Let $ABCD$ be a convex quadrilateral with the lengths $a, b, c, d$ and the area $S$. The main result in our paper equivalent to:

\begin{equation} a^2+b^2+c^2+d^2 \ge 4S + \frac{\sqrt{3}-1}{\sqrt{3}}\sum{(a-b)^2}\end{equation}

where $\sum{(a-b)^2}=(a-b)^2+(a-c)^2+(a-d)^2+(b-c)^2+(b-d)^2+(c-d)^2$ and note that $\frac{\sqrt{3}-1}{\sqrt{3}}$ is the best constant for this inequality.

Similarly with the same form above apply to n-polygon, I propose a conjecture that:

Let a convex polygon $A_1A_2...A_n$ with the lengths are $a_1, a_2, ...a_n$ and area $S$ we have:

\begin{equation} \sum_i^n{a_i^2} \ge 4\tan{\frac{\pi}{n}}S + k\sum_{i < j}{(a_i-a_j)^2}\end{equation}

I guess that $k=tan{\frac{\pi}{n}}-tan{\frac{\pi}{n+2}}$

I am looking for a proof of the inequality above.

Note that using Isoperimetric inequality we can prove that:

\begin{equation} \sum_i^n{a_i^2} \ge 4\tan{\frac{\pi}{n}}S\end{equation}

Case n=3:

\begin{equation}a^{2} + b^{2} + c^{2} \geq 4 \sqrt{3}S+ (a - b)^{2} + (b - c)^{2} + (c - a)^{2}\end{equation}

Case n=4: our paper, $k=\frac{\sqrt{3}-1}{\sqrt{3}}$, $k=tan{\frac{\pi}{4}}-tan{\frac{\pi}{6}}$

\begin{equation} a^2+b^2+c^2+d^2 \ge 4S + \frac{\sqrt{3}-1}{\sqrt{3}}\sum{(a-b)^2}\end{equation}

See also:


2 Answers 2


The conjectured inequality, with $k=\tan{\frac{\pi}{n}}-\tan{\frac{\pi}{n+2}}$, is false for $n=3$. More specifically, the constant factor $k=1$ is optimal in the Hadwiger--Finsler inequality: e.g., consider $a=b=1$ and $c\approx0$.

  • $\begingroup$ I will chesk by my computer now. Thank You very much. $\endgroup$ Jul 10, 2018 at 20:50
  • $\begingroup$ I check in long time, but I can not find the optimal formular of $k$ $\endgroup$ Jul 11, 2018 at 2:25
  • $\begingroup$ So far, we don't even seem to have a good expression for the optimal $k$ covering the cases of $n=3$ and $n=4$. $\endgroup$ Jul 11, 2018 at 11:44
  • $\begingroup$ I think optimal of $k \approx \tan\frac{\pi}{n}- \tan\frac{\pi}{n+2}$ $\endgroup$ Jul 11, 2018 at 12:16

The keyword you are looking for is "quantitative isoperimetric inequalities". The case of polygons was solved in the following paper: https://arxiv.org/abs/1402.4460

The "quantitative" term in the paper above is not necessarily of the form you seek. It involves the standard deviation of the radii and central angles.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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