I'm writing this in part because there is scant little free information on cupolas on the internet. I don't know if nobody's bothered to, or if it's a conspiracy with Stewart Marshall and Steve Chastain :-p, but I'm going to try my best anyway.
Not wanting to risk my silicon carbide crucible in my wood-fired furnace, but more and more wanting to melt bronze, I designed this. Possibly the world's smallest, only a 4" bore (smallest I've heard of yet is 6 or 8"). The intention was something small, to suit my small foundry. Not to mention I'd rather not handle a large load of metal, though at this temperature, it takes very little to be dangerous.
The theory: metal is heated by burning coke. Solid turns to liquid, and as the droplets fall through the burning coke bed they are superheated (heated past their melting point). When they reach the bottom, they fill it up with metal. When the resivoir is full, the tap hole is opened (while melting, it's plugged with a bod made from clay and sand) and discharged into a waiting ladle.
Details... The cupola itself is basically a refractory cylinder with a few holes in it: one all the way through, where the heat goes :), at least one tuyere (I have two - this is a small design), and the tap hole. It might also have a slag hole, which is kept open at all times, and used as an indicator for tapping. There may also be a tuyere which is set lower than the others - known as a safety tuyere, it should have a lead plate in the bottom so if metal somehow gets up to this level, it is semi-sacrificed, protecting all those on the wind belt. Oh, that's another thing - the wind belt is used to supply air to the tuyeres. In my lazy case I used some spare steel pipe instead.
Now to get a little more complicated. When it's started, the whole lower half, from the sand base to about 4 inches above the tuyeres, must be filled with coke. As metal fills the resivoir the coke floats up on top of it, so it won't run out of fire. The contents of a cupola are usually presented as alternating layers of coke and metal, and maybe some flux, piled high from bottom to top. It occurs to me that this image is somewhat misleading; once the first layer of metal melts, you get a dose of coke. That's all fine and dandy, but as you melt the next hit of metal, the coke bed floats up even higher! I doubt that's intentional, so I'm guessing you tap off after each layer. This also means, though it is called a continuous-melting furnace, it still comes in, and out, in batches - though this could be useful if you keep track of which charge is where, and are pouring a lot of heavy castings of differing weight - each metal charge is tailored to its respective casting. But if I'm wrong, please shoot me an e-mail.
Ok, enough talk, I'll put in some photos. ;)
Click for full-size image.
This is the cupola all setup in the melting area, and, in fact, already burning. Just twigs in there, tap hole open, slight breeze in the tuyeres.
Twigs burned down to coals, some metal and coke added, bod inserted, blast turned up. We be meltin'! The white smoke coming off is burnt zinc from the brasses. If you get this, STAY OUT OF THE SMOKE! Zinc, like anything else, is harmful in large quantities. In the full-size pic, that's Mark giving a Metallica-esque salute...not sure where his pinky went though..
After a little while of doing this, I decided it was time to try tapping it. I grabbed a screwdriver and started picking away at the bod. I'm through, I don't see any liquid metal, not even any hot glow. Hrm..... We decide to drop it (that is, drop the contents by removing the sand base, and everything that follows..), and nothing comes out! So I tip it over, and only the coals on top fall out. This is really starting to suck. So I grab an angle iron and start bashing out the charge from the top. 0_o After ample driving, I finally free the contents from the hot refractory cave.
Here's the dump, such that it is.. The circled chunk is what I drove out from the top; it was stuck just above tuyere level. It weighs several pounds, and except for some unmelted copper wire, is pertty much solid. The refractory shows glassy signs that it was hottest about 4" up from the tuyeres, so my coke bed must've been far too low (since the chunk got stuck only an inch or two above). During the heat, I saw a lot of heat energy from this thing, I mean like welding power, pieces of bronze a pound in weight getting orange hot in under a minute, it's an awesome sight. I have no doubt that a cupola is The Way to go for fast power melting.
Coke bed was way too low. Should've added more coke after burning down twigs, before adding metal.
Lack of understanding in operation; I now think each charge of metal is supposed to fill the resivoir, and the charge of coke is to take up the slack after tapping the molten charge. Here I was hap-hazardly adding metal and coke, keeping it more-or-less topped off...
Nonetheless, I'm calling this a learning experiment. It didn't accomplish anything (I ended up recycling the sand mold I had prepared, not feeling like chopping more wood (for the crucible furnace) to pour it in aluminum), but I learned a lot.