Today I was playing with some techniques for making a square blank into a round blank on the rotary. I know that if you move primarily in the X axis (mine is set X+), rotating a bit between passes, you end up with an N-gon instead of a cylinder (i.e. the result is faceted).
So I tried moving primarily in the A axis - spiralling from X0A0 to X5A12960 (or the equivalent). After tweaking that a bit, and pondering the not-flatness, it occurred to me that router bits are not flat on the bottom - the cutting edges are angled a bit so that the center doesn’t rub/burn. But that means if you run the bit right at Y=0 you actually mill a triangular spiral, like a very flat thread, because the center of the path is slightly higher.
I ended up offsetting the bit in the Y direction a little, and using smaller passes, so that I’m milling away from the center of the bit, but it’s not perfect.
I suspect there’s no intersection of “normal router bit” and “desired cylinder” that’s straight… so is this a fool’s errand? Or is there some magical bit or technique that will do this job correctly?
The best way I’ve seen this done is using the spiral toolpath you mentioned, but with a fairly large flattening bit. With a small stepover, it presents a large flat cutting edge that peels away the outside layer of the desired cylinder.
Vectric has a couple of gadgets that do a nice job with a rounding toolpath on a rectangle. The first couple of cuts take off the corners and then it does a spiral toolpath for rounding. Certainly, a piece of cake for Fusion too.
I found the best method for creating rounds it to use a pocket based on the corner to corner of the square dimension, cutting to final round diameter in 3 passes. The first pass will cut some air as it cuts off the corners. The final cut is smoother than using a spiral cuts or offset cuts with a planer bit.
The nice thing about a pocket is you don’t have to allow for the bit offset.
I use a 1/2" 3-flute spiral bit.
Most of my rounds are used in making two piece walking sticks (see Vectric’s Aspire Gallery). I started making sticks in 2000, these were profile cuts and full length.
The last 15+ years they are 2-piece with 3d art carvings. I have completed hundreds of them, gifts to people with medical needs without charge.
10% (could be higher with good sharp bits) usinng a down-cut 1/2 3-flute spiral bit
18K+ rpm
150% - 175% feed rate
Different woods, different settings…
~ .25" cut, 3 passes
Dale
Ah, that explains it. If you try closer to 90% you’d see the effects I was talking about; at 10% you’re only cutting with the tip of the bottom edge, but taking 9 times as long to finish.
You are probably wondering about all the sawdust when I have a vacuum system.
Running a 5 HP vacuum while cutting with small bits makes for an expensive cut.
I made a box using 3/16 Luan plyywood, full size under the rotary.
Then I put a strip of cardboard in the groove on the machine frame which deflects most of the sawdust into the box. It leans outward at about a 60 degree angle so it clears the spindle and dust head.
To make the box with thin wood I use quarter round on the inside corners
You could make one using the Vectric box gadget but I would still add quarter round
True, but my original query was more theoretical - a 100% pass on a flat pocket gives a flat pocket. A 100% pass on a flat cylinder gives a ridged cyilnder. Reducing the pass size makes the ridges arbitrarily small, but not flat, because only the tips of the cutter are at Z=0, with the rest angled up slightly.
