search

Dhondtguido / CalculiX

This repository contains the source files of CalculiX, a three-dimensional Finite Element Program (www.calculix.de).
GNU General Public License v2.0
86 stars 21 forks source link

Eigenvalues of shell first order not good in relation to other codes by adapy fem verification project #87

Open stefano2734 opened 4 months ago

stefano2734 commented 4 months ago

See adapy fem verification project with calculix In docX

comparison of abaqus, Code Aster, Sestra and Calculix 2.21

https://github.com/Krande/adapy/releases/tag/ADA-FEA-Verification-Report-240626

„weakness“ of tet4 is in all codes in relation to hex8. the frequencies of tet4 are much higher than hex8.

Shell with tri3 and quad4 is not in good shape against the other codes and the second order results of tri6 and quad8. They are too „hard“.

So here is a „weakness“ in code by „harder“ shell first order elements in eigenvalues.

Beam elements are not tested by ccx in this comparison.

disla2 commented 4 months ago

Hi, @stefano2734

I give it a quick check to the one having largest discrepancy. Shell 1st order elements. Beam(“MyBeam”, [0.0, 0.5, 0.5], [3.0, 0.5, 0.5], “IPE400”, “S420”) My results are quite different and do not show that large discrepancy ADA is reporting. Inp attached for anyone else to try. (it's in txt format. inp is not allowed.Just change extension to inp.)

imagen ADA_FEA_Verification_Report.txt

stefano2734 commented 4 months ago

To disla2 Than it is an error in inp-file or run for Calculix at this Report. I will make an issue at Ada-Report in GitHub.

thank you for your effort.

can you also check here the example with shell tri3?

stefano2734 commented 3 months ago

See new docx report https://github.com/Krande/adapy/releases/tag/ADA-FEA-Verification-Report-240812

With some significant differences to abaqus 2024 by quad reduced integration 1st order and tri full and reduced elements.

some problems are solved, but not all.

perhaps it should be a new example problem for calculix example bibliography with different solid, shell and beam elements with link to adapy comparison of other codes.

stefano2734 commented 1 month ago

New report with some changes, but with results by calculix 2.21 https://github.com/Krande/adapy/releases/tag/ADA-FEA-Verification-Report-241010

stefano2734 commented 6 days ago

Actual report: https://github.com/Krande/adapy/releases/tag/ADA-FEA-Verification-Report-241121

I see here some differences to abaqus quadR and Calculix Quad QuadR at first order: 2 frequencies with 96 and 126 Hz are not calculated.

mostly solid and shell elements are same like abaqus and Code Aster.

Good beam elements with Doppel-T section are not simple available in Calculix.