PROPULSION

ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

Contents

Part 1. Notes and Solutions for Rocket Propulsion Elements by George Sutton and Oscar Biblarz 2
1. Definitions and Fundamentals 2
1.1. Definitions 2
2. Nozzle Theory and Thermodynamic Relations 2
2.1. Isentropic Flow Through Nozzles 2
3. Flight Performance 2
3.1. Gravity-Free, Drag-Free Space Flight 2
3.2. Forces Acting on a Vehicle in the Atmosphere 3
3.3. Basic Relations of Motion 3
Problems 3

Part 2. AE121 4
4. Isentropic Flow Eqns. with Area Change 5
5. PSs 6
5.1. PS2 6
5.2. Lagrangian point of view for gravity-free, drag-free rocket 7
5.3. PS4 9
5.4. PS5 10

Part 3. Basic Feeling 12

6. Box with a hole rocket; bottled (box) rocket 12
Kinetic Theory of the Ideal Gas Law 12

Part 4. Notes and Solutions for Thermal Physics by Ralph Baierlein 14

7. Background 14
7.1. Heating and Temperature 14
7.2. Some dilute gas relationships 14
7.3. The First Law of Thermodynamics 14
7.4. Heat capacity 14
Problems 15
8. Phase Equilibrium 15
8.1. Latent heat 15
8.2. Conditions for coexistence 16

Part 5. Notes and Solutions on Fundamentals of Thermodynamics, 8th Edition, by Claus Borgnakke, Richard E. Sonntag 16
9. 16
Date: 13 novembre 2015.
Key words and phrases. Propulsion, Rocket Propulsion, Thermodynamics, Fluid Flow, Fluid Mechanics.

1
2 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

10. Properties of a Pure Substance 16

Part 6. Thermodynamics (Revisited) 17

11. Heat Capacity 17
12. Phase Equilibrium 17
References 18
13. Code listings 19
14. Automatic generation of documentation 20

2. Nozzle Theory and Thermodynamic Relations

Abstract. Everything about Propulsion, with a focus on rocket propulsion.
I also look at (rocket) propulsion for engineers from a (theoretical and mathematical) physicists’ point of view. I would like to 2.1. Isentropic Flow Through Nozzles. Subsection 3.3 “Isentropic Flow Through Nozzles” of Biblarz and Sutton (2001) [1]
seek more cross-polination between physicists and mathematicians and engineers in thermodynamics and fluid mechanics. The Bernoulli invariant for compressible flow 1 gives us this:
1 1
q
h1 + v12 = h2 + v22 or v2 = 2(h1 − h2 ) + v12
2 2
    γ−1 !   γ−1
Cp γ τ1 τ2 γ τ1 p2 γ γRT1 p2 γ
=⇒ h1 − h2 = (τ1 − τ2 ) = 1− = 1− = (1 − )
MN γ−1M τ1 γ−1M p1 γ−1 p1
so that v
u
u 2γRT1   γ−1 !
p2 γ
v2 = t 1− + v12
γ−1 p1
When chamber section is large compared to nozzle throat section, chamber velocity or nozzle approach velocity comparatively
small and v12 can be neglected.[1]
Chamber temperature T1 at nozzle inlet; under isentropic conditions, T1 differs little from stagnation temperature or (for
chemical rocket) combustion temperature.[1]
Ch. 4 Flight Performance of Biblarz and Sutton (2001) [1] should probably be read before Ch. 3 Nozzle Theory and Ther-
modynamic Relations or before. In AE121 Fall 2015, the material for Ch. 4 was covered in lectures (I think, from the problem
Part 1. Notes and Solutions for Rocket Propulsion Elements by George Sutton and Oscar Biblarz sets) before Thermodynamics and Nozzle Theory.
[1]
1. Definitions and Fundamentals
Ch. 2 Definitions and Fundamentals of Biblarz and Sutton (2001) [1] 3. Flight Performance
cf. Chapter 4 Flight Performance of Biblarz and Sutton (2001) [1]
1.1. Definitions. Section 2.1. “Definitions”, Ch. 2 Definitions and Fundamentals of Biblarz and Sutton (2001) [1]
Rt
total inpulse Is = 0 Fthrust dt 3.1. Gravity-Free, Drag-Free Space Flight. Let
t ≡ burning time ≡ tp mp ≡ (total) propellant mass (initially)
tp ≡ propellant burning duration time
specific impulse Is , total impulse per unit weight of propellant
m(0) = 0
Rt
0
Fthrust dt It Let mass of rocket + propellant M = M (t) = M (0) − m(t) s.t. M (0) = M0 + mp
Is = R =
g0 ṁdt mp g0 m(tp ) = mp
M (tp ) = M0
ṁue tp ue If ṁ =
mp
(assume constant propellant flow rate),
Is = = tp
ṁgtp g      
mp mp t M (0) − mp t
If Fthrust = ṁue , and constant propellant mass flow, M = M (0) − t = M (0) 1 − = M (0) 1 − 1 −
tp M (0) tp M (0) tp
1“Euler equations (fluid dynamics)”, Wikipedia
 PROPULSION 3

cf. Eq. 2-7 of Biblarz and Sutton (2001) [1].

m
Define some quantities: mass ratio = mf0 where F CD
cos (ψ − θ) − ρAu2 − g sin θ
u̇ =
mf := final mass (after rocket operation had consumed all usable propellant), which is M0 above =⇒ m 2m
m0 := mass before rocket operation, which is M (0) F CL
mp
uθ̇ = sin (ψ − θ) + ρAu2 − g cos θ
propellant mass fraction M (0) cf. Eq. 2-8 of Biblarz and Sutton (2001) [1]. m 2m
For thrust Fthrust , CD , CL are functions of velocity or Mach number (!!!)
mp
For actual trajectory analyses, perturbation effects in Sec. 4.6 must be considered
tp dt 3-body theory considered
Z
du ṁdt du ∆u mp
Fthrust = ṁue = M =⇒ = or = m = − ln (M (0) − t) = when propellant flow and thrust not constant
dt M ue ue M (0) − tpp t tp
!   from optical or radar tracking data, thrust or actual specific impulse during actual vehicle flights determined from accurately
mp M (0) M (0) observed trajectory data.
= − ln (M (0) − tp ) + ln (M (0)) = ln m = ln
tp M (0) − tpp tp M (0) − mp make assumption or measurement on propellant flow (which usually varies in a predetermined manner)
If L = 0 (wingless rocket projectile), ψ − θ = 0 (flight direction θ same as thrust direction), and
Thus     M (t) = M (0) − ṁ; assume constant ṁ = ttp mp
M (0) M (0) tmp mp t mp
∆u = ue ln = ue ln or M (t) = M (0) − = M (0)(1 − with ξ ≡
M (0) − mp M0 tp M (0) tp ), M (0) propellant mass ratio
   
∆u M (0) M (0) M0 −∆u F CD ue ṁ CD ρAu2
exp = = or = exp u̇ = − ρAu2 − g sin θ = − − g sin θ =
ue M (0) − mp M0 M (0) ue M 2M M 2M
Also remember that F = ṁue = ṁIsp g0 , so ue ≡ effective exhaust velocity can be related directly to the specific impulse Isp . ue ξ ttp CD ρAu2
mp
Also note that for propellant mass fraction M (0)
mp
< 1, M (0) = M (0)−M 0
= 1 − MM(0)
0 = − − g sin θ
M (0) 1 − ξ ttp 2M (0)(1 − ξ ttp )
and
3.2. Forces Acting on a Vehicle in the Atmosphere. assume starting and stopping transients very short and neglected,
F = ue ṁ uθ̇ = −g cos θ
m m Example 4-1. Consider “a simple-stage rocket for a rescue flare has the following characteristics and its flight path nomen-
if mass rate of propellant consumption ṁ constant, ṁ = tpp so F = tpp ue
drag D opposite to flight path due to resistance of body to motion in fluid clature is shown in the sketch.”
lift L normal to flight path Neglect drag “since flight velocities are low,” assume no wind, assume local acceleration of gravity to be equal to sea level
1 g0 , invariant throughout flight.
L = CL ρAu2
2 Problems. Problem 1. Recall that
1
 
M0 −∆u
D = CD ρAu2 = exp
2 M (0) ue
cf. Eqns. (4-10), (4-11) of Biblarz and Sutton (2001) [1] and so (in Python)
density of earth’s atmosphere can vary by factor up to 2 (for altitudes of 300 to 1200 km) import sympy
depending on solar activity and night-to-day temperature variations. major unknown in drag from sympy import ∗
Assume neglect variation of gravity with geographical features and oblate shape of earth, >>> exp ( − 1 6 0 0 / 2 0 0 0 . )
0.449328964117222
GMe m
F = 2 Problem 2.
r2 R02

R0
GMe
=⇒ g = g0 2 = g0 mp M (0) − M0 M0 1
mg0 = m R R0 + h = =1− = 1 − = 4/5 = 0.8
R02 M (0) M (0) M (0) 5
Problem 3. dragless projectile, so D = 0.
3.3. Basic Relations of Motion. cf. Section 4.3 “Basic Relations of Motion,” Chapter 4 Flight Performance of Biblarz and
Sutton (2001) [1] ue m0 t
u̇ = −g0 + =⇒ ∆u = −g0 t − ue ln (1 − )
tp (M (0)/mp − t/tp ) M (0)tp
mu̇ = F − D − mg Plugging in ue = 2209 m/sec, mp /M (0) = 0.57, tp = 5.0 sec, u0 = h0 = 0,
Propulsion.py
mu̇ = F cos (ψ − θ) − D − mg sin θ
>>> i n t e g r a t e ( f l i g h t p a t h d i r e c t i o n . s u b s ( Drag , 0 ) . s u b s ( p s i , t h e t a ) . s u b s ( t h e t a , p i / 2 ) . s u b s ( F t h r u s t , m p/ t p ∗ u e ) .
=⇒ dθ s u b s (M, M c o n s t a n t f l o w ) . s u b s ( m p , M0∗ 0 . 5 7 ) . s u b s ( t p , 5 . ) . s u b s ( u e , 2 2 0 9 . ) . s u b s ( g 0 , 9 . 8 ) . f a c t o r (M0 ) . r h s , ( t , 0 , 5 . 0 ) )
mu = F sin (ψ − θ) + L − mg cos θ 1815.32988528061
dt
ψ = direction of thrust angle from horizontal reference Problem0403 = f l i g h t p a t h d i r e c t i o n . s u b s ( Drag , 0 ) . s u b s ( p s i , t h e t a ) . s u b s ( t h e t a , p i / 2 ) . s u b s ( F t h r u s t , m p/ t p ∗ u e ) .
Consider perturbation effects (cf. Sec. 4.6, listed), drag and gravity s u b s (M, M c o n s t a n t f l o w ) . s u b s ( m p , M0∗ 0 . 5 7 ) . s u b s ( t p , 5 . ) . s u b s ( u e , 2 2 0 9 . ) . s u b s ( g 0 , 9 . 8 ) . f a c t o r (M0 ) . r h s
4 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

