More fun in 2016, Part 5

Geometry Level 5

Does there exist a right triangle with three integer sides whose area is 2016? If so, find the largest possible hypotenuse of such a triangle. If not, enter 666

Bonus question (very hard): Does there exist a right triangle with rational sides whose area is 2015?


The answer is 130.

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.

4 solutions

Otto Bretscher
Nov 24, 2015

We can use Euclid's formula for Pythagorean triples, a = 2 m n a=2mn and b = m 2 n 2 b=m^2-n^2 for the legs and c = m 2 + n 2 c=m^2+n^2 for the hypotenuse. The only integer solution of m n ( m 2 n 2 ) = 2016 mn(m^2-n^2)=2016 is m = 9 , n = 7 m=9, n=7 , with c = 130 c=\boxed{130} .

sir, have you found those integer solution by comparing?

Dev Sharma - 5 years, 6 months ago

Log in to reply

You can quickly see that m m and n n have to be odd, which leaves only a few cases to check.

Otto Bretscher - 5 years, 6 months ago

Log in to reply

would you please provide solution to bonus problem?

Dev Sharma - 5 years, 6 months ago

Log in to reply

@Dev Sharma See the discussion with Mr Mendrin below. Maybe I will pose this as a separate problem.

Otto Bretscher - 5 years, 6 months ago
Mark Hennings
Nov 25, 2015

An integer which is the area of a Pythagorean triangle with rational sides is called congruent . The bonus question asks if 2015 2015 is congruent. The Congruent Number Problem, determining whether a particular integer is congruent or not, is one of the big unsolved problems in Number Theory.

Assuming the Birch-Swinnerton Dyer Conjecture, it has been proved that all square-free integers congruent to any of 5 , 6 , 7 5,6,7 modulo 8 8 are congruent. This would make it highly likely that 2015 2015 is congruent. The number 2015 2015 is listed as congruent by Noe in a link that can be followed from OEIS, A003273.

Although I do not have an explicit Pythagorean triad to demonstrate it, I am led to believe that 2015 2015 is congruent.

Thank you for this very nice brief summary on congruent numbers, Dr. Hennings!

A student at Claremont- McKenna College, Jonathan Star, has written a fine Senior Thesis discussing these topics in detail, including an example of a rational right triangle with area 2015 (on page 2).

Otto Bretscher - 5 years, 6 months ago
Aareyan Manzoor
Nov 24, 2015

guys, one thing you should know is as one side gets bigger, the other gets smaller, but notice that h u g e 2 + s m a l l 2 s o m e 2 huge^2+small^2\neq some^2 .we will start from the middle hence. first, let adj=a,opp=b, hyp=c. then a b 2 = 2016 a b = 4032 \dfrac{ab}{2}=2016\Longrightarrow ab=4032 so lets check from middle: a , b = ( 64 , 63 ) c N ( a , b ) = ( 192 , 21 ) c N ( a , b ) = ( 32 , 126 ) c = 130 \begin{array}{c}&\quad\quad\quad\quad\quad\quad a,b=(64,63)\Longrightarrow c\notin \mathbb{N}\\ (a,b) = (192,21)\Longrightarrow c\notin\mathbb{N}\quad (a,b)=(32,126)\Longrightarrow c=130\end{array} it is obvious that further down either path will not result in any more rational 'c's. so we conclude 130 \boxed{130} is the only and the biggest 'c'(hypotenuse).

This method is computationally a bit tedious since 4032 has 42 divisors... but it works!

"Starting from the middle" isn't the right strategy here if you are in a hurry... you get a longer hypotenuse if one leg is really short and the other is really long.

Otto Bretscher - 5 years, 6 months ago

Log in to reply

sir, i meant that we will not get rational as we get up the road. for ex 403 2 2 + 1 \sqrt{4032^2+1} is not rational. the greatest pythagorean triplet for any int N is when n 2 + N 2 = ( n + 1 ) 2 n^2+N^2=(n+1)^2 for odd N,we can see that for 21 this is 22 0 2 + 2 1 2 = 22 1 2 220^2+21^2=221^2 for 32 it is the one that gave us answer. we can conclude since the lenghts are increasing and decreasing in a quick manner, that starting from middle will help. i am not certain of anything i just wrote, so correct me if i am wrong,sir.

Aareyan Manzoor - 5 years, 6 months ago

Log in to reply

If you "start in the middle", you have to examine all divisors to be certain that you have found the longest possible hypotenuse, but if you start "at the end" you can stop once you found a solution... it's just a minor point of strategy, really.

Otto Bretscher - 5 years, 6 months ago

Log in to reply

@Otto Bretscher that indeed is true sir. i saw opportunity that the values are ordered in such a way that i can reduce my work. but if i thought this was not possible, i would have started from the end. although i have a feeling that this included a lot of hypothetical guess work, so might be rejected perhaps,sir?

Aareyan Manzoor - 5 years, 6 months ago

i will do the bonus problem after solving that of more fun in 2016, part 4.

