General Physics Notes

李永丞

June, 2021
Contents

1 Preliminary (Math Tools) 4

1.1 Introduction to conic section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.1 Ellipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.2 Parabola . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.1.3 Hyperbola . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.2 More on ellipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.2.1 Conjugate diameters of ellipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.2.2 Focus and Directrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 Kepler’s law and Newton’s law 10

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.2 Laplace-Lenz-Runge vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.3 Conservation of mechanical energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2.4 Application of Laplace-Lenz-Runge vector . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.5 Properties of Laplace-Lenz-Runge vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.5.1 Inner product and outer product . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.5.2 Polar coordinate for planar motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

1
 2.5.3 Properties of Laplace-Lenz-Runge vector . . . . . . . . . . . . . . . . . . . . . . . . 25

2.6 Aspect from calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.7 Conservation laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.7.1 Conservation of Mechanical Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.7.2 Application of the conservation of mechanical energy . . . . . . . . . . . . . . . . . 35

2.7.3 Conservation of momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

2.7.4 Conservation of angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

3 Thermodynamics 53

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.1.1 Constant volume gas thermometer . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

3.2 Kinetic theory of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.3 Maxwell–Boltzmann distribution and Equipartition Theorem . . . . . . . . . . . . . . . . 58

3.3.1 The distribution of the atmosphere pressure . . . . . . . . . . . . . . . . . . . . . . 58

3.3.2 Maxwell speed distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.4 Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.5 First Law of Thermodynamics and its application . . . . . . . . . . . . . . . . . . . . . . . 64

3.5.1 Measure of the specific heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2
 3.5.2 Reversible adiabatic process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

3.5.3 Reversible isothermal process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

3.5.4 Adiabatic process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3.6 Carnot heat engine and Carnot cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

3.7 Second Law of Thermodynamics and Clausius inequality . . . . . . . . . . . . . . . . . . . 70

3.7.1 Second Law of Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

3.7.2 Clausius inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

3.8 Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

3.8.1 Computation of the difference of entropy . . . . . . . . . . . . . . . . . . . . . . . . 76

3
1 Preliminary (Math Tools)

In the first, we introduce some math tools which will use often in the future.

1.1 Introduction to conic section

1.1.1 Ellipse

We have two methods to construct ellipse, which are cutting cylinder and cutting cone. We now prove
that the curve which cut out is exactly a ellipse.

A First put balls under and above the plane which

tangent to the plane and both sides of cylinder.
We have known that a point outside a ball have
F1 F2 same length to any tangent point. Hence we have

P F1 = P A P B = P F2 .
P
On the other hand, P A + P B is a constant.
Hence P F1 + P F2 is a constant.

4
 Let B1 H = y, OB11 = b. From high school, we
have y 2 = V2 H ¨ HV1 . Observe that

△V1 V2 W „ △V1 HB △V2 V1 Z „ △V2 HA

combining y 2 = HA ¨ HB, we get

V1 H ¨ V2 W ¨ V1 Z ¨ V2 H
y2 = 2
V1 V2
B1 V1
Z If B1 coincides B11 , then y = b. Denote V1 V2 by
A H B 2a, we have
B11
O B2 V1 Z ¨ V2 W = 4b2
V2 W
Let OH = x, then we have

4b2
y2 = V1 H ¨ V2 H
(2a)2
y2 1
ñ 2 = 2 (a2 ´ x2 )
b a
x2 y 2
ñ 2 + 2 =1
a b

1.1.2 Parabola

Next we show that if we cut parallel to the generating line, we will get a parabola.

5
 A Let V H = y, BH = x. Similar to above, we
have
x2 = CH ¨ DH
Observe that
CH y
=
V CD AD
B Hence
AD AD x2
C H D y= CH =
CD CD DH
AD 1
Note that is a constant. Hence it is
CD DH
proved.

1.1.3 Hyperbola

Finally, we show that the vertical cut will get a hyperbola.

Let OG = b, GV2 = a, GH = x, CH = y.
As before, we have

y 2 = AH ¨ BH

V2 and
b a
=
AH x+a
O On the other hand,
G
HB x´a
=
V1 b a
Combining all together,
C
A B b b b2
H y 2 = (x + a) ¨ (x ´ a) = 2 (x2 ´ a2 )
a a a
x2 y 2
ñ 2 ´ 2 =1
a b

6
1.2 More on ellipse

1.2.1 Conjugate diameters of ellipse

Q
(x2 , y2 )
P (x0 , y0 )

C(xc , yc )
C1 L
O

C1
(x1 , y1 )
P1
Figure 1

Let P be any point on the ellipse, and L be a tangent line through P . Define P 1 by extending OP and
cross the ellipse on the other side. Draw a parallel line of L through O and get two points C and C 1 . We
called P P 1 and CC 1 conjugate diameter. The first goal is getting the slope of L. We calculate it as
follow.
Method 1
First, we draw a parallel line of P P 1 as Figure 1. We denote the slope of L by m. Since (x1 , y1 ) and (x2 , y2 )
are on the ellipse, we have
$ 2 2
& x1 + y 1 = 1
’
(x1 ´ x2 )(x1 + x2 ) 1
a22 b22 ùñ = 2 (y1 ´ y2 )(y1 + y2 )
x y
% 2 + 2 =1
’ a 2 b
a2 b2
Observe that the slope of C1 O also equals to m, hence
 
y1 + y2
´0
2 b2 (x1 ´ x2 ) b2 (x0 ´ 0) b2 x 0
m=   =´ 2 =´ 2 =´ 2
x1 + x2 a (y1 ´ y2 ) a (y0 ´ 0) a y0
´0
2

7
Method 2
We can also get m by calculus. We do implicit differentiation on the equation of ellipse.

x2 y 2
+ 2 =1
a2 b
2x 2y dy
ñ 2 + 2dx = 0
a b
dy b2 x
ñ =´ 2
dx a y
Next, we will show a special property of ellipse. We need to calculate the coordinate of C.
x0 x y 0 y
Since we know that the equation of CC 1 is 2 + 2 = 0, hence
a b
x2c yc2
1= + 2
a2 b
x2c 1 b4 x 2
= 2 + 2 4 20 x2c
a b a y0
2
 2 
b y0 x20 1 2 b2 1 2
= 2 + x = x
a b2 a2 y02 c a2 y02 c
 
a b
So, we get the coordinate of C is ´ y0 , x0 .
b a
Then the area of OP QC is

x0 y0  2 
b 2 a 2 x0 y02
a b = x0 + y0 = ab + 2 = ab
´ y0 x0 a b a2 b
b a
This means that the area generating by conjugate diameter is independent of the choice of conjugate
diameter.
Finally, we give a small formula.

h h
ùñ
A H B A1 H1 B1

Figure 2

8
For circle, we know
h2 = AH ¨ HB
We want to get some similar formula for ellipse. Note that the ellipse is a expand of a circle. Hence we
a a
have AH ¨ = AH 1 and HB ¨ = H 1 B. We get the formula
b b
h2 AH HB AH 1 H 1 B
= ¨ = ¨
b2 b b a a

1.2.2 Focus and Directrix

We define the directrix by the construction of the first method of ellipse (or see wikipedia). Define directrix
to be a cross line that the plane of ellipse and the plane of bottom or top of the cylinder.

P
H

θ
F s

L
Figure 3

Let L be a directrix of the ellipse. We define eccentricity

PF
ϵ=
PH
By simple calculation, we will get
1 1 1
r(1 + ϵ cos θ) = sϵ and = + cos θ
r sϵ s

9
2 Kepler’s law and Newton’s law

2.1 Introduction

We now introduce how Kepler use the data to find the orbit of Earth or Mars.

• Looking the trajectory of Earth from Mars

M0
β1
EN E0

E1 α1
S

Figure 4

Let TM be a period of the revolution of Mars. Let M0 be the location of Mars at t = 0, Ei be the
location of Earth at t = i ¨ TM . Write SM0 = R0 , SEi = ri , and =SM0 Ei = βi , =SEi M0 = αi .
From the law of sines, we know
R0 ri R0
= ùñ ri = sin βi
sin αi sin βi sin αi
The results have been shown in the Chapter 30, New Astronomy by Kepler. He found that the orbit
of Earth is NOT a circle. (Eccentricity is about 0.03)

• The orbit of Mars

10
 M1

β1,1
E0,1
E1,1
M0
E0
α1,1
S

Figure 5

Let Top be the period of Mars opposition. For i = 0, 1, let ti,j = i ¨ TM + j ¨ Top . The location of Mars
at ti,j is Mj and location of Earth is Ei,j . Also let SMj = Rj , SEi,j = ri,j and =SE1,j Mj = α1,j ,
=SMj E1,j = β1,j . Similarly, by law of sines

Rj r1,j r1,j
= ùñ Rj = sin α1,j
sin α1,j sin β1,j sin β1,j

Also, Kepler observed that there are some rules in the revolution in 1609, and 1618.

Kepler’s second law Equal times sweeps equal area.

Kepler’s first law Orbit is an ellipse, sun sits at its focus.
a3
Kepler’s third law 2 is a constant
T
where a is semi-long axis, and T is period.
Kepler tried to explain the reason. He thought that sun and planet are something like magnet and there
is some gravitation between them without contact.
After a long time, Galileo did some experiment and found some results.

11
 Figure 6

Figure 6 is a sketch of the experiment that Galileo did. Galileo found the ball will climb up to same height
as the height that ball fall down, and keep rolling if there is no uphill. After Descartes adapted,
á
Law of inertia Under no action force, the body keeps its constant velocity v .

Galileo also found that acceleration is independent of mass by throwing ball with different mass from Tower
of Pisa, and the distance of uniform accerlerated motion is given by the formula
1
S(t) = at2 . (1)
2
Then Newton gave his result in Mathematical Principles of Natural Philosophy on 1687,

Newton’s first law Law of interia

á d  á á
Newton’s secong law F = mv = ma
dt
Newton’s third law Action equals to reaction

and
á GM m á
Universal Gravitation F = e ár
r2
Now, we prove Kepler’s second law.

Proof. First we consider that there is no action force.

12
 á
v ∆t A5
O á
v ∆t A4
á
v ∆t A3
á
v ∆t A2
á
v ∆t
A1
A
Figure 7

á
It is clear that Area △OAA1 =Area △OA1 A2 = ¨ ¨ ¨ . We now consider if F is central force.

