Paint-tin furnace for melting aluminium

September 2013

I came across the concept of a home-made furnace when I was researching building my own lathe (via the Gingery lathe). The lathe project has been shelved but my little head was spinning at the possibilities of melting and casting aluminum.

This post is about making a small furnace with crucible, and melting aluminum. I don’t cover casting in this post.

After looking at all the options online, I decided to go small and cheap as a tester, and as it would be quick to do. I went for pearlite and fire cement as the furnace lining, understanding that this mix has its limitations in terms of temperature and longevity. The furnace will be fired with ordinary charcoal BBQ briquettes, and a hot air gun. All of these materials are cheap and readily available online or at your local hardware store.

The best online resource for home-made furnaces is: Read his excellent explanation of how furnaces work – as with most things there are trade-offs to be made in the design of a furnace. I bought his plans for an oil-fired furnace capable of melting iron – this is a project for another day but the plans are well worth the price (also I think it only fair to fling him some cash as he has been very generous in sharing his knowledge).

Parts list

  • a 5 lt metal paint tin (not plastic) – Mrs Naut is forever painting things, a quick rummage yielded a mostly-empty tin
  • two 10lt bags of pearlite (from B&Q)
  • two 1kg pots of ready-mixed fire cement (Wickes)
  • a hot air gun, also called a paint stripping gun – I got mine new off eBay for £10
  • an empty small aerosol can of about 40mm diameter for the inlet pipe
  • some self-tapping screws to attach the inlet pipe
  • some wire as a support matrix for the lining inside the lid – you may not need this if you can find a decent lid
  • ~40cm length of 110mm PVC pipe to use as an internal former – a leftover from a disgusting DIY job (don’t ask, this pipe is also called ‘poo-pipe’)
  • ~40cm length of 38mm PVC pipe as a former for the air inlet and outlet
  • a terracotta plant-pot-saucer-thingo to use as a lid – it was all I could find, you can do better
  • some thin plastic sheet (you can also use carrier bags) and tape
  • mixing bucket, a tray, piece of wood and hammer for ramming, and heavy gloves


Part 1 – inlet pipe

First clean out the paint tin. Next work out how thick you want the lining on the bottom – you will need this before making the air inlet hole at the bottom of the tin. I settled on about 40mm.

I used a straight inlet pipe (that enters the furnace at right angles) as this will be a charcoal-fired furnace. For gas-fired furnaces, the inlet would enter the furnace at an angle, to create a swirl of hot gas inside the furnace (see theartfulbodger’s website).

Next make a hole in the side of the tin, near the bottom – this will be the air inlet, so you need to size and position this hole for the size of the inlet pipe you are using. The bottom of the inlet pipe needs to be at the same level as the furnace lining – here’s a sketch to explain:

inlet pipe

Using a knife, cut a cross (vertical & horizontal line) in the side of the paint tin. My inlet pipe (the aerosol can) was 40mm in diameter, so I made a 40mm cross. Fold the metal inside the can – mind your fingers. You should now have a square hole.

Next cut the top and bottom off your donor (empty!) aerosol can. Slice four cuts up the side, about 40mm long. Insert the cut end into the hole you just made in the side of the paint tin and bend the metal inside the tin until it’s flat. Take your self-tapping screws and screw each of the bent tabs of the aerosol can into the tin – I added some pieces of scrap sheet metal (from a computer case) for the screws to bite into, this helped make it nice and solid. Having your screws exposed inside the can is good, as it will give the lining some support.

Here are a couple of pics which will explain:


Here’s a closeup of the screws on the inside in case you are a bit thick:



Part 2 – formers

The formers are used to create the internal shape of the lining, and must be solid enough to withstand ramming of the lining without deforming. PVC pipe is perfect, and easy to cut.

We will need two formers for the furnace body: a large one (110mm) for the main cavity, and a smaller one (38mm) for the air inlet. The length of the formers is not critical, just make sure that you have enough exposed so that you can pull the formers out after ramming – this requires some effort.

