Casting-Ready Jewelry Files, Engineered for the Pour
Watertight, castable models engineered for the pour — correct wall thickness, sprues and gates in place, shrinkage compensated and geometry validated. Built for casting houses and manufacturers who need production files, not screen models.
Built to cast, not just to look good.
A model can render flawlessly and still fail on the bench. Casting-ready is a different job — and it’s the core of what the studio does.
A picture-perfect mesh can hide walls too thin to survive burnout, edges that aren’t watertight, prongs that warp as the metal cools, and no feed plan at all — problems you only discover after the tree comes out of the flask. In lost-wax casting the model becomes a printed pattern, set in investment plaster and burned out, then filled with molten metal, and every one of those stages is unforgiving of geometry that wasn’t built for it. Casting-ready means every one of those failure points is engineered out of the file, before it ever reaches your printer or your caster.
Everything we deliver is built for one thing: taking the file off the list of reasons a piece fails to cast. We work the way a caster works — thinking in wall minimums, feed and gate placement, burnout behaviour and metal shrinkage — so what you receive is production geometry, not a screen model that lands back on your desk for re-cutting. It sits at the centre of the full range of our CAD services, and every other service feeds into it.
Our clients are casting houses and manufacturers across the US, and they judge a CAD file by what it does at the flask. So every file is checked against the ways a casting can fail — and the ones that live in the geometry are engineered out before it’s marked ready.
Every axis a casting can fail on.
Before a file is marked casting-ready it runs through a fixed set of checks. None are optional, and each one maps to a way castings actually fail in the flask.
Wall-thickness minimums
Every surface is held to a castable minimum — typically from 0.8mm, adjusted for metal, span and mass — so nothing boils away in burnout or fills short on the pour.
Sprues & gates
Feed points and gates are placed where metal should enter and sized to the piece, so the cavity fills fully and cleanly — less porosity, fewer misruns.
Shrinkage compensation
Geometry is scaled for the metal’s shrinkage from wax to cast, so the finished piece lands on your target dimensions and stones still seat.
Watertight geometry
Meshes are closed and manifold — no flipped normals, holes or self-intersections — so your printer slices the file the first time without manual repair.
Metal-weight report
Every file ships with an estimated metal weight across common metals and karats, so your quoting desk prices the pour without a test cast.
Stone count & size report
Exact stone counts and sizes come with the model, so setting, sourcing and costing are locked before a single stone is ordered.
Miss any one of these and the cost lands downstream — a short fill, a porous shank, a piece that comes out under-size, or a mesh your slicer refuses. Catching them in CAD is far cheaper than catching them at the flask.
How the file is engineered for the pour.
Wall thickness, held to the metal’s minimum
Thin walls are the most common reason a piece fails to cast. Below a certain thickness the wax burns out unevenly and molten metal either fills short or never reaches, leaving gaps and rough, incomplete surfaces. We hold every surface to a castable minimum — commonly from around 0.8mm, raised for longer unsupported spans, softer metals and larger pieces. Prong tips, knife edges, filigree and open galleries get particular attention, since those are exactly where a design tends to thin out and a caster tends to lose the piece. Where a design pushes below what will pour, we flag it and thicken discreetly rather than let it fail silently on your tree. The target is a wall you can actually cast and finish, not the thinnest number that still looks good on screen.
Sprues and gates, planned for a clean fill
A model isn’t casting-ready until the metal has a way in. We plan sprue and gate placement as part of the file — where feeds attach, how thick they are and how they route to the piece — so the cavity fills completely before the metal freezes. Feeds are sized so the sprue stays molten slightly longer than the piece it serves, letting metal keep flowing inward as everything solidifies. Good feed geometry means fewer misruns, less shrinkage porosity, and cleaner surfaces where the sprue is cut off. Placement is matched to the piece’s mass and to the way you tree up, so what we deliver fits how your shop actually pours instead of forcing your team to re-sprue every model by hand.
Shrinkage compensation, so the piece is on-size
Metal shrinks as it cools from a molten pour to a solid casting, and different metals shrink by different amounts. Ignore it and rings come out a half-size small while stone seats no longer fit. We compensate for that shrinkage in the geometry, scaling the model so the finished, cast piece lands on the dimensions you specified — ring size, gallery height, seat diameters and all. When you tell us the exact metal you’ll pour, the compensation is matched to it rather than to a generic average, which matters most on tight-tolerance pieces like eternity bands and calibrated settings. Stones meant to drop in still drop in, and your setters aren’t left reworking a piece that came out under-size.
