MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Is the following inequality known? I believe it's true, but I could find no reference.

For any convex body $C$ in the plane we have $$\left(4-\frac{8}{\pi}\right)area(C)\leq > diam(C)(per(C)-2diam(C)).$$

If true, this would be tight, with equality when $C$ is a disk. If it turns out not to be true, then it still makes sense to look for the best constant with the $area(C)$ term.

The following similar inequality $diam(C)(per(C)-2diam(C))\leq\frac{4}{\sqrt3}area(C)$ (equality when $C$ is an equilateral triangle) is definitely known and is proven here.

share|cite|improve this question
The paper, "Inequalities for Convex Sets," by Scott and Awyong, Journal of Inequalities in Pure and Applied Mathematics, might help, although I don't see your inequality there (but maybe I missed it amidst the welter of notation). PDF download: – Joseph O'Rourke Feb 1 '13 at 16:53
You should also check the book "Geometric Inequalities" by Burago and Zalgaller. – Misha Feb 1 '13 at 18:16
I had checked before both references (by Scott and Awyong and by Burago and Zalgaller), but I couldn't find what I need. – filipm Feb 1 '13 at 19:04
up vote 6 down vote accepted

This inequality is not true. Consider the rectangle on $\mathbb R^2$ with vertices $(\pm 1, 0)$, $(0, \pm \varepsilon)$. Then on the left you have $2\varepsilon(4-8/\pi)$ on the right you have approximatively $4\varepsilon^2$.

share|cite|improve this answer
Your numbers don't look right to me, but this does look like a counterexample. – Deane Yang Feb 2 '13 at 0:06
On second thought, this does not look like a counterexample. The left side appears to be equal to $4\epsilon(4−8/\pi)\simeq5.8\epsilon$ and the right side appears to be equal to $2\sqrt{1+\epsilon^2}(4(1+\epsilon)-4\sqrt{1+\epsilon^2}) \simeq 8\epsilon$. – Deane Yang Feb 2 '13 at 0:20
Sorry, there was a misprint in my answer. now it is correct – Dmitri Feb 2 '13 at 0:45
I also called rectangle what should be called a rombus :) – Dmitri Feb 2 '13 at 0:52
Yes. I misinterpreted your answer. – Deane Yang Feb 2 '13 at 1:53

Here's a variation on Dmitri's idea that works as a counterexample: Take the rhombus with long diagonal $2$ and short diagonal $2\epsilon$. Then the area (LHS) grows like $\epsilon$, but the perimeter minus twice diameter goes like $$4\sqrt{1+\epsilon^2}-4$$ which grows like $\epsilon^2$.

The idea is that by moving out $\epsilon$ in the "center" rather than the edges will give quadratic growth of the perimeter rather than linear.

share|cite|improve this answer
Connor, that is of course correct. In fact, I was thinking of exactly this example but for some reason (I guess to make the answer as short :) ) as possible put the vertices of the rombus in $\pm 1, \pm varepsilon$) instead of what I had in mind. – Dmitri Feb 2 '13 at 0:55
Yes, in restrospect I should have posted this as a comment on your answer- cheers! – Connor Mooney Feb 2 '13 at 1:16

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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