Mini-Foundry to Melt Aluminum, Construction and First Attempted Use

My friend Phil pointed me towards this YouTube video in which Grant Thompson (GT) demonstrates how to build a mini-foundry that you can use to melt Aluminum at home and out of common materials, so I decided to try it for myself.  This post will describe how I built it, largely following what GT did in his video demonstration, and a first attempt at using it to melt aluminum.

The mini-foundry basically consists of a metal bucket with a refractory lining and an air supply hole on the side, a lid made of the same refractory material, and a steel crucible which fits inside the refractory-lined metal bucket under the lid.  The refractory lining is basically an insulating material to retain heat.  With the crucible in the center of the bucket, you then pack fuel (e.g. charcoal) between the crucible and the refractory lining, light it, and then use the air supply hole to force fresh air into the fire, so that with the lid on there is oxygen still provided to the fire.  The steel crucible keeps the aluminum separate from the charcoal / fuel for the fire and should survive the temperatures needed to melt aluminum.  "Should" being the operative word ... !

Starting with a run to the hardware store, I found pretty much the exact items I needed, as GT lists them in the video:

• plaster of paris
• playground sand
• 10 quart steel bucket
• 2.5 quart plastic bucket - for measuring and other uses
• 1" x 24" steel pipe, threaded on each end
• 1" x 31" PVC pipe (bought as 120", cut down to 31")
• PVC coupling - 1 side has threading for 1" steel pipe, other side is slip adapter for 1" PVC pipe
• two 4" U-bolts

I found an old steel fire extinguisher in the basement and I had another one donated from my friends Emma and Shawn - thank you!

I then basically followed the instructions in the video:  First, I made the mixture for the refractory lining.  The ingredients:

• 1.75 buckets plaster of paris (4.375 quarts; 21 cups)
• NB:  the conversion from 4.375 quarts to 21 cups assumes dry quarts, a dry quart being 4.65 cups.  From:  http://www.traditionaloven.com/culinary-arts/volume/convert-quart-dry-qt-us-to-cup-us.html
• 1.75 buckets of play sand
• 1.25 buckets water (3.125 quarts; 15 cups)

I put the plaster and sand in a standard 5 gallon plastic bucket, then added the water and mixed by hand until the consistency was smooth - as GT says, get all the dry powder wet and break down any lumps, and you have to work efficiently - you can't walk away - since it will start to harden in approximately 15 minutes.  I then poured this mixture into the 10 quart steel bucket.  With the measuring bucket filled with water, I pushed it down into the center of the plaster-sand-water mix in the steel bucket, until its top was level with the mix.  I tried holding it in place for several minutes waiting for the mix to harden and hold it in place, but impatience led me to getting a brick and which I then used to weigh / hold down the measuring bucket in the mix.  Here's a picture after the mixture had hardened enough for me to remove the brick:

As in the video, I was able to remove the plastic measuring bucket from the refractory bucket using channel locks to grip the upper lip and then twisting and pulling.  

I was then able to use the 1 3/8" hole saw and a drill to cut through the steel bucket and the plaster to make the air supply port.  Interesting side note, GT notes in the video how easy it is to cut the plaster - not only was it easy, but in my case when the hole saw hit the plaster it liquified!  My friend Doug had a neat explanation - the fundamental crystal structure of the plaster contains water, and a bit of friction from the hole saw broke that down and released the water.  As GT says in the video, the hole cut with the 1 3/8" hole saw nicely accommodates the 1" steel pipe.

Next I built the lid.  As in the video, the ingredients are the same as for the lining for the steel bucket, but a half recipe (0.875 buckets plaster; 0.875 buckets sand; 0.625 buckets water).  Instead of using the wide mouthed bucket in the video, I used the same 5 gallon bucket I used to mix the plaster, sand and water.  When the mixture was ready, I stood the U-bolts up in it as GT described, but I also put a large glass beer bottle in the center to create the hole in the center, rather than have to drill it out later:

I waited a week to remove the lid, however when I tried to remove it, it was stuck.  I tried hammering the outside of the bucket with a mallet, but I ended up hitting it too hard and cracked the lid into 3 pieces.  The wide mouth bucket GT used in the video appeared to be much more flexible and thus probably avoids this problem.  Also, the bottle in the middle may have weakened the lid.

To create the crucible, I took an old fire extinguisher I found in our basement and released the pressure.  It was a chemical one, and when I pulled the pin and squeezed the handle/trigger, not much came out.  I determined the location of my cut by putting the extinguisher in the refractory bucket, and making a mark where the top of the lining.  I started cutting it with a hacksaw, but when I was about half way through, yellow powder started spilling out.  I drained the yellow powder into a garbage bag for awhile, then resumed cutting.  I was able to cut further with minimal spilled powder until I could pry it further apart and dump out the remaining powder, then finishing cutting.  Here's a picture of the 2 halves of what was the fire extinguisher, the bottom half - the crucible - having been used:

At this point I had the individual components ready, I just need to assemble and use.  That opportunity came several months later and happened when our friend Dan F. was visiting, providing much needed help.  Here's a picture of the assembled mini-foundry in use:

Bottom left is the steel bucket with the plaster refractory material lining, capped by the lid in 3 pieces.  The steel pipe is coming out of the bucket, attached via the coupling to the PVC.  Partially visible is an old hair dryer that Dan is holding against the end of the PVC.  Not visible is the crucible which is inside the foundry under the lid.