up = 1815 m/sec
Eperm0406 = Gconst ∗ M earth ∗ ( − 1 / ( 2 ∗ ( ( R e a r t h+Decimal ( 5 0 0 ) ) ∗ 1 0 ∗ ∗ 3 ) ) + 1 / ( R e a r t h ∗ 1 0 ∗ ∗ 3 ) )
>>> i n t e g r a t e ( i n t e g r a t e ( Problem0403 , ( t , 0 , t ) ) , ( t , 0 , 5 . 0 ) ) # Energy p e r mass
3890.37850288891 ’ %.6E ’ % Eperm0406 # 3 3 . 5 1 MJ/ kg

hp = 3.89 × 103
Problem 4. How to estimate A of the projectile? v = 7611 m/sec T = 5678 s or 1.58 hours 33.51 M J/kg
Problem 5.
Now
M (t)a = Fthrust
Fthrust tp Fthrust tp Part 2. AE121
Isp = mp =
g0 tp tp g0 mp Let’s translate γ
between physicists and engineers:
γ−1
mp
and for M (t) = M (0) − tp t,
Cp

g m
 
g0 m p
   γ CV Cp cp M N cp kB
Isp 0tp p Isp tp Isp g0  1 = N
= = =
γ−1 N N R
a= m ≤ a(t = tp ) =⇒ a(t = tp ) = M0
=
tp M (0)
 CV
M (0) − tpp t mp −1
2
Plugging a = 50 m/sec and solving for tp , 3.3.1. Speed of sound. From pp. 179, Problem 10 “Isnetropic relations of ideal gas” of Chapter 6: Ideal Gas of Kittel and
>>> 2 6 0 . ∗ 9 . 8 / 5 0 . ∗ ( 1 / ( 1 / 0 . 8 8 − 1 ) ) Kroemer [3], recall isentropic bulk moduli Bσ
373.70666666666637

pi Viγ
 
(a) ∂p
Bσ := −v =γ = γp
373.71 sec ∂V σ Vγ
maximum allowable burn time, assuming steady propellant mass flow p Vγ
(b) with p = Vi γi .
Isp g0 ξ/tp Very little heat transfer in sound wave. For velocity (magnitude) i.e. speed of sound, a
a= =⇒ ∆u = −Isp g0 ln (1 − ξt/tp )
1 − ξt/tp  1/2  1/2
so Bσ γp
a= =
∆u = 5402.4 m/sec ρ ρ
for maximum velocity relative to the launch vehicle
Problem 6. Satellite in circular orbit Now
Nτ
r
GMe m mv 2 GM0 p=
F = 2
= =⇒ =v V
(R0 + h) (R0 + h) R0 + h
p Nτ 1 τ
2π(R0 + h) 2π(R0 + h)3/2 =
=
= v =⇒ T = √ ρ V MN
V
M
T GMe
      
1 −GMe m −GMe m 1 GMe GMe GMe −1 1 Now p = ρRT (outside of theoretical physics, people use the so-called universal gas constant R).
mv 2 + − =m − + = m GMe +
2 R0 + h R0 2 R0 + h R0 + h R0 2(R0 + h) R0
Then run Propulsion.py which now imports (import) in Physique, a small package with the NIST (National Institute of Nτ MN τ
p = ρRT = = R
Standards and Technology) Fundamental Constants FundConst, NIST SI conversions conv, and NASA Planetary Fact Sheet V V kB
plnfacts as Python pandas DataFrames.
and so
M earth = p l n f a c t s . l o c [ p l n f a c t s [ ’ P l a n e t ’ ]==”EARTH” , ” Mass ( 1 0 2 4 kg ) ” ] . v a l u e s [ 0 ] ∗ 1 0 ∗ ∗ ( 2 4 ) # i n kg
R e a r t h = p l n f a c t s . l o c [ p l n f a c t s [ ’ P l a n e t ’ ]==”EARTH” , ” Diameter (km) ” ] . v a l u e s [ 0 ] / Decimal ( 2 )
kB
(1) R :=
Gconst = FundConst [ FundConst [ ” Q u a n t i t y ” ] . s t r . c o n t a i n s ( ” g r a v i t a t i o n ” ) ] . l o c [ 2 4 3 , ” Value ” ] M
v0406 = s q r t ( Gconst ∗ M earth / ( ( R e a r t h + Decimal ( 5 0 0 ) ) ∗ 1 0 ∗ ∗ 3 ) )
# v e l o c i t y o f s a t e l l i t e v o f Chapter 4 , Problem 6 o f B i b l a r z and S u t t o n
so, as Polk says, R is different for different gases.
7611.17633707692
And so
T0406 = ( 2 . ∗N( p i ) ∗ f l o a t ( ( R e a r t h + Decimal ( 5 0 0 ) ) ∗ 1 0 ∗ ∗ 3 ) ∗ ∗ ( 3 . / 2 ) ) / f l o a t ( s q r t ( Gconst ∗ M earth ) ) γp γτ
# 5678 s e c s . o r 1 . 5 8 h o u r s =
ρ M
 PROPULSION 5

4. Isentropic Flow Eqns. with Area Change Use continuity:

γ
ρ 1 v 1 A1 = ρ 2 A2 v 2
  γ−1
p τ
= Thus
pi τi 1/2
T1 ρ21

τ
 γ−1  γ−1
Vi ρ A2 ρ 1 v1 M1
= = = =
τi V ρi A1 ρ 2 v2 M2 T2 ρ22
 γ  γ
p Vi ρ EY : 20151120 One can also relate a point in the flow, 1, to another point “downstream” to the flow, 2:
= =
pi V ρi
cf. “Isentropic relations of ideal gas” of Kittel and Kroemer [3] T1 T1 /T0
=
cf. 20151105 AE121 Polk T2 T2 /T0

v2 Substitute for T1 /T2 and p1 /pe , e or exh for exhaust,

h0 = h +
2  γ+1 1/2
v2
! γ−1
γ−1 2
Cp T0 = Cp T + A2 M1  1 + 2 M2
2 (3) = γ−1 2

A1 M2 1+ 2 M1
T0 v2
=1+
T 2Cp T To get thrust, we need an expression for mass flow rate
Define M = av , Mach #
sound speed a = (γRT )1/2 ṁ = ρvA
ṁ (Recall) a = (γRT )1/2
= ρv
T0 γ−1 2 A
(2) =1+ M  1/2 " #1/2
T 2 1/2 1/2 T γRT0
v = Ma = M(γRT ) = M(γRT0 ) =M
 γ
 γ−1 T0 1 + γ−1
2 M
2
p0 γ−1 2
= 1+ M
p 2 Now
 1
 γ−1 1
ρ0 γ−1 2   " # γ−1
= 1+ M ρ 1
ρ 2 ρ = ρ0 = ρ0 γ−1
ρ0 1+ 2 M
2
(EY : 20151118 says
H ṁ ρ0 M(γRT0 )1/2
h := ρv = =
1
MN A 1+ γ−1 2

1/2
1 + γ−1 2 γ−1
u2 2 M 2 M
Cp h0 = h +
cp := 2 Using the ideal gas law
MN =⇒
MNu2 u2
H0 = H + cp τ0 = cp τ + ṁ p0 γ 1/2 1
2 2 = M γ+1
A (RT0 )1/2 1+ γ−1 2

2(γ−1)
H = Cp τ 2 M

and now At the throat, M = 1,

u u
M= = Cp = CV + N
a (γτ /M )1/2 γ+1
p0 γ 1/2
  2(γ−1)
ṁ 2
 γτ 1/2  1/2 N (4) =
γkB T =⇒ γ − 1 = A∗ (RT0 )1/2 γ+1
a= = = (γRT )1/2 CV
M M
and so Thus, from Eq. 4 above (giving the mass flow from throat area and stagnation p0 , T0 ), and plugging this into thrust (force)
τ0 u2 1 γτ γM2 M2 γ−1 2 equation,
=1+ =1+ M2 =1+ =1+ =1+ M
τ 2cp τ 2cp τ M 2Cp /N 2CV /N 2
T = ṁve + (pe − pa )Ae =
) v
γ+1   γ−1 !
Now we need to get M in terms of area so we can apply these to a nozzle. (5)
u
A∗ p0 γ 1/2
  2(γ−1)
2 u 2γRT0 pe
γ
= t 1− + (pe − pa )Ae
(RT0 )1/2 γ+1 γ−1 p0
6 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

From physics, equating kinematics, dynamics M u̇ to the external force, Fthrust ,

