Has anyone encountered the issue where cases from version 8.1 do not converge in version 8.2?
Thread Content
It worked fine with version 8.1; when using version 8.2, the computation does not converge, and this unconvergent state can persist for a long time – it takes half a minute to stop even after pressing stop: L
You can first export the .inp file under 8.1 and then import it under 8.2; this might solve the problem
Guiding it this way is useless. I don’t know why.
Send me the input file and I’ll give it a try; I’m using version 8.2 and haven’t encountered this issue before.
It’s not convenient to upload the original file; I’ll create an example addressing this issue and upload it when I have time. The basic situation is that there is a circulating flow in the middle, and convergence can be achieved without requiring initial values at 8.1. It failed to converge at 8.2; eventually, convergence was achieved by using the result obtained from 8.1 as the initial value. Slightly worse initial values also failed to lead to convergence. The problem is, when using 8.2 alone, where can one find such a precise initial value? :L
Be careful about this issue in the future; thanks to the original poster for sharing their experience.
Can the guy still find this 8.1 case? I’d like to give it a try
This is normal; for example, if I use 8.3, it converges, but with 9.0 it does not. Just adjust the settings again, especially the circulation backflow, and that will do. The higher the version, the greater the precision, and some rules also differ.