For our initial attempt, I had the idea to try to use bamboo instead of charcoal, since I had collected a lot of dead bamboo from a nearby grove, and I had noticed that although it takes longer to ignite than wood, it appeared to burn hotter.  Once we got the bamboo lit in the foundry, and kept feeding in a somewhat steady supply of bamboo, we were able to slowly melt some cans.  We discovered that we had to have the lid of the foundry on in order to have some limited success.  Here are some pictures of the hot crucible in the foundry during the bamboo run:

One measure of the importance of the lid was that we only saw the crucible begin to glow red when we had the lid on.  And when the crucible was glowing we began to see some melting of aluminum - poking around in the crucible at one point we did see a small spot of liquid aluminum, but we weren't able to get more than that.  In general it was not as efficient as what GT demonstrated in his video.  At that point we thought the difference was mainly due to the bamboo - because of the length of the bamboo, we could not put the cover on until it had burnt down about halfway, at which point there was a limited amount of time before we would have to take the cover off and add more bamboo.  The bamboo burned much faster than the charcoal, causing greater temperature fluctuations which meant we didn't consistently have the crucible hot enough.  Ultimately using bamboo was unsuccessful.  

Next we made a run to local hardware store and picked up some charcoal.  We found that for our foundry-crucible combination, whole pieces of charcoal would not fit between the crucible and the foundry's refractory lining if the crucible was centered.  We broke the charcoal up by hitting it with the back of an axe head or a mallet, and then we were able to pack the pieces in around the crucible.  It took awhile to get the charcoal lit, but once we did we had much better results.  The crucible got red-hot and cans would melt in a few seconds.  However, we did not get the nice liquid aluminum pool we were hoping for.  When we began to run low on cans, I began to use pieces from a broken tent frame:

These pieces were thicker than the cans, and yet the foundry melted them without too much trouble.  At the peak of the operation, we would add one of these and the end in contact with the bottom of the crucible would immediately go soft and the stick would slowly slide into the foundry.  I was able to add 4 pieces at once and have them all melt.  I'm sorry I don't have a picture or video to share of this!

Although the lid was cracked during the process of making it, it cracked again when I attempted to remove one of the pieces of the lid from the foundry with my bare hands.  The metal was hot and I dropped it and it broke into another pair of pieces.  I didn't get burned, I was just surprised since I had thought the handle (the U-bolts) would be insulated from the heat, since they were embedded in the plaster.  In hindsight, the bottom of the U-bolts were at the bottom of the bucket when the lid was formed, so their metal was only covered by the thinnest of plaster.  For the next version of the lid I'll try to keep them suspended above the bottom so that there is more insulation between them and the heat. 

One important thing we noticed was that the air being blown into the foundry was cooling off the crucible.  The crucible was glowing red everywhere except where the air inlet was.  For future work we plan on adding a diverter to cause the air to be directed sideways, perhaps by effectively capping the end of the steel tube and then drilling holes in the side of the cap.  Another idea Dan had was to fit more charcoal between the crucible and the refractory lining, and looking at back at GT's video, it appears that his crucible has a smaller diameter, allowing him to pack regular, full pieces of charcoal into the foundry around the crucible.  We also talked about making the crucible shorter so that we could pack charcoal underneath it and still have the lid on.  My initial thought was that the charcoal underneath would quickly burn away, and then we would be back to square one; however it now occurs to me that if we create legs for the crucible then as we add more charcoal to the top of the foundry, hot ash and burning charcoal will be able to continually fill in under the crucible and keep it hot.

Here's a video of the foundry in operation:

When we decided to call it quits, we quickly removed the crucible from the foundry and dumped the contents into a muffin tin.  Although it appeared to be mostly fragments of solid aluminum, I did see a small pool of liquid aluminum hit the pan.  Here are pictures of the piece aluminum that formed from that drop:

It is not smooth, however the part that was liquid is visible in its rounded shape.  Another thing we noticed is that a big chunk of material that fell out of the crucible formed into the shape of the muffing tin:

This piece did not appear liquid, however it clearly must have been soft since it took the shape of the container without us apply any pressure / force. 

Here's a picture of the muffin tin containing the scraps of aluminum that came out of the crucible:

At the end of the day, we didn't see a lot of liquid aluminum like we had hoped, but we did see some and we did have a good time, and we've got lots of ideas to try to get that liquid aluminum.

Summary of ideas to try next time:

• new lid that is in one piece
• keep U-bolt handle further away from bottom of lid so they are better insulated
• shorter crucible to fit better under lid and allow packing of fuel underneath it
• crucible with "legs" to allow hot coals to settle underneath it and heat bottom of crucible
• smaller diameter crucible to allow packing more charcoal around it e.g. solid pieces
• diverter at end of air supply tube to prevent cooler inlet air from flowing over crucible directly