mp1
M u̇ = Fthrust = ṁ1 (Isp1 g0 ) = Isp1 g0 so
tp1
mp1
tp1 Isp1 g0
 
M2 (t) + Mpayl + M01 − mp1
u̇ = mp1 =⇒ ∆u1 = −Isp1 g0 ln
M2 (t) + Mpayl + M01 − tp1 t M2 (t) + Mpayl + M01
Also,  
Mpayl + M02 − mp2
∆u2 = −Isp2 g0 ln
Mpayl + M02
(a) g s t d = FundConst [ FundConst [ ” Q u a n t i t y ” ] . s t r . c o n t a i n s ( ” g r a v i t y ” ) ] . l o c [ 3 0 3 , : ] . Value
# get standard a c c e l e r a t i o n of gravity
M 0 = Symbol ( ’ M 0 ’ , p o s i t i v e=True )
D e l t a u = −I s p ∗ g 0 ∗ l n ( ( M 0 −m p ) / M 0 )
# part ( a )
Deltau . subs ( I sp , 2 6 8 . 8 ) . subs ( g 0 , gstd ) . subs (M 0 , 8 0 5 3 0 9 . ) . subs (m p , (1 −0.1396)∗586344)
# 2 5 9 5 . 7 4 5 2 1 0 3 4 1 0 1 m/ s
5. PSs
∆u1 = 2595.7 m/s
5.1. PS2. Problem 1: Ares I launch vehicle. Consider a 2-stage rocket. It’s also interesting to explore the properties of
polybutadiene acrylonitrile (PBAN) (first stage solid rocket motor fuel), liquid oxygen/liquid hydrogen (LOX/LH2) (second (b) Deltau . subs ( I sp , 4 5 2 . 1 ) . subs ( g 0 , gstd ) . subs (M 0 ,183952+35013.). subs (m p , (1 −0.1110)∗183952)
stage liquid rocket engine fuel). # 6 0 9 0 . 6 8 7 1 6 7 3 0 3 1 8 m/ s
, according to wikipedia 2, is the “ratio between the empty mass of the stage, and the combined empty mass and propellant
mp ∆u2 = 6090.7 m/s and so
mass”, which is, in the notation of Biblarz and Sutton (2001) [1], ζ ≡ M (0) .
Neglect drag and earth’s gravity. u2f = 8686.4 m/s
Assume constant mass flow. (c) Now
Let Fthrust ṁ1 ue1
= = 1.5
W g0 M (0)
M (t) = M2 (t) + Mpayl + M01 − m1 (t)
is the thrust to weight ratio at the instant of takeoff.
with M = M (t) being the total mass of the entire system during the first stage, with m1 = m1 (t) ∈ R s.t. Now ue1 = Isp1 g0 and so
m1 (0) = 0 1.5M (0)
ṁ1 =
m1 (tp1 ) = mp1 (total mass of propellant of stage 1) Isp1
tp1 ≡ burn time of first stage and so
M01 ≡ total mass of empty stage 1 + propellant mass for stage 1 >>> 1 . 5 ∗ 8 0 5 3 0 9 . / 2 6 8 . 8
M01 − mp1 ≡ mass of empty stage 1. # 4493.911830357143
Also,
so
M01 − mp1 mp1
1 = =1− ṁ1 = 4493.9 kg/s
M01 M01
Over 4 tons of propellant reactants is dumped out per second!
and Problem 2. Continuous staging
R
Fthrust dt Fthrust tp ue (a) Mpayl ≡ payload mass; Mpayl ∈ R+ .
Isp = = =
W g0 ṁtp g0 Assume structure mass discarded at 0 velocity.
structure mass continuously jettisoned during the burn.
Now for the second stage, M = M (t) ∈ C ∞ (R) represents mass of structure undiscarded at time t, i.e. system of structure + propellant, at
“control volume” at time t. So consider
(
M02 if 0 ≤ t ≤ tp1
M2 (t) = M = Mpayl + ms (t) + mp (t) s.t.
M02 − m2 (t)
ms = ms (t) ∈ C ∞ (R), mass of structure not yet thrown out, and propellant mass mp = mp (t) ∈ C ∞ (R)
If we assume constant propellant burn (out), constant propellant flow rate, then
−mp
for m2 (tp1 ) = 0 ṁp (t) =
tp
m2 (tp2 + tp1 ) = mp2
2https://en.wikipedia.org/wiki/Multistage_rocket
 PROPULSION 7
+
with mp ∈ R total mass of propellant import sympy
tp ∈ R+ burn time of propellant fuel. from sympy import ∗
>>> M payl = Symbol ( ‘ ‘ M payl ’ ’ , p o s i t i v e=True )
>>> M s = Symbol ( ‘ ‘ M s ’ ’ , p o s i t i v e=True )
m
mp (t) = mp − tpp t = mp (1 − ttp ) >>> r a t i o b i = ( M payl + M s ) / ( M payl + M s + M p )
>>> p a y l o a d f r a c = Symbol ( ‘ ‘ p a y l o a d f r a c ’ ’ , p o s i t i v e=True )
Assume constant dead mass ratio δ = ms (t)/mp (t) (EY: my intuition is we’re throwing out as much propellant out >>> d e l t a = Symbol ( ‘ ‘ d e l t a ’ ’ , p o s i t i v e=True )
in fixed proportion to structure being discarded out; the name “dead” refers to what’s still left that’s being propelled >>> r a t i o b i n e w = r a t i o b i . s u b s ( M payl , ( M s + M p ) / ( 1 / p a y l o a d f r a c − 1 ) ) . s u b s ( M s , M p∗ d e l t a )
forward by thrust, I think (???)) >>> r a t i o b i n e w . expand ( ) . f a c t o r ( M p ) . s i m p l i f y ( ) . f a c t o r ( )
( d e l t a + payloadfrac )/( d e l t a + 1)
ms (t) = δmp (t) = δ(mp )(1 − t/tp )
Assume Isp constant R and so
Fthrust dt (−ṁp )ue tp ue  
Isp ≡ = = 1+δ
Wpropellant mp g0 g0 (∆u)i = Isp g0 ln
P λ+δ
Consider the instantaneous rest frame of spacecraft + propellant fuel system α ∆piα = 0
P
Consider before and after, after an instant. So for α pf α , (ii)
   
Now M = M (t) = Mpayl + ms (t) + mp (t) Mpayl + Mp 1
(∆u)ii = Isp g0 ln = Isp g0 ln
M (t + δt) = M (t) + Ṁ δt Mpayl λ
Then the momentum of the part that’s going to be propelled by the thrust at time t + δt is
(iii) For 1 (total number of) stage(s),
M (t + δt)u(t + δt) = (M (t) + Ṁ δt)(u(t) + u̇δt) = (M (t) + Ṁ δt)(0 + u̇δt) = M u̇δt + O((δt)2 ) 
1+δ

∆u = Isp g0 ln
The momentum of the propellant expelled out + structure that’s discarded is λ+δ
(−ṁp dt)(−ue ) + (−ṁs dt) · 0 = −ṁp dt(−ue ) as in part (b)(i).
m
p m u
p e
Let N be the number of stages.
ṁp ue tp ue tp
=⇒ M u̇δt + ṁp ue δt = 0 =⇒ u̇ = − = = t
Mpayl + ms (t) + mp (t) Mpayl + (δ + 1)mp (t) Mpayl + (1 + δ)mp (1 − tp )
5.2. Lagrangian point of view for gravity-free, drag-free rocket. Let M = M (t) = M (0) − m(t) s.t. m(0) = 0
" #
−ue

t

Isp g0 Mpayl + (1 + δ)mp where
=⇒ ∆u = ln (Mpayl + (1 + δ)mp (1 − )) − ln (Mpayl + (1 + δ)mp ) = ln t m(tp ) = Mp
1+δ tp 1+δ Mpayl + (1 + δ)mp (1 −tp )
h i Mp is the total mass of the propellant to be used, and M (0) represents the (initial) mass of the propellant + spacecraft or
Isp g0 Mpayl +(1+δ)mp
For t = tp , ∆u = 1+δ ln Mpayl payload or the interesting part we want to launch out. tp is the burn time of the propellant.
Rt
3
(b) Payload fraction is defined as the weight of payload over the takeoff weight (cf. wikipedia . Then for part (a), the ∆u, Then kinetic energy of M at instantaneous time t is 12 M u2 . Also note that 0 p ṁdt = Mp .
∆ua , is Now take the exterior derivative of m: dm = ṁdt. Consider the infinitesimal piece of mass dm at the (instantaneous) time
t; its kinetic energy will be
 
Isp g0 1 −Isp g0
(∆u)a = ln = ln λ
1+δ λ 1+δ 1
dm(u − ue )2
2
(i) M = M (t) = Mpayl + Ms + mp (t) s.t. mp (0) = Mp if assume constant mass flow out. mp (t) = Mp (1 − ttp ) Note that its velocity is (u − ue ) because mass dm is being expunged out of the rocket at constant exhaust velocity ue relative
to the rocket (i.e. change, transform, or “boost” into the instantaneous inertial reference frame of the rocket, so that the rocket
mp (tp ) = 0
has 0 velocity in this frame; now return back to the “lab” frame-the propellant expunged has gained velocity u as well).
Mp
tp ue Now
M u̇ = −ṁp uexh =⇒ u̇ = Z t
1 t
Rt Z
t 1
Mpayl + Ms + Mp (1 − tp ) 2 0 dt 1 2
dm(u − ue ) −−−→ dm(u − ue ) = ṁdτ (u − ue )2
So 2 2 0 2 0
t The full Lagrangian at time t is
u(t) = −Isp g0 [ln (Mpayl + Ms + Mp (1 − )) − ln (Mpayl + Ms + Mp )]
tp
1 t
Z
1 2

Mpayl + Ms + Mp
 L = M u + ṁdτ (u − ue )2
∆u = Isp g0 ln 2 2 0
Mpayl + Ms
∂L
Notice that L in this specific case does not depend on position. So ∂x i = 0.
Then it’s just algebra to put the ratio of the masses above in terms of λ, δ (there’s 3 unknowns, Mpayl , Ms , Mp
M +Ms +Mp Now
masses, and we’re given λ, δ and a ratio we want, payl ). Instead of doing the algebra entirely by hand,
Mpayl +Ms ∂ 12 M u2 d
dt d ∂ 21 M u2
let’s use Python and the sympy library: = M ui −→ = Ṁ ui + M u̇i = −ṁui + M u̇i
∂ui dt ∂ui
3 “Payload fraction”, Wikipedia https://en.wikipedia.org/wiki/Payload_fraction
8 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

and # t o u s e s c i p y . i n t e g r a t e . ode , t h e ODE syst em must be d e c l a r e d a s a ” c a l l a b l e ” , a Python f u n c t i o n

∂ 1 t 1 t
Z Z Z t
def d e r i v ( t , u ) : # r e t u r n d e r i v a t i v e s o f t h e a r r a y u
2
ṁdt(u − ue ) = ṁdτ 2(ui − ue ) = ṁdτ (ui − ue ) ”””
∂ui 2 0 2 0 0 c f . h t t p : / / b u l l d o g 2 . r e d l a n d s . edu / f a c u l t y f o l d e r / deweerd / t u t o r i a l s / T u t o r i a l −ODEs . p d f
d ∂ 1 t d t
d
Z Z
2
−dt→ ṁdt(u − ue ) = ṁdτ (ui − ue ) = ṁ(t)(ui − ue ) ”””
dt ∂ui 2 0 dt 0 uxdot = ( 7 . 8 5 3 ∗ 1 0 ∗ ∗ ( − 5 ) ) ∗ exp ( −u [ 3 ] / ( 1 0 0 0 0 . ) ) ∗ ( u [ 0 ] ∗ ∗ 2 + u [ 1 ] ∗ ∗ 2 ) ∗ ∗ ( 0 . 5 ) ∗ ( − u [ 0 ] )
Rt uydot = −9.8 + ( 7 . 8 5 3 ∗ 1 0 ∗ ∗ ( − 5 ) ) ∗ exp(−u [ 3 ] / ( 1 0 0 0 0 . ) ) ∗ ( u [ 0 ] ∗ ∗ 2 + u [ 1 ] ∗ ∗ 2 ) ∗ ∗ ( 0 . 5 ) ∗ ( − u [ 1 ] )
for clearly, if 0 ṁdτ (ui − ue ) = F (t) − F (0), then applying the total time derivative, will result in F 0 (t) and so we can read off return a r r a y ( [ uxdot , uydot , u [ 0 ] , u [ 1 ] ] )
the antiderivative, namely ṁ(t)(ui − ue ) at time t.
∂L d ∂L
i − dt ∂ ẋi = 0 and so
The Euler-Lagrange equation tells us that ∂x # i n i t i a l conditions
u0 = [ 3 0 0 . ∗ c o s ( 5 0 . / 1 8 0 . ∗N( p i ) ) , 3 0 0 . ∗ s i n ( 5 0 . / 1 8 0 . ∗N( p i ) ) , 0 , 0 ]
−ṁui + M u̇i + ṁui − ṁue = 0 or M u̇i = ṁue
# d e c l a r e t h e ODE t o be i n t e g r a t e d
ṁue Prob0203 = ode ( d e r i v ) . s e t i n t e g r a t o r ( ’ d o p r i 5 ’ ) # Problem 3 from Problem S e t 2 f o r AE121 F a l l 2015
=⇒ u̇i = # c f . h t t p : / / s t a c k o v e r f l o w . com/ q u e s t i o n s / 2 6 7 3 8 6 7 6 / does−s c i p y −i n t e g r a t e −ode−s e t −s o l o u t −work
M Prob0203 . s e t i n i t i a l v a l u e ( u0 )
Problem 4. Ballistic trajectories with atmospheric drag
t1 = 41.575
(a) Now dt = 0 . 0 0 5
M0 u̇ = −M0 g + Fd # p r i n t o u t t h e s o l u t i o n t o t h e ODE f o r v a r i o u s t i m e s a s a s a n i t y c h e c k
while Prob0203 . s u c c e s s f u l ( ) and Prob0203 . t < t 1 :
Fd Prob0203 . i n t e g r a t e ( Prob0203 . t+dt )
u̇ = −g + print ( ” %g ” % Prob0203 . t )
M0
print Prob0203 . y
In components, 
−ux
Fd # s t o r e t h e s o l u t i o n s i n a Python l i s t f o r p l o t t i n g
u̇x = Prob0203 . s e t i n i t i a l v a l u e ( u0 )
|u|
M0
  Prob0203 solution = [ ]
Fd −uy while Prob0203 . s u c c e s s f u l ( ) and Prob0203 . t < t 1 :
u̇y = −g + P r o b 0 2 0 3 s o l u t i o n . append ( [ Prob0203 . t+dt , ] + l i s t ( Prob0203 . i n t e g r a t e ( Prob0203 . t+dt ) ) )
M0 |u| # take the transpose of a l i s t of l i s t s
And so, for P r o b 0 2 0 3 s o l u t i o n = map( l i s t , zip ( ∗ P r o b 0 2 0 3 s o l u t i o n ) )
1 1
Fd = ρCD u2 A = ρCD (u2x + u2y )A plt . figure (1)
2 2 plt . plot ( Prob0203 solution [ 3 ] , Prob0203 solution [ 4 ] )
and plt . x l a b e l ( ’ x (m) ’ )
ρ = ρ0 e−y/λ plt . y l a b e l ( ’ y (m) ’ )
  plt . t i t l e ( ’ C a n n o n b a l l t r a j e c t o r y w i t h Drag : V a r i a b l e d e n s i t y ’ )
1 −y/λ 2 2 −ẋ
ẍ = ρ0 e CD (ẋ + ẏ )A
2M0 (ẋ2 + ẏ 2 )1/2 Horizontal range is 6.11 km
 
