Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Skin Effect and Specified Total Current through Global Constraint

Please login with a confirmed email address before reporting spam

Hi,

I'm trying to determine the steady state temperature of an XLPE insulated copper cable carrying a specified current using a 2D model. I've got the model working for DC, by first modelling the natural convective cooling of the cable, and then using the calculated heat transfer coefficient in a second model that uses the Heat Transfer (ht) and Magnetic and Electric Field (mef) interfaces. The final temperature for the DC case agrees with predictions.

To expand the model for AC, I have made the following changes:

1) Changed the equation form in mef from study controlled to frequency domain. I then specify the frequency after selecting user defined.

2) Under Model > Definitions, created an integration operator (intop1) on the domain that represents the copper area of the cable.

3) Under mef, replaced the external current density with a global constraint "intop1(mef.Jz)-500". I am only modelling a quarter of the cable, so the 500 represents a quarter of the total cable current of 2000A.

At very low frequencies, the AC model looks like the DC case, but a I increase the frequency the results become less and less like what I would expect. I've attached images showing the current density distribution in the copper at 5Hz. When I check the total current through the cable after the simulation, the global constraint does appear to be working with regards to the total current, but the distribution does not show the skin effect, but just some kind of simulation artefacts.

If anyone can suggest what I am doing wrong, it would be a great help.

Cheers,

Mark


3 Replies Last Post 2015年10月9日 GMT-4 03:03

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 2015年7月1日 GMT-4 15:43
I am not sure how much help this can be, but the skin depth is frequency dependent. In this case the skin depth is in the centimeter range which means that currents may be inside the conductor if it is large enough.
I am not sure how much help this can be, but the skin depth is frequency dependent. In this case the skin depth is in the centimeter range which means that currents may be inside the conductor if it is large enough.

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 2015年7月3日 GMT-4 06:29
After much trial and error I managed to solve this problem, so I'll briefly outline my solution below in case anyone else has the same problem.

1) The rather strange solution seemed to be caused by some kind of corruption in the file, as I was unable to make any progress until I created a new file and started from scratch.

2) Add the Magnetic Fields (mf), Heat Transfer (ht), and Global ODEs and DAEs (ge) physics.

3) Create a frequency study and deselect the ht physics. Then add a stationary study and deselect the mf and ge physics.

4) Show the default solver and add Other > Store Solution after Stationary Solver 1: this is needed to get the results from mf, such as current density.

5) Under Global Definitions > Parameters add I_target 2000A.

6) Under Model 1 > Definitions create an integration operator for the conductor cross section.

7) Under ge > Global Equations 1 add I_ext intop1(mf.Jz)-I_target.

8) Add mf > External Current Density and set the z component of Je to I_ext/intop1(1).

Everything else was as before, and this seems to be working perfectly.

I hope this is useful to someone else.
After much trial and error I managed to solve this problem, so I'll briefly outline my solution below in case anyone else has the same problem. 1) The rather strange solution seemed to be caused by some kind of corruption in the file, as I was unable to make any progress until I created a new file and started from scratch. 2) Add the Magnetic Fields (mf), Heat Transfer (ht), and Global ODEs and DAEs (ge) physics. 3) Create a frequency study and deselect the ht physics. Then add a stationary study and deselect the mf and ge physics. 4) Show the default solver and add Other > Store Solution after Stationary Solver 1: this is needed to get the results from mf, such as current density. 5) Under Global Definitions > Parameters add I_target 2000A. 6) Under Model 1 > Definitions create an integration operator for the conductor cross section. 7) Under ge > Global Equations 1 add I_ext intop1(mf.Jz)-I_target. 8) Add mf > External Current Density and set the z component of Je to I_ext/intop1(1). Everything else was as before, and this seems to be working perfectly. I hope this is useful to someone else.

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 2015年10月9日 GMT-4 03:03
Hi Todd

Could u upload the file of the model?
I'm studying skin effect too, but 3d model for me.

Xiaoxuan
Hi Todd Could u upload the file of the model? I'm studying skin effect too, but 3d model for me. Xiaoxuan

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.