Back in 2018, I had a go at making various metal badges, using only cheap, easily accessible materials (and I guess a 3D printer). The main issues at the time were getting a good pour, and getting the mould to release the initial positive. I have now attempted to find solutions to these issues while casting actual three dimensional figures.
I will again note how the purpose of this project is minimizing cost of production. The mold release issues could potentially be solved by using actual molding silicone, but the cost of silicone would be about 4 times higher.

Attempt 1

With the knowledge that silicone does not stick to PP, I expected the same might be true for PETG. I printed a positive in PETG, as well as an outer shell for the mould to contain the silicone, and then filled it up with regular builders silicone. As it cured however, it became obvious that the entire thing was impossible to get apart, and as the outer shell was rounded, I couldn’t cut it out either.
Molds filled by silicone and completely stuck

Attempt 1.1

In the following attempt, I split the outer shell into two parts, and printed outer rings to keep it in place. Before filling it with silicone, I covered everything in petroleum jelly. The mold took many days to cure, as the outer shell restricted access to air. After curing, I cut out the mold from the positive and outer shell. While part of the mold did release easily, other parts were still stuck on, and tore on removal. This resulted in unwanted protrusions and a loss of detail in the final metal casting.
Molds that have been cut in half and the printed positive removed
The resulting casts from the molds
While not a complete failure, improvements would require changing the material of the printed positive.

Attempt 2

Potential materials for the positive were narrowed down to either PP, which should stick less than other plastics, or a filament which could be melted or dissolved out of the mold without affecting the silicone. With PP being a maybe at best, I looked across the Prusa materials page for filaments used as dissolvable support. The two most common types of dissolvable filament are PVA and HIPS. PVA is quite expensive, but can be dissolved in regular water, while HIPS is cheaper but only dissolves in limonene or acetone. I had access to a roll of PVA filament, so I gave it a go.

Printing PVA as actual structure instead of as support is a challenge, and requires several changes to the default PVA filament profile. Through a few failed and successful prints, I arrived at the following recommendations:

  • Whether using a powder coated or PEI sheet, the printer bed must be covered in a layer of school glue to have sufficient adhesion.
  • Using glue may cause very strong adhesion. To avoid breaking the fragile model on removal, it might be useful to print it on a raft structure.
  • Printing must be slowed down to avoid problems with infill and curling of the layer, especially for prints with small/fast layers. 30mm/s or slower worked nicely for these prints.
  • The default PrimaSelect PVA+ profile features no cooling, and a somewhat high extruder temperature. Lowering the temperature to 185C and enabling cooling at max 40% seemed to also improve the quality of overhangs. Layer adhesion may however decrease with too much cooling.

With the PVA printed positive, I also printed new shells in PLA with added holes to speed up the drying process.

A Knight printed in PVA
A Rook printed in PVA

As it turns out, the solubility of the PVA was not important, as the positive released from the mold with almost no issue. I am unsure if this would still be the case without any petroleum jelly.

Molds in one piece

While destructively removing the positives (due to the low layer adhesion), I discovered multiple air bubbles in the mold, which I tried patching without much success.
Pouring went without a hitch, and I even managed to pull out the cast with the mold still in one piece.

The resulting casts from the new molds

The casts using PVA positives came out much cleaner and detailed, though still with some protrusions due to the air bubbles in the mold.

Future work

Next on the list of things to try out:

  • Getting rid of those pesky bubbles in the mold. This could possibly be done mechanically, or through degassing. Are budget vacuum chambers for degassing a thing?
  • Accelerating the drying and curing of the silicone. Even with added holes in the outer shell, curing still takes upwards of a week. Catalysts might be a thing to check out.
  • Using metal casting for actual mechanical parts in projects.
  • Possibly giving actual casting silicone a try. It might just solve all these issues while still not being that much more expensive.