude from this observation that P0 A = constant for an adiabatic
ow. When stagnation pressure drops, the sonic area increases. Both the isentropic and non-isentropic solutions are shown in the following T s diagram for reference:
T
* T2,a = T* 2,b
P2,a
P2,b
T2,a = T2,b
Isentropic Solution NonIsentropic Solution
s2,a
s2,b
Comments: The ridiculously small test section area is not reasonable for conducting wind tunnel experiments. We can increase its size by { reducing the plenum pressure { increasing the plenum temperature { paying a higher cost to allow a higher mass ow rate 39
2. In what frame of reference are the shock jump relationships derived in? 3. How is the strength of a shock measured by? What does it depend on?
4. Derive the relationship between the entropy production within a shock and the corresponding loss in stagnation pressure across it.
5. Choose the best answer in the following question and explain it: The Mach number upstream of a shock can/cannot be less than unity. 40
6. Complete the following table with increases, decreases, remains constant for a ow across a normal shock:
P T V c M P0 0 T0 A P T
Trace the process path across a shock on a T s diagram. (Indicate clearly its position relative to the stagnation state and the sonic state.)
7. Moving Shock Problem: A shock wave is propagating at Mach 2.0 into still air which is at atmospheric condition. What is the change of stagnation pressure observed by someone standing on the ground? How about the change in stagnation temperature?
P0 = ? T0 = ? M = 2.0 Still air at atm. condition
Moving shock
41
8. Shock Location Specication Problem: Air enters a C-D nozzle which has an exit-tothroat area ratio of 1.8 (A5=A2 = 1:8). A normal shock occurs at a location where the cross-sectional area is 1.2 times that of the throat (A3=A2 = 1:2). The schematic is shown below:
(3) (4)
(1)
(2) Shock
(5)
9. Back Pressure Specication Problem: Let us return to the C-D nozzle (with an exitto-throat area ratio of 1.2) we worked with last week. It connects again the same air supply chamber at 300 K and 6.0 bars to the ambient.
P = 6.0 atm T = 300 K Pamb
CD Nozzle
42
(a) Last week, we have found that the rst critical point operates at Pamb =P0 = 0:78997 and the third critical point operates at Pamb =P0 = 0:25922. What is the ambient-to-chamber pressure ratio at the second critical point? (b) We have learned last week how to compute the mass ow rate for both choked and unchoked cases. Let us focus on some o-design operations of this C-D nozzle this time. Describe qualitatively what happens inside the nozzle for the following ambient pressure values: i. Pamb = 5:4 bars ii. Pamb = 4:5 bars iii. Pamb = 4:2 bars iv. Pamb = 2:4 bars v. Pamb = 0:6 bars (c) In the cases where standing normal shock occurs in the nozzle, compute its location in terms of area ratio with respect to the throat.
(d) Look back to your calculations, how do you compare the shock location for different ambient pressure? Do you expect the shock to be closer to the throat or to the exit for a high ambient pressure? 43
V = constant
(b) Momentum Equation: Apply the results from mass conservation, we can directly integrate the momentum equation dP = V dV to obtain the following results:
P + V 2 = constant
(c) Conservation of Energy: The shock process is adiabatic.
2 h0 = h + V2 = constant (d) Second Law of Thermodynamics: The shock process is irreversible. Entropy is produced. s > 0
2. In what frame of reference are the shock jump relationships derived in? The shock jump relationships are derived in the shock frame of reference. Recall that the ow is treated as steady in the derivation, which is made possible by a Galilean transformation from a stationary observer to one traveled with the shock. We need to be careful when we apply these jump relationships to problems in which the shock is propagating. A change of reference frame to that of the shock will be necessary. 44
Associated with this change of reference frame, the stagnation properties will be
dierent but the static properties will remain the same.
3. How is the strength of a shock measured by? What does it depend on? The strength of a shock can be measured by the pressure jump across it: = P2 P P 1 1 where P1 and P2 are the static pressure upstream and downstream of the shock respectively. The larger the pressure jump, the stronger the shock is. The strength of a shock is dependent on the upstream Mach number: P2 = 2 k M12 ( k 1 ) P1 k+1 M12 1 ) ) = 2 k (k +1 in the following gure (for the k = 1:4 case):
10
The variation of shock strength with upstream Mach number is shown graphically
(P2 - P1) / P1
M1
As shown in the gure, the strength of a shock increases with the upstream Mach
number in a non-linear manner. Much higher pressure jump can be achieved at high Mach numbers. More quantitative results are tabulated in the following table for further reference: M1 1.0 0.0000 1.5 1.4583 2.0 3.5000 3.0 9.3333 5.0 28.000 7.0 56.000 45
Remarks: We can treat sound wave (M = 1) as a shock wave with zero strength. 4. Derive the relationship between the entropy production within a shock and the corresponding loss in stagnation pressure across it. We can apply the Gibbs equation to evaluate the pressure drop between the upstream and downstream stagnation states. 1 dP T0 ds0 = dh0 0 0
The shock process is an adiabatic process, hence, the stagnation enthalpy (h0) is
constant across the shock
d h0 = 0
1 dP 0 T0 0 R dP P0 0
!
Direct integration of the above equation between the upstream State (1) and the
s0;2 s0;1 = P0;2 R log P 0;1 s0;2 = s2 P0;2 R log P 0;1
!
Since a stagnation state has the same specic entropy as its static state, i.e.
s0;1 = s1; s2 s1 =
we can also write the previous results as follows:
The above equation relates the stagnation pressure drop to the entropy increase across a shock. 5. Choose the best answer in the following question and explain it: The Mach number upstream of a shock cannot be less than unity. 46
From Question 4, we obtain the stagnation pressure drop across a shock in terms
of the associated entropy change: P0;2 = exp s2 s1 P0;1 R
Recall that the stagnation pressure drop across a shock depends only on the
P0;2 = P0;1
"
upstream Mach number. It can be expressed by the following equation (Equation 13.37 in Fox & McDonald): ( k + 1 ) M12 2 + ( k 1 ) M12
8" < : 2 #
k k
"
k+1 2 k M12 k + 1
"
1
k
We can equate the above two equations and express the change of entropy across
the shock in terms of the upstream Mach number as
s2
s1 =
R log
( k + 1 ) M12 + ( k 1 ) M12
k k
k+1 2 k M12 k + 1
1
k
9 = ;
To unfold the physics contained in this complicated equation, we can plot the non-
dimensional entropy change ( ( s2 s1 ) = R ) against the upstream Mach number (M1 ). Results are shown in the following gure:
1.0
0.5
(s2 - s1) / R
-0.5
-1.0 0.5
1.0
1.5
2.0
2.5
M1
(solid line), { entropy decreases across the shock for a subsonic upstream Mach number (dotted line). Apparently, the latter case violates the Second Law of Thermodynamics. Hence, we conclude that the Mach number upstream of a shock cannot be less than unity. This conclusion also reveals the fact that a shock process is irreversible. 47
In the above gure, we observe two totally dierent behavior: { entropy increases across the shock for a supersonic upstream Mach number
6. Complete the following table with increases, decreases, remains constant for a ow across a normal shock:
P T V c M P0 0 T0 A P T
Across a Normal Shock increases increases increases decreases increases decreases decreases decreases remains constant increases decreases decreases remains constant
Trace the process path across a shock on a T s diagram. (Indicate clearly its position relative to the stagnation state and the sonic state.)
T T0,1 = T0,2 (2)
* * T1 = T2
P0,1
P0,2
(1) s
s1
s2
The upstream State (1) is supersonic, hence, it is located below the sonic state
on the T s diagram. The downstream State (2) is subsonic, hence, it is located above the sonic state on the T s diagram. The shock process is an irreversible, adiabatic process. Entropy is produced from State (1) to State (2). Hence, State (2) is located to the right of State (1) on the T s diagram. 48
7. Moving Shock Problem: A shock wave is propagating at Mach 2.0 into still air which is at atmospheric condition. What is the change of stagnation pressure observed by someone standing on the ground? How about the change in stagnation temperature?
P0 = ? T0 = ? M = 2.0 Still air at atm. condition
Moving shock
properties are independent of the reference frame. Having the above basic concepts in mind, we can properly dene the upstream state in the shock frame of reference as follows:
P0,2 = ? T0,2 = ? M1 = 2.0 P1 = 1.0 atm. T1 = 288 K
Recall from Question 2: { the shock jump relationships are derived in the shock frame of reference, { the stagnation properties are dependent on the reference frame while the static
Stationary shock
For an upstream Mach number of 2.0, we obtain from the normal shock table:
M2 = 0:57735 P2 = 4:50000 ) P = 4:50 atm. 2 P1 T2 = 1:68750 ) T = 486 K 2 T
1
We have just found out the static pressure and temperature downstream of the
shock in the shock frame of reference. Since the static properties are independent of reference frame, an observer standing on the ground will also measure the same static pressure and temperature as the above values. 49
V1 = M1 pk R T1 V2 = M2 k R T2
M2 = 0.57735 P2 = 4.5 atm. T2 = 486 K V2 = 255 m/sec
server standing on the ground, we need to know the downstream Mach number in the ground frame of reference. This requires a change of reference frame from that of the shock to that of the ground. In the ground frame of reference, an observer measures a downstream velocity of V2 = 425 m/sec (to the left) Or equivalently V2 = 425 m/sec (to the right) This corresponds to a downstream Mach number of M2 = 0:962 in the ground frame of reference.
M2 = 0.962 P2 = 4.5 atm. T2 = 486 K V2 = 425 m/sec M1 = 0 P1 = 1.0 atm. T1 = 288 K V1 = 0 m/sec
For a Mach number of 0.962, we obtain from the isentropic
ow table: P2 = 0:55192 ) P = 8:15 atm. 0;2 P
T2 = 0:84382 ) T = 576 K 0;2 T0;2 Hence, the change of stagnation pressure and temperature is: P0 = 8:15 1:0 atm. = 7:15 atm. T0 = 576 288 K = 288 K
50
0;2
8. Shock Location Specication Problem: Air enters a C-D nozzle which has an exit-tothroat area ratio of 1.8 (A5=A2 = 1:8). A normal shock occurs at a location where the cross-sectional area is 1.2 times that of the throat (A3=A2 = 1:2). The schematic is shown below:
(3) (4)
(1)
(2) Shock
(5)
(a) Sketch the pressure variation with streamwise location. Since there is a shock in the diverging section of the nozzle, we can conclude that we achieve Mach 1.0 at the throat. Then the
ow goes supersonic from the throat to Station (3). As the
ow accelerates from Station (1) to Station (3), pressure decreases continuously. Across the shock, the
ow is compressed. Hence, pressure increases from Station (3) to Station (4). The
ow then becomes subsonic after the shock and decelerates in the diverging section. This leads to a pressure rise from Station (4) to Station (5).
