Problem T1. Stabilizing unstable states (11 iv.

(1 pt) The cyclist is able to balance himself by turning

points) the handlebar so that the line connecting the wheels will move
Part A. Stabilization via feedback (3.5 points) to the desire direction. For that line to move, the bike must
i. (1.5 pts) The moment of inertia of the rod is move forward to a distance which is of the order of inter-wheel
2
I = ml3 . The torque is mg φl
2 . (0.4 pts) separation. So we can require vm τ ≈ d, where τ is bike’s char-
so that the Newton’s 2nd law is written as acteristic falling time. (0.5 pts)
φl Note that with this equation we neglect the cyclists’ reaction
I φ̈ = mg
2 time (which makes balancing more difficult) but on the other
3g hand the line connecting the wheels moves slightly already at
φ̈ = φ. (1)
2l a twice smaller forward-displacement of the bike (which makes
(0.4 pts) balancing easier). Anyway, we are only making an estimate, so
If we take φ = Ae τ + Be− τ , then
t t
a mistake by a factor of 2 is perfectly OK.
φ̈ = τA2 e τ + τB2 e− τ = τφ2 .
t t
(0.3 pts) Then we get √
Substituting this into the equation of motion (1) we get 2L
d = vm τ = vm
φ 3g 3g
2
= φ
τ 2l (0.2 pts)
√
2l √
τ= . 3g
3g vm = d = 2.7 m/s
2L
(0.4 pts) (0.3 pts)
t
− τt
This means that φ = Ae + Be τ is the solution for the equa- Part B. Tightrope walker (3.5 points)
tion of motion. i. (1 pt) From the conservation of angular momentum
ii. (0.5 pts) The boy has to react before the rod falls over the
dα1 dα2
angle π2 . Boy notices that the rod is falling, and tries to react. m(1.4H)2 + mH 2 = Const.
dt dt
If the rod falls faster than his reaction time, he cannot keep it
(0.3 pts)
in balance. In the expression φ = Ae τ + Be− τ , the dominat-
t t

Partial credit 0.2 pts if the conservation is mentioned without

ing term is the first one (the second one decays in time), so we
t writing equation. This process is instantaneous, i.e. dα dt and
1
can put φ = Ae τ , where A is the angle at t = 0. Hence, the dα
1
are very large, much large than that constant at the right-
falling time t = τ ln(π/2A) depends on the initial angle A, but dt
hand-side (which is defined by the initial falling speed), hence
logarithmic dependence is very slow — the logarithm remains
we can put Const= 0.
always of the order of unity. So we can estimate the falling time
(0.2 pts)
just as the characteristic time of the rod. This means that
√ This simplifies into
2 lr
τr ≈ 1.96∆α1 = −∆α2 (2)
3g

(0.3 pts) (0.1 pts)

We also have
3g
lr = τr2 = 0.59 m
2 β = α1 − α2 = (α0 + ∆α1 ) − (α0 + ∆α2 ) = ∆α1 − ∆α2 (3)
(0.2 pts)
(0.2 pts)
iii. (0.5 pts) The bird won’t be able to rebalance itself when
Solving the equations (1) and (2) we get
it has fallen over π2 . Similarly to the previous question, we can
say that the bird’s reaction time must be equal to the charac- β
α1 = α0 +
teristic time τb . (0.3 pts) 2.96
Then we get (0.1 pts)
√ 1.96
2 lb α2 = α0 − β
τb ≈ = 0.065 s 2.96
3g (0.1 pts)
(0.2 pts) ii. (0.5 pts) In order to be able to straighten himself, the
walker’s centre of mass has to move leftwards, by a negative
— page 1 of 7 —
angle. (0.1 pts) 1.96 1.96
α2 = α0 − β0 ≈ β0
2.96 2.96
The torque can be expressed as
By changing the upper body’s angle by ∆α1 , the lower
body’s angle will change by ∆α2 = −1.96∆α1 . The centre M = 1.4mgHα1 + mgHα2 =
of mass will then move by
1.4 1.96 0.56
β0 mgH − β0 mgH = − β0 mgH
1.4H∆α1 + H∆α2 = 1.4H∆α1 − 1.96H∆α1 = −0.56H∆α1 2.96 2.96 2.96
(0.3 pts)
(0.3 pts) Writing out ∆L = M Tb we get

