Section Fabrication vs One-Piece Casting: When Splitting a Design Makes It Better

One-piece casting is the default answer in jewelry manufacturing — not because it is always the best route, but because it is the easiest one to quote. Here is when that default fails, and what splitting a design into fabricated sections actually costs.

Why one-piece is the default

Casting is the industry's reflex. One mould, one pour, one object — the shape arrives complete and every step after it is subtraction. A floor tuned for repeating work loves this: the same fixtures, the same sequence, the same polishing route, piece after piece. When a design reaches most workshops, the first question is "can this be cast?" — and when the answer is yes, the conversation usually ends there.

That is not laziness. It is what the equipment, the quoting and the whole production rhythm are built around. The trouble starts when the honest answer is "yes, but" — the shape can be cast, yet the piece still cannot be finished as one object.

Three situations where one casting fails

1. When the seat is inside the geometry

A pierced design holds its stones in seats that have to be cut from inside the form. When the piercings that make the design work are the same features that block the tool path to the seat, casting produces the shape but not a usable seat — the stone ends up shallow, and a shallow stone does not stay put. The way through is to stop treating the piece as one object: divide it so every seat sits exposed on its own section, fabricate the sections separately, then join them. That is section fabrication, and it is a routing decision, not a clever trick.

2. When finished work has to survive the last joint

Assembly order is a design decision, not a shop-floor detail. If a stone has to be set, or a surface brought up, before the final joint closes, then the joining method decides whether that earlier work survives. A route that heats the whole assembly forces finished work to be redone — or worse, quietly opens a joint that was closed an hour ago. A route that joins locally lets the sequence run in the order the design actually needs. That is what laser welding is for on our bench: energy into a spot measured in fractions of a millimetre, parent metal fused to parent metal, no filler, and the work around the joint left alone.

3. When the piece has to be opened again later

A solid one-piece casting is a closed system. If a joint fails in service, a size has to change, or one section is damaged, the repair is a remake. A piece built from sections can often be re-opened at the joint that matters, reworked, and closed again. The same property that makes complex openwork buildable also makes it maintainable — which matters more than most buyers realise at the order stage.

What splitting costs — the honest part

Splitting is not a free upgrade. Every joint is a place to get it wrong: alignment risk grows with the number of sections, each seam has to be dressed back until it disappears, and the full set has to be dry-fitted and reviewed as one piece before any joining starts. That checkpoint is not a formality — it is the last moment when six parts can still become one flower instead of six assemblies. On that six-section openwork ring, the dry-fit review is exactly where the set was approved before a single weld was made.

So the rule we work to is simple: split when the geometry demands it, never because it sounds more impressive. The number of sections is kept as low as the geometry allows, because more parts is not more craft — it is more risk, and it shows up in the quote.

How the decision actually gets made

Before anything is quoted, the design is read as a manufacturing problem, and three questions get answered:

• Can every stone seat be reached with a tool, on the section it belongs to?
• Does any step have to happen after another step that cannot take heat?
• Will the specified finish survive the assembly sequence?

Three yeses: the piece stays one object. Any no: the design is divided exactly where the answer changes — and no further.

This is also where the 3D model earns its keep. Wall thickness, seat depth and the clearance between sections are all visible in the model long before metal is committed, and a printed prototype lets the split be judged in the hand rather than on a screen.