2
$\begingroup$

The problem can be equivalently stated as

PROBLEM : A convex $n$ sided polygon has a circumcircle and an inscribed circle, its area is $B$, and the areas of circumcircle and inscribed circle are $A$ and $C$ respectively. Prove that $2B < A+C$.

I think this problem is very difficult. This is my attempt for special case of polygons i.e Regular polygons.

Naming of parameters :

$R$ be the radius of circumcircle of the polygon.
$r$ be inradius of the polygon.
$n$ be the number of sides of polygon. $\theta$ = $\frac{2\pi}{n}$ = angle subtended by a side of polygon at centre.
$a$ be the length of side of polygon.

Relations between $R,r,a,\theta$ :

$R^2 = \frac{a^2}{4} + r^2$, $a = 2R*sin(\frac{\theta}{2})$ and $r = R*cos(\frac{\theta}{2})$

We need to prove $2B < A+C$
$\Leftrightarrow \frac{2sin(\theta)}{3+cos(\theta)} < \frac{\pi}{n}$

This can be verified by showing that, inequality is true for $n = 3 $ and LHS decreases faster than RHS.

The method I used for regular polygons isn't applicable to all. There is too much freedom and ambiguity. But I don't have any idea to tackle generalized polygon. Can anybody help me?

$\endgroup$

1 Answer 1

4
$\begingroup$

It's actually simple. If the perimeter of your convex shape is $P$, then your inequality is equivalent to: $$\pi(r^2+R^2) > Pr,$$ where $R$ and $r$ are the circumradius and inradius, respectively. A trivial upper bound on the perimeter is: $$P < 2\pi R$$ where the latter is the length of the circumcircle. But now a simple AM-GM finishes the problem:

$$Pr < 2\pi Rr < \pi(r^2+R^2)$$

$\endgroup$
2
  • 1
    $\begingroup$ ........No words to say. Beautiful proof. I unnecessarily overcomplicated the problem. How stupid of me.. Thanks @dezdichado $\endgroup$ Commented Dec 5, 2020 at 18:21
  • 2
    $\begingroup$ @MathematicalCuriosity no need to call yourself stupid. This is pure experience - I did probably thousands of Euclidean geometry problems in secondary school for math contest and so I happen to know basic ideas, techniques that are applicable in most cases. $\endgroup$ Commented Dec 5, 2020 at 18:24

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.