How much weight difference between cavities is acceptable in a multi-cavity foam mould?

It should be agreed in the mould specification before T1 rather than assumed. Common practice on well-balanced tools is per-cavity mean part weights held within a few percent of nominal, measured on parts conditioned per ISO 291, with tighter windows as the programme matures. The study must separate cavity-to-cavity differences from ordinary shot-to-shot variation.

Can an unbalanced mould be fixed with per-cavity machine recipes?

Only partially. Trimming shot weight with orifice inserts is a legitimate tool-side fix, but recipes cannot put steam into a branch that arrives late or drain a low point holding condensate. Equalise what the tool can equalise - symmetric steam paths, drained branches, matched fill lines - then trim the small remainder, treating per-cavity recipes as a last resort.

Why does one cavity always make weak or under-fused parts?

A single consistently bad cavity almost always has a single consistent plumbing difference. Check that cavity's steam branch for path length or restrictions and for a low point without a condensate trap, whether its vents sit at its true last-to-fill point, its fill-line length and orifice, and its cooling flow. A pressure survey and thermal mapping usually locate the cause quickly.

How often should a multi-cavity mould be re-verified for balance?

Formally after any maintenance touching steam, water, venting or fill plumbing, using a cavity weight study of 25-30 conditioned parts per cavity. Informally, a monthly light sample of about ten parts per cavity catches slow drift. Tools with aggressive water or poor steam quality need this discipline most, because scale and fouling are rarely symmetric.

Do multi-cavity tools need more steam pressure or more steam flow?

More flow, not higher pressure. The steam phase needs mass delivered fast, which means adequately sized branches and a supply that holds pressure under peak draw. Raising header pressure instead usually worsens the problem and adds superheat risk at the cavities. Size pipework and supply for peak flow, not averages.

Multi-Cavity EPP and EPS Molds: Balancing Steam, Fill and Weight

Q: Does more cavities always mean better economics?

No. Cavity count is paid for in tooling cost, machine and utility capacity, and maintenance complexity. If the boiler cannot deliver the peak steam demand of the added cavities, the steam phase stretches and the promised output never materialises. Choose the cavity count that meets real demand with the available utilities.

Q: Is it better to buy two single-cavity moulds or one two-cavity mould?

A symmetric two-cavity tool halves machines and operators per part and holds balance well. Two single-cavity tools give scheduling flexibility and decouple maintenance. For one stable high-volume part with adequate utilities the multi-cavity tool usually wins; for mixed volumes or immature programmes, two tools reduce risk. Peak steam capacity and sustained demand decide.

2026-10-28Industry News19 min read

Quick answer

How multi-cavity EPP and EPS molds are balanced: symmetric steam and fill manifolds, condensate drainage, cavity weight studies and layout rules.

Quick answer: In a multi-cavity EPP or EPS mould, every cavity is a separate moulding cell that happens to share one steam chest, one fill manifold and one cooling circuit. Balance means all cavities receive steam, beads, vacuum and cooling water through paths of equal difficulty — so cavity 1 and cavity 8 produce parts of the same weight, density and dimensions. Imbalance is not primarily a steam-pressure problem: it is usually asymmetric plumbing — unequal steam path lengths, trapped condensate in dead legs, unequal fill-line lengths and vents that are not at each cavity's true last-to-fill point. The tool is designed for balance at the CAD stage, proven with a cavity weight study at commissioning, and kept in balance by cleaning and re-checking after maintenance.

CAD model of a multi-cavity EPS/EPP mould plate showing several cavities arranged on one steam plate — A multi-cavity mould plate: several cavities share one steam chest, one fill system and one cooling circuit. Every shared resource is an opportunity for cavity-to-cavity imbalance — and every one of them is decided in the tool design, not on the machine.

Key takeaways

Multi-cavity output is sold as one process capability, not per cavity. A 4-cavity tool whose cavities produce 4 different parts is worse than a single-cavity tool at the same total rate, because every downstream check now has a wider spread.
Balance is decided at design time. Symmetric steam and fill manifolds with equal path lengths, per-branch condensate drainage, and venting at each cavity's true last-to-fill point remove most imbalance before the first shot.
Peak steam demand scales with cavity count and is delivered in seconds, not minutes: a 6-cavity tool moulding 250 g parts needs roughly 2,000 kg/h of steam instantaneously during the steam phase, even though the hourly average is a fraction of that. Machine and boiler capacity — not floor space — often set the realistic cavity count.
Diagnose spatial vs temporal variation separately. If cavities differ from each other but each cavity is consistent shot-to-shot, look at the tool. If all cavities drift together over the shift, look at the process. Confusing the two wastes days.
The cavity weight study is the acceptance test for balance: 25–30 conditioned parts per cavity, mean and spread per cavity, compared against agreed limits — run at T1 and repeated after any maintenance that touches steam, water or fill plumbing.
Family tools (different parts in one mould) almost never balance and should be treated as a deliberate, documented compromise, not a default.
Imbalance returns slowly through fouling: scale in water channels, blocked vents and partially failed gaskets throttle cavities at different rates. A tool that was balanced at T1 needs periodic re-verification.

