EM Loss Not Importing to Icepak in AEDT

I have set up and run a 3d transient magnetics model in Maxwell, and assigned losses for eddy currents and electrical steel. I can plot the losses in time, and they are significant.

When I set up the EM losses in Icepak, the simulation just settles to ambient temperature. I know this is incorrect because manually assigning a loss source equal to the copper losses results in a significant temperature rise. In addition, I have simulated a similar machine in the past, and I know approximately what the final temperature should be.

Moreover, I have a previous model that I simulated last spring, and that doesn't work anymore, either. It's like the EM loss isn't being properly imported.

If I go and delete all the Maxwell results, and tag "simulate the source design as needed", then the Icepak model just runs as if it already has results. In the past, I have watched Icepak run the Maxwell simulation, so I know that should work.

I also tried the following, with no luck:

  1. Set up a really simple Maxwell model, just a loop of wire with no core
  2. Assign current and verify that there are losses
  3. Set up an Icepak model and copy the coil into the Icepak model
  4. Set up EM loss
  5. Run the Icepak model

The result is the same: 20° (ambient) all around. If I manually assign the losses as a source then the coil hits 1750°C. If I then go back to EM losses, back to 20°.

Finally, whether or not there are Maxwell results, if I try to right-click Analysis, Clear Link Data is ghosted.

Has anyone else had this problem? I'm really not sure where to start troubleshooting. How would one start troubleshooting where the link between the Icepak and Maxwell simulations is getting lost?



  • satyasatya Member

    What version of Icepak are you using? The 1750 C temperature obtained after assigning the losses manually is very high. If you are using the EMtoIcepak ACT, I would recommend starting with a manual coupling approach instead. Use the example provided with the software that can be found under File-> Open Examples.

  • The 1750° temperature is right. I put 500W into a 25mm diameter air-cooled loop, so obviously that's ludicrous. I just wanted to see if the solver was working. I'm using Icepak within AEDT, version 2020R2.

    I feel like an idiot...I didn't even know there pre-built examples for the EM/Icepak link. Regardless, coil_plate.aedt doesn't work, either, and the behavior has changed over the past couple days so I'm pretty sure this is a problem with our installation at the university.


Sign In or Register to comment.