I figured I'd finally put up a page about this subject. Most people just melt everything together that looks about the same, and that's fine, until they suddenly discover that their new alloy is so brittle it cracks on cooling, then blame it on the moon being aligned with Uranus (or theirs, whichever). I'll try to keep this simple, but go into as much detail as I know. Okay, maybe not. Honestly, there are PhD's on the subject, and I'm sure I'm only scratching the surface with what I know. I'll try to write this as two pages, one for the scientifically curious and one for those who just want to melt stuff right and maybe understand a little of the behind-the-scenes action going on.
Surely the number one cause of bad alloys is mixing things that don't belong together. So I'll tackle this first: how to properly identify your stock. Feel free to e-mail me (address is at the bottom), as I'm sure there's a lot I'm missing or getting wrong here. Keep in mind there's a bit of overlap so it's not clear cut. It is better than paying a few $100,000's for a spectrum analyzer, although that WOULD be very cool to have on hand.
These are grouped as basically anything with a melting point that doesn't glow. The metals we are interested in are lead, tin and zinc, with bismuth and cadmium on the back burner (not literally; cadmium fumes are poisonous). Indium, gallium, thallium and mercury also fit in this class but are all too rare, expensive or poisonous, and have little use for a metalcaster anyway, unless you want to make "trick spoons" which melt when used to stir a hot cup of coffee.
Tin used to be used in sheet form, hammered or rolled, and is where we get the terms "tin foil" and "tin cans". These have been replaced due to tin's high cost. Probably the best bulk form you'll see is plumbing solder, which due to fears of lead poisoning, is now lead free. Since pure tin would be too soft, they add 1 to 10% of antimony, silver or copper to the tin. It is useless to try reclaiming solder off removed plumbing, although these soldered joints can be melted for bronze. I hear fishing weights and possibly shot is also being made from tin, because more people think lead is poisonous.
Lead is often used in solders. With tin it forms a low melting point eutectic around 37% ("Sn63") which is a joy to use for electrical soldering. It is used in car batteries (best not to tamper with these acid-filled devices yourself), bullets, shot, and a whole lot of other things; mostly in combination with antimony, which hardens it. (Arsenic would be preferred, but it's a little too poisonous for these days.) I've found old cast iron drain pipes to be a good source of relatively pure lead, if you don't mind busting up the iron to get it. It's used anywhere a cheap weight is needed, wheel weights are a good example of this. I know a guy who buzzed around for a year collecting a few tons of lead weights from tire shops for almost peanuts, then melted it all and cast a keel for the boat he's building.
Zinc is the most important white metal to me because it is a major constituent of brass. It also forms hard, strong alloys with aluminum (known as Zamak), more commonly maligned as "pot metal". This may be the most common form you will find it in, and I can only recommend you collect a large pile of stock and melt it all together (yes, you'll need a big crucible or reverberatory furnace), cast ingots and decide if you like the castability, strength and machiniabilty of it. Generic potmetal just varies so much. Zinc is popular for die casting, but aluminum is too - you can usually distinguish by the weight; zinc is almost three times heftier. Zinc has also been used in the American cent since 1983 (they are now only coated with a copper layer to preserve the appearance and provide corrosion resistance; on melting, it dissolves and contributes 2.5% copper content).
Zinc also has the unusual property of an unusually low liquid range - in fact, zinc and magnesium are the only metals you are likely to encounter that you can actually boil without being blinded by the immense heat required to do it. Boiling zinc produces white smoke, thin whisps (which are actually an aerogel), and a deposit left in the vessel known as Philosopher's Wool. The fumes can give you the unpleasant fume fever, so don't overheat anything containing zinc if you indend on breathing the air around it.
This group includes such exotic names as yttrium, beryllium and scandium, but of all of them, you are most likely to find aluminum, magnesium and titanium. Titanium makes bright white sparks, great for show but is too reactive and melts too high to be of any use to the foundryman. This leaves...
The second most common construction metal in the world, aluminum is forged, wrought, rolled, stretched and of course cast. Although all the alloys can be cast in some rudimentary sense, you'll get the best results from cast alloys, which have between 5 and 20% silicon, which enhances flowability and as-cast strength. Other alloys, like the ubiquitous 6061, mostly used for rolled and extruded products, are okay but are a bit soft on casting. Most alloys will harden after a month or two to a T4 condition; this is improved if you shake out soon after the metal solidifes and quench with water. You can also speed up the hardening process by baking in the oven around 300°F for a day.
