Material made from expanded polystyrene beads (usually referred to as “styrofoam”, although that’s a brand name for a slightly different material) is very common for a number of reasons.
What if we did this with minerals?
The foamed nature of the resulting material makes it light, very stiff for its weight, very soft for its volume, forgiving of impact (I think much of the impact energy is actually taken up by compressing the gas), and an excellent insulator, and being some 1–2 orders of magnitude less dense than the source polymer, it is also 1–2 orders of magnitude cheaper per unit volume. It doesn’t have much tensile strength because the beads tend to pull apart, but its compressive strength is extreme, although it suffers from creep in both tension and compression.
Moreover, it is much more resistant to fracture propagation than pure polystyrene, which has an alarming tendency to shatter. Both cell walls and bead boundaries act as barriers to crack propagation.
One of its biggest advantages, though, is its moldability. The beads are made to expand inside a mold, in this way filling the mold. (I’m not familiar with the process but I assume it’s a matter of releasing the pressure at a temperature where the thermoplastic beads are plastic but not molten.)
Another interesting feature is its amenability to shaping processes. You can cut it with a hot-wire slicer with very little energy and low to very low side loading, producing a smooth and precise surface. The smoothest surfaces with this process come when the wire is touching the foam and transmitting its heat by conduction, rather than by radiation, but this produces some side loading and consequent surface imprecision. The “swarf” from this process is absorbed into the surface of the material by virtue of densification from collapsing cells. It’s also fairly easy to cut with more traditional processes like sawing and milling — much, much easier than the underlying polymer. These processes work at such low forces that they do not create significant heat, and thus do not create a heat-affected zone.
The extension at yield of the foam is somewhat larger than that of the bulk polystyrene material, which is relatively brittle, but not dramatically so.
In terms of material properties, probably the most extreme result of the foaming process is the decrease in the material’s Poisson ratio: the foamed material very nearly does not expand at all laterally when compressed, even plastically, or shrink laterally when extended. This is particularly favorable to forming processes: drawing on a styrofoam cup with a thumbnail, for example.
Why isn’t everything made from styrofoam? It’s very flexible for its volume, which is an advantage in many applications but a disadvantage where it needs to resist buckling — though it is commonly used as part of composites (see Sandwich theory, which also talks about buckling a lot). It isn’t transparent, it doesn’t resist high temperatures (because polystyrene creeps and then melts), and it isn’t hard. The molding process isn’t very precise, typically having surface roughness that approaches a millimeter in places, and I suspect it’s considerably slower than injection molding. It’s nonporous, and in a lot of situations where softness is desirable, porosity is too.
Finally, a lot of people just fucking hate it aesthetically, precisely because it’s so widely used and cheap; also, it got a bad rap in the 1980s when it was blown with CFCs and thus contributed to ozone depletion.
Presumably any desired properties of a random foam could be equaled or dramatically exceeded by a metamaterial with similar pore size but precisely controlled geometry, topology, and heterogeneous materials, in the same way that woven and knit fabrics exceed felt and nonwovens, and masonry arches and reinforced concrete skyscrapers can exceed random piles of rocks. So in a sense foams may be a bootstrapping step toward self-replicating machinery or a technique to use at scales too small for your available machinery to manipulate explicitly, not a long-term material. Living tissues generally do not contain unstructured, random foams at scales larger than a white blood cell, and once the humans’ fabrication technology is not so primitive, neither will their fabricated artifacts.
As I’ve discussed in some other notes, it’s very desirable for self-replicating machinery to be independent of organic feedstocks, including things like petroleum: inorganic material is much more abundant both on Earth elsewhere, it doesn’t create conflicts with objectives like reterraforming, and it avoids scaring the humans. However, most materials that are strong but not too brittle at or near room temperature are organic. It might be possible to improve this situation with metamaterials — glass fiber, basalt fiber, and steel coil springs are far from the limits of what can be done. One of the simplest “metamaterials” to fabricate is foam.
In addition to styrofoam-like processes, it’s worth thinking about infusion postprocessing, where after the foam is shaped its cells are filled with some other material. This probably requires the cells to be substantially open rather than closed. Also, think about painting (depositing a layer of a different material on the surface, such as a harder material, maybe a metal), use as a mold (for example, for metal), and defoaming the surface to form a harder surface layer.
Soda-lime glass commonly requires extreme measures to prevent it from foaming: ingredients to reduce its viscosity and long kiln times to allow bubbles to escape. (I’m not sure if precalcination of some of the ingredients is also used). Bubbles are one of the most common obstacles to getting good, clear glass from raw materials.
