There’s a few areas where I know that there’s currently a need for some sort of undiscovered material.
[ul]
[li]Space Elevator: Needs a material strong but light enough to support several miles worth of its own weight. (Possibly solved by the discovery of methods for linking carbon nanotubes into a greater structure?)[/li][li]Railgun: Needs a material that can transmit electricity, magnetizes well, can withstand large forces and extreme heights.[/li][li]Batteries: Needs materials that don’t get “dirty” across charges.[/li][/ul]
I presume there are some other known applications where we know that a material with X properties would enable us to do Y, but X hasn’t been found yet.
Since it would be too broad a question to ask what all possible cases like this there are and how likely we are to find the mystical material in each instance, I’ll restrict my question to the following:
In general, how much more can we still advance our material sciences on a scale of 1 to 10? Are we more honing in on what’s possible with the raw elements of this universe already (1), or can advances in “atomic 3D printing” or whatever lead to materials with properties beyond anything a comic book artist has ever come up with (10)?
“In general, how much more can we still advance our material sciences on a scale of 1 to 10”
It will happen, just no way of predicting when. Could be tomorrow, more likely one or two centuries. I was in high school in the early seventies, during the first of the oil embargoes. (When oil was $4.00 a barrel, and more than half the electric plants ran on oil). I don’t recall if it was my physics or chemistry teacher, or both, that explained the possibilities of a room temperature super-conductor. Back then the best material had to be about 8 - 12 degrees C above absolute zero, and they were all metals. In the late seventies, a lab was able to increase that by another 50 K using a ceramic material. Just a history lesson. (maybe if oil goes to $400.00 a barrel we will have a big advance in research and reach the goal in ten years).
Even if you assume no new discoveries, there’s a ridiculous amount of things yet to be done with the materials we already know about. Self replicating equipment is inevitable, and we live in a solar system with vast amounts of dumb matter that could be converted to machinery.
There’s also clever solutions to virtually everything you mentioned that don’t require undiscovered material science :
Space Elevator : actually, carbon nanotubes are strong enough. However, an elevator is a bad idea for a number of reasons. We could build rockets using fully automated factories that recycle the expended rockets back into new ones. Or use lasers to increase the ISP of the launch rocket to several thousand.
Railguns : coilguns don’t create barrel erosion. Already available superconductors are ‘good enough’.
Batteries : we could probably build self repairing batteries, or just develop a way to recycle bad batteries back into new ones very cheaply, again with automation and self replicating equipment.
And so on. Don’t get me wrong. I hope lots of new tricks are discovered. But even with what is already known, solutions exist to virtually every human problem short of the heat death.
I started a short thread quite some time ago in which I asked something similar. One of the things that surprised me is that the bonds between individual atoms in something like a metal are tens of thousands of times stronger than the actual materials strength we can produce today. This is due to a lot of microscopic flaws in crystalline structure, impurities in alloys, etc. Some of these flaws are introduced after construction, such as through stresses on the material or even just cosmic rays and background radiation.
So, it is theoretically possible that we could improve materials by orders of magnitude (tens, hundreds or even thousands of times) if we have some kind of “magical” tech that would let us assemble and then correct the material at the atomic level. If so, you’d actually wind up at a point where something like a car might work an awful lot like a person does, with its own little army of nanites working to repair damage, circulate raw materials and waste products, protect against “infection” and so on.
Trying to put a number on this between 1 and 10 is pretty much impossible. We just don’t know where the limit is.
Supposedly worm holes can be created if you have some mass with negative energy density. The latter sounds impossible, but it might not be due to the Casimir effect.
Nitpick: alloys are materials which inherently have “impurities”. For the most part, these impurities, when added in the correct proportions, change certain material properties; for instance, making steel corrosion resistant, or increasing the toughness of aluminum, or making titanium stronger.
That being said, all metals have an inherent crystalline-like structure within which the bonds (metallic bonding) is incredibly strong, and ingots are formed from a large collection of individual crystals (grains) which are more weakly joined at their boundaries. If the metallic structure is also resilient (e.g. doesn’t have any shear planes which in which the bonds can’t resist) then the pure crystalline form is strong indeed. We don’t know how strong a single “whisker” if pure iron would be since we can’t grow one long enough to test, but it is estimated to be somewhere above two orders of magnitude greater than normal low carbon steel. However, such a material would also be extremely hard and nearly impossible to form in any way, and would likely also be difficult to weld to (due to thermal conductivity) so the only practical means of forming or joining would be some kind of sintering process where individual crystals are formed together under heat and pressure (leaving you again with a composite structure which is weaker than the individual parts), or somehow building the crystal up in an as-cast form. We have absolutely no idea how to make such a thing, or if it can even be done with non-magical technology.
