Pearl size, density, stability — every lever Mark can pull
Short answer: Size is controlled by granulator mechanics and process, not chemistry. Chemistry sets the ceiling — how large you can build before self-fracture, and how durable the finished pearl stays long-term. The two are coupled but independently tunable.
1. Size levers (granulator + process)
| Factor | Direction for larger pearls | Mechanism |
|---|---|---|
| Pan diameter | larger (500 → 1000mm+) | More orbital path per revolution = more tumbling exposure to spray binder and fines |
| Pan angle | shallower (35° → 45°+) | Shallower tilt = longer residence on the disc before pellets roll off |
| Rotation RPM | just above cascade threshold (slower than "thrown") | Fast rotation causes attrition; pellets shear off layers instead of growing |
| Residence time | extend the feed + binder cycle | Every extra minute = more layers accreted |
| Binder feed rate | pulsed, saturation just below runaway | Over-wet = giant wet lumps that crack on cure; under-wet = pellets stall |
| Fines feed rate | higher early, taper late | Fines smooth and bulk the growing pellet's surface |
| Operator intervention | divert undersized pellets back; keep oversized alone | Hand-sort mid-run to bias toward big pearls |
Pan-granulator physical ceiling: ~20–40mm diameter before pellets self-fracture under their own rolling weight during growth. The ceiling is geometric — a rolling mass on a tilted plane eventually has enough centrifugal force to overcome the fresh bond strength. Chemistry can push it a few mm (stronger initial set) but not past the physics.
2. Density & solidity levers (chemistry + preparation)
| Factor | Direction for denser pearls | Mechanism |
|---|---|---|
| Particle size distribution of bone meal / cremains | bimodal (coarse + fines, ~4:1 mass ratio) | Fines pack into voids between coarse particles. Single-size close packing ≈ 64%. Bimodal ≈ 80–85%. |
| Sieve prep | Separate 200-mesh fines and 60-mesh coarse, blend before granulation | Controls packing ratio precisely |
| Binder-to-solids ratio | lowest that still fully wets | Excess binder = porosity after water leaves |
| Colloidal silica solids | 40%+ (vs diluted) | More silica per unit water; less porosity after cure |
| Cure RH | 40–55% in enclosed tent | Fast-dry = cracked outer skin, hollow-core pearl |
| Paraloid B-72 vacuum infusion | 5–7% w/v in acetone, mild vacuum | Pulls resin into remaining pores. ~5–10% density add. Museum-conservation standard. |
| Second binder pass | dilute spray on partial-cured pearls before full set | Fills micro-cracks before they lock in |
3. Stability levers (5–10 year outdoor durability)
- No sodium in matrix — colloidal silica, not water glass. Eliminates ASR + efflorescence per
Pearl_Method_Binder_Selection.md. - Matrix compatibility (Path B2) — dilute slurry of the same Portland used in the Marble Method cast. Zero thermal / alkalinity mismatch at the interface.
- Paraloid B-72 consolidation after cure — multi-decade conservation rating on ceramics and stone.
- Tung oil + beeswax outer — physical water-shed, reversible, renewable; no film-forming plastic.
- Phase 1 cycle test — 28-day bench cure + 50 freeze-thaw cycles before catalog release.
4. The size-ceiling workaround — layered core-build
Stop trying to exceed ~40mm in the granulator. Switch modes.
- Form a solid seed pearl in the granulator, tight spec (15–25mm diameter, bimodal bone meal, 40% colloidal silica, slow rotation, long residence, full 72 h cure).
- Build layers by dip-coat or brush onto the cured seed. Each layer 0.3–0.6 mm wet thickness, 45–90 min dry between at 40–55% RH.
- Repeat 30–80 layers → 40–80 mm finished pearl, potentially denser than pure granulation because each layer is slowly placed and controlled.
- Standard Paraloid + tung-oil / beeswax finish.
Lineage: Japanese nerikomi / neriage (colored-clay layering) ceramic tradition. Thomas Hoadley and Dorothy Feibleman are the Western practitioners most cited. Japanese neriage revival led by Aida Yusuke and Matsui Kousei (Living National Treasure, neriage). Nerikomi artists routinely build vessels 200–300 mm by slow layer accretion — the physics of wet-slurry layering scales to arbitrary size; the only hard constraint is drying-stress cracking, managed by controlled humidity.
5. Max-spec recipe pulling every lever
| Stage | Spec |
|---|---|
| Seed formation | 20mm pan-granulated. Bimodal bone meal (200-mesh fines + 60-mesh coarse, 1:4 mass). Colloidal silica 40% solids. Slow rotation, long residence, hand-sort. |
| Seed cure | 72 h at 45% RH, enclosed tent |
| Seed consolidation | Paraloid B-72 vacuum-infusion (5% w/v in acetone) |
| Layer build | 60 layers × 0.5mm avg = +30mm radius. 60 min dry between. Same slurry as seed. Pigment variation by layer for visible banding if desired. |
| Final consolidation | Paraloid B-72 dip 7% w/v |
| Finish | Wet-sand 400→3000 grit. Tung-oil / beeswax hand-rub. |
| Projected outcome | ~80 mm diameter. Dense (bimodal + infused). Stable (Paraloid + tung-oil + no-sodium chemistry). Labor: ~60 layer sessions over 5–6 weeks at 1 layer/day. |
6. Where the 80 mm number comes from
The 80 mm figure is a projection, not a cited benchmark. It derives from:
- Granulator seed: 20 mm — comfortably inside the 20–40 mm pan-granulation ceiling (Iveson et al. 2001; Litster & Ennis 2004)
- Layer build: 60 layers × 0.5 mm each = 30 mm radius add — layer thickness and drying window from the sodium-silicate + nerikomi literature; colloidal silica should behave similarly but needs Phase 1 calibration
- Finished diameter: 20 + 2×30 = 80 mm
No memorial-industry reference ships an 80 mm pearl-form memorial piece that I can cite directly. Parting Stone tops out around 40–50 mm per pebble. Wooby / Shinjusou pearls top out around 10–12 mm (biological pearl farming). Traditional nerikomi ceramicists build vessels much larger than 80 mm but as hollow forms, not solid pearls.
What Phase 1 needs to confirm:
- Is colloidal silica layer-dry time ≤90 min at 40–55% RH? (Cure kinetics not in public literature for this application.)
- Does each layer bond cleanly to prior cured Paraloid without chemical keying? (Needs test.)
- At what mass does self-weight during drying exceed fresh layer bond strength? (The real ceiling — probably 80–150 mm depending on schedule.)
7. References
- Iveson, Litster, Hapgood, Ennis (2001). "Nucleation, growth and breakage phenomena in agitated wet granulation processes." Powder Technology 117: 3–39.
- Litster & Ennis (2004). The Science and Engineering of Granulation Processes. Kluwer.
- Capes & Danckwerts (1965). "Granule formation by the agglomeration of damp powders." Trans. IChemE 43: T116.
- McGeary (1961). "Mechanical packing of spherical particles." J. Am. Ceram. Soc. 44(10): 513–522.
- Funk & Dinger (1994). Predictive Process Control of Crowded Particulate Suspensions. Kluwer.
- Podany, Garland, Freeman, Rogers (1999). "Paraloid B-72 as a structural adhesive and as a barrier within structural adhesive bonds." WAAC Newsletter 21(2).
- AIC Wiki — Paraloid B-72
- Ceramic Arts Network — Neriage vs Nerikomi
- Thomas Hoadley — US nerikomi practitioner
Pearl_Method_Binder_Selection.md(Drive, 2026-04-13) — binder selection rationale