APhO Theoretical Problems and Solutions 2010-2016

LE HONG PHONG HIGH SCHOOL FOR THE GIFTED
HO CHI MINH CITY
APhO Theoretical
Problems and Solutions
2010 2016
FOR INTERNAL USE ONLY
Question Number
Theoretical Competition
25 April 2010
Page 1 of 5
__________________________________________________________________________________________
Theoretical Question 1
Particles and Waves
This question includes the following three parts dealing with motions of particles and waves:
Part A. Inelastic scattering of particles
Part B. Waves on a string
Part C. Waves in an expanding universe
Part A. Inelastic Scattering and Compositeness of Particles

A particle is considered elementary if it has no excitable internal degrees of freedom such
as, for example, rotations and vibrations about its center of mass. Otherwise, it is composite.
To determine if a particle is composite, one may set up a scattering experiment with the
particle being the target and allow an elementary particle to scatter off it. In case that the
target particle is composite, the scattering experiment may reveal important features such as
scaling, i.e. as the forward momentum of the scattered particle increases, the scattering cross
section becomes independent of the momentum.
For a scattering system consisting of an elementary particle incident on a target particle,
we shall denote by the total translational kinetic energy loss of the system. Here the
translational kinetic energy of a particle, whether elementary or composite, is defined as the
kinetic energy associated with the translational motion of its center of mass. Thus we may
write
= i f ,
where i and f are the total translational kinetic energies of the scattering pair before and
after scattering, respectively.
In Part A, use non-relativistic classical mechanics to solve all problems. All effects due to
gravity are to be neglected.
(a) As shown in Fig. 1, an elementary particle of mass moves along the axis with
-component of momentum 1 > 0. After being scattered by a stationary target of mass ,
its momentum becomes 2 .
Fig. 1
Question Number
25 April 2010
Page 2 of 5
__________________________________________________________________________________________
From data on 2 , one can determine if the target particle is elementary or composite.
We shall assume that 2 lies in the - plane and that the - and -components of 2 are
given, respectively, by 2 and 2 .
(i) Find an expression for in terms of , , 1 , 2 , and 2 .

[0.2 point]
(ii) If the target particle is elementary, the momenta 1 , 2 , and 2 are related in a
particular way by a condition.
For given 1 , plot the condition as a curve in the 2 - 2 plane. Specify the value
of 2 for each intercept of the curve with the 2 -axis. In the same plot, locate regions
of points of 2 corresponding to < 0, = 0, > 0, and label each of them as such.
[0.7 point]
For a stationary composite target in its ground state before scattering, which region(s)
of contains those points of 2 allowed?
[0.2 point]
(b) Now, consider a composite target consisting of two elementary particles each with
mass 12. They are connected by a spring of negligible mass. See Fig. 2. The spring has a
force constant and does not bend sideways. Initially, the target is stationary with its
center of mass at the origin O, and the spring, inclined at an angle to the -axis, is at its
natural length 0 . For simplicity, we assume that only vibrational and rotational motions
can be excited in the target as a result of scattering.
The incident elementary particle of mass moves in the -direction both before and
after scattering with its momenta given, respectively, by 1 and 2 . Note that 2 is negative
if the particle recoils and moves backward. A scattering occurs only if the incident particle
hits one of the target particles and 2 1 . We assume all three particles move in the same
plane before and after scattering.
Fig. 2
(i) If the maximum length of the spring after scattering is m , find an equation which
relates the ratio = (m 0 )/0 to the quantities , , 0 , , , , 1 and 2 .
[0.7 point]
Question Number
25 April 2010
Page 3 of 5
__________________________________________________________________________________________
(ii) Let sin2 . When the angle of orientation of the target is allowed to vary, the
scattering cross section gives the effective target area, in a plane normal to the
direction of incidence, which allows certain outcomes to occur as a result of scattering.
It is known that for all outcomes which lead to the same value of 2 , the value
of must span an interval (min , max ) and we may choose the unit of cross section
so that is simply given by the numerical range (max min ) of the interval. Note
that min , max , and, consequently, are dependent on 2 . Let c be the threshold
value of 2 at which starts to become independent of 2 .
In the limit of large , give an estimate of c . Express your answer in terms of , ,
and 1 .
[1.1 points]
Assume =3m and in the limit of large , plot as a function of 2 for a given 1 . In
[1.1 points]
the plot, specify the range of and p2 .
Part B. Waves on a String

Consider an elastic string stretched between two fixed ends A and B, as shown in Fig. 3.
The linear mass density of the string is . The speed of propagation for transverse waves in
be . The string is plucked sideways and held in a triangular
the string is . Let the length AB
form with a maximum height at its middle point. At time = 0, the plucked string is
released from rest. All effects due to gravity may be neglected.
Fig. 3
(c) Find the period of vibration for the string.
[0.5 point]
Plot the shape of the string at = /8. In the plot, specify lengths and angles which serve
to define the shape of the string.
[1.7 points]
(d) Find the total mechanical energy of the vibrating string in terms of , , , and .
[0.8 point]
Question Number
25 April 2010
Page 4 of 5
__________________________________________________________________________________________
Part C. An Expanding Universe

Photons in the universe play an important role in delivering information across the cosmos.
However, the fact that the universe is expanding must be taken into account when one tries to
extract information from these photons. To this end, we normally express length and distance
using a universal scale factor () which depends on time . Thus the distance () between
two stars stationary in their respective local frames is proportional to ():
() = (),
(1)
where is a constant and (t) accounts for the expansion of the universe. We use a dot above
a symbol of a variable to denote its time derivative, i.e. (t) = ()/, and let () ().
Taking time derivatives of both sides of Eq. (1), one obtains the Hubble law:
() = ()(),
(2)
where () = ()/() is the Hubble parameter at time . At the current time 0 , we have
(0 ) = 72 km s 1 Mpc 1 ,
where 1 Mpc = 3.0857 1019 km = 3.2616 106 light-year.

Assume the universe to be infinitely large and expanding in such a way that
() exp(),
where is a constant. In such a universe, the Hubble parameter is a constant equal to (0 ).
Moreover, it can be shown that the wavelength of photons travelling in the universe will be
stretched in proportion to the expansion of the universe, i.e.
() ().
Now suppose that photons making up a Lyman-alpha emission line were emitted at e by
a star that was stationary in its local frame and that we as observers are stationary in our local
frame. When these photons were emitted, their wavelength was (e ) = 121.5 nm. But when
they reach us now at 0 , their wavelength is red-shifted to 145.8 nm.
(e) As these photons traveled, the universe kept expanding so that the star kept receding from
us. Given that the speed of light in vacuum has never changed, what was the distance
(e ) of the star from us when these photons were emitted at e ? Express the answer in
units of Mpc.
[2.2 points]
(f) What is the receding velocity (0 ) of the star with respect to us now at 0 ? Express the
answer in units of the speed of light in vacuum .
[0.8 point]
Question Number
25 April 2010
Page 5 of 5
__________________________________________________________________________________________
Appendix
The following formula may be used when needed:
1
( ).
Question Number
25 April 2010
Page 1 of 4
__________________________________________________________________________________________
Strong Resistive Electromagnets
Resistive electromagnets are magnets with coils made of a normal metal such as copper or
aluminum. Modern strong resistive electromagnets can provide steady magnetic fields higher
than 30 tesla. Their coils are typically built by stacking hundreds of thin circular plates made
of copper sheet metal with lots of cooling holes stamped in them; there are also insulators
with the same pattern. When voltage is applied across the coil, current flows through the
plates along a helical path to generate high magnetic fields in the center of the magnet.
In this question we aim to assess if a cylindrical coil (or solenoid) of many turns can serve
as a magnet for generating high magnetic fields. As shown in Fig. 1, the center of the magnet
is at O. Its cylindrical coil consists of turns of copper wire carrying a current uniformly
distributed over the cross section of the wire. The coils mean diameter is and its length
along the axial direction is . The wires cross section is rectangular with width and
height . The turns of the coil are so tightly wound that the plane of each turn may be taken as
perpendicular to the axis and = . In Table 1, data specifying physical dimensions of the
coil are listed.
/2
Figure 1
/2
Table 1
= 12.0 cm
= 6.0 cm
= 2.0 mm
= 5.0 mm
In assessing if such a magnet can serve to provide high magnetic fields, two limiting
factors must not be overlooked. One is the mechanical rigidity of the coil to withstand large
Lorentz force on the field-producing current. The other is that the enormous amount of Joule
heat generated in the wire must not cause excessive temperature rise. We shall examine these
two factors using simplified models.
The Appendix at the end of the question lists some mathematical formulae and physical
data which may be used if necessary.
Part A. Magnetic Fields on the Axis of the Coil

Assume so that one may regard the wire as a thin strip of width . Let O be the
origin of coordinates. The direction of the current flow is as shown in Fig. 1.
Question Number
25 April 2010
Page 2 of 4
__________________________________________________________________________________________
(a) Find the -component () of the magnetic field on the axis of the coil as a function
of when the steady current passing through the coil is .
[1.0 point]
(b) Find the steady current 0 passing through the coil if (0) i s 10.0 T. Use data given in
Table 1 when computing numerical values.
[0.4 point]
Part B. The Upper Limit of Current

In Part B, we assume the length of the coil is infinite and . Consider the turn of
the coil located at = 0. The magnetic field exerts Lorentz force on the current passing
through the turn. Thus, as Fig. 2 shows, a wire segment of length is subject to a normal
force n which tends to make the turn expand.
Figure 2
/2
(c) Suppose that, when the current is , the mean diameter of the expanded coil remains at a
constant value larger than , as shown in Fig. 2.
Find the outward normal force per unit length n /.
[1.2 point]
Find the tension t acting along the wire.
[0.6 point]
(d) Neglect the coils acceleration during the expansion. Assume the turn will break when the
wires unit elongation (i.e. tensile strain or fractional change of the length) is 60 % and
tensile stress (i.e. tension per unit cross sectional area of the unstrained wire) is b =
455 MPa. Let b be the current at which the turn will break and b the corresponding
magnetic field at the center O.
Find an expression for b and then calculate its value.
[0.8 point]
Find an expression for b and then calculate its value.
[0.4 point]
Part C. The Rate of Temperature Rise
When the current is 10.0 kA and the temperature of the coil is 293 K, assume that the
resistivity, the specific heat capacity at constant pressure, and the mass density of the wire of
the coil are, respectively, given by e = 1.72 108 m , cp = 3.85 102 J/(kg K)
and = 8.98 103 kg m3 .

(e) Find an expression for the power density (i.e. power per unit volume) of heat generation in
the coil and then calculate its value. Use data in Table 1.
[0.5 point]
Question Number
25 April 2010
Page 3 of 4
__________________________________________________________________________________________
(f) Let be the time rate of change of temperature in the coil. Find an expression for and
then calculate its value.
[0.5 point]
Part D. A Pulsed-Field Magnet

If the large current needed for a strong magnet lasts only for a short time, the temperature
rise caused by excessive Joule heating may be greatly reduced. This idea is employed in a
pulsed-field magnet.
Thus, as shown in Fig. 3, a capacitor bank of capacitance charged initially to a
potential 0 is used to drive the current through the coil. The circuit is equipped with a
switch . The inductance and resistance of the circuit are assumed to be entirely due to
the coil. The construct and dimensions of the coil are the same as given in Fig. 1 and Table 1.
Assume , , and to be independent of temperature and the magnetic field is the same as
that of an infinite solenoid with .
Figure 3
(g) Find expressions for the inductance and resistance .

[0.6 point]
Calculate the values of and . Use data given in Table 1.
[0.4 point]
(h) At time = 0, the switch is thrown to position 1 and the current starts flowing.
For 0, the charge () on the positive plate of the capacitor and the current ()
entering the positive plate are given by
0
() =
sin( + 0 ),
(1)
sin 0
() =
0
=(
)
sin ,
cos 0 sin 0
in which and are positive constants and 0 is given by
tan 0 = ,
0 < 0 < .
(2)
(3)
Note that, if () is expressed as a function of a new variable ( + 0 /), then ()

and its time derivative () are identical in form except for an overall constant factor. The
time derivative of () may therefore be obtained similarly without further differentiations.
Find and in terms of , , and .
Calculate the values of and when is 10.0 mF.
[0.8 point]
[0.4 point]
Question Number
25 April 2010
Page 4 of 4
__________________________________________________________________________________________
(i) Let m be the maximum value of |()| for > 0. Find an expression for m .
[0.6 point]
If = 10.0 mF, what is the maximum value 0b of the initial voltage 0 of the capacitor
bank for which m will not exceed b found in Problem (d)?
[0.4 point]
(j) Suppose the switch is moved instantly from position 1 to 2 when the absolute value of
the current | ()| reaches m . Let be the total amount of heat dissipated in the coil
from = 0 to and the corresponding temperature increase of the coil. Assume the
initial voltage 0 takes on the maximum value 0b obtained in Problem (i) and the
electromagnetic energy loss is only in the form of heat dissipated in the coil.
Find an expression for and then calculate its value.
[1.0 point]
Find an expression for and then calculate its value. Note that the value for must be
compatible with the assumption of constant and .
[0.4 point]
Appendix
1.
(2
= 2 2
2
3/2
( + 2 )1/2
+ )
2. sin( ) = sin cos cos sin
3. permeability of free space 0 = 4 107 T m/A
------------------------------------------------------ END ------------------------------------------------
Question Number
25 April 2010
Page 1 of 4
__________________________________________________________________________________________
Electron and Gas Bubbles in Liquids
This question deals with physics of two bubble-in-liquid systems. It has two parts:
Part A. An electron bubble in liquid helium
Part B. Single gas bubble in liquid
Part A. An Electron Bubble in Liquid Helium

When an electron is planted inside liquid helium, it can repel atoms of liquid helium and
form what is called an electron bubble. The bubble contains nothing but the electron itself.
We shall be interested mainly in its size and stability.
We use to denote the uncertainty of a quantity . The components of an electrons
position vector = (, , ) and momentum vector = ( , , ) must obey Heisenbergs
uncertainty relations /2 , where is the Planck constant divided by 2

and = , , .
We shall assume the electron bubble to be isotropic and its interface with liquid helium is a
sharp spherical surface. The liquid is kept at a constant temperature very close to 0 K with its
surface tension given by 3.75 104 N m1 and its electrostatic responses to the electron
bubble may be neglected.
Consider an electron bubble in liquid helium with an equilibrium radius . The electron, of
mass , moves freely inside the bubble with kinetic energy k and exerts pressure e on the
inner side of the bubble-liquid interface. The pressure exerted by liquid helium on the outer
side of the interface is He .
(a) Find a relation between He , e , and .

Find a relation between k and e .
[0.4 point]
[1.0 point]
(b) Denote by 0 the smallest possible value of k consistent with Heisenbergs uncertainty
relations when the electron is inside the bubble of radius . Estimate 0 as a function
of .
[0.8 point]
(c) Let e be the equilibrium radius of the bubble when k = 0 and He = 0.
Obtain an expression for e and calculate its value.
[0.6 point]
(d) Find a condition that and He must satisfy if the equilibrium at radius is to be
locally stable under constant He . Note that He can be negative.
[0.6 point]
(e) There exists a threshold pressure th such that equilibrium is not possible for the electron
bubble when He is less than th . Find an expression for th .
[0.6 point]
Question Number
25 April 2010
Page 2 of 4
__________________________________________________________________________________________
Part B. Single Gas Bubble in Liquid Collapsing and Radiation

In this part of the problem, we consider a normal liquid, such as water.
When a gas bubble in a liquid is driven by an oscillating pressure, it can show dramatic
responses. For example, following a large expansion, it can collapse rapidly to a small radius
and, near the end of the collapse, emit light almost instantly. In this phenomenon, called
single-bubble sonoluminescence, the gas bubble undergoes cyclic motions which typically
consist of three stages: expansion, collapse, and multiple after-bounces. In the following we
shall focus mainly on the collapsing stage.
We assume that, at all times, the bubble considered is spherical and its center remains
stationary in the liquid. See Fig 1. The pressure, temperature, and density are always uniform
inside the bubble as its size diminishes. The liquid containing the bubble is assumed to be
isotropic, nonviscous, incompressible, and very much larger in extent than the bubble. All
effects due to gravity and surface tension are neglected so that pressures on both sides of the
bubble-liquid interface are always equal.
Radial motion of the bubble-liquid interface

As the bubbles radius = () changes with time , the bubble-liquid interface will move
with radial velocity /. It follows from the equation of continuity of incompressible
fluids that the liquids radial velocity / at distance from the center of the bubble is
related to the rate of change of the bubbles volume by
(1)
= 4 2 = 4 2 .
This implies that the total kinetic energy k of the liquid with mass density 0 is
0
1 0
1
1 1
k = 0 (4 2 ) 2 = 20 4 2 2 = 20 4 2
(2)
2
0

where 0 is the radius of the outer surface of the liquid.
0
Fig. 1
T0
Question Number
25 April 2010
Page 3 of 4
__________________________________________________________________________________________
(f) Assume the ambient pressure 0 acting on the outer surface = 0 of the liquid is constant.
Let = () be the gas pressure when the radius of the bubble is .
Find the amount of work done on the liquid when the radius of the bubble changes
from to + . Use 0 and to express .
[0.4 point]
The work must be equal to the corresponding change in the total kinetic energy of the
liquid. In the limit 0 , it follows that we have Bernoullis equation in the form
1
0 m 2 = ( 0 ) n .
(3)
2
Find the exponents m and n in Eq. (3). Use dimensional arguments if necessary. [0.4 point]
Collapsing of the gas bubble
From here on, we consider only the collapsing stage of the bubble. The mass density of the
liquid is 0 = 1.0 103 kg m3 , the temperature 0 of the liquid is 300 K and the ambient
pressure 0 is 1.01 105 Pa. We assume that 0 , 0 , and 0 remain constant at all times and
the bubble collapses adiabatically without any exchange of mass across the bubble-liquid
interface.
The bubble considered is filled with an ideal gas. The ratio of specific heat at constant
pressure to that at constant volume for the gas is = 5/3. When under temperature 0 and
pressure 0 , the equilibrium radius of the bubble is 0 = 5.00 m.

Now, this bubble begins its collapsing stage at time = 0 with (0) = i = 70 ,
(0) = 0, and the gas temperature i = 0 . Note that, because of the bubbles expansion in
the preceding stage, i is considerably larger than 0 and this is necessary if
sonoluminescence is to occur.
(g) Express the pressure () and temperature () of the ideal gas in the bubble as
a function of during the collapsing stage, assuming quasi-equilibrium conditions hold.
[0.6 point]
(h) Let /i and = / . Eq. (3) implies a conservation law which takes the
following form
1
2
0 2 + () = 0.
(4)
Let i (i ) be the gas pressure of the bubble when = i . If we introduce the

ratio i /[( 1)0 ] , the function () may be expressed as
() = 5 [(1 2 ) 2 (1 3 )].
(5)
Find the coefficient in terms of i and 0 .
[0.6 point]
(i) Let m be the minimum radius of the bubble during the collapsing stage and define
m m /i . For 1, we have m m .
Find the constant m .
[0.4 point]
Question Number
25 April 2010
Page 4 of 4
__________________________________________________________________________________________
Evaluate m for i = 70 .
Evaluate the temperature m of the gas at = m .
[0.3 point]
[0.3 point]
(j) Assume i = 70 . Let be the value of at which the dimensionless radial speed
| | reaches its maximum value. The gas temperature rises rapidly for values of near .
Give an expression and then estimate the value of .

[0.6 point]
Let be the value of at = (m + )/2. Evaluate .
[0.4 point]
Give an expression and then estimate the duration m of time needed for to diminish
from to the minimum value m .
[0.6 point]
Sonoluminescence of the collapsing bubble
Consider the bubble to be a surface black-body radiator of constant emissivity so that

the effective Stefan-Boltzmanns constant eff = SB . If the collapsing stage is to be
approximated as adiabatic, the emissivity must be small enough so that the power radiated by
the bubble at = is no more than a fraction, say 20 %, of the power supplied to it by the
driving liquid pressure.
(k) Find the power supplied to the bubble as a function of .

Give an expression and then estimate the value for an upper bound of .
[0.6 point]
[0.8 point]
Appendix
1.

= 1
2. Electron mass = 9.11 1031 kg
3. Planck constant = 2 = 2 1.055 1034 J s
4. Stefan-Boltzmanns constant SB = 5.67 108 W m2 K 4
END
----------------------------------------------------------------------------------------------------------------
Question Number
25 April 2010
Page 1 of 8
(Document Released: 14:30, 4/24)

__________________________________________________________________________________________
Theoretical Question 1: Particles and Waves

SOLUTION
Part A. Inelastic scattering and compositeness of particles

(a) Let the momentum of the target particle after scattering be . The law of conservation of
linear momentum implies = 1 2 . The total translational kinetic energies of the
scattering system before and after scattering are, respectively,
=
f =
1 2
2
2 2 (1 2 )2 2 2
1
+
=
+
( 2 21 2 + 2 2 ).
2
2
2 2 1
(a-1)
(i) By definition, we have = i f , or equivalently,
1 2 2 2
1
( 2 21 2 + 2 2 )
= i f =
2 2 2 1
1
=
{( )1 2 ( + )2 2 + 21 2 }
2
=
2
+ 2
{
1 2 2 +
}
+ 1 2
2 +
+ 2
2
{
1 2 2 +
1 2 2 2 }
2 +
+
+
2 2
{[
+(
) ]1 (2
1 )2 2 2 }
2 +
+
+
+
2 2
{(
) 1 (2
1 )2 2 2 }
(a-2)*
2 +
+
(ii) If the incident and target particles are both elementary, their internal energies remain the
same before and after the scattering. By the law of conservation of energy, we must
=
have i = f , or = 0. Thus, we obtain from Eq. (a-2) the following equality

(
2 2
) 1 = (2
)2 + 2 2
+
+ 1
In the 2 -2 plane, this represents a circle centered at (1 /( + ), 0) with

radius 1 /( + ). The case < is shown in Fig. A1. The values of 2 at the
intercepts of the circle with the 2 -axis are

1
1 =
and
+
+
+ 1
1 +
= 1 .
+
+ 1
(a-3)*
Question Number
25 April 2010
Page 2 of 8

__________________________________________________________________________________________
Fig. A1
> 0
< 0
= 0
1
For a composite target in its ground state before scattering, the law of conservation of
energy implies
i = f + int ,
where int 0 is the change in internal energy (or excitation energy) of the target as a
result of scattering and i and f are given by Eq. (a-1). Thus, in this case, the total
translational kinetic energy loss is given by = i f = 0. For points on the
circumference of the circle in Fig. A1, we have = 0, i.e. elastic scattering. For the
interior of the circle, we have > 0, corresponding to inelastic scattering with the target
in an excited state after scattering.
The circle and its interior ( 0) are thus allowed by a composite target in its ground
state before scattering.
(b) Let be the angular momentum of the target about an axis through its center of mass and
normal to the plane of particle motions after scattering. By the law of conservation of
angular momentum,
1
= (2 0 sin )(1 2 )
(a-4)
where the + (or ) sign is implied if the target particle on the left (or right) in Fig. 2 is hit
by the incident particle.
(i) After scattering, the target may undergo vibrational and rotational motions. When the
spring reaches its maximum extension, the length of the spring is m = (1 + )0 and the
moment of inertia of the target rotating about an axis through its center of mass and
perpendicular to the spring is m = 14 m 2 = 14 0 2 (1 + )2 . The law of conservation

of energy implies
1
2
2
= (m 0 ) +
2
2m
(a-5)
Question Number
25 April 2010
Page 3 of 8

__________________________________________________________________________________________
where the last term represents the rotational kinetic energy of the target at the maximum
extension of the spring. According to Eq. (a-4), we have
2
0 (1 2 )2 2
1 2 (1 2 )2 2
= ( )2
sin = (
)
sin
2m
m
2
2
1 +
and therefore
Note that, since 2
1
1 2 (1 2 )2 2
0 2 2 + (
)
sin .
1 +
2
2
= 0 and 2 2 , we have from Eq. (a-2)
= i f =
+
2 2
{(
) 1 (2
)2 }
2 +
+ 1
(1 2 )
{( )1 + ( + )2 }
2
A scattering can occur only if 1 2 and 0, so from Eq. (a-8), we obtain
=
(a-6)
(a-7)*
(a-8)

2 < 1
(a-9)*
+ 1
where the equalities hold only if = 0.
---------------------------------------------------------------------------------------------------------------*An equation marked with an asterisk gives key answers to the problem.
(ii) The scattering cross section is given by the numerical range of = sin2 . For
given 1 and 2 , is a constant by Eq. (a-8), and the value of can be found from Eq.
(a-7) to be
2
1
2
= sin2 =
(1
+
)
(
0 2 2 ) 0.
2
(1 2 )
2
In the limit of large , the last inequality can hold only if is very small. Thus may be
neglected in the factor (1 + ) and we obtain
2
1
1
2 2
= sin2
(
)
=
(1
0 2 2 ),
(a-10)
0
2
(1 2 )
2
2
where
2
1
=
{( )1 + ( + )2 }
2
(1 2 )
(1 2 )
is nonnegative and the last equality follows from Eq.(a-8).
(a-11)
From Eq. (a-10), the minimum value min of = sin2 is found to be zero for all 0,
and this occurs when
2
2 =
(min = 0).
0 2
Question Number
25 April 2010
Page 4 of 8

__________________________________________________________________________________________
Moreover, the maximum value max of = sin2 is seen to be given by

max
if 1 and = 0,
1
1
=
0 2 2 ) = .
1 if 1 and (1
(a-12)
Note that, from Eq. (a-11), it follows
2
,
+ 2 1
=
2
1 if 2
.
+ 2 1
1 if 2
Since min = 0, the cross section is given by = max min = max and, from Eq.
(a-12), we see that it becomes 1 and is independent of 2 when 1. Thus the threshold
value at which scaling of cross section starts is given by
2
.
(a-13)*
+ 2 1
For 2 below the threshold value, the cross section is equal to according to Eq. (a-12)
and, from Eq. (a-11), we have the following result
( )1 + ( + )2 (1 +2 )
= =
=
1. ( 2 ) (a-14)
(1 2 )
(1 2 )
=
Note that, from Eq. (a-9), we have 0 for 2 as expected. The cross section as a
function of 2 is shown in Fig. A2 which evidently shows scaling behavior.
1 =
+
2
= 3
1
Figure A2
Question Number
25 April 2010
Page 5 of 8

__________________________________________________________________________________________
Part B. Waves on a string
(c) The initial disturbance will propagate toward the fixed ends. It can be considered as the
superposition of two waves with wave forms ( ) and ( + ) , travelling
toward the right and the left, respectively. They will both be reflected out of phase at the
fixed ends. At = 0, the sum of their displacements must be equal to the initial wave form
(). Therefore
() + () = (), 0
(b-1)
()
Let
= /. At = 0, the string is at rest and the sum of velocities R = R
and y L = L of the two waves at must be zero. Thus we have
() () = 0,
0
(b-2)
Integrating Eq. (b-2) with respect to and combining with Eq. (b-1), we obtain
1
() = 2 (() + 0 ),
() = 2 (() 0 ) , , (b-3)
where 0 is a constant. (Note that this result may also be obtained by graphical
construction which takes initial conditions into account by superposing two pulses of
identical wave form but travelling in opposite directions.)
Since both waves, after being reflected once at each end and having travelled a
distance 2, return to its original position and state, the period is thus
=
2
.
[Another way to get the period ]:

Because the first harmonics (with the longest period) has = 2, we have
1 2
= = =
.

(b-4)*
(b-4')*
At = /8, 1/8 of each wave will have been reflected out of phase as shown below, where
the result for the wave traveling to the right is shown as dashed lines,
2
1
Question Number
25 April 2010
Page 6 of 8

__________________________________________________________________________________________
The displacements of all wave components may be added to give the resultant wave form
at = /8, as shown below.
tan = 2/
(b-5)*
[Another solution of (c)]:

Since the evolution of the string is periodic, waves that can simulate it must be periodic in
space. Furthermore, since the string is released from rest, we can consider the initial
configuration of the string as a superposition of two saw-tooth waves travelling in
opposite directions as shown below:
/2
/2
F
Both A and B will be fixed because two saw-tooth waves tend to move A or B in opposite
directions. Clearly, the period of motion is the time for the saw-tooth wave to travel the
distance 2. Hence we obtain
2
.
(b-4'')*
At any time , the shape of the string is determined by adding up the two waves as shown
below:
=
Question Number
25 April 2010
Page 7 of 8

__________________________________________________________________________________________
From the figure above, one sees that between 1 and 2 , two saw-tooth waves (line
marked by green and red line) have opposite slopes and hence their sum between 1 and
2 is constant with the height being given by /2 (the height of 1 or 2 ). Between
and 1 or and 2 , two saw-tooth waves have the same slope but move in opposite
direction, hence their sum simply reproduces the original saw-tooth shape. Thus, the shape
of string at time is
with
1 2 = 2. For = /8,
1 2 = /2 and
2
.
(b-5'')*
(d) To find the total energy, we note that the normal force which pulls the string sideways at
the midpoint is
tan =
2
,
(b-6)
where is the constant tension ( ) on the string and is the transverse

displacement at the midpoint. The work done by is the total mechanical energy given
to the string or
() = 2sin = 2
2
2
2
= ()d = 2 = 2
,
where use has been made of = /.
(b-7)*
[Another solution of (d)]:
Because c = / with being the tension on the string, we have = 2 . The total
mechanical energy at = 0 is the potential energy
2
1 2 2
1
2
= = =
= 2 2
2
/2
2 0
Here (, ) is the displacement of the elastic string.
(b-7')*
[Yet another solution of (d)]:
We consider a special moment when

and
move into
region completely. At
this moment, the string is flat so that total mechanical energy is equal to the total kinetic
energy. Since the velocity for each point on the string is 2 tan (downward) with
tan = /, we obtain
1
2
2
2
= (2 tan ) = 2
.
(b-7'')*
2
Question Number
25 April 2010
Page 8 of 8

__________________________________________________________________________________________
Part C. The expanding universe
(e) The photons were emitted at e , and are received now at 0 , so

(0 ) (0 ) 145.8
=
=
1.200
(e ) (e ) 121.5
(c-1)
On the other hand, the Hubble parameter can be derived as

()
() exp () () =
= ,
()
(c-2)
which is independent of time.
Within at some moment in the past, the photons traveled , which was
(e )
()
at time e due to the cosmic expansion. The photons were emitted at e so the distance of
the star from us at that time is
0
0
(e )
= exp [(e )]
(e ) =
e ()
e
=
(1 exp
[(e 0 )]).
(c-3)
We already know from Eq. (c-2) that
thus
exp
(0 ) (0 )
=
1.200,
exp
(e ) (e )
1
1
690 Mpc.
1.200
(f) Due to the cosmic expansion, the above distance is actually longer now:
(0 )
(0 )
(e )
(0 ) =
(e ) =
1
(e )
(e )
(0 )
(e ) =
(c-4)
(c-5)*
(c-6)
Thus according to the Hubble Law, we can compute the receding velocity of the star now:
v(0 ) = (0 ) =
(0 )
(e )
(0 )
1
=
1 0.200
(e )
(0 )
(e )
(c-7)*
Question Number
Page 1 of 6
25 April 2010

__________________________________________________________________________________________
Theoretical Question 2: Strong Resistive Electromagnets

SOLUTION
Part A. Magnetic Fields on the Axis of the Coil
(a) At the point on the axis, the magnetic field due to the current passing through the turns
located in the interval (, + ) is (see Fig. A1)
()
(/2)
= ( 0 )
(a-1)
4 (/2)2 + ( )2 (/2)2 + ( )2
() =
which, when summed over all turns of the coil, leads to the total magnetic field
() with
/2
0 2
() =
( )
2 2
[(/2)2 + ( )2 ]3/2
/2
0 2
=
( )
2 2
=
/2
/2
+ 2 ]3/2
[(/2)2
0
(/2)
(/2) +
2 (/2)2 + [(/2) ]2 (/2)2 + [(/2) + ]2
Figure A1
1
2
1
2
no. of turns in is /
(b) From Eq. (a-2), the magnetic field at O wth = 0 is

(0) =
(a-2)*
0
2(/2)
0
1
=
2 (/2)2 + (/2)2
1 + (/)2
If (0) is 10.0 T, then the current must be equal to
0 = (0) 1 + (/)2 = 1.7794 104 A 1.8 104 A

0
(b-1)
(b-2)*
Question Number
Page 2 of 6
25 April 2010

__________________________________________________________________________________________
Part B. The Upper Limit of Current
acting on the
(c) For an infinitely long coil with and , the magnetic field
current is the average of the fields inside and outside of the coil. The field outside is zero
and the field inside is the same as that at O, i.e. (0) in Eq. (b-1) with . Thus we
have
1

= = (0 + 0 ) = 0 ,
(c-1)
2
2
and the outward normal force on the wire segment of length is
0
n 0 2
n = = or
=
.
(c-2)*
2
2
As can be seen from Fig. A2, the resultant of the pair of tension forces at the ends of the
segment is given by
2
) t = t
.
(c-3)
2
This must be in equilibrium with the normal force n so that, by using Eq. (c-2), we have
2
n
0 2
n = t
or t =
=
.
(c-4)*
4
2
2t sin (
Figure A2
(d) At breaking, the tensile stress of the wire is, from Eq. (c-4),
t
0
= 2 b 2 = b = 4.55 108 Pa ,
4
and the tensile strain of the wire is
( ) ( )
=
= 60 % or = 1.60 .
From the last two equations, the current b at which the turn will break is
(d-1)
(d-2)
b
b
b = 2
= 2
= 1.737 104 A 1.7 104 A,
0
0 (1.60 )
(d-3)
0 b
0 b
= 2
= 10.914 T = 1.1 101 T,
(d-4)*
and the magnitude of the magnetic field at the center O, i.e. Eq. (b-1) with , is
b =
Question Number
Page 3 of 6
25 April 2010

__________________________________________________________________________________________
Part C. Rate of Temperature Rise
(e) When the current is 10.0 kA, the current density is given by
=
1.00 104
=
= 1.0 109 A/m2 .
(2.0 103 )(5.0 103 )
The power density is given by

2
e 2 = e = 1.720 1010 W/m3 1.7 1010 W/m3 .
(ALTERNATIVE)
(e-1)
(e-2)*
The volume and resistance (appearing also in Problem (h)) of the current-carrying
wire for a coil of length are given by
2
+ 2

= = ,
=
2
2
= e 2 = 1.9453 102 1.9 102 .