The whole notion of “atomic 3D printing”, “self-replicating equipment” and other notions of building up structures atom by atom are still largely the stuff of science fiction. Certainly, we can use the tools that nature has developed to build or modify long carbon chain molecules (hydrocarbons and carbohydrates), and, to a limited extent, even more complex structures (proteins), but synthesizing high strength materials or developing complex systems that have the capability to rebuild themselves runs into some very basic problems that are extremely difficult to solve, such as dealing with the intensity of electrochemical forces at the molecular and atomic level (especially metallic bonds), the difficulty in orienting and manipulating materials at this level, and the fundamental difficulty of providing sufficient power to these “nanobots” without overheating them or tearing them apart are well and beyond even conceptual capabilities at this point. There may be plenty of room at the bottom, as Dick Feynman once suggested, but the rules are also very, very different, and for the most part our ability to control chemical synthesis is still at a stochastic level.
My understanding is that we are most definitely reaching the limits. Technology can advance only so far, and then there’s nothing left to discover or invent.
The really amazing thing is that this has been the situation for several thousand years now, and it seems unlikely to change in the near - or distant - future.
It is true that we’ve very probably discovered all elements–certainly all naturally occurring chemical elements–which are sufficiently stable to be used for any structural capability. But it is absolutely untrue that we’ve discovered all of the materials that may be formed from existing elements, even prosaic ones like carbon, iron, aluminum, copper, titanium, et cetera. And far from being close to being in the position of “nothing left to discover or invent”, the field of material science is going gangbusters. Refined and new techniques for testing material properties, new processes for fabricating, forming, and attaching materials, improved modeling and simulation capabilities, biomimicry of natural material construction materials to make synthetic materials with higher robustness or unique properties, et cetera have resulted in a massive growth in material science research. There are literally a several hundred peer reviewed journals in material science. The difference in material capabilities between today and even just fifty years ago, much less “several thousand years”, is enormous, and yet innovations are integrated in such a low key fashion that most people don’t even realize how much new materials have affected modern life.
It is true that we don’t have some amazing robust thermally insulating material that allows spaceplanes to dive into the atmosphere at high angles of attack, or self-forming liquid metallic robots, or a material that converts ambient heat to laser energy, but that’s because these technologies are essentially magical to begin with and will likely never occur in the forms portrayed in movies and television.
Well yes, but the question wasn’t about whether such a thing could be achieved, it’s what the result would be if such a thing were possible.
E.g., can we model an arrangement of atoms in a computer and test its likely material properties, even though we have no way of achieving that arrangement in real life? If so, what does that tell us about how much further, potentially, material sciences can progress as time goes on?
There are certain properties–particuarly optical properties and thermal and electrical conductivity–which can be estimated from the lattice structure of a ‘pure’, homogeneous substance. The problem with trying to evaluate mechanical properties from this level is that mechanial properties are largely governed by configurations in the microstrcuture–not just flaws and dislocations, but the way elemental metals and metallic and metalloid substances change configuration in different phases, which is heavily determined by how the material is processed. For instance, wnen you add carbon to iron in order to make carbon steel, you don’t get a single configuration of iron atoms with a regular interstitial mix of carbon atoms; there are literally thousands of known configurations which are determined not only by the amount and distribution of carbon but also how the steel is heated, worked, cooled, et cetera. The process of this is to control the microstructure to give a particular set of properties, but are largely governed by the interstitial structure which is very difficult to estimate macro scale material properties from, and we discover and develop new materials largely as a matter of educated trial and error.
Returning to a material like crystallized iron with a body centered cubic lattice structure, it is true that it may have a very high tensile strength and high hardness, but like most crystals will tend to respond to resonant conditions and may fracture along shear planes, and so would not be an ideal structural material.
Communicating with neutrinos would be great. Direct line-of-sight even straight thru the Earth. Imagine being able to send a near-light speed signal anywhere on the planet or nearby at any time. The US Navy alone would love to have this for communicating with subs at a much higher data rate than current LW systems.
We can produce neutrinos in vast numbers easily enough. Modulation and such is a mere Engineering problem of no consequence.
It’s the other end that’s the problem. If some magical neutrino detection system could exist, the result would be amazing. Of course it would have to be directional (else the Sun’s neutrinos would swamp it), “tunable” to neutrinos of a set enery, yadda-yadda. But it would be way cool and a trillion dollar product.
I’ll get to work on this as soon as I perfect mass-produced high-strength Aluminum foam. Strong and very lightweight. I’ll leave the transparent part to Scotty.
It’s already done today (using metals other than iron, at least). Single-crystal turbine blades have been around for decades. They use a selective heating method that begins with a starter crystal, and allows the crystal to grow just behind the molten portion (not too different from how silicon ingots are formed). It’s a slow, involved process but it’s in widespread use.
