At the Makery lately, Brandon and I [with some help] have been working on a CNC Hot-wire Foam cutter quite a bit like this one. Ours is going to be a bit larger, capable of slicing up a 2×4 foot piece of foam (actually, the X is more like 5 feet than 4). For our Z axis, we are using a salvaged set of rails including a rack and pinion setup, but our X axis is entirely homebrew.
I’d initially lobbied for a belt setup similar to the X-axis on a Prusa Mendel, but Brandon objected on grounds of the cost for a 10 foot piece of belting. He proposed instead using a piece of all-thread as a rack, and a curved gear as the pinion. That was fine, but seemed like the gear would cost more than the belting. In the end, we decided to try making a gear, in the same way that extruder rollers are made for the 3d printers: Hobbing.
Tonight, I got around to trying to make the gear. We’d settled on 3/8-16 all-thread, as it’d be sufficiently beefy as to not bend with a gear pressing against it, and to not sag under the weight of a mostly unsupported 5 foot span. I grabbed a piece of (approximately) 1″ round aluminum from the scrap bin, and drilled a 3/16″ hole through. I bolted the cylinder to a bearing I had lying around, and chucked it into a V-jaw in the mill vise, with one end floating free. I chucked up a 3/8-16 tap in the spindle, and set the speed as low as it goes (I around 500rpm, I think).
From there, I aligned the tap so that the centermost full thread was parallel to the bolt through the work, so that I’d be cutting only on full threads. I used the Y-axis of the mill table to position the tap along the length of the work, and advanced the work onto the spinning tap by slowly feeding the X axis as the work turned. (The work spun freely in the bearing, powered by the tap cutting into the aluminum, like when a board lifts as a wood screw is driven.)
I initially had a bit of trouble with the work flexing in the chuck (as I was only supported from one end), but I overcame this by grabbing the outboard end of the bolt with my hand and keeping the bolt parallel to the vise jaws manually.
Overall, the process worked really well. The tap cut nice, deep, uniform teeth into the aluminum. If I were going to do it again, I’d find a better way to hold the work (supported by both ends) in the vise; I’d also pick a better piece of material, as I didn’t bother turning the surface imperfections of out this one before I started.