D3
C3

O
C2

B2
B1

á A1
A0 v ∆t

Figure 8

If no force action, the planet will be located at A1 at time ∆t. But there is a central force, the planet will be
at B1 at time ∆t with OA0 ∥ B1 A1 . Similar for OB1 ∥ C2 B2 , and so on. The true orbit is A0 Ñ B1 Ñ C2 Ñ
D3 . Since OA0 ∥ B1 A1 , we get Area △OA0 A1 =Area △OA0 B1 . Similarly, Area △OB1 B2 =Area △OB1 C2 .
Hence we get
Area △OA0 B1 = Area △OB1 C2 = Area △OC2 D3

Next, we show that acceleration is inversely proportional to the square of distance. (This method is based
on Newton)

13
 R
P
Q x
v

D I
E
S
C H

Figure 9

á
v is a point, but v is velocity. (Abuse notation)
As Figure 9, P is the location of planet, C is the center of the ellipse and S be the location of sun. Also,
P R ∥ Qv ∥ IH ∥ DC and RQ ∥ P S.
á á á
By law of inertial, P R = v ∆t, and by the equation (1), RQ = 12 A(∆t)2 .
Note that △P IH is an isosceles triangle since the =P IH = =P HI (see reflection property in Wikipedia).
Hence P H = P I. On the other hand, P H + P I + IE + ES = 2a. Oberserve that IE = ES since C bisects
SH.
Therefore, we get P I + IE = P E = a.
Since △P xv „ △P EC,
Px Pv Pv
= ñ Px = a
PE PC PC
á
Observe that QRP x is a parallelogram, which implies that QR = P x, P R = Qx = | v |∆t.
By the formula,
2 2 2
Qv P v ¨ Gv Qv CP
2 = 2 ñ Pv = 2 ¨
CD CP CD Gv

14
Combining all above,
2 2
1á 1 Qv CP
|A|(∆t)2 = QR = P x = a ¨ 2 ¨
2 P C CD Gv
2 2 2
1 1 P C Qv ¨ Qx
=a ¨ ¨ ¨
P C CD2 Gv Qx
2

2 2
1 1 CP Qv á 2
=a ¨ 2 ¨ ¨ | v | (∆t)2
P C CD Gv Qx2

Therefore,
2 2
á 2 1 CP Qv á 2
|A| = a ¨ 2 ¨ ¨ |v |
P C CD Gv Qx2
Now, let Γ be the area swept per unit time, and β = =SP R = =P IH = =P HI.
Then
1á á 2Γ
| v |∆t sin β ¨ SP = Γ∆t ùñ | v | =
2 SP sin β
2 2
á 2 1 CP Qv 4Γ2
ùñ |A| = a ¨ ¨ ¨ ¨
P C CD2 Gv Qx2 SP 2 sin2 β

Now, let ∆t Ñ 0, then Qx Ñ Qv, Gv Ñ 2CP . Hence

á 4Γ2
|A| = a 2 2
CD sin2 β ¨ SP
πab
By the definition of Γ, we know Γ = . On the other hand, CD and CP are conjugate diameters, we
T
know the parallelogram generated by CD and CP has area ab. Since RP ∥ CD, the area of parallelogram
generated by CD and CP is equal to the area of parallelogram generated by CD and P E. So we have

CD loP Eon sin β = ab ùñ CD sin β = b

omo
=a

Finally,
á 4Γ2
|A| = a 2 2
CD sin2 β ¨ SP
4π 2 a3 b2
= 2 2
T 2 ¨ SP ¨ CD sin2 β
4π 2 a3 1
=
T 2 SP 2

15
Now, Denote the mass of Earth and Sun by mP and MS respectively. By Newton’s second law,
á á
F S,P = mP A 9 mP

On the other hand, by Newton’s third law

á á
F S,P = ´F P,S 9 MS

So we get that
á
F S,P 9mP MS
By the previous argument,
á á 4π 2 a3 1
|F S,P | = mP |A| = mP
T 2 SP 2
4π 2 a3
Combining Kepler’s third law, it turns out that is a constant related to MS .
T2
4π 2 a3
Now, we suppose is proportional to MS , that is
T2
4π 2 a3
= GMS
T2
where G is called gravitational constant, and it is first measured by Cavendish in 1978. (see Wikipedia)
To sum up,
GmP MS á
Newton’s Universal Gravitation Law F S,P = ´ 2 e á
SP SP
á
where e á is the unit vector from S to P and the minus sign means the force is attraction.
SP
Now, we consider another problem

R
P
Q
v

Figure 10

16
This is similar to the previous one.
Similarly, $ á
&Qv = RP = |V H |∆t
’
2 2
Qv CP
%P v =
’
2 ¨
CD Gv
Hence á
2 2 2
1á Qv CP |V H |2 (∆t)2 CP
|AH |(∆t)2 = QR = P v = 2 ¨ = 2 ¨
2 CD Gv CD Gv
Similarly, let =CP R = α,
 
1 á πab á 2ΓH
CP sin α|V H |∆t = ∆t ùñ |V H | =
2 T
loomoon CP sin α
:=ΓH

We get
á 2 2
á 2|V H | CP 2CP 1 4Γ2H
|AH | = 2 ¨ = ¨ 2 ¨ 2
CD Gv Gv CD CP sin2 α
Let ∆t Ñ 0, Gv Ñ CP ,
á CP 4π 2 a2 b2
|AH | = 2 2 ¨
CD CP sin2 α T2
Similarly, CD ¨ CP sin α = ab, hence
á 4π 2
|A| = 2 CP = ω02 CP
T
2π
where ω0 = is natural frequency.
T

2.2 Laplace-Lenz-Runge vector

The main goal of this section is to construct the vector which can help us realize where does the planet
now since we don’t have any coordinate in universe. In other words, Laplace-Lenz-Runge vector is GPS of
planet motion.
As figure 11, let P be any point on the ellipse, and S be the sun, H be another focus of the ellipse. Let
SR be perpendicular to the tangent line of P , and RW = RS. Similar for H, HY is perpendicular to the
tangent line of P and HY = 2HQ. Let SN K HY , i.e. SN ∥ QR.
Notation :

17
 á
SR = h1 , HQ = h2 , SN = ℓ, HS = 2aϵ, P S = r(t), e K is the unit vector from H to Y (i.e. perpendicular
á
to the tangent line of P ), e r (t) is the unit vector from S to P , Γ is the area swept per unit time.

W
P
N
R

H S

Figure 11

It is not hard to see SHY W is an Isosceles trapezoid, and HW = 2a. Hence

(h1 + h2 )2 + ℓ2 = (2a)2 (2)

Look at the △HN S,

(h2 ´ h1 )2 + ℓ2 = (2aϵ)2 (3)
Let (2)-(3), we get
4h1 h2 = (2a)2 ´ (2aϵ)2 = 4b2 ùñ h1 h2 = b2
Now since HY = 2h2 , SY = 2a,
á á á á
2h2 e K ´ 2a e r (t) = HS := 2aLLR

18
 á
where LLR is called the Laplace-Lenz-Runge vector.
As before,
á πab
| v (t)|h1 = 2 = 2Γ
T
Hence, we get

á
2πab á
b2 2 | v (t)| T | v (t)| ¨ 2Γ
h2 = =b ¨ =
h1 2πab 2πab 4π 2 a2
T T T2
á á
| v (t)|Γ ¨ 2a | v (t)|Γ ¨ 2a
= 2 3 =
4π a GMS
T 2

Combining with the previous result,

á
á | v (t)|2Γ á á
LLR = e K ´ e r (t)
GMS
Observe that á
á v (t) á
eK = á ˆ ez
| v (t)|
To sum up,
á á
á| v (t)|2Γ v (t) á á
LLR = á ˆ e z ´ e r (t)
GMS | v (t)|
á á
mP v (t) ˆ 2Γ e z á
= ´ e r (t)
GMS mP

2.3 Conservation of mechanical energy

Keep following Figure 11, since △Y QP „ △Y N S, we have

h1 + h2 2a
=
h1 r(t)
h2 2a
ùñ 1 + =
h1 r(t)

19
 á
b2 1 | v (t)|
Remember that h2 = and = . Hence
h1 h1 2Γ

b2 2a
1+ 2
=
h1 r(t)
á
b2 | v (t)|2 2a
ùñ 1 + =
4Γ2 r(t)
á
b2 | v (t)|2 2a
ùñ 1 + 2 2 2 =
π ab r(t)
4 2
T
á
1 | v (t)|2 1
ùñ + 2 3 =
2a 4π a r(t)
2¨ 2
T
á
1 mP | v (t)|2 1
ùñ + =
2a 2GMS mP r(t)
á
GMS mP mP | v (t)|2 GMS mP
ùñ + =
2a 2 r(t)
á
mP | v (t)|2 GMS mP ´GMS mP
ùñ E := ´ =
2 r(t) 2a
á
mP | v (t)|2 ´GMS mP
where E is called mechanical energy, is called Kinetic energy and is called
2 r(t)
potential energy.
Note that the right hand side of the last equation is independent of the location of the planet. Therefore,
the equation tells us that the mechanical energy is conservative.

2.4 Application of Laplace-Lenz-Runge vector

Consider α particle scattering experiment done by Rutherford in 1911.

20
 1

Z|e|
á
LLRα

Figure 12

á á
á Kc Z|e|2|e| á á mα v (t) ˆ 2Γ e z á
Fα = 2
e r (t) and LLRα = + e r (t)
r (t) Kc Z|e|2|e|
where Kc is the Coulomb’s constant.
á
Let b be the aiming error, and the v0 e x be the initial velocity of α particle. First, suppose that there is no
aiming error i.e. b = 0. Let R0 be the minimal distance between α particle and the atom. At that point,
the kinetic energy will turn into potential energy since the velocity is 0.

Kc Z|e|2|e| 1
= mα v02
R0 2
Now, suppose b ‰ 0, by definition, 2Γ = v0 b. Observe that if α particle is far away from the atom, then
á á
e r = ´ e x . Hence
á á
á mα v (o) ˆ 2Γ e z á
LLRα = ´ ex
Kc Z|e|2|e|
 
mα v02 b á á
= (´ e y ) ´ e x
Kc Z|e|2|e|
´2b á á
= ey ´ ex
R0
á
Now, let ϕ be the angle between the LLR the negative x direction, θ = π ´ 2ϕ. We can easily find that
θ 2b R0 θ
cot = tan ϕ = ùñ b = cot
2 R0 2 2

21
 Z|e|
b
ϕ
ϕ θ

Figure 13

Suppose there is a bigger aiming error b + db, then there will a deflection of angle dθ. Hence there is an
annulus on the ball of radius 1 after α particle scattering.

db
Z|e|

dθ
Figure 14

Denote I0 be the flow rate of α particle into the annulus of db per time and per area, and I(θ) be the flow
rate of α particle out the annulus of dθ per time and per area. The number of particles flow in must equal
to the numbers of particles flow out. So we have the following equation

I0 2πbdb = I(θ)2π sin θdθ

22
Then we get
I(θ) |2πbdb|
=
I0 |2π sin θdθ|
| R20 cot 2θ | ˇˇ db ˇˇ
ˇ ˇ
=
| sin θ| ˇ dθ ˇ
| R0 cot 2θ | R0 θ1 R2 1
= 2 csc2 = 0 4 θ 9Z 2
| sin θ| 2 22 16 sin 2
which we can determine Atomic Number. Also, in this experiment, Rutherford found that mass and the
positive electric charge of atom is concentrated in the small area which is called Atomic nucleus.

2.5 Properties of Laplace-Lenz-Runge vector

First, we recall some mathematical tools.