The total power of Joule heat generated in the coil is
= e
(e-3)
(e-4)
= 2 = 1.9453 106 W = 1.9 106 W.

Thus the power density is
(e-5)
2
2
2
=
=
= e = e 2 .
e 2
This is identical to that obtained in Eq. (e-2).
(e-7)*
=
=
= 1.7 1010 W/m3 .
(e-6)*

Note that, by Eqs. (e-3) to (e-5), the expression for power density may also be written as
(f) The time rate of temperature increase of the coil is

=
e 2
e
2
=
.

At = 293 K and = 10.0 kA, we have

=
e
2
e 2
=
= 4.975 103 K/s 5.0 103 K/s.

(ALTERNATIVE)
The heat capacity of the coil is
= () = 3.9101 102 J/K 3.9 102 J/K.
From Eqs. (e-5) and (f-3), the time rate of temperature increase is
2
=
= 4.975 103 K/s 5.0 103 K/s.
(f-1)
(f-2)*
(f-3)
(f-4)*
Question Number
Page 4 of 6
25 April 2010

__________________________________________________________________________________________
Part D. A Pulsed-Field Magnet
(g) The magnetic flux through each turn is, in the limit , given by
2 0 2
= { lim (0)} =
.
2
The inductance of the coil is
0 2 0
=
=
= 2 2 = 1.0659 104 H 1.1 104 H.
2
4
The resistance of the coil is the same as given in Eq. (e-4). Thus
(g-1)
(g-2)*
= e 2 = 1.9453 102 1.9 102 .

(g-3)*

(h) According to Kirchhoffs circuit law, the change of electric potential around a closed
circuit must be zero and we have
+ + = 0.
(h-1)
In this question, we are given

0
0
0 (+0 )
0
sin ( +
() =
sin( + 0 ) = ( )
),
(1)
sin 0
sin 0
= e
() =
0
= (
)
sin ,
cos 0 sin 0
tan 0 =
(2)
(3)
Comparing the right sides of Eqs. (1) and (2), one sees that the current () = / is
obtained from () by changing the latters time variable to ( 0 /) or, equivalently,
changing ( + 0 /) to , and then multiplying its amplitude constant by a factor
0
(
).
cos 0
Since () in Eq. (2) has the same form as () in Eq. (1), we may apply the same rule
again to obtain its derivative / as
0
0
0
0
= (
) (
)
( ) sin ( )
cos 0
cos 0 sin 0
=(
2 0
)
sin( 0 )
cos 0 sin 0
(h-2)
Making use of Formula 2 given in Appendix, we may express the left side of Eq. (h-1) as a
linear combination of cos and sin so that
0
+ + =
( cos 0 sin + sin 0 cos ) = 0,
sin 0
(h-3)
Question Number
Page 5 of 6
25 April 2010

__________________________________________________________________________________________
which can be satisfied if and only if

2
1
(
) ( 2 ) + = 0,
(h-4)
cos 0
cos 0
2 1
(
) + = 0,
(h-5)
cos 0
Note that Eqs. (h-4) and (h-5) may be obtained more simply by considering Eq. (h-1) at
the moments when sin = 1 and 0, respectively. Subtracting Eq. (h-5) from Eq. (h-4),
we obtain
=
,
(h-6)
2
If we use the expressions given in Eqs. (g-2) and (g-3), we obtain
= 2 = 9.1249 101 s 1 9.1 101 s1 .

=
=
(h-7)*
2 2 0 0
22
Adding up Eqs. (h-4) and (h-5), we have, by Eq. (h-6) and Eq. (3),
1
2
=
=
= 2 (1 + tan2 0 ) = 2 + 2 .
2 cos 2 0
cos2 0
This may be rewritten as
2 = 0 2 2 =
and we obtain
( )2 with
0 =
= 9.7 102 rad/s,
(h-8)
(h-9)*
= 0 2 2 = 9.6428 102 rad/s 9.6 102 rad/s.

(h-10)*
(i) From Eq. (h-2), the maximum value of |()| appears at / = 0 when the time is
0
m = .
(i-1)
From Eq. (2), the maximum value of |()| is then given by
m = |(m )| = (
)0 0 .
(i-2)*
cos 0
From Eqs. (3), (h-7) and (h-10), we have
0
= 10.568,
0 = 1.4764 rad,
m =
= 1.531 103 s.
If m does not exceed b found in Problem (d), we must have
m = |(m )| b or (
) 0 0 b ,
cos 0
which implies that the maximum value 0b of 0 occurs when the equality holds
given by
b 0
0b =
cos 0 = 2.0623 103 V 2.1 103 V.
tan 0 =
(i-3)
(i-4)
and is
(i-5)*
Question Number
25 April 2010
Page 6 of 6

__________________________________________________________________________________________
[j] When |()| reaches its maximum at = m = 0 /, the voltage of the capacitor has
dropped from the initial voltage 0 = 0b to
(m )
0b 0
(m ) =
=
sin(20 ) = 20b 0 cos 0 .
(j-1)
sin 0
From = 0 to = m , the energy supplied by the capacitor bank to the circuit, in the form
of Joule heat and magnetic energy in the field of the coil, is
2
1
1
= 0b 2 [(m )]2 = 0b 2 1 4 0 cos2 0 .
2
2
(j-2)
By the law of conservation of energy, this entire amount of energy is eventually turned
into heat in the coil and we have
2
1
= = 0b 2 1 4 0 cos 2 0 = 2.0694 104 J 2.1 104 J. (j-3)*
2
If the heat capacity (as computed in Eq. (f-3) remains about the same as that at = 293 K,
then the temperature increase is
=
=
= 53 K.
(j-4)*
()
With such a temperature increase, the thermal and electrical properties of a metal such as
copper do not change substantially.
Question Number 3
Page 1 of 6
25 April 2010

__________________________________________________________________________________________
Theoretical Question 3: Electron and Gas Bubbles in Liquids

SOLUTION
Part A. An Electron Bubble in Liquid Helium
(a) Consider a half of the spherical interface (see Fig. A1 below). The condition for its static
equilibrium implies that the total force acting on it must be zero. This implies
2 (e He ) = 2
(a-1)
which leads to
e = He +
Fig. A1
(a-2)*
He
e
According to the de Broglie relation = / 1/, the non-relativistic kinetic energy

E k is inversely proportional to 2 , i.e.
2
const.
=
.
k =
2
2
By the work- energy theorem, we have
Thus,
or
e = k = (2)
(4 2 ) =
e =
2 3 k
const.
2

3
k
2

k
(a-3)
(a-4)
(a-5)
(a-6)*
[Alternative]
The state of an electron confined in the bubble corresponds to standing waves which
vanish on the interface. According to Part B of Question 1, these are equivalent to the
superposition of two travelling waves moving in opposite directions and continually being
reflected at the interface. They give rise to pressure on the interface and the relation
Question Number 3
Page 2 of 6
25 April 2010

__________________________________________________________________________________________
between the non-relativistic kinetic energy k and the pressure e for an electron inside a
bubble is similar to that obtained from the kinetic theory of gases. Thus we have
2 k
2
1
=
k =
3
3
4
2 3 k
(b) Let = /(2). From the uncertainty relations, we have
e =
2 ,
1
2
(a-7)*
1
2 .
From symmetry considerations implied by isotropy, we have

= = = 0,
=
=
= 0
2
2
2
2
2
2
() = = = () = () , ( )2 =
2 = ( )2 = ( )2 .
where denotes the mean value of the quantity . Therefore, we have
3(x)2 = (x)2 + (y)2 + (z)2 =
x 2 + y2 + z2 = r2 .
2 2
3( )2 = ( )2 + ( )2 + ( )2 =
2 +
+ = .
Thus we obtain (cf. 18th IPhO)
2 = 9()2 ( )2 9 2
2
4
(b-1)
(b-2)
(b-3)
(b-4)
(b-5)
(b-6)
and the kinetic energy must satisfy the following inequality:
2
1 92 1
k =
(b7)
2 2 4 2
The smallest possible kinetic energy 0 of the electron consistent with the uncertainty
relations is thus obtained if the mean-squared-radius 2 is set equal to its largest possible
value of 2 . This gives
92
8 2
(c) If k = 0 , it follows from Eqs. (a-2), (a-3), and (b-7) that we have
(b8)
k 0 =
For He
1
92
2
e =
=
=
+ He
0
2 3
165
= 0, this gives the following equilibrium radius of the electron bubble:

1
(c1)
1
4
92 4
9 (1.055 1034 )2
e =
=
32
32 9.11 1031 3.75 104
= (2.91674 1036 )1/4 m 1.31 nm

(c2)*
It might be of some interest to note that, from Eq. (c-1), the corresponding minimum
kinetic energy is
Question Number 3
Page 3 of 6
25 April 2010

__________________________________________________________________________________________
1/2
= 0.100 eV
(c3)
2
(d) The condition for stable local equilibrium of the electron bubble at radius is that
when is increased by a small amount > 0, the inward force pushing on the interface
must be greater than the outward force so as to decrease the radius. Thus, from Eq. (c-1),
we obtain
0 = 402 = 3
92
2
+ He >
(d1)
( + )
16( + )5
By keeping only terms linear in after both sides of the inequality are expanded as a
power series and making use of Eq. (c-1) to eliminate He , we obtain
2
92
>
(5)
(d2)
2
16 6
Note that the same inequality is obtained if we consider a small change < 0. Using Eq.
(c-1), we may express Eq. (d-2) in terms of He as
(1)
or equivalently,
2
2
< 5 He +
(d3)
5
(e) From Eqs. (a-2), (a-6), and (b-8), we have
He >
or equivalently,
(d-4)*
2
k
0
92
+ He = e =
2 3 2 3 16 5
(e1)
92
2
(e2)
5
16
The minimum of the right-hand side of the inequality occurs when its derivative vanishes,
i.e.
He
452
2
+ 2 =0
6
16
or
1/4
452
= th =
32
Substituting the last result back into Eq. (e-2), we obtain
He
92
2
1
2 8
16 2 1/4
th
=
5 = 5 1 = 5
5
452
16th
th
th
th
(e3)
(e4)
(e-5)*
Question Number 3
Page 4 of 6
25 April 2010

__________________________________________________________________________________________
For PHe < Pth , no equilibrium is possible for the electron bubble.
Part B. Single Gas Bubble in Liquid Collapsing and Radiation

(f) When the bubbles radius changes by , the volume of the liquid displaced
interface is = 4 2 . But the total volume of the incompressible liquid
change, so the change of the volume at the outer surface of the liquid must also
Thus the amount of work done on the liquid is
= 0 = ( 0 )4 2
From Eq.(2), the change in total kinetic energy of the liquid is, in the limit 0 ,
1 1
k = 20 4 2 = 20 3 2 .
0
Since k = , we obtain
with
by the
cannot
be .
(f-1)*
f2
1
m 2 = ( 0 ) n .
2 0
f3
m = 3,
n = 2.
(f-4)*
(g) The initial gas temperature is 0 . According to the ideal gas law, the initial gas pressure
i = (i ) is thus given by
i i3 = 0 03 .
Since the process is adiabatic, the radial dependence of the gas pressure P is
i3
i 5
i 5
0 3
() = 3 i = i = 0 .
i
and the temperature corresponding to the radius is given by
(1)
i3
() = 3
(h) From Eqs. (3) and (g-2), we have
i 2
0 = 0 .
1
0 0 3
3 2
=
= 1
2 2
0
0 0
In terms of = /i and = /i , the last equation may be rewritten as
1
0 i 3
3 2 =

1 .
2
2
0 i2 0
(g-1)*
(g-2)*
(g-3)*
(h-1)
This may be integrated to give
1 3 2
0 i 23
0 i 33 1 3 1
2
=
=
2
0 1 0
0 0 3(1 )
3
(h-2)
Question Number 3
Page 5 of 6
25 April 2010

__________________________________________________________________________________________
0
i
1
(1 2 ) + 2 (1 3 ).

2
30
0 ( 1)
h3
0 (1 2 )
2 (1 3 )
=

.
(1 2 )
3i2 5
h4
2
2
m
m
3
2
(1 m ) = m 1 +
.
=
2
1 m
1 + m
(i-1)
Since i /[( 1)0 ] and = 5/3, the last equation leads to

1
0
0 2 = () 2 5 [(1 2 ) 2 (1 3 )]
2
3i
Thus we obtain
0
.
(h-5)*
3i2
(i) The radius of the bubble reaches its minimum value when = i = 0. Thus, from Eq.
(h-4), we obtain
=
The last equality shows that m must be very small in order that 1. Thus
2
m
, or m , i.e. m = 1
For i = 70 = 35.0 m, we have, from Eq. (g-1),
Therefore
i
1
0 3 3 1 3
=
= = 0.00437.
0 ( 1) ( 1) i
2 7
m = = 0.0661,
m = m = = 0.0661 35 m = 2.31 m,
and from Eq. (g-3), the corresponding temperature m is
(i-2)*
(i-3)
(i-4)
(i-5)*
2
1 2
1
m = 0 =
300 K = 6.86 104 K.
(i-6)*
m
0.0661
(j) From Eq. (h-4), the maximum value of the radial speed occurs at =
where () is also at its maximum, i.e. the derivative of () with respect to must
vanish at = . Since
0
1
1
1
() =
1 ,
(j-1)
5 3
3
3i2
we have
0
5
3
3
=
5 3 3 = 0.
(j-2)
2
=
3
Thus
32
5
2
=
,
or
.
(j-3)
(5 32 )
3 1 +
Question Number 3
Page 6 of 6
25 April 2010

__________________________________________________________________________________________
which implies
=
= 0.0852.
3 1 +
(j-4)*
The radius midway between (corresponding to maximum speed) and m

(corresponding to zero speed) is given by
1
1
(m + ) (0.0661 + 0.0852) = 0.0757
2
2
From Eq. (h-4), the dimensionless radial speed at radius is

(j-5)
2
= = ( )
0
20 (1 2 )
2
=
1
+
= 5.52 106 .
2
2 3
(1
+
30 i
(j-6)*
Thus an estimate of the duration m for the radius of the bubble to diminish from to
the minimum value m is
( m ) (0.0852 0.0661)
=
= 3.45 109 s.
(j-7)*
5.52 106
(k) Suppose the bubble is a surface radiator with emissivity . By making use of Eq. (g-3), the
radiant power r of the bubble at temperature can be written as a function of , i.e.
1
r = (SB 4 )4 2 = 4i2 SB 04 6 ,
(k-1)
m =
where SB is the Stefan-Boltzmann constant. The power supplied to the bubble is

i
=
= i
= 4i3 i 3
(k-2)*
The assumption of an adiabatic collapsing of the bubble is deemed reasonable when the
radiant power is less than 20 % of the power supplied to the bubble at = . Thus we
4T
have
or
4i2 SB 04

20 %
i
i
6
3
i i

3
3 = 0 i 0 3 = 0.0107
5SB 04
5SB 04 i
(k-3)
(k-4)*
Theoretical Question 1: The Shockley-James Paradox

In the year 1905, Albert Einstein proposed the special theory of relativity to resolve the inconsistency between
Newtons mechanics and Maxwells electromagnetism. Proper understanding of the theory led to the resolution of
many apparent paradoxes. At the time, the discussion focused mostly on the propagation of electromagnetic waves.
In this question, we solve a paradox of a different type. For a fairly simple system of charges proposed by W.
Shockley and R. P. James in 1967, understanding the conservation of linear momentum requires careful relativistic
analysis. If a point charge is located near a magnet of changing magnetization, there's an induced electric force on the
charge, but no apparent reaction on the magnet. The process may be slow enough that any electromagnetic radiation
(and any momentum carried away by it) is negligible. Thus, apparently we get a cannon without recoil.
In our analysis of this system, we will demonstrate that in relativistic mechanics, a composite body may hold a nonzero mechanical momentum while remaining stationary.
Part I: Understanding the impulse on the point charge (3.3 points)
Consider a circular current loop of radius carrying a current , and a second, larger current loop of radius
concentric with the first and lying in the same plane.
a. (1 pt.) A current
the ratio
passing through loop 2 (the larger loop) generates a magnetic flux

. It is called the mutual inductance coefficient.
through loop 1. Find
b. (0.8 pts.) Given that

, obtain the total induced EMF in the larger loop as a result of a
variation
of the current in the smaller loop. Neglect the current in the larger loop. Hint: the induced
EMF is equal to the rate of change of the magnetic flux through the loop.
c. (0.5 pts.) The EMF you found in part (b) is due to the tangential component of an induced electric field. Obtain an
expression for the tangential electric field at radius as a function of the rate of change of the current.
Figure 1: A circular current loop and a point charge .

We now remove the larger current loop, and instead put a massive point charge at radius , as shown in Figure 1. It
may be assumed that the charge moves very little during the relevant time periods.
d. (1 pt.) Find the total tangential impulse
received by the point charge as the current in the small loop changes
from an initial value
to the final value
.
Part II: Understanding the recoil of the current loop (4.4 points)
We will now understand the origin of the recoil of the loop, using a loop of different geometry.
Page 1 of 3
e. (1.1 pts.) Consider a hollow tube with walls made of a neutral insulating material of length and cross section
carrying an electric current . The current is due to charged particles of rest mass and charge distributed
homogenously inside the tube with number density . Assume that the charged particles are all moving along the
tube with the same velocity. Find the total momentum of the charged particles in the tube, taking Special
Relativity effects into account.
f.
(3.3 pts.) Consider a square current loop with side . At a distance

from the loop, there is a point charge ;
see Figure 2. The loop carries current . We will model the current loop as a neutral tube, as in part (e). The charge
carriers can move freely along the loop, colliding elastically with the walls and making elastic right turns at the
corners. Neglect all interactions among the charge carriers. Assume also that all the charge carriers at a given
section along the tube always move with the same velocity. Assume that the loop is heavy and that its motion can
be neglected. Calculate the total linear momentum
of the charge carriers in the loop. It is called "hidden
momentum".
Figure 2: A square current loop and a point charge .

When the current stops, this linear momentum is transferred to the loop, and it gets an impulse equal to minus the
impulse received by the point charge . This is the missing recoil that we were looking for (note that in the initial state
there is also momentum in the electromagnetic field; this is important for conservation of the total momentum of the
entire system).
Part III: Summarizing the results (2.3 points)
g. (0.8 pts.) Current loops are often characterized by their magnetic moment
, where is the current and is
the loops area. Express the answer to part (d) in terms of , , and . Likewise, express the answer to part (f) in
terms of , , and . Note that the electric and magnetic constants are related by:
where is the speed of light.

h. (1.5 pts.) In a more realistic model, the current loop is a conducting wire, and the field of the point charge does
not penetrate into the conductor. We assume that the current is still conducted by charge carriers inside the wire.
Decide whether each of the following statements is true or false, and circle the correct option in the Answer Form.
Note: You may leave a statement undecided, but if you decide incorrectly, you will not get credit at all for part
(h).
A. (0.5 pts.) The linear momentum of the current loop is zero.
B. (0.5 pts.) As the total current in the loop changes from to zero, the charge carriers decelerate, causing induced
currents in the wires conducting material. Because of these induced currents, the point charge will not get a net
impulse.
Page 2 of 3
C. (0.5 pts.) The surface charges on the wire, induced by the presence of the external charge, will experience an
electric force as the current changes from to zero. This way, the loop will get the same impulse as found in part
(f).
Page 3 of 3
Theoretical Question 2: Creaking Door
The phenomenon of creaking is very common, and can be found in doors, closets, chalk squeaking on a blackboard,
playing a violin, new shoes, car brakes and other systems from everyday life. Here in Israel, a similar phenomenon
causes violent earthquakes with a period of several decades. These originate in the Dead Sea rift, located right above
the deepest known break in the earth's crust.
The physical mechanism for creaking is based on elasticity combined with the difference between the static and the
kinetic friction coefficients. In this question, we will study this mechanism and its application to the case of an
opening door.
Part I: General model (7.5 points)
Consider the following system (see Figure 1):
A box with mass is attached to a long ideal spring with spring constant , whose other end is driven at a constant
velocity . The static and the kinetic friction coefficients between the box and the floor are given respectively by
and , where
.
Figure 1: A general model for creaking

We would like to understand why this setup supports two different forms of motion:
1. The friction is always kinetic. This is known as a pure slip mode.
2. Kinetic and static friction alternate. This is known as a stick-slip mode. Stick-slip motion is the source of
the creaking sound commonly encountered.
a.
(1 pt.) Consider the case where at the initial time

, the box slides on the floor with velocity , and the
springs tension exactly balances the kinetic friction. Assume
. The springs elongation will
oscillate as a function of .
a1. (0.6 points) Find the period and the amplitude of these oscillations.
a2. (0.4 points) Sketch a qualitative graph of the springs elongation
for
.
b. (1.2 pts.) Now, consider the case where at

the box is at rest, while the initial spring elongation is the same
as in part (a). Sketch a qualitative graph of the velocity
of the box with respect to the floor for
,
where is the (new) period of the oscillations
. Motion to the right corresponds to a positive sign of .
Indicate approximately on your graph the horizontal line
.
c. (0.5 pts.) For the initial conditions of part (b), find the time-averaged value of the springs elongation after a
sufficiently long time has passed.
Page 1 of 3
d. (2.4 pts.) For the conditions of part (b), find the period
of the oscillations
Generically, stick-slip motion stops at high driving velocities . We will now discuss one of the possible mechanisms
behind this effect.
e. (2.4 pts.) Suppose that during each period , a small amount of energy is dissipated into heat in the spring, via an
additional mechanism. Let
be the fractional amplitude loss per period due to dissipation in pure-slip
motion. For
, find the critical driving velocity
above which periodic stick-slip becomes impossible.
The results of part (e) are not required for part II.
Part II: Application to creaking door (2.5 points)
A door hinge is a hollow, open-ended metal cylinder with radius , height and thickness . The lower end of the
cylinder lies on a metal base attached to the wall (the area of contact is a ring of radius ); see Figure 2. The static and
the kinetic friction coefficients between the cylinder and its base are and
respectively, with
. The upper
end of the cylinder is attached to the door, which can be regarded as perfectly rigid. A typical door hangs on two or
three such hinges, but its weight is concentrated on only one of them this is the hinge that will creak. The cylinder of
that hinge presses down on its metal base with the weight of the entire door, whose mass is .
Door
Wall
Figure 2: Schematic drawing of a door

Figure 3: The twisted hinge cylinder
hinge
The cylinder is not a perfectly rigid body it can twist tangentially without changing its overall form, so that vertical
line segments become tilted with a small angle ; see Figure 3. The elastic force on a small area element
of the
base due to this deformation is given by:
,
where
f.
is a material property known as the shear modulus. Use the values

,
,
,
,
,
,
. You may use the approximation
,
.
(1 pt.) We start rotating the door very slowly from equilibrium (zero torque). For small rotation angles, obtain an
expression for the torsion coefficient
, where is the torque which must be applied to rotate the door by
an angle .
g. (1.5 pts.) At very low angular velocity, when a transition from stick to slip occurs, a sound pulse is emitted. Find
the angular velocity of the door for which the frequency of these pulses enters the audible range at
.
Page 2 of 3
Assume that the frequency

and a numerical result.
of pure-slip oscillations in the hinge is much higher:
Page 3 of 3
. Provide an expression
Theoretical Question 3: Birthday Balloon
The picture shows a long rubber balloon, the kind that is popular at birthday parties.
A partially inflated balloon usually splits into two domains of different radii. In this
question, we consider a simplified model to help us understand this phenomenon.
Consider a balloon with the shape of a long homogeneous cylinder (except for the
ends), with a mouthpiece through which the balloon can be inflated. All processes
Figure 1: A partially
will be considered isothermal at room temperature. At all times, the pressure inside
inflated birthday balloon.
the balloon exceeds the atmospheric pressure by a small fraction, so the air may
be treated as an incompressible fluid. Gravity and the balloon's weight may also be neglected. The inflation is slow
and quasistatic. In parts (a)-(d), the balloon is inflated uniformly throughout its length. We denote by and the
radius and length of the balloon before it was inflated.
a. (1.8 pts.) The balloon is held by the mouthpiece, while its other parts hang freely. Find the ratio
between the
longitudinal surface tension (in the direction parallel to the balloons axis) and the transverse surface tension
(in the direction tangent to the balloons circular cross-section).
The surface tension of a rubber film is the force that adjacent parts exert on each other, per unit length of the
boundary.
Hooke's Law is a linear approximation of real-world elasticity for small tensions. Assume that the balloons length
remains constant at , while the surface tension depends linearly on the inflation ratio
:
(
b. (1 pt.) With these assumptions, obtain an expression for the dependence of the pressure inside the balloon on the
balloon's volume . Sketch a plot of
as a function of . What is the maximal inflation pressure
resulting from Hooke's elasticity approximation?
In reality, because the inflation ratio
is large (in Figure 1, typical values of about 5 can be observed), one must
consider non-linear behavior of the rubber and changes in the balloons length. These effects allow higher inflation
pressures than predicted by part (b). In a typical balloon, the graph of
is composed of three pieces:
1. For small inflation ratios,
grows in a Hooke-like manner.
2. At
the balloons length begins to increase, and
reaches a long plateau where it grows very
slowly.
3. At some large inflation ratio, the rubber starts strongly resisting any further stretch, which leads to a sharp rise
in
.
This behavior is depicted in Figure 2.
c. (1.3 pts.) Sketch a qualitative plot of the pressure difference
as a function of for a uniformly inflated
balloon that behaves according to Figure 2. Indicate any local extremum points on your plot. Indicate also the
Page 1 of 3
points corresponding to
accuracy.
Figure 2:
and
. Find the values of
for a realistic party balloon.
at these two points with 10%
Figure 3: A plot of equation (2).
To explore the consequences of the behavior you found in part (c), we approximate
balloon with a cubic function:
for a uniformly inflated
where
and are positive constants. Assume that the volume is larger than the balloons uninflated volume
and is large enough so that the function (2) is positive in the entire physical range
. See Figure (3).
The balloon is attached to a large air reservoir maintained at a controllable pressure . It may happen that some values
of are consistent with more than one value of the volume . If the balloon suffers occasional perturbations (such as
local stretching by external forces) while held at such inflation pressure, it may jump to a state of different volume.
This will happen when it becomes energetically favorable for the entire system, consisting of the balloon, the
atmosphere and the machinery maintaining the pressure . If the pressure is slowly increased from , and sufficient
perturbations exist at every step, this explosive volume jump will happen at a certain pressure where the energy
required to move between the two states is zero. Above this pressure, going from the smaller volume to the larger
volume branch releases energy, and vice versa. This type of discontinuity is often found in nature, and is sometimes
referred to as a phase transition.
d. (2.3 pts.) By considering equation (2), obtain the value of , the volume
the volume after the jump. Express your answers using
and .
of the balloon before the jump, and
A more realistic inflating agent, such as a birthday boy, is unable to supply enough air for the instantaneous volume
change described above. Instead, air is pumped gradually into the balloon, effectively controlling the balloons volume
rather than the pressure. In this case, a new type of behavior becomes possible. If it helps to minimize the total energy
of the system, the balloon will split (given sufficient perturbations) into two cylindrical domains of different radii,
Page 2 of 3
whose lengths will gradually change. The splitting boundary itself requires energy, which you may neglect. We shall
also neglect the length of the boundary layer (these assumptions are valid for a very long balloon.)
e. (1 pt.) Sketch a qualitative graph of the pressure difference
Indicate on your axes the pressure
and the volumes
f.
as a function of , taking the split into account.

and .
(1.4 pts.) The balloon is in the volume range that supports two coexisting domains. Find the length
of the
thinner domain as a function of the total air volume . Express your answer in terms of , and the radius of
the thinner domain.
g. (1.2 pts.) The balloon is in the volume range that supports two coexisting domains. Find the latent work
that must be performed on the balloon to convert a unit length of the thin domain into the thick domain.
Express your answer in terms of , , and the radius of the thinner domain.
Page 3 of 3
Theoretical Question 1: The Shockley-James Paradox

SOLUTION
a. The magnetic field created by the large loop at its center is:
Since
by:
, this is the field throughout the area of the small loop. Therefore, the flux through the small loop is given
The mutual inductance is then given by:
b. Since
, we have:
Taking the derivative with respect to time, this becomes:
c. The EMF is work per unit charge, while the electric field is force per unit charge. Therefore:
d. The electric field from part (c) leads to a force:
Integrating over
(and disregarding the sign), we get the impulse:
e. The current can be written as:
where
is the charge carriers velocity. We therefore have:

Page 1 of 2
The momentum is then given by:
where
) )
is the Lorentz factor associated with .
f. The hidden momentum is due to the charge carriers in the two vertical sides of the loop. Let be the mass of the
charge carriers, let be their charge, and let
be the potential energy difference for a charge carrier
between the two sides. Denote the longitudinal densities and velocities of the charges in the two sides by , ,
and . Let and be the appropriate Lorentz factors. From the constant value of the current, we have:
Energy conservation for the charge carriers passing from one side to the other reads:
(
The total momentum now reads:

(
Note that all the microscopic quantities
g) In part (d), the magnetic moment is
In part (f), the magnetic moment is
, ,
)
and
have dropped out.
, and we get:
, and we get:
We see that the results are identical.

h) The answer is (A)+(C). (A) is true because
between the near side and the far side of the loop vanishes. (B)
cannot be true, because the back-reaction of the induced charges on the external charge is a higher-order effect; for
instance, it involves higher powers of . Then the conservation of center-of-mass velocity requires that (C) is true.
Page 2 of 2
Theoretical Question 2: Creaking Door

SOLUTION
a1. The motion here is pure sliding under a constant kinetic friction. This is harmonic motion with a displaced
equilibrium point. The angular frequency is given by:
From here, the period is:
The initial slope is given by:

(
Therefore, the amplitude of oscillations is:
a2. The graph is sinusoidal, as shown below.
Page 1 of 5
The initial point is given by:
This is the equilibrium point of the sine function. The students are not required to find this equilibrium point, but they
are required to understand that it is positive.
b. This will be a stick-slip graph. The humps are sinusoidal, with a non-continuous derivative at their intersections
with the horizontal segments. The peaks of the humps are higher than
, since must be the average velocity of
the box. In fact, they are also higher than
, but this is not required from the students.
c. Lets pass into the reference frame of the driven end of the spring. The position of the box is then given by minus
the elongation . The motion is an oscillation around the equilibrium position . The slip phase is sinusoidal as in
part (a), while the stick phase corresponds to motion with a constant velocity . The stick phase ends when the
elastic force balances the static friction, i.e. at
, and starts again at the symmetric point with respect to
.
Page 2 of 5
We see that the average elongation is again the sines equilibrium point:
d. Again, let us pass into the reference frame of the driven end of the spring. During the stick phase, the box traverses
a distance of:
Its velocity during this phase is , so the duration of the stick phase is:
The slip phase is a sinusoidal motion around with angular frequency . The sinusoidal period is missing a phase
of , where is given by the ratio of initial position and initial velocity with respect to the equilibrium point:
Then the length of the slip phase is:

