Engineering Answers / Tooling & Cost
Most buyers ask for eight cavities because it sounds professional. The right number is the one that matches your annual volume to a tool you can still control.
Cavity count follows from annual volume divided by the number of shots a year can deliver, adjusted for yield. A tool needs enough cavities to ship the volume without running the press flat out, but every extra cavity raises tooling cost, adds alignment and flash risk, and makes cavity-to-cavity variation harder to hold. Below roughly 5,000 parts a year, a single- or two-cavity tool is usually correct; above 50,000, an 8- to 16-cavity tool normally pays back.
When a buyer asks how many cavities, they usually mean one of three things: how do I get the lowest unit price, how fast can you ship, or how much do I have to invest now. Those three pull in different directions, and the answer is a compromise between them — not a standard number.
Cavity count is set by one equation and three constraints.
The equation: annual volume ÷ (shots available per year × yield) = cavities needed.
The constraints: what the tool costs, how much quality risk you can carry, and whether the geometry can be filled evenly across that many cavities.
Take a part with 30,000 pieces a year. A single-cavity compression tool might deliver a few thousand parts a week on one press; the same part in a 4-cavity tool delivers four times as many per cycle. But the press does not run 8,760 hours a year — it runs the shifts you staff, minus changeovers, maintenance and downtime. Use realistic numbers, not theoretical ones.
A worked shape of the calculation, with the assumptions stated rather than hidden:
If the honest calculation says a single-cavity tool can already cover the year with spare capacity, more cavities buy you nothing except a larger capital number. If it says you would need to run three shifts, seven days a week to keep up, cavity count is the cheapest capacity you will ever buy.
Tooling cost scales with cavity count, but not linearly — the second cavity is cheaper than the first because the frame, ejection and heating are shared. The unit price falls for two reasons: tooling amortisation spreads over more parts, and press hours per part drop.
The steel decision sits underneath this. We build in NAK80, S136 stainless and H13 hardened tool steel at HRC 48–54, rated from 100,000 to 500,000+ cycles. A 100,000-cycle tool is correct for a two-year programme; a 500,000+ cycle tool only pays back if the programme actually runs that long. Specifying steel beyond your lifetime volume is the single most common way buyers overspend — and it usually arrives in the same email as an oversized cavity count.
A high-cavity LSR tool is a stiffness problem before it is a machining problem. Every cavity has to be shut off at the same time, at the same pressure, at the same temperature. A small parallelism drift across a large plate opens the shut-off faces on one side while the other side stays tight, and LSR will enter any gap under 0.005 mm. That is why flash control is the recurring warning in high-cavity tools.
Two practical consequences:
| Annual volume | Usual starting point | Why |
|---|---|---|
| Under 2,000 | Single cavity, soft tooling | Tooling dominates; volume never repays more cavities |
| 2,000–10,000 | 1–4 cavities, steel or staged | Enough capacity without over-investing |
| 10,000–50,000 | 4–8 cavities | Press hours start to matter more than tooling |
| Above 50,000 | 8–16 cavities, hardened steel | Tooling amortises; steel life matches programme |
These are starting points, not rules. A large part with a thick section can need a single cavity at 40,000 pieces a year because the press cannot hold more than one. A small O-ring can justify 32 cavities at 200,000 pieces.
The expensive mistake is committing to a high-cavity production tool before the design is proven. Aluminium rapid soft tooling produces 10 to 2,000 real injection-molded parts in 7 to 12 days for a fraction of steel tool cost. That run does three things: it proves the geometry, it measures the actual shrinkage instead of relying on the 2.5–3.5% book value, and it gives you parts for functional and regulatory testing.
Then the production tool is built once, at the right cavity count, with real data behind it. Custom mould development typically takes 15–25 days depending on complexity and cavity count — a schedule that gets much less painful when the design is already validated.
Not the cavity count — the annual volume and the expected programme life. Those two answers decide steel grade, cavity count and whether a staged path saves money overall. Buyers who arrive with a cavity number already chosen usually arrive with it chosen for the wrong reason.
Cavity count is capacity planning with a quality ceiling. Size it from annual volume and realistic press hours, not from what sounds impressive. Below a few thousand parts a year, one cavity is usually right. Above fifty thousand, the tool pays for its cavities quickly. In between, quote both and compare the total — including the cost of inspecting every cavity, because that is where the savings quietly disappear.
Yes, more often than buyers expect. If annual volume is under a few thousand pieces, or the part is large, or the tolerance is tight enough that you want every part coming from the same cavity, a single-cavity tool is cheaper to build, faster to qualify and easier to diagnose when something drifts. Paying for eight cavities to run 2,000 parts a year is capital that never earns anything back.
It changes the amortised tooling cost per part and the machine time per part, which is most of the unit price. Material per part barely moves. A 16-cavity tool spreads the tooling investment over more parts and reduces press hours per part, but the tool itself costs several times a single-cavity tool, so the benefit only appears once volume is high enough to consume the output.
Rarely in a way that saves money. Adding cavities means a new cavity plate, new alignment features and often a new frame, so the cost approaches a new tool while the old one becomes scrap. The practical version of "adding cavities later" is a staged plan: build a single-cavity or soft tool first to validate the design, then build the production tool at the final cavity count once the volume is confirmed.
Because clamp force and alignment have to be correct at every position at once. A small drift in parallelism or a difference in thermal expansion across a large plate opens the shut-off faces on one side of the tool while the other side stays tight. LSR will find any gap under 0.005 mm, so cavity count turns a tooling tolerance question into a tooling stiffness question.