AtChem/AtChem2

CVODE WARNING t + h = t

lucasce-mmn opened this issue · 11 comments

lucasce-mmn commented

Hey, in a simulation i get the warning:

[CVODE WARNING]  CVode
"Internal t = 38259.2 and h = 3.06996e-12 are such that t + h = t on the next step. The solver will continue anyway.

[CVODE WARNING]  CVode
  The above warning has been issued mxhnil times and will not be issued again for this problem."

I start my model at 36000 seconds. The problem stops when i take out some constraints but constraints that don't work are not much different from the ones that do, any ideas?

rs028 commented

HI @lcsmmn did the model complete its run and just printed out a warning, or did it crash before the end?

lucasce-mmn commented

It's stuck in the warning, no error or crash.

rs028 commented

It looks like a numerical issue, i.e. the system produces numbers that the solver struggles to deal with. You can try to play with the solver.parameters file, or look carefully at the input data (including the constraints) to see which one in particular is causing the problem. Hope this helps.

lucasce-mmn commented

Thanks, it was a error in the constraints file.

liuxiaoyu4545 commented

Thanks, it was a error in the constraints file.

Hello, I also encountered a similar problem. When I added the O3 to the constraints file, this problem occurred. How did you correct the error in the constraints file? I have tried to replace the ozone data and extend the data, but they have not been resolved. I hope to receive your reply. Thanks!

rs028 commented

@liuxiaoyu4545, see my comment above for possible approaches to solve the problem. The timestamp of the warning may give clues as to which section of your constraints dataset is creating the issue.

lucasce-mmn commented

I was editing my files on Windows, then running AtChem in a Linux VM, my problem got solved when i started editing my files also in linux.

liuxiaoyu4545 commented

@liuxiaoyu4545, see my comment above for possible approaches to solve the problem. The timestamp of the warning may give clues as to which section of your constraints dataset is creating the issue.
Thank you for your answer. Before the "model run" appeared, the previous information displayed some warnings,as shown in the figure. Does this have any impact on the subsequent operation?
1

2

liuxiaoyu4545 commented

I was editing my files on Windows, then running AtChem in a Linux VM, my problem got solved when i started editing my files also in linux.

I was editing my files on Windows, then running AtChem in a Linux VM, my problem got solved when i started editing my files also in linux.

Thank you for your reply. My constraint files were processed and edited in the Linux system. Unfortunately, there are still previous issues, which may be related to the solver.

spco commented

Hi @liuxiaoyu4545 the ifort warning you show in those screenshots is not a problem - you can safely ignore that warning.

liuxiaoyu4545 commented

Hi @liuxiaoyu4545 the ifort warning you show in those screenshots is not a problem - you can safely ignore that warning.
Okay, I understand. Thank you very much!