Re: [FEMM] Doubt in BELA

[David Meeker <dmeeker@xxxxxxxx>]

nicholas lim wrote:

> Dear Meeker,
>
> i found some doubt in BELA, so i hope u can explain to
> me. The doubltful that i found is in the femmplot the
> electric field intensity is (V/m)and how we define the
> distance(m) which is using to calculate the E.
>
> thank you.
>
> Regard,
> Nicholas
>  
Electric field intensity is defined as:

E = - grad(V) 

The solver solves for V, so the program has to take derivatives of V to 
get E.  There isn't any one distance that is used to calculate E.  To 
get a smooth E field (e.g. for plotting purposes), the program obtains 
nodal values of E for each element and interpolates linearly between the 
nodal values to get the E field at other points inside the element.  
Nodal values of E are determined by identifying a best fit plane that 
most closely (in a least-squares sense) passes through the node of 
interest and all other nodes that are neighbors of that node (i.e. all 
nodes that are in elements that also contain the node of interest).  The 
gradient of that best-fit plane is then taken as the nodal value of E. 
However, the program does subtly different things in the plethora of 
special cases associated with boundaries, corners, interfaces between 
materials, isolated charges or line conductors, etc, etc.

Dave.