(
))
And the total period is:
))
e. Consider again stick-slip motion in the reference frame of the driven end of the spring. During the sinusoidal slip
phase, the sines amplitude will decrease due to the dissipation. At the beginning of the slip phase, the velocity is ,
while the sine is at the phase , which we found in the solution to the previous part. Thus, the sines velocity
amplitude is
. For periodic stock-slip to occur, the sine must return to the slope . Due to the dissipation,
this will happen at a phase larger than
. In other words, dissipation shortens the stick phase. The critical case is
when stick phase shortens to zero. This will happen if the sine reaches the slope precisely at the equilibrium point,
i.e. at the phase . If the slope at
is less steep than
, the box will continue its damped sinusoidal motion
without ever reaching a stick phase again.
If it is to be killed by weak dissipation, the stick phase must be short to begin with. This corresponds to a large . The
slip phase then takes up almost an entire period of the sine wave. Thus, to a good approximation, the amplitude loss
during the slip phase is given by . The critical point is when the velocity amplitude drops from
to during
one period:
|
|
Page 3 of 5
where the LHS is the change in the amplitude due to dissipation over one period. Using the results from (d) in the
limit of small , we get:
Another derivation method based on the same reasoning is to use explicitly the initial amplitude
motion:
of the harmonic
,
A third method is to consider energy losses
in the reference frame of the springs driven end:
f. For small rotations, the lower edge of the cylinder will remain stuck to the base. When the cylinder is deformed by
an angle , a point on its upper edge shifts by a distance . This corresponds to a rotation angle
of the
door around the cylinders axis. The shear force on an area element
of the base is given by:
The corresponding torque is:
Summing over the contact area with the base, the total torque is:
Therefore, the torsion coefficient is:
The numerical result is not required from the student. Any expression which reduces to the one above in the limit
will be accepted.
g. We neglect the duration of the slip phase. Using the results of section (d) with
of linear motion, we get:
Page 4 of 5
instead of
and rotation instead
Any expression which reduces to the one above in the limit

different expressions may vary significantly, since
be checked against its expression.
will be accepted. Numerical results from such

is not really negligible. Each numerical result should
Page 5 of 5
Theoretical Question 3: Birthday Balloon

SOLUTION
a. Solution using forces:
Let the balloons radius be , and let be the pressure of the inside air. Consider the balloons rear half, and write
down the equilibrium of forces on it along the cylinders axis:
(
On the other hand, let us cut the balloon in half with a plane that runs along its axis, and consider a half-cylindrical
section of length . The equilibrium of forces in perpendicular to the cutting plane reads:
(
from which we derive
Solution using energies:

If we stretch the balloon longitudinally by length
If we inflate the balloon radially with an increment
, the energy cost is:
, the energy cost is:
The two deformations can be combined while keeping the volume fixed, if we take
i.e.
. The equilibrium state is the one where the combined energy cost
zero. This gives again the result
of such a deformation is
b. From part (a), we are reminded of the relation between surface tension and pressure:
(
The volume is related to the radius by:
So we get:
( )
The graph of
.
is a hyperbola-like function increasing from 0 at

Page 1 of 4
to an asymptotic value of
at
The maximal pressure is obtained at
c. The graph of
as a function of has the same qualitative form as
below. The graph rises from zero, then decreases, and then increases again. The points
the decreasing portion (and not on the local extrema).
as a function of , shown
and
lie in
The pressures at the two requested points are approximately given by:
(
d. The work done on the pressure-controlling mechanism during continuous inflation from volume
(
The work done on the atmosphere is:

(
)(
The condition for the jump is:
This translates into Maxwells equal-areas condition:

(
Or, equivalently:
Page 2 of 4
to volume
is:
The cubic function ( ) is symmetric around the point
The equal-areas condition is therefore satisfied at:
The volumes
and
are given by the points where:

(
This gives:
e. The range of volumes where a phase separation will occur is

. The pressure is constant throughout this
range, and equals the transition pressure . The graph of
as a function of is monotonous, with a rising piece,
a horizontal plateau at
,
, followed by another rising piece. At the start and end of the plateau, the
slope has a discontinuity, i.e. the graph has a kink.
f. The radii of the two domains correspond to the volumes and

volume of the thin domain changes linearly from to . We get:
(
Converting this into length, we have:
Page 3 of 4
. As the total volume increases from
to
, the
(
(
)
)
g. The increase in the balloons volume as a result of converting a length

(
into the thick phase is:

)
The corresponding work is:

(
Therefore:
Additional discussion (doesnt appear as part of the question):

During a realistic inflation, perturbations are not strong enough to keep the system in global equilibrium at all times.
The experimental graph increases up to , continues to increase some way beyond it, reaches a local maximum, then
decreases and settles on the plateau at . This over-increase of the pressure is responsible for the fact that inflating a
balloon is difficult during the first few puffs. After the plateau, the graph sharply increases as discussed above. The
decrease towards the plateau overshoots slightly again, reaches a local minimum and rises again to settle on the
plateau. This behavior is depicted in the graph below.
The illustration is taken from:

http://www.science-project.com/_members/science-projects/1989/12/1989-12-body.html
Page 4 of 4
13th Asian Physics Olympiad

May 01-07, 2012, New Delhi
Theory Question-I
Page 1 of 3
The Drag on a Falling Magnet

A clear and detailed discussion on eddy currents was first provided by the British
physicist Sir James H. Jeans (1877-1946) in his celebrated book The mathematical
theory of electricity and magnetism (1925). The present problem is based on electricity
and magnetism.
James H. Jeans
(1877-1946)
A small size magnet with dipole moment of

magnitude and mass is dropped through a very
long vertically held non-magnetic metallic tube as
shown in Fig. (1) (figure is not to scale). In general
the fall is governed by
=
Figure 1
(1)
Here is the acceleration due to gravity. Note

that the damping parameter is due to the
generation of eddy currents in the tube.
I.1
Obtain the terminal velocity ( ) of the magnet.
[0.5 point]
I.2
Obtain () , i.e. position of the magnet at time . Take ( = 0) = 0 and

( = 0) = 0.
[1.0 point]
We shall attempt to understand the dynamics of the fall. In order

to do this we consider in part (I.3) part (I.8) a simplified problem
of the magnet falling axially towards a fixed non-magnetic
metallic ring of radius , resistance and inductance as
shown in Fig. (2). In this problem, we shall ignore radiation
effects.
In our case it is convenient to change the reference coordinates to
a set of cylindrical ones (, , ) as shown in Fig. (2) where -axis
is the ring axis, the magnet is initially at rest at the origin and the
center of the ring is at distance 0 from the origin. Cartesian axes
(, , ) are also shown in the figure. The magnet has dipole
moment in the positive direction ( = ) where is unit
vector in direction. We will assume that during the fall,
magnetic moment remains in the same direction. The axial
component ( ) and radial component ( ) of the magnetic
field at an arbitrary point (, , ) when the magnet is at the
origin are given by
Figure 2

Theory Question-I
Page 2 of 3
0
4
3 2
1
2 + 2
2 + 2 2
0
3
=
2
4 ( + 2 )5/2
where 0 is the permeability of free space.
I.3
Let the instantaneous speed of the magnet be . Obtain the magnitude of the
induced emf ( ) in the ring.
[1.5 points]
I.4
This emf will give rise to an induced current () in the ring. Obtain the magnitude
of the instantaneous electromagnetic force ( ) on the ring in terms of .
[1.0 point]
I.5
What is the magnitude of the force on the magnet due to this ring?
[0.5 point]
I.6
Express the emf in the ring in terms of L, R and i. Do not solve for . [0.5 point]
I.7
As the magnet falls it loses gravitational potential energy. Identify the three main
forms of energy into which the gravitational potential energy is converted and
write down the expressions you would use to calculate each of the three
contributions.
[1.0 point]
Does the magnetic field of the magnet do any work in this process? Tick in the
appropriate box.
[0.5 point]
I.8
Next we will estimate the damping parameter due to the

pipe (see Eq. (1)). Take an infinitely long pipe with radius ,
small thickness , and electrical conductivity . For this and
later part, we take inductance of the pipe to be negligible. It
would help if you considered the pipe to be made of many
rings each of height , radius , small thickness and
electrical conductivity (see Fig. (3)). For simplicity, the two
ends of the pipe are at = and at = , respectively.
I.9
Obtain the resistance of an individual ring.

[0.5 point]
Figure 3

Theory Question-I
Page 3 of 3
I.10 Obtain the damping parameter due to the entire pipe in terms of , and
geometrical parameters of the ring. Since each ring is very thin, you may take
magnetic field to be constant over the thickness of the ring and equal to ( =
). Assume that at an instant , the magnet has a coordinate () with an
instantaneous speed . You should leave your answer in terms of a dimensionless
integral , involving a dimensionless variable = ( )/.
[2.0 points]
I.11 Assume that the damping constant depends on the following
= (0 , , 0 , )
where 0 is the effective resistance of a long pipe. Use dimensional analysis to
obtain an expression for . Take the dimensionless constant to be unity.[1.0 point]
The following integral may be useful:
2 + 2
1 2 + 2 1
+ Constant
2
1
(n > 1)

Theory Question II
Page 1 of 2
Chandrasekhar Limit
In a famous work carried out in 1930, the Indian Physicist Prof Subrahmanyan
Chandrasekhar (1910-1995) studied the stability of stars. The problem will help you
to construct a simplified version of his analysis.
You may find the following symbols and values useful.

Speed of light in vacuum
Plancks constant
Universal constant of Gravitation
Rest mass of electron
Rest mass of proton
= 3.00 108 m. s1
= 6.63 1034 J. s
= 6.67 1011 N. m2 . kg 2
= 9.11 1031 kg
= 1.67 1027 kg
S. Chandrasekhar
(1910-1995)
II.1. Consider a spherical star of uniform density, radius and mass . Derive an
expression for its gravitational potential energy ( ) due to its own gravitational
field (gravitational self energy).
[1.0 point]
II.2. We assume that the star is made up of only hydrogen and that all the hydrogen is in
ionized form. We consider the situation when the stars energy production due to
nuclear fusion has stopped. Electrons obey the Pauli exclusion principle and their
total energy can be computed using quantum statistics. You may take this total
electronic energy (ignoring the protonic energy) to be
2 3
3 7/3 5/3
=
2
10 42/3
where is the total number of electrons and = /2. Obtain the equilibrium
condition of the star relating its radius ( ) to its mass. This radius is called the
White Dwarf radius.
[2.0 points]
II.3. Numerically evaluate given that mass of the star is the same as the solar
mass (S = 2.00 1030 kg ).
[1.5 points]
II.4. Assuming that the electron distribution is homogeneous, obtain an order of
magnitude estimation of the average separation ( ) between electrons if the
radius of the star is as obtained in part (II.3).
[1.0 point]
II.5. Let us estimate the speed of electrons. For this purpose, assume each electron to
form a standing wave in a one-dimensional box of length . Estimate the speed of
electron () in the lowest energy state using de-Broglie hypothesis
[1.0 point]
II.6. Consider now a modification of the analysis in part (II.2). If we take electrons in the
ultrarelativistic limit ( = ), a similar analysis yields

Theory Question II
Page 2 of 2
2 3
= 4/3
5/3
4/3
Obtain the expression for the mass for which, the star can be in equilibrium in terms
of the constants provided at the beginning of the question. We call this the critical
mass ( ).
[1.5 points]
II.7. If the mass of the star is greater than the critical mass obtained in part (II.6),
state whether the star will expand or contract. Tick in appropriate box.
[0.5 point]
II.8. Calculate a numerical estimate of this critical mass in units of solar mass ( ).
(Note: Your answer may differ from Chandrasekhars famous result because of
the approximations made in this analysis)
[1.5 points]

Theory Question III

Page 1 of 4
Pancharatnam Phase
This problem deals with the two beam phenomena associated with light, its
interference, polarization and superposition. The particular context of the problem
was studied by the Indian physicist S. Pancharatnam (19341969).
S. Pancharatnam
(19341969)
Consider the experimental set up as shown in Fig. (1). Two coherent monochromatic light
beams (marked as beam 1 and 2), travelling in the direction, are incident on two narrow
slits and separated by a distance (1 2 = ). After passing through the slits the two
beams interfere and the pattern is observed on the screen . The distance between the slits
and the screen is and . Assume that the width of each slit 1 and 2 is much
smaller than the wavelength of light.
Figure 1
III.1. Let the beams 1 and 2 be linearly polarized at = 0. The corresponding
electric field vectors are given by
1 = 0 cos()
(1a)
2 = 0 cos()
(1b)
where is the unit vector along the -axis, is angular frequency of light and 0 is
the amplitude. Find the expression for the intensity of the light , that will be
observed on the screen where is the angle shown in Fig. (1). Express your answer
in terms of , , 0 , and where is the speed of light. Also, note that the
intensity is proportional to the time average of the square of the electric field. Here
you make take the proportionality constant to be . You may ignore the attenuation
in the magnitude of the electric fields with distance from the slits to any point on
the screen.
[1.0 point]
III.2. A perfectly transparent glass slab of thickness and refractive index is

Theory Question III

Page 2 of 4
introduced in the path of beam 1 before the slits. Find the expression for the
intensity of the light that will be observed on the screen. Express your answer
in terms of , , 0 , , , and .
[1.0 point]
III.3. An optical device (known as quarter wave plate (QWP)) is introduced in the
path of beam 1, before the slits, replacing the glass slab. This device changes the
polarization of the beam from the linear polarization state
1 = 0 cos()
to a circular polarization state which is given by
1 =
0 cos() + 0 sin()
(2)
2
where is the unit vector along the -axis.
Assume that the device does not introduce any additional path difference and that
it is perfectly transparent. Note that the tip of the electric field vector traces a
circle as time elapses and hence, the beam is said to be circularly polarized. We
assume that the angle is small enough so that intensity from slit one does not
depend on the angle even for polarization.
III.3.a. Find the expression for the intensity of the light that will be observed
on the screen. Express your answer in terms of , , 0 , and .
III.3.b. What is the maximum intensity ( )?
III.3.c. What is the minimum intensity ( )?
[2.0 points]
III.4.
Figure 2
Now, consider the experimental setup (see Fig. (2)) in which the beam 1 is subjected to
the device (QWP) described in part 3 and,
a linear polarizer (marked as I), between = and = which allows only
the component of the electric field parallel to an axis () to pass through. The

Theory Question III

Page 3 of 4
unit vector is defined as

= cos + sin
and,
another linear polarizer (marked as II) between = and = which
polarizes the beam back to direction.
Thus the beam 1 is back to its original state of polarization. Assume that the
polarizers do not introduce any path difference and are perfectly transparent.
III.4.a.
III.4.b.
III.4.c.
Write down the expression for the electric field of beam 1 after the first
polarizer at = [1 ( = )].
Write down the expression for the electric field of beam 1 after the
second polarizer at = [1 ( = )].
What is the phase difference () between the two beams at the slits?
[2.0 points]
The most general type of polarization is elliptical polarization. A convenient way of

expressing elliptical polarization is to consider it as a superposition of two orthogonal
linearly polarized components i.e.
= 0 cos cos() + 0 sin sin()
(3)
where and and this state of polarization are depicted in Fig. 3.
The tip of the electric field vector traces an ellipse

as time elapses. Here represents the ellipticity
and is given by
Semi-minor axis of the ellipse
tan =
Semi-major axis of the ellipse
Figure 3
Linear polarization (Eqs. (1)) and circular

polarization (Eq. (2)) are special cases of elliptical
polarization (Eq. (3)). The two parameters
( [0, ])and ( [/4, /4]) completely
describe the state of polarization.

Theory Question III

Page 4 of 4
The polarization state can also be represented by

a point on a sphere of unit radius called the
Poincare sphere. The polarization of the beam
described in Eq. (3) is represented by a point
on the Poincare sphere (see Fig. 4), then latitude
= 2 and longitude = 2. Here
is the center.
Figure 4
III.5.
Consider a point on the equator of the Poincare sphere.

Write down the electric field (Eq ) corresponding to this point.
What is its state of polarization?
[0.5 point]
Consider a point at the north pole of the Poincare sphere.
III.5.a.
III.5.b.
III.6.
III.6.a
III.6.b
Write down the electric field (NP ) corresponding to this point.

What is its state of polarization?
[0.5 point]
III.7. Now, consider the three polarization states of beam 1 as given in part 4. Let
the initial polarization (at = 0) be represented by a point 1 on the Poincare
sphere; after the optical device, let the state (at = ) be represented by point 2
and after the first polarizer (say, at = ), the state be represented by point 3 . At
= , the polarization returns to its original state which is represented by 1 .
Locate these points (1 , 2 , and 3 ) on the Poincare sphere.
[1.5 points]
III.8. If these three points (1 , 2 , and 3 from the part (III.7)) are joined by great
circles on the sphere, a triangle on the surface of the sphere is obtained (Note: A
great circle is a circle on the sphere whose center coincides with the center of the
sphere). The phase difference obtained in part 4 and the area of the curved
surface enclosed by the triangle are related to each other. Relate to .
This relationship is general and was obtained by Pancharatnam and the phase difference is called
the Pancharatnam phase.
[1.5 points]

May 01-07, 2012
I.1.
Detailed Answer &

Grading Scheme
Theory Question I
Page 1 of 3
Equation of motion for the magnet is

m
z = mg k z
For terminal velocity
z = 0
mg
vT = z =
k
which gives
I.2.
(1)
0.2 mark
0.3 mark
Rewriting Eq. (1)
dv
k
= g v(t)
dt
m
Given that v(t = 0) = 0; z(t = 0) = 0 which yields
mg
dz
(1 ekt/m ) =
dt
Z z
Zkt
mg
dz =
(1 ekt/m )dt
k
0
0
i
mg h
m kt/m
t + (e
1)
z(t) =
k
k
v(t) =
0.5 mark
0.5 mark
I.3. Method - I
: Because of the relative speed v between the magnet and the ring, in
~
the eld B = Bz k + B of the magnet, the induced emf is given by
Z
~ ~l
0.8 mark
ei = (~v B).d
ei = vBa 2a
where
Ba =
0
3pa(z0 z)
2
4 [a + (z0 z)2 ]5/2
0.4 mark
(2)
0.3 mark
Method - II
:Magnetic ux () through the ring is

Z
Bz 2d

Z a
0
p
3(z0 z)2
= 2
1 d
2
2 3/2 2 + (z z)2
0
0 4 ( + (z0 z) )
0 pa2
=
2(a2 + (z0 z)2 )3/2
d
d
Induced emf ei =
= v
dt
dz
2
3pa v(z0 z)
which gives ei = 02
2[a + (z0 z)2 ]5/2
=
0.2 mark
0.2 mark
0.3 mark
0.4 mark
0.4 mark

May 01-07, 2012
I.4.
Detailed Answer &

Grading Scheme
Theory Question I
Page 2 of 3
Bz component will cause a radially outward force on the ring and by symmetry this
0.4 mark
yields a null force.

Only B will contribute to
~
~
df
= i(d~l B)
em
~
fem = i2aBa
0.6 mark
where Ba is given by Eq. (2).

I.5.
By Newton's third law, equal and opposite force will be exerted by the ring on the
magnet. Hence the magnitude of the force on the magnet by the ring is fem .
0.5 mark
I.6.
ei = L
I.7.
Potential energy is converted to three parts:
I.8.
di
+ iR
dt
0.5 mark
(a) mv 2 /2 (kinetic energy)
0.3 mark
(b) Li2 /2 (magnetic energy)
0.3 mark
(c) i2 Rt (Joule loss due to the current in time t).
0.4 mark
The magnetic eld does no work in the process.

Yes
No
I.9.
0.5 mark
X
Resistance of the ring

R =
I.10.
2a
wz 0
0.5 mark
Now, the net force on the magnet, due to one ring at z 0 is given by
fem = (2a)iBa0
where
Ba0 =
0
3pa(z 0 z)
4 (a2 + (z 0 z)2 )5/2
0.3 mark
and i is the induced current in the ring which is given by

i=
ei
wei 0
=
z
R
2a
0.5 mark
Then the net force on the magnet due to the entire pipe is given by
Z
F =
fem =
Ba02 (2a)wdz 0 .z
0.2 mark

May 01-07, 2012
Detailed Answer &

Grading Scheme
Theory Question I
Page 3 of 3
Since the pipe is very long the limits of integration can be taken as and .
Substituting Ba0 , we get
F =
2
0
2 3
18p a w z
(z 0 z)2
dz 0
((z 0 z)2 + a2 )5
0.5 mark
Let u = (z 0 z)/a. Finally,

2 18p2 wz Z
u2
0
F =
du
2 5
4
a4
(1 + u )
Thus damping parameter

2 18p2 w Z
u2
0
k=
du
2 5
4
a4
(1 + u )
I.11.
0.5 mark
Given that,
k = f (0 , p, R0 , a)
Dimensions of various parameters involved are

[0 ] = I 2 M LT 2
[p] = IL2
0.2 mark
0.1 mark
[R0 ] = I 2 M L2 T 3
[a] = L
[k] == M T 1
0.2 mark
p2 2o
k= 4
a R0
0.5 mark
which gives

May 01-07, 2012
Detailed Answer &

Grading Scheme
Theory Question- II
Page 1 of 2
II.1. Consider a shell of width dr which is at the distance r from the centre of the star.
Let be the density of star. Gravitational Potential energy of shell is
(4r3 /3)(4r2 dr)
dEG = G
r
Z R
3 GM 2
EG =
dEG =
5 R
0
It is negative, i.e. gravitational force is radially inward.

II.2. Total energy of the star E = EG + Ee .
At equilibrium, at R = Rwd
dE
=0
dR
dEe
dEG
=
dR
dR
7/3
5/3
2
3 GM
~2 3
3
2Ne
=
2
3
5 Rwd
10me 42/3
Rwd

5/3
7/3
~2 3
2Ne
3
Rwd =
6Gme 42/3
M2
0.5 mark
0.3 mark
0.2 mark
0.2 mark
0.8 mark
1.0 marks
II.3. Since all the hydrogen is ionized, the number of protons (Np ) = Ne . Also mp me .
Hence
Ne = Np
Rwd
M
mp
~2 3
=
6Gme 42/3
0.5 mark
7/3
3
2
5/3
M 1/3 mp
Rwd = 2.28 104 km
1.0 mark
II.4. If rsep is average separation between electrons then

4 3
4 3
Ne rsep
Rwd
3
3
3
R
3
3 mp
rsep
= wd Rwd
Ne
M
rsep = 2.13 1012 m
0.2 mark
0.2 mark
0.6 mark
II.5. For a particle conned in a box of length rsep , its de Broglie wavelength dB for the

May 01-07, 2012
Detailed Answer &

Grading Scheme
Theory Question- II
Page 2 of 2
ground state can be written as

0.2 mark
dB = 2rsep
h
and momentum p =
dB
h
v
2me rsep
0.3 mark
0.2 mark
= 1.08 108 m.s1

Correction:
0.3 mark
If one takes relativistic momentum

dB = 2rsep
h
h
me v
p=
=
=p
dB
2rsep
1 v 2 /c2
h
v=r
h2
4m2e r2 + 2
c
8
= 1.06 10 m.s1
0.2 mark
0.3 mark
0.2 mark
0.3 mark
II.6. Similar to part II.2, at equilibrium

dEG
dE rel
= e
dR
dR
5/3
2
2
3 GM
3
~c 4/3
= 4/3
N
2
5 R
4
R2 e
0.3 mark
0.4 mark
For critical mass

3(53 )1/2
Mc =
16m2p
Alternatively
~c
G
3/2
0.8 mark
: Since the total energy

EG +
Eerel
3
2
GM 2 + 2/3
5
4
!
5/3
3
1
4/3
~cNe
must be minimized for equilibrium, one can argue that if coecient of 1/R is positive
then star would collapse otherwise it would expand.
II.7. For M > Mc
Expand
Contract
0.5 mark
X
II.8.
Mc = 1.36 1031 kg
= 6.8 MS
1.0 mark
0.5 mark

May 01-07, 2012
Detailed Answer
Theory Question-III
Page 1 of 3
III.1.
Let the phase dierence between two rays making an angle with z
direction be . Clearly
= t = d sin /c
(1)
[1.0 mark]
Then the intensity is given by

I () = |E1 + E2 |2
Here
|E1 + E2 |2 = |E0 cos(t) + E0 cos(t + )|2
= |E0 cos(t) (1 + cos ) E0 sin(t) sin |2

= E02 cos2 (t) 1 + 2 cos + cos2 + sin2 (t) sin2
2 cos(t) (1 + cos ) sin(t) sin ]
Since cos2 t = sin2 t = 1/2 and sin(t) cos(t) = 0, we get the intesity to be
I() = E02 (1 + cos )
where is given by Eq. (1).

(a) Alternate:
|E1 + E2 |2 = |E0 cos(t) + E0 cos(t + )|2
= |E0 cos(t) (1 + cos ) + E0 sin(t) sin |2
= |E0 A cos(t )|2
where
A2 = (1 + cos )2 + sin2
= 2 (1 + cos )
Since cos2 (t ) = 1/2, we have

I() =
2 2
E A = E02 (1 + cos )
2 0
where is given by Eq. (1).

III.2.
The beam 1 has travelled extra optical path = ( 1) w. Thus the

net phase dierence between two beams when they emerge at the angle is
[1.0 marks]
and I() = E02 (1 + cos ).
(d sin ( 1) w)
c

May 01-07, 2012
Detailed Answer
Theory Question-III
Page 2 of 3
III.3.
The two beams have travelled exactly same paths upto the slits. Thus
when they emerge at an angle , they have a net phase dierence of = d sin /c.
Then, notice
[2.0 marks]
|E1 + E2 |

2
E0

E
0
= i cos(t) + E0 cos(t + ) + j sin(t)
2
2

2

E2
1
= E0 cos(t) + cos + E0 sin(t) sin + 0 sin2 (t)
2
2

1
= E02 cos2 (t)
+ 2 cos + cos2 + sin2 (t) sin2
2

1
E2
+2 cos(t) + cos sin(t) sin + 0 sin2 (t)
2
2
Thus, after taking time averages, the intesity will be

1 1
1 2
E2
2
I() =
+ 2 cos + cos + sin + 0
2 2
2
4

1
= E02 1 + cos
2

d sin
1
2
= E0 1 + cos
c
2
h
i
The maximum value of the intensity is E02 1 + 12 .
h
i
The minimum value of the intensity is E02 1 12 .
E02
III.4.
[2.0 marks]
The electric eld at z = b is given by

E0
0
0
(cos (t kb) i i + sin (t kb) j i ) i0
E1 (z = b) =
2
1
= [E0 cos cos (t kb) + E0 sin sin (t kb)] i0
2
1
= [E0 cos (t kb )] i0
2
and at z = c
E0
E1 (z = c) = (cos (t kc ) i0 i) i
2
1
= cos E0 cos (t kc ) i
2
Where as
E2 (z = c) = E0 cos (t kc) i
So, the net phase dierence between beam 1 and beam 2 is now
Phase dierence = .
III.5.
[0.5 marks]

May 01-07, 2012
Detailed Answer
Theory Question-III
Page 3 of 3
(a) At equator e = 0. Then

E = EEq = i0 E0 cos (t) .
(b) Clearly the beam is linearly polarised along i0 .

III.6.
[0.5 marks]
(a) At north pole e = /4 and can be taken to be 0. Then i0 = i and j0 = j. The

electric eld
E0
E0
ENP = i cos (t) + j sin (t) ,
2
2
(b) Which represents circular polarisation.

III.7.
III.8.
[1.5 marks]
From gure it is clear that the area of the spherical triangle A1 A2 A3 is

2
= 2 and the phase dierence was . Thus the phase dierence is half
the area of the spherical triangle A1 A2 A3 on Poincare sphere.
Thus S = 2.
Here, the beam 1 passes through various states of polarization and returns to its
original state. Though there has been no additional path dierence, the beam has
picked up a phase with respect to the beam 2. This is called as Pancharatnam
phase.
[1.5 mark]
2
2
Problem
Conductors in Conducting Liquid

Theory
Question 1

page 1 of 1

A system consisted of two conductor bodies is immersed in a uniform dielectric
and weakly conducting liquid. When a constant voltage difference is applied
between both conductors, the system has both electric and magnetic fields. In
this problem we will investigate this system.

1. (0.4 pts) First consider an infinitely long line with charge per unit length
in vacuum. Calculate the electric field E(r) due to the line.

2. (0.4 pts) The potential due to the line charge could be written as
= + ,
where K is a constant. Determine f(r).

3. (0.7 pts) Calculate the potential in all space V(x,y,z)due to an infinitely long
line with charge per unit length at = , = 0 and another infinitely
long line with charge per unit length at = , = 0. Both lines are
parallel to the z-axis. Take = 0 at the origin. Sketch the equipotential
surfaces.

For the following questions, ignore any edge effects.

4. (2.0 pts) Now consider two identical conducting cylinders, both with radius
R = 3a in vacuum. The length of each cylinders are the same and much larger
than its radius ( ). The axis of both cylinders are on the xz-plane and
parallel to the z-axis, one at = 5 , = 0 and the other at = 5 , = 0.
An electrical potential difference of V0 is applied between the two cylinders
(the cylinder at = 5 has the higher potential) by connecting them to a
battery. Calculate the potential in all regions. Take = 0 at the origin.

5. (0.5 pts) Calculate the capacitance C of the system.

6. (1.0 pts) Now both cylinders are totally immersed in a weakly conducting
liquid with conductivity . Calculate the total current that flows between
both cylinders. Assume the permittivity of the liquid is equal to that of
vacuum, = ! .

7. (0.5 pts) Calculate the resistance R of the system. Calculate RC of the system.

8. (1.5 pts) Calculate the magnetic field due to the current in question 6.
Assume that the permeability of the liquid is equal to that of vacuum = ! .

!"#
!
Notes !! !! ! = arctan ! +
Problem
Relativistic Correction on GPS Satellite

Theory
Question 2

page 1 of 6

Global Positioning System (GPS) is a navigation technology which uses
signal from satellites to determine the position of an object (for example an
airplane). However, due to the satellites high speed movement in orbit, there
should be a special relativistic correction, and due to their high altitude, there
should be a general relativistic correction. Both corrections seem to be small but
are very important for precise measurement of position. We will explore both
corrections in this problem.
First we will investigate the special relativistic effect on an accelerated
particle. We consider two types of frame, the first one is the rest frame (called S
or Earths frame), where the particle is at rest initially. The other is the proper
frame (called ), a frame that instantaneously moves together with the
accelerated particle. Note that this is not an accelerated frame, it is a constant
velocity frame that at a particular moment has the same velocity with the
accelerated particle. At that short moment, the time rate experienced by the
particle is the same as the proper frames time rate. Of course this proper frame
is only good for an infinitesimally short time, and then we need to define a new
proper frame afterward. At the beginning we synchronize the particles clock
with the clock in the rest frame by setting them to zero, = = 0 (t is the time in
the rest frame, and is the time shown by particles clock).
By applying equivalence principle, we can obtain general relativistic
effects from special relavistic results which does not involve complicated metric
tensor calculations. By combining the special and general relativistic effects, we
can calculate the corrections needed for a GPS (global positioning system)
satellite to provide accurate positioning.

Some mathematics formulas that might be useful
! ! !! !!
sinh =
cosh =
tanh = !"#$ !
1 + sinh! = cosh!
sinh( ) = sinh cosh cosh sinh
! ! !! !!
!
!"#$ !
Problem

!"
!
!!! ! !
!"
!!! !
= ln
!
!!! !
!!!
!!!
Theory
Question 2

page 2 of 6

+
+
Part A. Single Accelerated Particle (2.8 points)

Consider a particle with a rest mass m under a constant and uniform force field F
(defined in the rest frame) pointing in the positive x direction. Initially (t == 0)
the particle is at rest at the origin (x = 0).
1. (0.5 pts) When the velocity of the particle is v, calculate the acceleration of
the particle, a (with respect to the rest frame).
2. (0.5 pts) Calculate the velocity of the particle () =
!(!)
!
at time t (in rest
frame), in terms of F, m, t and c.

3. (0.3 pts) Calculate the position of the particle x(t) at time t, in term of F, m, t
and c.
4. (0.7 pts) Show that the proper acceleration of the particle, ! = /, is
a constant. The proper acceleration is the acceleration of the particle
measured in the instantaneous proper frame.
5. (0.4 pts) Calculate the velocity of the particle (), when the time as
experienced by the particle is . Express the answer in g, , and c.
6. (0.4 pts) Also calculate the time t in the rest frame in terms of g, , and c.

