Soil model in FreeCAD FEM

[ebrahim raeyat]

Thanks @thschrader.

I think in this way it not produce the nodes with same location, not only different z, but different x, y coordinated. but I wanted exactly pair points with same location. As I said I can do it manually in opensees file from the generated mesh, but this way I must calculate area of each mesh and multiply it to soil modulus with kgf / cm3 (for example) unit to obtain each spring stiffness. also for each node I must sum all stiffness from all shell elements that this node share with those. I done it, but I think it is ugly.

[NewJoker]

I wanted a spring like constraint that share two nodes in the same location.

CalculiX has spring elements but they are not yet supported in FreeCAD so you would have to export the input file, define them manually (following the description in the documentation), run the analysis using standalone CalculiX solver and then import the results to FreeCAD.

The FEM ConstraintSpring available since 0.20 is unfortunately not yet usable since it was meant for the MoFEM solver which still hasn't been implemented in FreeCAD.

[thschrader]

...
but I wanted exactly pair points with same location.
...

What about the "Coherence Mesh = false" option?
No gap between the parts (!). As far as I understand this option,
the meshnodes in the compound are at equal location at the connected faces, but they are doubled.
When running the contact analysis again, you can see in the deformed plot, that the nodes can
glide. Maybe this helps. Or I totally misunderstand you...
File:

3point_noCoherenceMesh.FCStd

(39.22 KiB) Downloaded 34 times

noGap.JPG (114.01 KiB) Viewed 1360 times

[ebrahim raeyat]

CalculiX has spring elements but they are not yet supported in FreeCAD so you would have to export the input file, define them manually (following the description in the documentation), run the analysis using standalone CalculiX solver and then import the results to FreeCAD.

Thanks. You said me this before and I want to do the way you say with opensees. many thanks.

[ebrahim raeyat]

What about the "Coherence Mesh = false" option?

Yes, I think this is exactly what I want.

mesh1.png (156.31 KiB) Viewed 1247 times

Then, I tried to create an analysis with above two plate. I added loads on smaller plate and fixed the bigger plate. Then I used constraint contact for two plates and shell for both plates. But the displacement about z direction is zero. my question:

Can I model surface on soil with this approach?

[thschrader]

...
Can I model surface on soil with this approach?

Thats exactly what I want to know.
My commercial software uses springs for modelling soil-beam contact.
A very simple model. Input is soil-modulus x beam-width.
See result for self-weight structure alone:
The vertical displacement is the highest at the corner of the frame.
For me this is counter intuitive, I would expect max displ in the middle of the frame.

soilModulus_beams.JPG (119.66 KiB) Viewed 1182 times

vertical_disp.JPG (44.71 KiB) Viewed 1182 times

[thschrader]

...
The vertical displacement is the highest at the corner of the frame.
For me this is counter intuitive, I would expect max displ in the middle of the frame.

Result is correct, sorry.
I forgot the ballast on the frame (concrete plates 1)

ballast.JPG (28.51 KiB) Viewed 1148 times

[ebrahim raeyat]

Thats exactly what I want to know.

Maybe @bernd can response.