One almost completely unexplored avenue of materials science is in isotopically pure substances. Obviously, the nuclear industry is the big exception, but there has also been research into single-isotope silicon for semiconductors. A major benefit is significantly increased thermal conductivity.
I have to imagine that there are other benefits to single-isotope materials that are as yet unknown because it’s so expensive to purify elements. Maybe that will change one day.
I do agree. I’m certainly not looking for my self-healing nanite car to go on the market anytime soon. (Certainly, we’ll have fusion-power and flying cars first, and those are still fifty years away, right? )
Still, I see enough room for improvement there that I could see a very long curve of materials strength that might asymptotically approve some theoretical maximum, but with would nevertheless far surpass anything we can do now.
Stranger, you need to update your reference sources. Eric Drexler, commonly credited with popularizing the ideas behind nanotechnology, has several books on this subject. “nanobots” do not and will not work anything like you are describing. You thinking of a common science fiction cliche and not the actual proposals.
To summarize : a nanofabrication machine does not use any unattached components. It’s actually an array of attached parts, similar to existing MEMs devices but much smaller. You are completely correct that trying to create something like a continuous crystalline component of metal would be very difficult. The current proposals are to instead build much smaller, nanoscale components and then to assemble these “loose” tiny parts into small robots. These robots would be able to self assemble themselves into macroscale devices, but they would not be capable of construction or repair in the way it happens in science fiction because there is no way to store enough energy (or dissipate the heat)
A conglomeration of nanoscale robots, even if each robot were made of crystalline iron or something, would not be any stronger than modern materials. The killer advantage would be manufacturing costs and flexibility.
Self repair would also be possible, but for a very different reason. Basically, a car would be built using countless small robots. Let’s say each robot is the size of a eukaryotic cell for the sake of argument (they might be smaller or larger depending on a huge number of factors). Each robot resembles a small cube with a small internal computer and little legs or wheels that allow it to drive over other robots, in the same way that living cells do. To build a car, you’d need a few base types of robots. The most common form would simply be a cube, with little legs, and some kind of attachment device resembling velcro on each face. So you put the first cube down on the print plate. It is told it needs to go to coordinates “0,0,0” and by touching surfaces on the top of the print plate, it would be able to communicate with cubes inside the plate. It finds 0,0,0. Once in place a constant flood of cubes is manufactured and told their destination coordinates. Each cube inquires to it’s neighbors, who would in turn update their positions relative to the nearest “reference cube” they find (a reference cube is one that is already fixed and place and confident it knows where it is based on hop count).
Anyways, the cubes find their ways to the correct place. The attachment cubes would lock themselves to their neighbors to form a fairly strong bond. These cubes are the structural elements of the car.
Each cube has a small internal computer with a tiny amount of memory in order to do these maneuvers. Each one also needs it’s own tiny battery. The cubes share power and data with each other as their “legs” touch each other.
Another type of cube would have 5 faces of attachment points and one face with a smooth surface that can be lubricated. These cubes are for sliding joints. Another type would have a gear surface, and it would pair with a cube that has a motor (corresponding to actin/myosin). There would be a type with 5 attachment faces and one face with a material that can change color. These go at the surface of the body panels.
There would be some that have large internal batteries, and some with large internal computers. Networks of these would form the car’s battery bank and computing system.
Anyways, you could build a whole car with only ~10-20 subunit cube types. Each cube is several thousand tiny mechanical parts, assembled using convergent nanoscale assembly lines as shown in the video linked above. You could take the same 10-20 subunits, and build a huge number of other things using output from the same factory.
If the car is damaged, you have each cube check it’s neighbors. Cubes that you don’t have communication with, or report that they are damaged need to be replaced. An algorithm would determine the sections of the “car” that need to be removed, and the cubes that you do have communication with and are undamaged would detach themselves in sheets, allowing the damaged sections to fall free. Realistically, you would probably need a robot to go and scoop up the damaged sections, and then spray on fresh cubes. The fresh cubes would be assigned the roles previously performed by the broken pieces.
In theory, you could have the robots form themselves out of spare cubes on your vehicle and include the plasma furnace, feedstock plant, and nanofactory to build new cubes as well. This would be a critical capability in spacecraft.
Personally, whenever I hear anyone talking about “nanobots”, I just mentally replace the word with “cells” (which are, after all, nanobots). Can nanobots assemble into large structures like bridges? Sure, trees are of a comparable scale to bridges. But it’ll take decades or centuries to grow your nanotech bridge.
Read the post literally above yours. In no way would it takes decades or centuries. More like hours to days. The simple reason is that you would show up at the bridge site with tankers full of the robots needed to make it.
)