Part B. Flight Time (2.0 points)

The first part has not taken into account the flight time of the information to
arrive to the observer. This part is the only part in the whole problem where the
flight time is considered. The particle moves as in part A.
1. (1.2 pts) At a certain moment, the time experienced by the particle is .
What reading ! on a stationary clock located at = 0 will be observed by
the particle? After a long period of time, does the observed reading !
approach a certain value? If so, what is the value?
2. (0.8 pts) Now consider the opposite point of view. If an observer at the
initial point (x = 0) is observing the particles clock when the observers time
Problem

Theory
Question 2

page 3 of 6

is t, what is the reading of the particles clock ! ? After a long period of time,
will this reading approach a certain value? If so, what is the value?

Part C. Minkowski Diagram (1.0 points)

In many occasion, it is very useful to illustrate relativistic events using a
diagram, called as Minkowski Diagram. To make the diagram, we just need to
use Lorentz transformation between the rest frame S and the moving frame S
that move with velocity = with respect to the rest frame.

= ! + ! ,
!
ct
ct
= + ,
! = ,
x
! = .
=
1
!
1 !

Lets choose x and ct as the orthogonal axes. A point (x, ct) = (1,0) in the moving
frame S has a coordinate (x, ct) = , in the rest frame S. The line connecting
this point and the origin defines the x axis. Another point (x, ct) = (0,1) in the
moving frame S has a coordinate (x, ct) = , in the rest frame S. The line
connecting this point and the origin defines the ct axis. The angle between the x
and x axis is , where tan = . A unit length in the moving frame S is equal to
1 + ! =
!!! !
!!! !
in the rest frame S.
To get a better understanding of Minkowski diagram, let us take a look at this

example. Consider a stick of proper length L in a moving frame S. We would like
to find the length of the stick in the rest frame S. Consider the figure below.
Problem

Theory
Question 2

page 4 of 6

The stick is represented by the segment AC. The length AC is equal to
!!! !
!!! !
in
the S frame. The stick length in the S frame is represented by the line AB.

AB = AD BD
= cos sin tan
= 1 !
1. (0.5 pts) Using a Minkowski diagram, calculate the length of a stick with
proper length L in the rest frame, as measured in the moving frame.
2. (0.5 pts) Now consider the case in part A. Plot the time ct versus the
!"
position x of the particle. Draw the x axis and ct axis when ! = 1 in the
same graph using length scale ( ! ) and ( ! ) .

Part D. Two Accelerated Particles (2.3 points)

For this part, we will consider two accelerated particles, both of them have the
same proper acceleration g in the positive x direction, but the first particle starts
from x = 0, while the second particle starts from x = L. Remember, DO NOT
consider the flight time in this part.
1. (0.3 pts) After a while, an observer in the rest frame make an observation.
The first particles clock shows time at ! . What is the reading of the second
clock ! , according to the observer in the rest frame.
2. (1.0 pts) Now consider the observation from the first particles frame. At a
certain moment, an observer that move together with the first particle
Problem

Theory
Question 2

page 5 of 6

observed that the reading of his own clock is ! . At the same time, he
observed the second particles clock, and the reading is ! . Show that
!
sinh ! ! = ! sinh
,
where C1 is a constant. Determine C1.

3. (1.0 pts) The first particle will see the second particle move away from him.
Show that the rate of change of the distance between the two particles
according to the first particle is
sinh !
= !
,
!
cosh (! ! )
where C2 is a constant. Determine C2.

Part E. Uniformly Accelerated Frame (2.7 points)

In this part we will arrange the proper acceleration of the particles, so that the
distance between both particles are constant according to each particle. Initially
both particles are at rest, the first particle is at x = 0, while the second particle is
at x = L.
1. (0.8 pts) The first particle has a proper acceleration g1 in the positive x
direction. When it is being accelerated, there exists a fixed point in the rest
frame at x=xp that has a constant distance from the first particle, according to
the first particle thoughout the motion. Determine xp.
2. (1.3 pts) Given the proper acceleration of the first particle is g1, determine
the proper acceleration of the second particle g2, so that the distance
between the two particles are constant according to the first particle.
3. (0.6 pts) What is the ratio of time rate of the second particle to the first
particle
d 2
, according to the first particle.
d 1
Part F. Correction for GPS (2.2 points)

Part E indicates that the time rate of clocks at different altitude will not be the
same, even though there is no relative movement between those clocks.
Problem

Theory
Question 2

page 6 of 6

According to the equivalence principle in general relativity, an observer in a
small closed room could not tell the difference between a gravity pull g and the
fictitious force from accelerated frame with acceleration g. So we can conclude
that two clocks at different gravitational potential will have different rate.
Now let consider a GPS satellite that orbiting the Earth with a period of 12 hours.
1. (0.6 pts) If the gravitational acceleration on the Earths surface is 9.78 m.s-2,
and the Earths radius is 6380 km, what is the radius of the GPS satellite
orbit? What is the velocity of the satellite? Calculate the numerical values of
the radius and the velocity.
2. (1.2 pts) After one day, the clock reading on the Earth surface and the
satellite will differ due to both special and general relativistic effects.
Calculate the difference due to each effect for one day. Calculate the total
difference for one day. Which clock is faster, a clock on the Earths surface or
the satellites clock?
3. (0.4 pts) After one day, estimate the error in position due to this effect?

Problem
Physics of Spin

Theory
Question 3

page 1 of 4

All matters in the universe have fundamental properties called spin, besides
their mass and charge. Spin is an intrinsic form of angular momentum carried by
particles. Despite the fact that quantum mechanics is needed for a full treatment
of spin, we can still study the physics of spin using the usual classical formalism.
In this problem, we are investigating the influence of magnetic field on spin using
its classical analogue.
The classical torque equation of spin is given by
=
= .
In this case, the angular momentum represents the intrinsic spin of the
particles, is the magnetic moment of the particles, and is magnetic field. The
spin of a particle is associated with a magnetic moment via the equation
=
where is the gyromagnetic ratio.
In this problem, the term frequency means angular frequency (rad/s), which
is a scalar quantity. All bold letters represent vectors; otherwise they represent
scalars.

Part A. Larmor precession (1.6 points)

1. (0.8 pts) Prove that the magnitude of magnetic moment is always constant
under the influence of a magnetic field . For a special case of stationary
(constant) magnetic field, also show that the angle between and is
constant.
(Hint: You can use properties of vector products.)
2. (0.8 pts) A uniform magnetic field exists and it makes an angle with a
particle's magnetic moment . Due to the torque by the magnetic field, the
magnetic moment rotates around the field , which is also known as
Larmor precession. Determine the Larmor precession frequency ! of the
magnetic moment with respect to = ! .

Part B. Rotating frame (3.4 points)

In this section, we choose a rotating frame as our frame of reference. The
rotating frame ! = ( ! , ! , ) rotates with an angular velocity as seen by an
observer in the laboratory frame = (, , ), where the axes , , intersect
with , , at time = 0. Any vector = ! + ! + ! in a lab frame can be
written as = ! + ! + ! in the rotating frame . The time derivative
of the vector becomes
! ! ! ! ! !
!
!
!
=
+
+
+ !
+ !
+ !

Problem
Physics of Spin

where
!"#
!"#
+ ,
is the time derivative of vector seen by an observer in the lab
!" !"#
!
frame, and
Theory
Question 3

page 2 of 4
!" !"#
is the time derivative seen by an observer in the rotating
frame. For all the following problems in this part, the answers are referred to the
rotating frame S.
1. (0.8 pts) Show that the time evolution of the magnetic moment follows the
equation
= !"" ,
!"#
!
where !"" = ! is the effective magnetic field.

2. (0.4 pts) For = ! , what is the new precession frequency in terms of
! and ?
3. (1.2 pts) Now, let us consider the case of a time-varying magnetic field.
Besides a constant magnetic field, we also apply a rotating magnetic field
= cos + sin , so = ! + . Show that the new Larmor
precession frequency of the magnetic moment is
=
+ ! .
4. (1.0 pts) Instead of applying the field = cos + sin , now we

apply = cos sin , which rotates in the opposite direction
and hence = ! + cos sin . What is the effective magnetic
field !"" for this case (in terms of the unit vectors ! , ! , )? What is its time
average, !"" (recall that cos 2/ = sn 2/ = 0)?

Part C. Rabi oscillation (3.0 points)

For an ensemble of N particles under the influence of a large magnetic field,
the spin can have two quantum states: up and down. Consequently, the total
population of spin up and down obeys the equation
+ = .
The difference of spin up population and spin down population yields the
macroscopic magnetization along the axis:
= = ! .
In a real experiment, two magnetic fields are usually applied, a large bias field
! and an oscillating field with amplitude 2 perpendicular to the bias field
( ! ). Initially, only the large bias is applied, causing all the particles lie in the
spin up states ( is oriented in the -direction at = 0). Then, the oscillating field
Problem
Physics of Spin

Theory
Question 3

page 3 of 4

is turned on, where its frequency is chosen to be in resonance with the Larmor
precession frequency ! , i.e. = ! . In other words, the total field after time
= 0 is given by
() = ! + 2 cos ! .
1. (1.2 pts) In the rotating frame ! , show that the effective field can be
approximated by
! ,
which is commonly known as rotating wave approximation. What is the
precession frequency in frame ?
2. (0.6 pts) Determine the angle that makes with !"" . Also, prove that the
magnetization varies with time as
= cos .
3. (1.2 pts) Under the application of magnetic field described above, determine
the fractional population of each spin up = / and spin down
= / as a function of time. Plot () and () on the same graph vs.
time t. The alternating spin up and spin down population as a function of
time is called Rabi oscillation.

Part D. Measurement incompatibility (2.0 points)

Spin is in fact a vector quantity; but due to its quantum properties, we cannot
measure each of its components simultaneously (i.e. we can know both and
! as in above problems; but not all , ! , ! , and ! simultaneously). In this
problem, we will do a calculation based on the Heisenberg uncertainty principle
(using the relation ! ) to show how these measurements are
incompatible with each other.

1. (1.0 pts) Let us consider an oven source of silver atoms, which has a small
opening. The atoms stream out of the opening along y direction (see Figure
below) and experience a spatial varying field ! . The field ! has strong bias
field component in the z direction, where the atoms with different magnetic
moment ! = are split in the z direction. At a distance from the oven
source, a screen ! is put to allow only spin up atoms to pass (blocking spin
down atoms). Thus, at the instant after passing the screen, the atoms are
prepared in spin up states. After the screen, the atoms enter a region of non-
homogenous field ! where the atoms feel a force
! = ! .
The field ! has strong bias field component in the x direction, where the
atoms have magnetic moment ! = .
After that, the atoms enter a region of non-homogenous field B2 (x, y, z) = (B0 + Cx)i. In this region, the
atoms feel a force in the x direction, given by
Theory
Question 3

Physics
o
f
S
pin
Within the classical argument, the magnetic moment of the atoms are restricted topage
their 4
quantized
of 4 values
F =
dB
x
Pxroblem
dx
x = ~ due to the bias field in x direction.
= x C.

In order to determine ! by observing the splitting in x direction, show that
In order
tofollowing
determine condition
after
t leaving the screen SC1 , the following condition must be
x , show that
the
must
be time
fulfilled:
fulfilled:
11
| xCt
|x
1, 1,
~ !
where where
x is the
opening
width on screen
1.
is
the duration
after SC
leaving
the screen ! and is the opening
width
on atoms
! . are initially prepared in the spin up states right after leaving the screen, where
2. (0.8 pts)
The
z2.= (1.0
~ = |
theare
atoms
will precess
at ratesin covering
a range
of values
! after
with respect to
x |. This
pts)
The means
atoms
initially
prepared
the spin
up states
right
the field B2 .
leaving the screen, where ! = = ! . This means the atoms will precess
Prove that, if previous condition in measuring x is satisfied, the spread in precession angle !t is so
at rates
covering
a range
of values
with respect
to twe
he cannot
x component
f !, and
large that
the spin
z component
is completely
randomized
and hence
measure oboth
x
z
specifically
=
+
.
P
rove
t
hat
t
he
s
pread
i
n
t
he
p
recession
a
ngle

simultaneously.
In other
words,
the
measurement
of
destroys
the
information
on
.
x
z
!!
!
is so large and hence we cannot measure both ! and ! simultaneously. In
other words, the measurement of ! destroys the information on ! .
3
Theoretical 1: Solution
1. Using Gauss law
I
E dA =
q
.
0
(1)
From symetry we know that the electric field only has radial component. Choose a cylinder
(with a line charge as the axis) as the Gaussian surface, we obtain
E.2rl =
l
.
0
Simplify to obtain
E = r
.
20 r
(2)
2. The potential is given by

Z
V =
E dl
Zrefr
E.dr
ref
V =
ln r + K,
20
(3)
so f (r) = 2
ln r. where K is a constant.
0
3. The potential from both line charges is a superposition of both potential
!
!
y$
!
!
!
r1!
r2!
!
!
!!
!!
!
$b!
b!
!
!
Figure 1: System with two line charges
!
!
!
V =
ln r1 +
ln r2
!
20
20
p
!
(b x)2 + y 2
=
ln p
!
20
(b + x)2 + y 2
!
(b x)2 + y 2
!
V =
ln
.
40 (b + x)2 + y 2
!
!
!
Conductors in Conducting
Liquid
!
!
!
!
x$
(4)
(5)
Page 1 of 6
We can rearrange eq.(5) in to:
!

2

2
1+
1
+
x
+ y 2 = b2
1
1
1

where = exp 40 V . For an arbitrary potential V , Eq. (6) is an equation of circle.
Figure 2: The equipotential surfaces with b = 1, for = 12.35 (left) and =
1
12.35
(6)
(right)
4. From eq.(5) and eq.(6), we see that for any arbitrary potential V , the equipotential surfaces
of these two equal but opposite lines charge, are cylindrical surfaces. From this observation,
we can choose the specific position for each line charge in both cylinders so that the surface of
each cylinder is an equipotential surface.
!
Consider
the following figure
r1!
r2!
!!
l1!
R"
!!
!!
l2!
Figure 3: Two line charges with its equipotential surfaces

We would like to find a cylindrical equipotential surface enclose one line charge, let say the
(if we could find the surface, by symmetry, we surely can find the identical one that enclose
the line ). The potential is given by
ln r1 +
ln r2
20
20
=
ln(l12 + R2 2l1 R cos ) +
ln(l22 + R2 2l2 R cos ).
40
40
V =
(7)
Page 2 of 6
Since the surface of the cylinder has to be the equipotential surface, so the potential should
not depend on , i.e. V
= 0.
2l1 R sin
2l2 R sin
+
=0
40 l12 + R2 2l1 R cos 40 l22 + R2 2l2 R cos
l1
l2
= 2
2
2
2
l1 + R 2l1 R cos
l2 + R 2l2 R cos
(8)
l12 l2 + R2 l2 2l1 l2 R cos = l1 l22 + R2 l1 2l1 l2 R cos

l1 l2 (l1 l2 ) = R2 (l1 l2 )
l1 l2 = R2 .
(9)
l1 + l2 = 10a,
(10)
l1 l2 = 9a2 .
(11)
l1 = 5a 4a.
(12)
l1 = 9a,
(13)
l2 = a.
(14)
From the data in the problem, we have
Solve this quadratic equation to get

However, since l1 > l2 , we have
Using this results on eq.(5), we have

V =
(4a x)2 + y 2
ln
.
40 (4a + x)2 + y 2
(15)
This is the potential in all region except inside both cylinders. For cylinders at x = 5a, the
potential is constant and equal to
V (x = 2a, y = 0) =
(4a + 2a)2 + 02
ln
=
ln 3.
2
2
40 (4a 2a) + 0
20
(16)
For cylinders at x = 5a, the potential is constant and equal to

V (x = 2a, y = 0) =
(4a 2a)2 + 02
ln
ln 3.
=
2
2
40 (4a + 2a) + 0
20
(17)
The potential difference between both cylinders are

V =
ln 3 V0 .
0
(18)
Page 3 of 6
Substituting this results in the potential equation, the potential outside the two cylinders are:
V =
V0
(4a x)2 + y 2
ln
.
4 ln 3 (4a + x)2 + y 2
(19)
And the potential inside the cylinders are:

The potential inside the cylinder centered at (x = 5a, y = 0) is V = V0 /2.
The potential inside the cylinder centered at (x = 5a, y = 0) is V = V0 /2.
5. From eq.(18), we have
q
ln 3,
l0
(20)
q
l0
=
V0
ln 3
(21)
V0 =
so we get
C=
6. The electric field produces by both cylinders are

V0
4a + x
4a x
Ex =
+
.
2 ln 3 (4a + x)2 + y 2 (4a x)2 + y 2

V0
y
y
Ey =
.
2
2
2 ln 3 (4a + x) + y
(4a x)2 + y 2
(22)
(23)
The volume current density is given by

J = E
(24)
To calculate the total current, we may choose to calculate the current that flow through the
x = 0 plane. On this plane, there is no current in the y direction. The total current is given by
Z
I = J dA
(25)
Z
= Ex ldy
Z
8aV0
dy
= l
2 ln 3 (4a)2 + y 2
V0 l
I=
(26)
ln 3
7. The resistance is given by
R=
V0
ln 3
=
I
l
and therefore
RC =
0
(27)
(28)
Page 4 of 6
8. Since the system has a high symmetry, we may use Amperes law. The magnetic field should
not have any z dependence, since the current has no z dependence.
Figure 4 shows the current density J flow from one cylinder to the other cylinder. Choose an
Ampere loop on a constant x plane in a symmetrical way, so that the first path is pointing in
the positive z direction with constant y coordinate, the second path is pointing to the negative
y direction with constant z coordinate. The third path is pointing to the negative z direction,
but with constant y coordinate. The fourth path is pointing in the positive y direction with
constant z coordinate.
Having this path, we need to calculate the current that flow through the loop
Z
I = J dA
Z
= Jx ldy

Z
V0 l y
4a + x
4a x
=
+
dy
2 ln 3 y (4a + x)2 + y 2 (4a x)2 + y 2
!
!
!
!
!
!
!
!
!
!
!
!
z$
x$
2!
!
1!
J"
3!
4!
Figure 4: The Ampere loop

V0 l
I=
ln 3

arctan
y
y
+ arctan
4a + x
4a x
Using the Amperes law

I
B dl = 0 I

0 V0 l
y
y
2Bz l =
arctan
+ arctan
ln 3
4a + x
4a x

V0
y
y
B z = 0
arctan
+ arctan
2 ln 3
4a + x
4a x
(29)
(30)
(31)
Page 5 of 6
therefore
B = z
0 V0
2 ln 3

arctan
y
y
+ arctan
4a + x
4a x

(32)
Page 6 of 6
Relativistic Correction on GPS Satelitte
Part A. Single accelerated particle
1. The equation of motion is given by
d
(mv)
dt
mc
F =
=
(1)
(1 2 ) 2
F = 3 ma,
where =
1
1 2
(2)
and = vc . So the acceleration is given by

a=
F
.
3m
(3)
2. Eq.(3) can be rewritten as

d
F
= 3
dt
m
Z t
d
F
dt
3 =
mc 0
(1 2 ) 2
c
Z
0
Ft
p
=
mc
1 2
(4)
Ft
= q mc .
Ft 2
1 + mc
(5)
3. Using Eq.(5), we get

Z
F tdt
q

0 m 1 + Ft 2
mc
s

2
2
mc
Ft
x=
1+
1 .
F
mc
dx =
0
(6)
4. Consider the following systems, a frame S is moving with respect to another frame S, with
velocity u in the x direction. If a particle is moving in the S frame with velocity v 0 also in x
direction, then the particle velocity in the S frame is given by
v=
u + v0
0 .
1 + uv
c2
(7)
Page 1 of 10
If the particles velocity changes with respect to the S frame, then the velocity in the S frame
is also change according to
dv =
dv 0
u + v 0 udv 0
0 2
uv 0
c2
1 + c2
1 + uv
c2
dv =
1
dv 0
.
2 1 + uv20 2
c
(8)
The time in the S frame is t0 , so the time in the S frame is given by

ux0
t = t0 + 2 ,
c
(9)
so the time change in the S frame will give a time change in the S frame as follow

uv 0
0
dt = dt 1 + 2 .
c
(10)
The acceleration in the S frame is given by

a=
a0
1
dv
= 3
.
dt
1 + uv20 3
c
(11)
If the S frame is the proper frame, then by definition the velocity v 0 = 0. Substitute this to
the last equation, we get
a0
(12)
a = 3.
Combining Eq.(3) and Eq.(12), we get

a0 =
F
g.
m
(13)
5. Eq.(3) can also be rewritten as

d
g
= 3
d
Z
Z
d
g
=
d
2
c 0
0 1
!
1
g
p
+p
=
2
2
c
1
1
s
g
1+
=ec
1
g

g
g
g
e c + e c = e c e c
g
= tanh .
c
c
ln
(14)
(15)
(16)
Page 2 of 10
6. The time dilation relation is
dt = d.
(17)
g
1
= cosh .
=p
c
1 2
(18)
From eq.(16), we have
Combining this equations, we get

t
dt =
0
d cosh
0
g
c
c
g
t = sinh .
g
c
(19)
Part B. Flight Time

1. When the clock in the origin time is equal to t0 , it emits a signal that contain the information
of its time. This signal will arrive at the particle at time t, while the particle position is at
x(t). We have
c(t t0 ) = x(t)
s
2
c
gt
t t0 = 1 +
1
g
c
t0 2
t=
2 1
gt0
c
gt0
c
(20)
(21)
When the information arrive at the particle, the particles clock has a reading according to
eq.(19). So we get
gt
c
g
t0 2 c0
sinh
=
g
c
2 1 gtc0

1 gt0 2 gt0
g
g
0=
1 + sinh
+ sinh
2
c
c
c
c
gt0
g
g
= 1 + sinh
cosh .
c
c
c
(22)
Using initial condition t = 0 when = 0, we choose the negative sign

gt0
g
g
= 1 + sinh
cosh
c
c
c

c
g
t0 =
1e c .
g
(23)
As , t0 = gc . So the clock reading will freeze at this value.
Page 3 of 10
2. When the particles clock has a reading 0 , its position is given by eq.(6), and the time t0 is
given by eq.(19). Combining this two equation, we get
r

c2
2 g0
x=
1 + sinh
1 .
(24)
g
c
The particles clock reading is then sent to the observer at the origin. The total time needed
for the information to arrive is given by
c
g
c
=
g
c
t=
g
c
0 =
g
t=
g0 x
sinh
+
c
c

g0
g0
sinh
+ cosh
1
c
c
g0

e c 1

gt
ln
+1 .
c
(25)
(26)
(27)
The time will not freeze.

Part C. Minkowski Diagram
1. The figure below show the setting of the problem.
The line AB represents theq
stick with proper length equal L in the S frame.
2
The length AB is equal to 1
L in the S frame.
1+ 2
The stick length in the S frame is represented by the line AC
!
!
!
!
!
!
!
!
!
!
!
!
ct#
ct#
x#
C!
!!
A!
x#
!!
B!
3. The!position!of!the!particle!is!given!by!eq.!(5).!!
Figure 1: Minkowski Diagram
!
!
!
!!
p
AB
!
AC =
= 1 2 L.
!
cos
!
2. The position
of the particle is given by eq.(6).
!
!
!
!
Relativistic Correction
on GPS Satelitte
!
!
!
!
(28)
Page 4 of 10
!!
!" ! !!
!
!"!
!!
!!
! ! !!
!
Figure 2: Minkowski Diagram

Part D. Two Accelerated Particles
1. B = A .
2. From the diagram, we have
tan = =
ct2 ct1
.
x2 x1
(29)
Using eq.(6), and eq.(19) along with the initial condition, we get
c2
g
c2
x2 =
g
x1 =

g1
1 ,
c

g2
cosh
1 + L.
c

cosh
(30)
(31)
Using eq.(16), eq.(19), eq.(30) and eq.(31), we obtain

g2
g1
c
c
c
sinh
sinh
g
c
g
c
g1
tanh
=
c2

2
g
c
2
c
L + g cosh c 1 g cosh gc1 1
=
gL
g1
sinh
c2
c
gL
g1
sinh
c2
c
So C1 =
sinh gc2 sinh gc1

gL
c2
+ cosh gc2 cosh gc1

g2
g1
g2
g1
= sinh
cosh
cosh
sinh
c
c
c
c
g
= sinh (2 1 ) .
c
(32)
gL
.
c2
Page 5 of 10
g 2
g
sinh 1
c
c
!
!!
=
gL
g 2
g 1
+ cosh
cosh
c2
c
c
gL
g
g
g
g
g
sinh 1 = sinh 2 cosh 1 cosh 2 sinh 1
2
c
c
c
c
c
c
Theoretical
2:
Solution
Using!identity!relation,!the!last!equation!is!simply!to!
gL
g
g
Relativistic
!
(15)!
sinh 1 = Correction
sinh ( 2 1 ) !! on GPS Satelitte
2
c
c
c
sinh
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
t2!
!!
t1!
!!
x2!
x1!
!
Figure 3: Minkowski Diagram for two particles
3. From!the!length!contraction,!we!have!
x x
L' = 2 1
3. From the length contraction, we have 1
!
!!
x1 d
d
dx
dx
1
x
x
d
0 L x2
2
2
1
2
1
1
L ==
d0 1 d1 2 d 1 d1 1
12 d 1
dL
=
d1
dx2 d2 dx1
d2 d1
d1
1
x2 x1 d1
.
1
12 d1
(33)
(34)
Take derivative of eq.(30), eq.(31) and eq.(32), we get

g1
dx1
= c sinh
,
d1
c
dx2
g2
= c sinh
,
d2
c

gL
g1
g
d2
cosh
= cosh (2 1 )
1 .
c2
c
c
d1
(35)
(36)
(37)
The last equation can be rearrange to get

gL
cosh gc1
d2
c2
=
+ 1.
d1
cosh gc (2 1 )
(38)
Page 6 of 10
From eq.(29), we have
c
c(t2 t1 )
=
x2 x1 =
1
tanh gc1
c
g2
c
g1
sinh
sinh
g
c
g
c

.
(39)
Combining all these equations, we get

dL1
=
d1
!
gL
g
g2 c2 cosh c1
1
g1
g2
c sinh
c sinh
+ c sinh
c cosh gc (2 1 )
c
c
cosh gc1
1
c2
g2
g1
1
g
g1
sinh
sinh
sinh
g1
g1
2
g
c
c tanh c cosh c c
c
sinh gc2
dL1
gL
.
=
d1
c cosh gc (2 1 )
So C2 =
(40)
gL
c .
Part E. Uniformly Accelerated Frame

1. Distance from a certain point xp according to the particles frame is
L0 =
0
L =
x xp
c2
g1
(41)

cosh g1c 1 xp
cosh g1c
2
c
c2
g + xp
L =
1 g1 .
g1 cosh c
0
(42)
For L0 equal constant, we need xp = gc1 .

2. First method: If the distance in the S frame is constant = L, then in the S frame the length
is
s
1 + 2
Ls = L
.
(43)
1 2
So the position of the second particle is
x2 = x1 + Ls cos
s
s

2

2
c
g1 t1
g1 t1
1+
1 + L 1 +
=
g1
c
c
2
s

c
g1 t1 2 c2
x2 =
+L
1+
.
g1
c
g1
(44)
(45)
Page 7 of 10
!
2
Theoretical
c
g t 2: cSolution
x2 = + L 1+ 1 1 !!
Correction
c
g1
g1
Relativistic
on GPS
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
(17)!
Satelitte
t2!
L"
!!
t1!
!
x1!
x2!
!
Figure 4: Minkowski Diagram for two particles
the!time!of!the!second!particle!is!
ct 2 = ct1 + LS sin
g1t1
ct2 = ct1 L+ Ls sinc
s
2
gt 2
1 + L1+ 1 +1 1 p

=
ct
1 c 2 1!! + 2
!
= ct1 +

g1 L
ct2 = t1 c +
.
c1
Substitute eq.(47) to eq.(45) to get

gt
1+ 1 1
c
v
!
u
The time of the second particle is
2
u
g1 t2
c2
c2
t
g
L
1
x
=
+
L
1
+
2
ct
=
ct
1+
!
g1 L
!!
1
g1 2
g1
c 2 c 1 + c2
v
!2
2
u
Substitute!eq.(18)!to!eq.(17)!to!get!
u
c2
c
g1 t2
t
x2 =
+L
1+
.
g1
g1
1 + gc12L c
(46)
(47)
(18)!
(48)
From the last equation, we can identify

g2
g1
1+
g1 L
c2
(49)
Page 8 of 10
As for confirmation, we can subsitute this relation to the second particle position to get
s

2
g2 t2 2 c2
c
1+
(50)
x2 =
.
g2
c
g1
Second method: In this method, we will choose g2 such that the special point like the one
descirbe in the question 1 is exactly the same as the similar point for the proper acceleration
g1 .
For first particle, we have xp1 g1 = c2
For second particle, we have (L + xp1 )g2 = c2
Combining this two equations, we get
g2 =
g2 =
c2
2
L + gc1
g1
1+
g1 L
c2
(51)
3. The relation between the time in the two particles is given by eq.(47)

g1 L
t2 = t1 1 + 2
c

2
2
c
c
g1 L
g2 2
g1 1
=
1+ 2
sinh
sinh
g2
c
g1
c
c
g2 2
g1 1
sinh
= sinh
c
c
g2 2 = g1 1
d2
g1
g1 L
=
=1+ 2 .
d1
g2
c
(52)
(53)
Part F. Correction for GPS

1. From Newtons Law
GM m
= m 2 r
r2
2 2 13
gR T
r=
4 2
(54)
(55)
r = 2.66 107 m.
The velocity is given by

v = r =
2gR2
T
13
(56)
= 3.87 103 m/s.
Page 9 of 10
2. The general relativity effect is
dg
=1+
dt
dg
=1+
dt
U
mc2
gR2 R r
.
c2 Rr
(57)
(58)
After one day, the difference is

gR2 R r
T
c2 Rr
= 4.55 105 s.
(59)
g =
The special relativity effect is

r
ds
v2
= 1 2
dt
c
v
u
2 !

u
2 3
2gR
1
= t1
T
c2
2 !

1
2gR2 3 1
1
.
2
T
c2
(60)
(61)
After one day, the difference is

1
s =
2
2gR2
T
23 !
1
T
c2
(62)
= 7.18 106 s.
The satelites clock is faster with total = g + s = 3.83 105 s.
3. L = c = 1.15 104 m = 11.5km.
Page 10 of 10
Physics of Spin
Part A. Larmor Precession
1. From the two equations given in the text, we obtain the relation
d
= B.
dt
Taking the dot product of eq (1). with , we can prove that
d
= ( B) ,
dt
d ||2
= 0,
dt
= || = const.
(1)
(2)
Taking the dot product of eq. (1) with B, we also prove that
d
= B ( B) ,
dt
d
B
= 0,
dt
B = const.
(3)
An acute reader will notice that our master equation in (1) is identical to the equation of
motion for a charged particle in a magnetic field
dv
q
= v B.
(4)
dt
m
Hence, the same argument for a charged particle in magnetic field can be applied in this case.
2. For a magnetic moment making an angle of with B,
d
dt
d
|| sin
dt
d
0 =
dt
= B,
= || B0 sin ,
= B0 .
(5)
Part B. Rotating frame

1. Using the relation given in the text, it is easily shown that

d
d
=
dt rot
dt lab
= B k0

0
= B k
= Beff .
(6)
Note that k is equal to k0 as observed in the rotating frame.
Physics of Spin
Page 1 of 5
Physics of Spin
2. The new precession frequency as viewed on the rotating frame S 0 is
~ = (0 ) k0 ,
= B0 .
(7)
3. Since the magnetic field as viewed in the rotating frame is B = B0 k0 + bi0 ,

k0 + bi0 ,
Beff = B /k0 = B0
and
= |Beff | ,
s

2
=
B0
+ b2 .
(8)
4. In this case, the effective magnetic field becomes

Beff = B /k0

k0 + b(cos 2ti0 sin 2tj0 )

=
B0

which has a time average of Beff = B0 k0 .
(9)
Part C. Rabi oscillation

1. The oscillating field can be considered as a superposition of two oppositely rotating field:
2b cos 0 ti = b (cos 0 ti + sin 0 tj) + b (cos 0 ti sin 0 tj) ,
which gives an effective field of (with = 0 = B0 ):

Beff = B0
k0 + bi0 + b(cos 20 ti0 sin 20 tj0 ).
Since 0 b, the rotation of the term b(cos 20 ti0 sin 20 tj0 ) is so fast compared to the
frequency b. This means that we can take the approximation

Beff B0
k0 + bi0 = bi0 ,
(10)
where the magnetic moment precesses with frequency = b.

