Re: Scaling question

[rtickle@xxxxxxxx]

> (1). Are you scaling down the whole magnet, including the core
width, core height, pole width, pole gap, Ampere-turns in the coil
etc?

The entire design drawing was reduced to 1/4 scale - core width and
length, air gap, coil wire diameter. The amp-turns are a different
issue.
The number of turns are the same, only the wire diameter has been
scaled. By keeping the current density fixed in the first
calculation, I was effectively scaling the current down by a factor
of (1/4)^2, due to the change in the wire cross-sectional area.. I
think this was the problem.
Multiplying the current density by 4 in the 2nd calculation had the
effect of using the original full-scale current value scaled by
(1/4), i.e. it was like 1/4 scaling the number of amp-turns.
Apparently this is the correct approach, since this gave the same
values for fields along the centerline. I don't have an intuitive
sense of why this is the right parameter to scale, though, which is
what I would like to understand.

> (2). Generally while making the quartre scale(or some other scaled)
> prototype of a magnet only the length of the magnet is scaled down.

Not sure what you mean by this. Are you saying amp-turns should not
be scaled down? A smaller magnet with the same number of amp-turns
would be driven more closely to saturation generally, producing
larger fields (in a smaller air gap volume). I am interested in a
prototype which produces the same field magnitudes and distribution
within the proportionally smaller volume of the scaled airgap as the
full scale would have in its airgap.

> (3) If you are even scaling down the coil parameters by 1/4, then
you are bound to get 1/4 fields as can be seen even by simply using
the formula for field of a dipole magnet.

Not sure what you mean by this either. If I look at a current loop,
the field at the center is k*i/R. If I scale down both the current
and the radius, I get the same field at the center...