I have 2 curves.
1st curve have degree 3 and CP 22
2nd Curve have 5 degree and CP 6.
I created surface with lofting both curves.
Now what will be CP(Control Point) on lofted surface in U & V directions?
Thanks
In the V direction probably many (more than 22) - Rhino will have to fit a Degree 3 5 surface between the two curves that stays within tolerance on both edges.
In the U direction, if you have only two curves, the degree will be either 1 (straight line) or 3 depending on the option you choose in the loft type.
1 Like
Loft works the other way around. The loft surface degree in the V direction will be the highest of the input curve degrees with sufficient control points to match the input curves within the tolerance. Multi-knots may be introduced. In the OP’s example the loft surface will be degree 5 in the V direction.
1 Like
Thanks
I have 2 curves.
1st curve have degree 3 and CP 22
2nd Curve have 5 degree and CP 6.
I want make same control points on both curve and if possible same degree, to get good surface(Am I correct?)
Second thing, if I try rebuild 1st curve to make it same like 2nd curve, 5 degree and 6 CP.
It’s not possible by me because in reducing CP from 22 to 6 , it does alter with original blue print curve.
What should I do in such situation?
Any tips?
Rebuilding with fewer control points will alter the shape of a curve. Rebuilding with more control points does not (much, anyway). Changing the degree of a curve to a higher degree will also alter the curve.
Your best bet to keep the original curve’s shape might be to rebuild the degree 5 curve to match the degree 3 curve with 22 points. However, that is still a “noisy” curve with lots of control points. Another possibility is to duplicate the original degree 3 curve, lock one copy, rebuild the other with deg, 5 and 6 points (clean, single span) and then manually adjust the control points until you get a close approximation of the original curve.
Loft also has rebuild/refit options within the command.
This is the way to go of the shape allows it.
@Helvetosaur changing curve/surface degree upward should not change the shape (if Deformable=No)
-Pascal
Thanks, I have tried all ways,manually too but it do change still from my original blue print above tolerance limit
Can you post an example set of curves?
-Pascal
So pascal do we need to make deformation: yes
To get expected result (like of my case to fit my curve over blue print curve)?
macktest.3dm (434.8 KB)
i uploaed here : 2 curves(slightly currentlly there CP are changed)
i am looking for minimum CP to loft these curves without getting deviated from these bluepint curves.
ChangeDegree (not Rebuild) to a higher degree with Deformable=No will not change the shape of the curve, but it will introduce multi-knots at the knots of the original curve.
Deformable=No to retain the shape of the original curve.
1 Like
It appears that Loft in V7 when used with input curves of different degrees effectively changes the degree of the lower degree curves without deformation before creating the lofted surface.
mack’s example results in a lofted surface which exactly (within digital math precision) matches the input curves. CrvDeviation can be used to check the deviation from the input curves, and in this case it was 1.3e-12 and 0.9e-12.
How close to the blueprint curves do you need to be?
FitSrf can be used to simplify a surface. The “Fitting tolerance” in FitSrf is not exact so check the edges of the resulting surface to the input curves using CrvDeviation. If the deviation is too large Undo and use FitSrf again but with a smaller tolerance. If the deviation is much smaller than allowable but the are more control points than desired you can Undo and use FitSrf again but with a larger tolerance.
Added: If you plan to join the lofted surface to other surfaces then the coincident edges need to be within the “Absolute tolerance” (set in Document properties).
1 Like
Hi mack - no idea if I am close enough here but maybe this will get you started.
macktest_maybe.3dm (457.3 KB)
-Pascal
1 Like
pascal thanks, may i know what exactly you did to fit these curves?
thankyou
Hi mack - exactly as @Helvetosaur suggests - rebuild to the same degree and point count and adjust the points - I guessed at 12 points here because the high curvature and curvature changes at the ends - it may be that fewer could be used.
-Pascal
1 Like
Pascal if later in my design i want to split both curves in two parts ,than there would be again subset of original curve with same degrees, and thus will increase CP.
If i know it from start that later i need split both curves , how than i can stil let my 4 curves now remain with 6 Cp ? Is it possible?
thanks
Hi mack - If you start with a single span curve (Degree + 1 points) splitting will retain that structure for both parts and joining again will indeed have more points as well as a kink point at the join, so that is to be avoided if you want to retain the structure. Keep a copy.
-Pascal
offcourse than they will become two seperate surfaces and if join them than polysurface.
But if we matchsurf both adjacent edges of surfaces just to get correct continuity of surfaces, will it be a problem, like in case of hull design and fairing?
i also want know if i want check continuity(G0, G1,G2) between two adjacent surfaces,to get good fair(Zebra),is there is any command to check it?
i dont know if i apply GCon will it give asnwer about continuity of 2 adjacent surfaces?
Hi mack - use EdgeContinuity to check surface to surface coninuity.
-Pascal
1 Like