Aluminum Recycling Micro Factory

What if you could recycle your aluminum cans at home and could process a large volume of cans semi-autonomously?

Similar projects worth following
The concept behind this project is to explore the idea of directly recycling aluminum at home.

While electric forges for melting aluminum are fairly common, there are not any available that are capable of processing soda cans with minimal human interaction.

A user should be ideally be able to turn the Aluminum Recycling Micro Factory on and simply feed cans into a hopper.

The ArcticMicroFactory sub-reddit was started because I waned somewhere to post information that I found to track various examples of applied methods for recycling aluminum, copper wire and plastic.

In an ideal world, recycling should start and end at home. The Aluminum Recycling Micro Factory is the aluminum recycling portion of a larger, longer term project to make recycling possible in the Canadian arctic or anywhere that conventionally available recycling operators exist. That project is located here.

If you're interested in contributing, I'd be happy to collaborate. I could definitely use the help.

The Aluminum Recycling Micro Factory project will be borrowing a lot of hardware design principles from an open hardware design project called Precious Plastic. For the purposes of design, ideas will be borrowed from the Precious Plastic project's open hardware basic shredder and scaled down to a small enough size to fit within an automated backyard bulk aluminum can processor.

Here is the Precious Plastic community page. The CAD design files for a basic shredder can be found there.

However, I will also be releasing the design that I am working on as as an open hardware design as well.


Stand in for a aluminum soda can shredder. I'm trying to keep this to parts that anyone, anywhere can source themselves. In this case, I used the dimensions for commonly available leaf spring steel. Which is hardened and tempered. 80mm wide by 7mm thick. It'll need to be cut to 100mm to 107mm and welded together. However, the gears and a method for shredding the soda cans are both something I'm still trying to figure out.

application/x-zip-compressed - 60.08 kB - 06/02/2020 at 06:06



The model for the 2020 Corner Bracket. Can be exported as an STL and printed. Then, with the Aluminum Recycling Micro Factory, cast with PLA.

x-zip-compressed - 31.66 kB - 06/02/2020 at 06:05



Updated with 2020 90 Degree Angle Brackets, additional reinforcing corner sheets and the T-slots and machine screws to be added later.

application/x-zip-compressed - 224.07 kB - 06/02/2020 at 05:49


Bill of Materials.ods

Bill of materials so far. I live in the arctic, so shipping prices are pretty high.

spreadsheet - 14.08 kB - 05/29/2020 at 03:50


  • 4 × 1 Meter lengths of V-Slot 2020 Extrusion I've found the average price for 1M lengths to be about $15, shipping is about the same.
  • 1 × Windshield Wiper Motor You can salvage these from old vehicles
  • 1 × 8' x 4' Sheet of 3/8" plywood This will need to be cut to various dimensions with a skillsaw or table saw
  • 1 × 2kg Graphite Crucible
  • 1 × 1800 Watt ZVS Induction Heating Board Module

  • Most of the parts are in!

    hominidae07/29/2020 at 12:10 0 comments

    So, most of the parts for the Aluminum Recycling Micro Factory are in!

    I haven't posted much because I took three weeks off this July but now that I'm back most of what I need has arrived in the mail. Will be posting more updates soon.

  • Scrapped the use of a 20 lb propane tank

    hominidae06/02/2020 at 05:39 0 comments

    Updated the FreeCAD model.

    Removed the 20 lb propane tank. Will use that for an accessory resistance heating forge.

    I've ordered 4x hot swappable HP 750 Watt Server Power supplies (PN: 660183-001) and will instead be using induction heating to melt aluminum in a crucible instead.

    The general idea is going to be that I can use a crucible which can raise and lower into the work coil, melting the aluminum and using 3D printing hardware, dump the molten contents of the crucible.

    One potential issue is that the baking ingot trays would need to be heated. However, what if they didn't need to be? What if a steel pipe was used instead?

    The steel pipe would obviously get quite hot. But, a steel pipe could also radiate quite a bit of heat naturally. One issue with molten aluminum is that it is possible to trap moisture and steam causes an explosion.

    But, if the only direction that the molten aluminum had to go was up, would it splash out of a confined steel cylinder?

  • Different heating approach

    hominidae05/31/2020 at 23:12 0 comments

    I've decided to take a different approach for melting the aluminum.

    This is based on a couple of factors. The first is that we can use the principle of induction heating to raise the temperature of the aluminum well past it's melting temperature.

