Fluctuations are substantial, so every run is different.
The cyan curve in the previous pages (reproduced below) is a little atypical,
in that the whole 16kT of work was done unusually fast.
(It was the first one I ran, honest! I wasn't trying
The figure below makes clear that nearly all the
work done is done in a brief big push as the piston location
changes from essentially zero to about 1. The subsequent wandering of
the piston from 1 to 40 generates far less work. The last 2kT or
so of work is rather reversible, it comes and goes.
The upper graphs show the total potential energy of the system (including the two particles
riding up against the springy walls and the piston, as well as the 1-6 potential and
the linear potential).