13
$\begingroup$

How many triangles with integral side lengths are possible, provided their perimeter is $36$ units?

My approach:

Let the side lengths be $a, b, c$; now,

$$a + b + c = 36$$

Now, $1 \leq a, b, c \leq 18$.

Applying multinomial theorem, I'm getting $187$ which is wrong.

Please help.

$\endgroup$
7
  • $\begingroup$ Did you take the triangle inequality into account? $\endgroup$ – J. M. isn't a mathematician Jul 13 '12 at 12:55
  • $\begingroup$ Yes, I did. When one side is of $1$ unit, then the max possible values for other two will be $18$ i.e. $(1, 18, 18)$ $\endgroup$ – Bazinga Jul 13 '12 at 12:56
  • 5
    $\begingroup$ $$18+18+1\neq36$$ $\endgroup$ – J. M. isn't a mathematician Jul 13 '12 at 13:00
  • $\begingroup$ You didn't say how you did it. Possibly permutations of $a,b,c$ are considered to give the same triangle; did you account for that (after all drawing the the triangle elsewhere or in a different orientation does not count as giving a different triangle either, I suppose)? $\endgroup$ – Marc van Leeuwen Jul 13 '12 at 13:10
  • 1
    $\begingroup$ I counted the number of positive integer solutions $\{a,b,c\}$ to $a+b+c=36$; it is 108. But of course they are not all triangles. $\endgroup$ – i. m. soloveichik Jul 13 '12 at 13:18
18
$\begingroup$

The number of triangles with perimeter $n$ and integer side lengths is given by Alcuin's sequence $T(n)$. The generating function for $T(n)$ is $\dfrac{x^3}{(1-x^2)(1-x^3)(1-x^4)}$. Alcuin's sequence can be expressed as

$$T(n)=\begin{cases}\left[\frac{n^2}{48}\right]&n\text{ even}\\\left[\frac{(n+3)^2}{48}\right]&n\text{ odd}\end{cases}$$

where $[x]$ is the nearest integer function, and thus $T(36)=27$. See this article by Krier and Manvel for more details. See also Andrews, Jordan/Walch/Wisner, these two by Hirschhorn, and Bindner/Erickson.

$\endgroup$
2
  • $\begingroup$ What if the perimeter is 18? Then T(N) isn't an integer. $\endgroup$ – MCCCS Mar 30 '17 at 6:53
  • 1
    $\begingroup$ Did you notice the use of the nearest integer function? $\endgroup$ – J. M. isn't a mathematician Mar 30 '17 at 9:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.