As = b 0 , the magnetic moment does not feel the rotating term b (cos 20 ti0 sin 20 tj0 )
which averaged to zero.
Physics of Spin
Page 2 of 5
Physics of Spin
2. Since the angle that makes with Beff stays constant and is initially oriented along the z
axis, is also the angle between Beff and the z axis which is
tan =
b
B0
(11)
From the geometry of the system, we can show that (cos = z /):
sin2
2
1 cos
2
cos
2 sin
t
,
2
t
= sin2 sin2
,
2
1 cos t
= sin2
,
2
= 1 sin2 + sin2 cos t,
= 2 sin sin
cos = cos2 + sin2 cos t.

So, the projected magnetic moment along the z axis is z (t) = cos and the magnetization
is

M = N z = N cos2 + sin2 cos t .
(12)
Note that the magnetization does not depend on the reference frame S or S 0 (z has the same
value viewed in both frames).
Taking = 0 = B0 , the angle is 900 and M = N cos t.
3. From the relations
z
= cos ,
= 1,
P P =
P + P
Physics of Spin
Page 3 of 5
Physics of Spin
we obtain the results ( = 0 )
1 cos
2
1 cos2 sin2 cos t
=
2
1
cos
t
= sin2
2
t
b2
sin2
=
2
2
B0 + b2
P =
= sin2
and
P =
B0
t
,
2
b2
2
(13)
cos2
+
b2
t
t
= cos2
.
2
2
(14)
Part D. Measurement incompatibility

1. In the x direction, the uncertainty in position due to the screen opening is x. According to
the uncertainty principle, the atom momentum uncertainty px is given by
px
~
,
x
which translates into an uncertainty in the x velocity of the atom,

vx
~
.
mx
Consequently, during the time of flight t of the atoms through the device, the uncertainty in
the width of the beam will grow by an amount x given by
x = vx t
Physics of Spin
~
t.
mx
Page 4 of 5
Physics of Spin
So, the width of the beams is growing linearly in time. Meanwhile, the two beams are separating
at a rate determined by the force Fx and the separation between the beams after a time t
becomes
1 Fx 2
1
dx = 2
t =
|x | Ct2 .
2m
m
In order to be able to distinguish which beam a particle belongs to, the separation of the two
beams must be greater than the widths of the beams; otherwise the two beams will overlap and
it will be impossible to know what the x component of the atom spin is. Thus, the condition
must be satisfied is
dx x,
1
~
|x | Ct2
t,
m
mx
1
|x | xCt 1.
~
(15)
2. As the atoms pass through the screen, the variation of magnetic field strength across the beam
width experienced by the atoms is
B = x
dB
= Cx.
dx
This means the atoms will precess at rates covering a range of values given by
= B =
z
|x |
B =
Cx,
~
~
and, if previous condition in measuring x is satisfied,

t 1.
(16)
In other words, the spread in the angle t through which the magnetic moments precess
is so large that the z component of the spin is completely randomized or the measurement
uncertainty is very large.
Physics of Spin
Page 5 of 5
Question 1
The schematic below shows the Hadley circulation in the Earths tropical atmosphere around
the spring equinox. Air rises from the equator and moves poleward in both hemispheres
before descending in the subtropics at latitudes d (where positive and negative latitudes
refer to the northern and southern hemisphere respectively). The angular momentum about
the Earths spin axis is conserved for the upper branches of the circulation (enclosed by the
dashed oval). Note that the schematic is not drawn to scale.
X
Y
northward
+d
(a)
Equator
(2 points) Assume that there is no wind velocity in the east-west direction around
the point X. What is the expression for the east-west wind velocity uY at the
points Y? Convention: positive velocities point from west to east.
(The angular velocity of the Earth about its spin axis is , the radius of the Earth
is a, and the thickness of the atmosphere is much smaller than a.)
Page 1 of 4
(b)
(1 point) Which of the following explains ultimately why angular momentum is

not conserved along the lower branches of the Hadley circulation?
Tick the correct answer(s). There can be more than one correct answer.
(I) There is friction from the Earth's surface.
(II) There is turbulence in the lower atmosphere, where different layers of air
are mixed
(III) The air is denser lower down and so inertia slows down the motion
around the spin axis of the Earth.
(IV) The air is moist at the lower levels causing retardation to the wind
velocity.
Around the northern winter solstice, the rising branch of the Hadley circulation is
located at the latitude r and the descending branches are located at n and s as shown
in the schematic below. Refer to this diagram for parts (c), (d) and (e).
Q
Z
northward
n
(c)
Equator
(2 points) Assume that there is no east-west wind velocity around the point Z.
Given that r = 8, n = 28 and s = 20, what are the east-west wind
velocities uP, uQ and uR respectively at the points P, Q and R?
(The radius of the Earth is a = 6370 km.)
Hence, which hemisphere below has a stronger atmospheric jet stream?
(I) Winter Hemisphere
(II) Summer Hemisphere
(III) Both hemispheres have equally strong jet streams.
Page 2 of 4
(d)
(1 point) The near-surface branch of the Hadley circulation blows southward

across the equator. Mark by arrows on the figure below the direction of the eastwest component of the Coriolis force acting on the tropical air mass
(A) north of the equator;
(B) south of the equator.
N
Equator
near-surface
branch of Hadley
circulation
(e)
(1 point) From your answer to part (d) and the fact that surface friction nearly
balances the Coriolis forces in the east-west direction, sketch the near-surface
wind pattern in the tropics near the equator during northern winter solstice.
Suppose the Hadley circulation can be simplified as a heat engine shown in the
schematic below. Focusing on the Hadley circulation reaching into the winter
hemisphere as shown below, the physical transformation of the air mass from A to B
and from D to E are adiabatic, while that from B to C, C to D and from E to A are
isothermal. Air gains heat by contact with the Earth's surface and by condensation of
water from the atmosphere, while air loses heat by radiation into space.
blackbody radiation
cloud
B
surface heat
transfer
northward
Page 3 of 4
(f)
(2 points) Given that atmospheric pressure at a vertical level owes its origin to
the weight of the air above that level, order the pressures pA, pB, pC, pD, pE,
respectively at the points A, B, C, D, E by a series of inequalities.
(Given that pA = 1000 hPa and pD = 225 hPa. Note that 1 hPa is 100 Pa.)
(g)
(2 points) Let the temperature next to the surface and at the top of the atmosphere
be TH and TC respectively. Given that the pressure difference between points A
and E is 20 hPa, calculate TC for TH = 300 K.
Note that the ratio of molar gas constant (R) to molar heat capacity at constant
pressure (cp) for air, , is 2/7.
(h)
(2 points) Calculate the pressure pB.
(i)
For an air mass moving once around the winter Hadley circulation, using the
molar gas constant, R, and the quantities defined above, obtain expressions for
(A) (2 points) the net work done per unit mole Wnet ignoring surface
friction;
(B) (1 point) the heat loss per unit mole Qloss at the top of the atmosphere.
(j)
(1 point) What is the value of the ideal thermodynamic efficiency i for the
winter Hadley circulation?
(k)
(2 points) Prove that the actual thermodynamic efficiency for the winter Hadley
circulation is always smaller than i, showing all mathematical steps.
(l)
(1 point) Which of the following statements best explains why is less than the
ideal value? Tick the correct answer(s). There can be more than one correct
answer.
(I) We have ignored work done against surface friction.
(II) Condensation occurs at a temperature lower than the temperature of the
heat source.
(III) There is irreversible evaporation of water at the surface.
(IV) The ideal efficiency is applicable only when there is no phase change of
water.
Page 4 of 4
Question 2
The two-slit electron interference experiment was first performed by Mllenstedt et al, MerliMissiroli and Pozzi in 1974 and Tonomura et al in 1989. In the two-slit electron interference
experiment, a monochromatic electron point source emits particles at S that first passes
through an electron biprism before impinging on an observational plane; S1 and S2 are
virtual sources at distance d. In the diagram, the filament is pointing into the page. Note that
it is a very thin filament (not drawn to scale in the diagram).
The electron biprism consists of a grounded cylindrical wire mesh with a fine filament F at
the center. The distance between the source and the biprism is , and the distance between
the distance between the biprism and the screen is L.
Page 1 of 2
(a)
(2 points) Taking the center of the circular cross section of the filament as the
origin O, find the electric potential at any point (x,z) very near the filament in
terms of Va, a and b where Va is the electric potential of the surface of the
filament, a is the radius of the filament and b is the distance between the center of
the filament and the cylindrical wire mesh. (Ignore mirror charges.)
(b)
(4 points) An incoming electron plane wave with wave vector kz is deflected by

the biprism due to the x-component of the force exerted on the electron.
Determine kx the x-component of the wave vector due to the biprism in terms of
the electron charge, e, vz, Va, kz, a and b, where e and vz are the charge and the
z-component of the velocity of the electrons (kx kz). Note that
where h
is the Planck constant.
(c)
Before the point S, electrons are emitted from a field emission tip and accelerated
through a potential V0. Determine the wavelength of the electron in terms of the
(rest) mass m, charge and V0,
(i) (2 points) assuming relativistic effects can be ignored, and
(ii) (3 points) taking relativistic effects into consideration.
(d)
In Tonomura et al experiment,
vz
= c/2,
Va
= 10 V,
V0
= 50 kV,
a
= 0.5 m,
b
= 5 mm,
= 25 cm,
L
= 1.5 m,
h
= 6.6 x 10-34 Js,
electron charge, = 1.6 x 10-19 C,
mass of electron, m0 = 9.1 x 10-31 kg,
and the speed of light in vacuo, c = 3 x 108 ms-1
(i)
(ii)
(2 points) calculate the value of kx,

(2 points) determine the fringe separation of the interference pattern on the
screen,
(iii) (1 point) If the electron wave is a spherical wave instead of a plane wave, is
the fringe spacing larger, the same or smaller than the fringe spacing
calculated in (ii)?
(iv) (2 points) In part (c), determine the percentage error in the wavelength of
the electron using non-relativistic approximation.
(v) (2 points) the distance d between the apparent double slits.
Page 2 of 2
Question 3
Gravitational lensing is a phenomenon where light from a distant source may be deflected by
the curvature of space-time caused by a massive lensing object close to or in the line of sight
between an observer and a distant object. This was first directly observed during the solar
eclipse of 1919 where the observed positions of stars behind the sun differed from their
astrometric positions following Einsteins earlier predictions.
In the case where the observer, lensing object of mass and source are on a straight line,
light from the source is deflected by an angle (in radians) given by
) is the speed of light

where is the gravitational constant (
(
) and is the Einstein radius which is the least distance between the
lensing object and the apparent light path.
(a)
(4 points) Draw a diagram to describe the physical layout of an ideal (observer,

lens and point source in a straight line) lensing system. Draw the light path and
mark the quantities and . Also mark the angular Einstein radius
(the
angular deflection of the source image as seen from earth), and the other
quantities that an observer on earth can measure.
(b)
(2 points) Sketch the image of the source (such as a star), as seen by an observer
on earth, in the case where the source, lensing object and observer are on a
straight line.
(c)
(3 points) Sketch the image of the source (such as a star), as seen by an observer
on earth, in the non-ideal case where the source, lensing object and observer are
not in a straight line. Sketch the source-lens system to explain why this is so.
Page 1 of 3
Gravitational lensing has been proposed as a method to detect massive compact halo objects
(MACHOs) in our galaxy, which may be a candidate for dark matter. These objects are often
dark stellar remnants such as neutron stars and black holes. As stars and MACHOs orbit in
the galaxy, there is a chance that a lensing event may occur when a black hole or neutron star
passes in front of a background star.
(d)
(3 points) The Schwarzschild radius of a black hole defines the point of no return.
A correct expression for the Schwarzschild radius can be obtained by taking it to
be the radius where the escape speed is equal to the speed of light. This means
that something inside the Schwarzschild radius cannot escape the black hole.
Using Newtonian mechanics, derive the formula for the escape speed at a
distance away from a point object of mass . Hence, derive the Schwarzschild
radius for a point object of mass in terms of the gravitational constant and
the speed of light . Show your steps and reasoning clearly. (This happens to give
the correct expression for the Schwarzschild radius that comes from general
relativity.)
(e)
(1 point) In the case where the source, lens and observer are in a straight line,
given a measurement of and , how would you calculate the Schwarzschild
radius of the lensing object?
(f)
(2 points) Consider the case where we have a lensing object of the order of a few
solar masses (
a few
) in the nearby regions of the galaxy (distance
a few
away) and a source object somewhat further out (
a
few
). Which of the following apply in this case?
Choose the following conditions that apply to the case as described in the
question:
(g)
is large and tan , sin , cos
must be calculated exactly

is small and the small angle
approximations to tan , sin ,
cos are permissible
is irrelevant and need not be
calculated
is large and tan

, sin
,
must be calculated exactly
is small and the small angle
approximations to tan , sin ,
cos
are permissible
is irrelevant and need not be
calculated
cos
(3 points) Using the conditions in part (f), rewrite your expression in part (e) in
terms of measurable quantities (which are ,
and ) for a lensing object of
the order of a few solar masses (
a few
) and in the nearby regions
of the galaxy (distance
a few
away) with a source object
somewhat further out (
a few
). Show your working.
Page 2 of 3
(h)
(2 points) Suppose we have an event where a lensing object of

(3.0 solar masses),
away from earth passes in front
of a star
away from earth. This happens such that the ideal
configuration occurs during the event. What is the angular Einstein radius
(as seen from earth) during this event when the source, lens and observer
line up?
Page 3 of 3
Question 1
The schematic below shows the Hadley circulation in the Earths tropical atmosphere around
the spring equinox. Air rises from the equator and moves poleward in both hemispheres
before descending in the subtropics at latitudes d (where positive and negative latitudes
refer to the northern and southern hemisphere respectively). The angular momentum about
the Earths spin axis is conserved for the upper branches of the circulation (enclosed by the
dashed oval). Note that the schematic is not drawn to scale.
X
Y
northward
+d
Equator
Page 1 of 13
(a)
(2 points) Assume that there is no wind velocity in the east-west direction around
the point X. What is the expression for the east-west wind velocity uY at the
points Y? Convention: positive velocities point from west to east.
(The angular velocity of the Earth about its spin axis is , the radius of the Earth
is a, and the thickness of the atmosphere is much smaller than a.)
Solution:
As the problem is symmetric about the equator, we need only consider the
northern hemisphere as shown below.
cos +
equator
a
d
north
pole
Earth's
centre
Conservation of angular momentum about the Earth's spin axis implies that:
2 = (cos + ) cos
= (
1
cos
cos )
(1.5 point)
(0.5 point)
Page 2 of 13
(b)
(1 point) Which of the following explains ultimately why angular momentum is

not conserved along the lower branches of the Hadley circulation?
Tick the correct answer(s). There can be more than one correct answer.
(I) There is friction from the Earth's surface.
(II) There is turbulence in the lower atmosphere, where different layers of air
are mixed
(III) The air is denser lower down and so inertia slows down the motion
around the spin axis of the Earth.
(IV) The air is moist at the lower levels causing retardation to the wind
velocity.
Solution: (I) & (II)
(0.5 point each)
To discourage guessing, minus 0.5 point for each wrong answer.
The minimum points to be awarded in this part is 0.
Page 3 of 13
Around the northern winter solstice, the rising branch of the Hadley circulation is
located at the latitude r and the descending branches are located at n and s as shown
in the schematic below. Refer to this diagram for parts (c), (d) and (e).
Q
Z
northward
Equator
Page 4 of 13
(c)
(2 points) Assume that there is no east-west wind velocity around the point Z.
Given that r = 8, n = 28 and s = 20, what are the east-west wind
velocities uP, uQ and uR respectively at the points P, Q and R?
(The radius of the Earth is a = 6370 km.)
Hence, which hemisphere below has a stronger atmospheric jet stream?
(I) Winter Hemisphere
(II) Summer Hemisphere
(III) Both hemispheres have equally strong jet streams.
Solution:
The angular velocity of the Earth about its spin axis is:
2
= 7.27 105 1
24 60 60
=
so we have:
a = (7.27 105 1 )(6.37 106 ) = 463 1

Conservation of angular momentum about the Earth's spin axis implies that the
wind velocity u at latitude is:
2 cos 2 = ( cos + ) cos

cos2
= (
cos
cos )
(0.5 point)
The required east-west wind velocities are:

(0.5 point for each correct answer, but capped at 1 point maximum)
cos 2
cos 2 8
1
= (
cos ) = 463 (
cos 28) = 105 1
cos
cos 28
cos 2
cos2 8
= (
cos 0) = 463 1 (
1) = 8.97 1
cos 0
1
cos 2
cos2 8
1
= (
cos ) = 463 (
cos 20) = 48.1 1
cos
cos 20
Thus, the winter hemisphere (I) has a stronger atmospheric jet stream. (0.5 point)
Page 5 of 13
(d)
(1 point) The near-surface branch of the Hadley circulation blows southward

across the equator. Mark by arrows on the figure below the direction of the eastwest component of the Coriolis force acting on the tropical air mass
(A) north of the equator;
(B) south of the equator
N
Equator
near-surface branch
of Hadley circulation
Solution: (0.5 point for each correct arrow)
westward Coriolis force
Equator
eastward Coriolis force
near-surface branch
of Hadley circulation
Page 6 of 13
(e)
(1 point) From your answer to part (d) and the fact that surface friction nearly
balances the Coriolis forces in the east-west direction, sketch the near-surface
wind pattern in the tropics near the equator during northern winter solstice.
Solution:
As surface friction nearly balances the Coriolis forces in the east-west direction,
the east-west component of surface friction must act eastward and westward
north and south of the equator respectively. Since friction always opposes
motion, the east-west wind velocity near the surface must be westward and
eastward north and south of the equator respectively. So the resultant nearsurface wind pattern looks like below.
N
Equator
near-surface
wind pattern
(0.5 point for consistency with part (d), even if part (d) was wrong)
(0.5 point for correct answer)
Page 7 of 13
Suppose the Hadley circulation can be simplified as a heat engine shown in the
schematic below. Focusing on the Hadley circulation reaching into the winter
hemisphere as shown below, the physical transformation of the air mass from A to B
and from D to E are adiabatic, while that from B to C, C to D and from E to A are
isothermal. Air gains heat by contact with the Earth's surface and by condensation of
water from the atmosphere, while air loses heat by radiation into space.
blackbody radiation
cloud
B
surface heat
transfer
northward
Page 8 of 13
(f)
(2 points) Given that atmospheric pressure at a vertical level owes its origin to
the weight of the air above that level, order the pressures pA, pB, pC, pD, pE,
respectively at the points A, B, C, D, E by a series of inequalities.
(Given that pA = 1000 hPa and pD = 225 hPa. Note that 1 hPa is 100 Pa.)
Solution:
Since there is less and less air above as one climbs upward in the atmosphere,
atmospheric pressure must decrease upwards.
So,
pA pB pC and pE pD pC (0.5 point)
The process EA represents an isothermal expansion as heat is gained from the
surface. So,
pE pA (0.5 point)
Since the total heat gain must equal the total heat loss, more heat must be lost in
the isothermal compression CD than in the isothermal expansion BC. So net heat
loss occurs from B to D and hence
pD pB (0.5 point)
So with the values of the pressure at A and D, we deduce that:
pA pD (0.5 point)
Collecting all inequalities together,
pE pA pD pB pC
(g)
(2 points) Let the temperature next to the surface and at the top of the atmosphere
be TH and TC respectively. Given that the pressure difference between points A
and E is 20 hPa, calculate TC for TH = 300 K.
Note that the ratio of molar gas constant (R) to molar heat capacity at constant
pressure (cp) for air, , is 2/7.
Solution:
Since pE pA and pA = 1000 hPa, we have pE = 1020 hPa.
From the adiabatic compression from D to E, we have:
(1 point)
225 2/7
= ( ) = (
)
1020
300 = 195
(1 point)
Page 9 of 13
(h)
(2 points) Calculate the pressure pB.

Solution:
From the adiabatic expansion AB and adiabatic compression DE,
pA TH pB TC pA pB
(*) 1 point
pE TH pD TC pE pD
p
1000
pB A pD
750 hPa 220
735 hPa 1 point
225
pE
1020
(i)
For an air mass moving once around the winter Hadley circulation, using the
molar gas constant, R, and the quantities defined above, obtain expressions for
(A) (2 points) the net work done per unit mole Wnet ignoring surface friction;
(B) (1 point) the heat loss per unit mole Qloss at the top of the atmosphere.
Solution:
(A) Work done per mole in an isothermal process is generally given by
= = ( ) = 1 = ln + . (1 point)
Work done per mole in processes EA and BCD are respectively,
= ln + ln = ln ( )
= ln ( )
Work done in an adiabatic process is used entirely to raise the internal energy of
the air mass. Since the decrease in internal energy in process AB exactly cancels
the increase in internal energy in process DE because the respective decrease and
increase in temperature cancel, no net work is done in the adiabatic processes.
Page 10 of 13
So the net work done per mole on the air mass is:
= +
= ln ( ) + ln ( )

= ( ) ln ( ) + ln (
)
= ( ) ln ( ) ( ) ln ( )
using equation () in part (h)

(1 point)
(B)
The heat loss per mole at the top of the atmosphere is the same as the work
done per mole on the air mass because there is no change in internal
energy for an isothermal process.
(0.5 point)
= ln ( )
(0.5 point)
Page 11 of 13
(j)
(1 point) What is the value of the ideal thermodynamic efficiency i for the
winter Hadley circulation?
Solution:
= 1
=1
(k)
195
300
(0.5 point)
= 0.35
(0.5 point)
(2 points) Prove that the actual thermodynamic efficiency for the winter Hadley
circulation is always smaller than i, showing all mathematical steps.
Solution:
ln
(
)
1=
=
( ) ln ( )

ln ( )

=
( ) ln ( )
>
=
ln( )
( ) ln( )
as
>1
(1 point)
using equation () in part (h)
>1+
=

<
(1 point)
Page 12 of 13
(l)
(1 point) Which of the following statements best explains why is less than the
ideal value? Tick the correct answer(s). There can be more than one correct
answer.
(I) We have ignored work done against surface friction.
(II) Condensation occurs at a temperature lower than the temperature of the
heat source.
(III) There is irreversible evaporation of water at the surface.
(IV) The ideal efficiency is applicable only when there is no phase change of
water.
Solution: (II) & (III)
(0.5 point each)
To discourage guessing, minus 0.5 point for each wrong answer.
The minimum points to be awarded in this part is 0.
Page 13 of 13
Question 2
The two-slit electron interference experiment was first performed by Mllenstedt et al, MerliMissiroli and Pozzi in 1974 and Tonomura et al in 1989. In the two-slit electron interference
experiment, a monochromatic electron point source emits particles at S that first passes
through an electron biprism before impinging on an observational plane; S1 and S2 are
virtual sources at distance d. In the diagram, the filament is pointing into the page. Note that
it is a very thin filament (not drawn to scale in the diagram).
The electron biprism consists of a grounded cylindrical wire mesh with a fine filament F at
the center. The distance between the source and the biprism is , and the distance between
the distance between the biprism and the screen is L.
Page 1 of 6
(a)
(2 points) Taking the center of the circular cross section of the filament as the
origin O, find the electric potential at any point (x,z) very near the filament in
terms of Va, a and b where Va is the electric potential of the surface of the
filament, a is the radius of the filament and b is the distance between the center of
the filament and the cylindrical wire mesh. (Ignore mirror charges.)
( )
Writing out |E| =
(1 point)
Note that
( )
(= 0 at the mesh)
Also at the edge of the filament, Va = (
) , so
Giving together
( )
where
(1 point for final expression)
Page 2 of 6
(b)
(4 points) An incoming electron plane wave with wave vector kz is deflected by

the biprism due to the x-component of the force exerted on the electron.
Determine kx the x-component of the wave vector due to the biprism in terms of
the electron charge, e, vz, Va, kz, a and b, where e and vz are the charge and the
kz). Note that
z-component of the velocity of the electrons (kx
where
h is the Planck constant.

There are several ways to work out the solution:
A charge in an electric field will experience a force and hence a change in
momentum. Note that potential energy of the electron (charge =
) is
( ) Using impulse acting on the electron due to the electric field,
(2 points)
)(
(2 points for final expression)
The alternative solution is to write down the equations of motion for the
electrons (2 points) and determine the deflection of the electron as it passes
through the biprism:
Since
(2 points for final expression)
Page 3 of 6
(c)
Before the point S, the electrons are emitted from a field emission tip and
accelerated through a potential V0. Determine the wavelength of the electron in
terms of the (rest) mass m, charge and V0,
(i) (2 points) assuming relativistic effects can be ignored.
Equating the kinetic energy to eV0 (1 point)

(ii)
(3 points) taking relativistic effects into consideration.
Consider
Page 4 of 6
(d)
In Tonomura et al experiment,
vz
= c/2,
Va
= 10 V,
V0
= 50 kV,
a
= 0.5 m,
b
= 5 mm,
= 25 cm,
L
= 1.5 m,
h
= 6.6 x 10-34 Js,
electron charge,
= 1.6 x 10-19 C,
mass of electron, m = 9.1 x 10-31 kg,
and the speed of light in vacuo, c = 3 x 108 ms-1
(i)
(2 points) calculate the value of kx ,
Previous equation:
Plugging the relevant numbers into the equation gives:

(1 point for plugging the correct values)

(ii)
(2 points) determine the fringe separation of the interference pattern on

the screen,
Fringe separation is given by
(iii) (1 point) If the electron wave is a spherical wave instead of a plane wave, is
the fringe spacing larger, the same or smaller than the fringe spacing
calculated in (ii)?
Larger. (1 point for the correct answer)
Page 5 of 6
(iv) (2 points) In part (c), determine the percentage error in the wavelength of
the electron using non-relativistic approximation.
Non-relativistic:
Relativistic:
Percentage error:
or 2.4 percent.
(
(v)
(2 points) Calculate the distance d between the apparent double slits.
The double slit formula is given by

(
where m is the order and y is the distance for maximum intensity

from the central fringe.
In this case, since the fringe spacing is 907,
Page 6 of 6
Question 3
(a) (4 points) Draw a diagram to describe the physical layout of an ideal (observer, lens
and point source in a straight line) lensing system. Draw the light path and mark the
quantities and rE . Also mark the angular Einstein radius E (the angular deflection
of the source image as seen from earth), and the other quantities that an observer on
earth can measure.
Solution:
Apparent Source
rE
Source
rE
DS
Lens
E
DL
Observer
Other relevant quantities include the distances to the lens and source DL and DS .
(DL and DS need not be equal.)
1 point for correct layout
Correct answers should show that light is bent
Apparent source is not required
Accept answers that show the system as a thin lens approximation
(sharp deflection angles)
1 point for light direction correctly marked
Arrows on the light path
1 point for E , and rE correctly identified
1 correct: 0.4 points
1 point for DL and DS (observables; may have different notation)
0.5 points each
Notes:
and rE need not be identified, but may be useful in a later part.
rE should be perpendicular to the projected light path, but in our astronomical system, it makes no difference if it is perpendicular to the sourceobserver line since E is small. Accept answers that have rE perpendicular
to the source-observer line.
(b) (2 points) Sketch the image of the source (such as a star), as seen by an observer on
earth, in the case where the source, lensing object and observer are on a straight line.
Solution: The image of the source should be a symmetrical (1 point) and circular
(1 point) ring around the lensing object.
Notes:
Do not accept solutions that indicate a magnified image of the source. This
includes answers which state that the image is a filled in circle.
1 point for answers that have 2 source images symmetrically on either side
of the lens because the system should be considered in 3 dimensions instead
of 2.
Text answers (without any sketch or diagram) are not accepted. Correct
answers must have a sketch (as specified in the question).
(c) (3 points) Sketch the image of the source (such as a star), as seen by an observer on
earth, in the non-ideal case where the source, lensing object and observer are not in a
straight line. Sketch the source-lens system to explain why this is so.
Solution:
WHAT: (1.5 points) The observer will see light from one side of the lens but not
the other side. This means that the Einstein ring should be an arc instead of a
complete circle. The ring may be distorted or broken depending on how much
deviation from an ideal case. Correct answers should not be a perfect circle or
straight line.
Note:
Solutions that give 2 source images on either side of the lens (with asymmetry) are awarded 1 point instead of 1.5 point because the system should be
considered in 3 dimensions instead of 2.
Text answers (without any sketch or diagram) are not accepted. Correct
answers must have a sketch (as specified in the question).
WHY: (1.5 points)
One possible answer:
Lens
Source
rE
Observer
rE
DL
DS
For slight deviations from the ideal case, accept also the following diagram if rE1
Page 2
is smaller than rE2 .

Lens rE1
Source
Observer
rE2
DL
DS
In general, accept answers which show that the asymmetry in the system will cause
the observer to see something asymmetrical.
Notes:
The key concept in this question is asymmetry. Correct answers for either part
must demonstrate that departures from the ideal case will result in asymmetry in
the observed system, and that the asymmetry about the source-observer line is the
cause of the asymmetry in the observation.
(d) (3 points) The Schwarzschild radius of a black hole defines the point of no return. A
correct expression for the Schwarzschild radius can be obtained by taking it to be the
radius where the escape speed is equal to the speed of light. This means that something
inside the Schwarzschild radius cannot escape the black hole.
Using Newtonian mechanics, derive the formula for the escape speed at a distance r
away from a point object of mass M. Hence, derive the Schwarzschild radius for a
point object of mass M in terms of the gravitational constant G and the speed of light
c. Show your steps and reasoning clearly. (This happens to give the correct expression
for the Schwarzschild radius that comes from general relativity.)
Solution: By definition, the gravitational potential energy of a test mass m at a
distance r from the mass is (0.5 point)
=
GMm
.
r
To escape the gravitational potential, the total energy of the test mass needs to be
at least 0 so it should have a kinetic energy of (0.5 point)
K=
GMm 1 2
= mve .
r
2
Rearranging the above, the escape speed at distance from mass r is (1 point)
2GM
ve =
.
r
Substitute ve = c and rearrange to get (1 point)
rS =
2GM
.
c2
2 points for deriving escape speed (Any reasonable and physically sound method
based on Newtonian mechanics)
1 point for deriving the Schwarzschild radius from the escape speed.
Page 3
(e) (1 point) Using the formula for light deflection, write down an expression for the
Schwarzschild radius of a lensing object in the case where the source, lens and observer is in a straight line.
Solution: The Schwarzschild radius is
rS =
2GM
1
so rS = rE
2
c
2
Notes: Full marks for correct working.

(f) (2 points) Consider the case where we have a lensing object of the order of a few
solar masses (M a few 1030 kg) in the nearby regions of the galaxy (distance DL
a few 1018 m away) and a source object somewhat further out (DS a few DL ).
What can we say about and E in this case? (Choose your answer on your answer
sheet. Points will be deducted for wrong answers.)
is large and tan , sin , cos
E is large and tan E , sin E ,
must be calculated exactly.
cos E must be calculated exactly.
is small and the small angle approximations to tan , sin , cos
are permissable.
E is small and the small angle

approximations to tan E , sin E ,
cos E are permissable.
is irrelevant and need not be calculated
E is irrelevant and need not be calculated
Solution:
is small
E is small
Notes:
Choices pertaining to and E are to be marked independently (1 point
each).
The conditions are mutually exclusive so accept only one condition for each
quantity (, E ). Answers that select more than one condition for a quantity
(, E ) are wrong (no point to be awarded).
Reasoning: Working out the numbers, we can find that the Schwarzschild radius
is on the order of 104 m. Because has a maximum of 2 (largest possible angle),
this means the physical Einstein radius rE 104 m is very small compared to the
distance to the lens DL 1020 m. The geometry of the system therefore means
that is actually a very small angle.
Another approximation comes from the geometry of the system which sets bounds
on and E so that (see figure in part (a))
tan E =
rE
2r /
1016
= S
DL
DL
Page 4
which suggests that or E or both should be small.