2.5.1 Inner product and outer product

á á á á á á á á á á
Given two vector A = Ax e x + Ay e y + Az e z and B = Bx e x + By e y + Bz e z , the inner product A ¨ B is
defined by
á á á á
A ¨ B = Ax Bx + Ay By + Az Bz = |A||B| cos θ
where θ is the angle between the two vector. Also know that
á á á
A ¨ A = |A|2 = A2x + A2y + A2z
á á
The outer product A ˆ B is defined by
á á á
ex ey ez
á á á á á
A ˆ B = (Ay Bz ´ Az By ) e x + (Az Bx ´ Ax Bz ) e y + (Ax By ´ Ay Bx ) e z = Ax Ay Az
Bx By Bz
á á á á
Now, for another vector C = Cx e x + Cy e y + Cx e z ,

Cx Cy Cz
á á á á á á
C ¨ (A ˆ B) = Ax Ay Az = Vol(C, A, B)
Bx By Bz

23
 á á á á á á á á á á á á á
Since if C = A, then C ¨ (A ˆ B) = 0, which implies that A K (A ˆ B). Similarly, if C = B, C ¨ (A ˆ B) = 0,
á á á á á á
which implies that B K (A ˆ B). So we get (A ˆ B) is perpendicular to the plane generated by the A and
á á á á á á
B. Now, let θ be the angle between A and B, and β be the angle between C and A ˆ B. It is easily to see
á á á á á
that |C| cos β is the height. And we find that |A ˆ B| = |A||B| sin θ.

2.5.2 Polar coordinate for planar motion

á
Let R(t) be the position at time t, write
á á á á
R(t) = x(t) e x + y(t) e y + z(t) e z

We define the velocity

á
á dR(t) dx(t) á dy(t) á dz(t) á
V (t) = = ex + ey + ez
dt dt dt dt
and the acceleration
á á
á dV (t) d2 R(t) d2 á d2 á d2 á
A(t) = = 2
= 2
x(t) e x + 2
y(t) e y + 2
z(t) e z
dt dt dt dt dt
á
Now, we introduce the polar coordinate of plane. Let r(t) be the length of R(t) and θ(t) be the angle
á
between R(t) and the positive x direction. We can easily see the following equations:
a
r(t) = x2 (t) + y 2 (t)
y(t)
θ(t) = arctan
x(t)
x(t) = r(t) cos θ(t)
y(y) = r(t) sin θ(t)

Then we can write velocity in the polar coordinate form.

á dx(t) á dy(t) á
V (t) = ex + ey
dt dt   
dr(t) dθ(t) á dr(t) dθ(t) á
= cos θ(t) ´ r(t) sin θ(t) ex + sin θ(t) + r(t) cos θ(t) ey
dt dt dt dt
dr(t)  á á
 dθ(t)  á á

= cos θ(t) e x + sin θ(t) e y + r(t) ´ sin θ(t) e x + cos θ(t) e y
dt dt
dr(t) á dθ(t) á
= e r (t) + r(t) e θ (t)
dt dt

24
 á á á á á á
where we define e r (t) = cos θ(t) e x + sin θ(t) e y and e θ (t) = ´ sin θ(t) e x + cos θ(t) e y .
á á á á á
Observe that the area swept per time equals to 12 R(t) ˆ V (t) = 21 r2 (t) dθ(t)
dt
e r (t) ˆ e θ (t) = 21 r2 (t) dθ(t)
dt
e z.
Now, we can derive a useful formula by calculation,
dá d á d á
e r (t) = cos θ(t) e x + sin θ(t) e y
dt dt dt
dθ(t) á dθ(t) á
=´ sin θ(t) e x + cos θ e y
dt dt
dθ(t)  á á
 dθ(t) á
= ´ sin θ(t) e x + cos θ(t) e y = e θ (t)
dt dt
dá dθ(t) á dθ(t) á dθ(t) á
e θ (t) = ´ cos θ(t) e x ´ sin θ(t) e y = ´ e r (t)
dt dt dt dt

2.5.3 Properties of Laplace-Lenz-Runge vector

á
dLLR
First, we would show that Laplace-Lenz-Runge vector is independent of the time i.e. = 0. This can
dt
be easily proves by direct calculation.
á á
d á d mP V (t) ˆ 2Γ e z dá
LLR = ´ ez
dt dt GMS mP dt
á á
( d mP V (t)) ˆ 2Γ e z dá
= dt ´ ez
GMS mP dt
á á
F SP ˆ 2Γ e z dá
= ´ e r (t)
GMS mP dt
1 á á
´GMS mP r2 (t) e r (t) ˆ r2 dθ(t)
dt
ez dá
= ´ e r (t)
GMs mP dt
dθ(t) á á dá
=´ e r (t) ˆ e z ´ e r (t)
dt dt
dθ(t) á dá
= e θ (t) ´ e r (t) = 0
dt dt
the last equation holds by the formula in the last part of 2.5.2.
Next, we show that if there exists Laplace-Lenz-Runge vector, the orbit of planet will be the conic section.

25
 á á
First, we consider the inner product of R(t) and LLR.
á á á
R(t) ¨ LLR = |LLR|r(t) cos θ(t)
á  á

R(t) ¨ mP ˆ 2Γ e z á á
= ´ R(t) ¨ e r (t)
GM m
á S áP  á
mP R(t) ˆ V (t) ¨ 2Γ e z mP 4Γ2
= ´ r(t) = ´ r(t)
GMS mP GMS mP
Hence  
á mP 4Γ2
r(t) 1 + |LLR| cos θ(t) = (4)
GMS mP
á
Note that the equation (4) is an equation for conic section where |LLR| = ϵ is an eccentricity, and
mP 4Γ2
= sϵ where s is the distance between focus and directrix.
GMS mP
The main goal now is that the length of Laplace-Lenz-Runge vector since it is the value of eccentricity
which determines which conic section is.
á á á
á á á á m2 |V (t)|2 4Γ2 á mP V (t) ˆ 2Γ e z
LLR ¨ LLR = e r (t) ¨ e r (t) + P 2
´ 2 e r (t) ¨
(GMS mP ) GMS mP
á  á á

á
m2 |V (t)|2 4Γ2 R(t) ¨ mP V (t) ˆ 2Γ e z
=1+ P ´ 2
(GMS mP )2 r(t)GMS mP
á á  á
á
2 2
mP |V (t)| 4Γ 2 m P R(t) ˆ V (t) ¨ 2Γ e z
=1+ ´2
(GMS mP )2 r(t)GMS mP
á
m2 |V (t)|2 4Γ2 mP 4Γ2
=1+ P ´ 2
(GMS mP )2 r(t)GMS mP
2
 
2mP 4Γ 1 á
2 GMS mP
=1+ mP |V (t)| ´
(GMS mP )2 2 r(t)
2
8mP Γ E
=1+ = ϵ2
(GMS mP )2

where E is the mechanical energy.

Exercise 1. Consider the following figure, it is a circle

26
 MS

R0

mP
á á
V (t) = V0 e x
Figure 15

We can easily compute the V0 be the centripetal force formula.

V02 GMS mP
mP =
R0 R02
GMS
V02 =
R
c0
GMS
V0 =
R0
Try to fill the blank! (All answer must be written in the form of R0 , GMS , α, β)

á á á á
2Γ e z = R(0) ˆ V (t) E= ϵ = |LLR| the direction the minimum
á á
mP 2 GMS mP
2Γ =? 2
|V (t)| ´ r(t)
ϵ =? of LLR rmin
á á
R(0) = ´R0 e y
á á
V (0) = V0 e x
á á
R(0) = ´R0 e y
á á
V (0) = V0 (1 + α) e x
á á
R(0) = ´R0 e y
á á á
V (0) = V0 e x + βV0 e y

27
2.6 Aspect from calculus

We now will use the calculus to explain that we have just proved in 2.1. We keep assume that the Kepler’s
law and Newton’s law are true.
By Kepler’s second law and Newton’s second law

d á  á á á
á dV (t) dR(t)
0= R(t) ˆ V (t) = R(t) ˆ +
dt dt dt
á á á á
= R(t) ˆ A(t) + V (t) ˆ V (t)
á
á F S,P
= R(t) ˆ
mP
á á á á á á á á
Hence R(t) ∥ F S,P . Also we know R(t) ˆ V (t) is a fixed vector, and R(t) K (V (t) ˆ V (t)), V (t) K
á á
(R(t) ˆ V (t)), hence it is a planar motion.
Now, follow the notion in the 2.5.2. We have already known
á dr(t) á dθ(t) á
V (t) = e r (t) + r(t) e θ (t)
dt dt
á
Now, we compute acceleration A(t). Before we compute, we recall that
$
&dá
’
e r (t) =
dθ(t) á
e θ (t)
dt dt
%dá
’
e θ (t) = ´
dθ(t) á
e r (t)
dt dt
Then
 
á d dr(t) á dθ(t) á
A(t) = e r (t) + r(t) e θ (t)
dt dt dt
 2
d2 á dr(t) dθ(t) á dr(t) dθ(t) á d2 á dθ(t) á
= 2 e r (t) + e θ (t) + e θ (t) + r(t) 2 θ(t) e θ (t) ´ r(t) e r (t)
dt dt dt dt dt dt dt
 2 !  
d2 dθ(t) á dr(t) dθ(t) d2 á
= 2
r(t) ´ r(t) e r (t) + 2 + r(t) 2 θ(t) e θ (t)
dt dt dt dt dt
 2 !  
d2 dθ(t) á 2 dr(t) dθ(t) 1 2 d2 á
= 2
r(t) ´ r(t) e r (t) + r(t) + r (t) 2 θ(t) e θ (t)
dt dt r(t) dt dt r(t) dt
 2 !   
d2 dθ(t) á 1 d 2 dθ(t) á
= 2
r(t) ´ r(t) e r (t) + r e θ (t)
dt dt r(t) dt dt

28
  
d dθ(t)
2
Note that by Kepler’s second law, we have r (t) = 0. Hence
dt dt
 2 !
á d2 dθ(t) á
A(t) = 2
r(t) ´ r(t) e r (t) (5)
dt dt

Now, we compute the coefficient explicitly.

