Now on to mixing stuff. I suppose I'll go in the same order as the preceeding page. I will list a few characteristics, identify compatible metals and how to mix them, as well as some popular formulations and their characteristics. Each heading will only cover the alloys with that element as the main component.
Tin is found in many things, from bronze to solder to bearings. It always seems to be alloyed.
A common lead-free solder formulation contains one of: 1 to 8% antimony, 0.5 to 2% copper or 0.5 to 4% silver. All these are used to harden the metal (up to 8000 PSI tensile stength), but don't lowering the melting point much. Lead-free pewter is also in this range of composition. Leaded solders are often around the eutectic (at 63% Sn), because an eutectic composition melts at a low temperature and solidifies quickly, without a mushy state inbetween which is inconvienient for soldering. Lead and tin are fully compatible and you can happily use anything from 100% tin to 100% lead; antimony is also often added for strength. Common solder mixtures are 40/60, 50/50, 60/40 and 63/37 (tin/lead). Lead-rich alloys like 40/60 are favored for molding (as lead dent repair on auto bodies) because the slushy state (from 361 to 460°F) allows it to be molded without running off the panel. Bismuth is also compatible with tin much as lead is; the tin-bismuth eutectic is 58% Bi, 42% Sn and melts at 280°F.
Tin is also a major component in babbit bearing metal, although I don't know if I can say it is the primary component of typical mixtures. In general, a babbit alloy is a soft white metal such as tin or lead with something added to harden it, such as antimony. Copper, silver and zinc may also be used. I personally use approximately 33% of tin, lead and zinc for bearings and they seem to work just fine. Zinc and lead like to seperate so once melted, it must be stirred before use.
Although a large component of solder, this role is (unfortunately) being phased out due to lead's poisonous nature, which with similar misfortune is overrated: absorption is almost exclusively by digestion - none by contact. Most often, lead residue on the hands is transferred to the mouth by holding your lunch as you eat. This can be fixed by washing your hands before eating, like your mother always told you too. Additionally, adults absorb only 1% of any consumed. The real danger is to children, who are not only curious about encounting the world with all their senses, including tasting lead paint chips, but are also more sensitive to lead - half the lead eaten is absorbed and retained, deposited in the bones to cause problems later.
The largest consumption of lead is in car batteries, where it is used as a substrate for both electrodes due to its resistance to sulfuric acid. For strength, it is alloyed with antimony, sometimes up to 16% (with tensile strength up to 10 ksi). Pure lead is used for plumbing (as for filling cast iron pipe joints), wheel weights are around 3% Sb and battery lead is 12%. As mentioned under tin, lead is also used in babbit bearing metal and solders.
The only structural white metal, I suppose. It too finds use in solders, for aluminum mostly, which it resembles more than the other white metals. Tin can be alloyed with it for this purpose (one such solder is 91% Sn, 9% Zn), or used alone or with aluminum. The most common alloys are the ZA series, where the number is the amount of aluminum in the alloy. They usually come with 0.5 to 2% copper and a small amount of magnesium; ZA-8 is 8% Al, 0.02 Mg, 1% Cu and 91% Zn and reaches 35,000 PSI tensile strength. ZA-27 is the strongest at 62,000 PSI, and contains 2.2% Cu (copper is about proportional to aluminum content). Zinc is immiscible with lead and probably bismuth.
Pure aluminum is horribly useless in strength, so it's always alloyed. The best alloys for casting are ones which are hard out of the mold, since we can't very well work-harden them to final strength. The best way to accomplish this is silicon, which convieniently also reduces viscosity making for good castability. Often, 1-4% of magnesium and/or copper are added to increase strength and heat treatability. I have no idea where to get silicon for this, so you'll have to settle with more ordinary elements that don't cast as well. I haven't experimented much in this regard, but I imagine something in the range of 1-4% Cu, 1-6% Mg and 4-12% Zn. Nickel is also used on occasion, presumably for strength. Excess iron is to be avoided, because although it is soluble in liquid aluminum, it isn't in solid form. Hence, it precipitates as hard, sharp grains which weaken the structure.
The strongest alloys contain zinc (up to 80 ksi) or lithium (up to 2%, for a strength of 80-100 ksi, which is retained down to cryogenic temperatures making it useful for space vehicle fuel tanks).
When mixing your own, keep in mind that one plus one equals three: 5% of copper is comparable to 2% each of magnesium and copper. Also, aluminum has greater solubility for magnesium, so you can run a bit more, not too much or you'll simply make it brittle. Popcans give an alloy 4 to 6% Mg with 1% of other elements. It seems to me that magnesium alone isn't heat treatable until 8% or more, while copper needs 3 to 5%. Together, it is possible in the 5% Mg, 2% Cu range. Zinc is age-hardenable in the 5 to 20% range, though I haven't tried such an extreme concentration yet.
