Last updated: 17th September 2024 Stefan Ivanov

Physics Olympiad Errata

This file lists all the errors in Physics Olympiad solutions that I am aware of. Even at the international
level, solutions files have more errors than one would think. People can spend a lot of time searching
for a mistake in their work, when in fact it’s the official solution that is wrong. My hope is that hav-
ing all the errors in one place will make it easier for students to go through the problems by themselves.

All problems and solutions can be found in my archives.

IPhO
For reference, I still haven’t checked the papers from 2019, 2015, 2014, 2011, 2008, 2005 and everything
before that.

(1) 2017 2B.1 Depending on how you measure the arrival time, your answers in the next two parts
may diverge significantly from those in the original solutions. So, to be explicit – assume the
arrival of the P-wave corresponds to the instant when the seismomeder reading becomes nonzero.

(2) 2017 2B.2 Another comment – do not take any additional measurements for this part. Use
only your results from B1.

(3) 2017 2B.3 And yet another clarification – assume that the first signal to arrive at DNP is
indeed due to the wave travelling through the mantle.

(4) 2017 2B.4 The original solution is incomplete. We need to take the upper limit of the integral
2
p
for X, and only then do we find X = ap 1 − (pv0 )2 .

(5) 2017 2B.6 The result of this calculation is very sensitive to the denominators, and the original
solution hasn’t been careful with this. I get T = 184.1 s.

(6) 2017 2C.1 To make the problem statement less vague, you are being asked to find the potential
energy of the slab of height h with respect to the ocean level.

(7) 2017 3B.1/3B.2 These two parts of the problem are wrong, do not attempt them. They expect
you to work with a nonzero k and make use of the scale factor’s time dependence from A4. But
that was obtained using k = 0!

(8) 2017 3B.3 The “condition for inflation” you are expected to use is w = −1, not anything that
comes later in the text.

(9) 2017 3D.2 To be a bit more rigorous that the official solution, the observational constraints are
−5.19 > n > −6.69 and n > −1.81, and these cannot be satisfied simultaneously.

(10) 2022 2C The model used here isn’t self-consistent (as discussed in the solutions), and I think
it’s impossible to figure out what you are supposed to do if you are attempting the problem on
your own. The rest of the problem is really nice.

EuPhO
Nothing here yet!

APhO
(11) 2003 2A The solution is wrong because it doesn’t account for the relativistic correction on the
speed of light in a moving medium, which is of order ΩR/c too. This is the same correction as
the one in Fizeau’s experiment (for that, see Wikipedia or IZhO 2018.3). The final result should
2Ω
be ∆t = 4πR c2
as per this paper – the refractive index doesn’t matter.

1
 Stefan Ivanov

(12) 2003 2B ∆L = c′ ∆t isn’t the correct expression for the optical path difference. The OPD is
what you multiply by the wavevector in vacuum k to get the phase difference ∆ϕ, so
ω
∆ϕ = ω∆t = k∆L ⇒ ∆L = ∆t = c∆t.
k
Then in the next part we have ∆L = 3.0 × 10−12 m. After that we find

8π 2 R2 N Ω
∆θ = N ω∆t = . ?
cλ

(13) 2015 2C The ratio of the magnetic fields is 0.3, not 1.0.
q 
B0
(14) 2015 2D.ii The inequality should go the other way, θcr ≤ arcsin Bm .

IZhO
I think I’ve caught everything between 2018 and 2024, but the earlier papers are there for the taking.

(15) 2018 3.2.4 The numerical answer in the marking scheme is wrong. The one in the solution is
correct.