P / P0
0.528
(b) Locate the static and stagnation states of 1, 2, 3, 4 and 5 on a T s diagram. State (1) is subsonic, hence, its temperature is higher than the sonic temperature. State (2) is sonic. 51
State (3) is supersonic, hence, its temperature is lower than the sonic tem-
perature. State (4) is subsonic after the shock: { Its entropy is higher than that of State (3) because of irreversibility through the shock. { Its temperature is higher than that of State (3) because of shock compression. { Hence, it is located to the right of and above State (3). State (5) is subsonic. Its temperature is higher than that of State (4) because of
uid deceleration in the diverging section.
T T0
(5) (1) (4)
(2)
(3)
(c) What is the operating pressure ratio P5=P0;1? We can express the operating pressure ratio as a product of several pressure ratios: P5 = P5 P0;5 P0;4 P0;3 P0;1 P0;5 P0;4 P0;3 P0;1 Each of the pressure ratios on the right hand side of the above equation can be determined separately by knowing the local Mach number. Let us examine them individually: { P5=P0;5 depends on the local Mach number, M5 , only, which is an unknown up to this moment. { P0;5=P0;4 = 1 because the ow is isentropic from Station (4) to Station (5). { P0;4=P0;3 depends on the Mach number upstream of the shock, M3 . In this problem, we achieve the sonic state at Station (2) because the ow is supersonic between Station (2) and Station (3). A3 = A3 = 1:2 A2 A 3 52
which gives
M3 = 1:534
This gives a stagnation pressure loss of P0;4 = 0:91865 from normal shock table P0;3 (3). We simplify to the following equation: P5 = 0:91865 P5 P0;1 P0;5} | {z
f1 (M5 )
We need one more piece of information to complete this problem, namely the
local Mach number at Station (5). We can nd out the Mach number at Station 5 by relating the area ratio at dierent stations as follows: A5 = A5 A 5 A4 A4 A3 A2 A 5 A4 A4 A3 A2 Similarly, the area ratios in the above equation can be determined separately by knowing the local Mach number. Let us examine them individually: { A5=A2 = 1:8 as given. { A 5 =A4 = 1 because the ow is isentropic between Station (4) and Station (5). { A 4 =A4 depends on the local Mach number, M4 , which is governed by the Mach number upstream of the shock, M3 . For an upstream Mach number of 1.534, we obtain from the normal shock table,
M4 = 0:68939
which in turn gives an area ratio of A4 = 1:10230 from isentropic
ow table A 4 { A4=A3 = 1 because there is no area change across a shock. { A3=A2 = 1:2 as given. 53
which gives
CD Nozzle
(a) Last week, we have found that the rst critical point operates at Pamb =P0 = 0:78997 and the third critical point operates at Pamb =P0 = 0:25922. What is the ambient-to-chamber pressure ratio at the second critical point? The second critical point corresponds to an isentropic, supersonic ow all the way up to the C-D nozzle exit and then passes through a normal shock. For an exit-to-throat area ratio of 1.2, the condition just upstream of the shock, State (1), is given by the isentropic ow table: P1 = 0:25922 (third critical) M1 = 1:534 and P0;1 For an upstream Mach number of 1.534, the pressure jump across this shock can be obtained by the normal shock table as P2 = 2:57871 (shock jump) P1 54
0.790 0.668
0.259
Third Critical
dened by the nozzle geometry, namely the exit-to-throat area ratio. They take on dierent values for dierent nozzle geometry. We need to compare the following ambient pressure values with the rst, second and third critical pressure values before we conclude the ow behavior inside the nozzle. i. Pamb = 5:4 bars { Pamb =P0 = 0:9 in this case. It is higher than the rst critical point. { The nozzle is not choked. The ow remains subsonic inside the entire nozzle. 55
Please be reminded again that the rst, second and third critical points are
tion. ii. Pamb = 4:5 bars { Pamb =P0 = 0:75 in this case. It is between the rst and the second critical points. { The nozzle is choked. The ow reaches Mach 1.0 at the throat and goes supersonic for a portion in the diverging section. { A normal shock is expected to occur in the diverging section of the nozzle which terminates the supersonic ow. { The ow becomes subsonic after the shock. Pressure rises in the remaining portion of the diverging section. { Since the ow is subsonic, the ow comes out of the nozzle to match with the ambient pressure. iii. Pamb = 4:2 bars { Pamb =P0 = 0:70 in this case. It is between the rst and the second critical points again. { Similar ow behavior is expected within the nozzle. { But the shock location is dierent. iv. Pamb = 2:4 bars { Pamb =P0 = 0:40 in this case. It is between the second and the third critical points. { The ow remains isentropic and supersonic throughout the entire diverging section. { When the ow comes out of the nozzle exit, it is over-expanded (pressure is lower than the ambient value). { But a normal shock is too strong to match with the ambient pressure. Instead oblique shocks (weaker than normal shock) are expected to occur outside the nozzle to match with the ambient pressure. v. Pamb = 0:6 bars { Pamb =P0 = 0:10 in this case. It is lower than the third critical point. { The ow remains isentropic and supersonic throughout the entire diverging section. { When the ow comes out of the nozzle exit, it is under-expanded (pressure is higher than the ambient value). { Expansion waves (causes pressure drop) are expected to occur outside the nozzle to match with the ambient pressure. (c) In the cases where standing normal shock occurs in the nozzle, compute its location in terms of area ratio with respect to the throat.
{ Pressure drops in the converging section and rises in the diverging sec-
56
Case (ii) and (iii) in Part (b) will have a normal shock located in the diverging
section of the nozzle. The solution procedure is similar in both cases. They follow the same logic but the shock location will be dierent in the two cases. Let us rst label the states which will facilitate our calculations as follows:
P / P0 (1) 1.0 (5) (2) 0.528 (3) (4)
State (1) is the stagnation state in the supply chamber. State (2) is the sonic state, at the throat. State (3) is immediate upstream of the shock. State (4) is immediate downstream of the shock. State (5) is the exit state whose pressure has to match with that of the ambient (Subsonic ow has the obligation to match the imposed pressure.) When we examine the problem statement carefully, we are only given the operating pressure ratio and the exit-to-throat area ratio of the C-D nozzle. Hence, we need to work out the solution based on these two pieces of information. Let us express the pressure ratio in the following manner: P5 = P5 P0;5 P0;4 P0;3 P0;1 P0;5 P0;4 P0;3 P0;1
{ { { { {
{ P5=P0;1 is imposed by the problem statement (0.75 in Case (ii) and 0.70 { { { {
in Case (iii)). P5=P0;5 depends on the exit Mach number, M5 . P0;5=P0;4 = 1 because the ow is isentropic from Station (4) to Station (5). P0;4=P0;3 depends on the Mach number upstream of the shock, M3 . P0;3=P0;1 = 1 because the ow is isentropic from Station (1) to Station (3). 57
Hence, we simplify to the following relationship: P5 = P5 P0;4 (1) P P P 0 ; 1 0 ; 5 0 ; 3 | {z } | {z } | {z } f (M ) f (M ) Given Similarly, we can express the area ratio in the following manner: A5 = A5 A 5 A4 A4 A3 A3 A2 A 5 A4 A4 A3 A3 A2 { A5=A2 = 1:2 as specied in the problem statement. { A5=A 5 depends on the exit Mach number, M5 . { A5=A 4 = 1 because the
ow is isentropic from Station (4) to Station (5). { A4=A4 depends on the Mach number downstream of the shock, M4 . { A4=A3 = 1 because there is no area variation across a shock. { A3=A 3 depends on the Mach number upstream of the shock, M3 . { A3=A2 = 1 because the
ow is sonic at Station (2) before it reaches supersonic at Station (3) and it is isentropic from Station (2) to Station (3). Hence, we simplify to the following relationship: A 4 A3 1:2 = A5 (2) A A A 4 5 3 |{z} |{z} |{z}
1 5 2 3
i. Guess M3 (between unity and 1.534 (design exit condition)). { Look up the isentropic ow table to nd out A3 =A 3. { Look up the normal shock table to nd M4 { Look up the normal shock table to nd P0;4=P0;3 ii. With the value of M4 just obtained, look up A4 =A 4 in the isentropic ow table. iii. Substitute the above values into Equation (2) to nd out A5=A 5. iv. With the value of A5 =A5 just obtained, look up the corresponding M5 and P5 =P0;5 from the isentropic ow table. v. With P0;4 =P0;3 and P5=P0;5 gured out in the previous steps, we can substitute their values into Equation (1). vi. If the product between P0;4 =P0;3 and P5=P0;5 equals the value of P5=P0;1 as specied in the problem statement, the guess is correct in Step (i). Otherwise, keep guessing dierent values for M3 and repeat the above process until the pressure ratio matches with what is given in the problem. 58
(d) Look back to your calculations, how do you compare the shock location for different ambient pressure? Do you expect the shock to be closer to the throat or to the exit for a high ambient pressure? There are two mechanisms to obtain pressure rise in the diverging section of the nozzle: { compression due to normal shock, { subsonic pressure recovery after the shock. These two mechanisms work against each other. We need to evaluate the two alternatives separately: { If the shock is closer to the throat, the pressure jump due to the shock is less intense. But the subsonic ow (after the shock) has the entire diverging section to recover its pressure. It is operating close to the rst critical condition. { If the shock is closer to the exit, the pressure jump due to the shock is more intense. But the subsonic ow (after the shock) does not have much ow passage to recover its pressure. It is operating close to the second critical condition (lower than the rst critical). As a conclusion, the shock is expected to be closer to the throat for a higher ambient pressure.
59
3. Show that the Mach number corresponds to the maximum entropy point on a Fanno curve is unity.
4. Complete the following table with increases, decreases, remains constant for a ow through a constant-area, frictional duct:
P T V c M P0 0 T0 s A P T P + V2
M <1
M >1
60
5. A ow is supplied by a converging nozzle (unchoked). (a) Will the addition of a diverging section increase or decrease the mass ow rate? (b) What about adding a constant-area pipe? Will it increase or decrease the mass ow rate? (c) Sketch the variation of the mass ow rate, m _ , with back-to-stagnation pressure ratio, Pb=P0, for the above two cases on the same plot and highlight the dierences.
