I’m sitting in a café with a wall covered in embossed tiles, which I think are plastic painted to look like brass with a heavy patina. Some parts of the tiles have a sort of leather pattern, while other parts have raised floral designs and the like.
Despite the enthusiasm for 3-D printing via FDM, most of the things people make with it have very limited mechanical properties; instead they primarily are of interest because of their appearance — the infill is commonly a honeycomb or cross pattern with 60% to 90% empty space inside the outer shell. But appearance is mostly a function of the surface — entirely so, in the case of opaque objects. So perhaps a process that shapes only a surface may be of interest.
Single-point incremental forming, or “SPIF”, is an emerging industrial process for rapid prototyping, which is to say it barely works at all and it’s slow, but flexible. The idea is that you clamp a metal sheet around the edges and poke it with a stick; in cross section in ASCII art:
|clamp| | | |clamp|
------- U -------
----------------------------------
------- -------
|clamp| |clamp|
Where you poke it, you make a divot by stretching (“forming”) the metal out the other side of the sheet, and if you drag the stick along the surface without letting up the pressure, you can make a groove. (Generally you use a rounded-end tool and you lubricate and/or rotate it to keep it from sticking.) You can see the same process in action if you try to write on paper with a ballpoint pen with no ink.
Where this gets interesting is that if you make the groove, say, circular, the metal inside the groove has nothing pulling it up, so it gets pulled down to the level of the bottom of the groove, making a flat circular depression. Then you can do the process again inside there to make the depression deeper, and so on. The same considerations of thinning, wrinkling, springback, and work-hardening that apply to deep drawing apply here, but since SPIF is mostly used for one-offs where you can’t afford the investment to make stamping dies (though sometimes people use it with dies too) you need to use FEM software to simulate them.
(The low-tech approach to this whole thing is hammering sheet metal into shape over a form.)
I was thinking that maybe you could do this with aluminum foil with really minimal force. Aluminum foil also has the advantage that it’s thin enough that you can easily cut it with a spark; and, if you need to anneal it, you can be sure of a uniform temperature through the thickness of the material, and since aluminum doesn’t need a soak time at high temperatures to anneal it, it can be done very quickly. (This requires heating to close to the melting point; people sometimes use burnoff of carbon black on aluminum to indicate that it’s reached the right temperature.)
SPIF forming of a depression is normally done from the outside in, with an empty space under the workpiece, which is clamped only at the edges; this requires workpiece to transmit the load from the forming tool back all the way to the edges. A possible improvement may be to do an initial forming step on a resilient backing, such as a sheet of rubber, or a disposable one, such as a sheet of cardboard, creating many parallel grooves with a bit of separation between them; this produces an accordion-fashion section which can then be unfolded with much less force once the backing is removd.
This is not very far from what you might do with a beading machine to raise a rib to stiffen a sheet-metal surface, the difference being that you’re raising a lot of parallel ribs next to each other, and with the objective of selectively increasing compliance rather than decreasing it. The work-hardening of the metal obviously works against you here.
Electrotyping may be a particularly appealing next process step to apply, allowing the soft, easily melted aluminum to give its precisely-dimensioned form to metals like copper, brass, bronze, nickel, chromium, gold, or silver; I’m not sure how well electrotyped coatings will adhere to the aluminum’s passivating oxide layer, and I don’t think it’s likely for cathodic reduction to eliminate aluminum oxide in water, but if it doesn’t adhere well, that merely facilitates the removal of the aluminum for disposal.
Electrotyping is difficult to apply to alloys (whichever metal is easier to reduce tends to crowd out the other metals), although there are some processes that can electrodeposit some brasses and bronzes. But, by the same token, the electrodeposited metal may work well as a shell to fill with a harder alloy in the molten state. The easiest combination is presumably a copper or nickel shell filled with type-metal, which (as discussed in Flux deposition for 3-D printing in glass and metals and Hot oil cutter) melts at 241°, does not shrink upon solidifying, and does not dissolve iron or steel; I am guessing that it will not dissolve nickel either. It probably dissolves copper pretty well, since lead–tin solder does, but probably not very deeply in the time before it cools at the surface of the mold.
Copper doesn’t melt until 1084°, and nickel doesn’t melt until 1455°, meaning that in theory you could fill shells of them with materials of much higher melting points than the aluminum itself can survive — including, in the case of nickel, cast iron (see Flux deposition for 3-D printing in glass and metals), which will definitely dissolve it. Dimensional precision may suffer from thermal contraction, although I seem to recall that cast iron in particular shares type-metal’s happy property of neither expanding nor contracting upon solidification.
Various pot-metal alloys (Zamak, etc.) are also an option; in theory Zamak 2 melts at only 379–390° but has a yield strength of 361 MPa, better than ASTM A36 steel’s 290 MPa (according to Heckballs: a laser-cuttable MDF set of building blocks) or 250 MPa (according to Wikipedia’s A36 steel article), though the steel beats it at ultimate tensile strength. I think Zamak is more expensive than cast iron, though.
