Casting yield: the single number that compounds across every later station.

The casting station is where the metal goes in and the units come out. The yield at that station — how much of what went in becomes usable casting — is the foundation of your cost structure. Everything downstream (setting, polishing, finishing, QC) is calculated against the number of units the casting station produced. When the casting station produces more good units from the same metal input, every subsequent cost per unit falls.

This is a short field note, not a long essay. The concept is not complicated; the implementation discipline is.

The definition and the benchmark

Casting yield = (weight of accepted castings) ÷ (weight of metal input to casting), expressed as a percentage.

If you melt 1,000 grams of 18k gold and produce 920 grams of accepted finished castings, your casting yield is 92%. The remaining 80 grams — the sprue, the overflow, the rejects — is your yield loss. Some of that metal is recoverable (returned to the melt for the next run); some is not (oxidation loss, surface contamination, QC rejects that cannot be remelted cleanly).

Benchmarks by alloy and technique, from my production data across Thai factories:

• 18k yellow gold, centrifugal casting: 91–95% (median 93%)
• 18k white gold, centrifugal casting: 88–93% (median 90%) — lower because of alloy reactivity and higher rejection rates
• Sterling silver, centrifugal: 89–94% (median 92%)
• 18k gold, vacuum casting: 93–96% (median 94.5%) — higher because vacuum reduces porosity and rejects

If your operation is running below the median for your alloy and technique, that gap is recoverable margin. If you are at or above the median, the next improvement requires equipment or process investment rather than process discipline.

The compounding effect

Here is why casting yield matters more than almost any other single metric in a casting operation:

Assume a production run of 200 rings, metal input cost of $18,000. At 90% yield, you produce 180 accepted rings at an effective metal cost of $100 per ring. At 93% yield — a three-point improvement — you produce 186 accepted rings at $96.77 per ring. The difference is $3.23 per ring, or $645 across the run.

That $645 is only the direct metal saving. The indirect saving is larger:

• Six additional accepted castings that do not need to be recast — saving one casting cycle's worth of labor, energy, and setup time
• Fewer QC rejects entering the setting and polishing stations — reducing rework cycles at both stations
• Lower scrap-handling cost — less time processing and segregating rejected material

Across a year of production runs, a three-point improvement in casting yield typically produces an 8–14% reduction in effective unit cost for casting-intensive operations. This is not a modeled projection; it is the consistent result I have measured in operations where yield tracking was installed against a baseline.

How to measure it correctly

The common measurement error is weighing input and output at the wrong points in the process. The correct measurement:

• Input weight: Metal going into the casting machine, after alloying and homogenization, immediately before casting. Not the weight from the invoice — the actual melt weight at the machine.
• Output weight: Accepted castings after first-pass QC, before any further processing. Pieces that fail QC at casting are not output — they are yield loss, even if they are returned to the melt.

Measure by production run, by alloy, and by mould type. Aggregate yield is less useful than yield by mould — because different moulds have different yield profiles, and the moulds with consistently low yield are the ones that need attention first.

The three levers for improvement

If your casting yield is below benchmark, the root cause is almost always one of three things:

1. Tree design. The geometry of the casting tree — how many pieces are attached, at what angle, with what sprue diameter — determines how metal flows into each mould cavity. Poor tree design produces cold shuts, porosity, and misruns. A casting technician with experience in tree optimization can often improve yield 2–4 points without any equipment change.
2. Metal temperature and alloy consistency. Casting temperature that is too high oxidizes the metal and produces surface defects; too low produces incomplete fill. Alloy consistency — whether your recycled metal is being properly refined before reuse — affects both yield and QC rejection rate. Regular alloy assay on recycled metal is the discipline most operations skip and most regret.
3. Flask preparation. Investment burnout temperature and time directly affect how cleanly the metal fills the mould. Insufficient burnout leaves residue that contaminates the casting and raises rejection rates. This is the most easily corrected of the three levers, and often the fastest source of yield improvement in operations that have not systematically reviewed their burnout protocol.

The discipline is measurement. If you are not measuring casting yield by production run and by mould, you are managing by assumption. The number is available — the weight scale and a spreadsheet are all the infrastructure required — and the decisions it enables are worth the three minutes per run it takes to collect it.