Ringallets — Process

From PLA prototype to a sand-cast aluminum ring with milled and water-jet legs. Six steps, one design pivot, and a CA who couldn't wait to test it.

PLA prototyping Sand casting CNC mill + water jet Bead-blast finish

1. PLA Prototype (v1)

The first version was 3D printed in PLA. The ring was modeled as four quarter-circles that friction-fit at their joints, with two of those quarters carrying integrated round pegs. The legs printed as flat plates with round holes to receive the pegs. This was a faithful proof-of-concept of the assembly idea before committing to aluminum.

Assembled PLA prototype ring on white legs standing on a workbench.
v1 PLA prototype — full ring + legs, friction-fit assembly.
Four PLA quarter-sections of the ring laid out flat with the round pegs visible.
The four quarter-sections, two with integrated round pegs.
Single PLA leg with round peg-receiving hole.
v1 leg with the original round peg socket.

2. The Hex Pivot

Round pegs let the ring rotate freely in the leg sockets. If the friction fit wasn't perfect — and at PLA tolerances, it never was — the ring wobbled side to side under load. Switching to a hex cross-section killed the rotational degree of freedom outright: a hex peg can't spin in a hex hole, even with measurable clearance. The hex also made the joint more forgiving on tolerance, since the flats engage at six points instead of relying on a tight diametric fit.

v1: round pegs on the PLA quarter-sections.
v1 — round pegs.
v2: PLA ring with hex pegs visible.
v2 — hex pegs. Same form factor, no rotational wobble.

3. Sand Casting the Ring

With the hex design locked in, the ring needed to move from PLA to aluminum. Sand casting was the right process for two reasons: it handles the closed curved geometry of the ring without fixturing headaches, and the integrated hex pegs cast in one operation rather than being added later. The pattern was the PLA ring itself, split along a parting line.

Pattern preparation

PLA ring pattern half on a wooden parting board, dated and marked.
Pattern half on the parting board. The red marks are parting compound; the date and "Nick K 11/3" are mine.
Pattern with sprue and riser dowels installed before sand packing.
Sprue and riser dowels installed on both sides of the parting board before sand packing.

Mold making

Flask packed with sand, pattern still seated.
Sand packed around the pattern.
Sand mold cavity after the pattern was removed.
Cavity left after the pattern was lifted out.
Both halves of the sand mold opened, showing cope and drag cavities side by side.
Cope and drag opened, both cavities visible. The two halves clamp together to form the closed mold.

Pour

Assembled sand mold — cope on top of drag, clamped together and ready for pour.
Assembled sand mold, cope on drag, just after the pour. The dark surface is carbon residue.

Reveal

Raw aluminum casting on a steel table with the sprue, riser, and runner stubs still attached.
Raw casting, fresh out of the mold. Sprue, riser, and runner stubs still attached — this is what comes out before any cleanup.

Cleanup

Cleaning up the casting in a vise with hand tools.
Cleaning the casting in a vise — sawing and filing the gates and parting flash off.
Close-up of the as-cast hex peg in a vise before grinding to tolerance.
As-cast hex peg before grinding. Excess material on every flat needed to come off.

Surface finish

Bead blasting the cast aluminum ring inside a blasting cabinet.
Bead blasting in the cabinet — the satin finish you see in every later image comes from this step.

Tolerance check before committing

Before machining the actual legs, I cut hex test blocks at a few different sizes and fit them to a finished cast peg by hand. The goal was a peg that slides in firmly with finger pressure but doesn't rattle — a working friction fit. The test blocks let me dial in the hex hole dimension before cutting the real legs, where a single mistake would scrap a leg.

Hex test block.
One of the hex test blocks. Multiple iterations dialed in the fit.
Test block fit-checked onto the cast aluminum ring's hex peg.
Test block fit onto the actual cast peg. Goes/no-goes tracked by feel.
Two finished aluminum cast rings stacked on a stone surface.
Both cast rings, cleaned and bead-blasted.

4. Legs (Mill + Water Jet)

The legs started as a single block of 6061-T6 aluminum. I faced and squared it on the manual mill, then cut it into four matched leg blanks. Each leg got two tapped holes for the baseplate screws, then went onto the water jet to cut the hex hole that mates with the cast peg. Cutting the hex on the water jet rather than the mill kept the hole geometry exactly matched to the test-block dimensions established in §3.

6061 aluminum block on the mill bed with milled leg blanks resting on a red rag beside it.
Leg stock on the mill bed. The four cut blanks rest on the rag at left.
Operating the OMAX water jet to cut the hex hole in a leg.
OMAX water jet cutting the hex hole. Same machine that cut the baseplates.
Top view of all four leg blanks lined up showing tapped screw holes.
All four legs lined up, tapped holes facing up.
Side view of the four leg blanks with screws sitting in the tapped holes.
Same legs with screws threaded into the tapped holes for a fit check.

5. Baseplates

Baseplates were water-jet cut from aluminum sheet on the same OMAX machine, with the two through-holes for the leg screws cut in the same pass. Cutting the holes simultaneously with the outline guarantees they're concentric to the part — no secondary operation, no fixture error.

Water-jet-cut sheet metal showing the baseplate negative space and a finished baseplate piece beside it.
Water-jet output: the cut sheet (negative) on the left, a finished baseplate (positive) on the right.

6. Final Assembly

With all four legs, two baseplates per side, and two cast rings ready, the final assembly is just press-fit and screw-down: hex pegs into the hex holes, baseplates screwed to the legs.

Four finished legs with baseplates attached, lined up.
Four legs with baseplates attached, hex holes facing up.
Pair of Ringallets assembled with rings on legs but without baseplates yet.
Rings + legs assembled, baseplates not yet installed.
Final pair of Ringallets on a clean white background.
Final pair, fully assembled.

In the wild

Two early tests:

A course assistant doing a push-up on the Ringallets in the workshop.
The CA who'd been watching the build did push-ups on them as soon as the product was fully assembled.
A teammate from the Stanford gymnastics team doing a handstand on the Ringallets in a gymnastics gym.
I brought them to the Stanford gymnastics gym; a teammate wanted to try a handstand and ran one.