consider the system of the ship + the burnt • Relating energy and momentum with 4‐ pressed as

interstellar gas. momentum invariants (or equivalent, half

1. PHOTON ROCĸET (5 points) — Solution by of the points each for rocket and photons) dp d dv dm
The initial rest mass of the system is M + = (γv) m + γv
Taavet Kalda, grading schemes by Oskar Vallha­ 1 pts. d t d v d t dt
( M − m f ) = 2 M − m f = M i . This corresponds µ 2 ¶
gen and Konstantin Dukach. v d v
to an energy of E i = M i c2 and momentum • Solving the equation system in terms of en‐ = m γ3 2 + γ − γ vµ
i) (1 point) At non‐relativistic speeds, we can ergies and momentum 0. 5 pts. c dt
p i = 0 due to the gas and the ship initially be‐
apply classical momentum and energy con‐ • Solving for v f 0.5 pts. 3 dv d p ph
ing at rest. = mγ − γ vµ =
servation to find the acceleration in terms of dt dt
In the final configuration, we have the Non‐relativistic solution: s
the antimatter burning rate µ. In a time in‐ 1 − v/ c
space ship moving with a speed of v with • Realising that the problem can be solved = (γ + 1)µ c .
terval ∆ t, a mass of ∆m = µ∆ t antimatter an‐ 1 + v/ c
energy E f and momentum p f . We also by considering conservation of momentum
nihilates with an equal mass of matter. The
have a photon cloud with energy E p and and energy between initial and final state
resulting photons have an energy equal to We now change variable from time to mass
momentum p p , moving opposite to the dir‐ only 0.25 pts.
the annihilated rest energy ∆E = 2∆mc2 . For by noting that
ection of the space ship. Then the energy • Correct conservation laws for momentum
maximal acceleration, the photons have to
and momentum conservation simply read as and energy (half of the points for each) dv dv d m dv
be all emitted in the same direction (can be = =− µ,
E f + E p = E i and p i = 0 = p f + p p . We also 0.25 pts. d t dm d t dm
achieved, for example, using mirrors). The
have the 4‐momentum invariant for both the • Relating energy and momentum to v f
resulting photon cloud will then have a mo‐
space ship and the photon gas. They read 0.25 pts. yielding
mentum of ∆ p = ∆E / c. From the conser‐
E 2f − p2f c2 = m2f c4 and E 2p − p2p c2 = 0. Solving • Final answer (if commenting on unphys‐ s
vation of momentum, the space ship must 1 − v/ c
the 4 equations (with 4 unknowns), we get ical results, otherwise no points) 0.25 pts. 3 dv
get a momentum boost in the opposite dir‐ − mγ − γv = (γ + 1) c .
dm 1 + v/ c
ection, equal to ∆ p = M ∆v = M g∆ t. Com‐ M 2i − m2f M 2i + m2f Solution 2 by Oskar Vallhagen
bining everything, we get µ = M g/(2 c) = pf = c, Ef = c. An alternative way to solve the problem This equation can be solved by separation of
2M i 2M i
1.635 × 10−3 kg/s. is to find an expression for the reactive force variables according to
Grading: We could find the velocity by solving for v f felt by the rocket, and use it to derive an equa‐ Z Z
in the expression for the relativistic energy mf dm vf γ3
tion for the resulting velocity change. Dur‐ = − q dv,
• Introducing time ∆ t and expressing ∆m = E f = m f γ f c2 with γ f = (1 − v2f / c2 )−0.5 . A faster M m 0 1−v/ c
µδ t 0.2 pts.
ing a time d t, the energy released as photons γv + (γ + 1) c 1+v/ c
way, however, would be to use the expres‐ from the burnt rocket antimatter is γµd tc2
• Photon energy ∆E = 2∆mc2 0.2 pts.
sion for final momentum, p f = m f γ f v, and (including the kinetic energy of this mass) which can be simplified to
• Photon momentum ∆ p = ∆E / c 0.2 pts.
notice that and the energy released from the ordinary Z
• Rocket momentum ∆ p = M g∆ t (half point ³ mf ´ v f /c dx
M 2i − m2f matter needed for the annihilation is µd tc2 . ln = − p
if only force F = M g is stated) 0.2 pts. pf 2
vf = c = c As the photon momentum scales as the in‐ M 0 (1 − x)( 1 − x2 + x + 1)
• Final answer 0.2 pts. Ef M 2i + m2f
verse of the wavelength,
p pwhich is red‐shifted = I (v f / c) − I (0),
• Missing factor 2 accounting for both anti‐
(2 M − m f )2 − m2f by a factor of 1 + v/ c/ 1 − v/ c in the rest
matter and matter fom interstellar plasma = c frame compared to the rocket frame, the cor‐ with x = v/ c and I is the indefinite integral.
−0.2 pts. (2 M − m f )2 + m2f
responding change in photon momentum in To evaluate this integral, we first makep the
• Finding photon momentum by reasoning 180
= c ≈ 0.9945 c. the rest frame is substitution x = sin y. This gives 1 − x2 =
that the annihilated mass can be regarded 181 cos y and d x = cos yd y, so that the (indefinite)
as expelled with the speed of light −0.2 pts
integral becomes
ii) (3 points) The final speed is easiest to Grading:
Z
find by applying energy and momentum (4‐ • Realising that the problem can be solved s cos yd y
1 − v/ c I = − .
momentum) conservation in the initial and by considering conservation of momentum d p ph = (γ + 1)µd tc . (1 − sin y )(cos y + sin y + 1)
final configurations. If the final rest mass and energy between initial and final state 1 + v/ c
We now make the further substitution z =
of the ship is m f , then M − m f of antimat‐ only 0.5 pts.
tan y/2, which gives
ter is burnt throughout the acceleration and • Correct conservation laws for momentum
an equal amount of matter is burnt from the and energy (half of the points for each) 2z 1 − z2 2
interstellar space. Hence, it makes sense to 0.5 pts. The force acting on the rocket can now be ex‐ sin y = , cos y = , dy = d z,
1+ z 2 1 + z2 1 + z2
so that the integral becomes, after some ma‐ photon is observed by a stationary observer. Evaluated at x = L, this yields and hence,
nipulation, We can directly use the relativistic Doppler v
L2 ω u
Z shift effect to find u 1 + 3ρ a L2 1
dz v( x = L ) = . u 16ρ g t h 0
I= − = ln (1 − z). s
2h ω = ω0 u
t .
1− z 3ρ L2
1 + 16aρ g t h1
1 − v/ c m f
Substituting back, we use the trigonomet‐ f obs = f 0 = f0 Grading:
1 + v/ c M i
ric expressions sin y = 2 sin ( y/2) cos ( y/2) and mf 1 • Stating the idea of using conservation of We can see that limh→0 ω = 0, i.e. the air acts
cos y = 2 cos2 ( y/2) − 1 to rewrite z = tan ( y/2) = = f0 = f0. mass (or implicitly using it) to find the ve‐ as a cushion and stops the glass plate before
2M − m f 19
sin ( y/2)/ cos ( y/2) = p sin y/(2 cos2 ( y/2)) = locity 0.4 pts. it hits the stationary plate.
sin y/(1 + cos y) = x/(1 + 1 − x2 ), which after • Equation for the conservation of mass Grading:
insertion into the original integral gives Grading: 0.4 pts.
 
