I am working on an FE model containing beams that have a square (or rectangular) cross-section. As a first example, I used a beam with a square cross-section. Each side measures 20e-2m. When I export my model in LS-DYNA format, I see that a *SECTION_BEAM keyword is created. I understand well the first parameters from secid to itt, but I can't understand how the j parameter is computed.
Using Hypermesh for a square cross-section with a 20e-2m side, I have J=2.25310E-4m^4. However, when I try to find this value again using formulas from the wikipedia page (https://en.wikipedia.org/wiki/Torsion_constant), I can't find the exact same value. Depending on the used formula, I either find J = 2.25000E-4 m^4, J = 2.25600E-4 m^4 or J = 2.25333E-4 m^4. These values are close (especially the last one), but still they're not equal to the one given by Hypermesh. Right now, it might not be seen as a problem, but what if the values become too different if I use other cross-section values? That's why I need to understand what Hypermesh does.
I have searched the Hypermesh documentation but I didn't find any information about how this value is computed. And I need it to understand this because I want to run comparisons between commercial solvers and personal codes.
Does anyone have a hint about how to find this value? Thank you in advance,