1 −y/λ 2 2 −ẏ (b) def d e r i v b ( t , u ) : # r e t u r n d e r i v a t i v e s o f t h e a r r a y u
ÿ = −g + ρ0 e CD (ẋ + ẏ )A
2M0 (ẋ2 + ẏ 2 )1/2 ”””
c f . h t t p : / / b u l l d o g 2 . r e d l a n d s . edu / f a c u l t y f o l d e r / deweerd / t u t o r i a l s / T u t o r i a l −ODEs . p d f
Note that we can rewrite this as the following system of equations:
1 ”””
u̇x = ρ0 CD Ae−y/λ (u2x + u2y )1/2 (−ux ) uxdot = ( 7 . 8 5 3 ∗ 1 0 ∗ ∗ ( − 5 ) ) ∗ ( u [ 0 ] ∗ ∗ 2 + u [ 1 ] ∗ ∗ 2 ) ∗ ∗ ( 0 . 5 ) ∗ ( − u [ 0 ] )
2M0 uydot = −9.8 + ( 7 . 8 5 3 ∗ 1 0 ∗ ∗ ( − 5 ) ) ∗ ( u [ 0 ] ∗ ∗ 2 + u [ 1 ] ∗ ∗ 2 ) ∗ ∗ ( 0 . 5 ) ∗ ( − u [ 1 ] )
1 return a r r a y ( [ uxdot , uydot , u [ 0 ] , u [ 1 ] ] )
u̇y = −g0 + ρ0 CD Ae−y/λ (u2x + u2y )1/2 (−uy )
2M0 Prob0203b = ode ( d e r i v b ) . s e t i n t e g r a t o r ( ’ d o p r i 5 ’ )
ẋ = ux Prob0203b . s e t i n i t i a l v a l u e ( u0 )
Prob0203b . i n t e g r a t e ( 4 1 . 2 3 )
ẏ = uy
t1b = 4 1 . 2 2 5
In Propulsion.py, (just run it with python -i Propulsion.py in its directory) Prob0203b . s e t i n i t i a l v a l u e ( u0 )
import s c i p y Prob0203b solution = [ ]
from s c i p y import exp , a r r a y while Prob0203b . s u c c e s s f u l ( ) and Prob0203b . t < t 1 b :
from s c i p y . i n t e g r a t e import ode P r o b 0 2 0 3 b s o l u t i o n . append ( [ Prob0203b . t+dt , ] + l i s t ( Prob0203b . i n t e g r a t e ( Prob0203b . t+dt ) ) )
P r o b 0 2 0 3 b s o l u t i o n = map( l i s t , zip ( ∗ P r o b 0 2 0 3 b s o l u t i o n ) )
import m a t p l o t l i b . p y p l o t a s p l t
plt . figure (2)
M c a n n o n b a l l = ( 7 . 8 ∗ ( 1 0 ∗ ∗ 2 ) ∗ ∗ 3 / ( 1 0 ∗ ∗ 3 ) ) ∗ 4 . / 3 . ∗N( p i ) ∗ ( 1 5 . / 2 . / 1 0 0 . ) ∗ ∗ 3 plt . plot ( Prob0203b solution [ 3 ] , Prob0203b solution [ 4 ] )
( 1 . 2 2 5 ) ∗ ( 0 . 1 ) / ( 2 . ∗ M c a n n o n b a l l ) ∗ (N( p i ) ∗ ( 1 5 . / 2 . / 1 0 0 . ) ∗ ∗ 2 ) # 7 . 8 5 2 5 6 4 1 0 2 5 6 4 1 1 e−5 plt . x l a b e l ( ’ x (m) ’ )
plt . y l a b e l ( ’ y (m) ’ )
 PROPULSION 9

p l t . t i t l e ( ’ C a n n o n b a l l t r a j e c t o r y w i t h Drag : C o n s t a n t d e n s i t y ’ ) import P h y s i q u e
from P h y s i q u e import FCconv , KCconv , FundConst , conv , plnfacts , T C, T K, T F
Horizontal range is 5.89 km. This makes sense because the cannonball finds it easier to fly “through the air” at higher
k B o l t z = FundConst [ FundConst [ ” Q u a n t i t y ” ] . s t r . c o n t a i n s ( ” Boltzmann ” ) ] . loc [49 ,:]
altitudes, higher up the atmosphere, because the “air is thinner” in the “upper atmosphere.” >>> k B o l t z . Value
(c) For no drag, this can be solved analytically: Decimal ( ’ 1 . 3 8 0 6 4 8 5 2E−23 ’ )
>>> k B o l t z . Unit
u̇x = 0 x(t) = u0 cos θt
ux = u0 cos θ ’ J Kˆ−1 ’
=⇒ =⇒ 1 gt
u̇y = −g uy = −gt + u0 sin θ y(t) = − gt2 + u0 sin θt = t(u0 sin θ − ) So for T = 300K and T = 1000K,
2 2
 
u 2 >>> k B o l t z . Value ∗300∗ Decimal ( 1 . 5 )
So the horizontal range is x(tf ) = u0 cos θ 2u0 gsin θ = g0 sin (2θ) = 9044.m for Decimal ( ’ 6 . 2 1 2 9 1 8 3 4 0 0 0E−21 ’ )
>>> k B o l t z . Value ∗1000∗ Decimal ( 1 . 5 )
>>> 3 0 0 . ∗ ∗ 2 / 9 . 8 ∗ s i n ( 2 . ∗ 5 0 . / 1 8 0 . ∗N( p i ) ) Decimal ( ’ 2 . 0 7 0 9 7 2 7 8 0 0 0 0E−20 ’ )
9044.15283378558
or
#p a r a b o l a t r a j e c t o r y d a t a >>> k B o l t z . Value ∗300∗ Decimal ( 1 . 5 ) / JovereV . M u l t i p l y b y
P r o b 0 2 0 3 c x = [ i ∗10 f o r i in range ( 9 0 5 ) ] Decimal ( ’ 0 . 0 3 8 7 7 7 9 7 7 3 3 9 5 8 2 3 3 0 7 9 1 1 6 7 2 6 8 0 4 ’ )
P r o b 0 2 0 3 c y = [ t a n ( 5 0 . / 1 8 0 . ∗N( p i ) ) ∗ x − ( 9 . 8 / 2 . ) ∗ x ∗ ∗ 2 / ( 3 0 0 . ∗ c o s ( 5 0 . / 1 8 0 . ∗N( p i ) ) ) ∗ ∗ 2 f o r x in P r o b 0 2 0 3 c x ] >>> k B o l t z . Value ∗1000∗ Decimal ( 1 . 5 ) / JovereV . M u l t i p l y b y
Decimal ( ’ 0 . 1 2 9 2 5 9 9 2 4 4 6 5 2 7 4 4 3 5 9 7 0 5 5 7 5 6 0 1 ’ )
# plot a l l 3 t r a j e c t o r i e s together
plt . figure (3)
p l t . p l o t ( P r o b 0 2 0 3 s o l u t i o n [ 3 ] , P r o b 0 2 0 3 s o l u t i o n [ 4 ] , l a b e l=” Drag : V a r i a b l e d e n s i t y ” ) and so U/N = 6.213 × 10−21 J or 0.0388eV for T = 300K and 2.071 × 10−20 J or 0.129eV for T = 1000K
p l t . p l o t ( P r o b 0 2 0 3 b s o l u t i o n [ 3 ] , P r o b 0 2 0 3 b s o l u t i o n [ 4 ] , l a b e l=” Drag : C o n s t a n t d e n s i t y ” ) (c) Now
p l t . p l o t ( P r o b 0 2 0 3 c x , P r o b 0 2 0 3 c y , l a b e l=”No Drag ” )
 
∂U 3N
p l t . x l a b e l ( ’ x (m) ’ ) CV = =
p l t . y l a b e l ( ’ y (m) ’ ) ∂τ V 2
p l t . t i t l e ( ’ T r a j e c t o r i e s o f c a n n o n b a l l w i t h Drag o f v a r i a b l e d e n s i t y , Drag o f c o n s t a n t d e n s i t y , and no d r a g ’ )
plt . legend ()
From the definition of CV . Also, for enthalpy H, H = U + pV = U + N τ , for ideal gas law still holds,
   
∂H ∂U
If there was no drag, then the cannonball will fly out farther, and higher. It’s important to consider air resistance, CP = = + N = CV + N
∂τ V ∂τ V
as the horizontal range difference between drag and no drag is almost 3000 m (!!!). It’s important to consider variation
of atmospheric drag with altitude for horizontal range for precision landing (about a 300 m difference). and so
5N
5.3. PS4. Problem 1: Kinetic theory connection to thermodynamics properties. CP =
2
(a) These are all, above, physicists’ quantities. For engineers, specific heat capacities are useful with real-world material. Spe-
(b) Now the Maxwellian velocity distribution, P (v), where P (v)dv is the probability that the particle has speed in (v, v +dv), cific heat capacities are obtained from physicists’ heat capacities by dividing by what physicists would’ve deemed as M , the
is given by (straight-up) mass.
 3/2  2

