A few years ago, whilst still studying engineering, I wanted to cast an inlet manifold for an engine I was building. After searching around on the Web I soon found that a fair number of people were successfully casting aluminum at home.

I immediately started construction of my own furnace. Like so many projects, I spent a lot of time on the construction and even completed a test melt before other priorities took over. More recently I have come back to this project. Having access to a Mill has mad aluminum casting a more useful tool. Also the PCB Mill is a prefect candidate for custom cast brackets etc.

I'm publishing the details of my furnace in the hope that it may inspire others to have a go. There are certainly a number of plans etc. available on the web. But mostly you can build one by just reading and looking at photos. Importantly, don't be put off by those that tell you it won't work, plenty of my friends had that opinion, I had to remind them that humans have been melting a number of metals for quite a few centuries now.

I will be adding to this page as I do more with the furnace over the coming months. Especially I will be having a go at lost foam casting. For more information on this go to Dave Kush's excellent website www.BuildYourIdea.com.

This is how the furnace looked after its initial build. Following it's first and only melt I stored it in the shed for another day. This day came almost three years later.

The furnace is electric; many of the home built units described on the web are gas operated. I didn't feel comfortable with gas for a number of reasons: It explodes, it's expensive and not convenient (especially when your bottle runs out suddenly). Instead this unit operates on Australian domestic single phase power (240VAC 50Hz 10A).

There are, however, disadvantages to electric furnaces. Obviously it can still be dangerous (re. fatal). Also, even though it is more efficient because there is no need to vent any exhaust, the rate of energy input is lower. A melt with my furnace typically takes 3hr's - Gas units can get going in less than an hour (20min?).

Initially I started by purchasing plans from Dan Hartman (www.DansWorkshop.com). These were great plans, very well written. However, I ended up building the furnace to suit the materials that were most available rather than following the letter of the plans. Similarly the size was based on the materials I could source.

The main body is a 200 litre steel drum that I cut the top out of and welded on a frame with handles and wheels. A section from another drum was cut to create the lid.

The inside of the body and lid are lined with refractory bricks and cemented in place with kiln cement. This stuff is all available from large pottery supply companies. A word of warning though - the stuff is not cheap!

In this photo you can see one of the refractory brick that has been shaped to make a plug. These brick are easily formed using files\rasps. I still haven't worked out exactly what the vent is for?

The photo at left shows the crucible used to hold the molten aluminum. This was constructed from some scrap sections of steel pipe from a local steel merchant.

The bottom plate is 6mm steel plate that was 'gas axed' to size and welded in about 50mm up from the lower edge. Two holes bellow the floor are used to hook the crucible for pouring.

The hoops at the top are used to handle the crucible. A steel rod is paced through the hoops and used to lift & pour the aluminum. These hoops were checked deep into the tube and welded on both sides.

This crucible is very large, too large. It's good for large pours but on smaller jobs too much energy is used just to heat the steel. It could also be made less thick - less thermal mass. Above aluminum melting temperatures (~700°c) the steel is bright cherry red and almost translucent.

One future project is to build a frame that the crucible can sit in be pivoted for poring.

This picture shows the interior of the furnace. The wire that creates the heat sits in a helical channel in the refractory bricks. This channel was created on each brick with a rasp before installation in the drum.

The wire used was purchased from a specialist heating element manufacturer. (Thermal Eclectic Elements Australia - www.ThermalElectric.com.au). They supply all the data required to calculate the length and type of wire required.

Some other constructors have used Nichrome wire from other sources (even used mig welding wire). However, given the amount of current being drawn I decided to pay the extra for wire that I knew the properties of.

The wire was supplied wound and had to be stretched to the correct length before it was carefully inserted into the groove in the bricks. It is fairly important to get this spacing consistent for long wire life.

After applying the refractory cement it was left to air dry for around a week. I thought it was fairly dry and decided to fire the coil up in order to help out the drying process. The only problem was that obviously there was still allot of moisture as the effective coil resistance was far too low. There was a loud buzzing noise and eventually the circuit breaker on the controller popped. The part that disappointed me most was that the RCD (earth leakage device) didn't seem to kick in despite the fact that current was certainly leaking from the coil through the wet cement and to the grounded drum casing.