α0 0.56
Because the centre of mass will have to move by a negative −5.92mH 2 =− β0 mgHTb
τ 2.96
angle, ∆α1 needs to be positive, which means that the walker √ √
has to bow clockwise. (0.1 pts) α0 H α0 H 2.4 g α0 H
Tb = 31.29 = 31.29 = 28.18
β0 τ g β0 g 2.96 H β0 g
iii. (1 pt) We can write the equation of motion (0.1 pts)
Part C. Kapitza’s pendulum (4 points)
2.96α¨1 H = 2.4gα1
Throughout the entire problem, we use the system of refer-
this dif- ence of the suspension point.
Similarly to the question i. in part A, the solution for√
ferential equation is α1 (t) = Ae τ + Be− τ , where τ = 2.96
t t H
2.4 g .
i. (1.5 pts) During these periods of time when the suspension
(0.2 pts) point accelerates upwards (and force of inertia is downwards),
Because the time it takes to get to the vertical position is the equation of motion of the pendulum can be written as
t
infinite, the component Ae τ needs to be 0, meaning that d2 φ a0
α1 (t) = Be− τ .
t
(0.3 pts) = φ,
dt2 l
By taking time derivative, we obtain
where a0 = 2v0 /T is the frame’s acceleration. (0.4 pts)
α̇1 = − τ1 Be− τ .
t
(0.3 pts)
Incomplete attempts at writing Newton second law will be par-
For the instance when the boy straightened himself, t = 0, the
tially credited (0.2 pts).
equations take form α1 = B and α̇1 = − B τ . So, α̇1 = − τ , The relative change of φ is assumed to be small, so we can
α1

which can be rewritten as

√ approximate φ ≈ φ0 to obtain
α̇1 1 2.4 g
=− =− d2 φ 2v0
α1 τ 2.96 H 2
= φ0 .
dt Tl
(0.2 pts)
(0.2 pts)
iv. (1 pt) After the walker has straightened himself, the angle
During the rest of the time, the same equation can be used if
which he is at is still α0 , because during stage where he is bow-
a0 is changed to −a0 . (0.2 pts)
ing, the torque is much larger than when he is straightened,
Therefore, the graph consists of parabolic segments, as depic-
meaning that the change in angular speed is much larger than
ted in the Figure. (0.4 pts)
the change in the angle. (0.1 pts)
The amplitude is found as
As found in the previous subquestion, the speed before and
after the bowing are ατ0 and − ατ0 respectively. Then the change 1 v0 T
∆φ = φ0 .
4 l
in the angular momentum is
α0 (0.3 pts)
∆L = −5.92mH 2
τ
(0.3 pts)
Because during the falling stage the change in angle is minus-
cule, we can express the change in angular momentum as
∆L = M Tb , where M is the torque during bowing stage.
ii. (1.5 pts)
(0.2 pts)
The average torque ⟨M ⟩ = ⟨mla(t)φ(t)⟩ . (0.3 pts)
During the bowing stage, the angles of the body segments are
Let us note that ⟨a(t) ⟨φ⟩⟩ = ⟨a(t)⟩ ⟨φ⟩ = 0. (0.3 pts)
β0 β0
α1 = α0 + ≈ Therefore we can rewrite the average torque as
2.96 2.96
— page 2 of 7 —
 2v0 (0.2 pts)
⟨M ⟩ = ⟨mla(t)[φ(t) − ⟨φ⟩]⟩ = −ml ⟨|φ(t) − ⟨φ⟩ |⟩
T

