This thread reminds me of all the problems I had creating Fuselages for Aircraft with nurbs.
Gord
If you haven't already explored this option its worth trying using the loft, loose, exact method. See Andrei's video about 10 minutes in https://youtu.be/WBVUsr-eYdw
The trick with this technique is that all your hull cross section shapes must have the same number of points on them.
After you've lofted the hull shape, you can go back and tweak the points of you splines to create a nice smooth surface that matches your reference images & photos (as closely as possible).
Lofting also allows sharp changing of direction with nurbs without the artifacts and distortion that often occurs when using networks and sweeps.
Sometimes though, swept rails just won't boolean into another solid surface, whereas others had no problem. See the two rails on his one. And why, when you boolean does the object sometimes disappear?
> Sometimes though, swept rails just won't boolean into another solid surface, whereas others had
> no problem. See the two rails on his one.
Similar to how it is difficult to intersect a surface that skims right along another surface area it can also be difficult to have skimming edges like you've got here:
Another problem is in this area where you're trying to boolean there are some microstructures:
It will probably be difficult to do a boolean cut that has to try and intersect through that messy area.
> And why, when you boolean does the object sometimes disappear?
Part of the sequence of doing a boolean is it intersects the 2 objects and then needs to gather well formed closed intersection curve results which it then uses to partition objects into different pieces. If it was not able to get a good closed intersection then it can get confused about how to divide pieces up. The pieces that disappear were being classified as being in the volume that is to be discarded. If there are a lot of little tiny fragments in the area that is being intersected they can get partially glommed together and that can impede getting a well formed close intersection.
So it's not good to do repeated booleans of pieces that are kind of close to being aligned but just a little bit off, that can make those little slivery pieces.
That's basically a problem area to work with since these original hull surfaces didn't align very precisely here like I was talking about earlier above:
This area at the top of that juncture now has problems with edges not meeting up with each other very closely and that kind of loose tolerances will tend to increase "glomming together" in intersections through those spots.
If you're going to be doing more stuff along those joining areas then it's good to get them aligned cleanly from earlier on, it helps to avoid problems further down the road.
OK, thanks. I moved it away from the little parts as it worked. However, on another part, when I boolean, it suddenly deletes the structure for some reason! What happened?
> However, on another part, when I boolean, it suddenly deletes the structure
> for some reason! What happened?
If you post the model I can take a look. It's hard to guess what might be going on without the model.
But it's probably related to those problem spots with micro faces and edges that don't connect to each other very accurately like I was talking about above.
gord_microstructures1.jpg
