Problem with the Finite Element Method applied to Electrostatics

[Alan Lins Alves]

Hi!
I have a code that solve the poisson equation for FEM in temperature problems.
I tested the code for temperature problems and it works!

Now i have to solve an Electrostatic problem.
There is the mesh of my problem (img 01).

At the left side of the mesh we have V=0 (potencial).
There is a line charge located in (0.5,0) with charge density λ=1/(4*pi).

The poisson equation for electrostatics is:

My dificulty is how to solve it for a line charge (what is the best equation for ρ). I used the Dirac delta, and some aproximations for delta function but my code has wrong results.

 

[Alan Lins Alves]

Anyone can help me?
Please

 

[jim hardy]

wow
it's been fifty+ years and the term "Dirac delta" is unfamiliar
and in my day we didn't have computers with precoded solutions .

In slide rule days we'd integrate along the line.

sorry i can't be more help.

old jim

 

[BvU]

I tested the code for temperature problems and it works

Hi,

then how did you specify a pointlike heat source ?

 

[Alan Lins Alves]

wow
it's been fifty+ years and the term "Dirac delta" is unfamiliar
and in my day we didn't have computers with precoded solutions .

In slide rule days we'd integrate along the line.

sorry i can't be more help.

old jim

Ok,
Thanks old Jim!

 

[Alan Lins Alves]

Hi,

then how did you specify a pointlike heat source ?

This is the Poisson equation for temperature problem.
Note that temperature is in left side of equation.

 

[Henryk]

Alan,
I've done a lot of electrostatic calculations, analytical, numerical and FEA. However, I don't quite understand your question.
Could you elaborate on it? Draw a picture without a mesh to start with.
Are you trying to calculate the field of a charged wire next to a ground plane?