Calculix Vs Abaqus INP displacement value mismatch

1

Hello,
Greetings!

I am currently running a simulation on a cube with dimensions 10 mm × 10 mm × 10 mm, divided into 2 layers with a voxel size of 5 mm. I successfully converted an Abaqus .inp file to a CalculiX-compatible .inp file.

While the displacement field distribution visually matches between Abaqus and CalculiX, I noticed a discrepancy in the maximum displacement values. Specifically:

Abaqus reports a maximum displacement of 0.052 mm

CalculiX reports a maximum displacement of 0.031 mm

I have attached both the Abaqus and CalculiX .inp files for your reference. Could you please help me identify the cause of this discrepancy and suggest how to resolve it?

Calculix inp -

Abaqus inp -

Thank you for your support!

Best regards,
Mahesh

2

It’s hard to compare when the input files are created independently. It would be better to make CalculiX .inp file first and the modify it in text editor for use in Abaqus or vice versa. There’s a guide covering the most important differences here: A Guide to Modifying Abaqus Input Files for Use in CalculiX

Keep in mind that CalculiX doesn’t have the quasi-static application type for *DYNAMIC step.

1 Like
3

The CCX file uses C3D8 and the Abaqus one uses C3D8R. That could cause big differences in a coarse mesh.

1 Like
4

It rises up to 0.048

1 Like
5

0.048 vs 0.052 seems close enough given there might be some differences in how the two programs do hourglass control.

6

The simulation is converging for very coarse mesh, but diverging for fine mesh. When changed to *DYNAMIC, solution is converging for fine mesh, but values are not matching. Could you please help me in identifying which *STATIC solver setup would work for fine mesh ?