I wrapped my formers in plastic to aid removal – make sure that you do NOT tape the plastic to the former: when you remove the former it should slip out of the plastic. I didn’t take care to make sure that the wrapped plastic was smoothed out, which resulted in some creases in the plastic being transferred to the lining. Next time I would make sure it’s nice and smooth. I plugged the end of the big former with aluminum tape. I also drilled a hole in the side of the 110mm former to accept the 38mm inlet former – this is not necessary but helped make sure everything was lined up correctly.

Here is a pic of the inside (110mm) former, wrapped in plastic and with the bottom end plugged with tape, plus a pic of the inside of this former showing the cut-out I made to accept the smaller former for the inlet:


Here is a pic of the formers inserted into the furnace casing, ready for ramming the lining:



Part 3 – ram the lining

I used 4 parts pearlite to 1 part fire cement (ready-mixed) by volume, with approx. 1 cup of water added to aid mixing. At first I thought that this was going to be too crumbly, however after ramming it held together nicely. I am not sure that the water was necessary, as I mixed the lining for the lid without it and didn’t notice any difference.

The lining material is properly known as ‘refactory’ – however I find this word strangely unsettling so I will continue to call it ‘lining’.

Here are some pics of the raw materials (the fire cement is ‘natural buff’, just like me):


Get your heavy gloves on and get ready for some good old-fashioned mixing. First dump the contents of one of the 1kg pots onto a clean surface – for now we just want the container to measure out the pearlite, as the mixing goes much better if you put the pearlite in the bucket first. Next use the fire cement container to scoop five level containers of pearlite, and dump them into the bucket. Dump the 1kg of fire cement on top.

There really isn’t any other way of mixing this stuff other than by hand. The pearlite is so light it’s like polystyrene balls. The fire cement is thick like clay. Imagine trying to mix these two by mechanical means like one of those jobbies you attach to a drill to mix plaster… It took me a good 15 minutes effort to get this stuff mixed properly – here is a pic of what the mix looked like:


Take the furnace casing with the formers removed, and dump about 80mm of the mix into it – the exact depth of loose mix is not critical. Now take your ramming stick and the hammer, and ram the mix flat. Add additional layers until you reach the planned depth of the lining at the bottom of the furnace – this is the same level as the inlet hole we made earlier.

Now insert the formers: both the large internal one and the former for the inlet pipe. You will need to hold the big former as you ram the first few layers of mix, to prevent it moving out of position. After ~100mm is rammed you don’t need to hold it.

I found that 80mm loose mix rammed down into approx 1/4 height. I rammed the rest of the lining using layers of 80mm loose mix – I have no scientific, basis for this depth, but it worked OK for me. Just be aware that if you try to ram too much in one go, you may not get a tightly packed lining. Linings with air gaps are liable to crack or even even go ‘pop’ as the air expands as it warms up.

Keep going until you get to the top. I found it useful to do the ramming in a small plastic tray, as lots of loose mix spills out of the top – without the tray this would be wasted material. I used a hammer on the end of my ramming stick, hitting it a little softer than I would to hammer in a nail. Go easy for the first few layers after the formers go in otherwise you will dislodge the former.

Here is a pic of one of the last layers to be rammed, and also the finished top surface of the lining:



Part 4 – cure the lining

The fire cement we used in the mix is cured with heat. You could cure the cement by simply lighting a small fire inside the furnace, however a more controlled method is to heat it gently in an oven. I chose the oven method as I had used a little water in the mix, and if this is not dried out of the lining before it is fired then the water may turn to steam and either crack or ‘pop’ the lining (also Mrs Naut was out for the day).

First you need to remove the formers. I found it easiest to drill a hole across the exposed top, and use a bar to gently turn and pull the former – it should slip out of its plastic jacket, however getting it going will require some force. If the lining breaks as you remove the former, it is likely that you are being a bit brutal with former removal, or you have not been brutal enough with the ramming. If it does break just start again, and ram a little harder this time.

