Batch mode aborting at error messages

[thiesblaak]

Dear all,

I'm trying to run quite a complex nonlinear analysis using pymapdl in batch mode. I'm using a shell element which inherently attracts excessive distortion errors and excessive thickness errors. A few errors are not a big problem for my results. When running a single job in the APDL command prompt, I can suppress any errors that pop up using the /NERR command. I tried doing the same in the pymapdl module using the mapdl.nerr(abort='-1') command, in order to ignore any and all errors. The problem I'm having is that for some reason, my batch run still gets aborted when an error occurs. Does anyone have suggestions on an alternative fix for my issue? Or am I misusing the mapdl.nerr() command? I placed the command right after my launch_mapdl(), mapdl.clear() and mapdl.finish() commands.

With kind regards,
Thies

[mikerife]

Hi @thiesblaak
Please take care with the terminology, this gets a little tricky. PyMAPDL starts, or connects to, an instance of MAPDL running as a server. Not in batch or interactive - the two choices we have traditionally had. The as a server (aaS) is neither batch nor interactive but a combination of the two with uniqueness thown in as well.

A second point is that the /nerr command is already being used behind the scenes when PyMAPDL starts up MAPDL aaS. Just a FYI.

Instead of /NERR you may want to check out this mapdl method instead that sets the behavior to ignore errors.

Lastly please take care using this. I find when I am developing an PyMAPDL script not ignoring the errors to be more useful than ignoring them. But maybe that is just me...would be interested in hearing others experience on the matter.

Mike

[thiesblaak]

Dear Mike,

My apologies for the erroneous terminology. I'm quite new to Ansys. Regarding the method you recommended me, is there a specific place in the chronology of my script where this command has to be placed? I recall trying to use it before, but without any success.

Thies

[thiesblaak]

Also, it seems as if the error pop ups might not be the issue to why my simulation is aborting. I used the mapdl.list_error_file() function to see what was going wrong. It gave the following error:

"Element 1507 (type = 1, SHELL281) (and maybe other elements) has become highly distorted. Excessive distortion of elements is usually a symptom indicating the need for corrective action elsewhere. Try incrementing the load more slowly (increase the number of substeps or decrease the time step size). You may need to improve your mesh to obtain elements with better aspect ratios. Also consider the behavior of materials, contact pairs, and/or constraint equations. Please rule \n out other root causes of this failure before attempting rezoning or nonlinear adaptive solutions. If this message appears in the first iteration of first substep, be sure to perform element shape checking. *** WARNING *** SUPPRESSED MESSAGE CP = 258.719 TIME= 11:17:53 The unconverged solution (identified as time 1 substep 999999) is output for analysis debug purposes. Results should not be used for any other purpose."

The thing which confuses me is that this distortion is not an issue for generating a result in my originial MAPDL script. So why would it make this simulation crash?

[thiesblaak]

Dear @mikerife,

To continue spitballing on my issues. I found that the reason why my solution does converge in apdl, but not in pymapdl comes from the fact that when my solution fails in apdl, it automatically restarts from the previous load step. I'm not sure why this works, but is there perhaps a similar functionality in pymapdl?

With kind regards,
Thies

[germa89]

The automatically restart thing, I'm not sure how to address that. From my understanding, both PyMAPDL and MAPDL should behave similar. But because of the difference between Batch/Interactive and AaS, there might be some differences. Probably @mikerife knows better.

Regarding the mapdl.ignore_errors, you can use it as follows:

mapdl.ignore_errors = True
mapdl.solve()
mapdl.ignore_errors = False

You could set it from the very beginning if you wish.