What “balance” means — and what imbalance costs

A single-cavity mould is judged on its parts. A multi-cavity mould is judged on consistency between its parts: the customer receives a mixed stream, and any property that differs by cavity — part weight, density, fusion quality, wall thickness, dimensions — becomes spread in every lot. Typical consequences seen in practice:

How cavity imbalance surfaces downstream
Imbalance symptom	What the customer sees	What it usually traces back to
Part weight differs by cavity beyond the agreed window	Packing weight variation, strength outliers, failed audits	Unequal fill-line lengths or shot weights; steam path asymmetry
Density differs by cavity	Inconsistent cushion or stiffness performance in the same box	Condensate-filled steam branches throttling some cavities
Fusion quality differs by cavity	Bead pull-out or weak corners on only some parts	Vents in wrong positions per cavity; cold dead-leg branches
Dimensions differ by cavity	Assembly interference on a subset of parts only	Uneven cooling water flow; different shrink by cavity
One cavity cycles differently	Bottleneck: the whole mould waits on the slowest cavity	Vent blockage or scale throttling that cavity's steam or water

Note how this differs from the time-series variation covered in the in-process quality control guide: SPC charts track how a process drifts over time, while balance work asks whether the spatial design of the tool gives every cavity the same process. Both dimensions of variation exist, and reading one as the other is the most common diagnosis mistake — chasing machine parameters for a problem that lives in the manifold.

How many cavities is realistic?

Cavity count is usually chosen from output arithmetic (target rate ÷ cycle time), then discovered to be wrong because of the constraints below. All of them should be checked before the tool is quoted, because they cap cavity count regardless of how neatly the parts fit the platen.

Constraints that cap cavity count — and how to check each
Constraint	Why it caps cavities	How to check at RFQ stage
Usable platen area	Each cavity needs its own steam-chest wall; parts typically sit with a web of material between cavities, commonly in the 50–150 mm range depending on part size and chest pressure, plus edge margins for insulation and clamping	Lay out the real part footprints on the machine's usable plate drawing, including manifolds and fill-gun positions
Peak steam supply	All cavities demand steam simultaneously; the peak lasts seconds but the boiler and pipework must deliver it, or every cavity's steam time stretches together	Peak-flow calculation (worked example below) against the machine's steam connection and boiler capacity; see the steam supply sizing guide
Fill system capacity	More cavities need more fill lines and air, delivered inside a few seconds of the cycle without starving the last cavities	Sum the shot weight and convey-air demand per cycle against the fill-gun and air supply capacity
Vacuum capacity	Vent-assisted moulding draws on one pump sized for the whole platen; more cavities mean more leakage paths and more air to extract per cycle	Compare required extraction against effective pump capacity per the vacuum system sizing method
Cooling duty	Heat per cycle scales with cavity count; the moulding water circuit must remove it at the same cycle time	Flow × heat-capacity × temperature-rise arithmetic against the plant cooling circuit
Ejection and handling	All parts leave in one stroke; take-out robots and downstream conveyors must handle N parts per cycle, and a heavier platen changes machine clamp behaviour	Check take-out handling against parts-per-cycle and machine clamp specification
Tooling economics	Each cavity adds machining, testing and maintenance cost; there is an optimum where output meets demand without buying capacity the programme never uses	Cost-per-part comparison of 1/2/4/8-cavity options — see what drives mould cost

Worked example: cavity count vs peak steam

Steam is consumed during the steam phase, which is a few seconds of the cycle, so cavity count multiplies an already-intense instantaneous demand:

Peak steam ≈ (cavities × part mass × specific steam use) ÷ steam-phase duration
6 cavities × 0.25 kg × ~3 kg steam/kg ÷ 8 s ≈ ~2,000 kg/h instantaneous
Same tool, averaged over a 60 s cycle: 4.5 kg ÷ 60 s × 3600 ≈ ~270 kg/h average

The specific steam use (here an illustrative ~3 kg per kg of part) varies with density, wall thickness and machine type — measure it on comparable running tools rather than trusting a catalogue figure. The structural point is what matters: a tool can be oversized in cavity count for a boiler that was sized on averages. If the boiler cannot deliver the peak, the steam phase stretches, cycle time grows, and the added cavities return less output than promised. This is the same peak-vs-average logic used to size the steam plant in the boiler sizing article, applied here as a cavity-count decision.

