My Answer to IMOC Problems

A1
Let D(n) denote the sum of the digits of n when written in decimal
notation. Find all positive integers n such that

n2 − D(n)2 + 8D(D(n)) = 2024.

Answer: The only n that satisfy is 46.

Solution Let [A]max and [A]min denote the maximum and the minimum of
A. We first prove that n can only be two digits. It is obvious that when n
has only one digit, the equations will never hold. Let n have a ≥ 3 digits
(a ∈ N). This implies that n2 has at least 2a − 1 digits. We also need to
consider the following:

[D(n)2 ]max = 81a2 (n = 99 · · · ),

[D(D(n))]min = 1(n = 100 · · · ).

So we have to prove that

102a−1 − 81a2 + 1 > 2024 (∀a ≥ 3).

We prove it by induction. When a = 3 the inequality holds. Assume when

a = k, 102k−1 − 81k 2 + 1 > 2024. Then

⇒ 2(102k−1 − 81k 2 + 1) + 8 × 102k−1 − 82 > 2024

⇒ 102k − 81(k + 1)2 + 1 > 2024.

1
so when a = k + 1 the inequality also holds. So n must be a 2-digit number.
Now we find the maximum and minimum of n.

[D(n)2 − D(D(n′ ))]max = D(99)2 − D(D(19)) = 323,

n2 − 323 ≤ 2024 ⇒ n2 ≤ 2347 ⇒ n ≤ 48.

and
[D(n)2 − D(D(n′ ))]min = D(11)2 − D(D(99)) = −68,
n2 + 68 ≥ 2024 ⇒ n2 ≥ 2340 ⇒ n ≥ 45.
Taking n = 45, 46, 47, 48 into the equation we find that the only solution is
46.

2
A3
Find all a1 , a2 , · · · , an such that,
p

 a31 − a2 = a3 − 1,
 p
3
 a2 − a3 = a4 − 1,



..
 .
p 3
an−1 − an = a1 − 1,





p 3
an − a1 = a2 − 1.

Answer: ai = 1 ∀ 1 ≤ i ≤ n.

Solution: Notice that

q
0 ≤ a3i − ai+1 = ai+2 − 1 ∀ 1 ≤ i ≤ n − 2.

Next, we square both sides and add up all the equations to obtain
n
X n
X n
X
a3i − a2i + ai − n = 0,
i=1 i=1 i=1

n
X
(a3i − a2i + ai − 1) = 0,
i=1
n
X
(ai − 1)(a2i + 1) = 0 ∀ 1 ≤ i ≤ n.
i=1

Since a2i + 1 > 0 and ai − 1 ≥ 0, the only solution is ai = 1 ∀ 1 ≤ i ≤ n.

3
C1
Is it possible to have several points in the plane such that, no matter how
these points are divided into 3 classes, there are always at least two points
in the same class whose distance apart is 1?

answer Yes, it is possible

solution For n ≥ 7 it is possible. Suppose there exists a way to classify the

points in the diagram into 3 categories such that each category contains two
points that are 1 unit apart. In the figure below all segments are intended
to have length 1. Once you put A in one group, B and C must be in the
other two groups, so D must be in the same group as A. Similarly, D1 must
be in the same group as A but D and D1 are 1 unit apart. Which is a
contradiction.
D
C C1 D1

A B1

4
G1
In triangle △ABC, let the angle bisector of ∠BAC intersect the circum-
circle of △ABC at M . Let P be any point on the circumcircle of △ABC.
Let AM intersect BP at Q, and let AC intersect M P at K. Prove that KQ
is parallel to BC.

Solution ⌢
Case 1: P lies on BAC ⌢ ⌢
Since ∠BAM = ∠M AC, it follows that BM = M C. Connect P and C,
and we have
∠BAM = ∠M AC = ∠BP M = ∠M P C.
Let AC and BP meet at D, and we have ∠ADB = ∠P DC. Since we also
have ∠ABD = ∠P CD, hence △AQD is similar to △P KD and △ADB is
similar to △P DC.
AD DQ DB
= = .
DP DK DC
Since ∠QDK = ∠BDC, thus △QDK is similar to △BDC. So KQ is par-
allel to BC.

