Last updated 20 Feb 2017
This is a working document. Many of the pictures here are incorporated into the YouTube video here.
Here are some 24" and 36" gear wheels. The 36" wheels only have 8 segments versus the 20 segments of the 24" ones. Notice how much wood would be wasted if you cut off the corners of the 36" ones to make a circle.
It stands to reason that the more segments you have, the less waste you will have and the closer it will be to a true circle starting out
The joints are overlapped and staggered for maximum strength. Most of the strength will come from the side to side bonds of the wood and not the end to end junctions
Stacking the segments in groups of 5 or 20 makes it easy to tell at a glance how many wheel's worth of material is there.
At first we drew a circle on the table with a large compass, but later learned that if the cuts were done well enough, a simple ratchet strap tensioned up would bring them very close to being round
Using a tape measure to do diameter checks (similar to doing a diagonal check on a square to make sure it's square) at 0, 90, 45, and 135 degrees allowed us to fine tune it further and just nudge it the 1/4" it might be out by until the glue could dry
Every horizontal surface quickly became valuable (when are they not?). We tried to have as many drying/setting up at one time. The cold shop meant that what should have taken 10 minutes took over an hour
Eventually we learned to put a propane heater under the cast iron table saw table and this sped things up greatly.
We wanted the discs to dry flat, so we weighted them down as the glue set with whatever was handy. Some aluminum ingots and blocks, gear cutters, Lovejoy couplers, metal shaper cutters
We got all the tables at the same height, so we could glue on the table saw and let it get a basic bond with the glue. While it was too weak to pick it up, we could still slide it to the table to dry more while we did the next one
5 groups of 20, or 5 gear wheel's worth of segments stacked neatly here.
The involute gear cutters started arriving from various ebay purchases. They makae great weights.
The block in the center is the center point that we spun the compass on. You can't see it here, but there is a pencil circle on the table. We were initially making all the pieces tangent to the circle at their midpoint, then found that to be unnecessary.
Here are (7) 36" wheels done. Also shown is the compass as well as the datk stained wooden pulley from several years ago that we made with basically the same technique (minus the hub)
The gear wheels are not as fragile as they look, despite some gaps in the joints. The glue dries very strong and will get most of its strength from the side to side bonds. It will essentially be one solid piece of wood when we're done.
I knew that someone had to have done the work for this already, and sure enough, google didn't disappoing. There is an online segment calculator at http://www.woodturnersresource.com/extras/projects/segmentcalc/index.html I guess woodturners use this to make bowls.
15 Done of each 24" and 36". This will be enough for 3 gears of each, since we decided to copy our earlier design of the pulleys by having them 5 layers deep with 3 layer spokes centered on the three center layers.
Here are the actual numbers we used for the 18" wheels. Using 20 segments for a wheel this small makes them too small and unwieldly (and your hand has to get too close to the sawblade for comfort...)
This is what the 16 segment 18" one would look like. It's still pretty round with minimal wood waste
Now that we had a test ring to look at that looked good, the mass production would start...
Rather than work in the cold shop, I took the segments home for the 18" wheels. Here Minnie is playing with the automatic centerpunch
After checking the quality of the joinery between segments...
Minnie gives her approval. This step is optional.
One has to work fast since the time it takes to coat both ends with glue, nudge them together and go all the way around, one is pushing up against the 5-10 minute set time (depending on if you're using Titebond II or III)
Here are all 15 done in one evening (prior evening)
Next we had to notch out half of the thickness for the spokes to go in (which also had half of them notched). See the video for a better idea of this process.
Here is a sample hub done. When it's done, it will have another layer behind it also throughbolted and a center layer that gets sandwiched and glued between the two outer ones.
Stephen tried to dodge the camera, but I got proof that it was his arm holding the spokes (and the arm that assembled them too)
This should give a better picture of how everything goes together. It's missing the center layer for teh spokes, but you can see how the lap joints at the end overlap
Here's a closeup of the hubs after machining and drilling. Stephen asked me to include the drilling into the CNC milling program next time. I probably would have to agree that that's a good idea...
You can see some porosity in the castings in the cross section where it was milled. This was probably because we overheated many of the melts because we were multitasking and got back to it when it was red hot. (see later pics)
For the final glue-up, Pat had the idea to make a tent with a mover's blanket and heat the table saw table gently with a propane space heater. It made a nice dutch oven of sorts and greatly speeded up the drying process.
Here we go into the making of the cast and machined aluminum hubs
(more pics and captions coming soon!)