Recently I have been using Prepomax to run my simulation, and I thank you Matej for all the job on the software and all those that help the forum living, it is really interesting.
For my part I have an issue with results of displacement. I have a part and i want to check it’s mechanical strength to centrifugal load. But I am also interested in displacement of the part. But I have this result in term of displacement. Which seems weird normally those would be the same for a fixed radius. There should not be such a difference between positive z axis and negative z axis.
I set a boundary condition on one of the circle at one edge of the parts, because I am interested to know what displacements are in the internal surface.
At first I was thinking it was coming from the mesh disparity with the construction line generated by the .step file. But the displacement difference seems too large to be created by the disparity of the mesh only. So I am not sure where this is coming from and how could I solve it.
Thanks in advance for your answers.
Best regards.
Eliott
That’s just the displacement magnitude (resultant). Look at the 3 individual displacement components. And check the largely scaled deformed shape, it will show you (in an exaggerated way) how the part actually deforms. Then try changing the boundary conditions (e.g. supporting the rotor on both sides) and see how this influences the results.
I think it is the mesh problem. I changed the mesh from Netgen to Gmsh (I added a mesh setup item called Tetrahedral Gmsh), and the result is much better:
For a fully symmetric result, I would add an additional reference point at the other end of the hole and fix its radial translations only.
But it is not expected since the original mesh is very fine to start with. The internal structure also looks OK:
Indeed when I check the u1 displacement I see that most of the disparities is coming from this axis. But is it not suppose to be also similar even on this axis in each direction within a same radius.
An unrelated question…how did you get the section view you created (see snippet below) to display your full element (3D) and not just flat faces on the cut surface ?
I think the fundamental problem is fixing the ring of nodes on an edge rather than a surface. The stiffness of the elements at that kind of support isn’t well defined or physically sensible - it’s a stress singularity - so it’s sensitive to variations in shapes of the elements there.
