Matthieu Ancellin
Matthieu Ancellin
> @mancellin right now the `*.nc` file from Capytaine stores `hydrostatic_stiffness`, but not `disp_volume`, `center_of_mass`, or `center_of_buoyancy`. WEC-Sim needs all of them. Do you want those those parameters added to...
The problem comes probably from horizontal panels on the free surface (there is a division by z = 0). I have on my todo list to provide a more helpful...
For now: removing the panels on the free surface (the clipper of `body.keep_immersed_part()` does it IIRC) or sinking a little bit the mesh (`body.translate_z(...)`). In the long run, it might...
What have you tried exactly? The error message says you have panels above the free surface (or below the sea bottom).
The mesh is probably still too close from the free surface: the 1/z coefficients are huge and the linear solver (GMRES) has troubles inverting the matrix...
Sorry it wasn't clear. I'm assuming you've translated down your mesh by some very small epsilon, is that right? Then, the computer would be able to evaluate the 1/z singularity...
Oh, ok, that is not what I thought. Then, I don't know what is happening...
Just checking in to say that I did not have time yet, but I am planning to give it a try.
Coming back to this problem (better late than never...). In your first message, there might have been a first issue caused by panels on the free surface, leading to a...
It seems that such a function is missing. In the simplest cases, multiplying the coordinates of the vertices might be sufficient (not tested yet): ``` body.mesh.vertices = 50*body.mesh.vertices ```