M −M v
P (v) = 4π v 2 exp CV 3N
2πτ 2τ cV := =
M 2M
If N particles had the same kinetic energy, then the entire system of N particles would have a total internal energy of C P 5N
N 21 M v 2 , with M being the mass of a single particle. cP := =
M 2M
Thus, the total internal energy U is calculated by weighting by P (v) the above total kinetic energy, which in this
case of only 3 translational degrees of freedom, coincides with the total internal energy: N Avog = FundConst [ FundConst [ ‘ ‘ Q u a n t i t y ’ ’ ] . s t r . c o n t a i n s ( ‘ ‘ Avogadro ’ ’ ) ]
>>> c V = f l o a t ( Decimal ( 1 . 5 ) ∗ ( N Avog . Value ) ∗ ( k B o l t z . Value ) ) / M 0
>>> c P = f l o a t ( Decimal ( 2 . 5 ) ∗ ( N Avog . Value ) ∗ ( k B o l t z . Value ) ) / M 0
∞ Z ∞ 3/2 3/2 Z ∞ >>> c V . s u b s ( M 0 , 3 9 . 9 4 8 / 1 0 0 0 . )
−M v 2 −M v 2
Z        
1 MN M M
U =N dv M v 2 P (v) = dv · 4π v 4 exp = 2πM N dvv 4 exp = 312.198102337360
0 2 0 2 2πτ 2τ 2πτ 0 2τ >>> c V . s u b s ( M 0 , 1 3 1 . 2 9 3 / 1 0 0 0 . )
 3/2 √ √ 94.9912774647001
M 3 π 3 1 π 3τ N >>> c P . s u b s ( M 0 , 3 9 . 9 4 8 / 1 0 0 0 . )
= 2πM N 2 1/2
= 2πM N M

3/2
= 520.330170562267
2πτ 4 M 2 M 4 2τ 2π 2



2τ 2τ >>> c P . s u b s ( M 0 , 1 3 1 . 2 9 3 / 1 0 0 0 . )
158.318795774500
U 3τ
Also, = N 2 .
EY : 20151117 I want to reiterate that there must be a more systematic and sane and rational way of looking up So cV and cP for argon is 312.2J/(kgK) and 520.3J/(kgK), respectively, and cV and cP for xeon is 94.99J/(kgK) and
Physical Constants and other physical data than by manually looking it up a book or manually looking it up a website. 158.3J/(kgK), respectively.
People can make a mistake copying and pasting! Thus, I wrote the Python package Physique that you copy into your The so-called molar heat capacity is the amount of heat needed to increase the temperature of 1 mole of substance. Now,
working directory and can import in (this is all in the script Propulsion.py: physicists’ CV and CP is the heat capacities for the amount of heat needed to raise the temperature of a system of N number
10 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

of particles. Then certainly, dividing CV ,CP by N will result in the amount of heat needed to raise the temperature of a single •
particle. Use Avogadro’s number to convert between number of particles and moles. 3/2 3/2 Z ∞ M v2
!0
∞ ∞
−M v 2 e− 2τ
Z Z    
M 3 M 2
c= dvvP (v) =: hvi = dv4π v exp = 4π dvv −M
=
CV 3 0 0 2πτ 2τ 2πτ 0 τ
cV = = !∞ #
N 2 3/2 " M v2 M v2
Z ∞ # 3/2 " 3/2 2
e− 2τ 2τ e− 2τ
  
M M  M 2τ
CP 5 = 4π 0− dv2v −M = 4π M

= 4π =
cP = = 2πτ 2πτ M 2πτ M 2

0 −
N 2 τ τ

0
r r
1 τ 8τ
>>> Decimal ( 1 . 5 ) ∗ ( N Avog . Value ) ∗ ( k B o l t z . Value ) = 3/2
8π =
(2π) M πM
42 12.471689792172872460
Name : Value , dtype : object
>>> Decimal ( 2 . 5 ) ∗ ( N Avog . Value ) ∗ ( k B o l t z . Value ) Note
42 20.786149653621454100 3τ 1/2 1/2


vrms M 3π
= =
c 8τ 1/2 8


and so cV = 12472.J/(kmolK) and cP = 20786.J/(kmolK) πM
Problem 2: Mean thermal velocity. Problem 3: Ideal vs. real rocket analysis.
Wikipedia article “Thermal velocity” has 3 mean thermal velocities 4
 
• P (v) = 4π 2πτM 3/2 2


v exp −M v2 5.4. PS5. Problem 1: Nozzle flow in liquid rocket engines.
2τ
Viking series liquiad rocket engines used on first 2 stages of Ariane 4 launch vehicle.
Now
Rocket engines are storable propellant motors that use
Z ∞ Z ∞ 
exp (−αv 2 )
0 Z 2 Z ∞ nitrogen tetroxide and UDMH25 (unsymmetrical dimethyl hydrazine with 25 percent hydrazine hydrate) as propellants.
4 2 3 2 exp (−αv ) 3 2
dvv exp (−αv ) = v = 0 − 3v = v exp (−αv 2 ) = suitable mean value for molecular weight of the combustion product mixture 23 g/mol ratio of specific heats approximately
0 0 −2α −2α 0 2α
1.2
2
!0 " Z −αv2 # √
e−αv Combustion chamber temperatures for both 3350K approximately.
Z
3 3 e 3 π
= v = 0− = √ Viking 5C motor: chamber pressure 5800 kP a
2α −2α 2α −2α 4α2 2 α
propellant mass flow rate 275.2 kg/s
So nozzle expansion ratio 10.5
3/2 √ Viking 4B engine: chamber pressure 5850 kP a
Z ∞ 
2 2 M 3 π 1 3 1 3τ mass flow 278.0 kg/s
hv i = v P (v) = 4π = M =
M 2

q
0 2πτ 4 2τ M 2
2τ
2 M expansion ratio 30.8
2τ
(a) Use Eq. 4,
so γ+1
p0 γ 1/2
  2(γ−1)
ṁ 2

3τ
1/2
∗
=
vrms = A (RT0 )1/2 γ+1
M
Opening up NozzleTheory.py, using Python’s sympy library,
Now V i k i n g 5 C n o z z l e = m a s s f l o w r a t e E x p . s u b s (gamma , 1 . 2 ) . s u b s ( massflow , 2 7 5 . 2 ) . subs ( p 0 , 5 8 0 0 ∗ 1 0 0 0 ) . subs ( T 0 , 3 3 5 0 ) . \
 τ 1/2 s u b s (R, k B o l t z . Value / ( Decimal ( 2 3 / 1 0 0 0 . ) / N Avog . Value ) )
3τ V i k i n g 4 B n o z z l e = m a s s f l o w r a t e E x p . s u b s (gamma , 1 . 2 ) . s u b s ( massflow , 2 7 8 . 0 ) . subs ( p 0 , 5 8 5 0 ∗ 1 0 0 0 ) . subs ( T 0 , 3 3 5 0 ) . \
hv 2 i = 3hvx2 i = so hvx2 i1/2 = s u b s (R, k B o l t z . Value / ( Decimal ( 2 3 / 1 0 0 0 . ) / N Avog . Value ) )
M M
• s o l v e ( Viking5Cnozzle , Astar ) [ 0 ] # 0.0805128291046479
r s o l v e ( Viking4Bnozzle , Astar ) [ 0 ] # 0.0806368550290290
   
∂P (v) 2P (v) −M 2 M 2τ
= + P (v) v = P (v) − v = 0 if vmp = 0 or vmp = So the area of throat for Viking 5C and Viking 4B is 0.08051 m2 and 0.08126 m2 , respectively. Then, the expansion
∂v v τ v τ M
ratio gives the expanded area for Viking 5C and Viking 4B of 0.8454 m2 and 2.484 m2 , respectively.
where mp stands for most probable. (b) We need a relationship relating (cross-sectional) area ratio to Mach number, Eq. 3
 γ+1 1/2
! γ−1
γ−1
r r 2
2τ 3τ A2 M1  1 + 2 M2
vmp = < = vrms = γ−1 2

M M A1 M2 1+ 2 M1
4“Thermal Velocity”, wikipedia, https://en.wikipedia.org/wiki/Thermal_velocity
 PROPULSION 11
∗
At the throat, M = 1, and denote the area by A . Thus (d) Look at Eq. 5 again:
 γ+1 1/2
γ−1
! γ−1 T = ṁve + (pe − pa )Ae =
Ae 1  1 + 2 M2
= γ+1
 v
A∗ M   γ−1
u !
2 u 2γRT0 pe γ
= ṁt 1− + (pe − pa )Ae
Ae γ−1 p0
and so A ∗ is the expansion ratio.

This is implemented in NozzleTheory.py as a sympy object Eq:

# Area R a t i o t o Mach numbers
Mach 1 = Symbol ( ’ Mach 1 ’ , p o s i t i v e=True )
Mach 2 = Symbol ( ’ Mach 2 ’ , p o s i t i v e=True )
A 1 = Symbol ( ’ A 1 ’ , p o s i t i v e=True )
A 2 = Symbol ( ’ A 2 ’ , p o s i t i v e=True )
AreastoMachs = Eq ( A 2 / A 1 , Mach 1 / Mach 2 ∗ ( ( ( Rat ( 1 ) + (gamma−Rat ( 1 ) ) / ( Rat ( 2 ) ) ∗ Mach 2 ∗∗2 ) / \
( Rat ( 1 ) + (gamma−Rat ( 1 ) ) / ( Rat ( 2 ) ) ∗ Mach 1 ∗∗2 ) ) ∗ ∗ ( (gamma +1)/(gamma−1) ) ) ∗ ∗ 0 . 5 )

and one substitutes in the desired, given numbers (parameters):

Viking5CMachEq = AreastoMachs . s u b s ( Mach 1 , Rat ( 1 ) ) . s u b s (gamma , 1 . 2 ) . s u b s ( A 2 , A 1 ∗ 1 0 . 5 )
# 1 0 . 5 == ( 0 . 0 9 0 9 0 9 0 9 0 9 0 9 0 9 0 9 ∗ Mach 2 ∗∗2 + 0 . 9 0 9 0 9 0 9 0 9 0 9 0 9 0 9 ) ∗ ∗ 5 . 5 / Mach 2

Notice that now we have a so-called “root-finding” problem, with non-integer exponents. One should use a numerical
solver so that finding this root is done “efficiently”. Also, to do this in Python’s scipy, we have to create a Python
function object, and so I used sympy’s lambdify to turn a sympy expression into an actual Python function:
Viking5CMach = l a m b d i f y ( Mach 2 , Viking5CMachEq . r h s − Viking5CMachEq . l h s ) So if pe = pa , then
# Remember t o move a l l t e r m s t o 1 s i d e , and s o t h e o t h e r s i d e e q u a l s 0
v
u
u 2γRT0   γ−1 !
s c i p y . o p t i m i z e . newton ( Viking5CMach , 3 ) # Newton Raphson method pa γ

