Does CAM use NURBS to generate toolpath? 2

[mechie23]

Hi,

I have a doubt regarding toolpath generation in CAM software in general. Modern CAD/CAM systems represent surface and solid data in NURBS form. But while generating tool path does CAM software use NURBS data directly or does it convert NURBS data to Triangulation format? I heard that some CAM software convert in this way. If they do, what is the necessity for such conversion?

Many thanks.

 

[Tingsryd]

I believe NX does a combination of meshing of nurbs faces as well as exact offsetting of prismatic faces. It generally considered to be a fast way to generate CL data while giving the user the flexibility to control the tolerance, and the generation time, of the CL data. I know of some programmers using .005" intol/outtol for fast, low tolerance routing applications while other programers are using tolerances from .0005" to as low as .00002"-.00005" to attain very accurate cutter paths with excellent surface finishes.

NX8

 

[joycejo]

In NX there is an option for the motion output to be set to Nurbs. You will also need the post processor setup to format this correctly for your machine control.



John Joyce
N.C. Programming Supervisor
Barnes Aerospace, Windsor CT
NX6 & NX7.5

 

[JohnRBaker]

Please note that I think you will find that the option for outputting toolpath data in NURB format is specifically intended for support of so-called 'high-speed machining' processes, which requires, of course, machine tools and controllers explicitly designed for this very purpose.

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.

 

[Tingsryd]

Perhaps mechie23 should clarify exactly what information he needs. At first it sounds like he wants to know about NURBS nc file code output during post processing but the end of his question sounds more like he is questioning the difference between triangulated meshing vs. exact surface offsetting which happens during the tool path calculation.

NX8

 

[mechie23]

Many Thanks to Tingsryd & joycejo.

Can you explain what is this conversion to triangulation?

Also will the parameters generated by NURBS toolpath same as those that are defined for NURBS surface in CAD? Or will it reconstruct a new NURBS surface? I mean if I know the NURBS parameters from CAD model, can I directly use them in altering the NC code (using appropriate G code)?

Thank you.

 

[joycejo]

The machinie must have the "High Speed" option to acccept nurb format. The only on I have seen is on a Makino. I would recommend looking at the machine tool builders manual to review the imput required. Frankily I don't think it will be the same as what the CAD system uses as the tool radius must be taken into account.



John Joyce
N.C. Programming Supervisor
Barnes Aerospace, Windsor CT
NX6 & NX7.5

 

[markrief]

I believe the original question is why many CAM systems facet/triangulate/tesselate the original geometry instead of offsetting it directly.

The main reason is reliability, especially with complex topology or poor geometry with issues like gaps between edges, overlapping surfaces, etc. This does not affect the accuracy of the tool path.

In NX, we use both direct offsetting and faceting, depending on the geometry, operation, and tool axis.


Mark Rief
Product Manager
Siemens PLM

 

[mechie23]

Hi, thanks to all.

My original question is what markrief answered. Sorry for not being clear. However I got doubt on its usage on machine tool when joycejo posted his first reply. So I added that as well. I agree with what others said about requirement of machine tool. But I think more specifically it depends on control - as I've seen it on Siemens 840D few years back.

Now it's clear to me that to accommodate poor geometry CAM software does triangulation or tesselation. But why operation and tool axis influence the choice?

Thank you.

 

[markrief]

The numerous reasons why faceting is used are too complicated (and often proprietary) for me to explain in public. But I assure you that we use whatever methods we think are best to get the most reliable toolpaths with acceptable performance.

There has been a lot of research in this area, so maybe you will find the details you are looking for in a published paper somewhere.


Mark Rief
Product Manager
Siemens PLM

 

[Toost]

A good guess, without knowing anything about the needed math, is that if you calculate based on tesselation instead of the underlying math , the operation can ignore things like edges and faces of different types. Imagine a model with one analytic planar face , one complex curved B-surface and a Edge blend in between. ( Three different math types) Then you want to cut these three in a pattern which has nothing in common with the U and V directions of the model. As said, i would guess that it's far easier if done on tesselated representation.

Regards,
Tomas

 

[mechie23]

Thank you Toost and markrief. Useful info. I'll also try to search some research papers on this.