[femm] Re: Drawing circles & Another AC analysis question

[Dcm3c@xxxxxxx]

In a message dated 11/11/99 4:50:09 PM Eastern Standard Time, Marctt2@xxxxxxx 
writes:

> 
> << Now, inductance is:
> L=4*E/abs(i)^2
> The 4 gets in there (rather than a 2, like in the mstatic case) because 
we 
> are working in terms of amplitude, rather than rms quantities. Likewise, 
> resistance is: >>
> 
> I'm assuming that E is calculated over the whole problem area (i.e. over 
the 
> whole FEA region ?

Yes, it is calculated over the whole problem area.

> Also, when I do the integral over my coil volume to get total current, I 
get
> Total = Real + j(Imag) since there is a phase shift and amplitude change. 
> I'm assuming that abs(i)^2 = Real^2 + Imag^2, or the amplitude^2 of the 
> current. ?

Yes, that's right: abs(i)^2 = Real^2 + Imag^2

> Now, let's take the problem to one more level of difficulty:
> 
> I have a multi-turn coil .... I would like to measure AC inductance and 
> resistance of the coil.
> 
> Is there a way to insure that the individual loops have the same net 
current 
> in them ? If not, how do you calculate the inductance given that the 
> different loops will have different net currents (due to proximity 
effect)? 
> 
> Is there a smart way to add up the individual Em = 0.5LI^2 terms ?
> 
> In the problem that I'm working on, I have a 28 turn inductor that I would 
> like to figure out the AC resistance and the AC inductance in the DC-1 MHz 
> range. I'm thinking that perhaps I need to work out a 
> including self and mutual inductances .... perhaps there's an easier way.

Like I was saying before, I've been working on adding the options of applying 
this sort of constraints on harmonic problems, but you can't really solve 
this problem directly now. I wonder if there is a continuum way to model a 
coil with skin and proximity effects, perhaps by solving a second 
differential equation for the voltage gradient that is required to enforce 
constant current in the coils in the presence of skin and proximity effects. 
Anyone know of a reference?

I'd have to think about the about the 28x28 matrix approach. It seems like 
the current distribution that you induce via proximity effects is different 
than that induced by the self-inductance part, so this approach might only be 
approximate.

Dave.
--
http://members.aol.com/dcm3c