Originally published: 2015
This page will show you how I built a small home foundry and furnace for melting aluminium and brass for producing some custom metal castings.
This is one of those projects that started off as a desire to have the ability to cast my own metal items without actually knowing exactly what I was going to cast once I had built it. I had a few ideas for some model engines and a geared metal sculpture and even just for recycling waste turnings and offcuts. But I think I mainly just wanted to have my own home foundry… simple as that.
Now that I have built the furnace, set up the necessary equipment for the home foundry and have made a few successful castings, the more I find myself thinking of ideas for new castings for various jobs and projects and I can’t think of why I didn’t built one sooner! I have found that making the mould patterns is probably the most time consuming and skilful art of the entire casting process – practice and patience is a virtue here.
Designing a furnace
I did the usual research activity of trawling through various other documented builds from other like-minded individuals who had designed and built their own furnaces. There seemed to be a large plethora of furnace designs that varied in design complexity. Most of the furnaces that I looked at seemed to work great even with the many differences between them. The exact design of the furnace body isn’t all that critical, however, there are a few design criteria that are essential:
- the furnace body and lid needs to be made from suitable insulating materials to efficiently keep heat in to allow one to attain the high temperatures needed for melting metal.
- the materials used must be ones that can withstand the high temperatures within the furnace, typically 1500-2000°C – a proper refractory cement is the most suitable and frequently used material.
In my opinion attempting to build a furnace using common building cement or flue/chimney cement really isn’t worth it in the long run, even mixing in additive materials such as perlite, sand or even sawdust, the finished product just doesn’t withstand the continuous cycling of high temperatures when operating a furnace. The material will likely crack and crumble away after only a dozen or so heating cycles. Proper refractory cement that is designed for high temperature furnaces is easily obtained from specialist suppliers and should be selected as a priority. I used a company called Vitcas here in the UK – they were very helpful and offered a vast range of products suitable for building a foundry. I used 50kg of Vitcast 1700STD refractory cement rated for 1700°C designed to withstand abrasion and contact with molten metals.
Fuel and burner
The easiest and most common fuel that seems to be used is propane gas, and this is what I use. Butane could also be used but propane was the better candidate as it was easier to obtain in a variety of suitable bottle sizes and the ancillary equipment was available predominantly for propane including gas fittings, regulator, flash back, hose and valves. Propane seems to be used most frequently in commercial applications such as for tarring of roofs. Butane on the other had seems to be targeted at smaller portable burners for blow torches, barbecues, camping stoves etc.
Some furnaces have been successfully designed to run off waste oil. To me this seems messy and inconvenient as I would have to obtain and store waste oil, not to mention design a more complicated burner.
The general shape and dimensions of my furnace is shown in the diagram below.
Notice that the burner inlet port is not centred on the main body and it is positioned at the bottom of the furnace cavity. This is a critical part of the design. It encourages the burner flame to swirl around the furnace walls in a rising spiral until it reaches the top and escapes out the top of the furnace lid port.
The burner is probably the most critical when it comes to its design and is probably the one that takes a bit trial and error. I will only talk about propane gas fuelled burners here as they are the only ones that I have any experience for use in furnaces.
Gas burners for furnaces come in two main categories: forced-air and self-airing burners, the difference between these two burners is that one utilises a forced air intake that mixes large amounts of air with a large quantity of fuel – both the fuel and air flow have to be precisely set to achieve efficient burning of the fuel. The other, is one that relies on a similar principal to Bunsen burners – the burning fuel draws air in from a throttled air intake (or choke). As the flame burns down stream, the fuel stream flow can be set and the choke adjusted to set an efficient flame. An efficient flame or burning of fuel mentioned here is one that involves complete combustion – the flame produced is roaring, very hot and blue in colour, there is no (or little) waste carbon products, just water and carbon-dioxide. Incomplete combustion produces a sooty yellow-orange flame that has a much lower flame temperature.
Building the furnace
I decided to build a furnace that allowed me to cast small volumes of metal up to approximately 4 litres in volume. Deciding what items you want to cast and how much metal will be required is a good place to start when designing a furnace.
To obtain a liquid volume of approximately 4 litres of metal I would need a crucible with a diameter of about 115mm and a height of about 250mm – I had some scrap stainless steel tube suitable for this and proceeded to work out the dimensions of the main furnace body. I decided to have a clearance around the crucible of about 25mm. This meant the bore of the furnace needed to have a diameter of around 270mm. I gave an educated guess of about 75mm for a suitable furnace wall thickness. I had the following steel drum already in mind for the project and based on my requirements above, it was just right for what I wanted. The steel drum used used to contain solvents and I managed to pick it up from work as it was going to be scrapped.
Notice that it is quite a tall drum and that the rims of the drum allow it to be stacked with similar drums. This gave me a cunning plan… I decided to cut the top 150mm of the drum off to make the drum about the right height I wanted for the main furnace body and use the cut off top to form the lid of the furnace. Inverting the lid section allowed it to nestle nicely within the now open drum due to the reduced diameter on the rim. The photo below shows the drum once the top has been cut off using an angle grinder.
The base of the furnace needs a layer of refractory cement – I decided to fill the bottom of the drum with approximately 3″ of refractory cement. It was well agitated to removed any trapped air/bubbles.
To form the walls of the furnace I used a steel can with a diameter of 6.5″ and was long enough to protrude above the top of the main furnace height. I used a section of steel tube to form the burner port which is positioned off centre. I cut a hole in both the central tin and the main body of the furnace to allow the steel tube to fit tightly. The intersection hole patterns were designed in Excel.