Here is the bar (another PVC pipe) I used to turn and pull the big former out – the same approach was used for the smaller inlet former:


Before curing the lining, check for cracks. My lining was not cracked, but had a number of creases from the plastic I had wrapped around the former. This was quite disappointing. I had planned to use the second 1kg pot of fire cement to make a smooth internal surface (this is what is recommended by theartfulbodger), however after seeing the state of the lining I decided to keep this pot in case I needed to make the lining again.

As you’ll see later, the furnace appears to work fine without this smooth coat of pure fire cement. The insulating properties of the lining may be affected, but I could still melt aluminum using BBQ briquettes.

Here’s a shot of the lining with the creases from the plastic clearly visible:


Next bake the furnace in the over for 2 hours on gas mark 3, then turn it up to 11 for another two hours. Again no scientific basis for these timings, but it worked for me. I expected these timings to partly cure the fire cement, but more importantly slowly drive out the water that I had added to the mix.


Part 5 – the lid

I actually made this after the first unsuccessful attempt to melt aluminum. I thought that part of the problem was that too much heat was being lost out the top. I did this is a bit of a hurry, so I used the nearest shape I could find – the terracotta plant-pot-saucer-thingo is not ideal for many reasons, and has already cracked a little. It is the right diameter, has a central hole the right size for the exhaust, and the lip extends below the top of the furnace – so I used it. Find something better for your furnace.

I strung wire across the inside of the lid to help retain the lining. This appears to have worked well. Ram the lining as you did for the furnace body. Make sure you have an exhaust hole about the same size as your inlet hole.


Part 6 – crucible

A trip to the dump turned up a stainless steel vacuum flask that was a perfect fit for the furnace ID. I simply cut it to about 120mm height with an angle grinder, and put a couple of holes around the rim. I bent a coathanger to use as a removable handle.

Unfortunately I chose to use the outside wall of the flask, which has a 10mm insulating chamber at the bottom. I believe that this was the main reason my first attempt at using the furnace failed – theartfulbodger says heat is transferred into the metal to be melted through the crucible wall, if the wall isn’t getting up to the temperature of the fire in the furnace then you’ll be melting nowt. For the second, successful, attempt I used the inner part of the flask which had no chamber on the bottom.

Here’s a shot of the outside wall (larger one) and the inside wall – you can see the chamber at the bottom of the big one:


Part 7 – firing it up

As mentioned above, my first attempt to melt some scrap ally failed miserably. 45 minutes of full heat and only a little discoloration to show for it (the ally that is). Suggested reasons for this abject failure are:

  1. the insulating chamber on the bottom of the crucible – I blame my brain for not paying attention
  2. no lid – all that precious heat was just pumping out the top
  3. inadequate blower – I was using an old blower with a dodgy motor, it produced a gentle breeze rather than the howling gale I was hoping for…

The second attempt was a success. I used the lid, the inner part of the flask as a crucible, and also a hot air blower I bought off eBay for a tenner.

First I started a small fire using some scraps of pine. Then I dumped in two BBQ briquettes broken in two, then the crucible loaded with scrap ally, and then two more briquettes around the outside of the crucible. 10 minutes later all the ally was melted – this little event kept me smiling for days afterwards.

Here is a pic of the successful melt in progress – the plastic around the gun melted a little due to hot air blowing back through the inlet onto the gun, next melt I made sure there was no gap:

the melt

…and finally, without further ado, here is the product of all this effort – I have put it in my glory box and will treasure it for ever:

the blob

Online information resources

My primary resource was – there is so much information there, take a look.

For a more advanced furnace than the paint-tin furnace, but not as kick-arse as theartfulbodger’s melt-anything beast, check out myfordboy’s videos on youtube (start with this one). For some great information on casting have a look at his website:


Leave a Reply

Your email address will not be published. Required fields are marked *