• Noticing from the laminarity of the flow
• Correct relativistic expression for doppler • Correct final expression 0.2 pts.
³ mf ´ that the kinetic energy of the glass plate
 xf  shift (‐0.3 pts if non‐relativistic or only valid ii) (2.5 points) Since the air flow is laminar
ln = ln 1 − q  and its surrounding air is conserved. If the
M 1 + 1 − xf 2 for v f << c) 0.5 pts. and there is no diffusion, all of the lost ro‐
flow is assumed to be dissipative, and the
• Final answer (no points if unphysical, eg tational energy from the falling glass plate
1 xf problem is otherwise solved correctly, the
⇒ = 1− q negative frequencies, even if consistent will be converted into kinetic energy of air
maximum score is reduced by 0.3 pts from
10 1+ 1− x 2 with previous calculations) 0.5 pts. particles. As such, we have K + K = const.
f rot air this part. 0.6 pts.
180 2. GAſ AND FLUıD FLOWſ (10 points) — Solu­ We found from the previous parts that the air • Calculating the kinetic energy of the air, of
⇒ xf = . tion by Taavet Kalda, grading schemes by Joonas is pushed out at ever faster speeds (between
181 which: 1.2 pts,
Kalda, Mihkel Kree, Andres Põldaru, Oleg Košik. the plates, v( x) À Lω) from between the two – Noting that the majority of the air’s kin‐
Grading: plates. However, outside of the two plates, etic energy is between the two glass
i) (1 point) As the plate falls, it will rotate
the flow will diffuse fast in all directions, plates (no formal proof required) 0.6 pts.
• Correct photon momentum change in the around the bump without slipping and push
such that the vast majority of kinetic energy – Finding the air velocity at a distance x
rest frame 0.5 pts the air out from beneath it, at ever faster
will be concentrated between the two plates. from the pivot 0.2 pts.
• Expanding the momentum change of the speeds, the closer it gets to the bottom plate.
rocket in terms of v and setting up total mo‐ As such, part of the rotational energy of the The kinetic energy of the air between the – Correct setup for the integral for the kin‐
mentum (or force) balance 0.25 pts plate is transferred over to the escaping air two plates is etic energy 0.3 pts.
• Eliminate the time variable 0.25 pts molecules. Further, the pressure and tem‐ – Correct final expression for the kinetic
Z L
• Express solution as an integral using separ‐ perature of the gas is uniform, because of the hx 1 energy of air 0.1 pts.
K in = d x ρ a v( x)2
ation of variables 0.5 pts incompressibility condition. 0 L 2 • Calculating the kinetic energy of the glass
2 Z slab, of which: 0.4 pts,
• Simplifying the integral and first substitu‐ Since the problem is 2‐dimensional, the ρ a Lω L
= x3 d x – Correct expression for the moment of in‐
tion 0.5 pts mass, volumes, moment of inertia and other 8h 0
• Second substitution 0.5 pts ertia 0.2 pts.
quantities are per unit length of the system ρ a L 5 ω2
• Substitute back and calculate final answer = , • Final expression for angular speed as a
(on the figure, into the page). Let x mark the 32 h
function of h 0.3 pts.
0.5 pts distance from the bump and v denote the ho‐
Non‐relativistic solution: rizontal speed of air at x. There is a volume of and the rotational energy of the glass plate is iii) (3 points)The incoming water from ‘2’ will
air equal to V ( x) = xhx/(2L) between x = 0 and spread out axisymmetrically along the space
• Momentum balance 0.25 pts I ω2 m glass L2 ω2 L3 tρ g ω2 between the stone disk and the ceiling. After
x. As the plate falls down, V ( x) gets smaller K rot = = = .
• Eliminate time variable and express as in‐ 2 3 2 6 that, it will spread into the basin and eventu‐
tegral using separation of variables and as a result, air is pushed out. Consider a
small time interval d t. In that time interval, h Energy conservation then reads as ally leave through the outgoing pipe ‘3’. The
0.25 pts stone disk is kept up by the pressure differ‐
• Evaluate the integral 0.25 pts
changes by ḣd t = −ωLd t. From the conserva‐
tion of air particles, 0 = dV ( x) + v( x)hx/L with ences between the top and bottom of the disc,
• Final answer (if commenting on unphys‐ ρ a L 5 ω2 L 3 t ρ g ω2
dV ( x) = − x2 ω/2. Hence, + = const arising from the flow speed of the water dif‐
ical results, otherwise no points) 0.25 pts 32 h 6 fering on either sides. The fact that pres‐
iii) (1 point) The last photon is emitted when ρ a L5 ω20 L3 tρ g ω20 sures differ on boths sides can be seen dir‐
xLω = + ,
the space ship moves at a speed of v f . The v( x) = . 32 h 0 6 ectly by the application of the Bernoulli Prin‐
2h
ciple p + ρ w v2 /2 = const along a streamline or Grading: • Correct numerical answer 0.1 pts. Thus,
by noting that the flow speed gradients are • Speed dependence v( x) from continuity v) (3 points)Because the expansion of water is µ µ ¶ µ ¶¶
RT0 p1 γ T1
driven by pressure gradients. condition 0.5 pts. reversible, entropy is conserved. This means r= ln − ln = 0.114.
ML p0 γ−1 T0
The flow speed outside the gap is negli‐ • Dynamic pressure from Bernoulli equation that the change in entropy due to the expan‐
gible due to t ¿ R , hence we can take the 0.5 pts. sion of the vapour is balanced by the entropy
