GeometryDifficulty 7.1National Olympiad, round 2Prove itUnited States
Let ABC be an acute triangle with circumcircle Ω and orthocenter H. Points D and E lie on segments AB and AC respectively, such that AD=AE. The lines through B and C parallel to DE intersect Ω again at P and Q, respectively. Denote by ω the circumcircle of △ADE.
a. Show that lines PE and QD meet on ω.
b. Prove that if ω passes through H, then lines PD and QE meet on ω as well.
Solution
Solution to (a) Note that ∠AQP=∠ABP=∠ADE and ∠APQ=∠ACQ=∠AED, so we have a spiral similarity △ADE∼△AQP. Therefore, lines PE and QD meet at the second intersection of ω and Ω other than A.
Solution to (b) using angle chasing Let L be the reflection of H across AB, which lies on Ω. Claim — Points L,D,P are collinear. Proof. This is just angle chasing: ∠CLD=∠DHL=∠DHA+∠AHL=∠DEA+∠AHC=∠ADE+∠CBA=∠ABP+∠CBA=∠CBP=∠CLP.□
Now let K∈ω such that DHKE is an isosceles trapezoid, i.e. ∠BAH=∠KAE. Claim — Points D,K,P are collinear. Proof. Using the previous claim, ∠KDE=∠KAE=∠BAH=∠LAB=∠LPB=∠DPB=∠PDE.□ By symmetry, QE will then pass through the same K, as needed.
Solution to (b) by orthogonal circles (found by contestants) We define K as in the previous solution, but do not claim that K is the desired intersection. Instead, we note that: Claim — Point K is the orthocenter of isosceles triangle APQ. Proof. Notice that AH=AK and BC=PQ. Moreover from AH⊥BC we deduce AK⊥PQ by reflection across the angle bisector. In light of the formula "AH2=4R2−a2", this implies the conclusion. □ Let M be the midpoint of PQ. Since △APQ is isosceles, AKM⊥PQ and we conclude that MK⋅MA=MP2. So the circle with diameter PQ is orthogonal to ω. Combined with (a), this implies the result by Brokard theorem.
Solution to (b) by complex numbers Let M be the arc midpoint of BC. We use the standard arc midpoint configuration. We have that A=a2,B=b2,C=c2,M=−bc,H=a2+b2+c2,P=ba2c,Q=ca2b, where M is the arc midpoint of BC. By direct angle chasing we can verify that MB∥DH. Also, D∈AB. Therefore, we can compute D as follows. d+a2b2dˉ=a2+b2anddˉ−hˉd−h=−mb2=b3c⟹d=c(bc+a2)a2(a2c+b2c+c3−b3). By symmetry, we have that e=b(bc+a2)a2(a2b+bc2+b3−c3). To finish, we want to show that the angle between DP and EQ is angle A. To show this, we compute e−qd−p/e−qd−p. First, we compute d−p=c(bc+a2)a2(a2c+b2c+c3−b3)−ba2c=a2(c(bc+a2)a2c+b2c+c3−b3−bc)=bc(bc+a2)a2(a2c−b3)(b−c). By symmetry, e−qd−p=a2b−c3a2c−b3⟹e−qd−p/e−qd−p=a2bc3a2b3c=c2b2 as desired.
Want a route through all this instead of an archive? The track
puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.
Source: MathNet,
licensed CC-BY-4.0.
Statement reproduced verbatim; metadata (topic, difficulty) added by this project.