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 — 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.