Possible improvements that could be made to pymapdl

[germa89]

@germa89 @mikerife This issue suggests possible improvements that could be made to pymapdl:

• an option in pymapdl.solve() to do some validation before calling the SOLVE function on the back end (e.g. make sure that nlgeom is on if there are contact elements)

# suggestion
def solve(self, validation=None):
    if validation:
        self._validate_solve()
    return self.run("SOLVE")

def _validate_solve(self):
    # activate nlgeom if contact is involved.
    pass

• post-processing commands throw a helpful exception if there is no active result set loaded.
• ET commands can write a warning if elements are already defined with the given id

Originally posted by @koubaa in #2123 (comment)

Further additions:
Also add:

• mapdl.port Exposing more internal properties. #2246
• mapdl.exited Exposing more internal properties. #2246
• mapdl.check_status Exposing more internal properties. #2246
• mapdl.ip Exposing more internal properties. #2246
• Avoid mapdl.amesh in mapdl.aplot #3613

[mikerife]

I really feel that the SOLVE command needs to be left alone:

1. MAPDL already has heurestics built in to set some solution settings based on model type and features. I don't know that PyMAPDL needs to add to these.

2. It's not always correct to turn on large deformation, even with models with contact. At the very least it's hard to show the benefit of large deformation if we can't turn it off!

[koubaa]

@mikerife what if it just printed out some suggestions, rather than throwing any error or changing the behavior of solve?

[mikerife]

Care should be used when looking at adding a warning to multiple ET commands using the same ET ID. Reusing ET IDs is standard procedure in one way load coupled multiphysics models. It can also be SOP when using the MESH200 elements. Maybe an option to turn on 'enhanced warnings' that would pick up this type of usage?

[mikerife]

@koubaa that a good idea (printing suggestions)!