(16) 2018 3.3.2 Using the same approach (switching to the rest frame of the flow, using Snell’s law
there, and then switching back), I find

sin α v n2 − 1
 
sin β = + ,
n c n
2
implying a constant offset B1 = n cn−1 . I got the same result using the formulae on this site.
Then 3.3.3 would be wrong as well. The final two tasks are fine. ?
(17) 2020 1.1 I don’t think it’s allowed to use noninertial frames the way they do. My answer is
2
ω 2 r2

r1
ε=+ 1 = 1.5 × 10−7 .
g0 r2

I found this by comparing the tension forces on a pendulum in equilibrium at midday and at
midnight. The tension, summed with the Earth’s and the Sun’s gravitational forces, should
equal the centripetal force for the pendulum bob’s trajectory around the Sun. I made use of
r1 ω1 r1
r2 ≪ ω2 r2 ≪ 1. I’m not sure about my answer, but certainly one should get something that
depends on ω1 . ?
(18) 2020 1.3.1 The thickness of the border of the triangle should be ∆r = 2r1 = 2 mm.

(19) 2020 1.3.2 The thickness of the border of the star should be ∆r = 3r2 = 0.3 mm.

(20) 2021 1.1 The formulae for the frequencies are correct, but the values should be ω1 = 6.14 rad/s
and ω2 = 10.20 rad/s.

(21) 2022 1.3 The final answer should be

√
8 2ε0 mg
Q= .
σ
In going
R π/2 from equation 5R to equation 6, they’ve missed an extra factor of cos β. One needs to
π/2
find −π/2 cos2 β dβ, not −π/2 cos β dβ.

(22) 2022 2.9 The answer for α in the marking scheme is wrong. The solution has the correct one,
α = 0.014 K−1 .

2
 Stefan Ivanov

(23) 2022 3.6 The calculation is off by a hundred; The time is τ = 1.27 × 1010 s.
eV0
(24) 2023 3.9 There is a typo in the formula for T . We should have T ≪ kB = 8.70 × 105 K.

USAPhO
I’m only collecting errors from before 2007. The solutions of the papers from 2008 and onwards are
still maintained, and their most recent versions can be found here. If you find an error in those, email
Kevin Zhou.

(25) 1998 B2 In part A, the number of protons is Z = 56. In the original solution they plug in r1 =
rmax , but in reality r1 = rmax /β. This error propagates to part B, where I get λ = 5.8 × 10−9 m.

NBPhO
Nothing here yet!

InPhO
(26) 2014 7E One of the terms in β has an extra factor of 5/3:

1 + f 1/3 + (5/3)f + f 2 + f 5/3

β= .
(1 + f )5/3 (1 + f 1/3 )

Contributing to the list

There are many errors missing from this file, and a single person can’t hunt all of them down. This is
where I ask for your help! If you have found an error, please email me so that I can add it to the list.
Borrowing Donald Knuth’s idea, I will award physics money (i.e. Joules) for your troubles, as follows:

• Verifications. Worth 5 J. There are some errors here that I am not certain about. I’ve marked
them with a ? . I’d like someone else to double-check those. Message me with the number of
the error (e.g. 23) and attach some working which supports or disproves what’s written down
in the list. It doesn’t have to be neat, just legible.

• Wrong solutions. Worth 10 J. Some problems are correctly stated, but there are major issues
with their solutions. What I count as an error is something which leads to a wrong final answer,
either in the formula or in the numerical value. For example, a minus which disappears in one
line of the solution but reappears in the next is fine with me – this sort of typo is quite common
and not too harsh on the reader. Should you notice a significant error, please:
1. Explain why the official solution is wrong.
2. Show me what the correct answer is.

• Wrong problems. Worth 15 J. Occasionally there are problems which are so wrong that one
cannot patch up the solution and call it a day. One way this can happen is when a problem author
forgets about a key physical effect, and the setup actually does things which are completely
different from what the problem statement hints at (e.g. overdamping instead of oscillations).
If you think a problem is wrong, please outline why. There should be enough detail so as to
convince a fellow student aiming for IPhO.

Keep in mind that I am only tracking the competitions listed above, that is, IPhO, EuPhO, APhO,
IZhO, USAPhO, NBPhO, and InPhO. If maintaining the file proves easy enough, I might include a
few more in the future.

3
 Stefan Ivanov

Energy balance
Additions to the list are tracked and credited. If you want to stay anonymous, that’s alright too!

Alex Prodanov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 J
▷ (17)