FreeCAD produce different results under the same input parameters

[S-Y.Chen]

I have a very simple testing model here. It shows that FreeCAD FEM will produce different results under the same input parameters

The original model is 01.FCStd. The CAD parameters are linked to the Spreadsheet. After changing the input parameters in the Spreadsheet several times, and changed them back the the original value of 02.FCStd, FreeCAD produce difference results from its FEM result (for example, von Mises stress).

I suspected this behavior long time ago, so I decided to make a simple model to reproduce the problem

Unless I have done something wrong here, this is a serious problem of FreeCAD.

The two models after 7z are about 3MB in size. I have to update to the cloud. The link is proved below.

https://drive.google.com/file/d/1jO5-gc ... sp=sharing

[NewJoker]

The geometry is the same but the mesh generated by Gmsh is different. One has 5409 nodes and 3124 elements while the other has 5365 nodes and 3090 elements.

[S-Y.Chen]

This should not happen in an engineering software. an engineer program should produce the same result when then inputs are the same, and on the same computer.

All I do was change parameters, remesh and rerun. Finally I change the parameters to the original, but it produce different results.

I have reported this issue at
https://github.com/FreeCAD/FreeCAD/issues/7818

[NewJoker]

It seems to be a fault of Gmsh, it doesn't happen with Netgen. The same behavior can be observed with the newest version of Gmsh (4.11) so maybe this could be reported on their GitLab: https://gitlab.onelab.info/gmsh/gmsh/-/issues
However, I haven't managed to reproduce this using standalone Gmsh yet. Perhaps it's a matter of settings (I used default ones in Gmsh, apart from max element size).

[user1234]

It seems to be a fault of Gmsh, it doesn't happen with Netgen.

It is intended, see https://gmsh.info/doc/texinfo/gmsh.html ... RandomSeed, default is Mesh.RandommSeed = 1. Also it is to expect, when you change the version of gmsh or the numbers of CPU to mesh, that the result can vary.

@bernd: since you are the FEM guru, i have one question. Even when i switch off the Mesh.RandomSeed = 0, this happens sometimes. Could it be that it must somewhere specific in the *.geo file? At least i do not see it in the docs.

branch: https://github.com/grubertechnic/FreeCA ... d_property

Greetings
user1234