[Rodent]

I'm in the final stages of bringing a new product design to market that is on the older MadCAM 4.3 platform (I wish I could migrate this to 5.0, but that it not an option I'm allowed to consider until after initial product release)

I discovered last night that pocket tool paths are suddenly being computed in a totally different way than they were earlier this week. Previously all pockets cut into the center of a pocket area with a ramp, and then 'spiraled' out to the contour edge. This has been the stable behavior for the past 18+ months. What is now happening is that the ramp initiates in the center and then spirals out until it is about a cutter width away from the contour edge, then the tool path lifts and ramps back into the pocket with two spiral paths to cut the outermost boundary

I have not (knowingly) changed any of my Rhino or MadCAM settings, have not installed any Rhino or MadCAM updates, and the tools selected are unmodified since creating them eons ago. If I open an earlier model, the tool path generated for a pocket is like this new pattern instead of the previous way.


Does anybody know what could have changed in my settings to cause this behavior?

I'll add a couple before/after images and a reference model if need be. Let me know


Thanks for helping me get back on track!

All the best,

R

[Rodent]

added image of multi-lift pocket and reference model

all the best,

R

[svenakela]

I tried to reproduce your tool path in a mm world, but I couldn't. All paths became one down and spiral out to the edge. Can someone who work with inches try the same?
Does it matter if the path has a lift? Except for a bit longer cutting time though.

BTW, it seems that your mesh settings are untouched. Change them as written in the help file (I've said it many times, overwrite your default template as well). If it doesn't change the behaviour you will at least get a finer tool path.

[Rodent]

Thanks for taking a look at the part and for the reminder to also update the default template's settings so that it's impossible to forget to mod a file's settings

I changed my settings, but it did not resolve the lift issue. If this was only for cutting 1 - 10 units I'd guess I'd live with the added cutting time. Unfortunately this is for a production unit that will have hundreds of units cut, so the added minutes of cutting time have a cumulative impact on production potential

I would migrate to 5 today if I could, and look forward to having this part (and a couple others) out on the floor so I can turn my attention to testing our parts/workflows in 5 so it can be certified for all future work

all the best,

R

[svenakela]

I tried it in Rhino 4 with MC 4, I didn't get your result.

[Rodent]

I found an interim workaround that will get me to production this weekend

I spent time creating offset curves of different values until I discovered one value that didn't generate the pocket with a secondary lift. This lead me to believe that I could utilize my original pocket size and play with the Step Over value as a way to eliminate the secondary lifts. For the main cavity, I needed to adjust the Step Over from my standard 0.200" with a 0.50"D mill to 0.205". Since I'm cutting wood, this nuance should be a non-issue as far as tool feed/speed goes

I will now address the Step Over values on several of the smaller cavities which also have secondary lifts to verify I can massage the settings to acquire a single spiral cut pattern for each cutting pass.


If this shed any insight as to the root cause of this unexpected behavior change, I would still be most interested to know.

all the best,

R

[Rodent]

playing around with the overlap values got me where I needed to be. worst case was a .220" overlap vs. my standard .200" on a .500" spiral cut bit. since I clean up the visible cavity edges with a profile cut. the small bit of tearout O got was manageable while working and nonexistent in the final product

here's a couple images of the new design I was working and how it turned out on the first unit in each configuration. these parts are wiped with Naphtha to show how the wood figure will look when sprayed with a clear finish

[svenakela]

Damn! That five stringer looks awesome! What wood is it in the body?

[Rodent]

thanks

the top and back are bookmatched Camphor burl, the core is Mahogany, and there's a .125" thick Walnut accent laminate separating the top/back from the core. When cut, detail sanded, and a clear finish is sprayed, the resulting look is like this Redwood and Maple body in the attached image of a different model with 22 frets

Before Rhino/MadCAM I used to carve each body of this design by hand, so each instrument was unique to itself. with the new CNC rough carved version I can offer this at a significantly more consumer friendly pricing structure and also complete 4 bodies in less time than it took me to make just the one

MadCAM plays a significant part in my ability to bring more product to market, and to do so with an increased level of dimensional tolerance consistency between units that is only possible with good designs and CNC fabrication. I have gone from each unit being an island to itself to complete interchangeability between parts

all the best,

R

[John Coloccia]

FYI, this thread popped up in "related threads" when I submitted mine. I'm using Madcam 5, and I think the behavior might be even stranger there.

http://www.cnczone.com/forums/madcam...ool-paths.html