we are implementing a workflow for isogeometric analysis, and we are investigating the idea of working with SubD to be able to, later on, convert to NURBS, and perform our analysis. However, when building the topology we defined some Edge loops, marked in the first picture, that need to be respected when creating the NURBS. Nevertheless, this is a rule that is not always followed.
I suppose this happens because the surface is simple curvature and does not need to be subdivided, but for the continuity of our analysis it is necessary that the patches are all the same.
Any advice on how to ease this condition?
(In the ToNurbs command the SubDOptions G0… G2 options have been tested)
In that case it create single NURBS surfaces for every SubD face. In that case we cannot use it because, it is not a shared or common NURBS.
We could join them, but then, it is a polysurface…
@Kevin_Moreno_Gata There is no support for user-defined face packs in ToNURBS:
Is there any way that, in your IGA software, you could implement compatibility conditions across edges between patches of different knot structures? It feels like sooner or later you are going to run into a model where even the best face packs would lead to that situation.
Also, note the conversion of SubD to NURBS around extraordinary vertices is geometrically inexact, which might lead you to create NURBS models that are not an accurate representation of reality or that have poor qualities as the basis of a FEA. Since meshes obtained from the subdivision of SubDs are typically of very high accuracy and have a very good almost quad layout, have you tried using those meshes as the base of your FEA?
