I first read about the grinding machine in an archived usenet article that I no longer have. In it the author claimed to get about 20 grams of aluminum powder per day from a machine he built out of an old typewriter roller wrapped in fine grit sandpaper and partially submerged in water. If anyone has this article I would love to see it again so that I can give proper credit to the originator of this idea.
Based on the description in this article I built several machines, one of which was powered by a fan motor that was too weak, and another which splashed water everywhere. Then by late summer of 1996 I hit on the right formula and produced in a few weeks enough aluminum powder to last me for the rest of high school.
The basic design of the machine can be understood from the following diagram. A spinning roller is wrapped in sandpaper and the bottom of the roller sits in a bin filled with water. The material to be ground is pressed against the top of the roller and the sandpaper grinds away at the material. Flakes come off the sandpaper and settle on the bottom of the bin.
There are many ways you can go about building this. For the roller I have usually used the rubber rollers that come off the ringers of old washing machines. They are over 2" in diameter, have a 5/8" shaft and are often seen in salvage stores. You could also use a wooden rolling pin with a steel axle driven through it, or something much larger like a big sanding drum.
The roller must be geared to a motor to spin slow enough that water does not splash off it. Mine currently spins at 500 RPM and is about right. A larger diameter roller would require less RPM. The motor itself should be at least 1/4 horse-power.
I recommend ball bearings to support the roller shaft, as bushings tend to wear down fast with the applied pressure.
The following shows several completed grinding machines. The one on the left went together in an afternoon for less than 20 dollars. I used split babbit bushings which only lasted a few weeks, but they were cheap so it didn't matter. The steel rings you see are there to add pressure to the aluminum so that it grinds faster. The grinder in the middle is a much more compact version that I finished in June 2000, the black arm is a 2"x4" aluminum tube filled with cast lead, and on the right is a model that my friend Lee made.
| .jpg)
| 
|
You'll notice that all the grinders have a means of securing the material to be ground. I usually just clamp it to a pivoting arm that is attached to the machine. Lee made his with a built in clamping mechanism.
There are many ways to make more of a 'field expedient' type of grinder. Below I've sketched one possibility using an electric drill with a readily available sanding cylinder in the chuck. Another possibility is to rig a belt sander up so that it grinds down on a metal stock. These solutions would work but if you're serious it's worth building a dedicated machine.
Using the grinder is fairly straightforward. The first step is to wrap a spiral strip of sandpaper around the roller. Now there are many types of sandpapers out there. I like to use good quality coarse (24 to 60 grit) belt sander paper. This stuff will last for several days before loosing its edge. The original article recommended the 400-600 grit silicon carbide papers but my experience with these is that they don't last long and grinding speed is slow. I usually drive 1 or 2 small nails in each end of the spiral and then secure it with waterproof tape. After that the tray is filled with water, the metal is clamped to the arm, and the grinding is started.
The grinding rate will depend on a lot of factors, such as grit size, applied pressure, roller speed, and type of metal. In the past I usually got around a pound of aluminum per day, but my new machine with 40 pounds of lead weight produces 50 to 100 grams per hour.
Check on the grinder frequently to see if the metal stock needs to be moved down or if the tray needs water. When the roller spins dry the sandpaper usually clogs up and grinding ceases. After a day or two, the tray will be filled with metal powder. I let the particles settle, drain off the water, and then spread the powder on newspaper to dry.
I've ground aluminum, zinc, magnesium, brass, and even titanium with this setup, and I'm sure just about any soft metal will work. Here are some notes about certain materials:
- Magnesium
- Magnesium is somewhat problematic because of its reactivity with water. A spoonful of potassium dichromate added to the bath greatly suppresses the amount of bubbling that takes place. I've also used mineral spirits instead of water when grinding magnesium with very fine sandpapers, but this is less convenient.
- Aluminum
- Some aluminum alloys bubble a bit in water and others don't. It's no concern if it does but for good measure I frequently add a small amount of boric acid to the bath.
- Zinc
- Zinc grinds quite well but note that the flake is too coarse to use directly in zinc/sulfur rockets.
- Titanium
- Titanium wears down the sandpaper really fast, so it's difficult to make more than small samples at a time. Titanium also requires a fairly large pressure for it to grind at all.
- PVC
- Hard plastics grind okay with this setup and don't need a water bath. Unfortunately PVC pipe is not a good source for pyro applications. This plastic contains tons of filler materials and from my experience destroys coloured formulas.
