"Beam Element" and Explicit Time Integration

1

Hi there.

I’m new to calculix and I must say that I made a huge research of Open Source software for FEM. I’m familiar with Ansys and LS-Dyna and got a bit frustrated when I realised that the ‘Beam’ element (B32) is actually expanded to a hexaedron.

I’m simulating pretty long beams (such as risers or cables) with small cross section areas.

The problem arises when I create an explicit integration simulation. The expansion for a solid element creates quite small edges resulting in ridiculously small integration time steps. I may use Implicit time integration, but the computational cost may differ significantly.

Time steps for a Beam element shouldn’t be function of cross section area, as happening here. Does anyone implemented a true Beam element?

Thank you so much in advance!

Best regards

2

I think you could create your own user element. From page 140 of the current (2.17) manual,

An example for a 3D Timoshenko beam element (for static linear elastic
calculations and small deformations) according to [92] is implemented as element
“U1” in CalculiX. It is used in example userbeam.inp in the test suite. The
reader is referred to files resultsmech u1.f, e c3d u1.f and extrapolate u1.f for
details on how a user elements is coded."