It’s been a while since I played along during community challenges, so I’ll try something while you’re all working hard on your entries.
That puzzle cube art @WillAdams linked to sparked motivation to try a simpler one myself.
To cut corners I looked up puzzle cube ideas on Thingiverse, and found that one, with only 6 pieces:
Just importing the corresponding STLs was no fun, so I recreated the parts ni Fusion360 in a CNC friendly way, and with the intent to have flat sides when the cube is assembled, and filleted edges on the parts just where necessary to ensure machinability (I chose a 3.5mm radius on all filets, such that a 1/8" ball endmill would be able to cut them).
Here’s the assembled view,
Here’s the explosed view with parts layed out for the cut,
And here’s the victim (1.2" bamboo trivet, yes it’s a pity to cut that sweet endgrain trivet but it’s been sitting on a shelf for 6 months, I need to do something with it)
I’ll probably have to paint the parts different colors, we’ll see, I did not have thick enough stock from other woods.
Then I realized I don’t have any endmills with a 1.2" length of cut, nor any 1/8" endmill with enough reach for some of those fillets, so I ordered some and now we wait…
EDIT Aug 20th: received the endmills. I proceeded to lay out the parts as best as I could (close but leaving a large enough gap beween them for some adaptive clearing fun with a 6mm endmill)
I chose to do the roughing using a 3D adaptive toolpath, 6mm 3FL square endmill, 20k RPM, 90ipm, 0.04" optimal load, 0.6" depth of cut with spiral plunge. 0.5mm radial and axial stock to leave.
My stock was actually 30.15mm thick, I ran a surfacing pass to bring it to exactly 30mm.
The roughing pass went great. I just LOVE bamboo and how clean it comes out of the machine.
Check out that tiny shard that was left over by the toolpath, and still standing straight…what are the odds!
Anyway, that’s it for tonight, finishing pass coming up tomorrow.
EDIT Aug 23rd: well, after this little mishap, I had to redo the roughing part (luckily the stock is large enough to give me 4 tries at this). I then used two parallel finishing toolpaths, one at 0° and one at 90°. BUT since I don’t have a long enough 1/8" ballnose, I used a long 1/8" square endmill. Given the geometry, it could have worked, but is not optimal. Also, I had a number of problems with my G-code sender setup, and had to re-home between each toolpath, which did not help. It was still good to complete the cut:
Some walls are just horrible, others are fine:
And the resulting cube is… “yeah, I’ll give it another try”
I could probably spend time sanding the pieces to a better fit, but I won’t, I’d rather spend time trying better endmill/toolpaths combinations our of the machine. Now to find a 1/8" ballnose with a long enough reach…
EDIT: well this is the last day of the challenge, and I wanted to give this another go, but I don’t have a long enough ballnose yet, so I decided to do everything with the long reach 6mm square endmill, and use a ramp finishing toolpath:
which turned out to give a near perfect finish (considering the sub-optimal use of a square endmill for finishing…) right off the machine,
And a better fit:
I would still need to work on my tolerances, I have enough stock left for ONE final try.