Green Body Strength, Demolding, and the Freeze-Dry Advantage
Some morning thoughts reviewing the literature—
1. Porous Mold Research (SLS TPU)
The porous mold research Sparky found—especially the SLS TPU work (Anil & Nadimpalli, 2025)—seems quite promising. This is the only published example of a flexible slip-casting mold, which could be a real breakthrough for complex geometries with undercuts.
2. Agar Gelcasting — Weak Green Bodies
I read the gelcasting paper by Adolfsson. The green body strength of the agar slurry when demolding is quite weak:
“The agarose system is such a system where the cast materials have a relatively low wet green strength… it is important that the cast parts are not exposed to stresses that can cause deformation during demolding.”
This is a real concern for our use case where mold separation needs to happen cleanly.
3. The Freeze-Dry Advantage — Still Unmatched?
The original reason we were interested in the freeze-drying process: through freezing the part, it becomes rock solid, which allows you to tear off the silicone mold without worrying about the green body. Then after the drying process you fire it.
I was initially excited about agarose casting and starch consolidation for the same reason—hoping they would give similar rigidity—but they don’t seem (to me at least) as strong as the freeze-dry method. When you take off the silicone mold, you could destroy delicate features like overhangs.
Open question: Do any of these consolidation methods from the research actually harden enough at room temperature like the freeze-drying process does? This is the critical comparison we need to make.
4. LPCIM — Promising but Size-Limited
LPCIM seems promising for hardening parts at room temperature (wax solidifies immediately), but it only really works for smaller parts. The literature shows demonstrations up to ~300mm. For architectural-scale panels (500mm+), the wax migration path during debinding becomes impractical. Not a solution for our larger pieces.
5. The HPSC Elephant in the Room
There’s a big open question with the HPSC method that I haven’t dug into yet: for multiple overhangs and complex features, how do you determine how many molds to use and in what configuration?
I think there’s Siemens software (or similar) for mold-split planning, but the physical realization seems extremely difficult:
- How do you orient N mold pairs of various custom geometries?
- How do you make the HPSC system flexible enough to handle any part geometry?
- This seems like it could be a fundamental limitation of HPSC for complex ornamental work—the equipment is designed for standardized shapes (sanitaryware), not arbitrary architectural ornament.
This needs deeper investigation.
Overall: freeze-drying’s 20-hour cycle is painful, but the mechanical advantage during demolding might be harder to replace than we initially thought. We need to either (a) find a consolidation method that matches freeze-dry rigidity, or (b) redesign our demolding approach entirely.