(0.3 pts; if wrong sign 0.2 pts) iii. (1 pt) Gravity field does not affect the expression for the
It is easy to see that the average of |φ − ⟨φ⟩ | over the entire average torque of the force of inertia. So, we can use the result
period equals to the average over the time interval 0 < t < τ . of the previous question. However, it gives rise to an additional
Straightforward integration yields contribution to the average torque, equal to glmφ0 . (0.4 pts)
∫ ( ) Therefore, the equation of motion can be written as
2 T /2 4t2 2 1 v0 T ( )
⟨|φ − ⟨φ⟩ |⟩ = ∆φ 1 − 2 dt = ∆φ = φ0 . d2 φ0 1
T 0 T 3 6 l l2 2 = gl − v02 T 2 φ0 .
dt 3
(0.4 pts)
(0.4 pts)
Upon substituting this result into the previous expression we
The stability is ensured if the factor at the right-hand-side is
obtain
negative, i.e. if 3gl < v02 . (0.2 pts)
1
⟨M ⟩ = − mv0 φ0 .
2
3

— page 3 of 7 —
Problem T2. Gravitational waves (10 points) ii. (0.8 pts) The dimensions for A need to be
Part A. Dipole radiation (2.4 points) 1 kgm2 1 1
For the sake of convenience, let us denote k = 1 W = = 3 2
4πϵ0 kg2 m4 s3 kg2 m4 s m kg
i. (1.4 pts) The total dipole moment can be expressed as
(0.2 pts)
d⃗ = −⃗r1 e + ⃗r2 Ze, (0.2 pts)
A can only consist of the gravitational constant G, speed of
where ⃗r1 and ⃗r2 are the position vectors of the electron and
¨ ⃗ ⃗
light c and angular velocity ω. (0.2 pts)
nucleus respectively. Then d⃗ = −r⃗¨1 e + r⃗¨2 Ze = − m
F
e − mF1 Ze,
We can write a system of equations for every unit, m, s and kg.
(0.1 pts)
If we take the power of G, c and ω to be a, b and c respectively,
but because the mass of the nucleus m1 is much larger than m,
we can write out for metres
we can neglect the second term. Then
  3a + b = −2,
   F⃗  kZe3
 ⃗¨  
d  = − e =
 m  mr2 (0.1 pts)
(0.3 pts) for seconds
−2a − b − c = −3,
¨2 ( )2
d⃗ kZe 3
1 (0.1 pts)
Ped = =
4πϵ0 c3 mr2 6πϵ0 c3 and finally for kilograms
(0.2 pts)
We can express λ as λ = c
= 2πrc −a = −1.
f v . √
(0.2 pts)
mv 2 2
We get v from the relation r = k Ze
r 2 , then v = e
kZ
mr .
(0.1 pts)
(0.3 pts) Solving the equations we get a = 1, b = −5 and c = 6, which
Finally √ means that
2πrc 2πrc mr ω6
λ= = A=G 5
v e kZ c
(0.1 pts)
(0.1 pts)
⃗ ∑
ii. (1 pt) We know that dg = i ⃗ri mi . Let us recall that the
iii. (0.8 pts) The solution for this subquestion is the same as
distance to centre of mass is ∑
⃗ri mi for the previous one. The units for K are
⃗rcm = ∑i .
i mi W kg
= 3
Then (0.3 pts) m2 s
∑ ∑ (0.2 pts)
d⃗g = ⃗ri mi = ⃗rcm mi
i i
K can only depend on G, c and ω, (0.2 pts)
(0.2 pts) if we take their powers to be a, b and c respectively, we can
⃗˙ ∑
(Equivalently one can notice that dg = vi mi the net mo- write out for metres
i⃗
mentum.) ⃗rcm is constant because there aren’t any external 3a + b = 0
forces acting on the system. (0.3 pts) (0.1 pts)
¨
⃗
This means that dg = 0 and Pgd = 0. (0.2 pts) for seconds
−2a − b − c = −3
Part B. Quadrupole radiation (7.6 points)
(0.1 pts)
i. (1 pt) The force acting on one of the stars is
M2 M v2 and finally for kilograms
F =G 2 =
4R R −a = 1
(0.6 pts)
From here we can express the star’s speed v. We can express (0.1 pts)
v
the star’s angular speed from the relation ω = R. (0.2 pts) Solving the equations we get a = −1, b = 3 and c = 2, which
√ means that
v 1 GM c3 ω 2
ω= = K=
R 2 R3 G
(0.2 pts) (0.1 pts)

