Why do We Fall? To Rise Again!

a a and b b are positive integers such that a b a \ge b and the expression below is an integer: a + 1 b + b + 1 a . \frac{a+1}{b}+\frac{b+1}{a}. Find the sum of all values of a a less than 1000 1000 that satisfy the above conditions.


Inspiration: [Problem #1, Spain Mathematical Olympiad, 1996]

a a and b b are positive integers such that a + 1 b + b + 1 a \frac{a+1}{b}+\frac{b+1}{a} is an integer. Show that the greatest common divisor of a a and b b is not greater than a + b \sqrt{a+b} .


The answer is 1380.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

1 solution

Kazem Sepehrinia
Jun 26, 2017

Relevant wiki: Vieta Root Jumping

We use Vieta jumping method for solving this problem. Write a + 1 b + b + 1 a = k a 2 + a + b 2 + b = k a b a 2 + ( 1 k b ) a + b 2 + b = 0 , \frac{a+1}{b}+\frac{b+1}{a}=k \ \ \ \ \Longrightarrow \ \ \ \ a^2+a+b^2+b=kab \ \ \ \ \Longrightarrow \ \ \ \ a^2+(1-kb)a+b^2+b=0, where k k is a positive integer and last equation is a quadratic in terms of a a . Suppose that ( a 0 , b 0 ) (a_0, b_0) is a solution to the above quadratic, with a 0 b 0 a_0 \ge b_0 . If we let b = b 0 b=b_0 , equation has two solutions: a = a 0 a=a_0 and a = a a=a^* . By Vieta's formulas a 0 + a = k b 0 1 a_0+a^*=kb_0-1 and a 0 a = b 0 2 + b 0 a_0 a^*=b_0^2+b_0 . Observe that a = k b 0 1 a 0 a^*=kb_0-1-a_0 is an integer, and a = b 0 2 + b 0 a 0 a^*=\frac{b_0^2+b_0}{a_0} is positive.

Therewith, if a 0 b 0 2 + b 0 b 0 = b 0 + 1 a_0 \ge \frac{b_0^2+b_0}{b_0}=b_0+1 , then a = b 0 2 + b 0 a 0 b 0 a^*=\frac{b_0^2+b_0}{a_0} \le b_0 , and we can choose ( a 1 , b 1 ) = ( b 0 , a ) (a_1, b_1)=(b_0, a^*) as a solution to the equation. Note that a 1 + b 1 = b 0 + a < a 0 + b 0 a_1+b_1=b_0+a^* < a_0 +b_0 . In a similar manner, we can construct an another solution ( a 2 , b 2 ) (a_2, b_2) such that a 2 + b 2 < a 1 + b 1 a_2+b_2<a_1+b_1 and etc.

Well-ordering principle tells us that we cannot repeat this procedure ( Fall! ) \left( \color{#D61F06} \text{Fall!}\right) forever! Therefore, we will encounter a solution ( a n , b n ) (a_n, b_n) where a n < b n 2 + b n b n = b n + 1 a_n< \frac{b_n^2+b_n}{b_n}=b_n+1 , and we can no longer jump downwards. Now we must find the possible values of ( a n , b n ) (a_n, b_n) , i.e., possible values of the smallest solution. Note that b n a n < b n + 1 b_n \le a_n <b_n+1 . It follows that a n = b n a_n=b_n , put this back in the original equation: a n 2 + ( 1 k a n ) a n + a n 2 + a n = 0 a n + 1 k a n + a n + 1 = 0 k = 2 ( a n + 1 ) a n a_n^2+(1-k a_n)a_n+a_n^2+a_n=0 \ \ \ \ \Longrightarrow \ \ \ \ a_n+1-k a_n+a_n+1=0 \ \ \ \ \Longrightarrow \ \ \ \ k=\frac{2(a_n+1)}{a_n} Therefore a n 2 ( a n + 1 ) a_n \mid 2(a_n+1) or a n 2 ( a n + 1 ) 2 a n = 2 a_n \mid 2(a_n+1)-2a_n=2 , which gives a n = 1 , 2 a_n=1, 2 with k = 4 , 3 k=4, 3 , respectively. So we must reconstruct solutions ( Rise! ) \left( \color{#D61F06} \text{Rise!}\right) using two cases: ( a n , b n ) = ( 1 , 1 ) (a_n, b_n)=(1, 1) with k = 4 k=4 and ( a n , b n ) = ( 2 , 2 ) (a_n, b_n)=(2, 2) with k = 3 k=3 .

Notice that ( a n , b n ) = ( b n 1 , k b n 1 a n 1 1 ) (a_n, b_n)=(b_{n-1}, k b_{n-1}-a_{n-1}-1) and since a n = b n 1 a_n=b_{n-1} we get ( a n 1 , b n 1 ) = ( k a n b n 1 , a n ) . (a_{n-1}, b_{n-1})= (k a_n-b_n-1, a_n).

Above recurrence relation gives the values of a {\color{#D61F06}a} less than 1000 1000 , for ( a n , b n ) = ( 1 , 1 ) (a_n, b_n)=(1, 1) with k = 4 k=4 , as ( 1 , 1 ) , ( 2 , 1 ) , ( 6 , 2 ) , ( 21 , 6 ) , ( 77 , 21 ) , ( 286 , 77 ) . ({\color{#D61F06}1}, 1), ({\color{#D61F06}2}, 1), ({\color{#D61F06}6}, 2), ({\color{#D61F06}21}, 6), ({\color{#D61F06}77}, 21), ({\color{#D61F06}286}, 77). And gives the values of a a less than 1000 1000 , for ( a n , b n ) = ( 2 , 2 ) (a_n, b_n)=(2, 2) with k = 3 k=3 , as ( 2 , 2 ) , ( 3 , 2 ) , ( 6 , 3 ) , ( 14 , 6 ) , ( 35 , 14 ) , ( 90 , 35 ) , ( 234 , 90 ) , ( 611 , 234 ) . ({\color{#D61F06}2}, 2), ({\color{#D61F06}3}, 2), ({\color{#D61F06}6}, 3), ({\color{#D61F06}14}, 6), ({\color{#D61F06}35}, 14), ({\color{#D61F06}90}, 35), ({\color{#D61F06}234}, 90), ({\color{#D61F06}611}, 234). Thus the answer is 1 + 2 + 3 + 6 + 14 + 21 + 35 + 77 + 90 + 234 + 286 + 611 = 1380 1+2+3+6+14+21+35+77+90+234+286+611=\boxed{1380} .

I gave 1388 as answer as i counted 2 and 6 twice since (2,1) ,(2,2), (6,2), (6,3) were distinct solutions.

Shivansh Kaul - 3 years, 11 months ago

Log in to reply

You've misread the question!

Kazem Sepehrinia - 3 years, 11 months ago

Log in to reply

Misinterpreted; but it could have been clearer

Shivansh Kaul - 3 years, 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...