If I were making the furnace again I would have made the grooves deeper to better protect the coil from the steel crucible (see why bellow in the thermocouple section).

This picture shows the connection between the mains flex and the heating element. The element is passed out through a hole in the drum. Exposed sections of the element are covered with spaghetti (fiberglass) insulation.

The terminal block is a ceramic type that is a little more tolerant of the temperatures in this area. Ideally I would have used high temp silicone wire but it's a little expensive and the outside only gets to around 60°c.

Here you can also clearly see the fiberglass insulation that is wrapped round the outside of the furnace. This is surprisingly effective, but unfortunately very itchy also.

The temperature in the furnace is monitored with a Type-K thermocouple. This is mounted in a stainless steel sheath and capped with a standard industrial aluminum housing. These parts were all purchased from Temperature Controls, Brunswick, Victoria. The thermocouple is held in place by a steel block that clamps on the sheath, the block is welded to the drum.

During a dry run of the system, prior to the first melt, I tried to adjust the insertion depth of the thermocouple while the coil was live. The shaft slipped in too far, pushing the steel crucible against the coils on the other side of the furnace. Not only did this short the coil circuit, it connected it to ground through the thermocouple shaft.

The noise & sparks from this little mishap certainly were impressive. I now have very strict rules about not making any adjustments while the coil is live. The other thing that is quite impressive is the mark on the tip of the thermocouple where it tried to weld it's self to the crucible.

Since then I have found that the thermocouple does not operate very well, I'm not sure yet if this is because of the damage or instead due to bad mounting\placement. As it turns out it isn't really necessary anyhow.

I really got caught out with the controller. I had dealt with temperature control before, but never with these kinds of temperatures. I had the control interface lying around. I planed to program it with tricky heating profiles etc. As it turns out there just isn't enough energy input to create a situation where any kind of control is required.

Here you can see the internals of the control box. On the front panels there are some switches and the circuit breaker. The device behind the control panel is a solid state relay. This is mounted to a heat sink.

Currently the heat output from the solid state relay is a little high. As there isn't any room to add any more heat-sinking I will be adding a small fan. Testing has already shown this to be effective in removing the heat generated.

After recently digging out the furnace again I set about a few modifications. One key change was the addition of the galvanized corrugated iron covering (see photo).

This was added for two reasons, firstly to protect the insulation that was looking tatty all too quickly. Secondly I hoped that this may help reflecting radiation. Clearly at this temperature radiation becomes a more prevalent factor, whether this has any impact I cant be sure.

This is what the metal looks like before a melt. For my first few melts I have been using recycled cast aluminum from machines I have scrapped. The biggest problem with this is that there is a lot of surface area and a lot of paint on that area. The paint mostly burns off with an acrid smoke that upsets the neighbors a little.

I think that for future melts I will be first using the ingots I have made and then blocks from the scrap yard. The metal re-cycling yard near my work has a lot of 50x50mm section and it's only AU$6.00 per Kg.

This is a picture of my latest melt. At the moment I am only making ingots in large tin cans. These will fit nicely back into the crucible later for another melt. The reason for the ingots is twofold. Firstly I am getting used to the melt process. Secondly I think that this helps to reduce impurities. Since this is recycled there is lots of dross (junk on the surface of the molten metal).

Here you can also see why I need to build a device to help me pour the metal. I spilt some and it's a good thing that it was contained in the metal box. The crucible is heavy and hot which makes aiming to pour difficult.

This photo shows what happens to the paint and other contaminates that don't leve the furnace as smoke. Called 'dross' the contaminates are lifted from the top of the melt using a tool like a large spatula

If you're interesting in building your own foundry take a look at the sites in my links section for more ideas. Or if you have a question, send me an email.

Hopefully soon I can show some pictures of something I have cast that's more interesting\useful than ingots.