To find the mass flow rate, we start by
pressure at the bottom side to be uniformly • Express force by integrating dynamic pres‐ change due to condensation. As discussed
finding the flow speed v of the outgoing li‐
p 0 . The flow speed in the gap at a distance sure over disk area. 0.5 pts. before, because entropy is a function of state,
quid/vapour. Since we know r , we can ap‐
x from the axis of symmetry can be found • Explanation why the jet’s impact force can it’s most convenient to calculate the entropy
ply energy conservation (necessary for the re‐
from the conservation of mass applied on a be ignored (or equivalently, an expression change by imagining n moles of water ( n will
versibility to hold) on a flowing water packet.
concentric cylinder of radius x and height t, for the impact force that is carried along in later cancel out) cooling and expanding from
This is most conveniently done by demand‐
giving 2π xtρ w v( x) = µ. Hence, the solution) 0.5 pts. T t , p t to T0 , p 0 and condensing rn moles of
ing energy conservation on the system as a
µ • Disk’s weight and buoyancy force 0.5 pts. water at the end. r is found by demanding
v( x) = . whole. Suppose that in some time interval n
2π xtρ w • Express answer from force balance 0.5 pts. that ∆S = 0 in this process.
moles of water vapour with a volume of Vt
Applying Bernoulli’s principle, we get p( x) + iv) (0.5 points) In the context of thermody‐ The entropy change of the vapour is are created at the boiler. Then at the out‐
ρ w v2 /2 = p 0 and so namics, entropy is defined in terms of its dif‐ found by applying the first law of thermody‐ let, in the steady state, due to conservation
ferential, such that the change in entropy of namics over a small temperature and pres‐ of particles, (1 − r )n moles of water vapour
ρ w v2 1 ³ µ ´2 a system is given by dS = dQ /T , where dQ is sure increment dT , d p:
∆ p = p 0 − p ( x) = = . at a volume of V0 , alongside rn moles of li‐
2 2ρ w 2π xt the heat entering the system, and T its tem‐ quid water, are removed. The outflowing wa‐
dQ dU + dW
We can hence find the force due to the pres‐ perature. Further, entropy in reversible ther‐ dS vapour = = , ter has an additional kinetic energy of nµv2 /2.
T T
sure differences in the gap by integrating modynamic processes is a state function, i.e. where dU = nc v dT is the change in internal
The energy change of the whole system due
from x = r to x = R . First note that the res‐ it only depends on the current (equilibrium) energy of the vapour and dW = pdV is the
to both steps must cancel each‐other out due
ulting force will be pointing vertically up, be‐ thermodynamical state of the system. This work done by the vapour. Importantly, we
to energy conservation. We can write this as
cause p( x) < p 0 . Integrating, means that when calculating the entropy dif‐
neglect the volume of water compared to the
Z x= R ference of one mole of vapour and liquid, the 0 = Win − Wout + Uin − Uout − nMv2 /2 = 0,
vapour, as that allows using the ideal gas
F1 = 2π x d x ∆ p temperature at which the phase transition
x= r to simplify the work differential. Using the where Win = p t Vt = nRT t and Wout = p 0 V0 =
Z took place does not affect the final result.
µ2 x= R
dx ideal gas law, we then have (1 − r ) nRT0 are the works done by the in‐
= As such, it’s most convenient to consider µ ¶
4π t 2 ρ w
x= r x nRT dp coming and outgoing packet and similarly,
µ ¶ the two final states as only differing by the li‐ pdV = pd = nR dT − nRT .
µ 2
R p p U in = c v nT t and Uout = c v nT0 − rnML + rnRT0
= ln . quid undergoing condensation at t0 = 100 °C. are the internal energies of the incoming and
4π t ρ w
2 r The final temperature is t0 = 100 °C because Hence,
outgoing packet. Notice that for the outgo‐
Note that the water entering through the that’s when water vapour pressure is equal to dT dp ing internal energy, we have an extra term of
dS vapour = n( c v + R ) − nR
pipe will slow down, pushing the disk fur‐ p 0 (i.e. boiling temperature at atmospheric T p r p 0 V0 = rnRT0 . This is because when talk‐
ther down. The net force from this, however, pressure). This corresponds to a heat of ∆Q = ing about the latent heat of vaporisation, it in‐
and we can integrate to get
turns out to be negligible due to the condi‐ mL = 1 mol · ML entering the vapour system, µ ¶ µ ¶ cludes the work done in order to expand the
compared to the liquid one. Hence, the en‐ T0 p0
tion r À t. To see this, one can argue that ∆S vapour = nc p ln − nR ln vapour into the volume it’s supposed to oc‐
the said force is of order µvpipe ∼ µµ/(ρ w r 2 ) ∼ tropy difference between one mole of vapour T1 p1
cupy. Therefore, we should subtract the said
µ2 /(ρ w r 2 ) ¿ F1 . and liquid is given by where c p = c v + R = R ( i + 2)/2 = R γ/(γ − 1) is work from the latent heat of vaporisation, in
Further, we have the gravitational force the heat capacity at constant pressure. The order for it to capture the actual change in
∆Q 1 mol · ML
F g = − mg = −πR 2 hρ s g pulling the disk down ∆S = = = 110 J/K. net change in entropy is then the internal energy of the water. Combining
