error information of HFSS

[DrunkBear]

"Recovering discrete solution, process abc3d died unexpectedly"
This error jumped several mins after it started calculating and it seemed the system had already finished a few "passes".
What's more strange is the simulation went well in my classmate's computer--but i cannot get his software immediately.
p.s This HFSS modal was imported with a .SAT converted by a CST .mod, namely, the modal was originally made in CST.

 

[krmkr]

Well, I met this kind of errors several times. It's unfortunately very hard to estimate what the real problem is.

Let me suggest you to add a "surface approximation" to the circular/cylindrical/spherical objects. This is how I overcame a "Recovering discrete solution" situation...

Let me know the result

 

[DrunkBear]

Thank you for your reply!
I'm studying on the help file to see what "surface approximation" is and try to fix it .
But don't you feel it extramely strange that it performed differently in different PCs?
I'll inform you as soon as i get the latest news and hope to continue the discussion.
Thank you!

 

[DrunkBear]

Well, I met this kind of errors several times. It's unfortunately very hard to estimate what the real problem is.

Let me suggest you to add a "surface approximation" to the circular/cylindrical/spherical objects. This is how I overcame a "Recovering discrete solution" situation...

Let me know the result

I changed the "Maximum Aspect Ratio" in the "surface approximation" and it did give me the result after about one and a half day(30+ hours), though the result is a bit different from my expectation and needed to be further studied.
But what i'm wondering is, there are only 4 small vias which are of cylindrical shape and they should not have had such a great influence on the simulation.
In addition, the "maximum aspect ratio" was changed to be 3, other than the default 10 and it added a lot of simulation duration. How could i determine this value more proparly?