Been struggling to get any surface that isn’t like a starving horse look on what should be a ‘simple’ hull. Tried 2 rail sweep with limited effect. Loft is good until it gets to the ends and won’t do the chine crease or the bow. A combo looks like it works, but then I can’t get the panels to join into a smooth surface without getting creases at the joints. Hair pulling out time again!
calshot lines.3dm (175.6 KB)
Putting aside the bow for now - the reason why this approach doesn’t work well (despite what that old YouTube tutorial would have you believe!) is that you’re trying to do something with one single surface, that should be done with a collection of surfaces. Take the sides for instance - a big portion of the sides of your hull are flat, or near flat - those should be made first. Same with the bottom - a decent portion of the middle of your hull on the bottom is flat. So in that region - you should have one single, simple surface for the side, one for the bottom, and then a Blend surface between the two. So the short answer is that your “patch layout” as it would be called is the problem - you’re trying to do everything with one surface, and this is the typical result of that approach.
In the linked video Gerard Petersen of Rhinocentre shows an alternative approach to creating a ship hull model. Rhino Tutorial Rapid Ship Hull Modeling - YouTube
all loft, sweep and similar functions work way better, when the cross curves have the same number of control points.
for smoother results use less curves and skip the ones not absolutely shape changing.
I (command: rebuild) rebuilt it with increasing number of control points, until the deviation is acceptable and then lofted it again:
boat.3dm (221.0 KB)
hope this helps
Was that done with network or loft? How many points was it rebuilt to?
Thanks, but how would I align the surfaces to smooth into the next ‘panel’ without a kink at the joint? I tried cutting them back with a gap, and using tween or blend, but I then get a flat strip.
it was made with loft and if I remember correctly 18 points. in the rebuild menu you can preview the result and see the deviation of the new curve from the old one. I think it was about 3/100mm of deviation when I accepted.
which panel are you talking about? I don’t understand your question, could you provide an image?
usually you can choose a starting surface in some functions, but I’m not sure this info helps you, as I didn’t really understand you question.
I think you need to spend a bit more time fine-tuning your lofting curves. Use the ‘CurvatureGraph’ to inspect the continuity. Alignmentment with ‘SetPt’ and adjustment of the control points before lofting will yield better results.