In article ,
Duane Bozarth wrote:
George wrote:
...
How about this:

The free-wheeling upper wheel gets ahead of the driven, but loaded lower,
causing the blade to bunch into the gap. Doesn't take much difference in
speed to start the process, which then increases in effect as the bunched
part slows....

The upper wheel isn't free-wheeling, it's driven/pulled by the
blade...the blade makes the wheel move, not the other way round...

On the other hand, the upper wheel has rotational inertia. It's an
interesting dynamics problem to figure out exactly what happens if you
get the whole system up to speed and then place drag on the
downward-moving blade.