Does it sound intuitively dubious to waste good mech. energy to increase friction to increase heat transfer to increase the thermo efficiency of an engine?
The numbers say it won't work, at least not for a spinning cylinder heat transfer situation.
The thing that must be much less than one is the (heat recovered by the regenerator / net power output) X (Clausius number X the Coefficient of friction / Nusselt Number)
The U^3 term doesn't help. Things just keep getting better for lower and lower rotational speeds.
It was a dumb idea to begin with.
Bret Cahill