Where imbalance comes from: four mechanisms

Understanding the mechanism matters because each one leaves a different fingerprint, and each has a different design fix.

Four mechanisms of cavity-to-cavity imbalance
Mechanism	What happens	Fingerprint on parts	Design fix
Steam path asymmetry	Cavities far from the steam inlet sit behind longer, smaller or more throttled paths, arriving at lower pressure or later	Density and fusion degrade progressively with distance from the inlet	Symmetric manifold with equal path lengths; center-fed or reverse-return layout; generous branch sizing
Condensate and air pockets	Dead legs and undrained low points fill with condensate or air; steam cannot reach the beads behind them	Specific cavities always weak or under-fused, regardless of position	Per-branch condensate drainage with traps at every low point; slope branches back to drains; purge dead legs
Fill and vent asymmetry	Unequal fill-line lengths give different shot weights; vents not at each cavity's last-to-fill point leave trapped air	Weight spread that correlates with fill-line length; defects at consistent part locations	Matched line lengths from a symmetric fill manifold; trim with orifice inserts; map and place vents per cavity
Thermal position effects	Edge cavities lose heat to the platen edge; cavities near hot manifolds run hot; neighbouring cavities heat-soak each other	A gradient or hot-spot pattern across the platen, stable over months	Insulation at plate edges; keep steam branches off cavity walls they would overheat; layout so every cavity has similar neighbours

A useful habit at the design review: trace each utility from its source to each cavity and compare the drawings, not the promises. If steam travels a different distance to cavity 1 than to cavity 4, or if one branch has a low point the others lack, the imbalance already exists on paper — where it costs nothing to fix.

Measuring balance: the cavity weight study

Because bead foam parts are cheap to weigh and part weight correlates tightly with density and fusion, a structured weighing of parts per cavity is the fastest quantitative balance test. It belongs in mould qualification (alongside the acceptance criteria in the T1 trial report guide) and is repeated after any maintenance that opens the steam or water circuits.

Procedure

Condition first. Collect 25–30 consecutive parts per cavity after the process is stable, and weigh them only after conditioning per ISO 291 — freshly moulded parts are still losing moisture and their weight reading is misleading.
Compute per-cavity mean and spread. Mean part weight per cavity, plus standard deviation or range, and the coefficient of variation (CV) as a comparable measure.
Compare within-cavity spread to between-cavity differences. This is the diagnostic core of the study.
Set acceptance windows per cavity before the trial — typical practice is a weight window of a few percent around nominal per cavity, tightened as the programme matures. State the number in the mould specification, not after the data exists.

Reading the two signature patterns

Two patterns a weight study can show — and what each means
Pattern	Example (4-cavity tool, nominal 250 g)	Meaning	Action
Spatial signature — cavities differ from each other, but each cavity is consistent shot-to-shot	Means of 248 / 251 / 249 / 261 g; each cavity's CV small	The tool gives cavity 4 a heavier shot — typically a longer fill line, larger orifice, or a hotter/stronger steam branch	Tool-side: equalise fill-line lengths, trim shot with orifice inserts, re-check that cavity's steam branch. Do not tune machine recipes per cavity if avoidable — the imbalance will drift as the tool ages
Temporal signature — all cavities within a few percent of each other, but the whole set drifts across the shift	All cavities 248–252 g in the morning, 255–259 g after lunch	A process-level shift — bead prep, steam pressure, ambient conditions — not a tool problem	Process-side: work the drift with the SPC discipline in the process QC guide

Beyond weight, the study should sample density (per ISO 845) and fusion quality on a few parts per cavity, because two cavities can match on weight yet differ in fusion if their venting differs. Where dimensions matter for assembly, measure conditioned parts per cavity too — cooling-water imbalance shows up there first, as unequal shrink between cavities.

Designing for balance: the manifold and the layout

Most balance is won in three design decisions: how steam reaches the cavities, how beads reach the cavities, and where the cavities sit on the plate.

Steam manifold: equal paths, drained branches

Symmetric distribution. A center-fed or reverse-return (“H”-pattern) arrangement gives every cavity a steam path of similar length and restriction, instead of cavity 1 receiving full pressure and the far corner receiving the leftovers.
Size the branches for steam, not for convenience. Branch cross-sections are chosen so steam velocity stays in the usual saturated-steam distribution range (roughly 20–30 m/s in mains as a common design practice), keeping pressure drop between header and cavities small and measurable.
Trap every low point, per branch. Condensate does not flow uphill and does not share drains; a branch without its own trap at its low point fills with water and quietly starves its cavity. This is the single most common cause of “that one cavity” problems on otherwise symmetric tools.
Keep steam branches off cavity walls they would overheat, and insulate plate edges so perimeter cavities do not run colder than the centre.