# 3.3123573073570207 T = ṁt 1−
γ−1 p0
s c i p y . o p t i m i z e . b i s e c t ( Viking5CMach , 3 ,4) # B i s e c t i o n method
# 3.312357307356251
(e) Look at this webpage: http://www.engapplets.vt.edu/fluids/CDnozzle/cdinfo.html for a good recap of the
AIMS Senegal had a nice, easy introduction to scipy’s root finding methods 5. physics of shocks, accompanying the lecture by Polk in AE121.
Next, the static temperature, static pressure, static density at the nozzle exit can be easily calculated from the Also “Choked flow” wikipedia article https://en.wikipedia.org/wiki/Choked_flow
isentropic relations: Consider normal shock at the exit of the nozzle. Then before, M1 > 1 (supersonic), and right after the shock, M2 < 1
γ
p
  γ−1
τ (subsonic)
 −1 =
τ γ−1 2 p0 τ0 p0 Problem 3: Duct flow with heating.
= 1+ M 1 ρ0 =
τ0 2 ρ

τ
 γ−1 RT 0 (a) In this case, I think that the correct heat capacity to use is Cp because the process occurs at the constant ambient
=
ρ0 τ0 pressure (that ambient pressure, external to the (constant area) duct, remains constant and on the duct).
Then the static pressure, static temperature, static density for Viking 5C and Viking 4B engines are Now
Mexh = 3.312 Q = Cp dτ
p = 68.2 kPa
Cp
T = 1597.4 K For cp := MN , q = mcp dT (experimental physicists’ and engineers’ expression). Thus
3
ρ = 0.1181 kg/m Z
∆q̇
q̇ = ṁcp dT =⇒ q̇ = ∆q̇ = ṁcp (Tf − Ti ) or = Tf − Ti
Mexh = 4.057 γ ṁcp
p = 17.0 kPa
T = 1265.9 K Now this heating raises the temperature of the stagnation (enthalpy) temperature, because we’re considering this heating
as a (adiabatic?) thermodynamic process, separate from what’s going on with the flow.
ρ = 0.0372 kg/m3 For instance, along the flow, at a point, the thermodynamic property is (T, p). We can always “pull” this “back” to
(c) the stagnation properties, whether hypothetical or real:
5
Roots finding, Numerical integrations and differential equations, AIMS Senegal, AIMS Senegal
12 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

This is implemented in NozzleTheory.py:

upstream downstream T 01=Symbol ( ’ T 01 ’ , r e a l=True )
heataddTvsMachEq=Eq ( T 0 / T 01 , ( ( Rat (1)+gamma∗ Mach 1 ∗ ∗ 2 ) / ( Rat (1)+gamma∗Mach ∗ ∗ 2 ) ∗ Mach/ Mach 1 ) ∗ ∗ 2 ∗ \
( Rat ( 1 ) + (gamma−Rat ( 1 ) ) / Rat ( 2 ) ∗ Mach ∗ ∗ 2 ) / ( 1 + (gamma−Rat ( 1 ) ) / Rat ( 2 ) ∗ Mach 1 ∗ ∗ 2 ) )
(T0 , p0 ) (T, p)
heataddTvsMachProb0503=heataddTvsMachEq . s u b s (gamma , 1 . 4 ) . s u b s ( T 0 , 6 7 5 ) . s u b s ( T 01 , 6 7 5 + ( 1 . ∗ 1 0 ∗ ∗ 6 ) / \
( 4 0 . ∗ 1 5 . 6 ) ) . s u b s ( Mach , MachProb0503 )
T0 γ−1 2 Mach1Prob0503lamb=l a m b d i f y ( Mach 1 , heataddTvsMachProb0503 . r h s −heataddTvsMachProb0503 . l h s )
=1+ M p l o t ( heataddTvsMachProb0503 . r h s , ( Mach 1 , 0 , 1 0 ) )
T 2
  γ Mach1Prob0503=s c i p y . o p t i m i z e . newton ( Mach1Prob0503lamb , 0 . 1 )
p0 γ − 1 2 γ−1
= 1+ M
p 2 where M was obtained for the acceleration by the converging nozzle, T0 is the stagnation temperature that was given,
Thus, when one considers heat addition, the thermodynamic process, ∆q̇ occurs on stagnation temperature, and then and T01 is obtained from part (a).
can be related to the actual, physical fluid flow and its properties, through formulae we’ve derived: Thus,
h0 = h1 + 12 u21 M1 = 0.2024
(T0 , p0 ) (T1 , p1 )
x0 The flow through the channel isn’t thermally choked as M doesn’t become 1 by the heat addition.
(c) After exiting the ducts, the converging-diverging nozzle results in a Mach number given by
u ∆q̇  γ+1 1/2
! γ−1
γ−1
Ae 1  1 + 2 M2
h0 = h1 + 12 u21 = 
A∗ M γ+1
x1 (T0f , p0f ) (Tf , pf ) 2

(1.∗10∗∗6)/(40.∗15.6) and from the Mach number definition and energy equation (Bernoulli invariant), we can get the exhaust velocity
# 1602.5641025641025
s
So
p T0
ue = Ma = M γRT = M γR
∆T0 = 1602.6 K 1 + γ−1
2 M
2

(b) Now consider the converging section at the inlet to each 1 cm diameter channel, that “accelerates the flow to a relatively
low Mach number and produces a flow rate of 40 g/s”.
Mass continuity (conservation) still holds: Part 3. Basic Feeling
√
p p p √ p0 γ γ − 1 2 −2(γ−1) γ+1 6. Box with a hole rocket; bottled (box) rocket
ṁ = ρAu = AM γRT = √ AM γ = √ AM(1 + M )
RT RT RT0 2 Recall Ch. 14: Kinetic Theory, Section “Kinetic Theory of the Ideal Gas Law” of Kittel and Kroemer [3].
where the “pullback” to the stagnation properties for each point of the flow, before the converging section and after the
converging section, was used: Kinetic Theory of the Ideal Gas Law. Consider molecule strike unit area of wall.
 γ
Let vz ≡ velocity component normal to plane of wall.

p0 γ − 1 2 γ−1
= 1+ M Suppose molecules, of mass M , reflected specularly (mirror-like) from wall,
p 2
T0 γ−1 2 ∆pz = −2M |vz |
=1+ M
T 2
massconsEq = Eq ( massflo w , p 0 ∗ s q r t (gamma/ (R∗ T 0 ) ) ∗A∗Mach∗(1+(gamma−Rat ( 1 ) ) / Rat ( 2 ) ∗ Mach ∗ ∗ 2 ) ∗ ∗ \ Let a(vz )dvz , number of molecules per unit volume with z-component of velocity between vz and vz + dvz .
( ( gamma+1)/(−Rat ( 2 ) ∗ ( gamma− 1 ) ) ) ) Z
massconsProb0503 = massconsEq . s u b s ( massflo w , 4 0 . ∗ 1 0 ∗ ∗ ( − 3 ) ) . s u b s (gamma , 1 . 4 ) . s u b s ( p 0 , 6 . 8 ∗ 1 0 ∗ ∗ ( 6 ) ) . \ N
s u b s ( T 0 , 6 7 3 . ) . s u b s (A, N( p i ) ∗ ( 1 0 ∗ ∗ ( − 2 ) / 2 . ) ∗ ∗ 2 ) . s u b s (R, k B o l t z . Value / ( Decimal ( 2 . 0 1 5 9 ∗ 1 0 ∗ ∗ ( − 3 ) ) / N Avog . Value ) )
a(vz )dvz = =n
V
where for R, R = kMB and where for M , I used 2.0159g/mol for H2 , and Avogadro’s number, NA = 6.022140857 × 1023 , a(vz )vz dvz number of molecules in (vz , vz + dvz ) velocity range that strike unit area of wall in (per) unit time
which is number of particles per mole. So in this example, R = 4124.4 for H2 as Z ∞ Z ∞
>>> k B o l t z . Value / ( Decimal ( 2 . 0 1 5 9 ∗ 1 0 ∗ ∗ ( − 3 ) ) / N Avog . Value ) pressure p = 2M vz a(vz )vz dvz = M vz2 a(vz )dvz = M nhvz2 i
Decimal ( ’ 4 1 2 4 . 4 4 0 6 2 7 7 3 3 8 0 7 6 1 9 8 6 8 5 1 5 6 9 5 ’ ) 0 −∞

1
Let’s try to use the derived relation, relating the stagnation temperatures before and after (denoted 1) heat addition: 2
2 M hvz i = 21 τ by equipartition of energy (Ch.3)
2 " #
1 + γM21 M 1 + γ−1 2
 
T0 2 M Nτ
= p = nM hvz2 i = nτ =
T01 1 + γM2 M1 1 + γ−1
2 M1
2 V
 PROPULSION 13
−n


Maxwell Distribution of Velocities. cf. Ch.6. distribution function of ideal gas f (n ) = λ exp τ
Recall, Ch. 6, Sec. “Classical Limit” of Kittel and Kroemer [3], an ideal gas is defined as a system of free noninter- Z Z π/2 
M
3/2 Z ∞
1

−M 2

3
acting particles in the classical regime. hv out i = v cos θ sin θdθP Maxwell (v)dv = 4π v exp v dv =
0 2πτ 2 0 2τ
f () ≡ average occupancy of an orbital at energy  (7) 3/2  3/2  2 
 ≡ energy of orbital occupied by 1 particle; not energy of system of N particles 21/2  τ 1/2
   
M 1 1 1 M 1 4τ
1
Fermi-Dirac and Bose-Einstein distribution f () = exp [(−µ)/τ ]±1 = 4π 0− = 4π 2
= 1/2
2πτ 2 −2α α 2πτ 4 M π M
In order for f ()  1 ∀ orbitals, exp [( − µ)/τ ]  1 ∀ .
µ for (doing the integration by hand)
2 2
=⇒ f () ' exp [(µ − )/τ ] = λ exp (−/τ ) λ ≡ exp (e−αv )0 = −2αve−αv
τ 2 2 2

f (), average occupancy of orbital of energy , is classical distribution function. (v 2 e−αv )0 = −2αv 3 e−αv + 2ve−αv
πn 2
1

1 1

2 −1 2 2
particle in a box: n = 2M (for, recall 2M i∂ ψ = 2M ∂ ψ = Eψ) 2 −αv 2 e−αv 0 2
L
number of orbitals in range of quantum number (n, n + dn), probability such orbital is occupied (v e + ) = −2αv 3 e−αv
α
Armed with this mean velocity hvout i out of a hole of a box, we want the thrust that results on a box if we had a box of air,
1 1 at some pressure, and at some temperature, and then we punch a hole at one end.
( πn2 dn)f (n ) = πn2 λ exp (−n /τ )dn
2 2 What’s happening? Air that’s swirling around the box is then accelerated out of the hole. There’s fluid flow out. For low
1 2 1  πn 2 2 (M L)2 2 M Lv enough velocities, use Bernoulli’s equation, and assume at the starting point of the air’s streamline, the velocity is 0:
Mv = or n = v or n =
2 2M L π2 π 1 2 p 1 pf
u + = u2f +
Consider system of N particles in volume V . 2 ρ 2 ρf
Let N P (v)dv number of atoms with velocity magnitude in range dv at v p 1 2 pf
3 =⇒ = uf +
1 2 dn 1