— page 4 of 7 —
 P
√
qg
iv. (1 pt) The flux at distance L is S = 4πL 2, (0.4 pts) 1 GM
2 ω= ;
on the other hand we can express flux as S = Kh0 . This means 2 Rs3
that
Pqg AM 2 R4 (0.2 pts)
Kh20 = =
4πL2 4πL2 substituting Rs we obtain
(0.4 pts) √
1 c6 c3
and finally ω= GM = √
2 8G3 M 3 2 2GM
√ √
AM 2 R4 Gω 6 G M 2 R4 Gω 2 M R2 G2 M 2 (0.2 pts)
h0 = = = √ = √
K4πL2 c5 c3 ω 2 4πL2 2 πc4 L 8 πc4 LR The mass of a black hole is

(0.2 pts) c3
M= √ = 1.45 × 1032 kg = 72 M⊙
2 2ωG
v. (1 pt) At distance Rs from the black hole, it takes all (0.2 pts)
the energy mc2 for an object to overcome the potential energy In reality the masses of the black holes were 36 M⊙ and 29 M⊙ ,
−G MRms
. This means that where M⊙ is the mass of Sun.
vii. (1.5 pts) As mentioned previously the maximal strain is
2 Mm
mc = G . when the black hole’s orbital radius is Rs , reading from the
Rs
graph we get that the strain is h0 ≈ 0.9 × 10−21 . (0.4 pts)
(0.7 pts) Then, using the result of question iv,
2
(If non-relativistic energy mc /2 is used, deduct 0.3 pts.) From
G2 M 2
here we can express Rs h0 = √ 4 ;
8 πc LRs
GM
Rs = 2
c (0.4 pts)
Substituting Rs we obtain
(0.1 pts)
This answer is 2 times smaller than the correct one, which G2 M 2 c2 GM
h0 = √ 4 = √ 2
means that 2GM 8 πc L 2GM 16 πc L
Rs =
c2 (0.3 pts)
(0.2 pts) The distance to the black hole is

GM
vi. (1.5 pts) We can estimate the mass of the black holes by L= √ ;
16 πc2 h0
finding the orbital frequency and then using the expression we
(0.2 pts)
found in the first subquestion (Kepler’s III law). (0.3 pts)
numerically
The strain is maximal when the orbital radius for the black
L = 4.21 × 1024 m.
holes is Rs . (0.3 pts)
Reading from the graph, we get that the orbital period is (0.2 pts)
T ≈ 0.006 s (0.3 pts) This can be also expressed in megaparsecs, L = 136 Mpc. In
and frequency ω = 2π
T = 1000 rad/s. Then reality, the distance was L = 410 Mpc ± 170 Mpc.

— page 5 of 7 —
Problem T3. Magnetars (11 points) iii. (1 pt) During the collapse there is no torque on the
i. (1.5 pts) If we change the frog’s height by ∆h, the change star, this means that the angular momentum remains constant.
in potential energy needs to be smaller than the change in mag- Thus
netic energy. (0.2 pts) 2 2
M Rs2 ωs = M Rn2 ωn
5 5
Note that for every point in frog, the change in magnetic energy
is the same, thus we can express it as (0.6 pts)
1/3
Recall that Rs is inversely proportional to ρs . (0.2 pts)
∆(B 2 )χ B 2 χ∆h
∆E = −V =V 0 Now we can express ωn
2µ0 2h0 µ0 ( ) 23
R2 ρn
(0.6 pts) ωn = ωs 2s = ωs
The change in potential energy is Rn ρs

∆Π = V ρg∆h (0.2 pts)

iv. (1.5 pts) After time t, the neutron star has turned by an
(0.3 pts) angle β = ωn t. (0.2 pts)
Then
∆E + ∆Π < 0
The magnetic fields pass any radial line from the centre of the
B 2 χ∆h β
V 0 + V ρg∆h < 0 neutron star on average N = 2π = ω2π
nt
times. (0.3 pts)
2h0 µ0
The total magnetic flux entering the outer shell remains con-
(0.2 pts)
stant, and thus is always equal to Φ = 2πR0 B0 , (0.3 pts)
This means that √ where R0 is the radius of the outer shell. This means that the
2h0 µ0 ρg
B0 > − flux through any radial line is ΦN . (0.4 pts)
χ
Then
(0.1 pts) BR0 = 2πR0 B0 N = R0 B0 ωn t
and numerically
(0.2 pts)
B0 = 5.32 T.
And finally
(0.1 pts) B = B0 ωn t

