In spite of our understanding of chemistry, metallurgy still has quite a few mysteries in store for us. Some of them go back thousands of years. For instance we still have not been able to come up with a convincing reproduction of Damascus steel, the prized material that middle eastern swords were built of http://en.wikipedia.org/wiki/Damascus_steel. Another supposedly related and "lost" steel is wootz steel. http://en.wikipedia.org/wiki/Wootz_steel Perhaps a good term to describe the study of these materials would be forensic metallurgy. My closest brush with metallurgy was in a high-school chemistry text-book, which to be honest bored me to death. But at a different level it instills respect for the ingenuity and persistence of humans. Those methods are by no means simple and the bronze-age and iron-age men figured it out.
Among other things that make recreating these materials hard, is that in cooking them you not only need to get the ingredients right, but also get the sequence correct, sometimes to the minutest detail. Impurities in parts per million or the lack of can have significant effects in determining the properties of the material. Same goes for the process of making it.
We understand properties of individual components well enough but the physics that dictates the properties of alloys is different, difficult, non-smooth and very non-linear. Statistical physics have played a huge role in trying to model these behavior.
Another historical-metallurgical curiosity is the surprisingly corrosion resistant iron pillar in Delhi. It has been out there, in the open, facing the corroding elements of tropical weather for 25 centuries, but with little or no corrosion damage.
Indeed. The other problem with metallurgy is that it's awfully complicated on all sorts of scales from the atomic to the micron scale.
Most materials and chemicals out there we now have a pretty darn good understanding of, thanks to being able to sit there and simulate the atomic structure of the material in the computer. But metals like steel are far too complicated since you're not just interested in hundreds or thousands of atoms at a time. To understand steel and predict its properties you need to understand all the impurities and the grain boundaries -- these things occur on scales which are far too large for simulation.
So metallurgy remains somewhat like baking a cake. We can bake all sorts of tasty cakes, but there are other cake recipes out there yet to be discovered which might taste even better and we have no idea how to get there from here. (And for this analogy to not completely break down let's just assume that "tastiness" is both objective and measurable even though it isn't...)
No need to assume that! Some people like key lime, some people like pecan -- just like some people like their steel galvanized, and some people like their chrome-nickel-steel.
The interesting thing about steel is that the tiny grains that make it up are immensely strong, far stronger than any other material we have.
But we have no known method of making steel out of just those grains, it's always mixed with other weaker parts.
That's why spider silk is so incredibly strong - it's manufactured (essentially) one atom at a time.
If we could manufacture steel in the same way it would be far stronger than spider silk.
Here's a list of some of the various grains in steel:
Pearlite, Cementite, Bainite, Austenite (there are many more).
The steel making process is all about encouraging a specific type (and especially, mixture) of these grains to form. But we can do so only at low efficiency and in a random fashion.
This is partly misleading. It's accurate that steel is made up of grains, analogous to grains of sand, but the things you mentioned are more correctly phases, several of which can exist within a grain.
e.g. Pearlite, which you mention, consists of two phases, cementite and ferrite. And these phases can form at the edges of grains, or within grains, depending on growth conditions.
But this isn't the important thing. The important thing is that it's meaningless to talk about how strong "single grains" can be, and how wonderful it would be if we could make large single crystals. And it's meaningless for two reasons: firstly, there are always trade-offs. If you make a large single crystal (i.e. a single grain) you'll sacrifice toughness, because cracks will be able to easily propagate straight through the crystal, without any grain boundaries to stop them. And secondly, you have entropy. Whenever you try and make a large version of something small, entropy dictates you will always have defects. This is why, for example, carbon nanotubes will never scale (which is not to say they won't be useful, they will be!) but if you try and make a centimetre long nanotube it'll be full of defects and will never, ever, be as good as a single microscopic nanotube which is almost perfect.
Anyway, the point is, there's no perfect material. No "good parts" and "bad parts," there's always a trade-off. Sometimes the "bad" parts can do very useful things. As an example, if you make a very strong steel, good luck trying to shape it into anything useful! Imagine if you have a lump of the "best" material in the world, if you can't find a chisel to carve it into something, then it's not going to do anyone any good...
And claims like those given in the parent article are always a little dubious. It's hard for people to imagine the scale of the steel industry, and I personally know of processes that are great in the lab for both steel and other metals which have wonderful efficiencies but just can't be scaled. A kg of steel should cost about the same to make as a bottle of water. If the equipment/size/energy-to-run-equipment investment is just a little too high per kg, it's not going to be useful for anyone at the kinds of scales it needs to be...
That said, despite the difficulties there has been huge progress made in the last few decades, and this might be one of them! Without more information I couldn't say (and it's not my area of expertise anyway), so I wish them luck.
Loosely stated, imagine that atoms can be arranged in a very large number of configurations. Only one of those will be "perfect", i.e. a perfect crystal structure. The rest will be defected in some way or other. As the size of the system increases, the amount of defected configurations also increases and the greater the chance it will have a defect. This is due to entropy/thermodynamics, depending on your interpretation.
That's it in a nutshell, I'm sure someone else can clarify.
Among other things that make recreating these materials hard, is that in cooking them you not only need to get the ingredients right, but also get the sequence correct, sometimes to the minutest detail. Impurities in parts per million or the lack of can have significant effects in determining the properties of the material. Same goes for the process of making it.
We understand properties of individual components well enough but the physics that dictates the properties of alloys is different, difficult, non-smooth and very non-linear. Statistical physics have played a huge role in trying to model these behavior.
Another historical-metallurgical curiosity is the surprisingly corrosion resistant iron pillar in Delhi. It has been out there, in the open, facing the corroding elements of tropical weather for 25 centuries, but with little or no corrosion damage.
Edit: http://en.wikipedia.org/wiki/Iron_pillar_of_Delhi