Circle Inversion Fractals

Inversion Limit Sets

Convergence to the Limit Set

Suppose the circles C1, ... , CN bound discs D1, ... , DN having disjoint interiors.

Starting with a point z0 outside the discs, we pick a circle Ci1 at random and invert z0, obtaining a new point

z1 = Ii1(z0).

Then we pick another circle Ci2 at random, and invert z1, obtaining a new point

z2 = Ii2(z1) = Ii2(Ii1(z0)).

Continue in this way,

zk = Iik(zk-1) = Iik(Iik-1(...Ii1(z0)...)).

with the restriction that no ij equals ij+1.

Sketch

The limit set as the nested intersection, recalling the Cantor set construction, and the standard deterministic IFS for the gasket.

Because the inverting circles are disjoint, the radii converge to 0.

A standard adderss argument shows the limit set of any orbit is identical with the intersection of this nested sequence.

Return to circle inversion fractals.