A1.  P is a point inside the triangle ABC. D, E, F are  the feet of the perpendiculars from P to the lines BC, CA, AB  respectively. Find all P which minimise:          BC/PD + CA/PE + AB/PF.                 
A2.  Take r such that 1 ≤ r ≤ n, and consider all subsets of r  elements of the set {1, 2, ... , n}. Each subset has a smallest element.  Let F(n,r) be the arithmetic mean of these smallest elements. Prove  that:          F(n,r) = (n+1)/(r+1).                 
A3.  Determine the maximum value of m2 + n2, where m and n are integers in the range 1, 2, ... , 1981 satisfying (n2 - mn - m2)2 = 1.                 
B1. (a)  For which n > 2 is there a set of n consecutive  positive integers such that the largest number in the set is a divisor  of the least common multiple of the remaining n - 1 numbers?  (b)  For which n > 2 is there exactly one set having this property?                 
B2.  Three circles of equal radius have a common point O and lie  inside a given triangle. Each circle touches a pair of sides of the  triangle. Prove that the incenter and the circumcenter of the triangle  are collinear with the point O.                   
B3.  The function f(x,y) satisfies: f(0,y) = y + 1, f(x+1,0) =  f(x,1), f(x+1,y+1) = f(x,f(x+1,y)) for all non-negative integers x, y.  Find f(4, 1981). 
Solutions 
Problem A1  
P is a point inside the triangle ABC. D, E, F are the feet of the  perpendiculars from P to the lines BC, CA, AB respectively. Find all P  which minimise:  
        BC/PD + CA/PE + AB/PF.  
Solution  
We have PD.BC + PE.CA + PF.AB = 2 area of triangle. Now use Cauchy's inequality with x1 = √(PD·BC), x2 = √(PE·CA), x3 = √(PF·AB), and y1 = √(BC/PD), y2 = √(CA/PE), y3 = √(AB/PF). We get that (BC + CA + AB)2 < 2 x area of triangle x (BC/PD + CA/PE + AB/PF) with equality only if xi/yi = const, ie PD = PE = PF. So the unique minimum position for P is the incenter. 
Problem A2  
Take r such that 1 ≤ r ≤ n, and consider all subsets of r elements of  the set {1, 2, ... , n}. Each subset has a smallest element. Let F(n,r)  be the arithmetic mean of these smallest elements. Prove that:  
        F(n,r) = (n+1)/(r+1).  
Solution  
Denote the binomial coefficient n!/(r!(n-r)!) by nCr.  
Evidently nCr F(n,r) = 1 (n-1)C(r-1) + 2 (n-2)C(r-1) + ... + (n-r+1)  (r-1)C(r-1). [The first term denotes the contribution from subsets with  smallest element 1, the second term smallest element 2 and so on.]  
Let the rhs be g(n,r). Then, using the relation (n-i)C(r-1) -  (n-i-1)C(r-2) = (n-i-1)C(r-1), we find that g(n,r) - g(n-1,r-1) =  g(n-1,r), and we can extend this relation to r=1 by taking g(n,0) = n+1 =  (n+1)C1. But g(n,1) = 1 + 2 + ... + n = n(n+1)/2 = (n+1)C2. So it now  follows by an easy induction that g(n,r) = (n+1)C(r+1) = nCr  (n+1)/(r+1). Hence F(n,r) = (n+1)/(r+1).  
A more elegant solution by Oliver Nash is as follows  
Let k be the smallest element. We want to evaluate g(n, r) = ∑k=1 to n-r+1  k   (n-k)C(r-1). Consider the subsets with r+1 elements taken from 1,  2, 3, ... , n+1. Suppose k+1 is the second smallest element. Then there  are k   (n-k)C(r-1) possible subsets. So g(n, r) = (n+1)C(r+1). Hence  F(n, r) = (n+1)C(r+1) / nCr = (n+1)/(r+1), as required. 
Problem A3  
Determine the maximum value of m2 + n2, where m and n are integers in the range 1, 2, ... , 1981 satisfying (n2 - mn - m2)2 = 1.  
Solution  
Experimenting with small values suggests that the solutions of  n2 - mn - m2 = 1 or -1 are successive Fibonacci numbers. So suppose n > m is a solution. This suggests trying m+n, n: (m+n)2 - (m+n)n - n2 = m2 + mn - n2 = -(n2 - mn - m2)  = 1 or -1. So if n > m is a solution, then m+n, n is another  solution. Running this forward from 2,1 gives 3,2; 5,3; 8,5; 13,8;  21,13; 34,21; 55,34; 89,55; 144,89; 233,144; 377,233; 610,377; 987,610;  1597,987; 2584,1597.  