ML

−M v 2
 ρ 2 ρf
2
N P (v)dv = πn λ exp (−n /τ ) dv = πλ v exp dv
2 dv 2 π 2τ ρ is really MVN and by the ideal gas law (still applies), pV = N τ and τp = N V but the point is the gas density didn’t change
 2 3/2 much.
cf. Ch. 6, λ = nnQ = LN3 2π~Mτ

 3/2  3  3/2
1 N 2π ML N
π = 4πN
2 L3 Mτ π 2πτ
3/2
−M v 2
  
M
(6) =⇒ P (v) = 4π v 2 exp
2πτ 2τ
P (v) is Maxwell velocity distribution, P (v)dv is probability particle has speed in dv at v

Experimental verification. velocity distribution of atoms which exit from slit of oven.
exit beam weighted in favor of atoms of high velocity at expense of those at low velocity.
weight factor is velocity component v cos θ normal to plane of hole. Thus,
p − pf 1
π/2 = u2f
2πR3 − cos 2θ 
Z Z Z 
1 2π 3 sin 2θdθ ρ 2
(cos θ)drrdϕr sin θdθ = ( (2π)R3 ) cos θ sin θdθ = ( R ) = =
3 3 2 3 4 
0 The thrust is going to be given by the difference in pressure against the walls, the wall in front of the box opposite to the wall
2πR3 1 + 1
 
2πR3 1
  with a hole. You don’t need the area of the hole. This thrust is (p − pf )
= =
3 4 3 2 1 MN 1 2 MN 1 2
(p − pf )A = ρ u2f A = u A= u
Probability atom leaves hole will have velocity in (v, v + dv) : Pbeam (v)dv 2 V 2 f L 2 f
2 τ
Now from Eq. 7, u2f = π M, and so
−M v 2
 
Pbeam (v) ∝ vPMaxwell ∝ v exp 3 Nτ pV pL2
2τ (p − pf )A = = =
  πL πL π
M 3/2 2
with, recall PMaxwell = 4π 2πτ


v exp −M v2 Thus, the thrust on this box is given by
2τ
Nτ pL2
Pbeam distribution of transmission through a hole is Maxwell transmission distribution. Fthrust = =
πL π
14 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM
∂
Part 4. Notes and Solutions for Thermal Physics by Ralph Baierlein Suppose Q(γ̇) = Q(t)dt(γ̇ ∂t ) = Q(t)γ̇.
[2] Z Z Z
pV̇ dt = Q(t)γ̇dt − U̇ dt
7. Background
=⇒ p∆V = ∆Q − ∆U
7.1. Heating and Temperature. Heating: keep in mind 3 different types of heating for energy exchange between two systems:
(1) Heating by conduction - literal contact, molecules jiggle faster from molecules jiggling faster by bouncing off them p∆V interpreted as work done by gas. ∆Q is heat transferred to gas system. −∆U is the drop in internal energy of gas system
(2) Heating by radiation - em waves from hot source strike and excite target as it does work.
(3) heating by convection - energy transport by flow (perhaps a fluid)
This all relates to
Q 7.4. Heat capacity.
7.2. Some dilute gas relationships. 
∂Q
 
∂Q

Q = Q(τ, V ) = dτ + dV
Pressure according to kinetic theory. i.e. some kinetic theory ∂τ V ∂V τ
F ≡ force on area A due to molecules  
∆p ≡ momentum transferred to wall per collision So define CV ≡ ∂Q or interpret CV as energy input by heating at constant volume over ensuing change in temperature.
∂τ
n ≡ number of collisions in time ∆t ∂U

V
In this case, ∂τ V .
Thus
(∆p)n For the case of a monatomic gas, U = d2 τ N , ∂U d
∂τ = 2 N .
F =
∆t
Now ∆p = 2mvx since ∆p = mvx − (−mvx ) = 2mvx (elastic collision with momentum conversation) d
vx ∆tA is a volume, inside of which gas molecules can be within distance vx ∆t toward the wall. CV = N
2
N
V number density of molecules
Assume equal distribution of velocities: Thus 12 N ≡ number of molecules.
1N Now
n = (vx ∆tA)
2V Q = Λp dp + Cp dτ Q ∧ dp = Cp dτ ∧ dp
=⇒
F (2mvx )(vx ∆tA) 21 N N Q = ΛV dV + C V dτ Q ∧ dp = ΛV dV ∧ dp + CV dτ ∧ dp
P = = V
= mvx2
A A∆t V
Suppose hv 2 i = hvx2 i + hvy2 i + hvz2 i = 3hvx2 i = dhvx2 i. Now from the thermodynamic identity, Q = W + dU = pdV + dU ,
mN 2 21 N
=⇒ P = hv i = mhv 2 i Q ∧ dp = pdV ∧ dp + dU ∧ dp
dV d2 V
An empirical gas law. Now
Nτ and from (empirical) ideal gas law, pV = N τ (which defines a hypersurface on M ),
P = (empirical)
V
dpV + pdV = N dτ =⇒ pdV ∧ dp = N dτ ∧ dp
1 d
=⇒ mhv 2 i = τ
2 2
so then
7.3. The First Law of Thermodynamics.
dU = Q − W or Q = dU + W Q ∧ dp = N dτ ∧ dp + dU ∧ dp
Consider W = Q − dU .
In the case of the monatomic gas, U = d2 τ N , and so dU = d2 N dτ = CV dτ and so comparing all the equations above, one recovers

Consider path in M , γ, γ : R → M = (U, V ) 2+d

γ(t) = (U (t), V (t)) Cp = N + CV = N
2
∂ ∂
γ̇ ∈ X(M ), γ̇ = U̇ ∂U + V̇ ∂V
EY: 20151019 I’m curious to know how this all generalizes for CV , CP heat capacities, regardless of the type of molecule we
W (γ̇) = pdV (γ̇) = pV̇ = Q(γ̇) − dU (γ̇) = Q(γ̇) − U̇ consider.
 PROPULSION 15

The adiabatic relation for a classical ideal gas. Consider the adiabatic expansion (or contraction!) of a classical ideal gas. s o l v e ( FCconv . s u b s ( T C , Prob0104ansC ) , T F ) [ 0 ] # 1 0 9 9 . 1 6 0 5 4 5 4 3 5 2 6 F
This means that Q = 0; there is no heat exchange to or from the gas system.
Recall Q = dU + W . The final temperature is 866.01 K or 592.87 C or 1099.16 F
If Q = 0, and supposing W = pdV , then 0 = dU + pdV . (b) Now from the ideal gas law, which is obeyed at all thermodynamic states
EY : 20151019 Either by definition, or the thermodynamic identity, τ dσ = dU + pdV , then CV := ∂U


∂τ V . My question is Pf Vi τf
this: for manifold of thermodynamic states M , M = (U, V ), i.e. U is a global coordinate and V is a “local” coordinate. One =
Pi Vf τi
can make a Legendre transformation such that M is parametrized by (τ, V ), where τ is the temperature. In general, one should
Pf
say that U = U (τ, V ) ∈ C ∞ (M ), and so dU = ∂U ∂U 1
∂τ dτ + ∂V dV , dU ∈ Ω (M ).
and so Pi = 44.31
However, for this adiabatic process, we want
Q = 0 = dU + W = dU + pdV = CV dτ + pdV 8. Phase Equilibrium
∂U
which implies that dU = CV dτ . What happened to the ∂V dV
? Is it that in this adiabatic process, the internal energy of the
gas system goes to either doing work (expansion) or increases due to work being done on it (contraction), and is characterized 8.1. Latent heat. p, T const., liquid → gas, e.g. p = 1 atm
completely by a drop or increase in its temperature, respectively? And so dU = CV τ , and pdV completely describes what’s T = 373 K
going on with work done or work done on it? latent heat of vaporization Lvap (by def.) amount of energy supplied by heating.
Nevertheless, using the (empirical) ideal gas law, pV = N τ , Q = dU + W
Nτ  := U/N = average (internal) energy per molecule
0 = CV dτ + pdV = CV dτ + dV
V v := V /N = volume per molecule
Lvap = d + pdV
Consider a path γ : R → M ∂
in M , so that γ̇(t) = τ̇ ∂τ + V̇ ∂
∈ X(M ). d(γ̇) = vap − liq
∂V
dv(γ̇) = vvap − vliq > 0
γ(t) = (τ (t), V (t))
if p const., Lvap = d( + pV ) = dh, h := H/N = U +pV
N
Thus,
EY : 20151031 A better way to think about it is this: recall that
Nτ τ̇ N V̇
0 = Cv τ̇ +
V̇ so + Q = dU + W = dU + pdV = τ dσ
V τ CV V
τf N Vf H = U + pV so dH = dU + pdV + V dp = Q + V dp
R
dt N
−−→ ln + ln = 0 or ln τ V CV = const.
τi CV Vi
Now
N CP − CV Consider a path γ ∈ M s.t. dτ (γ̇) = 0 (constant τ, p)
= =γ−1
CV CV dp(γ̇) = 0
which is true, assuming the (empirical) ideal gas law, the thermodynamic identity, and, surely, for the case of a monatomic gas.
Q(γ̇) = dH(γ̇) − V dp(γ̇) = dH(γ̇) − 0 = dH(γ̇) Q(γ̇) = τ dσ(γ̇)
Thus Z Z
γ−1 γ−1
τf V f = τi Vi Q(γ̇) = τ dσ(γ̇) = τ (σg − σl )
Problems. Problem 4. Adiabatic compression.
Thus
A diesel engine doesn’t have a spark plug to ignite and explode the fuel. Instead, the air in the cylinder is compressed so R
Q(γ̇) Hg Hl
highly that the fuel ignites spontaneously when sprayed into the cylinder. = τ (sg − sl ) = −
N N N
(a) R
Q(γ̇)
γ−1 γ 1
τf Vf Pf Vf L ≡ τ (sg − sl ) = = (Hg − Hl )
= γ = 1 N N
τi Viγ−1 Pi Vi
 γ−1 Latent heat versus heat capacity. Take slow, reversible process.
Vi Now
τf = τi
Vf
 
∂σ
CV := τ
Run the Python script thermo.py to do the calculations. Here is (some of) the code from thermo.py for doing so ∂τ
(one still needs to import the necessary libraries):  V
∂σ
roomtemp K = KCconv . s u b s ( T C , 2 0 ) . r h s # room t e m p e r a t u r e i n K e l v i n CP := τ
∂τ P
Prob0104ans = a d i a t V . s u b s (gamma , 1 . 4 ) . s u b s ( V f , 1 ) . s u b s ( V i , 1 5 ) . s u b s ( t a u i , roomtemp K ) # answer t o Problem 4 o f Chapter 1
It’s stated in Kittel and Kroemer (1980), pp. 166, Equation (37), Chapter 6: Ideal Gas, Subsection “Heat capacity” [3], that
     
Prob0104ans = N( Prob0104ans . l h s ) # 8 6 6 . 0 1 6 9 6 9 6 8 6 2 5 3 K ∂σ ∂U ∂V
Prob0104ansC = s o l v e ( KCconv . s u b s ( T K , Prob0104ans ) , T C ) [ 0 ] # 5 9 2 . 8 6 6 9 6 9 6 8 6 2 5 3 C (8) CP = τ = +p
∂τ P ∂τ P ∂τ p
16 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

Suppose σ = σ(τ, V ). With τ dσ = dU + W = τ dσ = dU + pdV , 9.