T0 T0
and buoyancy force F b = πR 2 hρ w g pushing ∆S = 0 = ∆S + ∆S everything, we get
tot vapour condens
disk up. The force balance then reads F g + Grading: ML
F b + F1 = 0. Solving the equation, we find = ∆S vapour − rn 0 = nRT t − (1 − r ) nRT0 + c v nT t
s • ∆S = ∆TQ 0.1 pts. µ ¶
T0
µ ¶ − c v nT0 + rnML − rnRT0 − nMv2 /2
hρ w (ρ s − ρ w ) • ∆Q = LM 0.1 pts. T0 p0 ML
µ = 2πRt = nc p ln − nR ln − rn .
g.
• Understanding that T = T0 0.2 pts. T1 p1 T0 = ( c v + R ) n(T t − T0 ) + rnML − nMv2 /2,
ln(R / r )
 dφ
and hence, Note: using simple Bernoulli’s equation, which Grading: of m 1 ω2 r = − m 1 dr acting on an object of
s µ ¶
does not consider phase changes, gave 0 pts • Finding the trajectory in an inertial frame mass m 1 . The rope takes a shape that minim‐
c p ∆T since the problem is about deriving what happens 8 izes this potential energy. For this part and
v= 2 + rL = 906 m/s. 15 pts
M in that case. 8 the next, we will just be figuring out prop‐
• Finding the radial velocity 15 pts
3. ROTATıNG ſPACE ſTATıON (13 points) — • Finding the tangential velocity 8 erties of a rope that minimizes this poten‐
The density of air at the outlet is found from 15 pts tial energy (and other than that, we can for‐
ideal gas law ρ = p 0 M /(RT0 ). The mass flow Solution by Kaarel Hänni, grading schemes by • Correct velocity addition 4
15 pts get about the rotation). Consider cutting off
Adam Warnerbring, Kaur Aare Saar, Maksim • Correct numerical answer 2
rate of water vapour is then 15 pts. a tiny piece of rope of length ℓ from point
Pokrovskiy and ....
A p 0 Mv iii) (2 points) When the balloon comes to a C , then pulling the rope tight at C and glu‐
µvapour = A ρ v = , i) (0.5 points) Letting ω be the angular velo‐ stop, it is in equilibrium in the rotating frame. ing it back together (doing work ℓTC ), then
RT0 city, the acceleration experienced by people An object at radius R 0 of mass m has a ficti‐
¡ ¢2 1 cutting the rope open at A , letting it slip by
but we also have liquid water flowing out, on the “ground” is ω2 R = 2τπ R . Setting this tious radial (downward) force of m 1 ω2 R 0 act‐ length ℓ (doing work −ℓT A ), and finally mov‐
such that the total flow rate is given by equal to g gives ing on it in this frame. This force on the ing the tiny piece from point C to point A (do‐
µ ¶2 s mass m is mω2 (R − H + l ). The upward force
2
ing work ℓλ(φ(R ) − φ(R − h)) = − ℓλω h2(2R −h) ),
1 A p 0 Mv 2π R
µ= µvapour = = 59 g/s. R = g =⇒ τ = 2π ≈ 63.437 s. on the balloon is the difference between this filling the gap of length ℓ. The state of the
1−r (1 − r )RT0 τ g force for the balloon and the buoyant force rope is now the same as initially, so the total
in the rotating frame, which is 34 π r 3 ( M − work done should be 0:
Grading: Grading:
M 0 ) Vn ω2 (R − H ) = 34 π r 3 ( M − M 0 ) RP T ω2 (R − H ).
Finding r 1.5 pts • Finding the acceleration at the ”ground” Putting all this together, we can write down
G ℓλω2 h(2R − h)
ℓT C − ℓT A − = 0,
0.2 pts 2
• Idea that entropy is conserved. 0.2 pts the condition that the radial force is 0 in equi‐
• Idea to calculate entropy change of the • Correct expression for the period 0.2 pts librium from where
whole gas from T t , P t to T0 , P0 . 0.2 pts • Correct numerical answer 0.1 pts.
4 P λ2π2 h(2R − h)
• Equation dS = dU +TdW 0.1 pts, and ii) (2 points) Let us consider the motion of m ω2 ( R − H + l ) = π r 3 ( M − M 0 ) ω2 ( R − H ) T A − TC = − .
cdT + pdV
3 RG T τ2
dS = 0.1 pts. the ball in the non‐rotating (inertial) frame of
T
the center of mass of the spaceship. As the 4 3 0
3 π r ( M − M )P (R − H )
• Using the ideal gas law to calculate dS us‐
travel time is t = τ/2, the throwing point on =⇒ m = ≈ 110.8 kg. See 200 More Puzzling Physics Problems,
ing two of the variables T , V , P . 0.1 pts RG T (R − H + l )
the ground will rotate by exactly half a circle problem 78 (and its hint and solution) for a
• Integrating 0.2 pts and expressing entropy
between the ball being thrown and the ball longer explanation of this idea.
difference in terms of P and T 0.2 pts. Grading: If l is neglected, i.e. the mass
• Using or deriving c v = 3R . 0.1 pts being caught. In our inertial frame, the tra‐ and balloon are considered to be at equal Grading: Idea: move piece of rope from
• Correct entropy change due to the phase jectory of the ball will just be a straight line heights, this problem is marked out of 1.0 C to A/B 0.2 pts. Idea: work done by tension
change 0.2 pts between these two diametrically opposed pts maximum. Idea: balance of fictitious ra‐ 0.1 pts and correct expression 0.3 pts. Idea:
• Express the correct result for r 0.1 pts points; this has length 2R . In this inertial dial force and buoyant force 0.8 pts. Correct change in radial potential in rotating frame
frame, the ball thus travels with a constant expression for force on balloon 0.5 pts and OR kinetic energy in lab frame 0.1 pts and
Finding µ 1.5 pts R
velocity vinertial = 2τ/2 = 4τR (in the radial dir‐ mass 0.2 pts. Force balance correct 0.2 pts. correct expression 0.3 pts. Total work done
