Is it possible to predict the properties of metallic alloys?

Without actually making samples of test alloys, that is?

Hmmm, while we’re here, I’m not sure if the word metallic isn’t superfluous; is there such a thing as a non-metallic alloy? (or partially non-metallic?)

The properties you see listed at Matweb, inMil-Handbook-5, in vendor supplied data, etc. are all based on historical test data. Various military specs will require that a certain alloys exhibit certain material properties, for example 6061 Aluminum, spec QQ-A-200/8. However, these requirements are based on historical test data as well (obviously, it would be unreasonable to require that 6061-T6 bar stock have an ultimate tensile strength of 75 ksi if historically it has exhibited a UTS between 38 and 42 ksi).

Now, it is possible to predict some, but not all, properties of “new” alloys, based on knowledge of similar alloys. For example, if I were given a new aluminum alloy and told nothing about it besides the fact that it’s an aluminum alloy, I could predict that it would have a Young’s modulus of about 10E6 psi, a Poisson’s ratio of about 0.33, and a density around 0.1 lb/in^3, but I wouldn’t be able to tell you what the ultimate, yield, and fatigue strengths are.

And, yeah, “metallic” is superfluous. From Webster’s:

Although I tend to refer to alloys by specific metals, e.g. aluminum alloys. If I’m talking about alloys in general, then I leave “metal” or “metallic” out.

As Strainger says, Young’s modulus, Poisson’s ratio, and density are probably pretty easy to calculate for most alloys.

For alloys of steel, the strengths depend on three things: carbon content, heat treatment, and cold working. I’ll ignore the cold work for now. The heat treatment alters the strengths of the steel because the size and form of the grains in the material, and how the carbon precipitates out of the iron, depends on what temperature the material is heated to and how quickly it is cooled. A big effect of adding alloying elements other than carbon to the steel is to change when and how fast the microstructure changes during cooling.

I took a materials class from a guy who collected a lot of empirical data on the effects of different alloying elements and heat treatments. In fact, he wrote a book on the various effects of heat treatment and cold working. The upshot of this is, using the data in the book (and I wouldn’t be surprised if other authors have published similar data), one can predict ultimate, yield, and fatigue stengths of any steel, given the alloying elements and heat treatment. In principle, the same predictions could be made for aluminums or brasses or whatever if you have the data to base it on.

Additionally, I would think that predicting other material properties, like electrical conductivity, thermal conductivity, and coefficient of thermal expansion, ought to be fairly straightforward, although I admit I don’t know for sure.

As a related aside:

An acquaintance once said he didn’t go into superconductor research – which was looking extremely promising – because it was basically a random trial-and-error process. Try something, see if it works. If it doesn’t, try something else.

These guys design alloys in a computer before they ever manufacture them. (For the most part, anyway.)

Yeha, what happened to Questek anyway? They seemed to be donig somethign veyr interesting, and the site has no updateds in two years. Or three or something.

Great answers so far, thanks everyone.

So what is it about alloys that can make them stronger/tougher/harder (or sometimes softer/more flexible, I suppose) than any of their constituent metals? - Is it something to do with packing of molecules or is it something to do with altering the crystal structure of the metals (or is that the same thing anyway)?