Based on the geometry of the setup and what we have already established (
small), we then have the following cases:
E large means that DL is small which is not the case here.
small, E small is the only valid outcome here
The result and constraints in the question suggests that and E are both small
Because E is small, the Einstein radius rE 104 m is very small compared to the
distance to the lensing object DL 1020 m or source DS . We can therefore take
the small angle approximation where and E is involved.
(g) (3 points) Using the conditions in part (f), rewrite your expression in part (e) in terms
of measurable quantities (which are E , DS and DL ) for a lensing object of the order
of a few solar masses (M a few 1030 kg) and in the nearby regions of the galaxy
(distance DL a few 1018 m away) with a source object somewhat further out (DS
a few DL ). Show your working.
Solution: Adding up exterior angles, we see that = E + so E =
where is small ( and rE defined on the following diagram). Also note that rE is
approximately perpendicular to the source-observer system because E is small.
Apparent Source
rE
Source
rE
Lens
DS
E
Observer
DL
Using the small angle approximation for and E , we can write

rE
rE
rE
= tan E E and
=
= tan
DL
DS
DS DL
This gives (1 point)
=
So that (1 point)
rE
rE
+
DL DS DL
1
1
rS = rE = rE2
2
2
DS
DL (DS DL )
To write this in terms of E , DL and DS , we use rE = E DL to get (1 point)
1 2
DS DL
rS = E
2
DS DL
Notes:
1 point for
1 point for rS
1 point for final equation
Page 5
(h) (2 points) Suppose we have an event where a lensing object of 6.01030 kg (3.0 solar
masses), 2.61018 m away from earth passes in front of a star 9.21018 m away from
earth. This happens such that the ideal configuration occurs during the event. What is
the angular Einstein radius E (as seen from earth) during this event when the source,
lens and observer line up?
Solution: The Schwarzschild radius of the lens is
rS =
2 (6.673 1011 ) 6.0 1030

= 8.9 103 m
8
2
(3.0 10 )
From the previous part, the angular Einstein radius is given by
DS DL
2
E = 2rS
DS DL
(9.2 2.6) 1018

3
= 2 8.9 10
(9.2 1018 ) (2.6 1018 )
= 4.9 1015
Thus the angular Einstein radius is
E = 4.9 1015 = 7.0 108 radians = 0.014 arcseconds

(1 point for correct answer, 1 point for correct working)
Notes:
Students are expected to use the formula derived in part (g) to answer this
question.
For the final answer:
0.5 point off for missing units. While angles are mathematically dimensionless, a good student should be cognisant of the fact that there
are different physical units for angular measurement, and that units for
angles should be specified.
0.5 point off for final answers given to 1 significant figure or less.
1 point off if the final number is incorrect.
Page 6
Question 1
The fractional quantum Hall effect (FQHE) was discovered by D. C. Tsui and H.
Stormer at Bell Labs in 1981. In the experiment electrons were confined in two
dimensions on the GaAs side by the interface potential of a GaAs/AlGaAs
heterojunction fabricated by A. C. Gossard (here we neglect the thickness of the
two-dimensional electron layer). A strong uniform magnetic field was applied
perpendicular to the two-dimensional electron system. As illustrated in Figure 1, when
a current was passing through the sample, the voltage H across the current path
exhibited an unexpected quantized plateau (corresponding to a Hall resistance H =
3/ 2 ) at sufficiently low temperatures. The appearance of the plateau would imply
the presence of fractionally charged quasiparticles in the system, which we analyze
below. For simplicity, we neglect the scattering of the electrons by random potential,
as well as the electron spin.
(a) In a classical model, two-dimensional electrons behave like charged billiard balls
on a table. In the GaAs/AlGaAs sample, however, the mass of the electrons is
reduced to an effective mass due to their interaction with ions.
(i) (2 point) Write down the equation of motion of an electron in perpendicular
= .
electric field = and magnetic field
(ii) (1 point) Determine the velocity s of the electrons in the stationary case.
(iii) (1 point) Which direction is the velocity pointing at?
(b) (2 points) The Hall resistance is defined as H = H /. In the classical model,
find H as a function of the number of the electrons and the magnetic flux
= = , where is the area of the sample, and and the
effective width and length of the sample, respectively.
(c) (2 points) We know that electrons move in circular orbits in the magnetic field. In
the quantum mechanical picture, the impinging magnetic field could be
viewed as creating tiny whirlpools, so-called vortices, in the sea of
electronsone whirlpool for each flux quantum / of the magnetic field,
where is the Planck's constant and the elementary charge of an electron.
For the case of H = 3/ 2 , which was discovered by Tsui and Stormer, derive
the ratio of the number of the electrons to the number of the flux quanta ,
1/4
known as the filling factor .
Figure 1: (a) Sketch of the experimental setup for the observation of the FQHE. As indicated, a
current is passing through a two-dimensional electron system in the longitudinal direction with
an effective length . The Hall voltage H is measured in the transverse direction with an effective
width . In addition, a uniform magnetic field is applied perpendicular to the plane. The
direction of the current is given for illustrative purpose only, which may not be correct. (b) Hall
resistance H versus at four different temperatures (curves shifted for clarity) in the original
publication on the FQHE. The features at H = 3/ 2 are due to the FQHE.
(d) (2 points) It turns out that binding an integer number of vortices ( > 1) with
each electron generates a bigger surrounding whirlpool, hence pushes away all
other electrons. Therefore, the system can considerably reduce its electrostatic
2/4
Coulomb energy at the corresponding filling factor. Determine the scaling

exponent of the amount of energy gain for each electron () .
(e) (2 points) As the magnetic field deviates from the exact filling = 1/ to a
higher field, more vortices (whirlpools in the electron sea) are being created.
They are not bound to electrons and behave like particles carrying effectively
positive charges, hence known as quasiholes, compared to the negatively charged
electrons. The amount of charge deficit in any of these quasiholes amounts to
exactly 1/ of an electronic charge. An analogous argument can be made for
magnetic fields slightly below and the creation of quasielectrons of negative
charge = /. At the quantized Hall plateau of H = 3/ 2 , calculate the
amount of change in that corresponds to the introduction of exactly one
fractionally charged quasihole. (When their density is low, the quasiparticles are
confined by the random potential generated by impurities and imperfections,
hence the Hall resistance remains quantized for a finite range of .)
(f) In Tsui et al. experiment,
the magnetic field corresponding to the center of the quantized Hall plateau
H = 3/ 2 , 1/3 = 15 Tesla,
the effective mass of an electron in GaAs, = 0.067 ,

the electron mass, = 9.1 1031 kg,
Coulomb's constant, = 9.0 109 N m2 /C2 ,
the vacuum permittivity, 0 = 1/4 = 8.854 1012 F/m,
the relative permittivity (the ratio of the permittivity of a substance to the vacuum
permittivity) of GaAs, = 13,
the elementary charge, = 1.6 1019 C,
Planck's constant, = 6.626 1034 J s, and
Boltzmann's constant, B = 1.38 1023 J/K.
In our analysis, we have neglected several factors, whose corresponding energy
scales, compared to () discussed in (d), are either too large to excite or too
small to be relevant.
(i) (1 point) Calculate the thermal energy th at temperature = 1.0 K.
(ii) (2 point) The electrons spatially confined in the whirlpools (or vortices) have
a large kinetic energy. Using the uncertainty relation, estimate the order of
magnitude of the kinetic energy. (This amount would also be the additional
energy penalty if we put two electrons in the same whirlpool, instead of in
two separate whirlpools, due to Pauli exclusion principle.)
(g) There are also a series of plateaus at H = / 2 , where = 1, 2, 3, in Tsui et
al. experiment, as shown in Figure 1(b). These plateaus, known as the integer
quantum Hall effect (IQHE), were reported previously by K. von Klitzing in 1980.
Repeating (c)-(f) for the integer plateaus, one realizes that the novelty of the
FQHE lies critically in the existence of fractionally charged quasiparticles. R.
3/4
de-Picciotto et al. and L. Saminadayar et al. independently reported the

observation of fractional charges at the = 1/3 filling in 1997. In the
experiments, they measured the noise in the charge current across a narrow
constriction, the so-called quantum point contact (QPC). In a simple statistical
model, carriers with discrete charge tunnel across the QPC and generate
charge current IB (on top of a trivial background). The number of the carriers
arriving at the electrode during a sufficiently small time interval obeys Poisson
probability distribution with parameter
P( = ) =
e
!
where ! is the factorial of . You may need the following summation
e =
k=0
(i) (2 point) Determine the charge current B , which measures total charge per
unit of time, in terms of and .
(ii) (2 points) Current noise is defined as the charge fluctuations per unit of time.
One can analyze the noise by measuring the mean square deviation of the
number of current-carrying charges. Determine the current noise due to
the discreteness of the current-carrying charges in terms of and .
(iii) (1 point) Calculate the noise-to-current ratio /B , which was verified by R.
de-Picciotto et al. and L. Saminadayar et al. in 1997. (One year later, Tsui
and Stormer shared the Nobel Prize in Physics with R. B. Laughlin, who
proposed an elegant ansatz for the ground state wave function at = 1/3.)
4/4
Question 2
How are aurora ignited by the solar wind?
Figure 1
The following questions are designed to guide you to find the answer one step by one step.
Background information of the interaction between the solar wind and the Earth's magnetic
field
It is well known that the Earth has a substantial magnetic field. The field lines defining the
structure of the Earth's magnetic field is similar to that of a simple bar magnet, as shown in Figure
2. The Earth's magnetic field is disturbed by the solar wind, whichis a high-speed stream of hot
plasma. (The plasma is the quasi-neutralionized gas.)The plasma blows outward from the Sun and
varies in intensity with the amount of surface activity on the Sun. The solar wind compresses the
Earth's magnetic field. On the other hand, the Earth's magnetic field shields the Earth from much
of the solar wind. When the solar wind encounters the Earth's magnetic field, it is deflected like
water around the bow of a ship, as illustrated in Figure 3.
Figure 2
Question 2
Figure 3
The curved surface at which the solar wind is first deflected is called the bow shock. The
corresponding region behind the bow shock andfront of the Earth's magnetic field is called
themagnetosheath. The region surroundedby the solar wind is called the magnetosphere.The
Earth's magnetic field largely prevents the solar wind from enteringthe magnetosphere. The
contact region between the solar wind and the Earth's magnetic field is named the magnetopause.
The location of the magnetopause is mainly determined by the intensity and the magnetic field
direction of the solar wind. When the magnetic field in the solar wind is antiparallel to the Earth's
magnetic field, magnetic reconnection as shown in Figure 4 takes place at the dayside
magnetopause, which allows some charged particles ofthe solar wind in the region "A" to move
into the magnetotail "P" on the night side as illustrated in Figure 5. A powerful solar wind can
push the dayside magnetopause tovery close to the Earth, which could cause a high-orbit satellite
(such as a geosynchronous satellite) to be fully exposed to the solar wind. The energetic particles
in the solar wind could damage high-tech electronic components in a satellite.Therefore, it
isimportant to study the motion of charged particles in magnetic fields, which will give an answer
of the aurora generation and could help us to understand the mechanism of the interaction between
the solar wind and the Earth's magnetic field.
Question 2
Figure 4
Figure 5
Numerical values of physical constants and the Earth's dipole magnetic field
Speed of light in vacuum: c 2.998 108 m/s ;
Permittivity in vacuum: 0 8.9 10 12 C 2 / ( N m 2 ) ;
Permeability in vacuum: 0 4 10 7 N /A 2 ;
Charge of a proton: e 1.6 10 19 C ;
Mass of an electron: m 9.1 10 3 1 kg ;
Mass of a proton: mp 1.67 10
27
kg ;
Boltzmann's constant: k 1.38 10 23 J/K ;

Gravitational acceleration: g 9.8m/s 2 ;
Plancks constant: h 6.626 10 34 J s
Earth's radius R E 6.4 10 6 m .
The Earth's dipole magnetic field can be expressed as
B R3
Bd 0 5 E [3 xzx 3 yzy ( x 2 y 2 2 z 2 ) z ]
r
where r
, r R E
(1)
x 2 y 2 z 2 , B 0 3.1 10 5 T , and x , y , z are the unit vectors in
the x , y , z directions, respectively.

Questions
(a) (3 Points)
(i) (1 Point) Before we study the motion of a charged particle in the Earth's dipole magnetic field,
Question 2
we first considerthe motion of an electron in a uniform magnetic field B . When the initial electron
velocity v is perpendicular to the uniform magnetic field as shown in Figure 6, please calculate
the electron trajectory.The electron is initially located at (x,y,z)=(0,0,0).
e-
y
V
Figure 6
(ii) (1 Point) Please determine the electric current of the electron motion and calculate the
r
magnetic moment IA , where A is the area of the electron circular orbit and the direction
of A is determined by the right-hand rule of theelectric current.
(iii) (1 Point) If the initial electron velocity v is not perpendicular to the uniform magnetic field,
i.e, the angle between B and
v is 0 0 < 90 0 , please give the screw pitch (the distance along
the z-axis between successive orbits) of the electron trajectory.

(b) (4 Points) In the uniform background magnetic field as shown in Figure 6, theplasma density
isnonuniform in x. For simplicity, we assume that the temperature and the distribution of the ions
andelectrons are the same. Thus, the plasma pressure can be expressed as
p ( x ) kT [ n i ( x ) n e ( x )] 2 kT n ( x ) 2 kT ( n 0 x ) ,
Where B, T, k, n0 , and are positive constants, ni ( x ) and n e ( x ) are the number densities of the
ions and electrons.
(i) (2 Points) Please explain the generation mechanism of the electric current by a schematic
drawing.
(ii) (2 Points) If both the ions and electrons have a Maxwellian distribution, the ion distribution is
mi
f i ( x, v , v|| ) ni ( x )
2 kT
3/ 2
mi ( v2 v||2 )/ 2 kT
please calculate the constant in the magnetization M n( x)
kT
, where the magnetization
B
Question 2
Mis the magnetic moment per unit volume. (Hint: We have x exp( x ) dx 1 and
0
exp( x
) dx .)
(c) (1 Point)Now let's go back to the Earth's dipole magnetic field. Please apply the result from
Question(b) to calculate the ratio of the diamagnetic field and the Earth's dipole magnetic field in
Equation (1) at the position (x=10REy=0, z=1 RE). The plasma pressure is assumed to be
10
p ( z ) p0 e ( z / a ) , where p0 310 pa and a 2 R E .The magnetic field around this
2
position is also assumed to be uniform. Be aware of the difference in the coordinate systemsin
Questions (b) and (c). (Hint: The diamagnetic field is given by B m x o M .)
(d) (4 Points) From Figures 2, 3, and 5, it can be clearly seen that the Earth's magnetic
fieldstrength along a magnetic field line is the largest at the poles and the smallest in the equatorial
plane.Since the Earth's dipole magnetic field is axially symmetric and slowly varying along a
magnetic field line, it can for simplicitybe treated as a magnetic-mirror field as shown in Figure
7.The magnetic field strength along a magnetic field line is the smallest ( B0 ) at the point "P2" and
the largest ( Bm ) at the points "P1" and "P3". An electron with an initial velocity v is located at the
point "P2" and drifts towards the point "P3". The angle between the initial velocity v and the
magnetic field at the point "P2" is 00 900 . For the magnetic-mirror field
B Br r Bz z (with
B r B z ), we can assume
dB dB
dB
, where
is the spatial
dz ds
ds
derivative of B along a magnetic field line. Since there is no evidence of the existence of
magnetic monopoles, we have Br
of B r and rc is the electron gyroradius.
1 dB
rc Bz , where
2 dz
Br
is the gyro-average
Question 2
Figure 7
(i) (3 Points) Please givethe gyro-averaged magnetic-field force along the magnetic field lines on
an electron and show that the magnetic moment is a motion constant, i.e.,
d
0 , based on the
dt
law of the total kinetic energy conservation.

(ii) (1 Point) Based on the motion constant of magnetic moment, please determinewhat the
condition should be satisfied for the angle between the initial electron velocity v and the
magnetic field at the point "P2"ifan electron will not escape from the magnetic mirror field.
(e) (1 Point) Earth's dipole magnetic field lines (blue lines) are shown in Figure 8. The spiral
trajectory of a charged particle (red curve) is assumed to be confined in the y=0 plane since the
gradient and the curvature of the magnetic field can be ignored. If a charged particle with the mass
m, charge q, and velocity v is initially located at the equatorial position [x=6RE, y=0, z=0] and the
angle between the electron velocity v and the magnetic field is initially, please determine what
the condition should be satisfied for if the charged particle arrives below 200km of its altitude
at the latitude 600.
Question 2
Figure 8
(f)(5 Points) As shown in Figure 5, when magnetic reconnection takes place at the dayside
magnetopause, reconnected magnetic field lines drift towards the nightside region because the
solar wind flows tailward. Thus, some solar wind electrons in the region "A" also move towards
the magnetotail in the region "P". After the electrons arrive in the region "P", some electrons can
be accelerated to around 1keV. If energetic electrons drift down to the thermosphere (The altitude
of the thermosphere is about 85km-800km.), energetic electrons can collide with the neutral atoms,
which could cause the neutral atoms to jump into excited states. A photon is emitted when the
higher excited state of aneutral atom returnsto its lower excited state or ground state. Splendid
aurora (Figure 1) is generated in the aurora oval due to photons with different wavelengths. It is
found that the aurorais mainly resulted fromphotonsemittedby oxygen atoms.The energy levels in
the first and second excited states relative to the ground stateare 1.96eV and 4.17eV, respectively.
The lifetimes of the two excited states of an oxygen atomare 110s and 0.8s as shown in Figure 9.
Figure 9
(i) (2 Points) Please give the atmospheric density as a function of the altitude and the ratio of the
oxygen density at the altitudes H=160km and H=220km. For simplicity, we assume that the
atmospheric temperature is independent of the altitude and the air is an ideal gas.
( 0 g P0 0.13 / km , where 0 and P0 are the atmospheric density and pressure at sea level.)
(ii) (3 Points)Please give the colors of auroras at the altitudes H=160km and H=220km. (Hint: The
dependence of the collision frequency of atmosphericmolecules on the atmospheric density
is 0 0 where
0 109 / s
is the collision frequency of atmosphericmolecules at sea
level.The excited oxygen atom will lose a part of its energy when it collides with other neutral
molecules. )
(g) (2 Points) As mentioned above, a powerful solar wind can push the dayside magnetopause
tovery close to the Earth, which could cause a high-orbit satelliteto be fully exposed to the solar
wind. The energetic particles in the solar wind could damage high-tech electronic components in a
Question 2
satellite.For simplicity, the Earth's dipole magnetic field is assumed to remain unchanged when the
solar wind compresses it and that the plasma density is ignorable in the magnetosphere. Please
give the minimum solar wind speed to cause a damage of a geosynchronous satellite if the
9
magnetic field strength and the plasma density of the solar wind are Bs 5 10 T and
s 50 proton / cm3 , respectively. (Hint: The force per unit area associated with the magnetic
field is
f B2 20 . We only consider the variation in x for all physical quantities, i.e., the
physical quantities are independent of y and z.)
Question 3
Figure 1 shows a Fabry-Perot (F-P) etalon, in which air pressure is tunable. The F-P etalon
consists of two glass plates with high-reflectivity inner surfaces. The two plates form a cavity in
which light can be reflected back and forth. The outer surfaces of the plates are generally not
parallel to the inner ones and do not affect the back-and-forth reflection. The air density in the
etalon can be controlled. Light from a Sodium lamp is collimated by the lens L1 and then passes
through the F-P etalon. The transmitivity of the etalon is given by T
1
, where
1 F sin 2 ( / 2)
4nt cos
4 R , R is the reflectivity of the inner surfaces,
is the phase shift of two
1 R
neighboring rays, n is the refractive index of the gas, t is the spacing of inner surfaces, is the
incident angle, and is the light wavelength.
Gas in
Pin valve
t o vacuum pump
n
Microscope
N
B
S
t
Na Lamp
Fringes
L1
F1
L2
F-P Etalon
Figure 1
The Sodium lamp emits D1 ( 589.6nm ) and D2 ( 589nm ) spectral lines and is located
in a tunable uniform magnetic field. For simplicity, an optical filter F1 is assumed to only allow
the D1 line to pass through. The D1 line is then collimated to the F-P etalon by the lens L1.
Circular interference fringes will be present on the focal plane of the lens L2 with a focal length
f=30cm. Different fringes have the different incident angle . A microscope is used to observe
the fringes. We take the reflectivity R= 90% and the inner-surface spacing t=1cm.
Some physical constants: h 6.626 1034 J s , e 1.6 1019 C , me 9.11031kg , c 3.0 108 ms1 .
Page 1 of 4
Question 3
a 3points The D1 line ( 589.6nm ) is collimated to the F-P etalon. For the vacuum
case (n=1.0), please calculate (i) interference orders mi , (ii) incidence angle
i and (iii) diameter
Di for the first threei=123fringes from the center of the ring patterns on the focal plane.
b3 points As shown in Fig. 2, the width of the spectral line is defined as the full width
of half maximum (FWHM) of light transmitivity T regarding the phase shift . The resolution of
the F-P etalon is defined as follows: for two wavelengths and , when the central
phase difference of both spectral lines is larger than , they are thought to be resolvable;
then the etalon resolution is / when . For the vacuum case, the D1 line
( 589.6nm ), and because of the incident angle 0 , take cos 1.0 , please calculate
ithe width of the spectral line.
iithe resolution
of the etalon.
T
0.5
2m
Figure 2
c1 point As shown in Fig. 1, the initial air pressure is zero. By slowly tuning the pin valve,
air is gradually injected into the F-P etalon and finally the air pressure reaches the standard
atmospheric pressure. On the same time, ten new fringes are observed to produce from the center
of the ring patterns on the focal plane. Based on this phenomenon, calculate the refractive index of
air nair at the standard atmospheric pressure.

d (2 points) Energy levels splitting of Sodium atoms occurs when they are placed in a
magnetic field. This is called as the Zeeman effect. The energy shift given by E
m j gk B B
where the quantum number mj can be JJ-1-J+1-JJ is the total angular quantum number,
Page 2 of 4
Question 3
gk is the Land factor,
he
is Bohr magnetonh is the Plank constante is the electron
4me
chargeme is the electron mass, B is the magnetic field. As shown in Fig. 3, the D1 spectral line is
emitted when Sodium atoms jump from the energy level 2P1/2 down to 2S1/2. We have J
1
for
2
both 2P1/2 and 2S1/2 . Therefore, in the magnetic field, each energy level will be split into two levels.
We define the energy gap of two splitting levels as E1 for 2P1/2 and E2 for 2S1/2 respectively (E1
<E2). As a result, the D1 line is split into 4 spectral lines (a, b, c, and d), as showed in Fig. 3.
Please write down the expression of the frequency ( ) of four lines a, b, c, and d.
mj
1/2
-1 /2
P 1 /2
E 1
58 9.6 nm
1/2
a b
S 1 /2
c d
-1/2
E 2
Figure 3
(e) (3 points) As shown in Fig. 4, when the magnetic field is turned on, each fringe of the D1 line
will split into four sub-fringes (1, 2, 3, and 4). The diameter of the four sub-fringes near the center
is measured as D1 D2 D3 and D4 . Please give the expression of the splitting energy gap E1
of 2P1/2 and E2 of 2S1/2.
m
m-1
m-1
B=0
3 1
D4
D2
3
B0
D1
D3
2 4
Figure 4
Page 3 of 4
Question 3
f3 pointsFor the magnetic field B=0.1Tthe diameter of four sub-fringes is measured as
D1 3.88mm D2 4.05mm D3 4.35mm and D4 4.51mm . Please calculate the

Land factor gk1 of 2P1/2 and gk2 of 2S1/2.
g (2 points) The magnetic field on the sun can be determined by measuring the Zeeman effect
of the Sodium D1 line on some special regions of the sun. One observes that, in the four split lines,
the wavelength difference between the shortest and longest wavelength is 0.012nm by a solar
spectrograph. What is the magnetic field B in this region of the sun?
(h)3 points A Light- Emitting Diode (LED) source with a central wavelength 650nm
and spectral width 20nm is normally incident ( 0 ) into the F-P etalon shown in Fig. 1.
For the vacuum case, find (i) the number of lines in transmitted spectrum and (ii) the frequency
width of each line
Page 4 of 4
Question 1
The fractional quantum Hall effect (FQHE) was discovered by D. C. Tsui and
H. Stormer at Bell Labs in 1981. In the experiment electrons were confined in
two dimensions on the GaAs side by the interface potential of a GaAs/AlGaAs
heterojunction fabricated by A. C. Gossard (here we neglect the thickness of the
two-dimensional electron layer). A strong uniform magnetic field B was applied
perpendicular to the two-dimensional electron system. As illustrated in Figure
1, when a current I was passing through the sample, the voltage VH across the
current path exhibited an unexpected quantized plateau (corresponding to a Hall
resistance RH = 3h/e2 ) at sufficiently low temperatures. The appearance of the
plateau would imply the presence of fractionally charged quasiparticles in the
system, which we analyze below. For simplicity, we neglect the scattering of the
electrons by random potential, as well as the electron spin.
(a) In a classical model, two-dimensional electrons behave like charged billiard
balls on a table. In the GaAs/AlGaAs sample, however, the mass of the
electrons is reduced to an effective m due to their interaction with ions.
(i) (2 point) Write down the equation of motion of an electron in perpen~ = Ey y and magnetic field B
~ = B z.
dicular electric field E
Solution: An electron with charge e (e > 0) experiences the Lorentz force
due to the perpendicular magnetic field and the electric force
m

d~v
~ +E
~
= e ~v B
dt
where ~v is the velocity of the electron.

Grading: 1 point for writing down the electric force and the magnetic force
correct, and 1 point for writing down the effective mass and the acceleration
correct.
(ii) (1 point) Determine the velocity vs of the electrons in the stationary
case.
Solution: In the stationary regime, the acceleration vanishes. Hence
~ +E
~ =0
~vs B
The velocity can be expressed as
~vs =
~ B
~
E
B2
whose magnitude is simply vs = E/B.

Grading: Either writing down the correct magnitude of the velocity or its
vector form is sufficient for the 1 point.
(iii) (1 point) Which direction is the velocity pointing at?

Solution: The velocity ~vs should be perpendicular to both the magnetic field
~ is in the z direction and E
~ in the y direction, as
and the electric field. If B
given by the problem, ~vs is in the x direction, generating a charge current
in the x direction.
Grading: 1 point for the correct direction.
(b) (2 points) The Hall resistance is defined as RH = VH /I. In the classical

model, find RH as a function of the number of the electrons N and the
magnetic flux = BA = BW L, where A is the area of the sample, and W
and L the effective width and length of the sample, respectively.

Solution: The Hall voltage VH = Ey W . The current in the x direction is
I=
Q
Ne
Ne Ey
N
=
=
= e VH
t
L/vs
L B
Therefore,
RH =
1
VH
=
I
eN
Grading: 1 point for the final expression and 1 point for writing down the
expression for relating I with the number of electrons and their stationary
velocity (hence the electric field and the magnetic field).
(c) (2 points) We know that electrons move in circular orbitss in the magnetic
field. In the quantum mechanical picture, the impinging magnetic field B
could be viewed as creating tiny whirlpools, so-called vortices, in the sea of
electronsone whirlpool for each flux quantum h/e of the magnetic field,
where h is the Plancks constant and e the elementary charge of an electron.
For the case of RH = 3h/e2 , which was discovered by Tsui and Stormer,
derive the ratio of the number of the electrons N to the number of the flux
quanta N , known as the filling factor .
Solution: The Hall resistance can be rewritten as
RH =
1
h /(h/e)
h N
= 2
= 2
eN
e
N
e N
At the plateau, = N/N = 1/3.

Grading: 1 point for the final expression.
(d) (2 points) It turns out that binding an integer number of vortices (n > 1)
with each electron generates a bigger surrounding whirlpool, hence pushes
away all other electrons. Therefore, the system can considerably reduce
its electrostatic Coulomb energy at the corresponding filling factor. Determine the scaling exponent of the amount of energy gain for each electron
U(B) B .
Solution: The average distance between electrons can be written as f l0 ,
where
s
s
s
LW
h
=
=
l0 =
N
NB
eB
and f is a dimensionless constant that is determined by the electron distribution (or, quantum mechanically, wave function). Binding multiple vortices
with an electron effectively reduces the probability of other electrons getting
close. Therefore, the electrons optimize their distribution in such a way that
their average distance increases from f1 l0 to f2 l0 (f1 < f2 ). One expect the
Coulomb energy gain per electron is proportional to
e2
1
e2
1
=
40 r (f1 l0 ) 40 r (f2 l0 )
f1 f2
Therefore, (B) 1/l0 B, or = 1/2.
e2
40 r l0
Grading: The key point here is to realize that the energy scale is determined
by the Coulomb interaction, which scales inversely with a length scale (e.g.,
the magnetic length) that characterizes the mean electron distance (and its
change). 1 point for the final expression and 1 point for writing down the
correct relation between the length scale and the magnetic field.
(e) (2 points) As the magnetic field deviates from the exact filling = 1/m
to a higher field, more vortices (whirlpools in the electron sea) are being
created. They are not bound to electrons and behave like particles carrying effectively positive charges, hence known as quasiholes, compared to
the negatively charged electrons. The amount of charge deficit in any of
these quasiholes amounts to exactly 1/m of an electronic charge. An analogous argument can be made for magnetic fields slightly below and the
creation of quasielectrons of negative charge e = e/m. Assume the sample has an area A. At the quantized Hall plateau of RH = 3h/e2 , calculate
the amount of change in B that corresponds to the introduction of exactly
one fractionally charged quasihole. (When their density is low, the quasiparticles are confined by the random potential generated by impurities and
4
imperfections, hence the Hall resistance remains quantized for a finite range
of B.)
Solution: The flux change due to the change of the magnetic field is
= B(W L) =
h
e
Therefore, B = h/(eW L).

