Current in PCB trace -> temperature rising

Numerical methods and mathematical models of Elmer
Post Reply
gclairet
Posts: 4
Joined: 22 Jan 2021, 11:30
Antispam: Yes

Current in PCB trace -> temperature rising

Post by gclairet »

Hi all,

I am trying to simulate the heating due to the current in PCB trace (or a wire). I have used StatCurrentSolve and HeatSolve and it works. But the result comes from the potential. In my case, I need to use the current, what can I do?
I tried to use current density (Neumann condition) but since I want to simulate heating for different geometries of the trace it is not the right way for me.

Basically, I want to set the geometry and current flow, and have Elmer calculate the trace resistance, the power dissipated, and finally the temperature going up for me.
Is there an easy way to achieve this?

Thank you in advance.
When power electronic meets sailing...
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Current in PCB trace -> temperature rising

Post by raback »

Hi

If you know the current density, i, as a field and the electric resistivity, rho, can you not just use "Heat Source" that is "rho*i^2"?

-Peter
gclairet
Posts: 4
Joined: 22 Jan 2021, 11:30
Antispam: Yes

Re: Current in PCB trace -> temperature rising

Post by gclairet »

hi raback,
Thank you for your answer.

I don't know the current density but I can calculate it.
(About the formula, I'm ok with you if you talk about resistance and current : P=R*I^2. But with current density in Ω.m and resistivity in A.m^(-2) the result will be in A^2.Ω.m^(-3) or W.m^(-3), it lacks the dimensions (area and length) in m to get watts.)

Edit: I had a too quick reading of the Elmer Models Manual. It seems HeatSolve use current density and so the unit is W.m^(-3) (viewtopic.php?p=22427#p22427 and Eq. 17.9).

I am a true newbie in FEM and more in Elmer. So maybe I make a BIG mistake.

The final purpose is to interface Elmer with Electronic CAD tool. But first step is to make a simple simulation with geometry, current and material caracteristics.

Maybe I can get Elmer to calculate the current density from the "input" dimensions (the area of the boundary) and a given current. Is there a way to get the boundary area and write a custom formula to the sif file?

Gclairet
Last edited by gclairet on 22 Jan 2021, 15:25, edited 1 time in total.
When power electronic meets sailing...
gclairet
Posts: 4
Joined: 22 Jan 2021, 11:30
Antispam: Yes

Re: Current in PCB trace -> temperature rising

Post by gclairet »

Thank you Kevinarden,
I have already read this topic but it uses potential and I want to use current.
If I have a good understanding of the log, Elmer calculates the effective resistance from geometry and resistivity, and use it with the potential to calculates the dissipated power. I just want reverse the Ohm's law or use current in the joule effect.
When power electronic meets sailing...
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Current in PCB trace -> temperature rising

Post by raback »

Hi

Yes, the unit for heat source should be W.m^(-3) and the rho*i^2 should be consistent with that.

Is the current density formula i=I/A not rather limited to simple geometries? This is the reason why typically we see workflows where the current density is computed from a differential equation.

There is a way to divide the parameter on BC by the area that is applicable to any keyword. You just say,

Code: Select all

Boundary Condition 1
  ...
  Current Density = 1.23
  Current Density Normalize by Area = Logical True
and the primary keyword is automatically normalized by the area on which it is given.

But I guess you would like to compute the area and use it without solving any equation. It would be possible, for example, to write a small UDF that when 1st called would compute the area and later just give the heat source as defined above.

-Peter
gclairet
Posts: 4
Joined: 22 Jan 2021, 11:30
Antispam: Yes

Re: Current in PCB trace -> temperature rising

Post by gclairet »

Hi Peter,

Thank you for the explanation. FEM and 3D modelling is a new world for me.

I will re-read (and understand) the solver manual, learn how to use "Nomalize by area", what is a UDF file and follow your recommendation.

Gclairet.
When power electronic meets sailing...
Post Reply