Converging Nozzle
Frictional Pipe
61
Consider the above setup. (a) Where can Mach 1.0 be realized? (b) Unchoked Case In this part, we operate the above setup at a pressure ratio of Pb=P0;1 = 0.60. What is the ow conditions at the exit (P; T; M )?
(c) Choked Case How much do we need to lower the back-to-stagnation pressure ratio to reach the choking condition? What does this ratio depend on? (Recall that it depends on the exit-to-throat area ratio for a C-D nozzle.)
(d) L > Lmax After we lower the pressure ratio to the value we computed in Part (c), the ow is choked. What do you expect to happen if an extra two meters of pipe section is added to the existing system? Do you expect the ow is still choked at the exit? Explain your answer by showing the corresponding states on a T s diagram.
62
Pb
(1)
(2)
(e)
(a) Slightly dierent from the last problem, there are two possible locations at which Mach 1.0 is attainable in the above setup. Where are they? (b) L = Lmax In the supersonic operation mode, determine the pipe length Lmax which gives a sonic ow right at the pipe exit.
(c) L < Lmax i. For L = 2:0 m, determine the range of the back-to-stagnation pressure ratio, Pb=P0;1, over which A. a normal shock appears in the diverging section of the C-D nozzle
63
ii. For each of the above cases, A. sketch the process path from the nozzle inlet to the pipe exit on a T s diagram.
B. sketch the pressure distribution along the streamwise location from the nozzle inlet to the pipe exit.
C. outline the solution procedure to locate the shock position in cases where normal shock appears.
64
(d) L > Lmax (Common case) i. For L = 5:0 m, qualitatively describe the ow in the system for the following pressure ratios: A. Pb=P0;1 = 0:50
B. Pb=P0;1 = 0:10
ii. For each of the above cases, A. sketch the process path from the nozzle inlet to the pipe exit on a T s diagram.
B. sketch the pressure distribution along the streamwise location from the nozzle inlet to the pipe exit.
65
The Fanno curve is derived from the conservation of mass and the conservation of
energy principles. All states on the same Fanno curve have the same { mass
ux, m=A _ { stagnation enthalpy, h0 Eects of changing the mass
ux and stagnation enthalpy on the Fanno curve are displayed in the following gures: Changing m=A _
increasing m A
T
T0
Changing h0
increasing h0
T*
2. Trace out a Fanno curve on a T s plane and locate the corresponding stagnation and sonic states on the same diagram. stagnation states lies on the horizontal line (T = T0). The ratio of sonic to stagnation temperature is a constant for any local static state T0 = 1 + k 1 T 2 Hence, the sonic temperature is also a constant. All the sonic states lies on the horizontal line (T = T ).
Since the stagnation temperature is constant for a Fanno ow (adiabatic), all the
66
T
T0
10 1 (static)
Locus of all stagnation states Subsonic regime Locus of all sonic states
T*
1*
Supersonic regime
3. Show that the Mach number corresponds to the maximum entropy point on a Fanno curve is unity.
(a) Derive the equation of a Fanno curve on the T s plane. (b) Dierentiate the equation with respect to temperature. (c) Solve for the maximum entropy point: ds = 0 dT We rst start with the Gibbs equation: P d T ds = du 2 d R (ideal gas) ds = Cv dT T = Cv dT + R dV ( V = constant) T V Integrate the above equation (assume constant Cv ), we obtain
(3)
T)
2 Cp (T0
(perfect gas)
T)
(4)
Combine Equation (3) and Equation (4), we obtain the equation of a Fanno curve
on the T s plane:
T ) ] + constant
We conclude that the ow speed at the maximum entropy point equals the speed of sound. Hence, it corresponds to a Mach 1.0 point. 4. Complete the following table with increases, decreases, remains constant for a ow through a constant-area, frictional duct:
P T V c M P0 0 T0 s A P T P + V2
M <1 M >1 decreases increases decreases increases decreases increases increases decreases decreases increases increases decreases decreases decreases decreases decreases remains constant remains constant increases increases increases increases decreases decreases decreases decreases remains constant remains constant decreases decreases
68
5. A
ow is supplied by a converging nozzle (unchoked). (a) Will the addition of a diverging section increase or decrease the mass
ow rate? The addition of a diverging section will increase the mass
ow rate. The reasons are outlined as follows: { Pressure rises in the diverging section for a subsonic
ow which demands the pressure at the nozzle throat to be lower than the back pressure. { This leads to a higher Mach number at the nozzle throat, hence, a higher mass
ow rate in the system. (b) What about adding a constant-area pipe? Will it increase or decrease the mass
ow rate? The addition of a constant-area pipe will decrease the mass
ow rate. The reasons are outlined as follows: { Pressure drops along the pipe for a subsonic
ow which demands the pressure at the nozzle exit to be higher than the back pressure. { This leads to a lower Mach number at the nozzle exit, hence, a lower mass
ow rate in the system. Pressure distribution of the above three cases can be compared in the following gure:
P / P0 Converging Nozzle Only Pipe Section (Pressure drops)
Pb / P0
(c) Sketch the variation of the mass ow rate, m _ , with back-to-stagnation pressure ratio, Pb=P0, for the above two cases on the same plot and highlight the dierences. The C-D nozzle is choked at a much higher pressure ratio (determined by the exit-to-throat area ratio) than the frictional pipe (determined by fLmax=D). The mass ow rate out of a C-D nozzle is higher than that out of a frictional pipe. Mass ow rate of the above three cases can be compared in the following gure: 69
. mnozzle .
Pb / P0
Converging Nozzle
Frictional Pipe
As a rst step of our analysis, let us choose a control volume as indicated above
by the dotted line. We need to turn to the momentum equation Z Z @ ~ ~ ~ ~ V ~ dA ~ F + F = V ( dV ) + V S B | {z } @t CV {z | } {z } | CS external forces net momentum out
ow storage (acceleration)
and examine the dierences in { momentum ux across the control surfaces { pressure forces on the control volume before we conclude the motion of the cart. As we start lowering the back-to-stagnation pressure ratio, the exit plane pressure at the converging nozzle and the pipe are equal to the back pressure (subsonic ow). Hence, the net pressure force on the cart is zero. 70
Pe V 2 A e Ve2 Ae = R Te e e for both cases. The two exit states can be represented on the T s diagram as follows:
T
T
Pb = Pe
T
Nozzle Exit
Pipe Exit
From the above diagram, we observe that { Pe;nozzle = Pe;pipe { Ve;nozzle > Ve;pipe { Te;nozzle < Te;pipe Hence, we can conclude that the momentum
ux out of the converging nozzle is greater than that out of the frictional pipe. As a conclusion, The cart will move to the right. 7. Fanno Flow in Subsonic Regime
P0,1 , T0,1 Converging Nozzle Pb
(1)
(e)
Consider the above setup. (a) Where can Mach 1.0 be realized? 71
Let us examine the individual terms in the above equation: { Pb=P0;1 = 0:60 as given. { Pb=Pe = 1 because exit pressure equals back pressure for subsonic ow. depends on the exit Mach number, Me (from Fanno ow table). { Pe=Pe;f =P = 1 because Station (1) and Station (e) are on the same Fanno { Pe;f 1;f
curve and are driven to the same reference f state. { P1;f =P1 depends on the inlet Mach number, M1 (from Fanno
ow table). { P1=P0;1 also depends on the inlet Mach number, M1 (from isentropic
ow table). We then simplify to the following relationship: P 0:60 = Pe 1;f P1 (5) Pe;f P1 P0;1 (M1 ; Me). We can relate them by using the pipe geometry:
! !
The pressure ratios in Equation (5) depend on two unknown Mach numbers
f L = f Lmax D D
1
f Lmax D
where { fL=D = 0:5 as given by system specication. { (fLmax =D)1 depends on the inlet Mach number, M1 (from Fanno ow table). 72
f Lmax max (6) 0:5 = f L D 1 D e The solution procedure is an iterative one. It is outlined as follows: i. Guess M1 . ii. Look up the Fanno
ow table to nd out (fLmax=D)1. iii. Calculate (fLmax =D)e from Equation (6). iv. Look up the Fanno
ow table to nd the corresponding Me. v. With these two Mach numbers (M1 ; Me), { look up the Fanno
ow table to nd Pe=Pe;f { look up the Fanno
ow table to nd P1;f =P1 { look up the isentropic
ow table to nd P1 =P0;1 vi. If the product of the above pressure ratios equals 0.60 (Equation (5)), the guess is correct in Step (i). Otherwise, keep guessing dierent values for M1 and repeat the above procedure until the product of pressure ratios converges to 0.60.
(c) Choked Case How much do we need to lower the back-to-stagnation pressure ratio to reach the choking condition? What does this ratio depend on? (Recall that it depends on the exit-to-throat area ratio for a C-D nozzle.) The solution procedure for the choked case is simpler than the unchoked case because we know { Me = 1 { Pb = P1;f This also implies that ! f L = f Lmax = 0:5 D D 1 From the Fanno
ow table, we obtain the inlet Mach number to be M1 = 0:5978 With M1 = 0:5978, { the Fanno
ow table gives P1 = 1:7705 P1;f 73
table). We have
P1 = 0:78538 P0;1 The choking pressure ratio can then be computed as Pb = P1;f P1 = 1 (0:78538) = 0:4436 P0;1 P1 P0;1 1:77051 This choking pressure ratio depends on the value of fL=D of the pipe.
