3D Carving and Cleanup

[Brad Shipton]

This is a work in progress. I cut this with a roughing pass and final pass with a 1/4"x1/8" tapered ballnose (Vortex 2215). In the final pass I used a 15% stepover (time problem) and that was too large. I am curious what others might use?

For finishing prep it seems like it is just a matter of a lot of hand work. I have tried different dremel bits, and other than some of the sanding bits it does not seem to help much. The rifling rasps, and sand paper seem to be the best to avoid mottling all of the nice details. I would like to hear any tips if you are willing to share. I have done quite a few small signs, and those are easy. The larger pieces I am doing lately a bit more time consuming to get ready for nice glazing highlights.

Thanks too all.

CABIN_TABLE -sm.jpg

[bobby milam]

I've had some success with a flapwheel in a high grit but for the most part I haven't really found anything that worked well in a reasonable amount of time. Some times it is faster in the long run to let the machine run a second pass if it is going to need too much of a cleanup IMO.

[Jim Becker]

While I understand the time issue, I'd probably run the finish tool path again with a 9% step-over with the grain and about a thou or two deeper to clean it up. I've moved to using an Amana 46286-K for 3D surfacing for "carving" type work, such as the recent "magnolia" themed commission I posted about not long ago. The .0625" tip does make for longer run times, but the surface is excellent.

[Brad Shipton]

I have a 1/16" tapered ball nose as well, but I dread using the little thing. I will give that one a try next time. I have another carving in the machine now and it will be done in the am. I reduced the stepover in this case was .008", so it will be better. I might run the current carving again with the 1/16" anyway. I noticed I have lost some details I would like to keep using the 1/8". I am a bit paranoid about leaving the machine unattended, so 24hr+ carve times are difficult to plan with a real job.

This one is 80% fixed now. I found the little plastic abrasive dremel brushes actually work quite well, and you can use them to bring out details too. https://www.dremel.com/en_US/product...ssories/538-01 I tried the flap wheels, but it seemed quite easy to sand the details away.

Thanks all.

[Bruce Page]

For fine detailed carvings I use a 1/16 tapered ball nose with an 8% stepover. It makes for a long finish cut but I’d rather let the machine do the finish work if I can.
This GOT mirror frame needed minimal cleanup, mainly just a light touch up on the bottom edge.
For a large piece like yours the 3M Radial bristle discs and a Dremel work really well.

[Richard Gonzalez]

Have you tried the bristle type wheels (not flap type)?
5FA79273-B5AD-4342-B0B7-EF592F0FBC68.jpeg

I have all 4: 180, 240, 320 and 400 grit. They don’t seem to eliminate detail, just get rid of the “nubs”.

[Brad Shipton]

I know you are correct Bruce. Those little 1/16" bits look so fragile and so very sloow.

Richard, I looked at the bristle ones, but did not buy those. I will give them a try.

[David Falkner]

Have you tried the bristle type wheels (not flap type)?
5FA79273-B5AD-4342-B0B7-EF592F0FBC68.jpeg

I have all 4: 180, 240, 320 and 400 grit. They don’t seem to eliminate detail, just get rid of the “nubs”.

Where do you get those, Richard?

David

[Jim Becker]

I have been able to successfully stop and restart in the middle of a long 3D cut job, but it's not ideal...and the machine needs to stay "on" (electrically but not with the spindle running) for best results with a restart at a later time.

[Ted Reischl]

Stepover. Cusp Height. How much?

A while back this drove me to create some charts that show how various step overs effect finish.

Pay attention to the "cusp target". That is the height of the material that will be left when stepping over a ball nose.






The original thread is on the Vectric forum here:

http://forum.vectric.com/viewtopic.p...228634#p228634

[Brad Shipton]

Jim, did you do that with a stop and code edit? WinCNC outputs the line in process, so it seems easy to edit the code to the delete the lines and then re-start. When reading I see some packages allow you to break the code into time targets. I can't imagine the targets are all that close unless you can enter the acceleration of the various axes. With 3D work there is so much movement that the acceleration becomes key in time estimates.

Thanks Ted. I guess the question is what is a reasonable cusp target for a nice finish? Do most avoid sanding? Sanding the details was not actually all that bad, but sanding the flourish's was not fun.

[Gary Campbell]

Brad...
I haven't done but a handful of 3D projects over the years. That said, each of them was a good paying commissioned job. I was always a believer of going with a low stepover value at a longer file run cost vs. a quicker file run and more sanding. Reason was: I can do other tasks while the cnc is running, I cannot when I am sanding. YMMV

[Jim Becker]

I agree, Gary. The one commissioned 3D job I've done so far paid nearly 10% of my original machine startup cost.

Brad, for those few times I just plain had to do the stop/start thing (like having to take my daughter to work) I waited for the machine to lift up for a rapid move and aborted, taking note of the file line number. I left things sit like that (spindle off) until I returned and then used the WinCNC re-start, stepping the line number back a small amount so that it would result in a continuous job. I didn't exactly like doing this, but really had no choice when things went WAY long on me and I had to take care of obligations.

[Brady Watson]

A 1/32" ball is about as small as you can go before you can't hear it cutting over a quiet spindle. When you get down into the 1/64" and smaller end mill sizes, you can't hear the cutting sound - you have to visually look (sometimes with a flashlight) to see if that little puppy is still intact...otherwise the machine will keep going and no carving will be done because the tip was snapped off somewhere 'way back there'...when you should have checked it

A lot of guys don't use bits smaller than 1/8" because they don't properly rough out to clear for the smaller diameter tools that they say 'snap too easily'...If I have to run a 1/32 or 1/64" ball to resolve detail, I also run progressively smaller roughing operations, including say a 3D finishing toolpath with a 1/16" ball @ 25-30% stepover to clear away the big chunks for the smaller tools. It takes a lot longer...but you don't break tools. Only by hammering on your CAM previewer can you really see what needs to be done. Don't take anything for granted...Quality extra long ball end mills are $40+.

When you get below 1/16" diameter...the grain of the wood really plays a role in how closely it will resolve the detail. Some woods look like cable bundles in comparison to the detail you are trying to put in.

Here's a little frame I scanned and made out of scrap walnut. I literally just super glued together and carpet taped to the bed (testing machine performance)...Final 3D Finishing toolpath with a 1/64" ball @ 0.0015 stepover. It's about the size of a switch plate cover.

[Bill George]

Have you tried the bristle type wheels (not flap type)?
5FA79273-B5AD-4342-B0B7-EF592F0FBC68.jpeg

I have all 4: 180, 240, 320 and 400 grit. They don’t seem to eliminate detail, just get rid of the “nubs”.

Richard where did you get those???