bob012345 said:
I worked up the general case for integer ##n## overlapping circles. Here is the Desmos page;
The red circle is the generator of the ##n## overlapping circles.
The formula ## f=(\pi/5+\sin72^\circ/2-2\sin72^\circ\sin^236^\circ)/(\pi/5+\sin72^\circ/2)\\ ## from post #6 can be transformed into a general case.
## \begin{align}
f=&(\pi/n+\sin(2\pi/n)/2-2\sin(2\pi/n)\sin^2(\pi/n))/(\pi/n+\sin(2\pi/n)/2)\nonumber\\
=&(\pi/n+\sin(2\pi/n)/2-\sin(2\pi/n)(1-\cos(2\pi/n))/(\pi/n+\sin(2\pi/n)/2)\nonumber\\
=&(\pi/n-\sin(2\pi/n)/2+\sin(4\pi/n)/2)/(\pi/n+\sin(2\pi/n)/2)\nonumber\\
=&(2\pi-n\sin(2\pi/n)+n\sin(4\pi/n))/(2\pi+n\sin(2\pi/n))\nonumber\\
\end{align} ##
The final result is the same as the one in post #8.
The formula produces the fraction ## f=1 ## for ## n=2 ##; therefore, the formula is not valid for ## n=2 ## because the fraction equals ## 0 ## in this case.
Also, the formula produces the fraction ## f=(4\pi-6\sqrt3)/(4\pi+3\sqrt3) ## for ## n=3 ##; therefore, the formula is not valid for ## n=3 ## because the fraction equals ## (2\pi-3\sqrt3)/(4\pi+3\sqrt3) ## in this case.
The formula holds for ## n\ge4 ##.
It can be shown that the fraction approaches ## 1 ## as ## n ## approaches infinity, which is obvious.