Vieta's Formula Only Plays A Small Role

Let a , b , c 2016 a,b,c\leq2016 be positive integers that satisfy the equation a 2 + b 2 + 2 = a b c . a^{2}+b^{2}+2=abc. What is the largest possible value of b b ?


The answer is 571.

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

Brandon Monsen
Oct 19, 2016

Lemma 1

Suppose for a fixed c c that a , b a,b satisfies the equation. Then b , b 2 + 2 a b,\frac{b^{2}+2}{a} is also a solution

Proof 1

We know that a 2 a b c + b 2 + 2 = 0 a^{2}-abc+b^{2}+2=0 . Let a = x a=x . Then our polynomial becomes x 2 b c x + b 2 + 2 = 0 x^{2}-bcx+b^{2}+2=0 and has solutions a , d a,d for some value of d d .

We know by Vieta's Formula that a + d = b c a+d=bc , and since b c a bc-a is an integer, d d must also be an integer.

Vieta's formula also tells us that a d = b 2 + 2 ad=b^{2}+2 , so we know that d = b 2 + 2 a d=\frac{b^{2}+2}{a} . Hence, b , b 2 + 2 a b,\frac{b^{2}+2}{a} is also a valid solution \square .

Lemma 2

If a > b 3 a>b\geq3 , then 0 < d < b 0<d<b .

Proof 2

The fact that d = b 2 + 2 a d=\frac{b^{2}+2}{a} is a positive integer divided by a positive integer tells us that d > 0 d>0 .

d = b 2 + 2 a < b 2 + 2 b = b + 2 b < b + 1 d=\frac{b^{2}+2}{a}<\frac{b^{2}+2}{b}=b+\frac{2}{b}<b+1 . If d d is an integer, and b + 1 b+1 is an integer, then d < b + 1 d<b+1 implies that d b d\leq b .

If d = b d=b , then substituting this into the original equation gets that 2 b 2 + 2 b 2 c = 0 2 = b 2 ( c 2 ) 2b^{2}+2-b^{2}c=0 \rightarrow 2=b^{2}(c-2) , and it is easy to see that our only solution will be d = b = 1 , c = 4 d=b=1, c=4 , but b 3 b\geq3 , a contradiction.

Hence, 0 < d < b 0<d<b \square .

Now , from the above information we can conclude that if a solution exists, call it A k , B k , C 1 A_{k},B_{k},C_{1} , such that A > B 3 A>B\geq3 , we can "reduce" this solution to another solution A k 1 , B k 1 , C 1 A_{k-1},B_{k-1},C_{1} , and continue this process until we reach A 1 , B 1 , C 1 A_{1},B_{1},C_{1} , where b < 3 b<3 . Hence, we only need to check for possible values of c c at b = 1 , 2 b=1,2 .

b = 1 b=1

a 2 + 1 a c + 2 = 0 a 2 + 3 = a c a + 3 a = c a^{2}+1-ac+2=0 \rightarrow a^{2}+3=ac \rightarrow a+\frac{3}{a}=c . Since in order for c c to be an integer, 3 a \frac{3}{a} must also be an integer, so we get that a = 1 , 3 a=1,3 are possible solutions. Either case gives us c = 4 c=4 .

b = 2 b=2

a 2 + 4 2 a c + 2 = 0 a 2 + 6 = 2 a c a + 6 a = 2 c a^{2}+4-2ac+2=0 \rightarrow a^{2}+6=2ac \rightarrow a+\frac{6}{a}=2c . By the same logic as before we can conclude that a = 1 , 2 , 3 , 6 a=1,2,3,6 are all possible values of a a , and we can then conclude that 2 c = 5 , 7 2c=5,7 . In either case, c c is not an integer, so we can reject this case.

Lastly , since we know that c = 4 c=4 is the only possible value of c c , we can simply work our way back up a chain A n . B n , 4 A_{n}.B_{n},4 until one of the values surpasses 2016. We do this by using the same relation of our decreasing sequence in reverse, being b n + 1 = a n 2 + 2 b n b_{n+1}=\frac{a_{n}^{2}+2}{b_{n}} , and interchanging a a and b b as we go.

( A 1 , B 1 ) = ( 1 , 1 ) ( A 1 . B 2 ) = ( 1 2 + 2 1 , 1 ) = ( 3 , 1 ) ( A 2 , B 2 ) = ( 3 , 3 2 + 2 1 ) = ( 3 , 11 ) ( A 2 , B 3 ) = ( 1 1 2 + 2 3 , 11 ) = ( 41 , 11 ) ( A 3 , B 3 ) = ( 41 , 4 1 2 + 2 11 ) = ( 41 , 153 ) ( A 3 , B 4 ) = ( 15 3 2 + 2 41 , 153 ) = ( 571 , 153 ) ( A 4 , B 4 ) = ( 571 , 57 1 2 + 2 153 ) = ( 571 , 2131 ) \left(A_{1},B_{1}\right)=\left(1,1\right) \\ \left(A_{1}.B_{2}\right)=\left(\frac{1^{2}+2}{1},1\right) = \left(3,1\right) \\ \left(A_{2},B_{2}\right)=\left(3,\frac{3^{2}+2}{1}\right)=\left(3,11\right) \\ \left( A_{2},B_{3}\right)=\left( \frac{11^{2}+2}{3},11\right)=\left( 41,11\right) \\ \left(A_{3},B_{3}\right)=\left( 41,\frac{41^{2}+2}{11}\right)=\left( 41,153\right) \\ \left(A_{3},B_{4}\right)=\left(\frac{153^{2}+2}{41},153\right)=\left(571,153\right) \\ \left(A_{4},B_{4}\right)=\left(571,\frac{571^{2}+2}{153}\right)=\left(571,2131\right)

Note that each of the a = 1 a=1 solutions appear in this chain, so we can conclude that the other chain will give us the same results.

Since 2131 > 2016 2131>2016 , our answer is 571 \boxed{571} .

When you found that for c = 4 c = 4 , there are 2 possible chain starting values, you have to investigate the other chain.

In particular, it brings up the question of "In lemma 2, we didn't deal with scenarios of the form a = b a = b ". Also, if you look further, all that you shown is d b d \leq b , while what your lemma stated is 0 < d < b 0 < d < b .

This can be easily fixed. Do you see how?

Calvin Lin Staff - 4 years, 7 months ago

Log in to reply

I rejected the d = b d=b case since it ended up having one solution, being 1 , 1 , 4 1,1,4 , and didn't fit the assumption that a > b 3 a>b\geq3 .

I think that the second chain starting value would be 1 , 1 , 4 1,1,4 , whereas I started with 1 , 3 , 4 1,3,4 ?

Luckily that ends up being the same chain, so the answer would remain the same?

Brandon Monsen - 4 years, 7 months ago

Log in to reply

Oops, didn't read the next line of the lemma for d < b d < b . The a = b a = b concern still applies though. The way around it is to either
1. Explain that if a b 3 a \geq b \geq 3 , we can conclude that a > b a > b
2. When explaining the chain of solutions and what makes a solution "smaller", we say the size of ( a , b ) (a,b) is a + b a + b (as opposed to max ( a , b ) \max (a,b) )

Right, since it's the same chain, we end up with the same answer. Had it lead to a different chain, we could have a different answer. Hence, just check / state it, for completeness.

Calvin Lin Staff - 4 years, 7 months ago

Log in to reply

@Calvin Lin I changed the chain part to include the a = b a=b case, and also explained how this chain must be the same as the 1 , 3 , 4 1,3,4 chain

Brandon Monsen - 4 years, 7 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...