12 Posts • Page 1 of 1
 |
 |
Author |
 |
Message |
 |
 |
 |
 |
 |
 |
Pascual2005
Navier-Stokes Equations


Offline Joined: 24 Mar 2004 Posts: 1160 Location: colombia
|
2 circles Romanian IMO Team Selection Test TST 1999, problem 12; 17-th Iranian Math. Olympiad 1999/2000
Two circles intersect at two points and . A line which passes through the point meets the two circles again at the points and , respectively. Let and be the midpoints of the arcs and (which do not contain the point ) on the respective circles. Let be the midpoint of the segment . Prove that .
|
Posted: Sat May 01, 2004 3:39 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
grobber
Birch & Swinnerton Dyer

Offline Joined: 07 Apr 2003 Posts: 7862 Location: Romania
|
I know I once gave a pretty neat solution, but I definitely can't remember it now so I'll try to post the outline of a not-so-nice solution (it has no ingenious ideas ).
(1) Try to find the locus of , the midpoint of , as th line varies. You'll see that it's the circumcircle of , where is the midpt of (the centers of the 2 circles).
(2) Find the locus of , the midpoint of , as varies. It's the circle with diameter , so it also has center .
(3) Show that and we're done, because is on the perpndicular bisector of (from (1)), and from (3) we find that the line is the perpendicular bisector of , so , so is on th circle centered at which passes through , and that's precisely what we want, because (this is because ).
|
Posted: Sat May 01, 2004 10:43 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
Thanhliem
Poincare Conjecture

Offline Joined: 19 Mar 2004 Posts: 134
|
solution
O1O2 is center of two circle meet A,B O1M cut CB at P , O2N cut BD at Q then
Then <NAM=90 then KP/NQ=BP/QK,then <PKM=<QNK => KNQ=90+<CPQ+<QDK=>MKN=90 
|
Posted: Wed May 12, 2004 1:54 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
E. Leung
P versus NP

Offline Joined: 01 Sep 2003 Posts: 25
|
A different solution:
Let the mid-points of BC and BD be M' and N' respectively. Then M'KN'B is a parallelogram, and .
. Thus the two right triangles BNN' and MBM' are similar.
Then, , giving . In addition,
Thus .
So,

|
Posted: Sat Apr 16, 2005 9:37 am |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
darij grinberg
Birch & Swinnerton Dyer


Offline Joined: 10 Feb 2004 Posts: 5806 Location: Karlsruhe / Munich
|
Re: solution
Although we already have 3 solutions for the original problem now, let me show a rather uncommon solution using inversion:
Problem. Two circles c and d intersect each other at two points A and B, and a line l passing through the point A meets these two circles at the points C and D (apart from the point A). Let M and N be the midpoints of the arcs BC and BD which do not contain A on the circles c and d, respectively. Let K be the midpoint of the segment CD. Prove that < MKN = 90°.
Solution. The crucial point of the solution will be an inversion with center A. But before we undertake this inversion, we transform the problem into a more convenient form:
The point M is the midpoint of the arc BC on the circle c. Thus, the two chords BM and CM of the circle c have equal length. Hence, the chordal angles of these two chords must also be equal; in other words, we have < BAM = < CAM. Hence, the point M lies on the angle bisector of the angle BAC. Similarly, the point N lies on the angle bisector of the angle BAD. But, of course, the angle bisector of the angle BAD is the exterior angle bisector of the angle BAC; hence, since the interior and the exterior angle bisector of an angle are always perpendicular to each other, the angle bisector of the angle BAC and the angle bisector of the angle BAD are perpendicular to each other. Hence, < MAN = 90°.
The problem requires to prove < MKN = 90°. Since we know that < MAN = 90°, this will trivially follow if we show < MKN = < MAN. But this is equivalent to proving that the points M, N, A and K lie on one circle. Hence, in order to solve the problem, it is enough to prove that the points M, N, A and K lie on one circle.
Now, perform an inversion with respect to an arbitrary circle centered at the point A. Denote by B', C', D', M', N' and K' the images of the points B, C, D, M, N and K under this inversion. What properties do these new points B', C', D', M', N' and K' have?
- We know that the circle c passes through the points A, B, C and M. Since this circle passes through the point A, which is the center of the inversion, it is mapped by the inversion to a line; this line must pass through the images B', C' and M' of the points B, C and M under our inversion. Thus, the points B', C' and M' lie on one line. Similarly, the points B', D' and N' lie on one line.
- We know that the point M lies on the angle bisector of the angle BAC. Since our inversion has the center A, the images B', C' and M' of the points B, C and M under this inversion must lie on the rays AB, AC and AM, respectively, and thus we can state that the point M' lies on the angle bisector of the angle B'AC'. Similarly, the point N' lies on the angle bisector of the angle B'AD'.
- The point K is the midpoint of the segment CD. This seems to be a property hard to deal with using inversion. However, there is a workaround: We denote by the infinite point of the line l (this is the point of intersection of the line l with the line at infinity). Then, the image of this point under our inversion is the point A (since an inversion maps infinite points to the center of the inversion).
Also, since every segment is divided harmonically by its midpoint and the infinite point of the line containing the segment, the points K and divide the segment CD harmonically (since the point K is the midpoint of the segment CD, and the point is the infinite point of the line l = CD). In other words, the points C, D, K and on the line l form a harmonic quadruple. Now, it is well-known that if four points lying on a line through the center of an inversion form a harmonic quadruple, then so do their images under the inversion. Since the line l passes through the center A of our inversion, we thus conclude that the points C', D', K' and A, being the images of the points C, D, K and under our inversion, must also form a harmonic quadruple. Thus, the point K' is the harmonic conjugate of the point A with respect to the segment C'D'.
- Finally, let's see how the inversion transforms the assertion that we have to prove. We have already seen that, in order to solve the problem, it suffices to show that the points M, N, A and K lie on one circle. In other words, it suffices to show that the points M, N and K lie on one circle through the point A. Since the point A is the center of our inversion, and since an inversion maps circles through the inversion center to lines and conversely, it is clear that proving that the points M, N and K lie on one circle through the point A is equivalent to showing that their inversive images M', N' and K' lie on one line.
Hence, in order to solve the problem, it will be enough to show that the points M', N' and K' lie on one line.
Now, everything is simple: Since the points B', C' and M' lie on one line, and the point M' lies on the angle bisector of the angle B'AC', the point M' is the point of intersection of the angle bisector of the angle B'AC' with the line B'C'. Since the angle bisector of an angle of a triangle divides the opposite side in the ratio of the adjacent sides, we have . Similarly, the point N' is the point of intersection of the angle bisector of the angle B'AD' with the line B'D', and we conclude that . Finally, since the point K' is the harmonic conjugate of the point A with respect to the segment C'D', we have . Thus,
.
Hence, by the Menelaos theorem, the points M', N' and K' lie on one line. The problem is solved.
EDIT: This problem comes from the 24th Russian Mathematical Olympiad 1997.
Darij
|
Posted: Wed Apr 27, 2005 11:19 pm Last edited by darij grinberg on Mon Apr 09, 2007 7:43 pm; edited 4 times in total |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
Armo
Riemann Hypothesis

