Elmer Discussion Forum

Posted: **26 Aug 2019, 03:35**

Hello,
Just trying to do a very simple static current conduction problem to make sure I have things working properly, but I'm not getting quite what I expect.
I have a rectangular copper bar with dimensions of 1"x1"x6" and want to calculate the voltage drop when 1.0 amps of current is flowing through it.

The analytical solution from R = rho*L/A and V=IR is about 4e-06 volts, but the simulation shows 2.6e-06 volts. Any suggestions on what I might try to get a simulation that matches the analytical sol'n? I' tried playing with mesh size and convergence settings but this hasn't worked yet. Thanks in advance!
Best,
Steve

Posted: **26 Aug 2019, 11:45**

Looks like you loaded the material properties from the library. The default stored values in elmer are in SI units. May want to check units for consistency.

Posted: **04 Sep 2019, 23:52**

Hi Steve,
How about putting a neumann boundary condition at inlet (current density=1.00) and a dirichlet boundary condition (e.g. ground voltage , potential =0) at the outlet? For laplacian of potential, an inlet current of 1 means , outlet of 1 , owing to the principle of continuity of charge.

Please add current density BC = true , when using the neumann boundary condition.
The logical function will activate the current density boundary condition.

Code: Select all

Boundary Condition 1
  Target Boundaries(1) = 5 
  Name = "current_in"
  Current Density BC = Logical True !!!!!Please add this line to the boundary condition
  Current Density = 1.000
End

Boundary Condition 2
  Target Boundaries(1) = 6 
 ! Name = "current_out"
 ! Current Density = -1.000
 Potential = 0.0
End

Boundary Condition 3
  Target Boundaries(4) = 4 1 3 2 
  Name = "current_0"
  Current Density BC = Logical True !!!!!Please add this line to the boundary condition
  Current Density = 0
End

Yours Sincerely,
Anil Kunwar

Elmer Discussion Forum

statcurrent benchmark problem

statcurrent benchmark problem

Re: statcurrent benchmark problem

Re: statcurrent benchmark problem