On 11/13/2011 1:39 AM, Harry Veeder wrote:
Even if it is a "loss", why is one direction better at turning motion
into heat?
Conventional theory predicts the same "loss".
This is a good question because I have heard the same claim - that of an
energy discrepancy which is different if you rotate it in one direction
from the other. Unfortunately we have no information or measurements on
what configuration produces this effect and so we can only guess.
However I am so sure that it is not anomalous, that I can't be bothered
taking the time to think it through properly and prove it. But I will
wave my hands around to show that it is, at least in principle, possible.
I believe the main reason for the energy loss in the case described in
Rice's report, is that, because of the anisotropy of the ferrite in the
rod (or because it is a long thin rod and not a sphere), the field in it
flicks very suddenly from one direction to the opposite direction. I
think that these sharp transitions with overshoot are what is seen in
chart 4251 at about 130 degrees and 150 degrees when the torque is
monitored continuously while being rotated. Remember that in this case
the ferrite is glued hard against the conducting surface of a neo magnet
- which means that these sudden changes of magnetization direction will
induce significant Eddy current losses in the close conducting surface.
This effect is non-linear - in the sense that if you were to cycle the
structure through the same field changes but ensure that it happens
slowly and not by a positive feedback avalanche mechanism, then you
would not loose nearly as much energy as the same cycle with the sudden
switch.
So the question is, is there a method of arranging permanent magnets and
anisotropic ferrite in such a manner that if you cycle it in one
direction, then the ferrite flips its magnetization in an avalanching
manner, whereas if rotated the other direction, then the magnetization
changes smoothly and without sudden flips. I would guess that there is,
although I must admit that it is not obvious how it might be arranged
with a single moving part. It would certainly be easy with two moving
parts. But the effect is certainly possible to achieve in principle,
and I don't want to spend time proving whether it can be done with a
single moving part when we don't even know if that is what has been done.