Offline Joined: 14 Aug 2004 Posts: 377 Location: Yerevan
|
As I remember a more general problem is G8 in the shortlist of IMO2002. I can post the original solution if it is needed.
[Moderator edit: Thanks for noticing! Now I have found various solutions for this problem G8 in the following sources:
http://www.kalva.demon.co.uk/short/soln/sh02g8.html
http://www.mathlinks.ro/Forum/viewtopic.php?t=22165
http://www.mathlinks.ro/Forum/viewtopic.php?t=22201
http://www.mathlinks.ro/Forum/viewtopic.php?t=17322
http://www.mathlinks.ro/Forum/viewtopic.php?t=15588 post #2 file IMO_2002_shortlist.pdf page 21 .]
|
_________________ It is better the river of my life to be wide and short than long and narrow.
Avicena
Posted: Fri Apr 29, 2005 10:23 am |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
mecrazywong
Yang-Mills Theory

Offline Joined: 07 Jul 2004 Posts: 605 Location: Hong Kong
|
Let and be the midpoints of and respectively. It is obvious that . Then . The result follows.
|
Posted: Sun Sep 25, 2005 8:37 am |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
jarod
Poincare Conjecture

Offline Joined: 14 Jul 2005 Posts: 186
|
Please clarify your solution . I think It's not obvious and I believe there is a simpler solution
|
Posted: Sun Sep 25, 2005 12:54 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
mecrazywong
Yang-Mills Theory

Offline Joined: 07 Jul 2004 Posts: 605 Location: Hong Kong
|
Basically my solution is the same as the one posted by E. Leung (even the notation!), so I'm not going to explain it in details.
Anyway, although it's not really that obvious, I personally believe that anyone associated with Math Olympiad should be able to solve it.
|
Posted: Sun Sep 25, 2005 5:00 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
Virgil Nicula
Birch & Swinnerton Dyer

Offline Joined: 22 Jun 2005 Posts: 4621 Location: Bucuresti (RO) Bradenton (FL)
|
A equivalent enunciation. Own
A equivalent enunciation of the proposed problem:
Let the triangle , the midpoint of the side and two isosceles triangles
such that the line separes the points , the line separates the points and
Prove that and .
Remark. After at least a hour I will prove with complex numbers !
|
Posted: Mon Sep 26, 2005 12:12 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
Arne
Birch & Swinnerton Dyer


Offline Joined: 07 Sep 2003 Posts: 3694 Location: Belgium
|
My solution was somewhat different, but I also used spiral similarity. [I am just solving the problem the way levi stated it.]
Let and be the reflections of and in and respectively. It is easy to see that the right-angled triangles and are similar (this follows from ). Hence, triangles and are similar, and there is a spiral similarity (angle ) mapping the to . This spiral similarity maps to , so and are perpendicular. Since , we are done.
|
Posted: Mon Sep 26, 2005 12:26 pm |
 |
|
|
 |
 |
 |
 |
 |
 |
 |
Virgil Nicula
Birch & Swinnerton Dyer

Offline Joined: 22 Jun 2005 Posts: 4621 Location: Bucuresti (RO) Bradenton (FL)
|
My solution. own
(See the last my reply). I note:
, i.e.
1. Remark. Let be the point of the line for which .
Then the are cyclic quadrilaterals and thus, we obtain the initial proposed problem !.
2. Remark. This problem admits a generalization (see www.mathlinks.ro/Forum/viewtopic.php?p=289313#p289313 ):
. Let be a triangle and three isosceles triangles
such that the lines separe the pointa from the points respectively and
. Then
.
For we obtain the previous problem !.
|
Posted: Mon Sep 26, 2005 1:22 pm |
 |
|
|
 |
 |
 |
 |
|
 |
 |
12 Posts • Page 1 of 1
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum You cannot post calendar events in this forum
|