Probably the most useless aluminum alloy ever is 1100 series pure aluminum (3003 is close behind). Basically, if it's soft as hell whether work hardened or annealed, and does not respond to heat treatment, it might as well be pure. But because it is pure, it is also the most valuable -- for alloying. I'm still working on how to make sevicable master alloys, such as silicon. If you are happy enough with a 200 series alloy, you can mix equal weights copper and aluminum wires and make a 50/50 master alloy, which has a low enough melting point that it can be dropped into molten aluminum and will dissolve readily (unlike raw copper wire which will take much longer to dissolve). A hint of magnesium will round it out (for some reason, 95% of aluminum alloys contain 0.5 to 2% magnesium as well).
Incompatibilities: I can't think of anything off the top of my head that doesn't work. This begs the question of how I have personally had scrap that looked pretty aluminum-ish actually shrink and crack on cooling! I would guess mixing potmetal (heavier) or magnesium (somewhat lighter, bubbles in vinegar while aluminum doesn't) is a likely cause. I would also suggest you keep your alloys segregated as best you can identify them, cast from wrought, welded or extruded parts. The generic alloys are acceptable for thick, overbuilt, noncritical castings, or something that needs to "give" a bit rather than breaking. Cast alloys (identifiable because the shapes could only be cast :-p ) are naturally hard and strong out of the mold, so are better to use for mission-critical parts.
I don't know much about the uses of this metal, besides Volkswagen transmission cases and galvanic anodes (for steel-hulled boats and water heaters, among other things I'm sure). As a metal, it's a bit lighter, weaker and more flexible than aluminum, and not really cheaper, but the lower density means you can use thicker sections and the stiffness and strength go up. As a result it is used to save weight. As mentioned, the only certain identification is chemical: vinegar (or most any other weak acid) will react with magnesium, producing bubbles of hydrogen gas; while aluminum is not soluble.
Magnesium is dangerous to melt because, like zinc, it has a low boiling point; but because it is so much more reactive, it burns brightly in air (hot enough to light thermite with ease) and can only be put out with inert gas (such as argon or sulfur dioxide). It is reactive enough that neither carbon dioxide nor nitrogen will put it out. I personally have had excellent results with a salt cover flux: equal parts sodium and potassium chloride melt at 1215°F, which is also a good drossing flux on aluminum. 33% MgCl2 to 66% KCl should melt at 793°F, providing protection before the metal melts. These form a thin film over the metal's surface, preventing oxidation and burning, as well as fluxing the oxides, keeping the melt clean.
Hehe, but hey, it's true. These are medium to high density, have a high melting point, are mostly unreactive, and are strong. Because of a wide range of activities, reactivities and affinities, they run from ductile to hard to strong, noble (like gold) to so reactive they protect themselves (zirconium, titanium and so on are all covered in a thin oxide layer, just as aluminum is). Several have melting points so high they were used for light bulb filaments (one of which is still in use today). It is truely a shame that many of these melt too hot, are too reactive, insoluble or expensive to be of much use to the foundryman.
This glowing fellow gets a section all its own. There are just too many classes of "brass" and "bronze" (as if there were a difference) to cover it at once.
Coppers are anything obviously "red" in color, which runs from 99%+ purity electrolytic copper (used for electrical wire and water tubing) to old U.S. cents (before 1982, they used 95% Cu 5% Zn) and more unusual things like beryllium bronze. I would suggest keeping the electrolytic copper seperate from anything iffy; the pure stuff is useful for alloying. (Hint: if you have some plumbing, cut lengths of clean pipe away from the soldered joints which contain tin.) If you get some old tinned wires or plumbing joints, you don't know the composition very well, it might be as much as 5% lead or tin for all you know. If you collect pennies, I'd also suggest putting them (the old ones) in this pile. Such a melt will still probably look quite red, but it might be only 95% copper, not clean enough if you want to make aluminum bronze with it. Perfect starting point for a conventional zinc/tin/lead bronze though.