Grading: 2 points for the final expression.
(f) In Tsui et al. experiment,

the magnetic field corresponding to the center of the quantized Hall
plateau RH = 3h/e2 , B1/3 = 15 Tesla,
the effective mass of an electron in GaAs, m = 0.067me ,
the electron mass, me = 9.1 1031 kg,
Coulombs constant, k = 9.0 109 Nm2/C2 ,
the vacuum permittivity, 0 = 1/(4k) = 8.854 1012 F/m,
the relative permittivity (the ratio of the permittivity of a substance to
the vacuum permittivity) of GaAs, r = 13,
the elementary charge, e = 1.6 1019 C,
Plancks constant, h = 6.626 1034 Js, and
Boltzmanns constant, kB = 1.38 1023 J/K.
In our analysis, we have neglected several factors, whose corresponding
energy scales, compared to (B) discussed in (d), are either too large to
excite or too small to be relevant.
(i) (1 point) Calculate the thermal energy Eth at temperature T = 1.0 K.
Solution: The thermal energy
Eth = kB T = 1.38 1023 1.0 = 1.38 1023 J
Grading: 1 point for the numerical result.
(ii) (2 point) The electrons spatially confined in the whirlpools (or vortices) have a large kinetic energy. Using the uncertainty relation, estimate the order of magnitude of the kinetic energy. (This amount
would also be the additional energy penalty if we put two electrons in
the same whirlpool, instead of in two separate whirlpools, due to Pauli
exclusion principle.)
Solution: The size of a vortex is of order
l0 =
h
=
eB
6.626 1034
= 1.66 108 m
19
1.6 10
15
According to the uncertainty relation, p p h/l0 . Therefore, the kinetic

energy is
h2 eB
h eB
p2
=
=
2m
2m h
2 m
34
6.626 10
1.6 1019 15
=
2 0.067 9.1 1031
= 1.3 1020 J
Grading: 1 point for the final numerical result and 1 point for relating the
characteristic length to the momentum through the uncertainty relation, and
hence the kinetic energy. Note this is an estimate problem, hence any final
numerical result within a factor of 2 can be regarded as correct.
(g) There are also a series of plateau at RH = h/ie2 , where i = 1,2,3,... in

Tsui et al. experiment, as shown in Figure 1(b). These plateaus, known
as the integer quantum Hall effect (IQHE), were reported previously by K.
von Klitzing in 1980. Repeating (c)-(f) for the integer plateaus, one realizes
that the novelty of the FQHE lies critically in the existence of fractionally
charged quasiparticles. R. de-Picciotto et al. and L. Saminadayar et al. independently reported the observation of fractional charges at the = 1/3
filling in 1997. In the experiments, they measured the noise in the charge
current across a narrow constriction, the so-called quantum point contact
(QPC). In a simple statistical model, carriers with discrete charge e tunnel
across the QPC and generate charge current IB (on top of a trivial background). The number of the carriers n arriving at the electrode during
a sufficiently small time interval obeys Poisson probability distribution
6
with parameter
k e
,
(1)
k!
where k! is the factorial of k. You may need the following summation
P (n = k) =
e =
k
,
k!
k=0
(2)
(i) (2 point) Determine the charge current IB , which measures total charge
per unit of time, in terms of and .
Solution: The current can be calculated by the ratio of the total charge carried
by the averaged n quasiparticles to the time interval .
hn i =
kP (k) =
k=1
k e
k=1 (k 1)!
P (k)
k=0
=
where we have used
P (k) = 1. Therefore,
IB =
e
hn ie
=
Grading: 1 point for correctly calculating the average charge under the Poisson distribution and 1 point for the final expression for the charge current.
(ii) (2 points) Current noise is defined as the charge fluctuations per unit
of time. One can analyze the noise by measuring the mean square
deviation of the number of current-carrying charges. Determine the
current noise SI due to the discreteness of the current-carrying charges
in terms of and .
Solution: Similarly, the noise can be related to the averaged charge fluctuations during the time interval .
D
(n hn )2
= hn2 i hn i2
=
k=1
"
k 2 P (k) 2
P (k) +
k=0
k=1
kP (k) 2
Therefore,
D
(n hn )2 (e )2
(e)2
Grading: 1 point for correctly calculating the charge fluctuations under the
Poisson distribution and 1 point for the final expression that relates the current
noise to the charge fluctuations.
SI =
(iii) (1 point) Calculate the noise-to-current ratio SI /IB , which was verified by R. de-Picciotto et al. and L. Saminadayar et al. in 1997. (One
year later, Tsui and Stormer shared the Nobel Prize in Physics with
R. B. Laughlin, who proposed an elegant ansatz for the ground state
wave function at = 1/3.)
Solution: The noise-to-current ratio SI /IB = e = e.
Grading: 1 point for the final expression.
Figure 1: (a) Sketch of the experimental setup for the observation of the FQHE.
As indicated, a current I is passing through a two-dimensional system in the longitudinal direction with an effective length L. The Hall voltage VH is measured in
the transverse direction with an effective width W . In addition, a uniform magnetic field B is applied perpendicular to the plane. The direction of the current is
given for illustrative purpose only, which may not be correct. (b) Hall resistance
RH versus B at four different temperatures (curves shifted for clarity), adapted
from the original publication on the FQHE. The features at RH = 3h/e2 are due
to the FQHE.
Suggested Solutions for Question 2

Solution for (a)
iThe electron motion equation

dv
m -ev B .
dt
Let c eB / m we have

dv
c v .
dt
Since B Bz , thus Equation (2) can be written to be
(1)0.2point
(2)
dv x
dt c vy
dv y
c vx
dt
dv z
dt 0
(3)
The general solutions of Equation (3) are
vx v cos(c t y )
vy v cos(c t y ) ,
vz vz (t 0)
(4)0.2point
where x , y are the initial phases. Due to v z (t 0) 0 , it is indicated that the motion of an
electron remains to be perpendicular to the magnetic field. By further solving Equation (4), we
obtain the motion trajectory of the electron,
x rc sin ct
y rc cos ct ,
z0
(5)0.2point
or
x 2 y 2 rc 2
,
z 0
where rc
0.4point
mv
is the gyro radius.
eB
ii The electric current generated by the circular motion of an electron are
e ec
e2B
T 2 2 m
0.5point
Based on the definition of the magnetic moment, we have
e2B
e 2 B m 2 v 2
mv 2
2
IA
B
rc
2 2
eB
2 m
2 m
2B2
0.5point
iiiIn the case, we have vz v cos , v v sin . Since the velocity of the electron in z is
not zero, the solution (5) in (i) becomes
x rc sin ct
x 2 y 2 rc 2
r
cos
t
,
or
.
c
c
z
vt
cos
z vt cos
(6)0.4point
The orbit equation (6) indicates that the electron has a spiral trajectory. The screw pitch is
h vzT v cos
mv
cos .
eB
0.6point
Solution for (b):

(i) Since the magnetic field and the plasma are uniform z, the orbits of ions and electrons can
project into in the x-y plane. From the results of (a), we know that an ion has a left-hand circular
motion and an electron has a right-hand circular motion. Due to the linear increase of the plasma
density in x, the number of ions with upward motion is less than that with downward motion at a
given x position, which leads a net upward ion flow. Similarly, electrons have a net downward
flow. Combining the ion and election flows, we have a net upward electric current as illustrated
below in schematic drawing.
2.0points
(ii) Based on
mv 2
B from Problem (a)the total magnetic moments per unit volume
2B2
(i.e., the magnetization) for ions and electrons are
M i ,e f i ,e ( x, v ) i ,e d 3v

n( x )(
mi ,e 3/2 mi ,e ( v2 v2 )/2 kT mi ,e v 2
) e
2 v dv dv
2 kT
2B
m
m
m v 2 /2 kT
m v 2 /2 kT mi , e v
n( x ) i , e e i , e
2 v dv dv ( i ,e )1/2 e i ,e dv
2 kT
2B
2 kT
0
kT
n( x )
B
1.5point
The total magnetization for a plasma
M Mi Me
2kT
p ( x)
n( x)
B
B
Therefore, we have
2 .
0.5point
Solution for (c)

Using the equation for the magnetization field Bmx o M and the x component of the Earth's
dipole magnetic field at (x=10REy=0, z=1RE), we have
0e0.25 P0
p( z )
Bmx / Bd 0 2
1.0 .
|z R
Bd (10 RE , 0, z ) E 28.5 108 B02
1.0point
Solution for (d)

iUnder the cylindrical coordinate system, the motion of an electron can project into the
r and planethe Lorentz force is Fz ev Br .

By taking a gyro average, the averaged Lorentz force becomes
1
dB
dB
=
.
ev rc
dz
ds
2
dB dB
Since
, we have
dz ds
dB
.
F|| Fz
ds
Fz
1.0point
The acceleration of the guiding center of the electron can be obtained by the Newton's law, i.e.,
dv / /
dB
,

ds
dt
(7) 0.5point
where v// ds / dt . Equation (7) can be written to be
d 1
dB
.
( mv//2 )
dt 2
dt
(8)
d 1
1
( mv//2 mv2 ) 0 ,
dt 2
2
0.5point
dB
d 1
d 1
d
dB
d
( mv//2 ) ( mv2 ) ( B)
B
dt
dt 2
dt 2
dt
dt
dt
(9) 1.0point
Since the total kinetic energy is conserved, i.e.,

we can obtain
Combining Equations (8) and (9), we get
d
0.
dt
(10)
iiFrom the motion constant of the magnetic moment of an electron, the perpendicular velocity
increases with increase of the magnetic field, which means the parallel velocity decreases due to
the conservation of the total kinetic energy. When the electron arrives at the point "P3", its parallel
velocity decreases to zero, then the electron will not escape from the magnetic mirror field. Thus,
the initial velocityshould be
v2 0 v 2
.
B0 Bm
0.5point
Since v 0 v sin we obtain
B0
Bm
cr arcsin
0.5point
i.e., this is the condition for the electron confined in the magnetic mirror field.
Solution for (e)

Since the guiding center of the electron motion is always confined in the y=0 plane, the Earth's
dipole magnetic field B
B0 RE3
( 3 xz x ( x 2 2 z 2 ) z ) . At the initial position 6 RE , 0, 0 of
r5
an electron, the magnetic field strength
B0 RE3 2
B
Bi 5 ( x ) 0 .
216
r
0.2point
At the altitude H=200km and the latitude
BL ( Bx2 Bz2 )1/2
L 60 , thus
13
B0 .
2
0.2point
With the same argument in (d) (ii), the condition for the electron to arrive in below the altitude
H=200km with the latitude
L 60 is
0.05 2.8 .
0.6point
Solution for (f)

(i)
Based on the balance between the gravitational force and the force from the difference of the
atmospheric pressure, we obtain
A[ p ( r dr ) p ( r )] Adp
GM ( Adr )
r2
or
dp
GM ( dr )
.
r2
11
For the ideal gas, we have
NkT NmkT kT
.
V
Vm
m
12
Since the atmospheric temperature is constant, Equation (12) can written to be
0.5point
p0
13 0.5point
Inserting Equation (13) into (11), we have
0GMdr
p0
14
By integrating Equation (14), we obtain
0 e
0GM 1
P0
1
( )
r RE
0 e
0 gRE
P0
(1
RE
)
r
0 e
0 gH
P0
15 0.5point
Thus, the ration of the atmospheric density at the altitudes H=160km and H=220km is
( H1 160km )
e
( H 2 220km)
0 g ( H 2 H1 )
P0
2.44 103 .
16 0.5point
iiAt the altitude H=160kmthe collision frequency of atmosphericmolecules is

0 0e
0
0 gH
P0
0.93 / s .
17 0.5point
Since the lifetime of the oxygen atom in the first excited state lasts about 110s, the oxygen atom
will collide with other molecules over one hundred times. Thus, the high-frequent collision
between particles can de-excite oxygen atoms before they have a chance to radiate. But, the
oxygen atoms in the second excited state will emit photons since their lifetime is so short. Thus,
aurora at H=160km is resulted from emission of the oxygen atoms in the second excited state. The
wavelength is
hc (6.626 1034 J s)(2.998 108 m / s)
562nm .
eV
(1.6 1019 J / eV )(2.21eV )
0.8point
The color for this wavelength is green.

At the altitude H=220kmthe collision frequency of atmosphericmolecules is

0 0e
0
0.2point
0 gH
P0
3.8 10 4 / s .
18 0.5point
All oxygen atoms in the first and second excited states have a chance to radiate since the collision
frequency is so low. Because the number of the oxygen atoms in the first excited state is much
larger than in the second excited state, we observe the aurora color is from the emission of the
oxygen atoms in the first excited state. The wavelength is
hc (6.626 1034 J s)(2.998 108 m / s)
633nm .
eV
(1.6 1019 J / eV )(1.96eV )
0.8point
The color for this wavelength is red.
0.2point
Solution for (g)

At the geosynchronous orbit, we have
mM
v2
m(h RE ) 2 .
m
(h RE )2
h RE
or
(h RE )3
Using
GM
gRE2
g
RE3 .
RE
2
2 / (24 3600s) , we have
h 5.6 RE .
0.5point
Using w F x Ek , we have
1
( f m f s )sx mvs2 .
2
or
1 2 Bs2
B2
vs
m .
2
2 0 2 0
19 1.0point
where is the mass density. At the position x RE h 6.6 RE , y z 0 , the Earth's dipole
magnetic field strength is
Bd
B0
~ 100nT
290
From Equation (19),
v 330km / s .
we obtain
0.5point
With this speed of the solar wind, the position of the dayside magetopause is 6.6RE at the
geosynchronous orbit. Thus, a geosynchronous satellite could be damaged by the solar wind.
Solution for Question 3
Figure 1 shows a Fabry-Perot (F-P) etalon, in which air pressure is tunable. The F-P etalon
consists of two glass plates with high-reflectivity inner surfaces. The two plates form a cavity in
which light can be reflected back and forth. The outer surfaces of the plates are generally not
parallel to the inner ones and do not affect the back-and-forth reflection. The air density in the
etalon can be controlled. Light from a Sodium lamp is collimated by the lens L1 and then passes
through the F-P etalon. The transmitivity of the etalon is given by T
1
, where
1 F sin 2 ( / 2)
4nt cos
4 R , R is the reflectivity of the inner surfaces,
is the phase shift of two
1 R
neighboring rays, n is the refractive index of the gas, t is the spacing of inner surfaces, is the
incident angle, and is the light wavelength.
Gas in
Pin valve
t o vacuum pump
n
Microscope
N
B
S
t
Na Lamp
Fringes
L1
F1
L2
F-P Etalon
Figure 1
The Sodium lamp emits D1 ( 589.6nm ) and D2 ( 589nm ) spectral lines and is located
in a tunable uniform magnetic field. For simplicity, an optical filter F1 is assumed to only allow
the D1 line to pass through. The D1 line is then collimated to the F-P etalon by the lens L1.
Circular interference fringes will be present on the focal plane of the lens L2 with a focal length
f=30cm. Different fringes have the different incident angle . A microscope is used to observe
the fringes. We take the reflectivity R= 90% and the inner-surface spacing t=1cm.
Some useful constants : h 6.626 1034 J s , e 1.6 1019 C , me 9.1 10 31 kg , c 3.0 108 ms 1 .
Page 1 of 10
a 3points The D1 line ( 589.6nm ) is collimated to the F-P etalon. For the vacuum
case (n=1.0), please calculate (i) interference orders mi , (ii) incidence angle
i and (iii)
diameter Di for the first threei=123fringes from the center of the ring patterns on the focal
plane.
Solution
The transmittivity of the F-P etalon is given by
T
1
1 F sin 2
For bright fringes, we have
T 1 i.e. sin 2
m
2
2nt cos m
For n=1.0t=1cm 589.6nm thus:
cos i
mi
mi
2nt 33921.3
(a1)1 point if Eqs. (a2-a3) are not correct.
Because of cos 1 so the orders of the first three fringes are:
m1 33921, m2 33920, m3 33919
(a2)1 point
The incident angles of the first three fringes are
1 0.2410 , 2 0.5020 ,3 0.6670
(a3)1 point
The fringe diameter is given by
Di 2 f tan i 2 f i
(a4)0.5 point if Eq(a5) is not correct.
For the focal length f=30cmthus:
D1 2.52mm, D2 5.26mm, D3 6.99mm (a5)1 point
Page 2 of 10
of the spectral line is defined as the full width

of half maximum (FWHM) of light transmitivity T regarding the phase shift . The resolution of
the F-P etalon is defined as follows: for two wavelengths and , when the central
phase difference of both spectral lines is larger than , they are thought to be resolvable;
then the etalon resolution is / when . For the vacuum case, the D1 line
( 589.6nm ), and because of the incident angle 0 , take cos 1.0 , please calculate
b3 points As shown in Fig. 2, the width
ithe width of the spectral line.

iithe resolution
of the etalon.
T
0.5
2m
Figure 2
Solution
The half maximum occurs at
2m
(b1)0.2 point if Eq.(b3) is wrong.
Given that T 0.5 thus
F sin 2
1 (b2)0.2 point if Eq.(b3) is wrong.
4
2(1 R) 2(1 0.9)
0.21 rad (or 12.03degree) (b3)1 point

F
R
0 .9
is given by:
4nt cos
For a small , thus

The phase shift
4nt cos
b4 (1 point if Eq. (b5) is wrong.

2
For and 589.6nm we get
nt F cos 3.14 1.0 1.0 102 360 1.0
1.01106 (b5) (2 points

9
589.6 10
1.5 point if the final value of Eq. (b5) is wrong.
Page 3 of 10
c1 point As shown in Fig. 1, the initial air pressure is zero. By slowly tuning the pin valve,
air is gradually injected into the F-P etalon and finally the air pressure reaches the standard
atmospheric pressure. On the same time, ten new fringes are observed to produce from the center
of the ring patterns on the focal plane. Based on this phenomenon, calculate the refractive index of
air nair at the standard atmospheric pressure.

Solution
From Question (a), we know that the order of the 1st fringe near the center of ring patterns is m=33921
at the vacuum case (n=1.0). When the air pressure reaches the standard atmospheric pressure, the order
of the 1st fringe becomes m+10, so we have:
nair
m 10 33931
1.00029 . (c1)1 point

2t
33921
0.2 point for appearing the term of (m+10) when the final value of Eq.(c1) is wrong.
Or
0.8 point for the correct final expression (including other correct forms) without the correct
value.
Page 4 of 10
d (2 points) Energy levels splitting of Sodium atoms occurs when they are placed in a
magnetic field. This is called as the Zeeman effect. The energy shift given by E
m j gk B B
where the quantum number mj can be JJ-1-J+1-JJ is the total angular quantum number,
gk is the Land factor,
he
is Bohr magnetonh is the Plank constante is the electron
4me
chargeme is the electron mass, B is the magnetic field. As shown in Fig. 3, the D1 spectral line is
emitted when Sodium atoms jump from the energy level 2P1/2 down to 2S1/2. We have J
1
for
2
both 2P1/2 and 2S1/2 . Therefore, in the magnetic field, each energy level will be split into two levels.
We define the energy gap of two splitting levels as E1 for 2P1/2 and E2 for 2S1/2 respectively (E1
<E2). As a result, the D1 line is split into 4 spectral lines (a, b, c, and d), as showed in Fig. 3.
Please write down the expression of the frequency ( ) of four lines a, b, c, and d.
mj
1/2
-1 /2
P 1 /2
E 1
58 9.6 nm
1/2
a b
S 1 /2
c d
-1/2
E 2
Figure 3
Solution
The frequency of D1 line (2P1/2 to 2S1/2) is given by 0 c /
589.6nm
When magnetic field B is appliedthe frequency of the line a,b,c,d are expressed as:
1) 2P1/2 (mj=-1/2) 2S1/2 (mj =1/2): frequency of (a) a 0
1
E1 E2 0.5 point
2h
1
E2 E1 0.5 point
2h
1
E2 E1 0.5 point
3) 2P1/2 (mj=-1/2) 2S1/2 (mj=-1/2): frequency ofc c 0
2h
1
E1 E2 0.5 point
4) 2P1/2 (mj=1/2) 2S1/2 (mj=-1/2): frequency ofd
d 0
2h
2) 2P1/2 (mj=1/2) 2S1/2 (mj=1/2): frequency ofb

b 0
The results maybe have other correct forms.

But, 0.4 point for each result without the coefficient of
1 / 2 .
Page 5 of 10
(e) (3 points) As shown in Fig. 4, when the magnetic field is turned on, each fringe of the D1 line
will split into four sub-fringes (1, 2, 3, and 4). The diameter of the four sub-fringes near the center
is measured as D1 D2 D3 and D4 . Please give the expression of the splitting energy gap E1
of 2P1/2 and E2 of 2S1/2.
m
m-1
m-1
B=0
3 1
D4
D2
3
B0
D1
D3
2 4
Figure 4
Solution
m 1, cos m 1
m2
2
2nt cos m m 1
(e1)0.2point if Eq. (e4) is wrong.
m2
2
m
2nt
, m m'
1
m'2
2
m2 m'2
2
2nt
m
2nt
2 f m Dm

Dm2 Dm' 2 m
8f 2
2nt
Dm2 Dm'2
8f 2
(e4)1 point
The lines a, b, c, and d correspond to sub-fringe 1, 2, 3, and 4. From Question (d), we have.
The wavelength difference of the spectral line a and b is given by:
D22 D12
8f 2
Page 6 of 10
E1 h( b a ) , E2 h( d b )
or E1 h d c , E2 h c a
(e5)
0.5 point for each subequation in Eq (e5) if Eqs. (e6) and (e7) are totally wrong.
The wavelength difference of the spectral line a and b is given by:
D22 D12
8f 2
Then we obtain
E1 h 1
hc
D22 D12 hc D22 D12
8f 2
8f 2
D42 D32 hc D42 D32

or E1 h 1
(e6)1 point
8f 2
8f 2
hc
Similarly, for E2we get
D42 D22
8f 2
D42 D22 hc D42 D22

E2 h 2
8f 2
8f 2
hc
D32 D12 hc D32 D12

or E1 h 1
(e7)1 point
8f 2
8f 2
hc
(Eqs (e6 and e7) have other correct forms which should be in terms of D1 D2 D3 and D4 )
(2.5 points for the final expressions only with the incorrect coefficients. )
Page 7 of 10
f3 pointsFor the magnetic field B=0.1Tthe diameter of four sub-fringes is measured as
D1 3.88mm D2 4.05mm D3 4.35mm and D4 4.51mm . Please calculate the

Land factor gk1 of 2P1/2 and gk2 of 2S1/2.
Solution
Given that B=0.1Tso we have:
heB
6.626 1034 0.1
5.79 10 6 eV (f1)0.2point if Eq. (f4) is wrong.

B B
31
4me 4 3.14 9.1 10
E1 g k 1b B
hc
D22 D12
(f2)
8f 2
or, E1 g k 1b B
hc
D42 D32
0.5point if Eq. (f4) is wrong.
8f 2
For the D1 spectral line 589.6nm so we can get:
hc
6.626 1034 3 108

2.11eV (f3)0.2point if Eq. (f4) is wrong.
5.896 10 7 1.6 1019
thus
D2 D2
2.11
2.11
4.05 10 3 3.88 10 3
g k1
2 2 1
6
6
5.79 10
8f
5.79 10
8 0.3 0.3
0.68
1.5 points
or g k1
D42 D32
2.11
2.11
4.51 10 3 4.35 10 3
5.79 10 6
8f 2
5.79 10 6
8 0.3 0.3
0.72
Similarly, we get:
gk 2
D2 D2
2.11
2.11
4.51 10 3 4.05 10 3
4 2 2
6
6
5.79 10
8f
5.79 10
8 0.3 0.3
or g k 2
1.99 1.5 points
D32 D12
2.11
2.11
4.35 10 3 3.88 10 3
5.79 10 6
8f 2
5.79 10 6
8 0.3 0.3
1.95
(2 points for the correct final expressions if the final values are wrong.)
*Commentthe theory value of g k 1 and g k 2 is 2/3 and 2
Page 8 of 10
g (2 points) The magnetic field on the sun can be determined by measuring the Zeeman effect
of the Sodium D1 line on some special regions of the sun. One observes that, in the four split lines,
the wavelength difference between the shortest and longest wavelength is 0.012nm by a solar
spectrograph. What is the magnetic field B in this region of the sun?
Solution
We have E1 g k 1 B B and E2 g k 2 B B ;
The line a has the longest wavelength and the line d has the shortest wavelength line. The energy
difference of the line a and d is
E E1 E2 g k1 g k 2 B B . (g1)0.5point if Eq. (g3) is wrong.

(g2)0.5 point
g k1 g k 2 B B
g30.5 point
h
he
4me
So the magnetic field B is given by:

B
4me c
2 g k1 g k 2 e
4 3.14 9.1 10 31 0.012 10 9 3 108
589.6 10
9 2
2.67 1.6 10 19
(g4)1 point
0.2772T
2772.1Gauss
0.5 point if the first line in Eq (g4) is correct.
Page 9 of 10
(h)3 points A Light- Emitting Diode (LED) source with a central wavelength 650nm
and spectral width 20nm is normally incident ( 0 ) into the F-P etalon shown in Fig. 1.
For the vacuum case, find (i) the number of lines in transmitted spectrum and (ii) the frequency
width of each line
Solution
The wavelength of transmitted spectral lines is given by:
2nt mm
(h1)0.5 point if Eq. (h2) is wrong.
mc
2nt
c
m
1.5 1010 Hz
2nt
(h2)1 point
The frequency width of the input LED is
3 10 20 10
650 10 9 2
8
(h3)
9
1.42 1013 Hz
0.5point if the first line in Eq. (h3) is correct.
So we have the number of transmitted spectral line
s
m
(h4)1 point
1.42 10
946
1.5 1010
13
The spectral width of transmitted spectral line is
2
nt F
, then we have
c
nt F
(h5)1 point
8
3 10
5.0 108 Hz
3.14 1.0 10 10 3 360

Page 10 of 10
Theoretical Question T1
Page 1 of 3
Mechanics of a Deformable Lattice
(Total Marks : 20)
Here we study a deformable lattice hanging in gravity which acts as a deformable physical
pendulum. It has only one degree of freedom, i.e. only one way to deform it and the
configuration is fully described by an angle . Such structures have been studied by famous
physicist James Maxwell in 19th century, and some surprising behaviors have been
discovered recently.
As shown in the figure 1, N2 identical triangular plates (red triangle) are freely hinged by
identical rods and form an N N lattice (N > 1). The joints at the vertices are denoted by
small circles. The sides of the equilateral triangles and the rods have the same length l. The
dashed lines in the figure represent four tubes; each tube confines N vertices (grey circles) on
the edge and the N vertices can slide in the tube, i.e. the tube is like a sliding rail.
The four tubes are connected in a diamond shape with two angles fixed at 60 and another
two angles at 120 as shown in Figure 1. Each plate has a uniform density with mass M, and
the other parts of the system are massless. The configuration of the lattice is uniquely
determined by the angle , where 0 60 (please see the examples of different angle
in Figure 1). The system is hung vertically like a curtain with the top tube fixed along
the horizontal direction.
The coordinate system is shown in Figure 2. The zero level of the potential energy is
defined at y = 0. A triangular plate is denoted by a pair of indices (m,n), where m, n = 0, 1, 2,
, N 1 representing the order in the x and y directions respectively. A(m,n), B(m,n) and
C(m,n) denote the positions of the 3 vertices of the triangle (m,n). The top-left vertex, A(0,
0) (the big black circle), is fixed.
The motion of the whole system is confined in the x-y plane. The moment of inertia of a
uniform equilateral triangular plate about its center of mass is I = Ml 2 /12. The free fall
acceleration is g. Please use Ek and Ep to denote kinetic energy and potential energy
respectively.
Figure 1
Page 2 of 3
Figure 2
Section A: When N=2 (as shown in figure 3):
Figure 3
A1
What is the potential energy Ep of the system for a general angle when N = 2?
2 points
A2
What is the equilibrium angle E of the system under gravity when N = 2?
1 point
A3
The system follows a simple harmonic oscillation under a small perturbation from
equilibrium. Calculate the kinetic energy of this system in terms of d()/dt. Calculate
the oscillation frequency fE when N = 2.
5 points
Page 3 of 3
Section B: For arbitrary N :
B1
What is the equilibrium angle E under gravity when N is arbitrary?
3 points
B2
Consider the case when . Under a small perturbation of angle , the change of
potential energy of the system is p 1 , the kinetic energy of the system is k 2 ,
and the oscillation frequency is E 3 . Find the values of 1, 2 and 3.
3 points
Section C: A force is exerted on one of the 3N2 triangle vertices so that the system maintains at m = 60.
g
Figure 4
C1
Which vertex should we choose to minimize the magnitude of this force?
1 point
C2
What are the direction and magnitude of this minimum force? Describe the direction in
terms of the angle F defined in Figure 4.
5 points
Page 1 of 4
The Expanding Universe
(Total Marks : 20)
The most outstanding fact in cosmology is that our universe is expanding. Space is
continuously created as time lapses. The expansion of space indicates that, when the
universe expands, the distance between objects in our universe also expands. It is
convenient to use comoving coordinate system = (, , ) to label points in our
expanding
universe,
in
which
the
coordinate
distance
= |2 1 | = (2 1 )2 + (2 1 )2 + (2 1 )2 between objects 1 and 2 does not
change. (Here we assume no peculiar motion, i.e. no additional motion of those objects
other than the motion following the expansion of the universe.) The situation is illustrated
in the figure below (the figure has two space dimensions, but our universe actually has three
space dimensions).
The modern theory of cosmology is built upon Einsteins general relativity. However, under
proper assumptions, a simplified understanding under the framework of Newtons theory of
gravity is also possible. In the following questions, we shall work in the framework of
Newtons gravity.
To measure the physical distance, a scale factor () is introduced such that the physical
distance p between the comoving points 1 and 2 is
p = (),
The expansion of the universe implies that () is an increasing function of time.
On large scales scales much larger than galaxies and their clusters our universe is
approximately homogeneous and isotropic. So let us consider a toy model of our universe,
which is filled with uniformly distributed particles. There are so many particles, such that
we model them as a continuous fluid. Furthermore, we assume the number of particles is
Page 2 of 4
conserved.
Currently, our universe is dominated by non-relativistic matter, whose kinetic energy is
negligible compared to its mass energy. Let m () be the physical energy density (i.e.
energyper unit physical volume, which is dominated by mass energy for non-relativistic
matter and the gravitational potential energy is not counted as part of the physical energy
density) of non-relativistic matter at time . We use 0 to denote the present time.
A
Derive the expression of m () at time in terms of (), (0 ) and m (0 ).
2 points
Besides non-relativistic matter, there is also a small amount of radiation in our current
universe, which is made of massless particles, for example, photons. The physical
wavelength of massless particles increases with the universe expansion as p (). Let
the physical energy density of radiation be r ().
B
Derive the physical energy density for radiation r () at time in terms of (), (0 ) and
r (0 ).
2 points
Consider a gas of non-interacting photons which has thermal equilibrium distribution. In

this situation, the temperature of the photon depends on time as () [()] .
C
Calculate the numerical value of .
2 points
Consider the thermodynamics of one type of non-interacting particle X. Note that the space
expansion is slow enough and thermally isolated such that the entropy of X is a constant in
time. Let the physical energy density of X be X (), which includes mass energy and
internal energy. Let the physical pressure be X ().
D
Derive dX ()/d in terms of (), d()/d,X (),and X ().
4 points
Page 3 of 4
Consider a star S. At the present time 0 , the star is at a physical distance p = (0 ) away
from us, where is the comoving distance. Here we ignore the peculiar motion, i.e. assume
that both the star and us just follow the expansion of the universe without additional
motion.
The star is emitting energy in the form of light at power e , which is isotropic in every
direction. We use a telescope to observe its starlight. For simplicity, assume the telescope
can observe all frequencies of light with 100% efficiency. Let the area of the telescope lens
be .
Derive the power received by the telescope r from the star S, as a function of , , e , the
scale factor (e ) at the starlight emission time e , and the present (i.e. at the observation
time) scale factor (0 ).
If there were no gravity, the expansion speed of the universe should be a constant. In
Newtons framework, this can be understood as that, without force, matter just moves
away from each other with constant speed and thus d()/d is a constant depending on
the initial condition.
Let us now consider how Newtons gravity affects the scale factor (), in a universe filled
with non-relativistic matter in a homogeneous and isotropic way.
As illustrated in the above figure, let us assume C is the center of our universe (this
assumption can be removed in Einsteins general relativity, which is beyond the scope of
this question). We slice matter into thin shells around C. Let us focus on one thin shell (the
sphere in the above figure) whose comoving distance from the center is (recall that this
comoving distance is a constant in time).
4 points
Page 4 of 4
Use the motion of the shell to find a relation between d()/d, () and the density of
mass energy (). (In the final relation, if you encounter a constant depending on the
initial condition, it can be kept as it is.)
5 points
Based on the model described in Part (F), is the expansion of the universe
(a) accelerating or (b) decelerating? Choose from (a) or (b).
1 points
For your information, in 1998, a new type of energy component of our universe is discovered. It actually
changes the conclusion in Part (G).
Page 1 of 5
Magnetic Field Effects on Superconductors
(Total Marks: 20)
An electron is an elementary particle which carries electric charge and an intrinsic magnetic
moment related to its spin angular momentum. Due to Coulomb interactions, electrons in
vacuum are repulsive to each other. However, in some metals, the net force between electrons
can become attractive due to the lattice vibrations. When the temperature of the metal is low
enough, lower than some critical temperature Tc, electrons with opposite momenta and
opposite spins can form pairs called Cooper pairs. By forming Cooper pairs, each electron
reduces its energy by D compared to a freely propagating electron in the metal which has
p2
energy
, where p is the momentum and me is the mass of an electron. The Cooper pairs
2me
can flow without resistance and the metal becomes a superconductor.
However, even at temperatures lower than Tc, superconductivity can be destroyed if the
superconductor is under the influence of an external magnetic field. In this problem, you are
going to work out how Cooper pairs can be destroyed by external magnetic fields through two
effects.
The first is called the paramagnetic effect, in which all the electrons can lower their energy by
aligning the electron magnetic moments parallel to the magnetic field instead of forming
Cooper pairs with opposite spins.
The second is called the diamagnetic effect, in which increasing the magnetic field will
change the orbital motion of the Cooper pairs and increase their energy. When the applied
magnetic field is stronger than a critical value Bc, this increase in energy becomes higher than
2D . As a result, the electrons do not prefer forming Cooper pairs.
Recently, a type of superconductor called Ising superconductors was discovered. These
superconductors can survive even when the applied magnetic field is as strong as 60 Tesla,
comparable to the largest magnetic fields which can be created in laboratories. You will work
out why Ising superconductors can overcome both the paramagnetic and the diamagnetic
effects of the magnetic fields.
Page 2 of 5
A. An Electron in a Magnetic Field
Let us consider a ring with radius r, charge e and mass m. The mass and the charge density
around the ring are uniform (as shown in Figure 1).
r
y
x
Figure 1
A1
What is the angular momentum

with angular velocity ?
(magnitude and direction) of this ring if the ring is rotating
2 points
The magnitude of the magnetic moment is defined as M IA , where I is the current and A is
A2 the area of the ring. What is the relationship between the magnetic moment
angular momentum of the ring?
Suppose the normal direction of the ring is

magnetic field as shown in Figure 2.
and the
2 points
and it makes an angle with the applied
For the ring described in Part (A1), what is the potential energy U of this ring if the ring is
placed in a uniform magnetic field Bz pointing to the z-direction? You should assume the
potential energy to be zero when = /2.
Bz
A3
2 points
Figure 2
Page 3 of 5
An electron carries an intrinsic angular momentum, which is called spin. We know that the
magnitude of spin in a particular direction is 2 , where

and h is the Plancks
constant.
A4
What are the values of the potential energy Uup and Udown for electrons with spins parallel and
anti-parallel with the applied magnetic field respectively?
Please express your results in terms of the Bohr magneton B
1 point
e
5.788 105 eV T 1
2me
and the magnetic field strength B.