This cantankerous metal is probably best left alone, not only because it likes oxygen with a vengance (if you light off a crucible of it, don't expect your crucible or furnace or burn pit to survive) but because it seems to be tricky. The first time I melted it, I poured one ingot and soon discovered once-ductile metal to be brittle with a vengance. I also can NOT get that stock to burn, even in small pieces. Still, some people cast it with satisfactory results. The most common cast alloys are AZ91D, AM60A and AS41A. The naming system identifies aluminum with an A, zinc with Z, manganese with M and silicon with S. The respective numbers are the percent content (AZ91 = 9% Al, 1% Zn) and the letter suffix represents which variation the alloy is. Your best bet is to find cast alloy and melt that. For quality castings, I hear a slot sprue (for low turbulence) with a coarse steel wool filter is the way to go. Iron is not soluble so you can melt in an iron or steel crucible. I personally recommend a salt flux cover to prevent it from burning.
A popular thing to do with magnesium is to melt it half-and-half with aluminum, which forms a brittle intermetallic known as magnalium which melts around 815°F. This is crushed up and used for white sparks or fuel in fireworks.
Coppers are alloys with less than 5% additional elements, and are distinctly coppery in appearance. These are not too useful for casting, except for alloying to bronze. Because of the narrow melting point, they don't cast very well. They also need deoxidizers, unless aluminum, zinc or another reactive metal is already present. A typical deoxidizer is phosphor copper, a low-melting mixture of phosphorous and copper used for brazing.
This class here is the real moneymaker, as beautiful for casting as aluminum is convienient. Most alloys have a wide melting point, requiring less gating than aluminum alloys. The classic semired leaded bronze 85-5-5-5 (each number refers to the percent copper, tin, lead and zinc, in that order) starts melting at 1570°F and is fully liquid at 1850°F. It reaches an ultimate tensile strength of only 37 ksi, but since it has relatively low alloy content it also has an elongation of 30%! Zinc allows great strength and ductility, as seen in C260 (30% Zn) which shows 43% elongation annealed; after sufficient rolling or drawing, elongation drops to 3%, with 98.6 ksi tensile strength. The ancient bronze age was brought on by adding 10 to 25% tin to copper, which hardens it. Bell bronze is in the 20% range; the Liberty bell cracked due to a combination of too much tin and uneven cooling. C913 (19% Sn) has almost no elongation, with ultimate strength of 30,000 PSI.
Silicon is also a popular component, the silicon bronzes (typically with 2-5% Si and 1-10% Zn, possibly with lead or tin as well) are purported to be wonderfully easy to cast with, almost impossible to mess up. Personally, I haven't had any trouble with my conventional bronzes so I am quite happy not paying 2 to $3/lb for ingot.
Nickel strengthens copper without altering the structure; it also lightens the color: US nickels are not pure nickel, but rather 25% Ni, 75% Cu. Manganese bronzes are strong, but I can't find much information about them. Aluminum bronzes are as strong as steel and some can even be heat treated almost the same way. A typical aluminum bronze such as C623 (containing 9% Al, and 3% Fe as a grain refiner) is rated for hot and cold forging, with 15% elongation and 87.7 ksi ultimate tensile strength. Probably the strongest is C630, an aluminum nickel bronze with 10% Al, 5% Ni and 3% Fe, reaching 110 ksi tensile. It can be hot forged, and heat treated by heating to 1100 to 1600°F, quenching and tempering in the 200-500°F range. I don't know what rates these should be held to, I assume an hour for the two soak steps. It also has a wider melting point of 1890 to 1930°F which should make it relatively easy to cast (all aluminum bronzes have shrinkage problems for this reason). Some aluminum bronzes have 10-20% Zn, for what purpose I don't know. They are all non-heat-treatable aluminum bronzes in the 5% Al range. I would guess the best of both worlds, perhaps good for rolled products. The large amount of zinc might stretch the melting point, making it easier to cast.
Fluxes for copper and its alloys consist of a glassy slag, typically something between silica, soda and lime. Window glass is amazingly close, so you can just toss some on your melt to keep it clean. I like a mixture of sand, borax (a strong flux), lime and clay (contains alumina, thickening it). Someone told me you should use a halogen flux for aluminum, like potassium chloride. I tried this when making a pound of C630 and it failed miserably: the salt has a very low boiling point and simply evaporates before it can do anything. I didn't try a glass slag, but I assume it would get sticky based on my experiences with aluminum and similar slags.
