Profile of Points Vs Profile of a Line

[Cumofo]

Dear Friends,

I weighty question here:

Has anyone noticed a difference in best fitting a profile of lines or arcs, as opposed to a profile of points? I have a part that seemes to produce different results based upon the weighting of the best fit elements.
Does Pc-dmis interpet the centroid of a line or arc as a single element (for puposes of best fit)? Assigning equal weight as a single point?

I may need to revise a program based opun our findings.

Gabriel

[craiger_ny]

OH no. Is this in regards to that dispute with a supplier?

Craig

[Cumofo]

Yes Lucky me, the new gunslinger.

OH no. Is this in regards to that dispute with a supplier?

Craig

Yes, I have be refereed in favour of the supplier. So it's wipe the egg of my face and I need to understand why my programming techique produces the results it does.

[CMMPGMR]

It will be different. I have encountered the same in profiling all around a part that has radius corners. If you generate a set utilizing circles it will apply to the center or location of the circle only. If you use individual point along the arc then it will apply them independently thus if your profile is .010 then your corner radius points could vary +/-.005 & still be in tolerance. I believe it would apply the centroid point of the line in the same way.

[JamesMannes]

I don't have anything to add except that it is a very good discussion. Printing begins now.

[Cumofo]

It will be different. I have encountered the same in profiling all around a part that has radius corners. If you generate a set utilizing circles it will apply to the center or location of the circle only. If you use individual point along the arc then it will apply them independently thus if your profile is .010 then your corner radius points could vary +/-.005 & still be in tolerance. I believe it would apply the centroid point of the line in the same way.

I'm unsure about the centroid theory in the evaluation mode, because I can clealry see tesalation in curves, and the deviation as scaled by the graphical analysis macro. I think I will go jump off a bridge now.

[John Kingston]

I'm unsure about the centroid theory in the evaluation mode, because I can clealry see tesalation in curves, and the deviation as scaled by the graphical analysis macro. I think I will go jump off a bridge now.

Just be like The Fonz and say "I was Wrooong.

[Paul Sarrach]

I'm unsure about the centroid theory in the evaluation mode, because I can clealry see tesalation in curves, and the deviation as scaled by the graphical analysis macro. I think I will go jump off a bridge now.

I've never seen any tesalation in my curves

[Cumofo]

http://mathworld.wolfram.com/Tessellation.html

in current CAD rendering and geometeric parameterization of least squares regression employed by regular joes like us, the actual form of any geometeric parameterization cannot be truly computed or displayed, therefore we experience tessellation.

A regular tiling of polygons (in two dimensions), polyhedra (three dimensions), or polytopes ( dimensions) is called a tessellation. Tessellations can be specified using a Schläfli symbol.

Tessellations of the plane by two or more convex regular polygons such that the same polygons in the same order surround each polygon vertex are called semiregular tessellations, or sometimes Archimedean tessellations
http://library.thinkquest.org/16661/index2.html