If you want more bubbles instead of none, it seems like it should be easy to achieve: leave out the viscosity-reduction agents; incorporate more gas-producing minerals such as boric acid, sodium carbonate and even bicarbonate, and calcium carbonate, and cool the melt rapidly.
The sodium silicate I have here dries to a glassy, mildly alkaline substance with a Mohs hardness around 3–4. Upon heating to around 200–300°, it foams up into a white foam with visible bubbles; presumably water inside of it is boiling out and the heat has plasticized the sodium silicate enough to permit bubble formation. The total volumetric expansion is maybe a factor of 10. (These are crude stovetop experiments; my temperature estimate is based on the fact that a grease spot on the same electric burner had started smoking shortly before, for example.) The resulting foamed material floats and leaves tiny silica-gel sparkles when rolled across my palm.
(I suspect that if I let the dried waterglass sit for a long enough time, it might become harder by absorbing CO₂ from the air.)
This phenomenon is behind what I hear is the common use of sodium or potassium silicate as a firestop — upon heating, it foams up to produce a non-oxidizable insulating layer that stops airflow and heat transport.
This suggests that it should be possible to use dried sodium silicate beads or perhaps silica-gel beads to mold and foam up in the same way as styrofoam.
Since the water doesn’t condense again until a lower temperature at which the waterglass is no longer plastic, the foam remains foamed after cooling.
These are commonly foamed minerals used in gardening and glassblowing, able to withstand much higher temperatures without softening than the materials mentioned above; vermiculite is used as an insulator in laboratory glassblowing of borosilicate glass, commonly in direct contact with the hot glass without sticking to it.
They are mostly open-cell foams, which is the reason for their use in gardening to improve soil drainage.
In both cases, the foaming action is produced by the escape of water vapor from the mineral under heating; the minerals are found in nature in their unfoamed form, having formed, I think, under pressure sufficient to prevent the water from escaping.
This rock is naturally foamed during volcanic eruptions; like obsidian, it is a glass, but it contained a large amount of dissolved gas which bubbled out of solution during the eruption. Typically the foam is sufficiently closed to allow the rock to float.
Pumice is not common, suggesting that unusual circumstances are needed to produce it; I don’t know if that’s a matter of mineral content, dissolved-gas content (which might depend on the mineral content), or rapid cooling (which might not be entirely adiabatic; obsidian typically depends on water cooling). More common natural volcanic mineral foams such as scoria cool much more slowly and typically have an open-cell macrostructure and a crystalline rather than glassy amorphous microstructure.
Of course foamed portland concrete is a thing. I ran into a pallet of stacked foamed-concrete blocks on the sidewalk the other day. This works by mixing detergent and air into the water before mixing the concrete. While the reduction in weight and consequent improvement in insulation properties is substantial, as I understand it, it’s rarely even a factor of 3, let alone the 10–100 of styrofoam.
You could conceivably puff up foamed concrete into a mold before the cement starts to set, either by initially mixing it under pressure and then releasing the pressure, or by mixing it at one atmosphere and then pulling a vacuum once it’s in the mold. But usually people just pour it into molds before it sets without fooling around with pressures.
Insulating refractory firebrick is also a common foamed mineral
product. Most commonly it is made by mixing the refractory clay with
a filler material which burns out during firing; any carbon or organic
matter will do (I made some using used yerba mate; see file
ceramics-notes), and I think sulfur would do as well. This material
is also amenable to molding in its plastic state.
I was able to get density reduction of up to a factor of 9, but beyond a factor of 4 (3:1 clay body to yerba) the material was noticeably friable. At a factor of 4, it felt solid but could be carved with a thumbnail in fired form. These foamed ceramics were enormously easier to cut than regular fired clay, which tends to shatter whenever you try to cut it, and enormously more resistant to thermal shock, which I suspect is partly because it’s softer and partly because of its insulating qualities.
To make this process amenable to execution without organic matter, you’d have to find a different filler. This could be something with a low boiling point (like zinc); something whose oxides have a low boiling point, as do carbon, hydrogen, nitrogen, and sulfur; or something that decomposes or substantially decomposes into gases at high temperatures, like nitrate or carbonate.
Alternatively, you might be able to use a styrofoam-like foaming process, where a gas such as nitrogen or CO₂ is dissolved into the clay’s water under high pressure, then bubbled out with a release of pressure.