Remember that r(t) satisfies the
sϵ
r(t) = (6)
1 + ϵ cos θ(t)
Also, Kepler’s second law implies that
dθ(t) dθ(t) 2Γ
2Γ = r2 (t) ùñ = 2
dt dt r (t)
Hence,
 
d2 d dr(t)
r(t) =
dt2 dt dt
1
!
d dr(t) d r(t) dθ(t)
= 1
dt d r(t) dθ(t) dt
 
d 2 1 ´ ϵ sin θ(t) 2Γ
= ´r (t) ¨ ¨ 2
dt sϵ r (t)
 
d ´2Γ ¨ (1 ´ ϵ sin θ(t))
=
dt sϵ
2Γϵ cos θ(t) 2Γ 4Γ2 ϵ cos θ(t)
= ¨ 2 =
sϵ r (t) sϵr2 (t)
Combining with (5), we get
 
á 4Γ2 ϵ cos θ(t) 4Γ2 á
A(t) = ´ r(t) ¨ 4 e r (t)
sϵr2 (t) r (t)
 
sϵ
´4Γ2 ´ϵ cos θ(t) + r(t) á ´4Γ2 1 á
= 2
e r (t) = ¨ 2 e r (t)
sϵr (t) sϵ r (t)
Hence, we get that the acceleration is inversely proportional to the square of distance.
sϵ
Now, look at equation (6), we now that as θ = 0, r(t) has minimum rmin = and as θ = π, r(t) has
1+ϵ
sϵ
maximum rmax = . Then
1´ϵ
sϵ sϵ 2sϵ
$
&2a = rmin + rmax =
’ + =
1+ϵ 1´ϵ 1 ´ ϵ2
2
%2c = rmax ´ rmin = sϵ ´ sϵ = 2sϵ = 2aϵ
’
1´ϵ 1+ϵ 1 ´ ϵ2

29
where a is semi-long axis and c is the distance between center and a focus. Now denote b as a semi-short
2πab
axis, and we have b2 = a2 ´ c2 = a2 (1 ´ ϵ2 ). On the other hand, 2Γ = . Hence
T
4Γ2 4π 2 a2 b2 4π 2 a4 (1 ´ ϵ2 ) 4π 2 a3
= 2 = =
sϵ T a(1 ´ ϵ2 ) T 2 (1 ´ ϵ2 ) T2

Finally, we get
á 4π 2 a3 1 á
A(t) = ´ e r (t)
T 2 r2 (t)
4π 2 a3
Similar to 2.1, by Kepler’s third law and Newton’s second and third law, we know that is a constant
T2
related to MS . Hence we get Newton’s Universal Gravitation formula
á ´GMS mP á
F S,P = e r (t)
r2 (t)

All of above are compatible to 2.1.

Now, we consider the question conversely. That is, if the force is inversely proportional to the square of
the distance, then what the orbit looks like? By the equation (5) and Newton’s second law,
á C á
F = e r (t)
r2 (t)
á
= mA(t)
2 !
d2 dθ(t) á
=m r(t) ´ r(t) e r (t)
dt2 dt
 2 
d 4Γ2 á
=m r(t) ´ 3 e r (t)
dt2 r (t)
C d2 4Γ2
ùñ = r(t) ´
mr2 (t) dt2 r3 (t)

1 dr dr du dθ du 2Γ du d2 r d2 u dθ ´4Γ2 d2 u
We now let u = , then = = ´r2 = ´2Γ , and = ´2Γ = =
r dt du dθ dt dθ r2 dθ dt2 dθ2 dt r2 dθ2
d2 u
´4Γ2 u2 2 .
dθ
Hence we get a differential equation
C d2 u
u+ = ´
4mΓ2 dθ2
C
Let W = u + , then W satisfies
4mΓ2
d2 W
W =´ 2
dθ

30
 C
Hence if we specified the initial value, then we will get W = B cos θ = u + .
4mΓ2
Therefore,
1 C
= u = B cos θ ´
r(t) 4mΓ2

C 1 1
We now let ´ 2
= and B = , we will find that r(t) satisfies the formula which holds for conic
4mΓ sϵ s
section.

Remark 1. The proof in 2.1 needs the assumption that law of inertia, but in the method of differentiation,
C
the constant 4mΓ 2 is disappear after taking derivative. Hence the proof in this section, it seems that we

does not need the law of inertia. Compare the two proofs : Newton’s method needs complicated analysis
about the geometry, and the calculus method needs solve differential equation. Laplace-Lenz-Runge vector
gives a simpler explain for this result. Recall that Laplace-Lenz-Runge vector holds for the force which is
inversely proportional to r2 . If C ă 0,
á á
á mP V (t) ˆ 2Γ e z á
LLR = ´ e r (t)
´C
If C ą 0,
á á
á mV (t) ˆ 2Γ e z á
LLR = + e r (t)
C
Then á á á
á á á m(R(t) ˆ V (t)) ¨ 2Γ e z
R(t) ¨ LLR = r(t)|LLR| cos θ(t) = ´ r(t)
C
Hence
4mΓ2
r(t) = C
á
1 + |LLR| cos θ
á
1 C C|LLR|
ùñ = + cos θ
r(t) 4mΓ2 4mΓ2

C 1 á
where = and | LLR| = ϵ.
4mΓ2 sϵ

We now consider the Hook problem

31
 P
(x(t), y(t))
ℓ
á á
x(t) e x + y(t) e y
C
P
C mg

Figure 16

As we compute before,
á ´4π 2 á
A(t) = CP
T2
Hence we have  
x(t) á y(t) á m4π 2 á m4π 2 á á
mg e x, ey =´ 2
CP = ´ 2
(x(t) e x , y(t) e y )
ℓ ℓ T T
c
ℓ
It implies that T = 2π.
g  mg 
á mg
Now we consider conversely, given A is ´ x(t), ´ y(t) . By Newton’s second law
ℓ ℓ
á á d2 á d2 á
F = mA = m 2 x(t) e x + m 2 y(t) e y
dt dt
Hence we get the differential equation
2 2
$ $
&´ mg x(t) = m d x(t)
’ &´ g x(t) = d x(t) = ´ω02 x(t)
’
ℓ 2
dt2 ùñ ℓ 2
dt2
mg d y(t) g
%´ y(t) = y(t) = ´ω 2 y(t)
d
%´ y(t) = m
’ ’
0
ℓ dt2 ℓ dt2
g
where ω02 = . Similarly, if we specified the initial condition, we can get the solution
ℓ
$
&x(t) = C cos ω t
0
%y(t) = D sin ω0 t

x2 (t) y 2 (t)
We implies that + = 1, and hence the orbit is an ellipse. On the other hand
C2 D2
á á á
ex ey ez
á á á
R(t) ˆ V (t) = C cos ω0 t D sin ω0 t 0 = ω0 CD e z
´ω0 C sin ω0 t ω0 D cos ω0 t 0

32
 á á 2πCD
and 2Γ = |R(t) ˆ V (t)| = ω0 CD. But by the definition of Γ, 2Γ = , by comparing these, we get
T
g 2π
ω0 = = which is also compatible with before.
ℓ T

2.7 Conservation laws

2.7.1 Conservation of Mechanical Energy

1 á GMS mP
Recall from 2.3, we know E = mP |V (t)|2 ´ is a constant independent to the position in the
2 r(t)
orbit. We now use another aspect to see this.

á
Definition 1. We define a work W done by F from A to B by
ż á á
W = F ¨ dℓ
γ

where γ is the path from A and B.

á
T
B
á
H
A

á
FG
Figure 17

Consider a pendulum which we pull the bob horizontal, then we get

á á á
H + T + FG = 0

33
 á á á
Note that T is perpendicular to d ℓ , hence the work done by T equals to 0. So,
ż á á ż á á
H ¨ dℓ + FG ¨ dℓ = 0
γ γ

á á á á á á á
But we can calculate the work done by F G easily. F G = ´mg e z and d ℓ = dx e x + dy e y + dz e z . Hence
ż á á
ż zB
FG ¨ dℓ = ´mg dz = ´mg(zB ´ zA )
γ zA

and ż ż
á á á á
H ¨ dℓ = ´ F G ¨ d ℓ = mg(zB ´ zA )
γ γ

Consider another problem,

m
á
F = ´kx

Figure 18

By definition, ż xB
´1 2 1
W = ´kx dx = kxB + kx2A =: U (xA ) ´ U (xB )
xA 2 2
1
where U (x) is the potential energy at x. We define U (0) = 0, hence U (x) = kx2 .
2
But notice that NOT all the force can define the potential energy. The force which can define potential
energy is conservative force, where define
ż á á
U (A) ´ U (B) = F ¨ d ℓ (7)
γ

where γ is the path from A to B.

We now consider the revolution of a planet, first consider if A, B and sun are on the same line.
żB
´GMS mP GMS mP GMS mP
U (A) ´ U (B) = 2
dr = ´
A r rB rA

34
We usually define U (B) = 0 as rB Ñ 8, hence
GMS mP
U (rA ) = ´
rA
For A, B and sun, just consider the following

B1
Figure 19

since they are circle the work from A to B is equal to A to B 1 .

On the other hand, we figure out that the meaning of work.
ż á á ż á á
F ¨ d ℓ = F ¨ dR(t)
γ γ
ż á
dV (t) á
= m ¨ dR(t)
γ dt
ż tB á
dV (t) á
= m ¨ V (t) dt
tA dt
d m á 
ż tB á m á m á
= V (t) ¨ V (t) dt = |V (tB )|2 ´ |V (tA )|2
tA dt 2 2 2
Put this into (7), then we will get the conservation of mechanical energy.

2.7.2 Application of the conservation of mechanical energy

We can use the conservation law to understand x(t). (Note that if we use Newton’s second law, we need
to solve the second order differential equation.)

35
Consider the Hooke’sproblem,
2 i.e. Figure 18.
1 dx(t) 1
We know E = m + Kx2 (t) where K is the characteristic of the spring. Hence we have
2 dt 2
following computation,
1 dx(t)
1= b

2
m
E ´ 21 Kx2 (t) dt
Integral both side,
żt żt
1 1 dx(t1 ) 1
t= 1 dt = b
1
dt
0 0 2
m
E ´ 12 Kx2 (t1 ) dt
ż x(t)
dx
= a
x(0) A2 ´ ω02
ż x(t)
1 1
= b dx
A x(0) ω02 2
1 ´ A2 x
ω0
ż x(t)
A 1 1
= ?
2ω
du
ω0
A
x(0) 1 ´ u 0
ˇ ω0 x(t)
1 ˇA
= arcsin uˇˇ
ω0 ω0
x(0) A

2E 2 K
where A2 = , ω0 = . Hence we have
m m
ω0
ω0 t = arcsin x(t) ´ α
A
for some angle α. Therefore, we get that

A dx(t)
sin(ω0 t + α) = x(t) ùñ A cos(ω0 t + α) =
ω0 dt
Now, we consider the revolution of a planet.
At rmax and rmin ,  2
mP 2 dθ(t) GMS mP mP 4Γ2 GMS mP
E= r (t) ´ = ´
2 dt r(t) 2 r2 (t) r(t)
implies that
Er2 (t) + GMS mP r(t) ´ 2mP Γ2 = 0
We get that a
´GMS mP ˘ (GMS mP )2 + 8mΓ2 E
r˘ =
2E

36
Hence
´GMS mP ´GMS mP
2a = r+ + r´ = ùñ E = ă0
E 2a
so r´ ą r+ , and
a d
2 2
´ (GMS mP ) + 8mΓ E 1 + 8mP Γ2 E á
r´ ´ r+ = 2aϵ = ùñ ϵ = = | LLR|
E (GMS mP )2

On the other hand,

 2  2 !
mP dr(t) 2 dθ(t) GMS mP
E= + r (t) ´
2 dt dt r(t)
 2 !
mP dr(t) 4Γ2 GMS mP
= + 2 ´
2 dt r (t) r(t)

Recall that
dr(t) dr dθ(t) dr 2Γ
= =
dt dθ dt dθ r(t)