    The effect that this has on the aluminum is that any impurities are burned away in the process. This is important as we don't need a separate heater for removing the BPA epoxy liner and the paint on the aluminum soda cans.

    It also allows for a much safer method of melting. Raising and lowering a graphite curcible within the work coil means we can effectively create a system that requires no human interaction.

    Remember that the whole the whole point of this project is to build something that an operator can dump a bag of aluminum cans into the hopper, turn the unit on and leave unattended. The system should ideally be capable of taking care of every step in the process on it's own.

    A Raspberry Pi and/or an Arduino controls the shredder, feeds the shredded aluminum into the crucible, and turns the induction heater on and off, all while monitoring current to detect jams and the state of the metal in the crucible.

    An interesting phenomenon exists with the induction heating method. Once the shredded aluminum is dumped into the crucible, the current rises until a certain point. When the aluminum melts, the current drops. This can be used as a signal to feed more shredded aluminum into the crucible.

    In order to detect when there's enough molten aluminum to fill an ingot, a load cell will be needed to detect the tare weight of the crucible before being filled and once that point is reached, stepper motors will lower the crucible platform and tip it in order to dump the contents into a mold.

    This process is quite a bit safer in terms of semi-automated operation,

  • Requirements for the electronics

    hominidae05/28/2020 at 23:36 0 comments

    This is a brief description of the electronics requirements:

    First, we'll need a Raspberry Pi or an Arduino to control things. Among the system's being controlled by the Raspberry Pi or Arduino will be a UL listed voltage relay and PID controller to supply power to the kiln heating wire inside the forge. This will have it's own temperature sensor to measure the temperature inside the forge during operation.

    Next, a motor controller will need to supply power to the Windshield Wiper motor shredder and it will need to have a sensor to detect the current being drawn by the windshield wiper motor. The current sensor will need to periodically measure for a surge in current which would indicate a jam. The motor controller will need to go through a jam clearing process that involves reversing polarity to go into reverse and back to forward and repeat this several times until the jam is cleared. The Arduino or Raspberry Pi will need to raise an alarm condition if the jam cannot be cleared automatically and user interaction is needed to clear the jam. Safety features will need to be setup so that in the process of clearing a jam any operator of the system is not put in danger of losing fingers or electrocution.

    Additionally, two stepper motors will be needed. One will be for moving a conveyor belt or gate for the shredded pop can material into the forge inlet and another stepper motor will be needed in order to raise and lower a lead screw that tilts a graphite crucible inside the forge and dumps the content of the crucible into a steel spigot that the molten aluminum pours into a revolving tray of bread pan baking trays at aluminum recycling miro factory at it's base. This might need to be another stepper motor, but it can also likely be accomplished using a more affordable low RPM DC motor and a position sensor.

    The entire system operation would entail pre-heating the forge with the PID controller, dumping the contents of a plastic bag full of popcan's into the hopper, turning the system on and the shredder would begin shredding and the stepper motor conveyor belt or gate would direct the contents into the forge inlet where it's dropped into the crucible where the contents are melted, it's likely that the stepper motor can be installed on some form of load cell to detect the weight of contents inside the crucible and once a certain threshold is reached, the lead screw can dump the contents onto the spigot plate and be directed into a series of the baking trays revolving at the system's base.

    It's likely that additional safety precautions are needed as well. However, the whole system should be able to have a camera and web based monitoring system, a physical warning bell/siren and other safety features involved.

  • Shredding Popcans

    hominidae05/28/2020 at 05:08 0 comments

    The Precious Plastic project has a number of CAD designs as part of their starter kit.

    Included in their starter kit is a design for a plastic material shredder.

    The Aluminum Recycling Micro Factory project will be borrowing a number of the design choices from there with a couple of exceptions.

    The first is that the first motor to be tested will be the nearly ubiquitous windshield wiper motor. It runs on 12 VDC and has enough torque to hopefully shred popcans.

    Since we're going a bit smaller scale than the plastic shredder used by Precious Plastic, the blades will be smaller but the overall design will still utilize plasma cut mild steel.

    The first iteration of this backyard recycler design will also use threaded steel rod instead too.

  • Phase I: Identify requirements

    hominidae05/26/2020 at 16:34 0 comments

    We need to shred pop cans.

    The reason we need to shred soda cans first is because melting tiny shredded bits of aluminum is quite a bit easier than it is to melt the soda cans as they are. There is also the BPA epoxy spray on liner inside the can and the paint on the exterior of the can that we need to get rid of first as well.

