A jewelry buyer in Tokyo ordered 400 slim 1.4 mm wedding bands in 14k rose gold. The first sample rang true. Production came back with 9% porosity on the inner shank, visible as pinholes under the microscope after polishing. The wax was fine. The burnout was fine. The sprue was the problem: a gate sized like the band draws metal too fast and freezes before the ring is fed. Thin bands punish spruing mistakes harder than any other ring form. Here is how our foundry gates them.
Why thin bands behave differently
A standard 2 mm shank has roughly 60% more cross-section to absorb shrinkage. A 1.2 mm half-round band does not. When metal solidifies it shrinks about 1.5% to 2% by volume, and that volume has to be fed by a still-liquid sprue. On a thick ring the gate stays liquid long enough; on a thin band the ring freezes first and the gate becomes a dead end, pulling a void into the inner wall. That void only shows up after finishing, when it is cheapest for the buyer to ignore and most expensive for us to repair.
Lower-karat 10k alloys freeze over a wider slushy range and need larger gates than tight-freezing 18k. We test every new band width on a short run of five rings before releasing a full tree. See how we size sprues on a normal production tree and what shrinkage allowance we build into CAD. For the metal itself, buyers can check fineness labeling against the Karat definition on Wikipedia.
Our thin-ring gating recipe
Gate diameter
For a 1.2 to 1.6 mm band we use a 2.0 to 2.2 mm round gate, not the 1.2 mm sprue a new caster reaches for. The gate is deliberately larger than the band so it freezes last. A gate smaller than the casting defeats the whole purpose. We attach it at the shoulder where the shank meets the basket, never on the high-polish outer face.
Gate length and angle
Keep the sprue short, 8 to 10 mm, and angle it 30 to 45 degrees upward so the ring sits higher than the point where metal enters. That angle lets gases escape away from the band and keeps the metal flow directed at the shank. Vertical straight-on feeds cause cold shuts on the opposite wall.
One gate versus two
Under 1.5 mm width we run one gate per slim stackable band, and add a second gate on the opposite shoulder only when the first run shows shrinkage. Two gates double feeding paths but also double cleanup time, so we do not add them by default. We never sprue a thin band from the bottom of the shank, because that leaves a dimple on the inside that customers feel when the ring turns on the finger.
| Band width | Gate dia. | Gates per ring | Held void rate |
|---|---|---|---|
| 1.0-1.4 mm | 2.0 mm | 1 | <1% |
| 1.5-2.0 mm | 2.2 mm | 1-2 | <0.5% |
| 2.1-3.0 mm | 2.5 mm | 2 | <0.3% |
We also extend the burnout hold by 15 minutes for thin trees so the investment reaches a uniform temperature and the metal does not chill on the way in. Compare that against our notes on common sprue and investment defects, and on why 3D-printed wax changes gate geometry.
What we do on first article
Before a 400-piece run we cast five test rings, section one through the shank under our microscope, and photograph the wall. If we see a shrinkage bubble within 0.5 mm of the inner surface, we widen the gate 0.2 mm and recast. That five-piece sample costs about three hours and prevents a four-hundred-piece recall. We send those photos with the quote so the buyer sees the wall before bulk starts.
Frequently Asked Questions
Why do thin rings get pinholes only on the inside?
Shrinkage voids form at the last point to solidify, which is the thickest section feeding from the gate. On a band that is the inner shank near the shoulder. Outer-face pinholes are usually gas or investment, not shrinkage.
Can I just cast thin rings thicker and grind them down?
You can, but you pay for the extra metal and the grinding time, and you risk breaking through a hollow gallery on detailed bands. Correct gating is cheaper than over-casting.
Does 18k need a bigger gate than 14k?
No. 18k freezes over a narrower temperature range and feeds more cleanly. Wide-slushy 10k and sterling silver need slightly larger gates because they stay pasty as they solidify.
How many rings fit on one thin-band tree?
We cap thin-band trees at 25 to 30 rings instead of the 40 to 50 we run on chunkier designs. Crowding chills the metal and raises the defect rate on the top tier.
Will you share the sprue plan with my CAD team?
Yes. On custom orders we send a gate placement sketch with the first-article photos so your designer keeps high-polish faces clear of gate marks.
Key takeaways
- Gate the band, not the face: gate diameter larger than the shank, attached at the shoulder.
- Two gates only when the first run shows shrinkage; one gate is faster and cheaper to clean.
- First-article sectioning catches voids before the production tree goes in.
Next step: send us your thin-band CAD or a sample to measure. We return a gate sketch, a five-piece test casting, and a per-piece quote with the defect rate we hold on that width. Custom manufacturing opens at 20 pieces per SKU, and we release bulk only after your first-article approval.
