pull down to refresh
For
O_2
(1 - \frac{1}{2} - r)^2 + 1 = 1 + 2*r + r^2
For
O_3
(1 - \frac{1}{2} - \frac{1}{6} - r)^2 + 1 = 1 + 2*r + r^2
(\frac{6}{6} - \frac{3}{6} - \frac{1}{6} - r)^2 + 1 = 1 + 2*r + r^2
(\frac{1}{3} - r)^2 + 1 = 1 + 2*r + r^2
\frac{1}{9} - \frac{2}{3}*r + r^2 + 1 = 1 + 2*r + r^2
\frac{1}{9} - \frac{2}{3}*r = 2*r
\frac{1}{9} = \frac{8}{3}*r
\frac{3}{9*8} = r
\frac{1}{24} = r
Diameter is twice that, so
\frac{1}{12}
No?
reply
That would give diameters
\frac{1}{2}, \frac{1}{6}, \frac{1}{12}, \frac{1}{20}, \frac{1}{30}, ...
For the sequence of divisors there are multiple candidates
reply
This is a great website. Didn't know about it. Made me check out the infamous https://oeis.org/A000127 sequence, for which there seem to be many more cases showing this unexpected change from 32 to 31.
reply
Aren't you using the same symbol
r
to denote the different radii of O_1
(as in the drawing) and O_2
(as in the right-hand side of the formula below)? And same comment for the subsequent O_3
equation?ForO_2
(1 - \frac{1}{2} - r)^2 + 1 = 1 + 2*r + r^2
ForO_3
(1 - \frac{1}{2} - \frac{1}{6} - r)^2 + 1 = 1 + 2*r + r^2
I used a different approach to solve this problem (I'll post it as a solution with the next iteration), but that one gives me
1/18
for O_3
.reply
For
O_3

(the top-left corner of the triangle is supposed to be the center of
O
)The hypotenuse is 1 +
r
, where r
denotes the radius of O_3
.The horizontal leg is simply 1 (this is true for all
O_n
).The vertical leg is
1
minus the sum of all diameters O_1
and O_2
minus r
of O_3
.So, Pythagoras gives
r
of O_3
.reply
Ok got it. This seems correct. I'll have to figure out why my approach is giving a different result. I likely must have applied it incorrectly. Please don't spend more time on this, I'm assuming the error must be on my side at this point.
reply
You're right, I incorrectly applied Descartes. 1/12 is correct. My bad...
reply
Oh, neat, I didn't know about Descartes' theorem!
So you're calculating
k_4 = k_1 + k_2 + k_3 + 2*\sqrt{k_1*k_2 + k_2*k_3 + k_3*k_1}
where curvature
k_n
is \frac{1}{r_n}
(the inverse of the radius).If I pick
O
, O'
and O_1
to calculate O_2
1 + 1 + 4 + 2*\sqrt{1*1 + 1*4 + 4*1} = 12
Now we can simply repeat with
O
, O'
and O_2
to calculate O_3
1 + 1 + 12 + 2*\sqrt{1*1 + 1*12 + 12*1} = 24
So radius
\frac{1}{24}
or diameter \frac{1}{12}
reply
And I guess a direct formula for circle
O_n
could bed(n) = \frac{1}{n(n+1)}.
Yes, exactly.
Related to Apollonian gaskets
a fractal generated by starting with a triple of circles, each tangent to the other two, and successively filling in more circles, each tangent to another three.
I'll check my notes tomorrow.
reply
O_3
though.