operator jay science forum Guru Wannabe
Joined: 29 Apr 2005
Posts: 142

Posted: Thu Jun 09, 2005 5:13 pm Post subject:
Re: inductance vs skin effect



"Francois Grieu" <fgrieu@francenet.fr> wrote in message
news:fgrieu605EBB.17592906062005@news5e.proxad.net...
Quote:  In article <DQWoe.2753$mN.31518@news1.mts.net>,
"operator jay" <none@none.none> wrote:
y =
e^(x) , 0<=x<=1
e^(x2) , 1<=x<=2
This would show a symmetrical current density profile in a finitely
thick
conductor of rectangular section
but there is a sharp discontinuity (change of sign) of the derivative
at x=1, when there is physically nothing special at that point.
That's why I suggest
e^(x) + e^(x2), 0<=x<=2
then I renormalize by integrating, just the way you do.
This gives nearly the same result for a conductor that is either
much thicker or much thinner than the skin depth, but a very
different result around the skin depth.
François Grieu

Adding the extra term bothers me. What do you do for a (square) four sided
section? Use four terms? Eight sided? Use eight terms? Circular?
I agree that discontinuity you spotted does seem out of place, but, there it
is in a round conductor too, I think, and the approximation (if that's what
it is) is still used.
Of course, this is just food for thought. Once again I really can't say
anything of certainty. Maybe you'd have to do the calcs all the way through
with differential equations to get the actual distribution.
j 