∂σ ∂σ dU p 10. Properties of a Pure Substance
dσ = dτ + dV = + dV
∂τ  ∂V τ τ
∂σ 1

1 ∂U
 EY : 20151030 Is the word vapor the same as gas? vapour, gaz, gas, vapore
dσ(γ̇) = + 0 = dU (γ̇) = For coexistence equilibrium,
∂τ V τ τ ∂τ V
Then it’s clear that     µg (p0 , τ0 ) = µl (p0 , τ0 ) and
∂σ ∂U
τ = = CV µg (p0 + dp, τ0 + dτ ) = µl (p0 + dp, τ0 + dτ )
∂τ V ∂τ V
Now suppose σ = σ(τ, p).
Consider also the enthalpy, H = U + pV , and so and so

dH = dU + V dp + pdV = τ dσ + V dp dp L
(9) =
Now for σ = σ(τ, p), dτ τ ∆v
σ = σ(τ, p)    
∂σ 1 ∂H where v ≡ NV
, the so-called vapor pressure equation or Clausius-Clapeyron equation.
∂σ ∂σ dH V =⇒ dσ(γ̇) = +0= −0
dσ = dτ + dp = − dp ∂τ τ ∂τ V dp L
∂τ ∂p τ τ p p For (2) approximations, ∆v = vg − vl ≈ vg = Ngg and idealize vapor as ideal gas, pV = N τ so dτ = τ 2 /p .
So Second, if L constant,
   
∂σ ∂H
Cp := τ =
∂τ ∂τ
 
p p p(τ ) −L0
(10) p(τ ) = p0 exp (−L0 /τ ) or ln =
Now       p0 τ
∂U ∂V ∂H
dH(γ̇) = dU (γ̇) + V dp(γ̇) + pdV (γ̇) = +0+p = so
∂τ p ∂τ ∂τ
     
p p cf. Ch. 10 Phase Transformations, pp. 278-284, “Derivation of the Coexistence Curve, p Versus τ ” of Kittel and Kroemer
∂σ ∂U ∂V (1980) [3].
Cp = τ = +p
∂τ p ∂τ p ∂τ p Eq. 10 explains the shape of the coexistence curve between solid and gas (vapor) (sublimation) and liquid and gas (vapor)
Kittel and Kroemer (1980) [3] argues that for ideal gas, (vaporization; vaporization curve).
    Saturation is this p = p(τ ) coexistence curve.
∂U ∂U Isotherms, Isothermals
=
∂τ p ∂τ V
since U = U (τ ). Recall G = F + pV and so dG = dF + V dp + pdV = dU − τ dσ − σdτ + pdV + V dp = −σdτ + V dp
F = U − τσ dG = −σdτ + V dp
8.2. Conditions for coexistence. Sec. 12.3 of Baierlein (1999) [2].
Recall
G = F + pV = U + pV − τ σ = G(τ, p, N ) So then G = G(τ, p)
Now G = G(τ, p, N )
∂G ∂G where the latter statement is when we include particle transfer, so that
G = G(τ, p, Nvap , Nliq ) =⇒ dG = dNvap + dNliq = µvap 1 + µliq (−1) = 0
∂Nvap ∂Nliq
=⇒ µvap (τ, p) = µliq (τ, p) dG = −σdτ + V dp + µdN for µ = µ(τ, p)
From Kittel and Kroemer (1980) [3], Example : N atoms in a box, Chapter 3: “Boltzmann Distribution and Helmholtz Free
Energy”,
For the ideal gas:
state of energy α (1) + β (2) + · · · + ξ (N ), α, β, . . . ξ denote orbital indces of atoms in successive boxes.
each entry occurs N ! times in Z1N (EY: 20151022) N ! ways to fill α, β . . . ξ orbitals with N distinguishable particles. Thus,
 n  
1 N 1 F (τ, V ) = F = −N τ ln
Q
+1
ZN = Z1 = (nQ V )N n
N! N!  3/2
Mτ
Part 5. Notes and Solutions on Fundamentals of Thermodynamics, 8th Edition, by Claus Borgnakke, Richard nQ =
where 2π~2
E. Sonntag p
n ≡ N/V =
[5] τ
n  n   3/2 !
Q Q Mτ τ
G(τ, p, N ) = −N τ (ln + 1) + N τ = −N τ (ln ) = −N τ ln
n n 2π~2 p
   3/2 !
∂G Mτ τ
so then = µ = −τ ln 2
∂N τ,p 2π~ p
 PROPULSION 17

For the Van der Waals gas 12. Phase Equilibrium

N 2a
   
nQ (V − N b) : Phase Equilibrium
F (vdW ) = −N τ ln +1 − : Phase Diagram
N V
: (Phase) Coexistence Curve
N 2a
 
∂F Nτ
p=− = − 2 : Antoine Equation (parameters)
∂V τ,N V − Nb V
dp L
NτV 2N a2

nQ (V − N b)
 For coexistence equilibrium, µg (p0 , τ0 ) = µl (p0 , τ0 ) and µg (p0 + dp, τ0 + dτ ) = µl (p0 + dp, τ0 + dτ ), and so dτ = τ ∆v with
G(τ, V, N ) = − − N τ (ln + 1) V
v ≡ N.
V − Nb V N V dp
Make the approximation that ∆v ≡ vg − vl ≈ vg = Ngg and idealize vapor as ideal gas, pV = N τ , so dτ = τ 2L/p .
Nevertheless, consider, when considering isotherms, isothermals, If L constant,
dG = −σdτ + V dp + µdN  
p(τ ) −L0
Consider a path γ on constant τ , constant total number of particles N , and so (11) p(τ ) = p0 exp (−L0 /τ ) or ln =
p0 τ
Z
dG(γ̇) = V dp(γ̇) = Gg − Gl = V dp Consider the Antoine Equation Parameters given in the NIST (National Institute of Standards and Technology) Chemistry
γ
WebBook 6

Part 6. Thermodynamics (Revisited)

(12) log10 (P ) = A − (B/(T + C))
11. Heat Capacity
From Kittel and Kroemer (1980) [3], pp. 165-166, Chapter 6: Ideal Gas, “Heat Capacity”, Then
 
B
A= T +C A B
− T +C A −B
Q = Q(τ, p) = Λp dp + Cp dτ = dU + W = dU + pdV P = 10 = 10 10 = 10 exp ln 10
T +C
Let c ∈ Σ s.t. dp(ċ) = 0 (constant pressure). And so
D D DC
Now T +C = T − T (T +C) , and so
Q(ċ) = 0 + Cp dτ (ċ) = τ dσ(ċ) = dU (ċ) + pdV (ċ)
   
(B ln 10)C −B ln 10
for c = (τ, 0), ċ = ∂
∂τ ∈ T Σ. P = 10A exp exp
T (T + C) T
     
∂σ ∂U ∂V Consider how much the pressure is changed due to this C parameter. Consider P0 , P1 , defined as such:
=⇒ Cp = τ = +p
∂τ p ∂τ p ∂τ p        
−B ln 10 (B ln 10)C −B ln 10 P0 (B ln 10)C
for heat capacity at constant pressure is larger than CV because additional heat must be added to perform the work needed to P0 = 10A exp P1 = 10A exp exp =⇒ = exp
expand volume of gas against constant pressure. T T (T + C) T P1 T (T + C)
Now recall for enthalpy P0
H = U + pV So a deviation can be estimated from 1 − P1
.
Open up thermochem.py. The saturation curve or coexistence curve for, for example, oxygen, from liquid to gas, can be
H = H(σ, p) reproduced. One needs to input in the Antoine parameters from the NIST website 7
Then
>>> O2coec = C o e x i s t C u r v e ( 3 . 8 5 8 4 5 , 3 2 5 . 6 7 5 , −5.667 )
dH = dU + pdV + V dp = Q + V dp >>> p l o t ( O2coec . c u r v e S I . r h s , (T, 5 4 . 3 6 , 1 0 0 . 1 6 ) )

Thus, for c ∈ Σ, dp(ċ) = 0 (constant pressure). Hence >>> N( O2coec . c u r v e S I . r h s . s u b s (T , 6 0 . ) )

  731.804072053687
∂H >>> N( O2coec . c u r v e S I . r h s . s u b s (T , 7 0 . ) )
dH(ċ) = = Q(ċ) + 0 = Cp 6253.42680398774
∂τ p >>> N( O2coec . c u r v e S I . r h s . s u b s (T , 7 5 . ) )
14494.1195824433
Hence,
   
∂H ∂Q
Cp = =
∂τ p ∂τ p
6Phase change data for Oxygen
7http://webbook.nist.gov/cgi/cbook.cgi?ID=C7782447&Mask=4#Thermo-Phase
18 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

References
[1] George P. Sutton, Oscar Biblarz. Rocket Propulsion Elements, 7th Edition. Wiley, 2001.
[2] Ralph Baierlein. Thermal Physics Cambridge University Press (July 28, 1999), ISBN-13: 978-0521658386
[3] Charles Kittel, Herbert Kroemer, Thermal Physics, W. H. Freeman; Second Edition edition, 1980. ISBN-13: 978-0716710882
[4] Bernard F. Schutz, Geometrical Methods of Mathematical Physics, Cambridge University Press, 1980. ISBN-13: 978-0521298872
[5] Claus Borgnakke, Richard E. Sonntag. Fundamentals of Thermodynamics, 8th Edition, Wiley, (December 26, 2012). ISBN-13: 978-1118131992
There’s an 8th edition of Biblarz and Sutton [1] for 2010 that I would like to have. If you find any of this material useful or if you’d like to help, email me or visit my Open/Tilt page ernestyalumni.tilt.com and donate to the crowdfunding campaign or
click on the PayPal donate button.
 PROPULSION 19

13. Code listings

20 ERNEST YEUNG ERNESTYALUMNI@GMAIL.COM

14. Automatic generation of documentation

Demontration using epydoc:
epydoc --pdf -o /home/fnielsen/tmp/epydoc/ --name RBBase wikipedia/api.py
This example does not use brede_str_nmf but another more well-documented module called api.py that are used to download material from Wikipedia.