Aareyan Manzoor - 5 years, 6 months ago

Log in to reply

As you can probably see by now, the problems posed in Parts 4 and 5 of "More Fun in 2016" are equivalent: Find three perfect squares in an arithmetic progression and find a Pythagorean triangle with a given integer area.

Otto Bretscher - 5 years, 6 months ago
Michael Mendrin
Nov 24, 2015

Using the classic generator for Pythagorean triples, this would have to be true

1 2 ( 2 k m n ) k ( m 2 n 2 ) = 2015 \dfrac { 1 }{ 2 } (2kmn)k({ m }^{ 2 }-{ n }^{ 2 })=2015

so that we'd quickly end up with this, since 2015 2015 is square free

m n ( m n ) ( m + n ) = 5 13 31 mn(m-n)(m+n)=5\cdot 13 \cdot 31

which won't work out

There certainly are no integer solutions, but I am asking for rational sides... that makes the problem a lot trickier.

Otto Bretscher - 5 years, 6 months ago

Log in to reply

In that case, first we observe that any rational non-integer Pythagorean triple are the same as an integer Pythagorean triple a , b , c a,b,c divided by a common square factor x x . If x x doesn't divide into both a , b a, b , then the area cannot be integer. For the area to be integer, one of the sides, say a a has to be a multiple of the square of x x . Without loss of generality, we can then state that a , b a,b has to have the form

a = 2 m x 2 n a=2m{x}^{2}n
b = m 2 x 4 n 2 b={m}^{2}{x}^{4}-{n}^{2}

so that

( m x 2 n ) ( m 2 x 4 n 2 ) = 5 13 31 \left(m{x}^{2}n\right)\left({m}^{2}{x}^{4}-{n}^{2}\right)=5 \cdot 13 \cdot 31

which, again, does not work out. This is just a brief sketch of a fuller proof.

Michael Mendrin - 5 years, 6 months ago

Log in to reply

I'm looking forward to the fuller proof.... I'm not convinced yet.

Otto Bretscher - 5 years, 6 months ago

Log in to reply

@Otto Bretscher Well, first, the major preliminary is that any "rational non-integer Pythagorean triple" can only be your usual integer Pythagorean triple divided by the same factor. I'll get to this later if you do want to expand on this subject.

Michael Mendrin - 5 years, 6 months ago

Log in to reply

@Michael Mendrin I'm interested in your argument because I came to the opposite conclusion, namely, that such a triangle actually does exist.

I don't understand your claim "If x x doesn't divide into both a a and b b , then the area cannot be integer"

Otto Bretscher - 5 years, 6 months ago

Log in to reply

@Otto Bretscher In that case, Otto, since that would make an excellent problem, don't reveal the solution here, and post it as another problem. Then I'll give my answer there.

I was going to give my analysis here, but I'll hold off.

Michael Mendrin - 5 years, 6 months ago

Log in to reply

@Michael Mendrin I posted a much easier version first, replacing 2015 by 5: "Inspired by Fibonacci". It has not been solved yet. You are invited to have some fun with that one!

Otto Bretscher - 5 years, 6 months ago

Log in to reply

@Otto Bretscher Yeah, I already saw that one. I think you should post another for 2015.

Michael Mendrin - 5 years, 6 months ago

Log in to reply

@Michael Mendrin I will, in due time. But the case of 2015 is almost impossible to solve at this level of "recreational math"

When you look at the case of 5, you see that your claim "if x x doesn't divide into both a a and b b , then the area cannot be integer " does not hold

Otto Bretscher - 5 years, 6 months ago

Log in to reply

@Otto Bretscher I did say that the proof wasn't complete! It was just a sketch of a proof, IF there really was one.

Besides, what's wrong with posting a really hard problem here, in Brilliant.org?

Michael Mendrin - 5 years, 6 months ago

Log in to reply

@Michael Mendrin To chime in, there is a rational right triangle with area 2015. Good luck finding the sides.

This (more generally) is known as the Congurent Number theorem. The general case hasn't been solved. Somewhat interestingly, 5 is a congruent number, as seen from the right triangle with sides 3 2 , 20 3 , 41 6 \frac{3}{2}, \frac{20}{3}, \frac{41}{6} (1/6 of the (9,40,41) triangle).

Calvin Lin Staff - 5 years, 6 months ago

Log in to reply

@Calvin Lin Thank you for "chiming in", Calvin. We continued the discussion of 5 here , and I cited a recent paper where the sides of a rational right triangle with area 2015 are given.

(Calvin edit: Let's continue the discussion on the linked solution.)

Otto Bretscher - 5 years, 6 months ago

@Calvin Lin As I said, "IF there really was [a proof]" Subsequently, and elsewhere, I've carried on this conversation with Otto, and agreed that a solution [likely] does exist, and it's likely to involve numbers so huge it can't fit in here.

In other words, I'm saying I'm wrong with my initial assessment about this.

Michael Mendrin - 5 years, 6 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...