@tfostermorphis - see if this helps at all - I don’t know what your curves are like, but if you can get a copy of the nice curve somewhat close to the Crv2View output (or any nasty curve), then run this script, it may clean up well enough.
Hi Tom - it looks to me like it will be hard to closely match the 3d shapes with the single span curve - at least in any fully automatic way, but the tool I posted can help you pull it closer if you get it somewhat close by point editing.
Trying to automate curve fairing for boat building, frankly, isn’t going to work. It’s an art not a science. The “solution space” of the problem is literally infinite, the rebuilding tools are too crude to get you more than partway. If it could be done it would have by now–maybe if you want to get into machine learning, I dunno maybe? Do you have 1000 clean hull models to train it on?–many many people make boats with Rhino. It just takes some practice to get the hang of, and time to do.
Well, no. There’s no NURBS math that says that this curve in top view + some other arbitrary curve in side view=a perfectly fair 3D curve. Even if there was it’s not physically possible to input them accurately enough. If you want a fair curve…just draw the curve in 3D?? What is the aversion to doing that? It takes minutes.
If I accept the C2V 3d line, as is, w/ variances in point count compared to 2d input, 6 pt to 50 or 60. Degree appears to remain same as input curves. Form appears accurate.
Maybe all 3d lines, regardless of number of points, could be made with existing C2V cmd and sections might be made at each station. Results would be that a point set, for each 3d line will be arrayed longitudinally with a point at each station/section.
If I re-draw thru these points, will the new curve conform to the desired shape w/ consistent degree and point count between pairs for surfacing? My intent is to surface w/ edge curves w/ minimum, matching, point count and degree.
I’ll give it a try as we try to get the script for C2VLOOSE, going.