Attached is an example of a bad surface intersection result.
Transom object was created in Maxsurf and imported to Rhino. Transom Extension was created in Rhino.
Note that the original Transom surface (also attached) had a single trimming border, which is not desirable. However, even un-trimmed it isn’t working properly.
Hi Ncik - if you insert a few knots in V on the vertical surface it works - the knot and point spacing on these surfaces is a little unusual… V5 seems to run into surface-surface intersection problems most often on sparse areas of the input surfaces.
V6 does not like this setup as is either, by the way - I’ll get that on the pile.
I worked around it by un-trimming the vertical surface and re-building the horizontal. Infact, the horizontal is probably the dodgy one because another intersection between it and another surface was bad also. After the rebuild both intersections worked. Unfortunately the horizontal surface was a rhino surface, I think it was an edgesrf.
EdgeSrf creates surfaces with knots which correspond to the edge curves, so problems with the result of EdgeSrf are likely to have been inherited from the edge curves. Your surfaces have several rows and columns of control points close to the edges without any abrupt changes in shape which might require them. Do you know why those control points are there?
Hi Nick - these one-edgers are pretty common when importing iges - apparently, the edges are defined that way in the iges file and Rhino does nothing (despite my lobbying over the years) to Rhino-ize them.DupBorder/Untrim/Trim is the way out… This file has a somewhat nasty looking trim curve, I would say, ( 690 points, with some abrupt spikes in curvature graph) I’d think about making a new curve there to trim with.
@davidcockey They’re probably a result of extendsrf being applied a few times to try and get the intersection to work. I noticed them too, hence the rebuild of it.
@pascal Standard dodgy Maxsurf trim between two fairly simple surfaces.
Thanks for the help. The problem at my end is sorted for this case, but I like to send these in to McNeel and discuss openly as a service to the Rhino community.