Watertight geometry, validated before it ships
Printers and slicers need a closed, manifold mesh. Holes, flipped normals, non-manifold edges and self-intersecting surfaces are the errors that make a file refuse to slice or print with defects — and they’re invisible in a nice render. Every file is validated to be watertight and print-ready, with wall thickness and stone seats given a final pass, so it slices cleanly the first time and your team never stops to repair a mesh. If you also want the piece signed off on screen before you commit metal, we can pair it with a photoreal render for sign-off — but the casting-ready file itself is judged on geometry, not looks. Every commission runs through our four-step studio workflow, so you always know exactly which stage your piece is at.
Weight and stone reports with every file.
A castable file is only half the job — costing the pour is the other half. Every delivery ships with the numbers your quoting desk needs.
The first report is an estimated metal weight, given across common metals and karats, so your desk can price gold, silver or platinum without weighing a test cast first. Because it’s derived from the model’s true volume, it also flags when a piece is heavier — and pricier to pour — than the design implied, while there’s still time to adjust. The second is an exact stone count and size list, so setting labour, stone sourcing and the final quote are locked before anything is ordered. Together they turn a model into something you can price in minutes rather than approximate and correct later. Here is the kind of spec line that lands with every casting-ready file:
Halo Setting · V04
Shipped alongside the STL / 3MF for this model.
Figures are illustrative — each file carries the real numbers for that exact model.
STL, 3MF and 3DM — matched to your pipeline.
You receive casting-ready STL and 3MF for printing and casting, plus an editable 3DM working file for when you’ll want to adjust the model later. If your printer or pipeline needs a particular format or setup, note it in the brief and we’ll match it. We model in Matrix, Rhino and 3Design, so deliverables drop into the tools most casting houses and manufacturers already run — no file headaches on your end.
Need a physical pattern as well as the file? We can supply printed castable patterns from the finished model as an add-on — tree-ready wax or resin, printed straight from the same geometry we validated. Whether you take the digital file, the printed pattern, or both, it comes from one source of truth, so nothing drifts between what you approve and what you pour.
Built for how your shop actually pours.
No two casting operations are identical, so a file that ignores your process still creates work. We build to fit how you actually pour. If you tree and cast in-house, sprue and gate placement is planned around your tree layout and flask sizes, so models group efficiently and feed the way your metal likes to run. If you print in-house and cast out, or send files to a service, the geometry and formats are prepared so nothing has to be re-fixed by the next pair of hands.
For repeat, batch and catalog work, consistency is the entire point. A collection modelled to one standard holds that standard across every piece — the same wall minimums, the same feed logic and the same report format on every model — so your bench isn’t re-learning each one. High-volume runs are scoped to the batch and carry the identical castability checks as a one-off, which is what keeps a fifty-piece order predictable rather than a gamble. The result is fewer file-side surprises at the flask: when the geometry already respects your metal, your minimums and your workflow, one major variable is off the table before you ever pour.
What to send to start a casting-ready file
The more we know at the brief, the tighter the file. Useful inputs are the metal and karat you’ll pour, the finished dimensions that matter — ring size, length, gallery height — and the stone list, with sizes and counts, so seats are cut to fit. A sketch, photo, hand drawing, existing CAD or a written spec are all fine to work from; if something we need is missing, we ask before modelling so nothing is assumed. From there you approve scope, and the file comes back checked, weighed and ready for the tree.
Casting-ready, answered
What file formats do you deliver?
Casting-ready STL and 3MF for printing and casting, plus an editable 3DM working file when you’ll want to adjust the model later. Metal-weight and stone reports come with the delivery. If your printer or pipeline needs a specific format or setup, tell us in the brief and we’ll match it.
Do you account for shrinkage and minimum wall thickness?
Yes — both, on every model. We hold each surface to a castable minimum (typically from 0.8mm, adjusted for metal, span and mass) so nothing fills short, and we compensate the geometry for the metal’s shrinkage from wax to cast so the finished piece lands on your target dimensions and stones still seat.
Do you provide a metal-weight and stone report?
Every casting-ready file ships with both: an estimated metal weight across common metals and karats, and an exact stone count and size list. Together they let your quoting desk price the pour and the setting without weighing a test cast or counting stones by hand.
What if a model doesn’t cast?
If a piece fails to cast because of a file-side error on our end — a wall we set too thin, a seat that’s off, a mesh fault we missed — we re-cut the model and re-deliver at no charge. That covers the geometry we’re responsible for. The pour itself — metal, temperature, investment and burnout — stays on your bench, which is exactly why we engineer the file to take our part out of the equation.
Send your model or sketch. Get a castable file back.
Tell us your stones, metal and deadline. We’ll reply the same day with a quote, a realistic turnaround — typically 1–2 business days — and a file that’s checked, weighed and ready for the tree.