• Idea to use energy conservation along the ection). The initial velocity vector in the ro‐ Expression for m correct 0.2 pts. Numerical is zero 0.2 pts. Correct answer for tension dif‐
flow. 0.2 pts tating frame is the difference of the velocity answer correct 0.1 pts. ference 0.3 pts.
• Correct expression for work in the energy vector in the inertial frame and the velocity
iv) (1.5 points) In the rotating frame, letting r v) (1.5 points)Let’s use the equilibrium condi‐
conservation. 0.2 pts vector of the throwing point in the rotating
be the distance from the axis of the cylinder, tion that torque around the center of the cyl‐
• Correct internal energy change. 0.5 pts frame compared to the inertial frame. So the
there is an effective radial potential of inder is 0 (in the rotating frame) for the left
• Correct energy change due to kinetic en‐ initial velocity vector in the rotating frame
Z r half of the rope. Note that the fictitious force
ergy. 0.1 pts has a radial component of 4τR and a tangen‐ 2 ω2 r 2 is radial, so that contributes nothing to the
φ( r ) = (−ω x) dx = − ,
• Express correct v 0.1 pts. tial component of 2πτR . The throwing speed s 0 2 force. So the only contributions are from the
• Equation for gas flow rate µ = vρ g A . 0.1 pts is the magnitude of this vector, which is tension in the rope on the two sides, so
• The ideal gas law with density 0.1 pts s where we have chosen the potential zero
µ ¶ µ ¶
• Correction factor 1−1 r for the whole mass 4R 2 2π R 2 2 R p level to be at r = 0. In other words, in TA R−h
flow. 0.2 pts s = τ
+
τ
=
τ
4 + π2 ≈ 117 m/s. this frame, there is a fictitious radial force RT A cos α = (R − h)TC =⇒ =
TC R cos α
.
 Grading: Idea: rope in equilibrium 0.2 for tension 0.4 pts. Solved only for T A : full viii) (2 points) Consider a charge q1 at rest This lets us evaluate the main integral as
pts. Using zero total torque as condition 0.2 marks with penalty ‐0.1 pts. Correct numer‐ at radius r in the rotating frame. The I I
pts. Eliminating radial force by choice of ical value for TC 0.2 pts. electromagnetic force applied to it is frame‐ E ⃗ = 2πµ0 Q ⃗r · d A
⃗ · dA ⃗
2π r Q τ 2L
point 0.4 pts. Some correct torque expres‐ vii) (2 points)The rotating charge on the walls independently q1 τ µ0 τL . This charge is not I
sion containing both tensions 0.2 pts. Cor‐ is making the spaceship into a solenoid car‐ moving in the rotating frame, so in its frame, 4πµ0 Q 1
= ⃗ = 4πµ0 Q V .
⃗r · d A
rect ratio 0.5 pts. Q
rying current I = τ . Inside of the spaceship, the force applied to it by the magnetic field τ L
2 2 τ2 L
vi) (1.5 points) Let the x‐axis be the diameter this creates a constant axial magnetic field of must be 0 , and so the electric field measured
in its frame must satisfy q1 E = q1 2πτ r µ0 τQL , Grading:
AB, with coordinates chosen to be in meters, magnitude
and with the coordinates of A being (−R,0). from where • Arguing that in the rotating frame there is
only an electric field 0.4 pts
⃗ = 2πµ0 Q ⃗r.
(So the coordinates of B are (R,0) and the co‐ I Q
B = µ0 = µ0 . E
ordinates of C are (0, − (R − h)). The unique L τL τ L
2 • Finding the electric field strength 0.4 pts
parabola y = ax2 + bx + c that goes through • Choosing a suitable Gauss surface 0.4 pts
these three points has b = 0 from symmetry The force this creates on the charged ball We now have an expression for E at each • Evaluating the surface integral and obtain‐
2πR Q point in theHrotating frame; what remains is ing the correct answer 0.8 pts.
across the y axis, c = −(R − h) from condsid‐ is − qvB = − q τ µ0 τL in the radial upward
R −h to evaluate ⃗ · dA
E ⃗ . For the sake of variety,
ering the point C , and then a = − R 2 from direction. For the ball to hover above the 4. STRETCHıNG GLOVEſ (8 points) — Solution
considering the point A . The derivative at A “ground” motionlessly, the acceleration cre‐ we will demonstrate two ways to evaluate this by Eero Uustalu, grading schemes by ....
is ddxy | x=−R = 2a · R = − 2(RR−h) , from which ated by this radial force has to be equal to the integral.
i) (1 point)
2
centripetal acceleration ω2 R = (2πτ)2 R : For the first option, note that our E ⃗ =
−d y 1 const · ⃗r is exactly the electric field of a uni‐ Cut a few strips of the material with the
cos α = p =r ³ ´
( d y)2 + ( dx)2 2
q 2πR Q (2π)2 R q 2π L formly charged cylinder (with the correctly same constant width and mark them with
1 + ddxy − µ0 = =⇒ = − .
m τ τL τ2 m µ0 Q chosen charge density), and so we could equi‐ two perpendicular lines close to the ends of
valently find the same integral around such the strip while leaving ample space for affix‐
1
=q ≈ 0.7106. a uniformly charged cylinder, which Gauss’ ing or binding. Then roll one strip length‐
2 Grading:
1 + 4(RR−h)2 theorem gives as V ρ /ϵ0 = V · c 1 . We can find wise into a cylindrical shape for stretch‐
Method 1:
an expression for the constant c 1 by consider‐ ing. Measure the initial length between
4
Using the equations from parts (iv) and • Finding the current 15 pts ing the simple case where the surface we are the marked lines l 0 . With a straight meas‐
(v), we now have a system of equations in two • Finding the magnetic field 0.4 pts dealing with is a cylinder of width 1 m itself, uring tape affixed to the surface of the