Colloquially, this term refers to a composition of only copper and tin, possibly with small amounts of other additions, say a few percent zinc. (A typical bell bronze is 24% Sn, balance Cu.) The bastardized truth, however, is that "bronze" and "brass" are basically interchangable. To sum it up, I will give this example: a standard semired leaded brass is composed of 85% Cu and 5% each zinc, tin and lead. You would expect such a combination to be termed bronze, right? Likewise, a gunmetal bronze might contain 90% Cu, 10% Zn -- clearly a red brass! These days, bronze is the term used for basically any alloy of copper that is strong enough to lighten the color (deep pink (red) is low alloy copper), with the major constituent specifying what it is, if something special ("aluminum bronze", "manganese bronze").
It is because of this range that, unfortunately, copper alloys could be the worst disaster of all. Surely the worst stock is keys, which might be made of a base of anything from brass to manganese bronze to aluminum bronze, maybe even beryllium copper, and so on. Not to mention they often come plated with chromium, which may or may not be a harmful addition (at the moment I'm guessing it is a grain refiner, but I'd like some confirmation on that first). Nonetheless, I know a fellow who melts this very stock and makes some nice investment castings with it. The bottom like for copper appears to be, if it has good castability, is strong enough and machines well, it fits your application just fine.
The only incompatibilities I am aware of is tin and lead in aluminum bronze. I haven't yet tried seeing what they do, but I would hazard a guess they weaken the structure. In all the aluminum bronze alloys I've looked over, none of them contain more than 0.1% of tin or lead, and are always listed as impurities. Aluminum itself, if you are not out for an aluminum bronze, may be considered an incompatibility. Aluminum bronzes, though strong as steel (but more corrosion resistant and heat treatable too), are difficult to cast because of the narrow melting point causing shrink defects. I haven't personally tried a casting yet, but I have made a one pound ingot of C630 and I don't like the way it dimpled (from shrinkage) on solidifying.
These are similar to copper, with their own quirks of course, not the least of which is being more expensive. Uh, I don't really have much to say, other than that! Being precious, these tend not to be left lying around like aluminum and copper.
Wow, there's a big class. Since so few people can melt and cast this, I might not even bother, but I'll try to touch on a bit anyway.
The ubiquitous metal, quite literally all around you. It comes in grades from "shovel the shop sweepings into the furnace" (also known as A36) on up to high quality, vacuum induction melted tool steels. Strength ranges from weak in the 20kSI (1kSI = 1,000 PSI) on up to 250kSI and above (for the iron-based (ferrous) superalloys). Stainless contains large amounts of chromium for corrosion resistance and nickel to temper the hard nature. Manganese toughens steel, and in fact almost all grades of iron and steel contain around 1% of it. Fortunately, there is one easy way to determine what a steel alloy is -- the spark test. There are websites with more detail, in the short: orange, forked sparks are mild steel (<0.3% carbon). The ends start to burst and turn into yellow-white sparkles as carbon rises to 1%. Alloys have anything from orange globs to short streamers, depending.
Steel isn't very useful for casting, not just because it isn't easy to melt, but it generally has a narrow melting point so probably suffers from shrinkage as aluminum bronze does. It is economical for relatively noncritical industrial castings simply because there's so damn much of it around in liquid form. It's just another step to pour the ladle into a sand mold rather than the usual billet molds.
This is within the reach of some backyard metalcasters. Those who have done it say it is a great material with high flowability. It also has a very, very wide melting point, meaning risers and gates can be very conservative without fear of shrink defects. The typical composition is 2 to 4% carbon and 1 to 3% silicon. The silicon keeps the carbon from combining with the iron to form hard, brittle cementite particles characteristic of white cast iron, instead forming graphite flakes characteristic of gray iron. More silicon is needed for thin sections because the faster the iron solidifies, the more likely it is to form the cementite. Because the graphite particles are soft, they act almost like empty space within the iron structure. We're all familiar with how, once a crack starts in a piece of glass, it'll just keep on going: this is known as a stress raiser. Well all those little thin, sharp flakes of graphite are stress raisers and prevent the iron from deforming much. They also encourage it to break across the flakes, thus giving the distinctive gray color seen. Small amounts of magnesium and/or cerium (in the 0.03% range) can be added to form ductile iron, which forms graphite balls rather than flakes. The round globs can't raise stress around them, resulting in iron of nearly the same composition able to be smashed and bent much, much more before breaking.