Combining two equation above, we get

1 dr 2Γ
c   =1
2 GMS mP 2mP Γ 2 dθ r2
mp
E+ r(t)
´ r2 (t)

Integral over θ both side,

ż
1 2Γ
c   dr = θ(t) ´ θ(0)
2 GMS mP 2mP Γ2 r2
mP
E+ r
´ r2

1
Let u = ,
r
ż ż
´2Γ ´1
b du = b du
2 2E 2GMS u
mP
(E + GMS mP u ´ 2mP Γ2 u2 ) 4mP Γ2
+ 4Γ2
´ u2
ż
´1
= c 2 du
2E GMS

2  
GMS mP
4mP Γ2
+ 4Γ2
´ u´ 4mP Γ2

37
 2E


GMS 2 GMS mP
Let B 2 = 4mP Γ2
+ 4Γ2 , and W = u ´ 4mP Γ2
. Then
ż  
´1 W
θ(t) ´ θ(0) = ? dW = arccos
B2 ´ W 2 B
Hence
GMS mP
!
u´ 4mP Γ2
θ(t) ´ θ(0) = arccos ´β
B
for some β. Let ϕ(t) = θ(t) ´ θ(0) ´ β. We get
1 GMS mP GMS mP 1
B cos ϕ(t) = ´ 2
ùñ 2
+ B cos ϕ(t) =
r(t) 4mP Γ 4mP Γ r(t)
d  2
2E GMS mP 1 GMS mP 1
where B = + = and = . Therefore, we get
4mP Γ2 4mP Γ 2 s 4mP Γ 2 sϵ

2
d  2 d
sϵ 4mP Γ 2E GMS mP 8EmP Γ2 á
ϵ= = + = 1 + = | LLR|
s GMS mP 4mP Γ2 4mP Γ2 (GMS mP )2

2.7.3 Conservation of momentum

á á
Definition 2. The momentum is defined to be P = mV .

Consider the two particle collision.

m1
m2
á
á V1
V2
ùñ

m2 m1

á1 á1
V2 V1
Figure 20

38
We consider the center of mass. The velocity of the center of mass is
á á
á m1 V 1 + m2 V 2
V cm =
m1 + m2
And then $ á á
’ á á á m (V ´ V 2)
& u 1 = V 1 ´ V cm = 2 1
’
’
má 1 + m2
á
á á
and m1 u 1 + m2 u 2 = 0
’
’ á á á m1 (V 2 ´ V 1 )
% u 2 = V 2 ´ V cm =
’
m1 + m2
After collision, we can get similar,
$ á á
1 1
’ á á á m ( V ´ V 2)
&u1 1 = V 1 1 ´ V cm = 2 1
’
’
á á
má1 + má 2 and m1 u1 1 + m2 u1 2 = 0
1 1
’ á á á m ( V ´ V 1)
%u1 2 = V 1 2 ´ V cm = 1 2
’
’
m1 + m2
We look this at the system of the center of mass,
ùñ

Figure 21

We compute the kinetic energy before and after collision.

1 á á 2
 1 á 
á 2
E = m1 V cm + u 1 + m2 V cm + u 2
2 2
á á á á
1 á 1 m22 |V 1 ´ V 2 |2 1 m21 |V 2 ´ V 1 |2
= (m1 + m2 )|V cm |2 + m1 + m2
2 2 (m1 + m2 )2 2 (m1 + m2 )2
1 á 1 m1 m2 á á
= (m1 + m2 )|V cm |2 + |V 2 ´ V 1 |2
2 2 m1 + m2

39
 m1 m2
We called reduced mass. And similar for kinetic energy after collision.
m1 + m2
1 á 1 m1 m2 á1 á
E 1 = (m1 + m2 )|V cm |2 + |V 2 ´ V 1 1 |2
2 2 m1 + m2
á á
|V 1 2 ´ V 1 1 |
Hence the kinetic energy will be decreased or not is depend on the value e = á á , and this is defined
|V 2 ´ V 1 |
to be coefficient of restitution.
We called the collision is elastic collision if e = 1 and inelastic collision if e ‰ 1.
Example 1. We consider a ball fall down from height h, and bounce to height h1 . At the moment that
ball touch the floor, all of the potential energy transforms into the kinetic energy. And after bounce to the
height h1 , all of the kinetic energy transform to the potential energy. Hence we can compute the velocity
before and after bounce, $
& 1 mV 2 = mgh c
2 |V 1 | h1
ùñ e = =
% 1 mV 12 = mgh1 |V | h
2

where we assume the velocity of Earth is 0 since its mass is far larger than a ball.

á
We now consider a simple case that V 2 = 0, and e = 1. Hence we have
$ á
’ á m 1 V 1
’
’ V cm =
m + +m
’
’
’ 1 2
’
’ á
’
&á
’ m1 V 1
u1 =
m1 + á m2
’
’
’ á ´m1 V 1
u2 =
’
’
má1 + m2
’
’
’
% á1 á
’
’
|V 2 ´ V 1 1 | = |V 1 |
First case :
á
V 11

m1

m2 á
V 12

Figure 22

40
Similarly, we see in the system of the center of mass.

á á
u1 u2

ùñ
á
u1 1

á
u1 2

Figure 23

á á á á
Let v 1 1 ˚ V 1 2 = |V1 | n. Then
$ á á á á
1 1
’ á á á m V
1 1 m2 ( V 1 ´ V 2 ) m m V
1 1 1 m1 m2 á á
&P 1 1 = m1 (V cm + u1 1 ) = m1
’
’ + m1 = + |V 1 | n
m1 á
+ m2 m1 + m2 m1 + m2 m1 + m2
’ á á á mmV m1 m2 á á
%P 1 2 = m2 (V cm + u1 2 ) = 2 1 1 ´
’
’ |V 1 | n
m1 + m2 m1 + m2
á á á
á m1 m1 V 1 á m2 m1 V 1 á m1 m2 á á |AO| m1
Let AO = , OB = , and OC = |V 1 | n. Note that á = .
m1 + m2 m1 + m2 m1 + m2 |OB| m2
Then if m1 ą m2 ,

á
C
1
P 1 á
P 12

A O B

Figure 24

41
 á
And note that |P 1 2 | attains maximum when AC is the tangent line, and
á
á
1 2m1 m2 V 1
P 2,max =
m1 + m2
Similarly for m1 ă m2 ,

A O B
á
P 11
C

Figure 25

This is why Rutherford found that there are some α particle will bounce back.

Remark 2 (The discovery of neutron). Neutron is found in the experiment by doing particles collision
and calculate the velocity after collision. It turns out that the mass of the neutron is about 1.15 times the
mass of the Hydrogen ion.

2.7.4 Conservation of angular momentum

Definition 3. Given an object P with mass m, then the angular momentum with respect to O is
á á á
L = mR(t) ˆ V t
á
where R(t) is the vector from O to P at time t.

We now consider a pendulum.

42
 á
e θ (t)

á
e r (t)

θ(t)
ℓ

T
á
V (t)
á
mg

Figure 26

Note that r(t) = ℓ is a constant.

á á á á dr(t) á dr(t) á
R(t) = r(t) e r (t) = ℓ e r (t), and V (t) = e r (t) + ℓ e θ (t). So,
dt dt
á á á dθ(t)
L(t) = mR(t) ˆ V (t) = mℓ2
dt
Recall from here, we have
"  2 #  
á d2 r(t) dθ(t) á 1 d 2 dθ(t) á
mA(t) = m ´ r(t) e r (t) + m r (t) e θ (t)
dt2 dt r(t) dt dt

Hence we have
 2
á dθ(t) á mℓ2 d2 θ(t) á
mA(t) = ´mℓ e r (t) + e θ (t)
dt ℓ dt2
á á
= (mg cos θ(t) ´ T ) e r (t) ´ mg sin θ(t) e θ (t)

So we have to solve the equation

mℓ2 d2 θ(t)
= ´mg sin θ(t)
ℓ dt2
d2 θ(t)
ñmℓ2 = ´mgℓ sin θ(t)
dt2

43
 dθ(t) d2 θ(t)
Here, we call angular velocity and angular acceleration.
dt dt2
Now note that

d á á á dV (t) á
á
 á á á á á
m(R(t) ˆ V (t)) = mR(t) ˆ = R(t) ˆ m g + T = R(t) ˆ m g =: τ
dt dt
á
where τ is called torque. (The equation means the rate of change of the angular momentum is equal to
the torque)
Now we consider the multiple mass point problem : Suppose there are n mass point with mass mi in the
system. Then the center of mass is ÿ á
mi r i (t)
á
i
Rcm (t) = ÿ
mi
i

and hance we get

! ! á
d ÿ á d ÿ á ÿ dRcm (t) ÿ á
mi Rcm (t) = mi r i (t) ùñ mi = mi V i (t)
dt i
dt i i
dt i

which says that the sum of momentum is equal to the momentum at the center of the mass. And also,
! á " ! #
2
ÿ d Rcm (t) ÿ d á ÿ ÿ á á ÿá
mi = m i V i (t) = f j,i + F i,ext = F i,ext
i
dt2 i
dt i j‰i i

á
where F i,ext is denoted the exterior force acted on the i-th mass point, and the last equality holds by the
Newton’s third law. (This equation means that the all force acted on any mass point can be viewed as the
sum of the force acted on the center of mass.)
On the other hand,
! á !
d ÿ á á ÿ á dV i (t)
mi r i (t) ˆ V i (t) = mi r i (t) ˆ
dt i i
dt
" ! #
ÿá ÿá á
= r i (t) ˆ f j,i + F i,ext
i j‰i
ÿá á
= r i (t) ˆ F i,ext
i

á á á á á
the last equation holds since f i,j = ´ f j,i and r i (t) ´ r j (t) ∥ f j,i . (The equation means the sum of the
angular momentum can be viewed as the angular momentum at the center of mass.)

44
 á á
á á á d rri (t) dRcm (t) á
Now, let rri (t) + Rcm (t) = r i (t), and + = V i (t). Hence
dt dt
 á 
ÿ á  á
á á d rri (t) dRcm (t) 
mi rri (t) + Rcm (t) ˆ 
ÿ á
mi r i (t) ˆ V i (t) = +
i i
dt dt
á á
ÿ á d rri (t) ÿ á dRcm (t)
= mi rri (t) ˆ + mi Rcm (t) ˆ
i
dt i
dt
á á
ÿ á dRcm (t) á ÿ d rri (t)
+ mi rri (t) ˆ + Rcm (t) ˆ mi
i
dt i
dt
looooooooooooomooooooooooooon looooooooooooomooooooooooooon
=0 =0

á á
á d rri (t) á dRcm (t)
We call i mi rri (t) ˆ spin angular momentum and i mi Rcm (t) ˆ orbital angular
ř ř
dt dt
momentum. By before,
 á á

d  ÿ á d rri (t) ÿ á dRcm (t)  ÿ á á
mi rri (t) ˆ + mi Rcm (t) ˆ = r i (t) ˆ F i,ext
dt i
dt i
dt i
!
ÿá á á ÿá
= rri (t) ˆ F i,ext + Rcm ˆ F i,ext
i i

and notice that á ! !

d ÿ á dRcm (t) á ÿá
mi Rcm (t) ˆ = Rcm ˆ F i,ext
dt i
dt i

Hence we get the nontrivial result,

 á

d  ÿ á d rri (t)  ÿ á á
mi rri (t) ˆ = rri (t) ˆ F i,ext
dt i
dt i

which means that the spin angular momentum is the angular momentum caused by exterior force action.

