Discussion Forum

Hi

I'm also in trouble with spontaneous crashes today, I suspect a java upate from yesterday that has interferred with the one of COMSOL. Now I even cannot relaunch COMSOL further than the splash ;)

Time to update, the conference in Stuttgart is starting tomorrow (still have to finish todays work first)
--
Good luck :(
Ivar

Hi Ivar,
I'm sad to see that you have such a problem as well. However, my program starts, and it does run other models and work fine with them. In fact I am now working on a different set of coils in another save file and it works perfectly. It seems to be that specific model - I suspected that the complex Geometry - two helixes - might be the cause. But then again, it is only my thinking, unbased.

If anything comes up, a solution, update or any info at all - please update me :)
Thanks, and hopefully it will be solved for us both,

Shoval

Hi

then you have another issue, I'm rather sure mine is linked to Java, just now I need to reboot for each launch of COMSOL. Single shot computing, I had the same trouble back in June or July, a few days, then it disappeared ;)

Computing ... Will eventually find out .

Now its time to run to catch the train, Conference time, perhaps we meet in in Stuttgart ? ;)

--
Good luck
Ivar

Well no, we probably won't meet, I don't have the funds to get there from here :)
However, if you solve your problem, can you try and have a look at the model I attached a few posts above? It still makes my comsol 4.2 crash, and other models work perfectly. I really need that model :(

Thanks!
Shoval

How long does it take to rebuild the model?
I had files that got corrupted and I had to rebuild.
I tried opening your file and my COMSOL is stuck and my computer is trying hard and I don't know what it is doing though. All I know is, one CPU is at its full capacity and the other one is almost there. My RAM is at 60% usage. I only have 4GB RAM so I am not sure if this is a memory issue or if it is a corrupt file.

Goodluck!
Sirisha

I have already thought about that. I have rebuilt the model twice and recieved the same results before posting here. Reconstructing the geometry really isn't so bad, it's two identical helixes, so it's Helix+copy+rotate and you're done. My Mac always got stuck when I right-clicked the physics node, trying to add a Ground/Terminal.

Thank you for your response tho. Hopefully by tomorrow something will come up.

Shoval

Does it work when you have only one helix? I mean does it still freeze if you have only one helix?

Sirisha

In the model file you posted, the "Form Union" node of the Geometry sequence reports an error: "Could not resolve domain relations." You did not mention that in your original post.

When you click on any node outside the geometry sequence, such as the Electric Currents node, Comsol automatically analyzes the geometry as defined, trying to deduce domain relations. That's when it numbers the domains, boundaries, etc. You can manually trigger this analysis by executing the "Form Union" node. The error message means that in your case Comsol has failed. It doesn't understand your geometry.

And neither do I. The helix you've defined has lots of overlapping volumes. I don't know why you chose that value for "radial pitch", but it seems unphysical to me. But maybe I'm missing something. In any case, it's no wonder Comsol takes forever to analyse that geometry...

Thank you john for showing me my problem :) I probably measured my existing coil incorrectly; because I needed for some reason to have 17 laps of 1mm of diameter each, inserted into a 15mm space. Definitely impossible! I switched some numbers there. Now it works fine.

I also thank you very much for the explanation of the Form Union - this is something I did wonder about at times, now I know :)

Problem solved!

hi , im trying also to simulate coils , and my comsol also gets stuck or takes for ever to analyse the geometry... if you have one model working , can you please attatch it so i can see what im doing wrong.. would help me a lot

thank you very much =)