3D Parallel odd behavior

[parkerledwards]

Hi Folks!

I'm in the middle of working out some best practices for machining topography into wood. I have a small section of canyon that I'm using as a guinea pig to figure out good toolpaths. Being that these operations are enormously long and detailed I'm working to get the simplest solution possible. For this reason I would like to do everything using 3D parallel understanding that are other operations that may be better suited in certain areas.

I'm running into a sort of 'rounding' issue with these 3D parallels though. The model is quite detailed and sharp, however the operation is calculated in a way that heavily rounds off fine detail. I've tried everything available in the toolpath settings however I can't keep the toolpath from calculating in this way. 

As of now:

Turning off and on Slope, rest machining, fillets, smoothing, and machine steep areas, all with various settings within each have not made any difference. 

Direction/ Up+Down milling are set to both. 

Any ideas on how to get this 3D parallel to machine in detail? 

Attached is a good example, this sort of model-to-toolpath relationship is consistent across the model. 

https://a360.co/3do8nSD

Best

[johnswetz1982]

The only thing you can really do I think is use a really small (tapered) bit with a very small stepover and a small (.0005) tolerance. You can also use [Compare/Edit] once the toolpath is made and look at changing the tessellation there also.

But a word of advice, dont try and do this all at once with a small toolbit you will break it. I would use a 3D adaptive with .010-.015 stock to leave and then go back parallel with a medium sized ball or tapered mill then a final pass with your smallest bit. It is more steps but less than trying to remove a bunch of material with a bit that cant handle it and keeps breaking.   

[parkerledwards]

Hi John,

Here's the same path with the .0005 stepover and .0004 tolerance. Yeah this is the last finishing pass of a 3 tool rough to finish operation. 

Thanks for the tip, I opened compare and edit and don't see a tessellation option, is that what you meant? 

[mattdlr89]

In Compare and Edit there is a Surface Triangulation option. Making this finer may give you a better toolpath in this case. IIRC by default it is set to half of your tolerance. Perhaps have a play with this setting. 

[parkerledwards]

Hey Matt, thanks for the idea. I set the surface triangulation tolerance down to one ten thousandth with the same result unfortunately. It's odd, it just seems to persistently want to not exceed a certain radius of cut if that makes sense. I've noticed that that roundness I have pictured above seems to be the limit for how tight of a move it will make. It's making the entire toolpath unnecessarily smooth. I also experimented with making the tool ridiculously tiny to no effect. 

[mattdlr89]

Do you have an updated file with the toolpath in to look at?

Sometimes tool paths can be deceiving as the point of contact is changing as the tool rolls over an edge. Have you tried to cut the part at all to see what the toolpath actually looks like?

[daniel_lyall]

To get any decent detail in the part you will have to use a tinny cutter like this tool https://www.toolstoday.com/global/v-12539-46280-u.html or smaller.

The smaller the cutter the greater the detail and yes it takes a lot longer but if you want detail.

[johnswetz1982]

To reiterate what @mattdlr89 said, the toolpath looks smooth because of where on the ballmill is the cutting portion tangent to the surface. In other words, it starts cutting with the side "rotates" to cutting with the bottom then continues to "rotate" to cut with the backside of the tool. You will generally never have anything sharp with parallel toolpaths other than the linking moves. 

[BjoernJohnsson]

Hi Parker,

As Matt (mattdlr89) and John (johnswetz1982) already said, the roundness over the peak in you example is from the tool moving over it, always touching the peak. The limitation causing the roundness is the radius of the tool, not the toolpath calculation.

I've looked at that section of the toolpath toolpath, and it is moving over the peak, cutting as closely to the model as is possible with that tool.