Re: Inductance calculations

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Antonio Carlos M. de Queiroz by way of Terry Fritz <teslalist-at-qwest-dot-net>" <acmq-at-compuland-dot-com.br>

Tesla list wrote:
 >
 > Original poster: "Peter Lawrence by way of Terry Fritz 
<teslalist-at-qwest-dot-net>" <Peter.Lawrence-at-Sun.COM>
 >
 > Antonio,
 >          I've been sort of following your recent posts, are you working on
 > a corner case where the spacing increment per turn is large compared to the
 > wire diameter and/or coil diameter.

Certainly a corner case. The methods differ for quite weird coils.

 > I would expect "formulas" like Wheeler, Lundin to be off, but why would the
 > Finite Element Analysis methods in ACMI, etc not be correct?

The problem appears to be in some detail about the wire diameter.
The values given by Acmi, Fantc, and Mandk appear to consider too
thick wire. I have not examined the methods in Grover's book in enough
detail to see what may be wrong, but it appears to follow Maxwell.
Looking at the code in Acmi, it appears that the program computes
inductances as mutual inductances between two closely spaced coils,
what is correct, but it's not clear what is the separation used
in all cases (half of the pitch of the winding? Paul?).
My program uses:
spacing =  radius of the wire * exp(-0.25)
that is what Maxwell calculated for round wires with uniform current.
A test:
The coil in the Fantc example:
5 turns
height = 0.2
radius = 0.5
wire diameter = .002
Acmi and Fantc give 39.35 uH. Inca gives 46.43 uH
(Wheeler: 37.93, Lundin and Lorenz: 39.47, Snow: 46.38)
The mutual inductance between two of these coils separated by
a vertical distance of 0.2/5/2=0.02 Inca calculates as 37.13 uH,
or 37.12 uH, more exactly.
With spacing of 0.001*0.7788=0.007788 Inca calculates the 46.43 uH.
So, I think that the problem is in spacing between the windings
in the calculation of inductance as a mutual inductance between
closely spaced coils.

I have implemented Lorenz formula in the program too. Identical
to Lundin, and fails for thin wire too.

Antonio Carlos M. de Queiroz