cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Have a PTC product question you need answered fast? Chances are someone has asked it before. Learn about the community search. X

Nc tool path cpu usage

JP_10024045
3-Visitor

Nc tool path cpu usage

Hi,

I am a student at a university working on a year+ long project that requires us to cut out a semi complex aerodynamic part (around 2x0.5x0.5m for reference in size) on a 4th axis cnc machine. I am working on the nc tool paths to cut this part however I have gotten to the point where it takes 15+ minutes to calculate the tool paths (around 5 cut line operations) which makes it very hard to diagnose problems and becomes very time costly. I have heard about creo parametric not being able to use all of the cpu cores and as I am in creo 6 (for compatibility and IT reasons) I wanted to know if the newer versions of creo support using more cores for nc tool paths or if there is a way to get around it, I have a Ryzen 2600x (6 cores, 3.6 GHz) and 16GB of ram)

9 REPLIES 9

There have been discussions of this in the past, and a quick search brought me to this document:

 Article about multithreading 

It doesn't mention manufacturing at all, unfortunately.

I have found that I can often speed up toolpath calculation significantly by defining specific part surfaces as check surfaces rather than including the entire model. Perhaps that is something to consider. For example, if I'm surface milling it will be much faster if I specifically pick the surface and any adjacent surfaces, for a total number of check surfaces in the tens, rather than the hundreds of surfaces it will use if stick with the default "use reference parts" or whatever that setting is.

Thanks for your response! 
unfortunately this doesn’t seem to help much, I have surfaces in the tens selected with check surfaces however my suspicion is that the complexity of the surfaces (not the number of them) is the main problem. this is a part that will be used as a mold to create an aerodynamic part so I can’t just simplify it’s shape or suppress features that take up computing power. My main concern is that when these toolpaths are being computed, about 25% of my cpu is being loaded and I’m not sure if this is simply because it’s not possible to utilise all of my cores or if it’s because I’m using an earlier version of creo. I have looked at a few discussions about multithreading including the one you linked but these are from years ago with earlier versions I’m using.

As far as I know, they never really utilized multiple cores for NC path generation. I don't know if it's a type of problem that would lend itself to being parceled off to multiple cores. A toolpath calculation would seem to be something that is algorithmically "calculate this position, based on that, calculate the next position, etc." It's not like solving a whole warehouse-full of linear equations where matrix inversions can be chunked out to a bunch of "calculators".

Ah, complex geometry. Have had tons of fun with that. I used to design the internal passages of airfoils. They are such an adventure to work with. In fact, another thing I've had cause lots of slowness in path generation was when I got surface geometry from a customer that was horribly discontinuous at surface boundaries (terrible tangencies, gaps, etc.). I'd see it when I was simulating the toolpath.

Do you have much experience with using custom workpieces? To be more specific on my project, we are building a human powered vehicle to hopefully take to battle mountain this year so we are basically cutting out a big aerodynamic tictac, to save a lot of time and material cutting I have made a sort of oversized Lego version of the part instead of just cutting down a big block of foam. Do you know if the complexity of the workpiece/stock model changes the toolpaths and could possibly make it slower?

I've used everything from simple blocks to a previously machined rough part as workpiece/stock model. For simulation, when using ModuleWorks or NC Check, you can bring in a previous simulation you've saved to check if your new toolpaths are going to work the way you expected, etc.

The complexity of the workpiece, as far as I know, doesn't affect the toolpath generation. I think by default your workpiece is ignored during the toolpath calculations.

Daniel_E
6-Contributor
(To:JP_10024045)

It does seem to use multiple cores on my system (running creo 9).  When I simulate, it looks like at least 6 (of 16) cores light up. 

 

But have you tried increasing the "interpolation tolerance" setting?  This speeds things way up for me... or if I set it too small it slows things way, way down. 

Simulation is different from the calculation of the toolpaths. The toolpath calculation is done first, to write a file, then whatever simulation program you have chosen uses that as its input. I don't know if the different simulation packages are using multiple cores or not. Maybe they do?

What the original discussion was about is the calculation of the toolpaths. For complex parts this can take lots of time, particularly for surface milling. It will just slowly tick through percentages, crawling towards 100%...

Daniel_E
6-Contributor
(To:KenFarley)

I see.  Thanks for clarifying.  

Hi JP, I invite you to try our new toolpaths released in Creo 9, both for 5 Axis Roughing and 5 Axis Finishing (Geodesic). You can share your models with me if you like and we can evaluate it together.... in the mean time, if your problem is that you are adjusting the toolpath and you know that very likely you will need a few iterations, you can loosen up a bit the machining tolerance parameter. Is not that this will solve your issue, but can speed up the computation a bit.

my email: jcoronado@ptc.com

Top Tags