(d) L > Lmax After we lower the pressure ratio to the value we computed in Part (c), the
ow is choked. What do you expect to happen if an extra two meters of pipe section is added to the existing system? Do you expect the
ow is still choked at the exit? Explain your answer by showing the corresponding states on a T s diagram. After the
ow is choked, addition of extra pipe section will reduce the mass
ow rate inside the pipe. This corresponds to switching to another Fanno curve with a smaller mass
ux value (m=A _ ) on the T s diagram. Since the exit pressure cannot be greater than the back pressure in subsonic
ow (no shock mechanism), the
ow leaves the exit subsonically with Pe = Pb. In other words, the addition of extra pipe section unchokes the system. This adjustment of
ow conditions within the system is possible because subsonic
ow can communicate. The addition of extra pipe section downstream can aect the pipe inlet condition upstream. In this case, it reduces the local Mach number at the pipe inlet. As we shall see in the next question, supersonic
ow does not have this communication means. It can only adjust to extra pipe section by shock/expansion mechanism. The above conclusion can be summarized graphically in the following gure:
T
T0
(2) (1)
unchoked (higher entropy due to longer pipe)
Pb
T*
choked
smaller
m A
74
Pb
(1)
(2)
(e)
(a) Slightly dierent from the last problem, there are two possible locations at which Mach 1.0 is attainable in the above setup. Where are they? Mach 1.0 can be realized at { the throat of the C-D nozzle { exit of the frictional pipe. (b) L = Lmax In the supersonic operation mode, determine the pipe length Lmax which gives a sonic ow right at the pipe exit. Under supersonic operation mode, the inlet Mach number, M2 , is governed by the area ratio of the C-D nozzle. For an area ratio of A2=A1 = 2:0, we obtain from the isentropic ow table
M2 = 2:197
= 0:36017
Lmax = 3:602 m
(c) L < Lmax i. For L = 2:0 m, determine the range of the back-to-stagnation pressure ratio, Pb=P0;1, over which 75
A. a normal shock appears in the diverging section of the C-D nozzle There are two limiting cases to consider here: Case I Case II
2a 2b
(1)
(2)
(e)
(1)
(2)
(e)
Case I: upper limit { With the normal shock right at the nozzle throat, the ow goes sub-
sonic in the diverging section of C-D nozzle. { Pipe inlet Mach number, M2 , is governed by the area ratio of the C-D nozzle (subsonic solution from isentropic ow table):
{ { {
P2 = 0:93712 P0;2 With M2 = 0:3060, the Fanno
ow table gives ! f Lmax = 5:031 P2 = 3:5479 P2;f D 2 This further gives the value of fLmax=D at Station (e) ! ! f Lmax = f Lmax f L = 4:831 D e D 2 D From the Fanno
ow table, this corresponds to an exit Mach number of 0.3105 and a pressure ratio Pe = 3:4947 Pe;f This gives the back-to-stagnation pressure ratio for Case I to be P P P Pb = Pb Pe Pe;f 2;f 2 0;2 P0;1 Pe Pe;f P2;f P2 P0;2 P0;1 1 = (1) (3:4947) (1) 3:5479 (0:93712) (1) = 0:9231 M2 = 0:3060;
76
Case II: lower limit { With the normal shock right at the pipe inlet, the ow goes super-
sonic in the diverging section of C-D nozzle until it hits a shock at the nozzle exit, then goes subsonic right before entering the pipe section. { The Mach number upstream of the shock is governed by the area ratio of C-D nozzle (supersonic solution from isentropic ow table):
P2a = 0:093936 P0;2a { With M2a = 2:197, the normal shock table gives P2b = 5:4656 M2b = 0:54744 P2a { With M2b = 0:54744, the Fanno ow table gives ! P2b = 1:9438 f Lmax = 0:74305 P2b;f D 2b M2a = 2:197
max = fL D
2b
f L = 0:54305 D
B. a normal shock appears in the pipe There are two limiting cases to consider here: Case II
2a 2b
Case III
ea eb
(1)
(2)
(e)
(1)
(2)
(e)
Case II: upper limit It has already been studied in the previous part. Case III: lower limit { With the normal shock right at the pipe exit, the ow remains su-
personic from the nozzle throat all the way up to the pipe exit just before the shock, then exits subsonically after the shock. { The pipe inlet Mach number, M2 , is again governed by the area ratio of C-D nozzle (supersonic solution from isentropic ow table). It has been found from the previous part:
P2 = 0:093936 P0;2 { With M2 = 2:197, the Fanno ow table gives ! f Lmax = 0:36012 P2 = 0:35567 P2;f D 2 M2 = 2:197
f Lmax = f Lmax f L = 0:16012 D ea D 2 D { From the Fanno
ow table, this corresponds to a Mach number of Mea = 1:566 and a pressure ratio Pea = 0:57292 Pea;f { With Mea = 1:566, the normal shock table gives Peb = 2:6944 Meb = 0:6790 Pea
78
P P P Pb = Pb Peb Pea Pea;f 2;f 2 0;2 P0;1 Peb Pea Pea;f P2;f P2 P0;2 P0;1 1 = (1) (2:6944) (0:57292) (1) 0:35567 (0:093936) (1) = 0:4077 Hence, a normal shock appears in the pipe section when 0:4077 < Pb < 0:4764 P0;1
For any back-to-stagnation pressure ratio which is lower than the critical value corresponds to Case III, the ow within the C-D nozzle and pipe section will be unaected. All the pressure adjustment will take place outside the pipe. We will expect oblique shocks if the back pressure is higher than the design condition oblique expansion waves if the back pressure is lower than the design condition The back-to-stagnation pressure ratio corresponds to the design condition (free of shock/expansion) is P P P Pb = Pb Pe Pe;f 2;f 2 0;2 P P2 P0;2 P0;1 P0;1 Pe Pe;f 2;f 1 (0:093936) (1) = (1) (0:57292) (1) 0:35567 = 0:1513 C. oblique shocks appear outside the pipe Oblique shocks appear outside the pipe when the back-to-stagnation pressure ratio is between the design condition and the critical value corresponds to Case III: Pb < 0:4077 0:1513 < P 0;1 D. oblique expansion waves appear outside the pipe Oblique expansion waves appear outside the pipe when the back-tostagnation pressure ratio is below the design condition: Pb < 0:1513 P0;1 79
ii. For each of the above cases, A. sketch the process path from the nozzle inlet to the pipe exit on a T s diagram.
T
T0
(2) (e) (b)
T
T0
T*
(1)
T*
T
T0
T
T0
T*
(1)
Pe < P b
(e)
T*
(1)
P e > Pb
(e)
Pb
(2) (2)
Remarks: State (a) and State (b) are the upstream and downstream states of the normal shock respectively.
80
B. sketch the pressure distribution along the streamwise location from the nozzle inlet to the pipe exit. Normal Shock in Nozzle
a b
(1) P P0
0.528
(2)
(e) P P0
(1)
(2)
(e)
a b
0.528
a b
(1)
(2)
(e)
(1)
(2)
(e)
(1) P P0
0.528
(2)
(e) P P0
0.528
(1)
(2)
(e)
(1)
(2)
(e)
(1)
(2)
(e)
C. outline the solution procedure to locate the shock position in cases where normal shock appears. Similar to our previous procedure in locating a normal shock in the diverging section of a C-D nozzle, we need to solve this problem iteratively. The solution procedure is outlined below for the case with a normal shock standing in the diverging section of C-D nozzle (more dicult case): 81
a b
(1)
(2)
(e)
A. Guess the Mach number upstream of the shock, Ma . B. Obtain the pressure ratio, Pa=P0;a, from the isentropic ow table (function of Ma ). C. Obtain the Mach number downstream of the shock, Mb , and the pressure ratios (Pb=Pa; P0;b=P0;a) from the normal shock table (function of Ma ). D. Relate Mb to the Mach number at the nozzle exit, M2 , as follows: A2 = A2 A 2 Ab Aa A1 A 2 Ab A a A1 { A2=A1 = 2:0 as given by system geometry. { A 2 =Ab = 1 because the ow is isentropic from Station (b) to Station (2). { A b =Aa = P0;a =P0;b because the shock process is adiabatic. This ratio depends on Ma . { A a =A1 = 1 because the isentropic ow from Station (1) to Station (a) is choked at Station (1). E. We then obtain the simplied equation: A2 P0;a 2:0 = A P0;b 2 F. With the pressure ratio, P0;a=P0;b, determined in Step (C), the above equation gives the value of A2 =A 2 which denes the Mach number at the nozzle exit, M2 (from the isentropic ow table). G. Obtain P2=P0;2 from the isentropic ow table (function of M2 ). H. Obtain P2=P2;f ; (fLmax =D)2 from the Fanno ow table (function of M2 ). I. Relate the Mach number at pipe exit, Me, with the Mach number at pipe inlet, M2, as follows: ! ! ! fL f Lmax = f Lmax D e D 2 D 82
J. With fL=D = 0:2 (given by system specication), the above equation gives the value of (fLmax=D)e which denes the Mach number at the pipe exit, Me (subsonic solution from the Fanno ow table). from the Fanno ow table (function of Me ). K. Obtain Pe=Pe;f L. The overall back-to-stagnation pressure ratio can then be computed as follows: P P P P P Pb = Pb Pe Pe;f 2;f 2 0;2 0;b 0;a P P2 P0;2 P0;b P0;a P0;1 P0;1 Pe Pe;f 2;f { Pb=Pe = 1 because exit pressure equals back pressure for subsonic exit. =P = 1 because Station (2) and Station (e) are on the same { Pe;f 2;f Fanno curve and are driven to the same reference f state. { P0;2=P0;b = 1 because the ow is isentropic from Station (b) to Station (2). { P0;a=P0;1 = 1 because the ow is isentropic from Station (1) to Station (a). M. We then simplify to the following relationship: P2;f P2 Pb = Pe P0;b P0;1 Pe;f P0;2} P0;a} 2} |P {z | {z | {z | {z } Step K Step H Step G Step C All the pressure ratios have been determined in the previous steps. We just need to multiply all these pressure ratios together and check if their product equals the given back-to-stagnation pressure ratio. If it is, the guess is correct in Step (A). Otherwise, keep guessing dierent values of Ma and repeat the above procedure until the product of pressure ratios converges to the required value. The case with a normal shock standing in the pipe section can be analyzed in a similar manner and is easier! (d) L > Lmax (Common case) Since the fLmax =D values for supersonic ow is much smaller than those of subsonic ow, for most applications, L > Lmax . Since L > Lmax and supersonic ow cannot communicate with downstream, a shock is unavoidable within the pipe section. The ow goes subsonic after the shock. It can communicate with the downstream condition. It either exits the pipe subsonically, matching the back pressure or sonically with an exit pressure higher than the back pressure. To decide between these two possible situations, we need to compare the back pressure, Pb , with the sonic pressure on the Fanno curve, Pf. 83
expected to occur outside the pipe to adjust to the lower back pressure. In our present case, the sonic pressure on the Fanno curve, Pf, can be found as follows: i. From Part (b), we know that M2 = 2:197 in supersonic operation mode. ii. From the isentropic ow table, we obtain P2 = 0:093936 P0;2 iii. From the Fanno ow table, we obtain P2 = 0:35567 P2;f iv. Combine these two pressure ratios, we obtain P2;f P2;f P2 P0;2 = P0;1 P2 P0;2 P0;1 1 (0:093936) (1) = 0:35567 = 0:2641 i. For L = 5:0 m, qualitatively describe the ow in the system for the following pressure ratios: In both cases, there is a normal shock within the system because L > Lmax . A. Pb=P0;1 = 0:50 Since Pb > Pf, the ow exits subsonically with exit pressure equals the back pressure. No pressure adjustment is necessary outside the pipe. B. Pb=P0;1 = 0:10 Since Pb < Pf, the ow exits sonically with exit pressure higher than the back pressure. Pressure adjustment in the form of oblique expansion is expected to occur outside the pipe.