Alternatively, you might be able to cast directly into the aluminum foil, as long as the pressure isn’t too high. According to Filling hollow FDM things with other materials, pure aluminum doesn’t melt until 660°, and the alloy used for aluminum foil isn’t too much below that. Casting will clearly involve some pressure that can deform the aluminum, but this may be small enough not to matter. Solder, type metal, ABS, PLA, lead, hard candy (sugar syrup), paraffin wax, thermoset resins such as silicone or PMMA, and Zamak should all be moldable directly in aluminum foil; some of these will stick permanently to it, but probably all of them will stick to it well enough to require deforming the aluminum foil irreversibly (plastically) for demolding.
Casting crystalline materials involves substantial loss of surface detail from crystallization, both from crystal grain growth directly distorting the surface and from the contraction or expansion that usually accompanies the phase transition. Glasses such as hard candy avoid this problem.
I deformed some aluminum foil (thin, about 10 μm†, times or divided by 2) with a ballpoint pen on top of some paper by drawing a sort of face on it, washed it with 96% ethanol, laid it on a glazed porcelain floor, and then melted some 60–40 tin–lead solder (183°–190°) on top of it, using a shitty non-temperature-controlled soldering iron. The solder was able to pick up most of the contours of the drawing accurately, but something black matter was stuck to the bottom of the solder and screwed up the impression. I think it came from rosin contamination of the tool that carbonized instead of boiling off, but it could also be a chunk of metal oxide from the tool, either copper or iron. The foil peeled easily off the solidified solder. The solder surface showed solder’s usual dull finish, despite the bright finish of the aluminum foil that had molded it; I think this is probably due to wrinkling as the surface cooled before the interior and would be eliminated by adding antimony, as in type metal.
An earlier attempt using the same blob of solder was much less successful, because the bottom of the solder blob was full of bubbles, obliterating most of the submillimeter-scale contours I was trying to pick up. These might have resulted from oil contamination of the aluminum foil (e.g., from my hands) or from the rosin in the solder, which does volatilize significantly at the temperatures needed to melt the solder, let alone the higher temperatures the tool was presumably reaching.
Specks of the black material were also evident on the upper surface of the solder, and upon breaking it with pliers, apparently also in its interior. I was hoping to draw some sort of conclusion about its density or solder-wettability, but I don’t know what to make of this.
The solder blob was about 1.5 mm thick. If the solder’s density is 9 g/cc, which I haven’t measured but should be about right (see A phase-change soldering iron) this works out to about 130 Pa of pressure on the aluminum, or 0.02 psi in folk units, which should put fairly small forces on the aluminum foil, not deforming it much except in very flat areas; the total weight of the solder over a square centimeter would be about 1.4 grams. However, I melted the solder directly on the surface by resting the soldering iron on top of it, so in this case the aluminum foil experienced much greater forces than the weight of the solder. Pouring already molten solder would avoid this problem
† I folded a piece of it in half 8 times and measured it at almost 2 mm with my caliper. If I knew where to find the battery for the caliper I’d be able to give a more precise measurement, but I don’t think it’s that critical — I’d expect anything from 10 μm to 125 μm (a Coke can I had lying around, though that includes the epoxy and paint) to give roughly similar results.
Suppose you use SPIF to cast some simple stamps in, say, solder, Zamak, or type metal, some 100 times thicker than the aluminum foil, and thus 100 to 10'000 times as rigid, depending on the deformation mode you’re considering. (Well, before you account for the difference in moduli of rigidity between the aluminum and the other materials, but those are small by comparison.) You can then use these stamps to stamp or incrementally form more foil, which you can then use to cast or electroform further shapes, either in positive or in negative.
To take a simple example, you could use the SPIF process to form stems, serifs, and bowls of letterforms in foil, which you can then cast into stamps, which you can then stamp into an aluminum foil matrix to form letters, which you can then cast or electrotype into a font; the additional step of making up a stereotype for a page of text allows you to reuse the same letters for each page without losing the ability to reprint previous pages. A recent analysis of Gutenberg’s books suggests he probably used such a process, although of course he wasn’t using aluminum foil for the expendable soft matrix. (Likely materials that occur to me include clay bodies, plaster of Paris, and gold foil; I think papier-mâché, as used for stereotypes, wouldn’t have worked for reasons of temperature limitations. Later printers used a thick, durable, work-hardening copper matrix instead, cutting a single punch from steel for each letter.)
But other shapes might also work reasonably. For example, you could use a tiny spherical SPIF tool to make a mold for a larger spherical SPIF tool; you could quite reasonably make stamping dies (at least for soft metals and slow stamping) to put ribs in things, stamping the ribs themselves with a series of stamps from torus-shaped rib positive and negative dies; and even features like countersunk holes to cast into thin sheets might be feasible. To cast half of a screw thread, you could possibly use a stamp with the form of a single turn of the screw thread, repeatedly; two such molds clamped together could cast an entire screw.
In cases where the cast material releases easily from the aluminum-foil matrix, you may be able to cast both a positive and a negative die simultaneously from the same sheet, thus ensuring a fairly close fit even if the sheet deforms significantly during molding.