Example 2. The diver spin in the air, but consider the exterior force (only gravitation) acted on the diver
!
ÿá á ÿ á á
r i (t) ˆ mi g = mi rri (t) ˆ g = 0
i i

which means that the diver just need to focus on himself.

45
 á
Next, we calculate the angular momentum of rigid body. From now on, we denote r i (t) is respect to the
center of mass (without tilde) for convenience.
Consider a board rotate about the axis passing through the center of mass.

á a
n 2

á
αi r i (t)
b
CM

Figure 27

á dθ(t) á á dθ(t)
Let ω = n, where |w| = . Then
dt dt
ˇ á ˇ á
ˇ d r (t) ˇ dθ(t) d r i (t) á á
ˇ i ˇ á á á
= | r i (t)| sin α i = | r i (t)| sin α i | ω| ùñ = ω ˆ r i (t)
ˇ dt ˇ dt dt
ˇ ˇ

Then
á ÿ á
á
d r i (t) ÿ á á 
L= mi r i (t) ˆ = mi V ri (t) ˆ ω ˆ r i (t)
i
dt i

Note that
!
á
ω ÿ á
há á i 1 ÿ á á  á á 
á ¨ mi r i (t) ˆ ω ˆ r i (t) = á mi ω ˆ r i (t) ¨ ω ˆ r i (t)
| ω| i | ω| i
á
!2
1 ÿ d r i (t)
= á mi
| ω| i dt
1 ÿ á á ÿ á
= á mi | r i (t)|2 sin2 αi |ω|2 = | ω| mi | r i (t)|2 sin2 αi
| ω| i i
looooooooooomooooooooooon
=:I

46
where I is called moment of inertia. Hence we get
á á
L = Iω

Note that this formula only holds for the object is symmetry over the rotation.
We now calculate the moment of inertia I,
a
á 2
ω á a
ω

x
a
d= 2
b b

dx
M M

Figure 28

For the moment of inertia of left hand side,

a
a2 M
ż
2 M
I= x2 b dx =
´ a2 ab 12

But for right hand side, ża

M a2 M
I= x2 dx =
0 a 3
Observe that
a2 M a2 M  a  2 a2 M
= + M= + d2 M
3 12 2 12
So we have the following

Theorem 1 (parallel axis theorem). Let Icm be the moment of inertia of the object of mass rotated about
the axis passing through the center of mass. Then the moment of inertia Id rotated about the axis which
is parallel to the previous one and has distance d is Icm + md2 .

47
Proof. Look at the following figure.

Icm Id

d
(xd , yd )
(0, 0)

Figure 29

ÿ  
Id = mi (xi ´ xd )2 + (yi ´ yd )2
i
 
mi x2i + mi x2d + mi yi2 + mi yd2 ´ looomooon 
ÿ
= 2mi xi xd ´ 2m i yi yd
looomooon
i =0 =0
ÿ
ÿ
= mi x2i + yi +
2
(mi )(x2d + yd2 ) = Icm + md2
i i

Now, we calculate the moment of inertia of cylinder.

dr r

M h

Figure 30

48
 ża
2πr M 2
I= r2 M 2
h dr = a
0 πa h 2
which is independent of the height h. Hence if a coin also has mass M and diameter 2a, the moment of
M 2
inertia is still equal to a.
2

Iy

x
y
Ix

Figure 31

M 2
We know that Iz = a and
2 ż
Iz = (x2 + y 2 ) dm = Ix + Iy

where ż ż
2
Ix = x dm and Iy = y 2 dm

M 2
Hence we know Ix = Iy = a.
4
Now we calculate the inertia of sphere. At first, we calculate for the spherical shell.

dm (x, y, z)
r

Figure 32

49
where x2 + y 2 + z 2 = r2 .
ż
Ix = y 2 + z 2 dm
ż
Iy = x2 + z 2 dm
ż
Iz = x2 + y 2 dm

and
ż
Ix + Iy + Iz 2 2
Ix = Iy = Iz = = x2 + y 2 + z 2 dm = r2 m
3 3 3
Therefore, for spherical with radius a and mass M ,
ża
2M a 4 2a2
ż
2 2 2 M
Iz = I = r 4πr 4π 3 dr = 3 r dr = M
0 3 3
a a 0 5

Exercise 2.

a VA = V0

ωA
ñ Vcm

ωA+B
VB = 0

Figure 33

(I) VA = V0 , VB = 0 and ωA = 0. After collision, and stick together, the angular velocity ωA+B . Compute
ωA+B .

(II) VA = V0 , VB = 0 and ωA = ω0 ‰ 0. Compute ωA+B .

Exercise 3.

50
 L
V0 4 m

m L
4
ñ ω

M M
Figure 34

A stick with mass M length L. A mass point with mass m with velocity V0 hit the stick perpendicular at
L
the point which is distant from the center of the stick. Compute the angular velocity ω after collision
4
and stick together.

Now we consider the ball rotating

Vcm (0) = 0

M R
ω(0) ‰ 0
fk = M gµk

Vc (t) = Vcm (t) ´ Rω(t)

Figure 35

We have $
&M dVcm (t) = M gµk
$
’ &Vcm (t) = Vcm (0) + µk gt
dt ùñ
’
% d 2
M R2 ω(t) = ´M gµk R %Rω(t) = Rω(0) ´ 5 µk gt
dt 5 2
Hence
7 7 7
Vc (t) = Vcm (t) ´ Rω(t) = (Vcm (0) ´ Rω(0)) + µk gt = Vc (0) + µk gt = ´Rω(0) + µk gt
2 2 2

51
Then there exists t˚ such that Vc (t˚ ) = 0 where
2
t˚ = Rω(0)
7µk g
This is the follow shot of billiard, where the ball be accelerated by friction force after collision with another
ball. Similarly, we can do the converse where the ball will go backwards after collision.
Now we consider another classic problem, we assume rolling without sliding

ω(t)
Vcm (t)

fs ℓ

mg

Figure 36

We have the following equation

$
d
M Vcm (t) = M g sin θ ´ fs
’
& 2 M dVcm (t) = f
’ $
& dt
’
’
d2 ùñ 5 dt
s
M R2 ω(t) = Rfs
’
’
’ dt 5 %
Vcm (t) = Rω(t)
%Vc (t) = Vcm (t) ´ Rω(t) = 0
’

No sliding implies  
2 dVcm (t) dVcm (t) 5
M 1+ = M g sin θ ùñ = g sin θ
5 dt dt 7
We can also do this problem by the conservation of the mechanical energy,
1 2 12
E = M g sin θℓ(t) + M Vcm (t) + M R2 ω 2 (t)
2 25
where ℓ(t) is the distance between the bottom and the ball at time t. Differentiate over t, get
dℓ(t) dVcm (t) 2 ω(t)
M g sin θ + M Vcm (t) + M R2 ω(t) =0
dt dt 5 dt
Recall that
dℓ(t)
= ´Vcm (t) and Vcm (t) = Rω(t)
dt

52
Hence
7 dVcm (t) dVcm (t) 5
M g sin θ = M ùñ = g sin θ
5 dt dt 7
The result tells that there is not all energy distributed to the kinetic energy, i.e. there is some energy
changed into the angular kinetic energy. And there is no work done by friction force, which is bizard.

3 Thermodynamics

3.1 Introduction

Heat and temperature is the key we focus now.

• 1593 Galileo invented gas thermometer.

• 1643 Torricelli invented barometer.

• 1654 Guericke invented aspirator.

• 1662 Boyle found Boyle’s law - For a given amount of gas, at same temperature, the pressure times
volume is a constant P V = P0 V0 .

• 1783 Charles invented hot air balloon.

• 1802 Gay-Lussac’s law - For a given pressure, the volume and temperature has the equation Vt =
1
V0 (1 + βt) and β = 273.15 . Note that β is independent of the gas.

• 1808 Dulong-Petit found for a given volume, the volume and pressure has the equation Pt = P0 (1+βt).

By the results found by Dulong-Petit and Gay-Lussac’s law, we have

 
1
Pt Vt = P0 V0 (1 + βt) = P0 V0 β t + = P0 V0 β(t + 273.15) = P0 V0 βT
β

where T is the kelvin (the base unit of temperature).

53
 • 1800 Law of definite proportions - a given chemical compound always contains its component elements
in fixed ratio.

We call the quantity of a gram of hydrogen atom a mole. (The definition of a mole now is 12 grams of
carbon atom.)

• Avogadro’s hypothesis - equal volumes of all gases, at the same temperature and pressure, have the
same number of molecules N .

Combining above law’s,

PV
9 numbers (N, n)
T
where N is the numbers of molecules and n is the moles of molecules. We write
PV PV
= nR or = Nk
T T
and we called R ideal gas constant and k Boltzmann constant.

• 1906 J.Perrin measured the Avogadro constant NA = 6.02 ˆ 1023 , and NA k = R by definition.

3.1.1 Constant volume gas thermometer

For fixed pressure, the relation between temperature and volume is independent of the gas. Hence it is a
good way to measure the temperature.

54
 PT

V0

Figure 37

We have
Nk
PT V0 = N kT ùñ PT = T
V0
If the amount of gas and the volume is same, then the temperature have same meaning. Now we will talk
about the meaning of the temperature through kinetic theory of gases.

3.2 Kinetic theory of gases

Figure 38

The pressure is caused by the atom of gas collide against the wall.