{ If Pb > Pf, the ow exits subsonically and Pe = Pb. { If Pb < Pf, the ow exits sonically and Pe > Pb. Expansion waves are
84
ii. For each of the above cases, A. sketch the process path from the nozzle inlet to the pipe exit on a T s diagram.
T
T0
(b) (e) (1)
Pb > Pf
T
Pe = Pb
Pb < Pf
T0
(b)
T*
Pf*
T*
(1)
Pe = Pf*
(e)
Pb
(a) (2)
(2)
(a)
B. sketch the pressure distribution along the streamwise location from the nozzle inlet to the pipe exit.
Pb > Pf
a b
Pb < Pf
a b
(1) P P0
0.528
(2)
(e) P P0
a b
0.528
(1)
(2)
(e)
a b
(1)
(2)
(e)
(1)
(2)
(e)
85
2. Trace out a Rayleigh curve on a T s plane and locate the corresponding stagnation and sonic states on the same diagram.
(1)
(2)
(3)
(4)
(5)
(6) (7)
(8)
(9)
(a) Mark the following statements as true or false: P0;1A 1 = P0;7 A7 P0;6A 6 = P0;9 A9 (b) Sketch the process path for the above system on a T s diagram. Indicate both the static and stagnation states. Are State (2) and State (8) on the same Fanno curve?
86
(c) Sketch the pressure distribution along the streamwise location from Station (1) to Station (9).
4. Show the following results are true on a Rayleigh curve: p (a) M = 1= k at the maximum temperature point (b) M = 1 at the maximum entropy point 5. Complete the following table with increases, decreases, remains constant for a ow through a constant-area duct with heat transfer:
P T V M P0 0 T0 s A P T
6. In your opinion, which assumption(s) in the Rayleigh ow analysis may be potential source(s) of error in solving a real life problem? Remarks: The procedure for solving Rayleigh ow problems is very similar to the approach used for Fanno ow except that the link between any two ow states in Rayleigh ow is determined by heat transfer q=m rather than wall friction fL=D. 87
7. This problem examines the eects of heat transfer on the
ow pattern ahead of a combustor inlet.
Heat Addition
(1)
Combustion Chamber
(2)
(a) Unchoked Case Air at 250 K and 1.0 bar is moving at 100 m/sec towards the entrance of a combustion chamber. Determine the exit conditions if 300 kJ/kg is added to the ow during the combustion process.
(b) Just Choked Case How much more heat can be added without changing the conditions at the entrance to the combustor?
(c) Choked Case Let us add sucient fuel to the system so that the exit stagnation temperature is raised to 1500 K now. Assume that the receiver pressure is very low. What do you expect to happen in the system? Describe the ow both qualitatively and quantitatively.
88
8. A converging-diverging nozzle with an area ratio of 3.0, supplies air to a supersonic wind tunnel. If there is moisture in the air, it is possible for the water vapor to condense during the expansion process if the local static temperature drops below the saturation temperature. In practice, this condensation process occurs very rapidly, leading to an almost discontinuous change in the ow properties. Assume that the stagnation temperature of the air/water vapor mixture entering the nozzle is 600 K and that the mass fraction of water vapor in the stream is Yw = mwater =mmixture = 0:01. The saturation temperature for the air/water vapor mixture is 14 C and the heat of vaporization of water is 2470 kJ/kg.
(1)
Condensation Front
(2)
Test Section
(b) Determine the test section Mach number i. without condensation ii. with condensation How is the test section Mach number aected by the presence of the condensation process?
(c) Sketch the nozzle expansion process without and with condensation on a T diagram, including all appropriate stagnation and sonic states.
89
The Rayleigh curve is derived from the conservation of mass and the conservation
of linear momentum principles. All the states on the same Rayleigh curve have the same mass
ux (m=A _ ), and 2 impulse (thrust) function, (P + V ).
2. Trace out a Rayleigh curve on a T s plane and locate the corresponding stagnation and sonic states on the same diagram.
Along a Rayleigh curve, neither the stagnation temperature nor the stagnation
pressure is constant. Hence, the stagnation states are not as trivial as those of the previous
ows we have studied in class. Nevertheless, the conservation of energy principle tells us that the stagnation temperature always increases when there is heat addition into the
ow regardless the
ow is subsonic or supersonic. Based on this conclusion, we can draw the stagnation lines (with positive slopes) on the T s diagram in the following gure:
T
M=1/ k
We then need to address the question: \Which stagnation line corresponds to the
subsonic/supersonic regime?" To answer this question, it will be helpful to think about a normal shock process on a Rayleigh line. Justication: Across a shock, (a) V = constant (b) P + V 2 = constant 90
(c) T0 = constant Conditions (a) and (b) guarantee that the upstream and downstream states of a shock lies on the same Rayleigh line. It is represented as follows:
T
T0,1 = T0,2
(2) subsonic
(1) supersonic
T0;2 = T0;1 Hence, the stagnation states of (1) and (2) must be on the same horizontal line on the T s diagram. Because s2 > s1 (due to irreversibility within the shock), we can conclude that the upper stagnation curve corresponds to the supersonic regime and the lower one corresponds to the subsonic regime. To trace out the sonic states, we need to realize T0 = k + 1 = constant T 2 Hence, the shapes of the sonic curves resemble those of the stagnation curves. As a conclusion, all the stagnation and sonic states can be summarized in the following diagram:
T
M=1/ k
Supers
Locus of all stagnation states
onic
M=1
Su
ic bson
91
(1)
(2)
(3)
(4)
(5)
(6) (7)
(8)
(9)
(a) Mark the following statements as true or false: P0;1A 1 = P0;7 A7 P0;6A 6 = P0;9 A9 { The product, P0 A , is constant for a steady, adiabatic ow. { In our case here, the ow is adiabatic from Station (1) to Station (7) but heat addition occurs from Station (7) and Station (8). Hence, P0;1 A 1 = P0;7 A7 P0;6 A 6 6= P0;9 A9 { From conservation of mass, we can show that P 0 A p = constant T0 Since the heat addition from Station (7) to Station (8) raises the stagnation temperature at Station (9) relative to that of Station (6), we conclude that P0;6 A 6 < P0;9 A9
92
(b) Sketch the process path for the above system on a T s diagram. Indicate both the static and stagnation states.
T
T0
(1) (2) (8) (3) (7) (9)
T*
(4)
(6) (5)
Are State (2) and State (8) on the same Fanno curve? Since there is heat addition from Station (7) to Station (8), Station (2) cannot be on the same Fanno curve as Station (8) [h0;8 > h0;2 but (m=A _ )8 = (m=A _ )2:] (c) Sketch the pressure distribution along the streamwise location from Station (1) to Station (9).
Shock Heat Addition
(1) P P0 1.0
(2)
(3)
(4)
(5)
(6) (7)
(8)
(9)
0.53
93
4. Show the following results are true on a Rayleigh curve: (a) M = 1 at the maximum entropy point (b) M = 1= k at the maximum temperature point and dT=ds based on the following principles: { Gibbs equation { ideal gas equation { conservation of mass { conservation of linear momentum Then set the derivatives to zero to obtain the corresponding Mach number condition. Let us start with the Gibbs equation: P d T ds = du 2 d R (ideal gas) ds = Cv dT T through mass and momentum conservation. Let P + V 2 = C1 Momentum conservation V = C2 Mass conservation we have
2 2 = C P+C 1 2 2 = C RT + C 1
The strategy to prove the necessary results is to obtain the derivatives: ds=dT
We then try to relate the density change, d, with temperature change, dT ,
The above equation denes a one-to-one correspondence between density and temperature, i.e. knowing one enables us to nd the other. We can relate the dierentials in the above equation as follows: 2 2 d = 0 R dT + R T d C 2 or equivalently d = R dT V2 RT 94
We are ready to substitute the above results into the Gibbs equation now. It will
R2 dT ds = Cv dT T V2 RT Further rearrangement gives ds = Cv R2 dT T V2 RT dT = T (V 2 R T ) ds C V2 C RT
v p
give
(7) (8)
(a) At the maximum temperature point, dT = 0 ds Equation (8) implies V2 = RT T = kR k 2 = c k 1 2 ) M = k 1 or M = p k (b) At the maximum entropy point, Equation (7) implies
ds = 0 dT
V2
Cv = R2 T V2 RT 2 R T = RC T v R 2 V = RT 1 + C v C = R T Cp v = kRT = c2 M = 1
95
(Cp = Cv + R)
5. Complete the following table with increases, decreases, remains constant for a ow through a constant-area duct with heat transfer:
P T V M P0 0 T0 s A P T
M <1 (heating) decreases decreases p increases (0 < pM < 1= k) decreases (1= k < M < 1) increases increases decreases decreases increases increases increases decreases decreases increases
M >1 M <1 (heating) (cooling) increases increases increases increases p increases decreases (0 < M < 1 = k) p increases (1= k < M < 1) decreases decreases decreases decreases decreases increases decreases increases increases decreases increases decreases increases decreases decreases increases decreases increases increases decreases
6. In your opinion, which assumption(s) in the Rayleigh ow analysis may be potential source(s) of error in solving a real life problem?
Heat transfer causes the stagnation temperature to change signicantly in the
ow, which leads to a large variation of static temperature. The perfect gas assumption (constant specic heats) may not be appropriate. At higher and higher temperatures, more and more energy modes are activated within the molecules. In general, this causes the specic heats to rise with temperature. In cases where combustion occurs, chemical composition of the constituent gases changes signicantly. Reactant species will be consumed and product species will be produced. Their relative ratio changes as combustion proceeds. Values like gas constant, R, and specic heat ratio, k, will no longer be constant but depend on the extent of combustion.
