Re: [femm] Neumann boundary condition

[dcm3c@xxxxxxx]

In a message dated 11/14/01 3:40:32 AM Eastern Standard Time, sh9988@xxxxxxxxxxxxx writes:


> Dear David and All,
>
>   I have encountered a mixed boundary condition (see below) along the 
> outer boundary of domain in a 2D time harmonic simulation.
>
>    Ht = Jm * exp(-jPhi * x)
>
>    Where Jm is the line density of Current, Phi is the phase angle 
> and it is a function of x-coordinate. 
>
>   This condition can't be set directly to the boundary with the Mixed 
> boundary condition in FEMM. The only way I have thought is to split 
> the boundary into many small segments. Then manually calculate the 
> values on each segment's midpoint, set each segment's condition with 
> its midpoint's value. 
>
>   Apparently, it's not a good way and cause a lot of work. Is there a 
> better way or we can add the feature that support Mixed boundary 
> condition with a function of coordinates(like the Prescribed A 
> condition)?
>
> Best regards
>
> Si hang


This is probably something that would be good to include as an extra boundary condition type or something.  I end up wanting to do this sort of thing occasionally as well (especially to sanity-check analytical results that I come up with), but I haven't gotten around to writing it in yet.  For the moment, a number of different boundary conditions associated with different current sheet amplitudes is the only way to go.

Dave. 
--
David Meeker
http://members.aol.com/_ht_a/dcm3c