55
 v

vx A
vx ∆t
Figure 39

We know
á á
∆(m v ) = ´2mvx e x
We denote nvx be the molecule density of the velocity of x-direction is vx . Then
1 ÿ ÿ
2
ÿ
P = 2mv x Av x ∆tn v = 2mv n
x vx = (mvx2 )nvx
A∆t V ą0 loooooooomoooooooonx V ą0 all v
x x x
Total momentum

We assume that the velocity of the every direction is symmetric, hence the last equality holds. Now, we
let Nvx = nvx ¨ V which means that the number of molecules that have x-direction velocity vx . Then
ÿ Nvx
P = mvx2
all vx
V
 ÿ 
N 1 2 N
= mvx Nvx = mvx2
V N V
ÿ
where N = Nvx is the number of molecule, and mvx2 is the average of kinetic energy in the x-direction.
all vx
Hence we have
P V = N mvx2 = N kT
By symmetry,
1 2m 2 2
mvx2 = mvy2 = mvz2 = m(vx2 + vy2 + vz2 ) = (vx + vy2 + vz2 ) = K.E.
3 32 3
where K.E. is the average of kinetic energy. Then
3
K.E. = kT
2

56
 V
mA

vA
vB

mB

Figure 40

$
&P V = N kT
A A law of partial pressures
ÝÝÝÝÝÝÝÝÝÝÝÝÝÑ (PA + PB )V = (NA + NB )kT
%PB V = NB kT

We look at the center of mass,

á á á á
á mA V A + mB V B mA V 1 A + mB V 1 B
V cm = =
mA + mB mA + mB
where the velocity with prime is after collision.
á  á á
á
1
á
1 mA V 1 A + mB V 1 B á1 á
V cm ¨ V A ´ V B = ¨ (V A ´ V 1 B )
mA + mB
 
1 á
1
á
1
á
1
á
1
á
1
á
1
= mA V A ¨ V A ´ mB V B ¨ V B + (mB ´ mA )V A ¨ V B
mA + mB
á á á á
Note that this average must be 0, and mA V 1 A ¨ V 1 A ´ mB V 1 B ¨ V 1 B can be canceled since they are at same
temperature. This implies that
á á
V 1A ¨ V 1B = 0
which means there is no relation between the motions of two kind of gases. The average kinetic energy
after collision is
á á
1 á á 1 mA V 1 A + mB V 1 B
(mA + mB )V cm ¨ V cm = (mA + mB )
2 2 (mA + mB )2
 
1 1 2
á
1
á
1 2
á
1
á
1
á
1
á
1
= mA V A ¨ V A + mB V B ¨ V B + 2mA mB V A ¨ V B
2 mA + mB
1 1 3
= (mA 3kT + mB 3kT ) = kT
2 mA + mB 2

57
which means that if the two kinds of atoms combined together, then the average of the kinetic energy is
3
still kT .
2

3.3 Maxwell–Boltzmann distribution and Equipartition Theorem

3.3.1 The distribution of the atmosphere pressure

We assume the temperature is the same in every height. We denote the atmosphere pressure at the height
z by P (z) and the air density by n(z). By P V = N kT , we know P (z) = n(z)kT .

dz

A
Figure 41

By the definition of the pressure,

AP (z) = AP (z + dz) + gmn(z)dz

dP (z)
ñ ´ mgn(z)dz = P (z + dz) ´ P (z) = dz
dz
dP (z) mg
ñ = ´mgn(z) = ´ P (z)
dz kT
1 dP (z) mg
ñ =´
P (z) dz kT

58
integral both side
żz żz
mgz mg 1 dP (z)
´ = ´ dz = dz
kT 0 kT 0 P (Z) dz
żz
d log P (z)
= dz
0 dz
P (z)
= log P (z) ´ log P (0) = log
P (0)
Hence
mgz mgz
P (z) = P (0)e´ kT and n(z) = n(0)e´ kT

3.3.2 Maxwell speed distribution

At the height z, we can classify all molecule of gas by the velocity of z-direction vz . We denote the number
of molecule by N and the number of molecule whose velocity is between vz and vz + dvz by dN . We can
dN
view as a probability, and we can see
N
dN
9dvz
N
We write it as
dN
= f (vz ) dvz
N
where f (vz ) is the probability density function (p.d.f.).
Observe that the molecule at the height z will attains the height z 1 (z 1 ą z) if
1 2
mvz ě mg(z 1 ´ z)
2
We let u be the minimal velocity that the molecule can attain height z 1 i.e.
1
mu2 = mg(z 1 ´ z)
2

z1
0

z
u
Figure 42

59
Compute the number of both height z and z 1 , we get
ż8 ż8
1
n(z) f (vz )vz dvz = n(z ) f (vz1 )vz1 dvz1
u 0
Then
ż8
mg(z 1 ´ z) 1
n(z ) f (vz1 )vz1 dvz1
´ n(z 1 )
e kT = = ż 08
n(z)
n(z) f (vz )vz dvz
0
ż8 ż8
n(z) f (vz )vz dvz f (vz )vz dvz
u u
= ż 8 = 8 ż
n(z) f (vz )vz dvz f (vz )vz dvz
0 0
ż8
We know that f (vz )vz dvz is a constant, we write it C. Then
0

ż8 ´mu2
´
f (vz )vz dvz = Ce 2kT
u
ż8 mu2
d mu ´
ñ f (vz )vz dvz = ´C e 2kT
du u kT
mu2
mu ´
ñ ´ f (u)u = ´C e 2kT
kT
mu2
´
ñf (u) = C 1 e 2kT
where C 1 is another constant. Remember that f (vz ) is p.d.f., hence we have
ż8 ż8mu2 c c
1
´
1 2πkT 1 m
1= f (vz ) dvz = C e 2kT du = C ñC =
´8 ´8 m 2πkT
To sum up, the probability that the velocity of z-direction is between vz and vz + dz is
c ´mvz2
m ´
f (vz )dvz = e 2kT dvz
2πkT
By symmetry, $
’ c ´mvy2
m
’
’ ´
&f (vy )dvy = e 2kT dvy
’
’
2πkT
’ c ´mvx2
’
’ m ´
%f (vx )dvx = e 2kT dvx
’
’
2πkT

60
Then
c 3 m(vx2 + vy2 + vz2 )
m ´
f (vx )f (vy )f (vz ) dvx dvy dvz = e 2kT dvx dvy dvz
2πkT
a
means that the probability of the velocity between v = vx2 + vy2 + vz2 and v + dv and the probability of
speed between v and v + dv is
c 3 ´mv 2
m ´
e 2kT 4πv 2
2πkT
Now we consider not only velocity but also the height, the probability of the height is between z and z + dz
and the velocity is between v and v + dv is

n(z)dzf (vx )dvx f (vy )dvy f (vz )dvz

mgz c 3 mv 2 Ep + Ek
´ m ´ ´
=n(0)e kT e 2kT dzdvx dvy dvz 9e kT
2πkT
E
where Ep is potential energy and Ek is kinetic energy. The probability is relation to e´ kT is called
Boltzmann distribution law.
The equipartition theorem says that the energy is equally distributed to any degree of freedom. We see
some cases

• Monatomic gas :
1 1 1
Ek = mvx2 + mvy2 + m2z
2 2 2
and
1 2 1 2 1 2 1
mv = mv = mv = kT
2 x 2 y 2 z 2
• Diatomic gas : Let the mass of two atoms be m1 and m2 and the distance between two atoms be
á
R(t)

– The distance between to atoms is fixed :

á ! á !
1 2 2 2 1 m1 m2 dR(t) dR(t)
Ek = (m1 + m2 )(vcm,x + vcm,y + vcm,z )+ ¨
2 2 m1 + m2 dt dt

We know that
1 2
1 2
1 2
1
(m1 + m2 )vcm,x = (m1 + m2 )vcm,y = (m1 + m2 )vcm,z = kT
2 2 2 2

61
 á á
á á dR(t) dR(t) á á
Since |R(t)| is a constant, R(t) ¨ = 0 implies = ω ˆ R(t). Then
dt dt
1 m1 m2 á á  á á  1 1
ω ˆ R(t) ¨ ω ˆ R(t) = Ix ωx2 + Iy ωy2
2 m1 + m2 2 2
since the rotation about the z axis does not effect. The average energy of square terms is like
moving,
1 1 1
Ix ωx2 = Iy ωy2 = kT
2 2 2
Then
5
Ek = kT
2
– The distance between to atoms is not fixed : We can see as there is a spring between two atoms.
á
Write R(t) = ℓ0 + δ(t). We also have
á ! á !
1 2 2 2 1 m1 m2 dR(t) dR(t)
Ek = (m1 + m2 )(vcm,x + vcm,y + vcm,z )+ ¨
2 2 m1 + m2 dt dt

Similar from the previous case,

á !  2
1 m1 m2 dR(t) 1 1 1 m1 m2 dδ(t)
¨ = Ix ωx2 + Iy ωy2 +
2 m1 + m2 dt 2 2 2 m1 + m2 dt

Then
 2
1 á á 1 1 1 m1 m2 dδ(t)
Ek = (m1 + m2 )(V c,m ¨ V c,m ) + Ix ωx2 + Iy ωy2 + = 3kT
2 2 2 2 m1 + m2 dt
á
But since R(t) is not fixed, so there is potential energy.

K 2
Ep = δ (t)
2
where K is characteristic of the spring. Hence the average of total mechanical energy is
7
Ek + Ep = kT
2
So a spring contains kT energy in this sense.

We can also use this to measure the specific heat of sodium. (Dulong-Petit)

62
 Na

Figure 43

An atom of sodium has six springs out to connect other atoms, hence the atom is corresponding to three
springs averagely. If there are NA atoms of sodium, then there are 3NA springs. Then the average energy
is 3NA kT = 3RT , which means that molar specific heat of sodium is 3R.

3 5
Remark 3. The molar specific heat of hydrogen is R in low temperature and is R at 80K ´ 2000K and
2 2
7
is R in high temperature. Equipartition Theorem cannot explain the phenomenon, and it needs quantum
2
mechanics to know what happens.

3.4 Heat

The concept of heat and temperature was confused. Lavoisier thought that heat is an element. Mayer
gave the concept of the conservation of energy, and said that heat is the energy from high temperature to
low temperature.

Definition 4. The calorie is defined as the energy needed to raise the temperature from 14.5˝ C to 15.5˝ C

Joule measured the mechanical equivalent of heat 1cal= 4.18J

63
3.5 First Law of Thermodynamics and its application

First law of thermodynamics is a statement of the conservation of energy, which states that ∆U =
∆Q ´ ∆W where ∆U is the change of internal energy, ∆Q is the heat that system absorbed and ∆W is
the work done by the system on the surroundings.

3.5.1 Measure of the specific heat

The characteristic of ideal gas is the internal energy U only depend on the temperature T .
Molar specific heat at constant volume CV :
Since ∆V = 0, ∆W = 0. By the first law of thermodynamics, ∆U = ∆Q (the energy all absorbed by heat
transform to the internal energy). Hence ∆U = ∆Q = nCV ∆T

He Ne Ar O2 N2 H2
3R 3R 3R 5R 5R 5R
CV (300K) 2 2 2 2 2 2
3R 3R 3R 7R 7R 7R
CV (2000K) 2 2 2 2 2 2

Molar specific heat at constant pressure CP :

nR∆T
Now, pressure P is fixed, it is clear that ∆W = P ∆V = P = nR∆T .
P

∆U = ∆Q ´ P ∆V = nCP ∆T ´ nR∆T
Since internal energy only depend on temperature T and is independent of the volume V , ∆U = nCV ∆T .

nCV ∆T = nCP ∆T ´ nR∆T

ñCP = CV + R

CP R
Define γ = CV
=1+ CV
.

He Ne Ar O2 N2 H2
5 5 5 7 7 7
γ (300K) 3 3 3 5 5 5

64
3.5.2 Reversible adiabatic process

Adiabatic process means that ∆Q = 0. nCV ∆T = ∆U = ´∆W = ´P ∆V .

nCV ∆T P ∆T
=´
nRT PV
CV dT dV
ñ =´
R T V
1 dT dV
ñ =´
γ´1 T V
ż ż
1 dT dV
ñ =´
γ´1 T V
1
ñ log T = ´ log V + A
γ´1
1
ñ log T γ´1 + log V = A
1
ñ T γ´1 V = eA = A1

1
Hence T γ´1 V is a constant. Also,
1
T γ´1 V = A1
 1 γ´1
ñ T γ´1 V = A1γ´1 = A2
ñ T V γ´1 = A2
P V γ´1
ñ V = A2
nR
ñ P V γ = A3 = nRA2

P V γ is also a constant.