Remarks: The procedure for solving Rayleigh ow problems is very similar to the approach used for Fanno ow except that the link between any two ow states in Rayleigh ow is determined by heat transfer q=m rather than wall friction fL=D.
96
7. This problem examines the eects of heat transfer on the
ow pattern ahead of a combustor inlet.
Heat Addition
(1)
Combustion Chamber
(2)
(a) Unchoked Case Air at 250 K and 1.0 bar is moving at 100 m/sec towards the entrance of a combustion chamber. Determine the exit conditions if 300 kJ/kg is added to the ow during the combustion process. In this unchoked case, the chamber inlet can take in all the ow which comes towards it. Hence,
T1 = 250 K
V1 = 100 m/sec
For T1 = 250 K, c1 = 317 m/sec. This gives an inlet Mach number of From the isentropic
ow table, we obtain
T1 = 0:9805 ) T0;1 = 255 K T0;1 From the Rayleigh
ow table, we obtain T0;1 = 0:3763 T1 = 0:4427 P1 = 2:106 T0;1;R T1;R P1;R
T0;2 = T0;2 T0;1 T0;1;R T0;2;R T0;1 T0;1;R T0;2;R = (2:172) (0:3763) (1) = 0:8173 From the Rayleigh
ow table, this corresponds to an exit Mach number of (subsonic) M2 = 0:5985 and T2 = 0:9152 P2 = 1:598 T2;R P2;R The exit state is T T1;R 1 T1 = (0:9152) (1) 0:4427 T2 = TT2 T2;R T (250 K) = 517 K 2;R 1;R 1 P P1;R 1 (1:0 bar) = 0:7588 bar P2 = PP2 P2;R P P1 = (1:598) (1) 2:106
2;R 1;R 1
At Station (2),
(b) Just Choked Case How much more heat can be added without changing the conditions at the entrance to the combustor? For the just choked case, M2 = 1. This implies T0;2 = T0;1;R Apply the same energy equation in Part (a), we have q = Cp ( T0;2 T0;1 ) = Cp ( T0;1;R T0;1 ) T = Cp T0;1 ( 0;1;R 1 ) T0;1! 1 kJ 1 (255 K) = 1:004 kg K 0:3763 = 424 kJ/kg Hence, we can add an extra 124 kJ/kg into the ow before we choke it thermally. At this choked condition, the stagnation temperature at Station (2) is T0;1;R 1 T0;2 = T0;1;R = T T0;1 = 0:3763 (255 K) = 678 K 0;1 98
(c) Choked Case Let us add sucient fuel to the system so that the exit stagnation temperature is raised to 1500 K now. Assume that the receiver pressure is very low. What do you expect to happen in the system? Describe the ow both qualitatively and quantitatively. In this case,
(1)
(2)
Combustion Chamber
dierent Rayleigh line with a smaller m=A _ value. Since the receiver (back) pressure is very low, we can assume that the ow is choked at the exit station (2), i.e.
}
Spillage
M2 = 1
T1 = 0:2024 P1 = 2:275 T1;R P1;R Recall that the exit state is the R reference state due to choking, we can conclude the exit condition to be 1 (253 K) = 1250 K T2 = T1;R = TT T1 = 0:2024 1 (1:04 bar) = 0:457 bar P2 = P1;R = PP P1 = 2:275 This heat addition process can be described qualitatively by the following T s diagram:
1;R 1 1;R 1
T
exit inlet
smaller
m A
100
8. A converging-diverging nozzle with an area ratio of 3.0, supplies air to a supersonic wind tunnel. If there is moisture in the air, it is possible for the water vapor to condense during the expansion process if the local static temperature drops below the saturation temperature. In practice, this condensation process occurs very rapidly, leading to an almost discontinuous change in the
ow properties. Assume that the stagnation temperature of the air/water vapor mixture entering the nozzle is 600 K and that the mass fraction of water vapor in the stream is Yw = mwater =mmixture = 0:01. The saturation temperature for the air/water vapor mixture is 14 C and the heat of vaporization of water is 2470 kJ/kg.
Condensation Front
(1)
(c)
(2)
Test Section
(a) Determine the location of the condensation front, in terms of Acond=Athroat . Condensation occurs when Tc = 14C = 287 K. Tc = 287 = 0:4783 ) T0;c 600
ii. with condensation When the water vapor condenses, it releases the heat of vaporization to the air ow. The amount of heat release is q_ = m _ water hfg Since the condensation process occurs over a very short distance in space, we can assume that there is no signicant area variation during the heat addition process. Hence, we can apply our Rayleigh ow analysis to analyze this ow. Let us label the states as follows in our analysis:
cx cy
(1)
(c)
(2)
(0:01) 2470 kJ = 1:004 kJ (600 K) T0;cy 1 kg kg K T0;cx T0;cy = 1:041 ) T0;cx We then relate the Mach number upstream and downstream of the condensation front as follows: T0;cy = T0;cy T0;cy;R T0;cx;R T0;cx T0;cy;R T0;cx;R T0;cx 1 1:041 = T0;cy (1) T0;cy;R 0:73404 T0;cy = 0:7641 ) T0;cy;R 102
T0;cx = 0:73404 T0;cx;R Apply conservation of energy across the condensation front, we have q_ = m _ mixture Cp (T0;cy ! T0;cx) m _ water h = C T T0;cy 1 fg p 0;cx m _ T
mixture
!
0;cx !
Acy = 1:926 A cy
from the isentropic
ow table. After the condensation front, the
ow continues to expand isentropically from Station (cy) to Station (2). Their Mach numbers can be related by their area ratio. Acy = Acy A cy A2 A2 A cy A2 A2 2:264 = (1:926) (1) A 2 3 A2 A 2 ) A = 2:552
2
M2 = 2:465
How is the test section Mach number aected by the presence of the condensation process? Without condensation, M2 = 2:637. With condensation, M2 = 2:465. The condensation process reduces the Mach number in the test section. This is because the heat addition process slows down the supersonic
ow at the condensation front (Mcx = 2:335 ! Mcy = 2:154).
103
(c) Sketch the nozzle expansion process without and with condensation on a T diagram, including all appropriate stagnation and sonic states.
T T0 T*
(1)
104
2. The Reynolds Transport Theorem is the core basis in control volume analysis. It serves as a bridge between the control mass and the control volume approach. We can state it as: ! Z Z dN @ ~ dA ~ = ( dV ) + V dt @t For the following principles, what are the quantities N and ? (a) conservation of mass (b) conservation of linear momentum (c) conservation of angular momentum
system CV CS
105
3. What is the main criterion in choosing a suitable control volume in problem solving?
5. Choose the best answer in the following question: Pressure is always directed into / out of the control volume of interest. 6. The angular momentum principle can be expressed in the following two forms: Form 1: Z ~ ~shaft ~ r Fs + (~ r~ g) ( dV ) + T CV Z Z ~ ( dV ) + ~ V ~ dA ~ = @ ~ r V ~ r V @t CV CS Form 2: Z ~ ~shaft ~ r Fs + (~ r~ g) ( dV ) + T
Z
~ V ~ dA ~ = @t ~ rV CV CS (a) What is the main dierence between the above two forms?
(b) Give a verbal description to each term in the equations.
CV @ Z
_ ~ ~ +! ~ r 2~ !V ~ (~ !~ r) + ~ ! r ( dV )
CV
~ ( dV ) + ~ rV
106
7. A total water discharge of 200 cm3 is issued from a sprinkler as shown in the following gure:
Aj = 0.1 cm
2
10 cm 40 o
10 cm
10 cm
Wj
Wj
Wj 40 o
A
Wj Wj Wj 10 cm
10 cm
10 cm
Assume that the jet speed is the same from all the holes. (a) Static Case: Determine the torque that must be applied to the sprinkler arms to hold them from rotating.
(b) Frictionless Case: Determine the angular speed if the arms are free to rotate and there is no friction.
(c) Frictional Case: Determine the angular speed if there is a constant frictional torque of 1 N-m resisting rotation of the arms.
(d) Relative Motion: Determine the absolute velocity of the uid leaving Hole A in Part (b) and (c).
(e) Sketch the corresponding velocity vector diagrams for Part (d).
107
8. Refer to the schematic below, a \wye" joint splits a pipe
ow into two equal amounts, Q=2, which exit at a distance R0 from the x-axis. The system rotates about the x axis at a rate
.
Q/2 , Q R0
R0 >> Dpipe x
Q/2
(a) Inertial Frame Analysis: Apply the angular momentum principle in an inertial frame to i. determine the torque required to turn the pipe (constant speed).
ii. determine the additional torque which is required to generate an angular _ on the existing system (constant acceleration). acceleration
(b) Rotating Frame Analysis: Repeat the analysis in Part (a) in a rotating frame.
108
Pmass = 0
(b) conservation of linear momentum According to Newton's law of motion, we can change the momentum of a system by apply an external force on it. X ~ Plin:mom: = F
There are two main types of forces: { Surface force It is present along the control surfaces of your selected control volume. The best way to identify the surface forces is to trace along the entire
control surface and ask yourself the question \What force does my control volume experience along this surface?" Some examples of surface force are pressure force (normal direction) and friction (tangential direction). { Body force It is present due to the contents inside the selected control volume under the in uence of the surrounding force eld. Some examples of body force are gravitational force and electrostatic force.
(c) conservation of angular momentum If we draw the analogy between linear momentum in translational motion and angular momentum in rotational motion, external force will be analogous to external torque. Hence, X Pang:mom: = ~ 109
Since the angular momentum equation is derived from the linear momentum
~ , (both surface and body) in the linear equation, all the external forces, F ~ , on the same momentum equation are capable of generating torque, ~ rF control volume as long as the line of action of the force does not pass through the center of rotation.