    I've decided that a form factor that is approximately the size of a 3D printer is necessary because getting the materials for doing this means the parts can be sent via regular mail.

    We will also be using 2020 aluminum extrusion for the "Aluminum Recycling Micro Factory" because it's commonly available for prototyping.

    However, rather than rushing forward and ordering parts it makes sense to first identify what we need to cram into a small volume of space first.

    1. A hopper is needed to dump a single large plastic bag full of soda and beer cans
    2. The hopper needs to be directly above a shredder
    3. A conveyor belt or some form of conveyance needs to direct a stream of shredded aluminum into a stage that involves cooking the BPA liner and the paint off of the aluminum cans
    4. Both of the above stages will need to be controlled by an Arduino microcontroller or Raspberry pi
    5. Once enough shredded and pre-processed material is collected, it needs to be dumped into a graphite crucible
    6. The graphite crucible then needs to be heated to approximately 800 degree's celsius using resistive heating using kiln wire and a PID controller with a high temperature probe
    7. Once the aluminum is melted, it needs to be tipped over and dumped into a form of some kind, the most likely choice will be a conventional baking tray to create an ingot

View all 6 project logs

Enjoy this project?



sgyoyad wrote 03/29/2022 at 14:58 point

A Micro Factory that is working towards nothing less than a better environment. We are passionate about making a difference and we take pride in our resources recycling and all our processes are environmentally friendly. Aluminum casting Recycling is very easy with GSAL.

  Are you sure? yes | no

Jon wrote 09/05/2020 at 17:35 point

I've seen pucks mentioned in the affiliated site. Is there any particular shape which might be more optimal for storage and/or later repurposed uses? Can't wait to see some final product!

  Are you sure? yes | no

hominidae wrote 03/16/2021 at 05:55 point

Hi Jon, I think most of the metal recycling hardware I've seen use bricks.

I've been busy with grad school and work but I recently picked up a set of breakout boards for the HP power supplies that will do nicely. Especially if I upgrade to the 1200 watt HP server PSU's down the road.

The 1800 watt ZVS induction heater isn't quite rated for that kind of power yet, but perhaps with a large cooling solution and a future iteration it'll be possible.

The issue that I currently have on paper is combining the power supply voltage into 48VDC @ 62.5 Amps. Specifically in a manner that is completely modular in the event of component failures.

  Are you sure? yes | no

hominidae wrote 05/27/2020 at 23:44 point

In regard to whether or not cooking the BPA liner and paint off at 360 degree's works though, I don't know yet.

I will be testing that as soon as I can.

The additional use of some form of bath or chemical solution after cooking the BPA liner and paint at 360 degrees (the boiling point of BPA) might be a great way to retrieve that 4.7 grams of aluminum that is wasted as dross per can.

I also don't know how much additional aluminum would be retrieved if sodium chloride aka table salt was used as an additive flux.

It'd need to be done as a form of unit analysis. With salt, without salt, and with a 360  degree pre-cook with and without salt. The results would indicate what the best method is for processing large batches of aluminum cans and whether it's worth the expended energy.

  Are you sure? yes | no

hominidae wrote 05/27/2020 at 23:41 point

Hello Dan,

I made a sub reddit to keep track a bunch of links, video, and knowledge. In one of the videos, 1,000 cans were melted down.

Here's a link to the post about that video:

I did a quick/hasty analysis here (and did a quick market analysis for use here in the arctic):

If each can weighs between 13 and 14.9 grams, 10.2 grams of aluminum per can was effectively retrieved using the straight dump into a crucible method.

In other videos, the use of sodium chloride was added as flux agent which increased total yield. I don't have numbers for that yet.

However the paint and BPA liner is source of high temperature oxidization for the aluminum. When the BPA liner and paint is heated to higher than their boiling points, I suspect but can't yet prove that they are reacting with the aluminum to create additional dross.

If the BPA spray on expoy liner was removed and possibly cleaned before melting, it might result in greater yield per can.

The numbers aren't that great, but they do add up over millions of cans.

The BPA liner is probably quite a bit of that lost 4.7 grams. I'd wager a guess that the liner and paint weighs about 3 grams.

Here's what the spray on liner looks like once the 0.102 mm aluminum side is removed:

  Are you sure? yes | no

Dan Maloney wrote 05/27/2020 at 17:31 point

Why do you need to deal with the lining and paint in a separate step? I've seen can dropped right into foundries, and I assumed the coatings flashed off. Does baking them off beforehand reduce the need to skim off slag?

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates