Replies to This Discussion

Permalink Reply by Mary Ann Fugier on September 6, 2012 at 10:56am

Hi André,

My initial guess is that the isocurves are turned off on the surface whose edge you are extuding. it is likely far more complex than the shaded image leads us to beleive.

So to take the guessing out, please ExportSelected and send me just the surfaces in questions. You can attach the 3dm to this tread or email mary@mcneel.com .

Thanks for your help.

Sincerely,

Mary Fugier
Robert McNeel & Associates
Seattle, WA

 

• ▶ Reply

Permalink Reply by Pascal Golay on September 21, 2012 at 5:53pm

Hi Andre- if the edge is a trimmed one, or joined to a trimmed edge on an adjacent surface, it may well have a more complex structure than the surface it is attached to. Untrimmed edges will inherit the surface structure but generally not trimmed ones.

-Pascal

pascal-at-mcneel-dot-com

• ▶ Reply

Permalink Reply by André bauer on January 14, 2013 at 12:33pm

Hi there,

I'm still absolutely not satisfied with the situation.

Now there's a surface of 10x6 points, if I make a section curve through it, wether by sectioning with another flat plane or with the section curve command the curve will have not 10 points but between 122 and 132 points. 122 is the result if I trim the surface and copy the edge curve, 130 points is another result and 132 points has the section curve itself.

I feel quite fooled.

A section curve MUST have the same number of points as the base surface.

And again an offset curve has "off course" about 370 points.

The offset curve included has now 129 points, again can't see a difference between those two.

What for do I have the offset function if the curve needs to rebuild and fitted manually to the desired offset curve?

If I make a section curve, the curve needs to have exactly and automatically the same munber of points.

Fast modeling is only possible if the offset curve has again the same number of points as the section curve.

Only then I can build a edge surface on these curves.

If I have to rebuild the section and offset curves with fitting them to the surface again, how much time shall spent I on?

Only chance is to trim all surfaces against each other, like this the surfaces will have few points only, the trim edge will be complex.

This seems not to be a system one can effectively work with.

Best Regards,

André

Attachments:

• Rhino_surface section_2013-01-14.3dm, 429 KB

• ▶ Reply

Permalink Reply by James Carruthers on January 14, 2013 at 12:59pm

Andre,

If you want a section with the same number of points as your base surface, you use ExtractIsocurve, which actually takes a curve directly off the surface definition.

If you intersect something with a surface, then the result is a fitted intersection, there's absolutely no reason to expect it to even come close to having the same number of control points as one of the input objects, especially in the example you posted where there's no relationship between your section plane and the structure of your surface. There's not a CAD system on the planet that's going to give you a section curve with only 10 points and 0.001mm precision.

It's actually mathematically impossible, unless you're just working with straight lines and arcs, for an accurate offset curve to have the same number of control points as the input. The offset of an ellipse is NOT an ellipse, for a very simple example, try offset vs a scaled copy of one to see what I mean. Same applies to higher-degree freeform curves.

The question of course is what exactly are you actually trying to DO. If you don't want to build surfaces off of really dense intersection curves, the fact is most of the time you don't have to. Learning how to make efficient NURBS models is about how to structure your modelling workflow to avoid that as much as possible. I would recommend some training.

• ▶ Reply

Permalink Reply by André bauer on February 1, 2013 at 5:39am

Hi all,

VSR-Shape Modeling has functions which produce a hole lot less points than rhino itself whith the intersection command for example.

So it is possible to have curves from surfaces with less points.

And its not my missing understanding of nurbs as told by the support earlier.

Which is a affending thing saying to a client he's stupid.

The support should rethink to which company it belongs and who gives the money to buy the bread tomorrow.

A saying in germany:

§1 The client is king and always right.

§2 If not see §1

A way to work around is not the way to solve this, to much time consuming.

Also the problem is interesting to everyone, not only to me, also told by the support...

Everyone wants good fairing surfaces with minimum points with less work as possible.

VSR shows its possible!

The argument to the prize of rhino is a good one of course.

Just need to buy VSR, not use the rhino functions and the problems are smaller.

André

• ▶ Reply