Fit Point vs Control point Spline Issues

[tomwmarquis]

Hello, I recently started using control point splines and have been having trouble when lofting a series of closed spline profiles. I will get errors when lofting and if I do manage to get a successful loft I get errors when trying to do a simple fillet. It will suggest trying something smaller. But these shapes are very simple so I don't know why it's choking. While troubleshooting I took a look at the curvature display and noticed a sharp break in the curve flow. The endpoints are constrained properly with colinear and coincident constraints (i believe). Is that break in the curvature flow normal for control point splines that are closed? As you can see in the example on the left I dont have that problem with the fit point closed splines.... Thanks for any insight!

[jeff_strater]

@tom - this is one of the inherent differences between Fit Point and CV splines.  The fact that the Fit Point splines create a closed curve that is both direction and curvature continuous is a bonus of that type of curve.  CV splines are more like working with the bare metal of spline curves.

What you see with the curve in your image is that making the two adjacent control polygon lines colinear guarantees tangency at the point, but not curvature continuity.  To get curvature continuity, you'd need to make one span back also colinear.  You can do this by inserting more CV vertices:

[tomwmarquis]

Thanks so much Jeff! But I don't get the same results when I follow your instructions. The closed endpoints are still pinched. Is there some other constraint I need to add? Something to the actual point where the two ends are coincident? So that it behaves like the closed point of a fit point spline?

[jeff_strater · Accepted solution]

@tomwmarquis - you have to make all 4 spans colinear (or parallel, either works).  These two still need to be made colinear:

It won't be 100% like a fit point spline, but you will get curvature and direction continuity.

[tomwmarquis]

That works Jeff! Thanks so much. But it does not give me the control I need in that area of the curve. Which is fine, I just have parts already referencing it.  Next time I will plan to have the closed endpoints in an area that does not matter as much.