unknowns: • Finding the upward force on the charged in which case the integral is table, stretch one of the strips to the break‐
2 ing point while determining the maximum
( 2
ball 3 pts I
R − h)
T A − TC = − λ2π hτ(2 • Writing out the centripetal acceleration of ⃗ · dA⃗ = 1 m · 2π r 2 πµ 0 Q 4 πµ 0 Q length between the lines before the strip
2
. E r=V 2 .
RT A · 0.7106 = (R − h)TC the ball 4 τ 2L τ L breaks l max . Repeat this at least three times
30 pts
• Correct relation between force and acceler‐ Hence, c = 4πµ0 Q , and so to see if the data is reproducible. Calculate
1 τ 2L
It remains to solve this system of equa‐ ation 4
15 pts
the result ϵmax = ( l max − l 0 )/(l 0 ).
I
tions. The second equation gives T A = • Correct final expression
4
pts ⃗ =V 4 πµ Q In our measurement we got ϵmax = 5.5.
15 E⃗ · dA 0
.
TC R ·R0.−7106
h
≈ TC · 0.7107. Plugging this into • Wrong sign of the charge 4
− 15 pts τ2 L
Grading:
the first equation then gives Method 2: To briefly describe a second option for • Selection of the actual latex piece for the
4 evaluating this integral, note that by parti‐
λ2π2 h(2R − h) • Finding the current 15 pts sample. Strips have to be of high quality
TC = 2 ≈ 12631 N.
τ (1 − 0.7107) • Finding the magnetic field 0.4 pts tioning a 3D body into volume slices, with and not include the finger end. 0.1 pts
• Finding the upward force on the charged each slice bounded by an area element d A ⃗ • Selecting the strip length: 0.2 pts
Grading: Some correct parabola given ball 2
3 pts and with cylindrical radius vector ⃗r , and with – at least 8cm: 0. 2/0 .2 pts
some choice of axes 0.3 pts. Correct expres‐ • Writing out the force balance in the rotat‐ the volume of each slice being ⃗r · d A ⃗ /2, we – at least 4cm: 0.1/0.2 pts
sion for cos α 0.2 pts and correct numerical ing frame (with the correct expression of get that the volume of the body is the sum of – shorter, not specified 0/0.2 pts
value 0.2 pts. System of equations using ex‐ the upward force) 0.4 pts volumes of all such slices, – Remark. If a series of different lengths
pressions found in iv) and v), even if they • Correct final expression
4 I
15 pts 1 of strips is used, then the medium value
4 V = ⃗
⃗r · d A.
are incorrect, 0.2 pts. Correct expression • Wrong sign of the charge − 15 pts 2 should be considered. Additionally, if
 the graph for finding ϵmax is used by the thickness a of the strip such that d = k·a measurement of the rope of strips was made
including the point (0,0), then 0.1 pts where k remains constant for the strip while using force F stretching the second 7 mm
should be subtracted. stretching. Therefore wide 80 mm test strip to 519 mm .
• Method of measurement, of which: 0.4 pts But the 7 mm stripe broke when it was
– one end affixed with tape or held using a stretched to 599 mm .
ruler 0.1 pts. F Fmax
– lines marked on the strip, or the strip is σ= and σmax = . So the dmax 0 = 7 mm and l max = 599 mm .
a· a· k a max ·a max · k
affixed over the edge of the ruler 0.1 pts
(if both ends are affixed, 0.2 pts, but
can’t get more than 0.4 pts for this sec‐
tion). From V = const we can deduce that a·a = mm mm mm
a 0 ·a 0 · l 0
– using narrow strips 0.1 pts.
l and a max ·a max = a0l·max
a0 ·l 0
. We make many strips with different ( l − l 0 )/ l o l0 l d0 σ/σmax
– measuring only the part of the strip that’s widths, mark them with the same initial
If we could apply the same force F on 0.638 80 131 71 0.022
being stretched (between marks) 0.1 pts. length l 0 using perpendicular lines, measure
strips of different initial widths d = k·a, we 1.606 80 208,5 37 0.066
– Measuring the full length of the stip and record the initial widths d0 of each seg‐
could vary the tension for each strip. There‐ 1.888 80 231 30 0.090
0.1 pts.
fore we would be using different strips with ment, then roll them lengthwise into a cyl‐ 3.025 80 322 21,5 0.175
• Repetition of measurements: 0.2 pts
indrical shape and bind them one after an‐
different k values for each measurement 3.338 80 347 18 0.226
– measurement repeated at least 5 times other. The force will be the same for all the
F l 1 F l max 1
0.2/0.2 pts. σ= k · l 0 · a 0 ·a 0 and σmax = k max · l 0 · a 0 ·a 0 ,
4.225 80 418 14 0.349
segments, but the value of σ for each seg‐
– measurement repeated at least 3 times 4.925 80 474 11 0.505
where k = d /a and kmax = dmax /a max are ment will depend on the initial width of each
0.1/0.2 pts. the width to thickness ratios of the the re‐ 5.488 80 519 7 0.866
individual strip.
– if distortion of measured ϵmax values is spective strips used for F and Fmax . By com‐
low +0.1 pts. A range where dmin and dmax differ at
bining the above, we get
• Final numerical value. If the value lies in least 8 times is recommended.
range 4‐7 (depends on the material, even 3‐ Rolling the strip lengthwise into a cyl‐
σ
11 may be ok) 0.1 pts. indrical shape before binding gives the pos‐ And the graph, σmax
for the Y axis and ϵ
sibility to distribute the force evenly on all for the X axis.
ii) (7 points)
σ k max l the strips regardless of the width of the strip!
1,000