5
 ⌢
Case 2:P lies on BM C
Let M C and QK meet at D. We first prove that Q, M, D, and P are
concyclic. By the similarity of △QM P and △CP K, we have PQP K
= M
CP
P
.
Since ∠CAM = ∠BAM , we also have ∠CP M = ∠BAM . So △QP K is
similar to △CP M . This implies ∠KQP = ∠CM P , and the cyclicity is
proven.
Let AB and QK meet at F . From the cyclicity, we have ∠QDM =
∠QP M = ∠BAM , and so F, A, D, and M are cyclic. This implies ∠AF Q =
∠AM C = ∠ABC. And we are done.

6
G2
Given triangle ABC with its incircle touching sides BC, CA, AB at points
D, E, F respectively. Let AEF have a circumcircle intersecting the circum-
circle of ABC at P , and P D intersects the circumcircle of ABC at M . Prove
that AM bisects angle ∠BAC.

solution We have ∠P BA = ∠P CA and ∠P F A = ∠P EA, so △P BF is

similar to △P CE. Also, BF = BD, EC = CD. Therefore,
BF BD PB
= = .
EC DC PC
⌢ ⌢
Thus, P D is the angle bisector of ∠BP C, implying BM = M C, and we are
done.

7
N1
Prove that for all integers a, there exist only finitely many integers x ̸= −a
such that
x + a | x2 + 2x + 6.

Solution To prove this, consider the divisibility condition:

x2 + 2x + 6 (x + a)(x + 2 − a) + 6 − 2a + a2
= .
x+a x+a
This simplifies to:
6 − 2a + a2
x+2−a+ .
x+a
Let k = x + 2 − a. Applying the quadratic formula, we find:
√
−2 ± 4 − 4 · 6
x= .
2a
Therefore, a2 − 2a + 6 = 0 has no real roots, implying a2 − 2a + 6 is always
a nonzero integer. Thus, there are only finitely many integers x such that
x + a divides x2 + 2x + 6.

8
N2
Let n be a positive odd integer, and a1 , a2 , . . . , an be all odd numbers.
Prove that
 
a1 + a2 a2 + a3 an−1 + an an + a1
gcd(a1 , a2 , . . . , an ) = gcd + ,..., + .
2 2 2 2

Solution Let d = gcd(a1 , a2 , . . . , an ) and define an+1 = a1 . Thus,

d | ai for all i = 1, 2, . . . , n.
Therefore, d | ai + ai+1 for all i = 1, 2, . . . , n.
Since ai are odd numbers, d must be odd, and ai + ai+1 are even. Hence,
ai + ai+1
d| .
2
Next, we prove that there does not exist an integer D > d such that
D | ai +a2 i+1 . Assume there exists such a D, then 2D | ai + ai+1 , implying
2D | (a1 + a2 ) + (a2 + a3 ) + · · · + (an−1 + an ) + (an + a1 ).
Thus,
2D | 2(a1 + a2 + · · · + an ),
which leads to
D | a1 + a2 + · · · + an .
Since n is odd,
n
X
D| ai − [(a1 + a2 ) + (a3 + a4 ) + · · · + (an−2 + an−1 )].
i=1

Therefore,
D | an .
Similarly, we can show D | ai for all i = 1, 2, . . . , n. However, gcd(a1 , a2 , . . . , an ) =
d, which is a contradiction. Hence, there does not exist an integer D > d
such that D | ai +a2 i+1 . Therefore,
 
a1 + a2 a2 + a3 an−1 + an an + a1
gcd + ,..., + = d.
2 2 2 2
This completes the proof.