Concept v0.1. Numbers in print_check.py (run from the repo root). Everything [TO-MEASURE] until built.
Don't extrude wet mud upward like everyone else. Build the panel flat, on its steel grid, in the casting bed, by laying and compacting thin lifts of near-dry mix — then tilt it up.
Extrusion needs a wet, soft mix so it can pass through a nozzle. Wet earth is weak when dry, shrinks and cracks (hence the straw), and can't carry the next layer until it dries — so print speed is set by drying, not by the machine. And steel can't easily run across stacked layers.
| problem | extrusion printers | print-and-tilt |
|---|---|---|
| strength | wet, uncompacted, lower density | compacted at optimum moisture by a floating vibrating shoe — rammed-earth density [TO-MEASURE] |
| waiting to dry | each layer must stiffen first | none: compacted earth stands up immediately (why rammed-earth forms come off right after ramming); flat panels don't stack height anyway |
| drying the panel | air only | quicklime hot-mixed at the head binds 16–22% of the mix water chemically and warms the lift ~45–65 °C for the lime-pozzolan reaction — the way road crews dry wet clay; extra evaporation from thin lifts [TO-MEASURE] |
| steel | hard to place across layers | grid laid on the bed first, printed onto; it ends up on the bottom face = the interior face after tilting (exoskeleton, anchors) |
| stiffness per kg | solid walls | printed ribs: 3 in skin + ribs to 8 in = ~30% less earth, same stiffness, 40% lower lift stress; ribs to 12 in = 3.5× stiffer for tall walls |
| insulation | straw in the mix | rice husk packed between the ribs, lime-capped: thermal mass inside, insulation outside |
| curing | ambient | electrified bed: the grid heats the panel (warm cure) and, as the cathode over a sand drainage layer, can pull water down and out electrically |
| relief cavities | — | the head leaves them as printed voids: shrinkage relief, electrode wells, grout or tie points |
- Frame: a low gantry on two rails that run along the casting-bed edges (the bed edges are the rails). Spans an 8×16 ft bed; ~12–18 in of vertical travel. Steel tube, V-wheels.
- Head: hopper → auger feed → water/quicklime dosing at the outlet (hot mix happens here) → floating vibrating compaction shoe (a stripped-down plate compactor on a spring suspension), plus retractable pins to leave the relief cavities.
- Why the shoe floats: like a road paver's screed, the shoe rides on the mix. The gantry only tows and lifts it, so the frame never has to resist the compaction force — which keeps the frame light and cheap.
- Tool changer: the same gantry can carry a tiller head over native ground. That makes it the no-dig machine too (MISSION §15 Route A): spread binder, till in place, compact, press the grid, wire-cut, tilt — with one frame. With a hoist trolley on the same beam it is also the tilt-up gantry (MISSION §14): the pick point travels ~5.7 ft during a tilt, and a rolling gantry keeps the line vertical.
- Power: motors + vibrator ~1–3 kW; runs off the kit's solar system during the day.
- Control: open-source CNC controller; the "print" is a simple raster of lifts plus rib and void paths.
Head capacity ~2.7 m³/h (≈6 panels/h). The mixer is the bottleneck: a 350 L pan mixer feeds ~1 m³/h ≈ 2 panels/h. A house's 17 panels of material is a day or two of mixing, versus 4–6 crew days for the pit route. Curing and tilting are unchanged.
Frame and rails $1.5–4k · motors, drivers, controller $0.8–2.5k · hopper and auger feed $0.5–1.5k · floating shoe (repurposed plate compactor) $0.5–1.5k · dosing and sensors $0.3–1k → ~$3.6–10.5k, reusable across villages, replacing forms and most of the spreading/compacting labor.
- Hand-towed shoe on rails over a 4×8 ft test bed: does a floating vibrating shoe reach rammed-earth density in 5 cm lifts? Core it and measure. Gate.
- Hot-mix at the outlet by hand: dose quicklime and water at the pour; log temperature and moisture of each lift over 24 h. Gate.
- Ribbed panel: form ribs with simple guides, cure, tilt, and check cracking against
print_check.py. - Motorize X/Y, then add the auger feed and rib/void paths.
- Tiller head for the no-dig route.
- Shoe-compacted density or strength falls >30% below pit-route panels → change to a tamping (ramming) head before building anything else.
- Hot mix sets faster than lifts can be placed (cold joints between lifts) → reduce quicklime, or scarify and wet between lifts.
- Compacted hot-mix cubes swell or crack when soaked at 7 days (quicklime slaking after compaction) → mellow the mix before it reaches the shoe, or switch to hydrated lime at the head.
- Ribs crack at the skin junction during the tilt → add fillets or run grid bars up into the ribs.
WASP prints tall walls and domes by wet extrusion; this prints flat panels by compaction and tilts them. Different trade-offs. Their printed domes may still be the best earth roof in dry climates — worth a partnership conversation, not a race.