(0.1 pts)
ii. (1 pt) Let us observe a piece of the star with a volume V0
before the collapse and volume V1 after the collapse. The mass v. (1 pt) We can find the gravitational energy by integrating:
before and after are same. This means that we imagine removing the material layers of thickness dx one by
one, starting from the outermost one. The potential energy for
V0 ρs = V1 ρn
a hollow sphere with a thickness dx in the gravity field of the
(0.1 pts) matter inside it is
The radius of the star scales as V 1/3 , (0.1 pts) (4πx2 dxρn ) 34 πx3 ρn 16π 2 2 4
and the cross-sectional area as V 2/3 . (0.1 pts) dΠ = −G =− Gρn x dx
x 3
The total magnetic field through the volume is also the same
(0.2 pts)
before and after the collapse:
Integrating from x = 0 to x = Rn we get
2 2
Bs V0 = Bn V1
3 3
16π 2 2 5 3 GMn2
Π=− Gρn Rn = −
(0.4 pts) 15 5 Rn
Now we can express Bn (0.3 pts)
( ) 23 ( ) 23 A partial credit of 0.2 pts is given is if integration is substituted
V0 ρn
Bn = Bs = Bs by a simplifying product.
V1 ρs
This potential energy is equal to the magnetic energy
(0.2 pts)
and numerically 4 3 2 1 3 GMn2
Π= πR Bn =
Bn = 5.0 × 105 T 3 2µ0 5 Rn

(0.1 pts) (0.3 pts)

— page 6 of 7 —
Solving for Bn we get √ tainty principle. The uncertainty of the momentum is
M µ0 G
Bn = 3 2
R 10π ∆p = 2me v
(0.1 pts)
and numerically (0.3 pts)
Bn = 1.18 × 10 14
T and uncertainty of the coordinate

(0.1 pts) ∆x = 2R1

vi. (1 pt) The electron orbit will get distorted when the (0.2 pts)
Lorentz force becomes in the same order of magnitude as Cou- So we have
lomb force. (0.1 pts) 4me vR1 ≈ ~
The Coulomb force is
(0.2 pts)
1 e2 ~
F1 = 2 Substituting me v = 4R1 to the first equation we get
4πϵ0 RH
(0.2 pts) ~
= 4eBn
On the other hand, R12
me v 2
F1 = (0.1 pts)
RH
(0.2 pts) Then √
~
We can express the velocity of electron R1 =
√ 4eBn
1 (0.1 pts)
v=e
4πϵ0 RH me
The length of the cylinder will still remain in the order of mag-
(0.1 pts)
nitude of RH because the Lorentz force doesn’t act on the
Then the Lorentz force is
electron in that axis (parallel to the magnetic field). (0.5 pts)
F2 ≈ evB Then the ratio of the length and diameter is approximately
√
(0.2 pts) RH eBn
κ= = 2RH
Upon substituting v we obtain R1 ~
√
1
F 2 = e2 (0.1 pts)
4πϵ0 RH me
and numerically
From the condition F1 ≈ F2 we can express the magnetic field
κ = 39 ≈ 40 .
strength √
me (0.1 pts)
B= 3
4πϵ0 RH Note that if we were to make calculations for magnetars with
(0.1 pts) B = 1 × 1011 T, the orbital electrons would be ultrarelativistic.
and numerically
B = 2.56 × 105 T
(0.1 pts)

vii. (2 pts) Perpendicularly to the magnetic field, the Lorentz

force is much larger than the Coulomb force since the magnetic
field Bn is much larger than the magnetic field found in the
previous question. This means that in the perpendicular plane,
the electrons move along a circular cyclotron orbit. (0.2 pts)
Then we can write
me v 2
= evBn ,
R1
(0.2 pts)
where R1 = d/2 is the orbit’s radius. Now we apply the uncer-

— page 7 of 7 —