(d) conservation of energy We can change the total energy of a system by adding heat (Q) or doing work (W ) on the system Penergy = Qin + Win 2. The Reynolds Transport Theorem is the core basis in control volume analysis. It serves as a bridge between the control mass and the control volume approach. We can state it as: ! Z Z dN @ ~ dA ~ ( dV ) + V = dt system @t CV CS with a xed amount of substance (control mass approach). However, this approach is not easy to follow for a uid system simply because a uid can be deformed continuously as it moves around in space. The Reynolds Transport Theorem relates the rate of change of an extensive property N of a control mass system
Most physical laws are Lagrangian in nature, i.e. they are derived for a system
dN dt
system
with the rate of change of the same property in a control volume @ Z ( dV ) + Z V ~ dA ~ @t CV {z | } | CS {z } storage out ow - in ow
There are two major components in the above equation: { The rst one is the storage term which accounts for the rate of increase in
property N within the control volume. { The second one is the net out
ow term which accounts for the loss of property N due to the
uid motion in and out of the control volume. Only with the Reynolds Transport Theorem, we can then relate the physical laws to what we measure in a xed region in space (control volume approach). 110
For the following principles, what are the quantities N and ? The quantity N is an extensive property of the system whereas the quantity is its intensive counterpart. (a) conservation of mass
Remarks: With the results of Question 1 and 2, we can summarize all the conservation laws in the following form: @ Z ( dV ) + Z V ~ dA ~ = P @t CV CS 3. What is the main criterion in choosing a suitable control volume in problem solving? We should put the control surfaces at places where ~ ;P) { we know how the ow behaves, for example, (V { we want to know something about, for example, frictional shear, exit pressure. 4. From what physical principle is the angular momentum equation derived? The angular momentum equation is derived by taking the cross product between the position vector, ~ r, and the linear momentum equation. Hence, its main physics comes from Newton's law of motion. The main dierence is that the linear momentum equation governs the translational motion while the the angular momentum equation governs the rotational motion of the system. 111
5. Choose the best answer in the following question: Pressure is always directed into the control volume of interest. Pressure is a compressive force. Hence, it is always directed into the system of interest. 6. The angular momentum principle can be expressed in the following two forms: Form 1: Z ~ ~shaft ~ r Fs + (~ r~ g) ( dV ) + T CV Z @ Z ~ ~ V ~ dA ~ ~ ( dV ) + ~ rV rV = @t CS CV Form 2: Z ~shaft ~ (~ r~ g) ( dV ) + T ~ r Fs +
Z
~ V ~ dA ~ ~ rV = @t CV CS (a) What is the main dierence between the above two forms? Form 1 is derived in an inertial frame. Form 2 is derived in a rotating (non-inertial) frame. ~ , is consistent with the cor We need to make sure that the velocity vector, V responding choice of reference frame when we invoke the angular momentum principle. As long as we use the two forms in a consistent manner, they should give identical results.
(b) Give a verbal description to each term in the equations. ~s is the torque generated by surface force, F ~s. ~ r F R CV (~ r~ g)( dV ) is the torque generated by gravitational force. R ~ ) ( dV ) is the torque generated by Coriolis force. RCV ~ r (2~ !V r [~ ! (~ !~ r)] ( dV ) is the torque generated by centripetal force. RCV ~ _ CV ~ r (~ ! ~ r) ( dV ) is the \ctitious" torque due to angular acceleration of the rotating reference frame. R @ ~ ) ( dV ) is the rate of increase in angular momentum within the @t CV (~ rV control volume. R ~ ) ( V ~ dA ~ ) is the net out
ow of angular momentum caused by
uid CS (~ rV motion in and out of the control volume. 112
CV @ Z
i _ ~ ~ +! ~ r 2~ !V ~ (~ !~ r) + ~ ! r ( dV )
CV
~ ( dV ) + ~ rV
7. A total water discharge of 200 cm3 is issued from a sprinkler as shown in the following gure:
Aj = 0.1 cm
2
10 cm 40 o
10 cm
10 cm
Wj
Wj
Wj 40 o
A
Wj Wj Wj 10 cm
10 cm
10 cm
Assume that the jet speed is the same from all the holes. above. Let us solve this angular momentum problem using an inertial reference frame and see how the analysis works. The corresponding angular momentum equation is Z Z X @ ~ ~ V ~ dA ~ ~ = @t ~ r V ( dV ) + ~ rV CV CS We then examine every individual term in the above equation: { The sources of external torque in this problem come from shaft torque, ~ shaft frictional torque, ~ f { The storage term is zero because we are dealing with a steady problem. { The net angular momentum out ow term can be evaluated by rst considering the jet out of one hole only:
Wj
We rst choose a control volume to include the entire sprinkler arm as indicated
ri
ri
~ dA ~ = (r Vt) ( Wj Aj ) ~ V ek = (r Vt) Q ek 6 ~ where Vt is the tangential velocity component measured in an inertial frame.
~ ~ rV
113
Vt = Ut + Wt = ri ! + Wj sin ) r Vt = ri2 ! + ri Wj sin = ri2 ! + ri Q sin 6 Aj { Hence, the angular momentum out
ow from this hole is equal to ! Q Q 2 ri 6 A sin ri ! 6 ~ ek j { The total angular momentum out
ow can be obtained by summing the contribution from all six holes together
Q Q 2 ~ ~ ~ ek ~ r V V dA = 2 ri 6 A sin ri ! 6 ~ CS j i=1 where r1 = 10 cm, r2 = 20 cm, r3 = 30 cm. The entire angular momentum equation can then be simplied to " # Q Q 2 2 2 shaft + f = 3 6 A (r1 + r2 + r3) sin ! r1 + r2 + r3 j Z
3 " X
This general equation of motion forms the common basis for the following special
cases of interests. (a) Static Case: Determine the torque that must be applied to the sprinkler arms to hold them from rotating. In this static case, we have {!=0 { f = 0 The restraining torque is 2 ! Q shaft = 18 A (r1 + r2 + r3 ) sin j (b) Frictionless Case: Determine the angular speed if the arms are free to rotate and there is no friction. In this frictionless case, we have 114
{ shaft = 0 { f = 0
r1 + r2 + r3 sin ! = 6Q 2 + r2 + r2 Aj r1 2 3
(c) Frictional Case: Determine the angular speed if there is a constant frictional torque of 1 N-m resisting rotation of the arms. In this frictional case, we have { shaft = 0 The angular speed of the sprinkler arm is " # Q 3 1 f ! = r2 + r2 + r2 6 A (r1 + r2 + r3 ) sin Q j 1 2 3 which reduces to the results in Part (b) for f = 0. (d) Relative Motion: Determine the absolute velocity of the uid leaving Hole A in Part (b) and (c). ~ =U ~ +W ~ to In both cases, we can apply the relative motion equation V analyze the velocity components. { Radial component: Vr = Ur + Wr = 0 + Wj cos = Q cos 6 Aj { Tangential component: Vt = Ut + Wt = r3 ! + Wj sin = 6Q A sin r3 !
(e) Sketch the corresponding velocity vector diagrams for Part (d).
W= Q 6 Aj
Vr = Wr
Vt = U Wt
U = r3
8. Refer to the schematic below, a \wye" joint splits a pipe
ow into two equal amounts, Q=2, which exit at a distance R0 from the x-axis. The system rotates about the x axis at a rate
.
Q/2 , Q R0
R0 >> Dpipe x
Q/2
(a) Inertial Frame Analysis: Apply the angular momentum principle in an inertial frame to i. determine the torque required to turn the pipe (constant speed). ii. determine the additional torque which is required to generate an angular _ on the existing system (constant acceleration). acceleration
The angular momentum equation in an inertial frame is Z Z X @ ~ ~ V ~ dA ~ ~ = @t ~ r V ( dV ) + ~ rV CV CS Let us rst consider a control volume which includes the upper tube only: 116
CV1
Q/2
dr
, Q
R0
z
3 0
sin ~ = A R2 i cos ~ j 3 sin The storage term can then be obtained by take the time derivative of the above expression. (Notice that only the angular speed is a function of time.) 3 _ @ Z ~ 0
~ ( dV ) = A R2 ~ ~ r V sin i cos j @t CV 3 sin The angular momentum out
ow term can be obtained by evaluating the ~ expression at r = R0 = sin ~ rV " 2 Z # Q R
0 ~ ~ ~ ~ ~ 2 ~ r V V dA = sin sin i cos j CS
1 1
CV1
117
These procedures conclude the analysis of the upper tube. We can then
proceed to analyze the lower tube by following the same procedures.
