Problem with 2 separate beams in one simulation

[bnret]

Hello everyone,
I am trying to model a network of multiple beams which are not necessarily connected at each load step. I started by constructing a simple example involving 2 separate beams. This is the code I have come up with so far:

from ansys.mapdl.core import launch_mapdl
mapdl = launch_mapdl()
mapdl.clear()
mapdl.prep7()
mapdl.et(1, "BEAM188", "", "", "", "", "", 2)
mapdl.mp("EX", 1, 30e6)  # Elastic modulus
mapdl.mp("NUXY", 1, 0.25)  # Poisson's ratio

#define crosssection
mapdl.sectype(1, "BEAM", "CSOLID", "", 0)
mapdl.secoffset("CENT")
mapdl.secdata(0.2, "", "", 0, 0, 0, 0, 0, 0, 0, 0, 0)

#Create Geometry
mapdl.n(1, 0, 0, 0)
mapdl.n(2, 1, 0, 0)
mapdl.n(3, 2, 0, 0)
mapdl.n(4, 0, 1, -1)
mapdl.n(5, 0, 1, 1)
mapdl.n(6, 0, 1, 2)
mapdl.e(1, 2)
mapdl.e(2, 3)
mapdl.e(4, 5)
mapdl.e(5, 6)

#BCs
mapdl.d(1,'all',0)
mapdl.d(4,'all',0)
mapdl.d(6,"UY", -1.5)
mapdl.allsel()
mapdl.finish()

#Solve
mapdl.slashsolu()
mapdl.solve()
mapdl.finish()
result = mapdl.result

nnum,disp = result.nodal_displacement(0)
print(disp[:,1])


mapdl.exit()

Which yields the following output:
[6.36598737e-314 0.00000000e+000 0.00000000e+000 0.00000000e+000
0.00000000e+000 0.00000000e+000]

That is obviously incorrect, as the displacement at node 6 is set to be -1.5. This problem does not occur when I do the calculation in MAPDL, or when I delete the second beam. In those cases, the result is totally fine. I cannot see what goes wrong when I set the same algorithm that works in MAPDL into PyMAPDL. I am running PyMAPDL from PyCharm Community Version 2024.3.

Has anyone an idea how this can be fixed? Any help is much appreciated!
Kind Regards

[germa89]

Hi @bnret

You should use post_processing module:

>>> mapdl.post1()
>>> mapdl.set(1,1)
>>> nodal_displacements = mapdl.post_processing.nodal_displacement()
>>> print(nodal_displacements)
[0.         0.         0.         0.         0.73986644 1.5       ]

I am not sure why the results module does not give you the right answer. Maybe @mikerife knows something?

[mikerife]

@bnret @germa89 The results module does not take into account the orientation node, which messes everything up. It will also not take into account the auto-added interior nodes if the shape function (keyopt 3) was set to quadratic or cubic. Anyway if you look at the element nodal connectivity for element 1 it is nodes 1, 2, 0. Not 1, 2. It's a zero since a) you did not define an orientation node and 2) the beam was found to be aligned along global X. But that node connectivity still has a value. I don't think whoever developed PyMAPDL Reader way back when took this into account. So the result access routine is not quite right for beam elements and we get these types of results. As @germa89 said, better to use the post_processing module.
mike

[germa89]

Thank you @mikerife for your very useful answer. :)

[bnret]

Thank both of you very much for the fast reply, that helps a lot <3