I cut a slit in some MDF boards to aid with keeping the inner tin concentric with the main furnace. I fixed the inner tin in place at the base with plenty of duct tape. I also lightly greased the surface of the tine to allow it to be removed easily after the cement had cured. It was important to keep the central form fixed in place during filling with refractory cement as agitation and manhandling would easily cause the central tin to move off centre. The MDF supports were raised as the level of refractory cement increased.
The photo below shows the furnace filled almost to the top with refractory cement. I purposely left about 1/2″ gap at the top to allow the lid of the furnace to nestle nicely into this recess.
The photos below show how I have added lifting arms and a vent hole form to the lid ready for casting the refractory cement. I decided to reinforce the cement using some coarse wire mesh.
The lifting arms were some galvanised steel tubes that I had in stock. I created some paper templates using Excel to aid with cutting the correct shaped holes in the lid so that the lifting arms fitted neatly when inserted. The vent hole size was another educated guess and I used a short section of 2.5″ diameter PCV tubing.
The wire mesh was included to reinforce the refractory cement so that it would not fall apart too easily. As the cement in the lid is not supported underneath, if any cracks were to form later on it would likely end up crumpling and falling through – not good. The wire mesh was secured around the steel lifting arms and a total of two wire mesh layers were included.
I set the lid on top of a Mylar table top surface – the cement does not bond well to this surface. I secured it in position using duct tape. I also carefully positioned the vent hole form in the centre of the lid and secured it in position using duct tape (you can just see the tape on the inside of the tube form in the photo below). When casting the refractory it should be agitated vigorously to remove any air pockets and bubbles trapped under or within the cement.
The diagram below shows a section view of my burner design.
On the left is a fan module salvaged from an old hair dryer. I mounted this into an aluminium tube and brought out the wires to power the motor through a grommeted hole in the side of this tube.
The propane nozzle support coupling is the central tubular insert in the burner. This reduces the previously described fan tube diameter down to approximately the outer diameter of the main flame tube of the burner. It has clearance with the flame tube to allow a press fitted layer of heat resistant wadding to be positioned between the two, creating a thermal barrier from the flame tube as this gets moderately hot when running the furnace. This keeps heat away from the motor/fan section of the burner. It also supports the propane nozzle pipe which is a 2″ long brass tube with a threaded BSP end blank. This blank has a 1.2mm diameter nozzle hole drilled into it. This allows me to change the nozzle size easily. This insert also needs to allow sufficient air to flow through it – hence the elaborate pattern of holes seen in the pictures below. The air inlet side of the nozzle support coupling has a hose barb fitting to allow a standard sized gas hose to be attached. This gas hose is brought through the side of the aluminium tube through a grommeted hole, next to where the motor wires come through.
The hair dryer fan is rated for 24VDC. I built a variable speed controller unit that would allow me to adjust the air intake flow rate of the furnace as required. This allows me to slowly pre-heat the furnace before a full temperature run to drive out any absorbed moisture since it was last used. It can also allow the furnace to be cooled slowly to avoid cracking of the refractory cement or even just allow the furnace to tick over between multiple full temperature runs. I have included a circuit diagram of the circuit below.
speed controller pics
The furnace weighs quite a bit and I wanted to be able to move it around easily by my self. I decided to mount the furnace on a wheeled base. I took an old office chair with a five pointed caster wheel base and decided to modify it slightly to suite the furnace. The length of each spur on the chair base was too long so I simply removed the castor wheels and cut the spurs to the desired length and remounted the castor wheels. I also ground some notches in the top of each spur to allow the furnace base rim to locate on.
The finished furnace is shown in the pictures below.
Foundry moulding table
To make some castable moulds I needed a foundry moulding table where the moulding sand and equipment could be stored and prepared. I built a wooden box to fit on top of a rigid metal cabinet. This is where I would store my moulding sand and prepare moulds using moulding patterns.
The wooden box was made to fit on top with a recessed bottom so that it could not slide about on top of the cabinet but can be lifted up off from he cabinet if ever required. The box was made from 12mm MDF board and given many, many coats of polyurethane varnish (MDF is much more porous than I thought it would be with varnish!). The varnish would preserve the MDF board from absorbing moisture and oil from he moulding sand that I would be using. The box is about 8″ deep allowing me to easily store 50kgs of sand. You can see in the photos a well secured wooden rail mounted at both the front and back of the box. This is to rest a 18mm thick varnished plywood platform to support a flask where the moulding patterns are rammed up to form a mould.
The shelves in the cabinet below are a very convenient place to store the rest of my foundry equipment. It came out a a nice compact tidy unit that I am very happy with.
The moulding sand I am using is an oil based moulding sand from a manufacturer called Mansbond. I purchased the sand from an on-line UK foundry shop: ArtisanFoundry.co.uk. I much prefer this sand to making green sand by mixing sand with clay and water. The oil based sand does not require prolific vent holes to allow for vaporising water and gives an excellent finish – far superior to what I have seen with traditional green sand. It is also ready to use with no need of continual adding of binding and wetting agents as the oil does not evaporate out of the sand like water does from green sand. When new, the sand has the orange-red colour seen in the pictures above. As the oil bonded sand is used in castings, the sand in contact and in close proximity to poured molten metal has the oil content burnt out. This sand turns black and the whole batch I have will eventually all turn black. The sand can be easily rejuvenated by mulling the ‘burnt’ sand back in with the rest of the sand and occasionally adding a small amount of oil to replenish the oil content when needed.
* coming soon, including the following topics *
- crucible types
- crucible lifting and pouring tongs
- degassing plunger
- degassing compounds and gasses
- flux compounds
- drossing spoon
- alloys and casting temps
- pattern making
- mould making and flasks (cope, drag)