I’m trying to create the hull of a catamaran. I lofted some profiles but the isolines show it is off. To create the profiles I drew over a shape with Crv:interpolate points. Did some trimming and curve blends and the profiles are essentially the correct shape, but they have a lot of nodes. Is this what’s causing the problem? In addition, any idea how to form the front piece? (Edit: Networksrf worked for the front. No a perfect transition, but I can get a closed polysurface)
HULL.3dm (3.2 MB)
Hello - for a clean loft, make much, much simpler curves and make them all the same (i.e. copies of an original- you’ll get a much cleaner curvature graph and a far better surface.
-Pascal
Thanks. I reworked/rebuilt the profiles and am getting a much better loft. However, I stuck again with the front. Can’t get Networksrf to work again. Not sure that is the best method anyway.
HULL_2.3dm (1.6 MB)
Hello - you could do still better on the surface by:
To make the nose, ExtendSrf what you have, and then trim back with an offset of the bow curve:
Then a Sweep2 (preserve height set) will make a complex but decent looking surface -
You can use that surface as a basis for a much cleaner set of surfaces if needed,
-Pascal
Wow! That was quick. My brain actually went to that procedure, but I had no idea how to achieve it. Thanks.
So this is probably good enough for my purposes. But there is an angle where the front joins the hull. Tangency problem? Anyway, If I wanted to blend/smooth it out, how could I do it?
HULL_3.3dm (2.4 MB)
You can use MatchSrf with curvature continuity along the long edges - but - the result will probably not be satisfactory. The front surface has a very large number of control points and MatchSrf with curvature continuity will only modify the control points along the edge and the adjacent two rows of control points. Therefore the changes to the surface will be confined to a narrow strip along the edge. The number of control points in the V direction of the front surface needs to be reduced from 56 to 6 or 7.
My guess is the reason for the large number of control points in the V direction is that the ends of the notches in the main surface do not coincide with isocurves on of the main surface. ExtractIsocurve at the desired locations, split the isocurves and use the split portions to trim the surface. Then build the blend surfaces.
Thanks for the response. I followed Pascal’s direction but somewhere I suddenly get all those control points. I think my extruded trim piece had 7, but the shape after the trim has lots. (56? How does one get a count?) I understood the excessive control points was a problem and tried all manner of editing but could not solve it. I also switched on “Surface Isocurves” in the display mode, but nothing shows. Unfortunately, I’m not exactly understanding your second paragraph.
What command.
One of the contributors to the large number of control points is the use of millimeters as the units with an absolute tolerance of 0.001 mm given the size and shape of the object. You might try mm and a tolerance of 0.01 or use larger units.
The edge of a cut surface usually cannot be represented exactly by a NURBS curve unless the edge is an isocurve of the surface. So Rhino uses enough control points to represent the edge within the absolute tolerance. Sometimes that requires lots of control points. Several strategies for reducing the number of control points required by the edge curve:
Go into Properties for the surface and click “Show surface isocurves”.
Ah! The tolerance. This is modeled to actual size. (37m)
One of the contributors to the large number of control points is the use of millimeters as the units with an absolute tolerance of 0.001 mm given the size and shape of the object. You might try mm and a tolerance of 0.01 or use larger units.
The edge of a cut surface usually cannot be represented exactly by a NURBS curve unless the edge is an isocurve of the surface. So Rhino uses enough control points to represent the edge within the absolute tolerance. Sometimes that requires lots of control points. Several strategies for reducing the number of control points required by the edge curve:
Make the edge coincide with an isocurve.
I don’t know what you mean by this. Which edge. The sweep2 edges? I can’t find any way to reduce the count there, and I can’t draw the curve (I tried) as it curves in all directions (G3?) Will a change in tolerance reduce the count there?
Larger absolute tolerance.
Will see what that does.
Simplify the curve and live with a gap larger than the absolute tolerance.
I at a loss as to which gap.