Question about beam calculation with calculix

About the development of the FEM module/workbench.

Moderator: bernd

ickby
Veteran
Posts: 3116
Joined: Wed Oct 05, 2011 7:36 am

Question about beam calculation with calculix

Post by ickby »

Hello,

I'm calculating simple beam structures to analyse deformation and stress in profiles. The calculix solver has the option to choose 3d for output which allows for stress visualisation, compared to the 2d output. However, I also noticed different results dependent on that setting. For a very simple beam deformation is the same, no matter the output setting, but von mises stress is different. The "output 3D" value is correct according to normal beam theory, and the value for output=2d is wrong. Any idea why?

Example the file attached:

Deformation: 0.37mm
Stress 3d: ~6.3MPa (theory: 6.4MPa)
Stress 2d: ~0.73MPa
Beam.FCStd
(115.05 KiB) Downloaded 76 times
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Question about beam calculation with calculix

Post by thschrader »

mmhh..there seems an issue with the model itself.
The tangent at the deflection-line should be zero at points A+B (the fixings), but that is not the case.
You can check it with paraview (convert frd-results to vtk-format, use "plot over line")
For this model the momentum-fixing is too "soft" (it rotates), you underestimate the moment at the fixing
and overestimate the bending-moment in the field.
beam_1D_example.JPG
beam_1D_example.JPG (101.7 KiB) Viewed 3017 times
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Question about beam calculation with calculix

Post by bernd »

but they are fixed ...

Code: Select all

***********************************************************
** Fixed Constraints
** written by write_constraints_fixed function
** ConstraintFixed001
*BOUNDARY
ConstraintFixed001,1
ConstraintFixed001,2
ConstraintFixed001,3
ConstraintFixed001,4
ConstraintFixed001,5
ConstraintFixed001,6
BTW: Stefan ... quite a lot of nodes ... Normally 1/10 of them would even be faily much.
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Question about beam calculation with calculix

Post by bernd »

might be related to this issue #3056
UR_
Veteran
Posts: 1354
Joined: Tue Jan 03, 2017 8:42 pm

Re: Question about beam calculation with calculix

Post by UR_ »

Got this:

Screenshot 003.png
Screenshot 003.png (18.81 KiB) Viewed 2990 times

:roll:
ickby
Veteran
Posts: 3116
Joined: Wed Oct 05, 2011 7:36 am

Re: Question about beam calculation with calculix

Post by ickby »

UR_ wrote: Thu Jan 09, 2020 11:55 am Got this:
:roll:
Also when changing the analysis to 2d output and a rerun?
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Question about beam calculation with calculix

Post by bernd »

ickby wrote: Thu Jan 09, 2020 6:20 am
Deformation: 0.37mm
Stress 3d: ~6.3MPa (theory: 6.4MPa)
Stress 2d: ~0.73MPa

Beam.FCStd
confirmed for

OS: Windows 10 (10.0)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.19.19181 (Git)
Build type: Release
Branch: master
Hash: 2504247d65271b937dd5f033a0efff9c0d7bf375
Python version: 3.6.8
Qt version: 5.12.1
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: German/Switzerland (de_CH)


since the Displacement magnitude is 0.37 for both there really is something strange ...
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Question about beam calculation with calculix

Post by bernd »

but to be honest I have no idea what are the von Mises for a 1D output? In which point in the cross section is this stress?
fandaL
Posts: 440
Joined: Thu Jul 24, 2014 8:29 am

Re: Question about beam calculation with calculix

Post by fandaL »

ickby wrote: Thu Jan 09, 2020 6:20 am Deformation: 0.37mm
Stress 3d: ~6.3MPa (theory: 6.4MPa)
Stress 2d: ~0.73MPa
FreeCAD reads nodal values of stress and displacement. CalculiX expands 1D and 2D elements to make analysis with 3D elements. If 2D output is requested, ccx interpolates values from 3D elements back to original 1D or 2D mesh. So if you have purely bent beam, interpolated stress will be ~0. And by the way if you have other than rectangular cross-section 3D representation is still hexa element so that nodal stresses cannot represent real cross-section.
But you can obtain integration point stresses in *.dat file by output request:

Code: Select all

*EL PRINT
S
maybe there is some tool to display them in Paraview.
You can also ask for beam forces and calculate stresses on your own according to your profile. I think it was discussed quite some time ago https://freecadweb.org/tracker/view.php?id=3126 . It should be card

Code: Select all

*EL FILE, SECTION FORCES
S
CalculiX will output beam forces instead of stresses. From manual *EL FILE card
• xx: Shear force in 1-direction
• yy: Shear force in 2-direction
• zz: Normal force
• xy: Torque
• xz: Bending moment about the 2-direction
• yz: Bending moment about the 1-direction
or use specialized sw for beam calculations...
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Question about beam calculation with calculix

Post by bernd »

that makes sense ...

from calculix manual ...

Concerning the output, nodal quantities requested by the keyword *NODE PRINT are stored in the beam nodes. They are obtained by averaging the nodal values of the expanded element. For instance, the value in local beam node 1 are obtained by averaging the nodal value of expanded nodes 1, 4, 5 and 8. Similar relationships apply to the other nodes:

beam node 1 = average of expanded nodes 1,4,5 and 8
beam node 2 = average of expanded nodes 9,11,13 and 15
beam node 3 = average of expanded nodes 2,3,6 and 7


Screenshot_20200109_175255.png
Screenshot_20200109_175255.png (19.13 KiB) Viewed 2904 times
Post Reply