When adding zinc, do this: place a chunk on your stirring tool and hold it in the furnace, above the melt, until the zinc starts melting. As soon as it starts dripping, thrust it into the melt and stir vigorously. This must be done because copper has a melting point higher than zinc's boiling point and if you just throw a lump on the surface, it WILL boil and pop and explode and make your life, and at the very least your furnace, miserable. (My furnace has a black ring on the inside, above the height of the crucible's rim, from doing this incorrectly!) Tin and lead can be added as raw material with no ill effects; they both have high boiling points like metals are supposed to. Nickel doesn't form a eutectic, so you may have to superheat your copper to get it fully dissolved. I'm not sure about silicon, but I would assume it behaves in a similar way. Both have a similar melting point not far above copper's so at the worst, you can superheat to about 2500°F. Iron is insoluble above 3% so your best bet is probably adding it as filings so it can dissolve quickly.
Iron is the quintessential transition element: high melting point, weird behavior (pure iron has three allotropes, alpha iron (ferrite), gamma (austenite; why not beta, I don't know) and delta (ferrite structure, exists for a short range before melting), is ferromagnetic, combines with many elements to form weird things, including solid state reactions (useful for heat treatment) and through a wonderful coincidence, happens to be extremely common - 8% in the Earth's crust! It does have the unfortunate property of not being corrosion-resistant, but then again half the transition elements can only wish they weren't trapped underneath their own smothering oxide layer.
The most common iron alloy by far is with carbon, which being on the far side of the periodic table has a pretty concentrated effect - only 0.2% is necessary to change the characteristics of this metal. Anything above 0.4% can be heat treated by heating above the curie point (austenite is nonmagnetic) and quenching; more carbon concentrates this action, making higher carbon alloys stronger and harder. The most popular tool steel in this category is W1: the W meaning Water-hardening (water removes heat quickly, in other words a fast quench) and the number identifies a variation. This is basically 1080 to 1095 (the 1 identifies it as a carbon alloy, the two last digits represent the decimal amount of carbon; 1095 is 0.95% C), also used in "drill rod", files and music wire.
Carbon also has the property of easy motion through the iron matrix, without melting. This is known as solid state diffusion, but that doesn't matter as much as its consequence, case hardening. You can take a soft, easy to machine alloy such as 1020 or 12L14 (leaded for machinability), surround it with carbon (often bone char, carbonized leather, peach pits, etc. Supposedly a little cyanide helps, which is probably why peach pits are preferred) and heat to red to yellow heat for a number of hours. The carbon diffuses (apparently with oxygen as a catalyst, trafficing the carbon in the form of carbon monoxide) to a maximum depth of 1/16". After the carburization, the metal is quenched, tempered and ground (if needed) just like carbon steel, the difference being the soft core which will bend rather than snap, as it would if made of solid carbon steel tempered the same way.
Probably the single greatest metal to the foundryman. Not only is it an ancient metal, recent advances have also made it a worthy foe of other forged and welded shapes. It used to be limited to gray and white cast irons (where white is just ordinary gray cooled too quick (for the silicon content) for graphite flakes to form). Nonetheless, these were still very useful, gray iron for general castings where strength isn't a big issue (about 20-60 ksi tensile; much greater in compression) and white iron where abrasion resistance is needed (the two can also be combined, using chills in a mold). Gray iron also has great damping ability, combined with its density makes cheap, solid machine tools. Neither have any elongation or malleability; however, white iron can be annealed very slowly (over a period of days) to produce malleable iron, which can be worked. The annealing breaks down the hard, brittle carbides into small clumps of graphite, I've heard it described as a cluster of grapes. These round shapes allow the surrounding ferrite structure to deform, making it malleable.
Most cast iron is in the range of 2-4% carbon, 0-3% silicon. Some alloys have an excess of silicon, producing silicon precipitates which combine with the graphite forming silicon carbide particles. Ductile iron was discovered in 1946, with the addition of a small amount of magnesium (0.03% typical; due to impurities and loss, you have to add more like 0.07%). This causes the carbon to precipitate as balls (compared to malleable's "cluster of grapes", these are more like beach balls) rather than flakes, allowing the ferrite to move and slide without breaking. This gives higher strength with elongation.
If you start with raw steel, carbon is typically added by dripping it through carbon, as in a cupola, where the charge is supported by a bed of hot burning coke. In a crucible melt it probably dissolves quite slowly, so powdered carbon stirred into the melt would be your best bet. Silicon content originally comes from high-silica ores and fluxes reacting with the carbon in the blast furnace; these days it comes from ferrosilicon. Usually a few bits are added to bring it up by 1% or so, at the ladle or before pouring. It has a low melting point so disperses quickly.
Typical cast irons are 2-4% C (I forget what the difference is), 1 or 2% silicon for thick sections and 2-3% for thinner pieces. 1/4" or thinner may need 3-4%, I forget. Nickel can be added up to 60% for Ni-resist alloys. Magnesium (and cerium) are added to spheroidize graphite on solidification, but magnesium has a low boiling point so is usually added as a mixture with iron and/or nickel. I suppose it could be added with copper, which is also used in irons up to 3% to toughen it.
Next, the final part 3 -- why alloys work (phase diagrams and other sciencey crap).