But how do we know that there are no other solutions? The trick is to  run the recurrence the other way. For suppose n > m is a solution,  then try m, n-m: m2 - m(n-m) - (n-m)2 = m2 + mn - n2 = -(n2 - mn - m2)  = 1 or -1, so that also satisfies the equation. Also if m > 1, then m  > n-m (for if not, then n >= 2m, so n(n - m) >= 2m2, so n2 - nm - m2 >= m2  > 1). So given a solution n > m with m > 1, we have a smaller  solution m > n-m. This process must eventually terminate, so it must  finish at a solution n, 1 with n > 1. But the only such solution is  2, 1. Hence the starting solution must have been in the forward sequence  from 2, 1.  
Hence the solution to the problem stated is 15972 + 9872. 
Problem B1  
(a)  For which n > 2 is there a set of n consecutive positive  integers such that the largest number in the set is a divisor of the  least common multiple of the remaining n - 1 numbers?  
(b)  For which n > 2 is there exactly one set having this property?  
Solution  
(a)  n = 3 is not possible. For suppose x was the largest number in the  set. Then x cannot be divisible by 3 or any larger prime, so it must be a  power of 2. But it cannot be a power of 2, because 2m - 1 is odd and 2m - 2 is not a positive integer divisible by 2m.  
For k ≥ 2, the set 2k-1, 2k , ... , 4k-2 gives n = 2k. For k ≥ 3, so  does the set 2k-5, 2k-4, ... , 4k-6. For k ≥ 2, the set 2k-2, 2k-3, ... ,  4k-2 gives n = 2k+1. For k ≥ 4 so does the set 2k-6,2k-5, ... , 4k-6.  So we have at least one set for every n ≥ 4, which answers (a).  
(b)  We also have at least two sets for every n ≥ 4 except possibly n =  4, 5, 7. For 5 we may take as a second set: 8, 9, 10, 11, 12, and for 7  we may take 6, 7, 8, 9 ,10, 11, 12. That leaves n = 4. Suppose x is the  largest number in a set with n =4. x cannot be divisible by 5 or any  larger prime, because x-1, x-2, x-3 will not be. Moreover, x cannot be  divisible by 4, because then x-1 and x-3 will be odd, and x-2 only  divisible by 2 (not 4). Similarly, it cannot be divisible by 9. So the  only possibilities are 1, 2, 3, 6. But we also require x ≥ 4, which  eliminates the first three. So the only solution for n = 4 is the one we  have already found: 3, 4, 5, 6. 
Problem B2  
Three circles of equal radius have a common point O and lie inside a  given triangle. Each circle touches a pair of sides of the triangle.  Prove that the incenter and the circumcenter of the triangle are  collinear with the point O.  
Solution  
Let the triangle be ABC. Let the center of the circle touching AB and AC  be D, the center of the circle touching AB and BC be E, and the center  of the circle touching AC and BC be F. Because the circles center D and E  have the same radius the perpendiculars from D and E to AB have the  same length, so DE is parallel to AB. Similarly EF is parallel to BC and  FD is parallel to CA. Hence DEF is similar and similarly oriented to  ABC. Moreover D must lie on the angle bisector of A since the circle  center D touches AB and AC. Similarly E lies on the angle bisector of B  and F lies on the angle bisector of C. Hence the incenter I of ABC is  also the incenter of DEF and acts as a center of symmetry so that  corresponding points P of ABC and P' of DEF lie on a line through I with  PI/P'I having a fixed ratio. But OD = OE = OF since the three circles  have equal radii, so O is the circumcenter of DEF. Hence it lies on a  line with I and the circumcenter of ABC. 
Problem B3  
The function f(x,y) satisfies: f(0,y) = y + 1, f(x+1,0) = f(x,1),  f(x+1,y+1) = f(x,f(x+1,y)) for all non-negative integers x, y. Find f(4,  1981).  
Solution  
f(1,n) = f(0,f(1,n-1)) = 1 + f(1,n-1). So f(1,n) = n + f(1,0) = n + f(0,1) = n + 2.  
f(2,n) = f(1,f(2,n-1)) = f(2,n-1) + 2. So f(2,n) = 2n + f(2,0) = 2n + f(1,1) = 2n + 3.  
f(3,n) = f(2,f(3,n-1)) = 2f(3,n-1) + 3. Let un = f(3,n) + 3, then un = 2un-1. Also u0 = f(3,0) + 3 = f(2,1) + 3 = 8. So un = 2n+3, and f(3,n) = 2n+3 - 3.  
f(4,n) = f(3,f(4,n-1)) = 2f(4,n-1)+3 - 3. f(4,0) = f(3,1) = 24 - 3 = 13. We calculate two more terms to see the pattern: f(4,1) = 224 - 3, f(4,2) = 2224 - 3. In fact it looks neater if we replace 4 by 22, so that f(4,n) is a tower of n+3 2s less 3.
Read moreSolutions are also available in     Murray S Klamkin, International  Mathematical Olympiads 1978-1985 , MAA 1986, and in   István Reiman,  International Mathematical Olympiad 1959-1999, ISBN 189-8855-48-X.
, MAA 1986, and in   István Reiman,  International Mathematical Olympiad 1959-1999, ISBN 189-8855-48-X.  
 Labels:
IMO
Labels:
IMO

 
 Previous Article
 Previous Article