The Dry Process
You can omit with the water altogether with the coarse sandpapers. This is useful when small samples are needed in a hurry. However, there are a few drawbacks. For one, the finer particles tend to float around in the air and settle all over the place. Secondly, without the water bin as a coolant the metal and the roller get extrememly hot. This hasn't caused me any troubles yet, although it may damage the rubber roller. Thirdly, there is the chance that the sandpaper will clog up, but with coarse grits this doesn't seem to happen.
With coarse sandpapers the grinding machine will produce a pretty broad size distribution of particles. The flakes themselves are somewhat curly and tend to interlock a bit. Just a few minutes in the ball mill flattens them out and makes a denser, more pourable, and messier product.
| Sample A | Sample B | B milled for 2 hours | |
|---|---|---|---|
| Bulk Density | .85 g/mL | .20 g/mL | .67 g/mL |
| +60 mesh | 20% | 1% | -- |
| 60-100 | 28% | 7% | 2% |
| 100-200 | 29% | 26% | 9% |
| 200-325 | 13% | 18% | 12% |
| -325 mesh | 10% | 48% | 77% |
Sample A comes from a recently made batch, and B I've had lying around for several years. They were both made with 36 grit sandpaper but the pressure applied in A was much greater, and I think that is why we see a higher proportion of large particles and a much denser mix. Sample B was ball milled with 1/2" - 5/8" steel balls in a 6" PVC jar. You can see the effects of milling by looking at the photo from before and after.
Naturally you can use finer grit papers for a finer product, but the grinding rate will be slower. I prefer to make the coarse flake and then mill it when I need very fine sizes.
When ball milling aluminum dry it is important to continuously replace the oxygen in the mill so that the powder does not become pyrophoric. This happened to me once when I left the mill running for 4 days while I was away on a hike. After taking the lid off the jar and looking at the powder I went to the kitchen for some snacks, and about 5 minutes later the fire alarm was going off. The flaming PVC jar with white hot aluminum inside was not a pretty site. Now I open the lid every hour so that the aluminum can breath and I have not had the problem recur.
For screening the powder I use a set of stacking sieves that I made from ABS pipe and couplings. These fit snuggly together so that the fine dust doesn't puff out and turn your snot silver. Cut the screen so that there is 3/8" of overhang on the pipe, then apply glue and press the parts together as shown.
 |
 |
.jpg) |
With screens 100 mesh or coarser you may need to cut notches out of the overhang and prefold them, otherwise it can be difficult to push the pipe into the coupling. |
 |
To automate the screening process I simply clamped an orbital sander upside down on a workbench and screwed a piece of plywood to the sanding base. The screens are clamped tight with another piece of plywood and two threaded rods. Make sure that all the nuts have lock washers on them as the intense vibration will loosen anything, and spilt metal powder can be quite a mess. |
Homemade metal powders can be used in just about any pyro application. The broad size distribution of the as-is flake lends itself well to stars and fountains. My favorite mix is this:
| Aluminum | 60 % |
|---|---|
| Barium Nitrate | 40 % |
 |
That's it! The mixture burns extreemely vigorously and I've even made nozzleless rockets out of it. 3/8" unchoked fountains with this formula will easily overshadow 3/4" conventional black powder based fountains. I add 5% dextrin for use in stars or comets, and then use a stepped prime with BP to ignite it. |
Adding up to 20% homemade aluminum to black powder produces a standard Al fountain mix. Experiment with the ratio but I find that more than 20% aluminum seems to suffocate the fountain. The picture on the left is of 4 1/2" fountains and that on the right is a 3/4" fountain, both simple BP/Al mixes.
Unfortunately the aluminum is fairly weak in flash powders unless you ball mill it or sieve out the fines.
The magnesium powder I've made works excellent in coloured stars and produces a deadly flash powder even in coarse sizes. For mag stars I used PVC pipe glue as the binder and chlorine donor. When the mix is partially dry and the consistency of fruit leather, it can be cut into stars with a pair of scissors.
Nobody should have any trouble finding scrap aluminum to grind up, but more exotic metals like zinc and magnesium can be difficult to come by. I got my first hunk of magnesium from a motorcycle salvage yard in the form of an engine sidecover. This was a bit inconvenient to grind because of it's odd shape but at the time I didn't care.
Since then I have got most of my magnesium in the form of sacrificial anodes from scap metal stores. These are most often made of zinc for protecting ships and what-not against corrosion, but occasionally they're made of Mg. I recently bought 200 pounds of the round Mg anodes you see below, for $1 a pound!