Theoretical considerations. = · ,
σmax k l max 0,900

Our solution was to make two identical

We can assume the material has uniform strips with the smallest possible width and at‐
0,800

thickness. Knowing that σ = FA , V = const, tach them to one another at one end to form
0,700

and that any force applied to the strip affects a Y shape. The previously determined ϵmax
0,600

all directions perpendicular to the applied but since d0 = k·a 0 is the initial un‐ was used to estimate the possible maximum
0,500

force equally, then at any time for the same stressed width of one strip and dmax 0 = stretching almost to the breaking point of
0,400

strip k max ·a 0 is the initial unstressed width of an‐ the narrowest strip. In case of a breakage
0,300

other strip stretched to the breaking point, event the identical spare strip at the other
0,200

V = a·a· k· l = a 0 ·a 0 · k· l 0 = a max ·a max · k· l max , for building the graph it transforms to: 0,100

end of the Y shape could be used for meas‐

σ d l 0,000

where a 0 is the initial thickness of material σmax

= maxd 0 · l max for the Y axis and
0 urement. After affixing the multi‐width rope 0,000 1,000 2,000 3,000 4,000 5,000 6,000

of combined strips at its maximum stretch

with no force applied and l 0 is the initial l l
ϵ = −l 0 for the X axis. Approach 2
0 position, the values of l were recorded for
length of the strip with no force applied, a We make two identical elongated tri‐
Measurements. each individual section. Afterwards the rope
is the thickness when force F is applied and angles from the longest piece of material
was stretched till breaking to find the actual
l is the length of the strip when force F is ap‐ We can assume the material has uniform available. We mark the triangles with evenly
breaking length l max of the narrowest strip.
plied, a max and l max are the respective values thickness. We want to apply the same force spaced perpendicular lines. (The lines are
at the breaking force Fmax . on the strips with different initial widths. The 7 mm wide initially 80 mm long test
perpendicular to the central symmetry line
strip broke at length 522 mm. Therefore the
k is the ratio between the width d and Approach 1 of the elongated triangle.)
 mm mm mm ing (force may be mg) 0 pts.
σ/σmax l0 l d0 ( l − l 0 )/ l o • reasonable amount of repetitions (break‐ • stretching the strip to different lengths not
0.985 6 10 72.6 6.26 ing is not very well repeatable) knowing the actual force applied 0 pts.
0.605 8 10 59.5 4.95 BAD practice: • MOST OFTEN: Assumption that F=kx !
That is actually an alternative lightly modi‐
0.444 10 10 54.5 4.45 • intentionally variable length strips / short fied form of writing down the dependency
0.333 12.5 10 51 4.1 strips WE HAVE TO MEASURE, not assume! (and
0.266 15 10 49 3.9 • not marking the initial length on the strip the real dependency is not linear) 0 pts.
0.221 17.5 10 47.5 3.75 with a marker
We record the initial length l 0 and the av‐ Possible grading scheme for the solutions
0.183 20 10 45 3.5 • not stretched with perpendicularly uni‐
erage width d0 of each segment as defined given as example:
0.148 22.5 10 41 3.1 form strain
by the perpendicular lines. When stretched,
0.129 24 10 38 2.8 • not taking measurements only from actual part ii
all segments will have the same force applied
0.104 27 10 34.5 2.45 part to be stretched
but each segment will have its own σ. We theory:
0.074 32.5 10 29.5 1.95 • not accounting for possible friction (some
have to affix the base of the prolonged tri‐ • assuming that the material has initial uni‐
0.061 36 10 27 1.7 methods)
angle to the ruler with tape (it has to be a form thickness 0.2 pts.
good strong joint) and we affix the ruler to 0.051 39 10 24.5 1.45 part ii
0.043 41.5 10 22 1.2 • idea that as V = const then the dimensions
the table. A long strip of graph paper was other solutions proposed: in axes perpendicular to the applied force
taped to the table to have the rubber triangle 0.037 46.5 10 21 1.1
0.030 50 10 18.5 0.85 • multiple similar masses used (not in the change in a fixed ratio to each other 1.0 pts.