A3 = P V γ = P 1´γ (P V )γ
= P 1´γ (nRT )γ
= P 1´γ T γ nγ Rγ

P 1´γ T γ is a constant.

65
 P

isothermal

adiabatic
V

Figure 43

3.5.3 Reversible isothermal process

Suppose that the initial volume is Vi , and after getting heat, the volume becomes Vf . Since internal energy
only depend on temperature, ∆U = 0.

0 = ∆Q ´ ∆W = ∆Q ´ P ∆V

Hence ż ż Vf
nRT Vf
∆Q = P dV = dV = nRT log
Vi V Vi

66
3.5.4 Adiabatic process

∆U = nCV ∆T = ∆Q ´ ∆W
= ´∆W
ż
= ´ P dV
ż Vf 3
A
= γ
dV
Vi V
A3 1´γ Vf
= V | Vi
γ´1
1
= (Vf Pf ´ Vi Pi )
γ´1
1
= nR(Tf ´ Ti )
γ´1
= nCV (Tf ´ Ti )

3.6 Carnot heat engine and Carnot cycle

A Carnot heat engine consists of four steps :

(I) Isothermal expansion from volume Va to Vb at high temperature TH .

(II) Adiabatic expansion from volume Vb to Vc and temperature from TH to low temperature TL .

(III) Isothermal compression from volume Vc to Vd at low temperature TL .

(IV) Adiabatic compression from volume Vd to Va and temperature from TL back to the high temperature.

Note that the volume Vd can not choose arbitrary!

We can draw the procedure of pressure and volume in the P ´ V plane

67
 P
a(Va , Pa , TH )

IV I
b(Vb , Pb , TH )

d(Vd , Pd , TH ) II
III
c(Vc , Pc , TL )

Figure 43

We now compute the work

P a

∆WI
c
∆WII
V

Figure 44

68
 P a

∆WIV
c
∆WIII
V

Figure 45

(I) ∆UI = ∆QI ´ ∆WI , ∆U = 0. So ∆QI = ∆WI = nRTH log VVab .

(II) ∆QII = 0. ∆UII = ´∆WII = nCV (TL ´ TH ).

(III) ∆QIII = nRTL log VVdc = ∆WIII .

(IV) ∆UIV = ´∆WIV = nCV (TH ´ TL ).

W = ∆WI + ∆WII + ∆WIII + ∆WIV

Since ∆WII + ∆WIV = 0,
W = ∆WI + ∆WIII = ∆QI + ∆QIII
∆U = ∆UI + ∆UII + ∆UIII + ∆UIV = 0
∆Q = ∆QI + ∆QII + ∆QIII + ∆QIV = ∆QI + ∆QIII
The efficiency of Carnot heat engine is
∆W ∆QI + ∆QIII ∆QIII
η= = =1+
∆QI ∆QI ∆QI
And
∆QIII nRTL log VVdc ´TL log VVdc
= =
∆QI nRTH log VVab TH log VVab

69
Note that II and IV are adiabatic process,
$ 1 1
&T γ´1 V = T γ´1 V Vc Vb
L c H b
1 1 ùñ =
%T γ´1 V = T γ´1 V Vd Va
L d H a

Hence
TL
η =1´
TH
Also
∆QIII ∆QI
+ =0
TL TQ
this is the reason why we can use Carnot heat engine to measure entropy. And the work done by Carnot
heat engine  
TL
Wc = η∆QI = 1 ´ ∆QI
TH

3.7 Second Law of Thermodynamics and Clausius inequality

3.7.1 Second Law of Thermodynamics

There are two kind of statements of Second law of thermodynamics :

• Kelvin Planck : It is impossible to absorb energy in the form of heat from a single thermal reservoir
and to deliver an equivalent amount of work.

• Clausius : Heat can never pass from a colder to a warmer body without some other change.

A heat engine :

70
 TH QH

Wc = QH ´ QL

TL QL

Figure 46

Theorem 2. The two statements are equivalent.

Proof.

• Clausius ñ Kelvin Planck :

If not, that is, there exists a heat engine energy from single heat reservoir and transform all into
work. Then we can construct a heat engine that contradicts Clausius statement.

TH QH

Q1H

Q1L

TL

Figure 47

71
 where W = Q1H ´ Q1L and W = QH . Hence Q1L = Q1H ´ QH which means that this is a heat engine
that heat pass from a colder to a warmer body without anything change.

• Kelvin Planck ñ Clausius :

If not, that is, there exists a heat engine that heat can pass a colder to a warmer body without
anything change. Then given another heat engine than we can combine together to construct a heat
engine that absorb heat energy and give an same amount of work.

TH Q1H

QH

QL QL

TL

Figure 48

this is a heat engine absorb Q1H ´QH heat energy from heat reservoir and do the work W = Q1H ´QH
which contradicts Kelvin Planck statement.

Now we show that the heat engine between higher temperature and low temperature has the largest
efficiency when the heat engine is reversible.

72
 TH QH TH

Q1H

W Winv

QL Q1L

TL TL

η ηrev
Figure 49

We suppose that η ą ηrev , W = QH ´ QL = Wrev = Q1H ´ Q1L . Then η ą ηrev implies that Q1H ą QH .
And Q1L ´ QL = Q1H ´ QH ą 0 which means when we combine these two heat engine, then we construct a
heat engine that heat pass from a colder to a warmer body without anything change. So η ď ηrev . Then
η ď ηrev ď ηc ď ηrev . Hence ηrev = ηc = 1 ´ TTHL . And for any heat engine, η ď ηc .

QH ´ QL TL
ď1´
QH TH
QL TL
ñ1 ´ ď1´
QH TH
TL QL
ñ ď
TH QH
QH QL
ñ ď
TH TL
QH QL
ñ ´ ď0
TH TL
This is a Clausius inequality.

73
3.7.2 Clausius inequality

T2
T1
TN +1

Ti

Ti+1

Figure 50

We write the heat from Ti to Ti+1 by ∆Qi . And for each ∆Qi , we use a Carnot heat engine that absorb
QH,i heat and exhaust QL,i = ∆Qi heat with do the Wc,i work.
For this cycle A, ∆Ui = ∆Qi ´ ∆Wi for each i. Then
N
ÿ +1 N
ÿ +1 N
ÿ +1 N
ÿ +1 N
ÿ +1 N
ÿ +1
0= ∆Ui = ∆Qi ´ ∆Wi = QL,i ´ ∆Wi = ´WA
i=1 i=1 i=1 i=1 i=1 i=1

Hence WA = N
ř +1 řN +1
i=1 QL,i where WA = i=1 ∆Wi . (Caveat : ∆Wi ‰ ∆Qi )
Now we look N + 1 Carnot cycles as a system. Then
N
ÿ +1 N
ÿ +1 N
ÿ +1
WA + Wc,i = QL,i + (QH,i ´ QL,i )
i=1 i=1 i=1
N
ÿ +1 N
ÿ +1 N
ÿ +1
= QL,i + QH,i ´ QL,i
i=1 i=1 i=1
N
ÿ +1
= QH,i
i=1

74
 řN +1
Note that QH,i need to smaller than 0, otherwise it will contradict the second law of thermodynamics.
i=1
QH,i QL,i
For N + 1 Carnot cycles, = . So
TH Ti
N +1 N +1
ÿ QL,i ÿ ∆Qi
TH ď 0 ùñ ď0
i=1
Ti i=1
Ti

And this is Clausius inequality. Note that if A is reversible, then ´∆Qi ´ (´∆Wi ) = ´∆Ui and
N +1 N +1
ÿ ´∆Qi ÿ ∆Qi
ď 0 ùñ =0
i=1
Ti i=1
Ti

If we let N Ñ 8, then we write it as ¿

dQ
ď0
T
And if A is reversible then ¿
dQ
=0
T

3.8 Entropy

dQ
For reversible cycle, the integral of is 0. Then we can like conservation force to define something like
T
potential energy. So for a, b two point, we define the difference of entropy is
ż
dQ
S(b) ´ S(a) =
Γ T

where Γ is a reversible path from a to b.

75
3.8.1 Computation of the difference of entropy

a(Va , Pa , Ta )
b(Vb , Pb , Tb )



adiabatic c Vb , Pc , Tc = Va γ´1
Ta
Vb

Figure 51

Note that a ´ c is a reversible adiabatic process, dQ = 0, hence S(c) ´ S(a) = 0.

S(b) ´ S(a) = S(b) ´ S(c) + S(c) ´ S(a)

= S(b) ´ S(c)
żc
dQ
=
b T
ż Tb
nCV dT
=
Tc T
Tb
= nCV log
Tc
Tb Vbγ´1
= nCV log
T V γ´1
a a 
R Vb Tb
=n (γ ´ 1) log + log
γ´1 Va Ta
 γ´1 
Vb T Vb Tbγ´1
= nR log + log bγ´1 = nR log
Va Ta Va Taγ´1
We can also compute another way

76
 P

a(Va , Pa , Ta )

b(Vb , Pb , Tb )

isothermal
d(Vb , Pd , Ta )

Figure 52

S(b) ´ S(a) = S(b) ´ S(d) + S(d) ´ S(a)

ż Tb żd
nCV dT nRTa dV
= +
Ta T a Ta V
Tb Vb
= nCV log + nR log
Ta Va
R Tb Vb Tbγ´1 Vb
=n log + nR log = nR log γ´1
γ´1 Ta Va Ta Va
Free expansion : this process is adiabatic, i.e. dQ = 0. But the difference of entropy is not 0.

VA VB
ñ

VA VB

Figure 53

77
 a(VA , Ta ) real

isothermal Γ

b(VA + VB , Ta )

Figure 54

VA + VB VA + VB
S(b) ´ S(a) = nR log N k log
VA VA
This is wrong :
Since dQ = 0, then żb żb
dQ
S(b) ´ S(a) = = 0=0
a T a
The reason is that free expansion is irreversible. Suppose Γ is the reverse path from b to a. Then the
isothermal and Γ will contradict the second law of thermodynamics.
Now, the free expansion from a to b and the isothermal process from b to a form a cycle,
¿ ż ża
dQ dQ dQ
0ą = + = 0 + (S(a) ´ S(b))
T Γfree T b T

Hence S(a) ď S(b) which means that for the isolated system, the entropy tends to arise. And the equilib-
rium is at the maximal entropy for given restrict.
 N
S(a)´S(b) VA PA
e k = =
VA + VB PA+B
We can view it as a probability. And S(a) = k log PA . Then we have the Boltzmann formula

S = k log ΩA

ΩA PA
where ΩA+B
= PA+B
is the probability that all atoms are in A.

78
References

[1] 易富國, 普通物理學, 台大開放式課程 http://ocw.aca.ntu.edu.tw/ntu-ocw/ocw/cou/099S129

79