Inscribed Sector

Geometry Level 5

What is the maximum area of the right-angle circular sector bounded by the equilateral triangle of side length 1?


This problem is the fun twist of the following inspirations: (1) , (2) , (3) .
π 24 \dfrac{\pi}{24} π 6 \dfrac{\pi}{6} 3 π 32 ( 2 3 ) \dfrac{3\pi}{32}(2 - \sqrt{3}) None of the rest 3 π 16 ( 2 3 ) \dfrac{3\pi}{16}(2 - \sqrt{3}) 3 π 8 ( 2 3 ) \dfrac{3\pi}{8}(2 - \sqrt{3}) π 12 \dfrac{\pi}{12}

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.

1 solution

Michael Huang
Dec 29, 2016

There are few approaches for this problem.


Method 1: Large Equilateral Triangle Method


Consider a semicircle of radius r r and maximum area enclosed by the equilateral triangle.

Figure 1. Large equilateral triangle of the small triangle Figure 1. Large equilateral triangle of the small triangle

Since the semicircle bounded by the equilateral triangle has the maximum area, we can slice Δ A E I \Delta AEI with F J \overline{FJ} to obtain the smaller equilateral triangle and color triangle Δ F G K \Delta FGK to remove the unneeded area. The side length to decrease for Δ A E I \Delta AEI is determined by the side length of Δ B C K \Delta BCK , which is an equilateral triangle. This guarantees that the right-angle circular sector has the maximum area.

Observe that the quadrilateral C F G K CFGK is a rectangle, consisting of two 30 30 - 60 60 - 90 90 triangles. Since G K = C F = r |GK| = |CF| = r , then C K = F G = r 3 |CK| = |FG| = \dfrac{r}{\sqrt{3}} . Since Δ B C K \Delta BCK is equilateral, C K = B K = C B = r 3 |CK| = |BK| = |CB| = \dfrac{r}{\sqrt{3}} , which shows that to obtain the smaller triangle we decrease the side length by r 3 \dfrac{r}{\sqrt{3}} .

Notice that Δ C E F \Delta CEF and Δ F G K \Delta FGK since C F A G \overline{CF} \parallel \overline{AG} , C K E I \overline{CK} \parallel \overline{EI} and F K A E \overline{FK} \parallel \overline{AE} . Therefore, E G = 2 r 3 |EG| = \dfrac{2r}{\sqrt{3}} . Since Δ A E I \Delta AEI is symmetric about A G \overline{AG} , E G = G I = 2 r 3 |EG| = |GI| = \dfrac{2r}{\sqrt{3}} , which implies that E I = 4 r 3 |EI| = \dfrac{4r}{\sqrt{3}} . So by decreasing each side length by r 3 \dfrac{r}{\sqrt{3}} , we obtain 3 r 3 \dfrac{3r}{\sqrt{3}} .

Thus, for the problem, radius of circle = r = 3 3 \text{radius of circle} = r = \dfrac{\sqrt{3}}{3} which gives Area of sector = π 4 r 2 = π 12 \text{Area of sector} = \dfrac{\pi}{4} r^2 = \boxed{\dfrac{\pi}{12}}


Method 2: Direct Geometric Approach


Figure 2. Starting equilateral triangle Figure 2. Starting equilateral triangle

This problem can also be solved without considering the large equilateral triangle. For the sector to have maximum area, bound its side and vertices on the triangle's segments as illustrated above.

Because sector B D E BDE is cyclic and tangent to A F \overline{AF} , this forces D G F = 9 0 \angle DGF = 90^{\circ} . But since B D C F \overline{BD} \perp \overline{CF} at point D D , C D B = D G F = 9 0 \angle CDB = \angle DGF = 90^{\circ} . Noticing that Δ C D B \Delta CDB and Δ D F G \Delta DFG share common angles of 6 0 60^{\circ} as Δ A C F \Delta ACF , and that B D = D G = r |BD| = |DG| = r , by Angle-Angle-Side Postulate , this shows that Δ D F G \Delta DFG and Δ C D B \Delta CDB are the same triangles of same lengths and angles.

In that case, C D = G F = r 3 |CD| = |GF| = \dfrac{r}{\sqrt{3}} and B C = D F = 2 r 3 |BC| = |DF| = \dfrac{2r}{\sqrt{3}} , which implies that C F = 3 r 3 |CF| = \dfrac{3r}{\sqrt{3}} . We achieve the same result as we did by Method 1 .

Hm, it doesn't feel like your approaches are rigorous. In the linked problems, the reasoning is that "A in B is maximized if and only if X in Y is maximized", and there was a very simple / direct way to go from A in B to X in Y. You have not provided that connection quite as yet. IE In the first approach, how does "quarter circle in triangle" lead to a "semi circle in (larger) triangle"?

Calvin Lin Staff - 4 years, 5 months ago

Log in to reply

Ah. Seems to me that you provided the good point. :)

Let me see:

I have already proven that "semi circle in larger triangle" leads to "quarter circle in triangle". I need to prove that "quarter circle in a smaller triangle" leads to "semi circle in larger triangle".

Assume that the quarter circle in a smaller triangle has the maximum area. Extend the arc about the origin formed by the two sides of the sector. Call the angle of arc extension α \alpha . Then, the measure of the new sector is ( α + 90 ) (\alpha + 90)^{\circ} , where 0 < α 9 0 0 < \alpha \leq 90^{\circ} . But since this forces the sector to be unbounded, we need to bound the segment of the equilateral triangle that the top vertex of the sector touches. So as the arc angle increases, the length of all triangle sides increases.

We see that from the first approach, we decrease the side lengths bounded by the leftmost vertex of the large triangle, such that the arc segment is perpendicular to one of the segments. This direct relationship also holds for increasing the angle of the sector. However, if the angle measurement of the sector is between 15 0 150^{\circ} and 18 0 180^{\circ} , the side length of the equilateral triangle remains constant, which leaves the sector's area either smaller or bigger (depending on how you deal with the arc angle).

It would be interesting to consider many possible cases.

Michael Huang - 4 years, 5 months ago

Log in to reply

Not quite. Your first case is "semi circle in larger triangle with base on one side" leads to "quarter circle in smaller triangle". You have to do it more generally, like for tilted semicircles, while also ensuring that the ratio of larger triangle to smaller triangle is the same throughout. At a quick though, I'm not fully convinced this is possible.

Similar concerns (esp with the ratio) about your "quarter circle to semi circle".

Note: I believe you have the correct numerical answer, but I don't think your solution is rigorous. Even in method 2, you assume that the "best fit" must have a certain form, instead of showing that's indeed the best fit. Avoid arguments of "Clearly we cannot optimize better from this shape", which doesn't work well for such packing questions.

Calvin Lin Staff - 4 years, 5 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...