Spline through points

[bb1982]

Good morning to all members of the forum!

I was wondering which algorithm does UG use when drawing a spline through points: I'm studying the mathemathics behind Bezier curves, Bsplines or Nurbs in order to understand how it works but, for the time being, the only theoretical reference I could find deals with the opportunity to make the knots of the curve passing through the defining points, which should corresponds to the option "match knot position" in the studio spline command, but it's not what I'm looking for.
What about the "through points" option without the "match knot position"?Is it a piece-wise bezier curve?Is there any criterion for calculating the poles for a given set of points so that the spline crosses all of them?

Thank you in advance for your support.

 

[JohnRBaker]

Generally speaking, splines created thru points are NURBS made up of multiple Bezier segments. Optionally you can create Spline thru poles as a single segment Bezier curve. Whatever way you define a spline, it can be created in any degree upto and including degree 24.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[bb1982]

Thank you John for your reply.

Maybe I'm too specific now, but how are calculated both poles and the knot vector for a given set of points?

 

[cowski]

You can use some of NX's tools to find out more about a spline (Insert -> Datum/Point -> Point Set... to find the location of poles and/or knots, Information -> Object, make sure to select the spline curve and not the feature, to see info on: degree, number of segments, number and type of knot points, etc).

Unless the information you ask for is crucial to the use of the tool, I wouldn't expect anyone who knows the internal algorithms to divulge their intellectual property on a public internet forum.

www.nxjournaling.com

 

[bb1982]

Thank you cowsky,

the last part of your reply is actually what I mean: whether that algorithm is something "hidden" (i.e. intellectual property) or it is a "simple" application of formulas or methods publically available in the specialized literature.