HI
In the attached model (symmetric and symmetrically loaded structure) I applied a surface force equal to Fx (F1) =1 N (total value 1x 999 nodes = 999 N). The values of the forces (Fx)
on the bolts are incorrect. I don’t understand why
Thanks everyone for the help
Trave_Colon.pmx (6.3 MB)
Why don’t you use the Surface traction load instead ? Also, is the compression-only constraint really needed here ? It’s placed between two parts while it should serve as a boundary condition (support connecting the structure to the ground). Maybe you should replace it with a layer of compression-only material (using the Keyword Editor).
The actual load to be applied is composed of Fx (F1), F2 (Fy), F3 (FZ), Mx and My. The example is for testing only.
I also tried applying a HARD surface contact between the faces, but the problem remains.
Why are the forces on the bolts different and not symmetrical?
I can’t solve the problem.
Thank you
Ok, but you’ll need another way to apply moments (e.g. using rigid body constraint).
Yes, contact would be more appropriate here.
Disable the load and you’ll still see the reaction forces in the bolts. They come from contact interference adjustment (sort of interference fit). You could use tie constraints to eliminate this effect but then you would need a different way to check the forces in the bolts (like *SECTION PRINT).
For explanation @Agazzotti, you have meshed each part separately which leads to penetrating nodes and distorts the elements between curved surfaces.
To avoid this, simple create a compound part before meshing the assembly and change the mesh setup → Mesh operations → Merge compound parts: No
Thanks for the reply, now the results are more reliable.
I have one further observation
I analyzed the structure with 2 different contacts. The first considering a “hard” type contact between the surfaces and the second considering a “Compression_Only” type contact
By applying a force F1(Fx) = 1000 N
The sum of the forces on the bolts for both cases is always equal to 1000 N.
In the first case the forces on the bolts are substantially equal F=Fx/4 = approximately 250 N, and this is correct; in the second case the forces are very different:
_ FX(B1)= 317.38 N
_ FX(B2)= 319.00 N
_ FX(B3)= 182,70 N
_ FX(B4)= 180,90 N
I don’t understand why.
I compared the v2.0.7 and v2.0.8 versions and noticed that the calculation times were very different:
Case 1 = 111 sec (v2.0.7)
Case 1 = 332 sec. (v2.0.8)
Case 2 = 32 sec. (v2.0.7)
Case 2 = 88 sec. (v2.0.8).
Thanks for your help
As I’ve mentioned before, the compression-only constraint is not a type of contact but a support (boundary condition). It connects the surface to the ground while contact connects it to the other surface in front of it. Not to mention the different default spring stiffness values.
Thanks to the immediate response.
All clear.