stretched over it. • idea of changing σ by changing the width
0.022 56 10 15 0.5 equipment list)
• a lever and some mass d of the material when constant force is ap‐
• from the same latex material (a very long plied 1.0 pts.
σ
And the graph, σmax
for the Y axis and ϵ uniform and relatively thick strip is needed • idea of applying a common force to all the
The triangle was stretched until the nar‐ for the X axis. to be in proportional range, but then the different samples at the same time 1.0 pts.
row end segment was near breaking point, 1,000

measured force will be low) • finding how to calculate the value of ϵ to be

and the values of l were recorded for each 0,900

• multiple similar strips in parallel as op‐ plotted 0.2 pts.

segment. (Most easily done by using a 0,800

posed to using fever strips in row and break‐ • finding how to calculate the value of σ/σmax
marker on the graph paper while holding the 0,700
ing with even force to be plotted 0.6 pts.
narrow end of the triangle at a fixed position 0,600
• trigonometry and constant force measurements and data:
with the other hand). One of the triangles 0,500
• different width strip pairs (force applied • the strips/intervals used are as long as pos‐
was actually stretched to the braking point, 0,400
must remain constant for all pairs, in some
the other triangle was stretched to the break‐ sible (in our case at least 14cm) 0.1 pts.
0,300
solutions friction would need to be accoun‐ • data range covered ‐ value of the largest
ing point only after the measurement was 0,200
ted for)
made. measured σ differs from the smallest meas‐
ured σ at least 8 times
0,100

unsuccessful solving attempts / bad prac‐ 0.1 pts.

The narrow end segment of the first tri‐ 0,000
0 1 2 3 4 5 6 7
tices: • data points ‐ at least 8 points measured (
angle of length 10 mm and of average width 0.2 pts for each correctly measured point
6 mm broke at length 73.1 mm. Therefore Grading: • using equipment not listed
• calculated, not measured (experiment!) or 0.1 pts for each cor‐
the measurement of the second similar tri‐ GOOD practice: rectly measured point if less than 4 points
0 pts.
angle tested was made using force F stretch‐ • using the longest possible strips measured (have you seen graph with 3 data
ing the narrow end segment of the triangle • marking the initial length on the strip with • using a finger as a force measuring device
0 pts. points?), for maximum 8 measured points)
to 72.6 mm. The actual breaking of the nar‐ a marker or MAX 1.6 pts.
row segment of the second triangle occurred • measuring initial length to one straight af‐ • assuming the repeatedly applied force is
constant (which it might not be) 0 pts. • data evenly distributed 0.1 pts.
at length 73.7 mm . fixed surface to another parallel affixed sur‐ • at least one data point measured with aσ
• measuring by varying the length of the
So the dmax 0 = 6 mm and l max = 73.7 mm . face strip while keeping the applied force con‐ value at least 0,8 of σ max 0. 1 pts.
• rolling the marker‐marked strip lengthwise stant (stress created remains constant) • measurements for the breaking event val‐
into a cylindrical shape for uniform stretch‐ ues correctly done 0.2 pts.
• chart of measured and calculated values units) 0.2 pts. • If the out of list equipment used does not • If the out of list equipment does enable a
given 0.3 pts. If the results of any intermediate calcu‐ allow for any principally new methods of principally new method of solution, credit
graph: lations are excessively rounded, a deduction solution, any credit awarded for the parts will only be awarded for aspects of the solu‐
−0.1 pts for every of the solution that rely on the use of the tion that would be applicable in a solution
• graph correctly plotted with all measured from the final score of out of list equipment or data measured that does not rely on the out of list equip‐
points 0.3 pts. intermediate stage used.
with it will be reduced by 50%. ment.
• graph plotted in units asked (ratios, no using out of list equipment: