Relationship of surface continuity to toolpath generation

[DaSalo]

Hello,
A bit of an abstract question here: I was discussing surface quality with our senior NC programmer yesterday and he made a statement that seemed questionable to me at the time. He claimes that he experiences a direct correlation between the continuity numbers (c0,c1,c2,etc.) of the splines used to create a surface and the success that he will have generating clean contouring toolpaths for that surface. He associates these numbers as c0 or c1= good surface, c2 or greater = bad surface (i.e. designer screwed it up and made his job more difficult.)

My question is: What is the relationship of the continuity of a surface to the toolpath generation algorithms? Is it true that the manufacturing tools will always have a more difficult time processing higher order continuity or are problems more related to how well the spline mesh was organized rather than the continuity order of the splines themselves? Should it be a goal of the designer to always maintain the lowest possible order of continuity?

Assuming that there is not a direct correlation between surface continuity and toolpath quality is there another way to analyze a surface that will highlight areas that would cause toolpath issues so that they can be addressed before they leave the designer?

Any thoughts would be greatly appreciated!

Thanks,
Jeff

P.S. Our design work is done in NX4 or 6 and NC programming in NX6 or 7.

 

[hudson888]

Depending upon what software you're using to generate your toolpaths the answers may vary slightly. Assuming you're using NX then I believe the geometry is rendered to a faceted form for processing into tool paths but the considerations are pretty much universal.

If the edges don't meet (i.e G0 continuity) then the cutter may traverse the edge my making a small undesired move. You want to avoid that.

If the edges meet at some kind of technically significant angle then the cutter has to change direction. Occasionally this results in dwell marks which highlight the discrepancy and have to be polished out somehow. This relates mainly to tangency or G1 continuity.

Continuity above G1, (i.e. G2 or G3) relates to curvature and produces a kind of smoothness that enhances the appearance of styled surfaces. It rarely if ever has a direct impact on machining apart from being better in a way that generally exceeds the basic requirement for success.

The best way to model for success is to set your modelling tolerances to meet or exceed requirements and use examine geometry to check whether your models meet the set standards.

I work in metric and set the minimum standard for modelling to be a Distance tolerance of 0.02 and an Angular tolerance of 0.50 degrees. The Distance tolerance controls G0 continuity and that Angular tolerance governs the G1. It is set under Preferences Modelling to 0.01 and 0.25, exactly half of the target amount because over the course of creating complex models the tolerances may accumulate as features are overlaid one upon the other.

When I surface I set the requirements a little higher but what I don't do is to try and force continuity by screwing the tolerances right down. I have seen people use distances 0.001 and 0.01 degrees because they think that they're being technically excellent. What they in fact often create in surfaces are internal wrinkles near the edges because the system is trying to force continuity by increasing the complexity of the surface definition. This is a bad thing! It ALSO creates machining defects.



Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[DaSalo]

That is a very helpful reply Hudson. Thank you. The condition that you describe in your last paragraph is exactly the type of problem geometry that we see: Wrinkles, folds, etc. When you examine these surfaces a large number of elements come back as C2. This has been assumed to be an indicator of a problem surface but my feeling, a feeling that has been encouraged by your response, is that it is not the continuity itself that is problematic but rather the context in which it is applied:
If one tries to create a G2 or G3 continuous surface through an erratic spline mesh with tightly set tolerances, problems will abound. The same surface created through a carefully laid out mesh, with carefully considered tolerances that are directly related to the accuracy of that mesh, could be problem free.
In a way none of this matters to what we do; if the geometry is problematic we contact the customer and request a change. The motivation behind my question here is to try to understand this better from a design perspective so that when we do need to make that call we can precisely describe the problem and suggest specific remedies. Thanks for your help with that.

-Jeff