Krande / adapy

A python library for structural analysis and design
https://krande.github.io/adapy
GNU General Public License v3.0
96 stars 21 forks source link

Calculix 3D beam B31 and B32 possible for comparison of beam elements #102

Open stefano2734 opened 4 months ago

stefano2734 commented 4 months ago

Is your feature request related to a problem? Please describe. A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

see http://www.dhondt.de/ccx_2.21.pdf page 118 and 119

B31 and B32 are the elements for this calculation with beams in Calculix ccx.

Describe the solution you'd like A clear and concise description of what you want to happen.

Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered.

Additional context Add any other context or screenshots about the feature request here.

Krande commented 3 months ago

Hey @stefano2734!

Would you like me to add these beam elements for the Calculix ada-py FEM tests? If I recall correcly, the reason for not including these elements were because Calculix did not support eigenvalue analysis using these element types with the specific cross section. I can't recall exactly the error message I got when attempting to solve this, but I vaguely recall something about it being possible using either a different custom element type and/or using general sections.

If you have any advice on this, I would appreciate any help you can provide:)

Let me know if I've misunderstood your comment?

Best Regards Kristoffer

stefano2734 commented 3 months ago

Three points here.

  1. Calculix beam implementation ist not pure Timoshenko for b31 and b32 elements. But this is not a disadvantage, because with internal boost to C3D8 elements more accurate results are possible.

  2. Section General results was with errors before version 2.21. perhaps there is a another problem. https://calculix.discourse.group/t/problem-using-u1-element-general-beam-section/600

  3. I see in documentation of 2.21 at 6.2.46 only static solution for small displacement for u1. Perhaps here no modal solution possible. Test file is only static: see username.inp and results in userbeam.dat https://feacluster.com/CalculiX/ccx_2.18/doc/ccx/node484.php?limit=500

I will try calculation with calculix 2.21 with some beam elements, but I see here not the example for the modal analysis. I see only only https://github.com/Krande/adapy/blob/main/files/fem_files/calculix/u1general.inp where is the modal file here?

stefano2734 commented 3 months ago

I made a test with userbeam.inp: no modal analyze with u1 elements. So a change to other beam element like b31 or b32 or others in calculix is necessary.

https://feacluster.com/CalculiX/ccx_2.18/doc/ccx/input_deck_viewer.php?input_deck=userbeam.inp For run, density is necessary I made: *Density 1.0e-8 and change from static run lines to FREQUENCY 10

Krande commented 3 months ago

Right, and if I remember correctly there are a limited number of cross section types supported for calculix B31/B32 elements? (just pipe and box if I am not mistaken?)

So if we should switch to B31/B32 we would have to modify the cross section properties to match that of any other cross section?

I believe that is the core reason I ended up using the general section element type. It just seemed easier to directly input the cross sectional properties such as A, I11, I12, I22 etc..

Maybe it's possible to expand the number of supported cross section types beyond just pipe and box by writing custom types? (I don't have any experience with that though :( )

stefano2734 commented 3 months ago

beam section of b31 and b32 is limited in documentation to rectangular, elliptical, pipe and box without general. https://www.feacluster.com/CalculiX/ccx_2.18/doc/ccx/node60.html with internal expansion to C3D8I or C3D20 elements it should be solvable in future by the developers of calculix. but I will check this. sometimes the documentation is not actual.

another alternative is the Uxxxx Element with own values to beam element with section general for super advanced users. https://www.feacluster.com/CalculiX/ccx_2.18/doc/ccx/node72.html

stefano2734 commented 3 months ago

A fast workaround for beam elements b31 or b32 here is three rectangular sections for Doppel T.

here the example of a composite cantilever beam with two rectangular sections in beam elements with same nodes. https://feacluster.com/CalculiX/ccx_2.18/doc/ccx/input_deck_viewer.php?input_deck=beamcom.inp

combining of rect, circle, pipe and box to geometric added general section is possible. Price is multiple elements with multiple sections. Great value is more accurate calculation and more information of section result values.