, Q
r dr
y
R0
z
Q/2
CV2
vectors
to sum up the results from the upper and the lower tubes. Z @ Z ~ ~ ~ V ~ dA ~ r V ( dV ) + ~ r V @t CV CS 3 ! 2 A R 0 2 _ ~ = Q R0
+ 3 sin
i From the angular momentum equation in an inertial frame, we deduce that the applied torque is 3 ! 2 A R 0 2 _ ~ ~ = Q R0
+ 3 sin
i The rst term corresponds to the torque required to turn the pipe at constant speed,
, while the second term corresponds to the additional _ on the existing torque required to produce an angular acceleration,
, system. Hence, 2
~ ~
= Q R0 i 3 _ 2 A R0
~ ~
i _ = 3 sin 118
To obtain the global conservation equation for the entire system, we need
CS2
(b) Rotating Frame Analysis: Repeat the analysis in Part (a) in a rotating frame. The angular momentum equation in a rotating frame is
~
~ V ~ dA ~ ~ ( dV ) + ~ rV ~ rV = @t CS CV Let us rst consider a control volume which includes the upper tube only:
CV1
CV @ Z
h i _ ~ ~ +~ ~ r 2~ !V ! (~ !~ r) + ~ ! r ( dV ) Z
Q/2
dr
, Q
R0
z
Hence, the storage term and the net out
ow term both equal zero and do not contribute to the angular momentum balance in the rotating frame. Let us evaluate each cross product term in the \ctitious" torque carefully: ~ = 2
~ ~ ~ 2~ !V i 2Q cos i + sin j A
sin ~ ~ = Q i j A Q = A
sin ~ k h i ! ~ ~ r =
~ i r cos ~ i + sin ~ j = r
sin ~ i ~ j = r
sin ~ k 119
! ~ (~ ! ~ r) = = = _ ! ~ ~ r = = =
~ i r
sin ~ k 2 r
sin ~ i~ k ~ r
2h sin j i _~
i r cos ~ i + sin ~ j _ sin ~ i ~ j r
_ sin ~ r
k
Q
_ _ ~ 2~ !V + ~ ! (~ !~ r) + ~ ! ~ r = sin A + r
~ k r
2 sin ~ j and h i
+ r
_ ~ ~ +~ _ ~ ~ r 2~ !V ! (~ ! ~ r) + ~ ! r = r sin2 Q i A Q
_ j r2
2 sin cos ~ k r sin cos A + r
~ The combined \ctitious" torque can be obtained by integrating the above expression along the entire upper tube length Z
CV1 Z R0 = sin
0
"
h i _ ~ ~ +~ ~ r 2~ !V ! (! ~ ~ r) + ~ ! r ( dV ) Q
Q
_ _ ~ r i r sin cos A + r
~ j A + r
i r2
2 sin cos ~ k ( A dr)
sin2
2 cos Q
_ ! _ ! R R R Q
0
0
0 ~ ~ + i + j = ( A) 2 A 3 sin sin 2 A 3 sin 3
2 cos # R0 ~ k 3 sin2 These procedures conclude the analysis of the upper tube. We can then proceed to analyze the lower tube by following the same procedures. 2 R0
120
, Q
r dr
R0
z
Q/2
CV2
The main dierences will be in the expressions of position and velocity vectors
~ r = r cos ~ i sin ~ j ~ = Q cos ~ ~ V i sin j 2A After algebraic manipulations, we obtain Z h i _ ~ ~ +~ ~ r 2~ !V ! (~ ! ~ r) + ~ ! r ( dV )
CV2
"
2 cos Q
_ ! _ ! R R R
Q
0
0 0 ~ i + sin + 3 sin ~ j = ( A) 2 A + 3 sin 2 A 3
2 cos # R k + 0 2 ~ 3 sin To obtain the global conservation equation for the entire system, we need to sum up the results from the upper and the lower tubes. Z h i _ ~ ~ +~ ~ r 2~ !V ! (! ~ ~ r) + ~ ! r ( dV ) 2 R0
_ !# Q
R 0
~ = ( A) 2 A + 3 sin i From the angular momentum equation in a rotating frame, we deduce that the applied torque is Z h i _ ~ ~ +~ ~ = ~ r 2~ !V ! (! ~ ~ r) + ! ~ r ( dV )
2 2 R0
CV
"
_ !# R Q
0
~ = ( A) 2 A + 3 sin i The rst term corresponds to the torque required to turn the pipe at constant speed,
, while the second term corresponds to the additional torque required _ on the existing system. Hence, to produce an angular acceleration,
, 2
~ ~
= Q R0 i 3
_ 2 A R0 ~ ~
_ = 3 sin i
2 2 R0
CV
"
121
either the inertial or rotating reference frame to analyze angular momentum problems as long as the velocity vector is consistent with the corresponding chosen reference frame. In this problem, { the production, the storage and the net out ow terms are all active in the inertial frame analysis; { the production term and the \ctitious" torque are both active but the storage and the net out ow terms are both zero in the rotating frame analysis.
This example demonstrates that we can obtain the same results by choosing
122
2. Every ow is characterized by a Reynolds number. (a) How is it dened? (b) Give a physical meaning to it. (c) Why do we ignore its eects in the analysis of turbomachinery?
3. What is/are the necessary condition(s) for two ows to be dynamically similar?
123
(b) Can you give a physical meaning to specic speed? (c) It can be expressed by two common forms. What are they?
T = Q (r2Vt;2 r1Vt;1 ) Pw = Q (U2 Vt;2 U1 Vt;1) 1 (U V U1Vt;1 ) H = g 2 t;2 (a) What are the assumptions behind these equations?
(b) From which basic principles and denitions are they derived?
7. The least trivial quantity in the above Euler turbomachinery equations is the uid tangential velocity relative to an inertial frame, Vt. We often need to invoke the following equation to nd it out: ~ = U ~ +W ~ V How do you verbally describe the dierent velocity vectors in the above equation?
124
8. We encounter two common types of turbomachine: centrifugal
ow type axial
ow type Prociency in analyzing each of them is crucial in designing machines with better performance. We will examine both of them in this problem. (a) Centrifugal Flow Type:
Backward Curved Blade
i. Draw the velocity vector diagram at the outlet. ii. Which velocity component is responsible for carrying uid in and out of the machine? iii. Determine the following velocity components:
Ur Ut Wr Wt Vr Vt
= = = = = =
iv. Determine the torque and power input. v. What is the theoretical head? Give a verbal description to it. 125
Stator Blade 1
Rotor Blade
i. Draw the velocity vector diagram at the inlet. ii. Which velocity component is responsible for carrying uid in and out of the machine? iii. Determine the following velocity components:
Ux Ut Wx Wt Vx Vt
= = = = = =
126
by three non-dimensional parameters: { CH = g H=(!2 D2) head coecient 3 5 { CP = P=( ! D ) power coecient 3 { CQ = Q=(! D ) ow coecient Choosing the ow coecient to be the independent non-dimensional parameter, we can express the head coecient and the power coecient as:
CH = f1 (CQ) CP = f2 (CQ)
Remarks: There are two other non-dimensional parameters which can be formed by combining the above three non-dimensional parameters: { specic speed (to be discussed in Problem 4) { eciency, = CH CQ= CP 2. Every
ow is characterized by a Reynolds number. (a) How is it dened? Reynolds is dened as L Re = V where V is a characteristic velocity while L is a characteristic length of the
ow problem. Take the example of a uniform
ow over a sphere. The characteristic velocity will be the free stream velocity and the characteristic length will be the diameter of the sphere. (b) Give a physical meaning to it. The Reynolds number can be interpreted as the ratio of inertial to viscous force in a
ow. (c) Why do we ignore its eects in the analysis of turbomachinery? The higher the
ow Reynolds number gets, the less important the viscous eects are, compared with the inertial eects. As the
ow Reynolds number exceeds a certain limit, we can ignore the viscous eects without causing too much error in our analysis. Take the example of the Moody diagram. The friction factor is a function of both relative roughness (e=D) and Reynolds number (Re). 128
As the Reynolds number reaches 106 or above, the value of friction factor
3. What is/are the necessary condition(s) for two
ows to be dynamically similar? The two
ows must be geometrically similar. The independent non-dimensional parameters which characterize the
ow must be of comparable magnitude.
approaches nearly a constant value for relative roughness greater than 0.001 and becomes independent of the Reynolds number.
4. (a) What is the unique characteristic of specic speed as a non-dimensional parameter? The specic speed does not depend on the size of the system. (b) Can you give a physical meaning to specic speed? Specic speed can be interpreted as the operation speed of the machine at which it produces a unit head rise per unit
ow rate through the system. (c) It can be expressed by two common forms. What are they? Specic speed is another non-dimensional parameter which characterizes the machine operation. It can be formed by combining dierent non-dimensional parameters in such a way that the size of the system (D) is factored out of its denition. It can be formed by combining the
ow coecient and the head coecient as follows: C 1=2 ! Q1=2 Ns = Q 3=4 = H 3=4 CH or by combining the power coecient and the head coecient as follows:
1=2 P 1=2 C P Ns = 5=4 = !H 5=4 CH
5. What does the net head of a pump signify? The net head of a pump signies the amount of energy gained by the uid as it passes through the machine. It is related to the power required to drive the pump through the pump eciency. 129
T = Q (r2Vt;2 r1Vt;1 ) (9) Pw = Q (U2 Vt;2 U1 Vt;1) (10) 1 (11) H = g (U2Vt;2 U1Vt;1 ) (a) What are the assumptions behind these equations? The main assumptions are: steady state (no storage term) uniform
ow at each station (
ow variables do not vary with azimuthal direction) (Optional) In evaluating the tangential velocity component Vt, we often invoke one more assumption in the analysis: attached
ow is achieved on the blade surface (no
ow separation).
(b) From which basic principles and denitions are they derived?
7. The least trivial quantity in the above Euler turbomachinery equations is the uid tangential velocity relative to an inertial frame, Vt. We often need to invoke the following equation to nd it out: ~ = U ~ +W ~ V How do you verbally describe the dierent velocity vectors in the above equation? ~ is the uid velocity measured with respect to an inertial frame. V ~ is the blade velocity measured with respect to an inertial frame. U ~ is the uid velocity measured with respect to the rotating blade (non-inertial W frame). 130
8. We encounter two common types of turbomachine: centrifugal
ow type axial
ow type Prociency in analyzing each of them is crucial in designing machines with better performance. We will examine both of them in this problem. (a) Centrifugal Flow Type:
Backward Curved Blade
U2 = r2
V2
ii. Which velocity component is responsible for carrying uid in and out of the machine? The radial velocity component (Vr ; Wr ) is responsible for carrying uid in and out of a centrifugal machine. iii. Determine the following velocity components:
Ur = 0 Ut = r ! Q Wr = 2 rb 131
(mass conservation)
Wt = = Vr = = Vt = =
iv. Determine the torque and power input. By the Euler turbomachine equation,
132
Stator Blade 1
Rotor Blade
U1 = r W1 1 1
ii. Which velocity component is responsible for carrying
uid in and out of the machine? The axial velocity component (Wx; Vx) is responsible for carrying
uid in and out of an axial machine. iii. Determine the following velocity components: Ux = 0 Ut = r ! Wx = ( r 2 Q r 2 ) (mass conservation)
o i
133
Wt = = Vx = = Vt = =
where ri and ro denote the inner and outer radii of the rotor cross section respectively. Remarks: At the design point operation, the stator blade outlet angle 1 is set so that x;1 tan 1 = V Vt;1 to achieve ow attachment. iv. Determine the torque, power input and theoretical head. The results from the Euler turbomachine equation apply to both centrifugal and axial machines: T = Q (r2 Vt;2 r1 Vt;1 ) P = Q (U2 Vt;2 U1 Vt;1) 1 (U V H = g U1 Vt;1) 2 t;2 where subscripts 1 ;2 refer to the inlet and outlet stations respectively. v. What is/are your assumption(s) in determining the results in Part(iv)? Usually, we assume a mean radius value rm to evaluate the U and Vt terms in the above equations: ro rm = ri + 2 U = rm ! Vt = rm ! + (r2 Q r2) cot i o which is an approximation itself. By doing so, we can avoid integrating the angular momentum distribution along the radial direction at both the inlet and the outlet. But this assumption requires the dierence between the outer and inner radius, ro ri, to be small compared with the mean radius, (ro + ri)=2. 134
References
Fox, R. W. & McDonald, A. T., Introduction to Fluid Mechanics, Wiley, New York, 1973. Shapiro, A. H., The Dynamics and Thermodynamics of Compressible Fluid Flow, Volume I, Wiley, New York, 1953. White, F. W., Fluid Mechanics, McGraw-Hill, New York, 1986. Zucker, R. D., Fundamentals of Gas Dynamics, Matrix, Ohio, 1977.
135