According to quantum mechanics, the potential energy U up and U down are twice the values Uup
and Udown found in Part (A4). Assuming that the applied magnetic field is 1 Tesla. What is the
A5 potential energy U up and U down for an electron with spin parallel and anti-parallel to the
applied magnetic field respectively? In the rest of this question, you should use the
expressions for U up and U down for your calculations.
B. Paramagnetic effect of the magnetic field on Cooper pairs

In the question below, we consider the paramagnetic effect of an external magnetic field on
Cooper pairs (as shown in Figure 3).
Theoretical studies show that in superconductors, two electrons with opposite spins can form
Cooper pairs so that the whole system saves energy. The energy of the Cooper pair can be
expressed as
p12
p2
2 2 , where the first two terms denote the kinetic energy of the
2me 2me
Cooper pair and the last term is the energy saved for the electrons to form a Cooper pair.
Here, D is a positive constant.
B
Cooper pair
Figure 3
1 point
Page 43 of 5
B1
Assuming that the effect of the external magnetic field is only on the spins of the electrons,
not on the orbital motions of the electrons. What is the energy ES of the Cooper pair under a
uniform magnetic field
must have opposite spins.
? Recall that the electrons which form a Cooper pair
1 point
In the normal state (non-superconducting state), electrons do not form Cooper pairs. What is
the lowest energy E N for the two electrons under a uniform in-plane magnetic field
B2
pointing to the x-direction? Please use the U up and U down defined in Part (A5)
1 point
in your calculations and ignore the effects of the magnetic field on the orbital motions of the
electrons.
At zero temperature, a system will favor the state with the lowest energy. What is the critical
B3 value B in terms of , such that for B B superconductivity will disappear?
P
P
1 point
C. Diamagnetic effect of the magnetic field on Cooper pairs

In the question below, we are going to ignore the effects of magnetic fields on the spins of
the electrons and consider the effects of external magnetic fields on the orbital motions of
the Cooper pairs.
At zero temperature, the energy difference between the superconducting state and the
= (0,0, ) can be written as
normal state for a superconductor in a magnetic field
d 2 e2 Bz2 x 2
F
dx .
4me dx 2
me
Here y (x) is a function of position x and independent of y. 2 ()denotes the probability

of finding a Cooper pair near x. Here, 0 is a constant and it is related to the energy saved
by forming Cooper pairs. The second and the third terms in F are related to the kinetic
energy of the Cooper pairs taking into account the effect of the magnetic field.
At zero temperature, the system prefers to minimize its energy F . In this case, y (x) takes
1
2 4 x2
the form x
e , with l > 0 .

Find l in terms of e , Bz , and
The following integrals may be useful:

C1
3 points
Here a is a constant.
ax 2
dx
xe
2 ax 2
dx
1
.
2a a
Page 5 of 5
C2
Work out the critical value of Bz in terms of , at which the superconducting state is no
longer energetically favorable.
2 points
D Ising Superconductors
In materials with spin-orbit coupling (spin-spin couplings can be ignored), an electron with
momentum experiences an internal magnetic field
. On the other hand, an
electron with momentum
experiences an opposite magnetic field
. These
internal magnetic fields act on the spins of the electrons only as shown in Figure 4.
Superconductors with this kind of internal magnetic fields are called Ising superconductors.
Figure 4: Two electrons form a Cooper pair. Electron 1 with momentum
experiences
internal magnetic field
but electron 2 with momentum
experiences an
opposite magnetic field

arrows.
. The internal magnetic fields are denoted in dashed
D1 Then what is the energy EI for a Cooper pair in an Ising superconductor?
1 point
In the normal state of the material with spin orbit coupling, what is the energy || for the two
|| = ( , 0,0) ? (Here the internal
electrons under a uniform in-plane magnetic field
D2
|| . You should also ignore the effects of the
magnetic fields still exist and perpendicular to
in-plane magnetic field on the orbital motions of the Cooper pairs.)
2 points
|| | > I, || < I ?
D3 What is the critical value BI such that for |
1 point
Marking Scheme T1
Page 1 of 15
(Full Mark = 20)
Part
A1
Model Answer
Marks
The potential energy for N = 2 is:
p () = cm.(0,0) 4 + 2 (0.5 points)
- Eq. (1)
where
cm.(0,0) =
3
sin ( 6
3
+ ) (0.5 points)
- Eq. (2)
is the y coordinate of center of mass of triangle (0,0), and

= A(0,1) A(0,0)
= [sin ( 3 + ) + sin ( 3 )]
= 3 cos (0.5 points)
- Eq. (3)
is the translational difference of two neighbouring triangles in y-direction. Solving Eqs.

(1), (2) and (3), we obtain
2
p () = 3 (43 cos + 3 sin ) (0.5 points)
- Eq. (4)
A2
At equilibrium, the potential energy reaches a minimum, which gives:

p ()
=E
= 0 (0.5 points)
3 sin E + 3 cos E = 0
- Eq. (5)
- Eq. (6)
Marking Scheme T1
Page 2 of 15
or
E = tan1
A3
3
4
(0.5 point)
- Eq. (7)
If the total energy of the oscillation has the following form

1
(, ) = p + k = 2 ()2 + 2 ( )2 , (0.5 points)
- Eq. (8)
where Ep and Ek are the potential and kinetic energies of the system respectively, then the
motion is a simple harmonic oscillation with angular frequency = /. Here =
E . Under a small perturbation, the potential energy change is:
p
1 2 p
|
()2
2 d 2 =
E
1 2
= ( ) ( ) (43 cos E + 3 sin E )()2
2 3
=
57
()2
3
(1 point)
- Eq. (9)
The total kinetic energy of the system includes the translational kinetic energy of every
plate and the rotational kinetic energy of every plate relative to its center of mass
k = ktrans + krot
- Eq. (10)
The rotational kinetic energy is

1 2
krot = 4
2
12
( )2 = 2 ( )2 (0.5 points)
6
- Eq. (11)
ktrans can be obtained by considering the motion of the center of mass of each triangle and
setting N = 2.
c.m.(,) = (2 cos ) + (2 cos ) cos
c.m.(,) = (2 cos ) sin 3
+
cos ( + ),
3 3
6
sin ( + 6 ).
(0.5 point)
Differentiating and substituting

sin =
3
19
, cos =
4
19
33
, sin ( + 6 ) = 219 , cos ( + 6 ) = 219,
Marking Scheme T1
Page 3 of 15
7 3
c.m.(,) = (2 + + )
,
6 57
2
2
2
c.m.(,)
= c.m.(,)
+ c.m.(,)
=
trans
c.m.,k
=
c.m.(,) =
3(2 1)
219
(12+6+7)2 +27 2
( )2 ,
228
(1 point)
2
164 2
2
2
2
[c.m.(0,0) + c.m.(0,1)
+ c.m.(1,0)
+ c.m.(1,1)
]=
( )2 .
2
57
347
trans
ktrans = c.m.,k
+ krot = 114 2 ( )2 .
(1 point)
Alternatively, another way to get trans is based on the center of mass of the whole system:
trans
rot
k = c.m.,k
+ r.c.,k
(0.5 points)
- Eq. (12)
where
trans
r.c.,k
=
2
2
2
2
[r.c.(0,0)
+ r.c.(1,0)
+ r.c.(0,1)
+ r.c.(1,1)
]
- Eq. (13)
is the translational kinetic energy relative to the center of mass of the system and
trans
c.m.,k
=
4
2
2
c.m.
- Eq. (14)
is the translational kinetic energy of the center of mass of the system.

The center of mass of each of the 22 = 4 triangles always form diamond shape with
lateral length 2l cos . The center of mass of the whole system is at the center of the
diamond shape. Hence
r.c.(0,0) = r.c.(1,1) =
(3 cos )
|
=
E
r.c.(1,0) = r.c.(0,1) =
( cos )
=E
- Eq. (15)
Substituting Eqs. (14) and (15) into Eq. (13), we obtain

trans
r.c.,k
= 4 sin E2 2 ()2
trans
For c.m.,k
,
- Eq. (16)
Marking Scheme T1
Page 4 of 15
dc.m. 2
c.m. = (
) +(
dc.m. 2
) |
- Eq. (17)
=E
is the velocity of the center-of-mass of the four triangular plates, with

1
c.m. = c.m.(0,0) + (B(0,0) + A(1,0) )
2
=
3
cos (6
3
+ ) + 2 cos
- Eq. (18)
1
c.m. = c.m.(0,0) +
2
=
3
sin ( 6
3
+ )
3
cos
2
- Eq. (19)
Substituting Eqs. (17), (18) and (19) and into Eq. (14), we obtain
2
trans
c.m.,k
= (3 + 10 sin2 ) 2 ()2 (0.5 points)
- Eq. (20)
Combining Eqs. (12), (16) and (20), we obtain

trans
trans
k = krot + r.c.,k
+ c.m.,k
5
= ( + 14 sin2 ) 2 ()2
6
347
= 114 2 ()2
(1.5 points)
- Eq. (21)
According to Eqs. (8), (9) and (21),
1
2
57
3
347
2
114
1
2
3857 (0.5 points)

347
- Eq. (22)
[Note 1: 0.5 point should be deducted if there are numerical mistakes, but all steps are
correct.
Note 2: A rough estimate of ~ can get 0.5 points out of 5 points.]
Marking Scheme T1
Page 5 of 15
B1
For arbitrary N, the total potential energy
p = 1
,=0 p (, )
- Eq. (23)
where
1
p (, ) = 3 [A(,) + B(,) + C(,) ]
- Eq. (24)
(0.5 points for Eqs. (23) and (24))

and
A(,) = sin ( ) sin ( + ) = 3 cos

3
3
B(,) = A(,) sin = 3 cos sin
C(,) = A(,) sin (3 + ) = 3 cos sin ( 3 + )
- Eq. (25)
(0.5 points for all three correct coordinates)

Thus,
1
p (, ) = 3 [33 cos + sin + sin (3 + )]
- Eq. (26)
and
p =
p (, )
,=0
= 3 1
,=0 [33 cos + sin + sin ( 3 + )] (0.5 points) - Eq. (27)
Using the mathematical relations
1=
=0
1
=0
1=
and
1
1
=0 = =0 =
Eq. (27) becomes
(1)
2
- Eq. (28),
Marking Scheme T1
Page 6 of 15
1
33( 1) cos
p = 2 [
+ sin + sin ( + )]
3
2
3
or
At equilibrium,
= 3 2 [
dp
d
3(32) cos
2
+ 2 sin ] (1 points)
- Eq. (29)
= 0, therefore
33(1) sin E
+ cos E + cos ( 3 + E ) = 0
3
E = tan1 (32) (0.5 points)
- Eq. (30)
- Eq. (31)
[Remark: Increasing lowers each triangle relative to its vertex A, but globally raises the
system, i.e. the bottom tube is raised higher. When , the global displacement
dominates, consequently .]
B2
Under a small perturbation, the potential energy change, according to Eq. (29) is
1 2 p
p 2
d2 =
E
()2 ~ 3 or 1 = 3 (0.5 points)
3
- Eq. (32)
[Remark: There are N2 triangles and the y coordinate of the total center of mass is
proportional to N, hence ~ and = . Using this argument to derive the
correct can also get 0.5 points.]
The kinetic energy of a triangle includes the translational energy of its center of mass and
the rotational energy about its center of mass. Hence the total kinetic energy of the N2
triangles is
k = , c.m.(,) + , r.c.(,)
- Eq. (33)
where
1 2
r.c.(,) = 2
12
( )2 = 2 ( )2 ~1
24
- Eq. (34)
and
c.m.(,) =
=
2 c.m.(,)
( )2
2
dc.m.(,) 2
[(
dc.m.(,) 2
) +(
) ]
=E
(0.5 points)
- Eq. (35)
Marking Scheme T1
Page 7 of 15
Since
c.m.(,) = A(,) +
3
cos ( + )
3
6
3
= (2 + ) cos + cos
sin
2
6
and
c.m.(,) = A(,) +
3
sin ( + )
3
6
= 3 cos +
3
cos
6
+ 2 sin
- Eq. (36)
(0.5 points for correct x and y)
dc.m.(,)
1
3
= [(2 + ) sin sin
cos ]
d
2
6
dc.m.(,)
1
3
= [3 sin
sin + cos ]
d
6
2
we have
1
c.m.(,) = 2
2 ( )2
(42 + 42 + 4 + 2 + 2) sin2 E
+
23
3
( ) sin E cos +
- Eq. (37)
Since E ~ in Eq. (31), we have

1
c.m.(,) = 2 2 + + ~1 (0.5 points)
- Eq. (38)
According to Eqs. (33), (34) and (38), we have

k =
c.m.(,) +
r.c.(,) ~ 1~ 2
or 2 = 2 (0.5 points)
- Eq. (39)
Marking Scheme T1
Page 8 of 15
[Remarks: ~ because there are N2 triangles, each contribute r.c. (, )~
(relative-to-center-of-mass kinetic energy) and c.m. (, )~ (center-of-mass kinetic
energy).]
Note that r.c. (, )~1 is true for arbitrary while c.m. (, )~1 is only true for the
special case of E 0 or .
Therefore
p
E ~ ~
k
or 3 = 0.5 (0.5 points)

C1
- Eq. (40)
The minimum force should act on the farthest triangle (N 1, N 1), whose motion can be
decomposed into the motion of the center of mass and the rotation around the center of
mass: = c.m. + rot . As shown in the figure, rot of vertex C makes the smallest angle
relative to the direction of c.m. near m /3. Hence its displacement is the largest and
its corresponding force is minimum, i.e. the minimum force should act on vertex C(N 1,
N 1). (1 point)
[Remarks: A rigorous calculation is given in Appendix 3.]
Marking Scheme T1
Page 9 of 15
C2
At = m /3, a small change in will change the potential energy by:

p (m ) =
p
|
=
m
1 2
33
3
[(
3) sin m cos m ]
3
2
2
3
= 4 ( 1) 2 (1 point)
- Eq. (41)
The displacement of C(m,n) point is
C(,) = [(2 + ) sin m sin ( + m )]

3
=
(2++1)3
2
(0.5 points)
C(,) = [3 sin m cos ( + m )]

3
=
(3+1)
2
(0.5 points)
For C(N-1,N-1), = ()2 + ()2 = (3 2)() . (1 point)

Hence
min =
p (m )
max
3(1) 2
4(32)
(1 point)
- Eq. (42)
and
min = tan1 [
C(1,1)
]+
C(1,1)
= tan1
3
3
+ =
5
6
(1 point)
- Eq. (43)
[Remarks: This min is not perpendicular to the C(N-1,N-1)A(0,0) direction because

of the constraints of the tunes, e.g. A(1,0), A(2,0), A(3,0), , are also the holding
points.]
Marking Scheme T1
Page 10 of 15
N/A
Appendix 1:
(a) Calculation of the exact p , k and E in Parts (C), (D) and for arbitrary N
Under a small perturbation, the potential energy change is
1 2 p
()2
p
|
2 2 =
()2
1 2
33 23
3
= (
cos E + sin E )
3
2
2
2
=
3(32)2 +9
12
2 ()2
- Eq. (44)
The kinetic energy of a triangle includes the translational energy of its center of mass and
the rotational energy around its center of mass. Hence the total kinetic energy of the N2
triangles is
k = , c.m.(,) + , r.c.(,)
- Eq. (45)
where
1 2
r.c.(,) = 2
12
( )2 = 2 ( )2
24
- Eq. (46)
and
2
2 c.m.(,)
c.m.(,) =
=
()2
2
dc.m.(,) 2
[(
dc.m.(,) 2
) +(
) ]
=E
Since
c.m.(,) = A(,) +
3
cos
(
3
6
+ )
3
= (2 + ) cos + cos
sin
2
6
and
c.m.(,) = A(,)
3
sin ( + )
3
6
- Eq. (47)
Marking Scheme T1
Page 11 of 15
= 3 cos
3
cos
6
2 sin
- Eq. (48)
Hence,
dc.m.(,)
1
3
= [(2 + ) sin sin
cos ]
d
2
6
dc.m.(,)
1
3
= [3 sin +
sin cos ]
d
6
2
We have
1
c.m.(,) = 2
2 ( )2
(42 + 42 + 4 + 2 + 2) sin2 E
+
23
3
( ) sin E cos E +
- Eq. (49)
and
=
c.m.(,) +
r.c.(,)
1
5
= [ (11 1)( 1) sin2 E + ] 2 2 ( )2
6
24
=[
(111)(1)
2(32)2 +6
] 2 2 ( )2
24
- Eq. (50)
With Eqs. (44) and (50), we have

3(3 2)2 + 9 2

1
12
E =
2 (11 1)( 1) 5
[
+ 24] 2 2
2(3 2)2 + 6
23(32)2 +9
= 2 12(111)(1)
[
(32)2 +3
+5]
(b) Center of mass movement of the whole system

According to Eq. (48), we have
c.m.(sys.) () =
, c.m.(,)
2
- Eq. (51)
Marking Scheme T1
Page 12 of 15
3
, [(2 + ) cos + cos
2
6 sin ]
2
=(
3 2
3
) cos
sin
2
6
and
c.m.(,) () =
, c.m.(,)
2
, [3 cos +
3
6 cos + 2 sin ]
2
32
sin
= (
) 3 cos
- Eq. (52)
Eq. (52) is the trajectory of the center of mass for the whole system, which is not a straight
line.
Appendix 2: Calculation of the moment of inertia of a triangular plate
An equilateral triangle with lateral length l can be divided into four small equilateral
triangles with lateral length l/2. For the central small triangle centered at c1, its moment of
inertia is
1 =
( )
2
- Eq. (53)
For the non-central small triangle centered at2 , 2 and 2 ,

2 = 1 +
- Eq. (54)
where = 3/6 is the distance between the centers of triangles 1 and 2. The second term
is from the parallel-axis theorem. The moment of inertia of the whole triangle is the sum of
the moment of inertia of the four sub-triangles:
N/A
Marking Scheme T1
Page 13 of 15
2 = 4
( ) +3
4 2
- Eq.(55)
Thus
1
= 12
- Eq. (56)
Marking Scheme T1
Page 14 of 15
Appendix 3: The minimum force corresponds to the maximum displacement of the

exerting point of this force.
Consider the position of vertices A, B, C of a triangle (m,n) :
A(,) = (2 + ) cos m
A(,) = 3 cos m
B(,) = (2 + + 1) cos m
B(,) = (3 cos m + sin m )
C(,) = [(2 + ) cos m + cos ( + m )]

3
C(,) = [3 cos m + sin ( 3 + m )]
- Eq. (57)
Taking derivatives on on the above coordinates we get

A(,) = (2 + ) sin m =
A(,) = 3 sin m () =
(2 + )3
B(,) = (2 + + 1) sin m =
(2 + + 1)3
B(,) = (3 sin m + cos m ) =
3 1
(2 + + 1)3
C(,) = [(2 + ) sin m sin ( + m )] =
3
2
C(,) = [3 sin m + cos ( 3 + m )] =
(3+1)
2
- Eq. (58)
For = ()2 + ()2, we have

A(,) = 32 + 32 + 3()
B(,) = 32 + 32 + 3 + 3 + 1()
C(,) = 32 + 32 + 3 + 3 + 3 + 1()
- Eq. (59)
N/A
Marking Scheme T1
Page 15 of 15
Thus we nd
C(,) > B(,) > A(,)
- Eq. (60)
Therefore, we should choose point C of the triangle (N 1, N 1) to obtain

max = (3 2)
so that the force is minimal.
- Eq. (61)
Marking Scheme T2
Page 1 of 4
(Full Marks: 20)
Part
A
Model Answer
The physical volume is
Marks
2
p = 3 (). (0.5 points).

The comoving number density is a constant, thus the physical number density is
( ) 3
()
(0
= ( ()0 ) . (0.5 points)

)
The kinetic energy for non-relativistic particles are negligible, thus the energy density is
m () = (), (0.5 points)
where is the mass of a particle.
Thus
( ) 3
m () = m (0 ) ( ()0 ) (0.5 points)

[Remarks: It is acceptable if the student just writes / and full points will be given.]
The Einsteins energy relation for a massless particle is

= . (0.5 points)
From de Brogiles relation:
1/P 1/a(t). (0.5 points)
[Remarks: No point if only a(t) is written because already given.]
Thus
1/a(t). (0.5 points)
Physical number density is 1/3 .
Energy density is .
Thus
( ) 4
r () = r (0 ) ( ()0 ) (0.5 points)

[Remarks: It is acceptable if the student just write / .]
Marking Scheme T2
Page 2 of 4
C
The photons in thermal equilibrium satisfy Boltzmann distribution

(())
()
B ()
, (1 point)
where 1/a(t).
Condition of being non-interacting implies that there is no energy transfer. Thus the energy
distribution must be stable.
To be explicit, for two different comoving wavelengths,
(1 ())
(2 ())
= [2 ()1 ()]/[B ()] = const.
[Remarks: All the above steps can be replaced by the intuition of , if the students
realize it, the above 1 point can be given.]
Thus
() 1/, i.e. = 1. (1 point)
D
The 1st law of thermodynamics is
dX = X dp . (1 point)
Here no entropy term appears, because =const. No chemical potential appears, because
of particle number conservation.
Here p = 3 .
dp = 32 d. (1 point)
X = X p . (0.5 points)
dX = d(X p ) = 3 dX + 3X 2 d. (0.5 points)
Thus
d
dX + 3 ( )(X + X ) = 0. (0.5 points)
X + 3 ()(X + X ) = 0. (0.5 points)

[Remarks: 0.5 point for relating variation and time derivative no matter in which step it is
being used.]
Marking Scheme T2
Page 3 of 4
E
With lens area , we only receive part of the starlight. The area ratio is
/(42 (0 ) 2 ). (1 point)
The wavelength of each photon emitted from the star gets stretched. Thus energy per
photon is lowered, contributing a ratio
(e )/(0 ). (1 point)
The separation among the photons also increases due to cosmic expansion, contributing a
ratio
(e )/(0 ). (1 point)
As a result, the power that the telescope receives is
2 ( )
r = 44 ( e) 2 e . (1 point)
0
The kinetic energy and gravitational energy of the shell adds up to a constant:
1
(p )2
, (2 points)
where
=
4
3
p 3 2, (1 point)
(Note: energy conservation without evolving pressure requires the assumption of nonrelativistic matter.)
p = (), (1 point)
[Remarks: The point is given because the student understand that the shell is not pulled
gravitationally from the outside, because the force due to the mass outside cancels.]
Thus
2
2
= 2
8
3 2
2 . (1 point)
Alternative Solution:
For the gravitational force due to the mass inside:
p =
where is mass of shell.
p2
= 3 2 p , (2 points)
Marking Scheme T2
Page 4 of 4
p = (), (1 point)
[Remarks: The point is given because the student understand that the shell is not pulled
gravitationally from the outside, because the force due to the mass outside cancels.]
and = (0 )3 (0 )/3 ().

Thus
4
= 3 2 (0 )3 (0 )2 . (1 point)
Integrate the above equation. One gets
1
= 2 2
4
3 2
(0 )1 = 2 2
4
3 2
2 , (1 point)
where is an integration constant.

G
(b) decelerating. This is because gravity is attractive for the matter that we are considering
here. As a result, d()/d is a decreasing function of .
Appendix: Notes about the physics behind this set of problems:

To reduce students reading load, we have not mentioned in the problems, that those
problems set up the framework of researches in modern cosmology:
A theory of gravity (especially Einsteins general relativity) contains two aspects: Gravity
tells matter how to move (kinematics of matter motion in a gravitational field); and matter
determines the gravitational field (dynamics of the gravitational field). Parts (A)-(E) are
about kinematics and part (F) is about dynamics in this sense. The two key equations in
cosmology are derived in part (D) (this is known as the continuity equation, containing
parts (A) and (B) as special cases) and (F), upon which the whole theory of modern
cosmology is built.
The equation derived in part (F) is known as the Friedmann equation, which is
2
conventionally written as () 2 =
8
3
. This equation governs the dynamics of
cosmic expansion and actually not only applies for non-relativistic matter but also for
general matter components (which needs general relativity to derive). The constant is
related to the curvature of 3-dimensional space, which is observed to be vanishingly small.
Part (C) indicates that the universe was hotter at earlier ages. The hot universe in local
thermal equilibrium determines the whole thermal history of our universe, which answers
questions such as where the light elements come from, and when the universe becomes
transparent for light. Part (E) defines the luminosity distance, which relates the telescope
observations to the cosmic reality.
1
N/A
Marking Scheme T3
Page 1 of 7
Part
Model Answer
(Full mark = 20)
Marks
The angular momentum should be

A1
2
(1 point for the definition of angular momentum)
Here
is the unit vector pointing from the center of the ring to the mass point on the ring
and
is the unit vector parallel to the direction of the linear velocity at the mass point.
We know that v = w r , so finally we can get

, with
(1 point for the correct answer: 0.5 points for the
magnitude and 0.5 points for the direction)
For a current loop, the magnetic moment is defined as

A2
2
The current can be expressed as
I = -ef = -e
w
(1 point for the current expression)
2p
Finally
(1 point for the answer)
A3
For a current loop, under a uniform magnetic field the total torque should be
(0.5 point for the torque definition)
The work done by the magnetic field on the torque should be
W d
2
d
2
2 M B sin d
M B
(1.5 points for the work on the torque)
Marking Scheme T3
Page 2 of 7
(0.5 point for the answer)
A4
We assume that the magnetic field is along z direction such that
, then in general
The magnetic torque of an electron should be

Mz
e
Sz
2me
(0.5 points for the electron torque)
Thus
U M B
e
Sz B
2me
(0.5 points for the answer)

Sz B
1
B B
2
Here B
e
is the Bohr magneton.
2me
B 5.788 105 eV T 1
A5
Thus for spin parallel state
, we have
= 5.788 105 eV (0.5 points)
For spin anti-parallel state

= 5.788 105 eV (0.5 points)
, we have
Marking Scheme T3
Page 3 of 7
B1
In the superconductivity state, electrons forming a Cooper pair have opposite spins, thus
the external magnetic field cannot have any effect on the cooper pair. Thus the energy of
the Cooper pair does not change.
ES =
B2
p12 p22
+
- 2D ( 1 point for the answer)
2m 2m
In the normal state, the two electrons will align their magnetic moments parallel to the
external magnetic field. Therefore we have
EN
p12 p22 2B S1x Bx 2B S2 x Bx
2m 2m
Here the potential energy of electrons should be twice as the classical estimation
according to quantum mechanics. Because
can make the magnetic
moment aligned along x direction, eventually we have

p12 p22
2B Bx
2m 2m
(1 point)
p12 p22 e
Bx
2m 2m me
EN
B3
EN < ES 2Bx m B > 2D Bx >

Thus BP
mB
2me
(1 points)
e
Note: The above simple consideration for the upper critical field BP over estimates its
value. The strict derivation considering the Pauli magnetization and superconductivity
m
2 e .
condensation energy will give BP
e
2B
C1
Method 1:
Marking Scheme T3
Page 4 of 7
1
2 4 x2
Substituting x
e into the F y , we have

( )
e 2 Bz2 x 2 x2
x2
x2
x2
2 2 x2
e
x
e
e dx

4me
me
2
2 2
2
2 x2 e2 Bz2
2 2 x2
dx
x e

2me
me
me
2 2
e 2 Bz2
1
2me me
me 4
2
e 2 Bz2

4me 4 me
( )
(1.5 points for the correct expression of F y as a function of l )
( )
We can treat F y as a function of l . Thus we have
4me
2
e2 Bz2
e2 Bz2
dF
, and
.
4 me
d 4me 4me 2
F (y ) takes the minimum value when

2
4me
d2 F
dF
0 , thus
0 and
d 2
d
( )
e2 Bz2
0 ( 0.5 point for the way to minimize F y )
4me 2
Finally, we can get

(1 point for the correct answer)
We can check that
d2F
> 0 when
dl 2
minimum value when

Method 2:
( )
, which guarantees that F y takes the
Marking Scheme T3
Page 5 of 7
d 2 e2 Bz2 x 2
F
dx
2
4
m
d
x
m
e
e
2 2 2
2
2
e Bz x
d
(1 point)

dx
2
4me dx
me
H dx
In this way, for normalized wave function the F is simply the energy expectation
, the eigenvalue of the Hamiltonian
d 2 e2 Bz2 2
H
x
4me dx 2
me
2
The first two terms correspond to the quantum simple harmonic oscillator Hamiltonian.
Thus the ground state energy should be
Here
eBz
and ground state wave function becomes
me
1
2m 4
e e

1
4
2eBz
me
eBz
x2
(1 point)
x
Therefore, we have
. (1 point)
From Part (C1) we know Fmin
eBz
. At the critical value for Bz , it makes the
2me
energy difference zero. It means that the critical value Bz satisfies

C2
eBz
0 . (1 point for this equation)
2me
Consequently,
Marking Scheme T3
Page 6 of 7
. (1 point for the correct answer)
D1
EI
p12 p22
2 S B 2 S B
2 B 1 z z B 2 z z
2m 2m
Here
EI
D2
p12 p22
2 S B 2 S B
2 B 1 z z B 2 z z
2m 2m
p12 p22
2 B 2 B
2 B z B z
2m 2m
2
2
(1 point)
2
2
p1
p2
2 2 B Bz
2m 2m
p12 p22
e
2
Bz
2m 2m
me
In the normal state, the electrons will align the magnetic moment parallel to the total
magnetic field, thus
2
1
2
|| = 2
+ 2
+
1 2B 2
2
2B 1
For electron 1,
For electron 2,
Therefore,
and
can make the their magnetic moments parallel to the total magnetic field respectively.
(1 point for the correct expression of spins: 0.5 points for each respectively)
Finally
2
1
2
1
2
|| = 2
+ 2
2B 2 + 2 = 2
+ 2
2 + 2 (1 point for the answer)
D3
|| < Ising 2B 2 + 2 > 2 + 2B >
2 + 2B
(1 points)
B
Marking Scheme T3
Page 7 of 7
Another correct expression is:

APhO Theoretical Problems and Solutions 2010-2016

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

APhO Theoretical Problems and Solutions 2010-2016

Uploaded by

Copyright:

Available Formats

LE HONG PHONG HIGH SCHOOL FOR THE GIFTED

HO CHI MINH CITY

FOR INTERNAL USE ONLY

Part A. Inelastic Scattering and Compositeness of Particles

(i) Find an expression for in terms of , , 1 , 2 , and 2 .

of contains those points of 2 allowed?

Part B. Waves on a String

Part C. An Expanding Universe

where 1 Mpc = 3.0857 1019 km = 3.2616 106 light-year.

Part A. Magnetic Fields on the Axis of the Coil

Part B. The Upper Limit of Current

Part C. The Rate of Temperature Rise

and = 8.98 103 kg m3 .

Part D. A Pulsed-Field Magnet

(g) Find expressions for the inductance and resistance .

in which and are positive constants and 0 is given by

Note that, if () is expressed as a function of a new variable ( + 0 /), then ()

2. sin( ) = sin cos cos sin

3. permeability of free space 0 = 4 107 T m/A

------------------------------------------------------ END ------------------------------------------------

Part A. An Electron Bubble in Liquid Helium

uncertainty relations /2 , where is the Planck constant divided by 2

(a) Find a relation between He , e , and .

Part B. Single Gas Bubble in Liquid Collapsing and Radiation

Radial motion of the bubble-liquid interface

Collapsing of the gas bubble

pressure 0 , the equilibrium radius of the bubble is 0 = 5.00 m.

Let i (i ) be the gas pressure of the bubble when = i . If we introduce the

Give an expression and then estimate the value of .

Sonoluminescence of the collapsing bubble

Consider the bubble to be a surface black-body radiator of constant emissivity so that

(k) Find the power supplied to the bubble as a function of .

2. Electron mass = 9.11 1031 kg

3. Planck constant = 2 = 2 1.055 1034 J s

4. Stefan-Boltzmanns constant SB = 5.67 108 W m2 K 4

(Document Released: 14:30, 4/24)

Theoretical Question 1: Particles and Waves

Part A. Inelastic scattering and compositeness of particles

(i) By definition, we have = i f , or equivalently,

have i = f , or = 0. Thus, we obtain from Eq. (a-2) the following equality

In the 2 -2 plane, this represents a circle centered at (1 /( + ), 0) with

(Document Released: 14:30, 4/24)

perpendicular to the spring is m = 14 m 2 = 14 0 2 (1 + )2 . The law of conservation

(Document Released: 14:30, 4/24)

Note that, since 2

is nonnegative and the last equality follows from Eq.(a-8).

(Document Released: 14:30, 4/24)

Moreover, the maximum value max of = sin2 is seen to be given by

Note that, from Eq. (a-11), it follows

(Document Released: 14:30, 4/24)

Part B. Waves on a string

[Another way to get the period ]:

(Document Released: 14:30, 4/24)

[Another solution of (c)]:

(Document Released: 14:30, 4/24)

where is the constant tension ( ) on the string and is the transverse

where use has been made of = /.

[Another solution of (d)]:

Here (, ) is the displacement of the elastic string.

[Yet another solution of (d)]:

We consider a special moment when

(Document Released: 14:30, 4/24)

Part C. The expanding universe