Fill system: matched lines, trimmable shots

Equal line lengths from a common manifold. Fill guns arranged so every cavity sees the same resistance; where geometry forbids equality, trim at the gun orifice rather than accepting the spread.
Design for trim. Removable orifice inserts at the fill guns let commissioning equalise shot weights per cavity without re-plumbing — a cheap feature at build time and an expensive retrofit afterwards.
Fill pattern and vents are a per-cavity decision. Every cavity has its own last-to-fill point, and its vents belong there. Copying the vent pattern from a similar single-cavity tool is how trapped-air defects end up on one cavity only; the filling and venting logic is set out in the steam chamber design guide and the EPP mould design steps.
Fill gun selection matters at cavity counts: gun flow and nozzle sizing must suit the number of cavities and total shot weight, which is worth re-checking against the EPS fill gun specification when adding cavities to an existing line.

EPP automotive structural component mould showing one moulding cell within a mould assembly — Whether a programme runs one cavity or eight, each moulding cell needs the same steam, fill, vent and cooling quality. Multi-cavity design is the discipline of giving all of them exactly that.

Layout on the plate

Mirror symmetry beats raw density. A layout where each cavity has the same relationship to inlets, drains and edges — even at the cost of one cavity — usually outperforms a maximally packed plate with asymmetric plumbing.
Leave real webs between cavities (commonly 50–150 mm depending on part size and chest pressure) for steam-chest strength, gasket land and routing; cramming cavities together trades a tooling saving for years of balance trouble.
Same part, multiple cavities — not a family tool. Moulding different parts in one mould means one recipe must suit several geometries, which contradicts balance by construction. If a family tool is unavoidable, treat it as a documented exception with per-cavity acceptance windows agreed in advance.
Number the cavities. Engrave cavity numbers on the core inserts; they are the key that makes weight studies, defect maps and traceability records actionable, as described in the traceability chain.

Commissioning a balanced tool

Balance is proven in a sequence, because each step's data only makes sense if the previous step is clean:

Commissioning sequence for a multi-cavity tool
Step	What is checked	What “good” looks like
1. Pressure survey	Steam pressure at test ports near each cavity during a steam-only cycle	Comparable pressure and arrival time across cavities; outliers point at path restrictions
2. Thermal mapping	Surface temperature of each cavity through a steam phase (contact thermocouples or thermal imaging through opened tool)	No stable hot or cold cavity; gradients consistent with the design, not with plumbing
3. Drainage check	Condensate evacuation from every branch; no low point holds water after the steam phase	Branches clear; no gurgling or water carry-over at any cavity
4. Fill verification	Bead fill complete in every cavity, no starving or over-packed zones; shot weight per cavity	All cavities fill; per-cavity shot weights inside the agreed window after trimming
5. Cavity weight study	25–30 conditioned parts per cavity, weighed and compared	Between-cavity means within the agreed window; no spatial pattern
6. Fusion and dimensions	Density (ISO 845), fusion quality, conditioned dimensions per cavity	No cavity-specific defects; dimensional spread explained by design, not by cavity
7. Sustained running	The acceptance cycle-time and quality performance held over continuous running, per the T1 discipline	Balance holds when the tool is warm and the plant is at operating rhythm

Keeping balance over time

A tool that passed its weight study will not stay balanced by itself. Utilities foul and wear at slightly different rates per branch, so imbalance creeps back. The recurring checks are short:

Maintenance events that reopen imbalance — and the re-check each needs
Event	What it changes	Re-check
Water-side descaling	Removes scale unevenly; cooling behaviour shifts per cavity	Weight + dimension sample per cavity
Vent cleaning or vent replacement	Vents re-cut slightly differently change fill and fusion per cavity	Fusion inspection + weight sample
Gasket or steam-chest repair	Leak paths appear or disappear per branch	Pressure survey + weight study
Fill system work (guns, lines, orifices)	Shot weights and fill patterns shift	Per-cavity shot weight + weight study
Seasonal or plant-side utility changes	Steam pressure, water temperature or air supply at the machine changed	Quick all-cavity weight sample before full production

The most expensive form of imbalance is the slow one: a cavity that degrades over months gets absorbed into “normal spread” and quietly widens every lot's tolerance. A monthly 10-part-per-cavity weigh-in catches it for the cost of ten minutes of weighing.

Multi-cavity balance: review checklist

Symmetric steam distribution — equal path lengths, center-fed or reverse-return, branches sized for steam velocity
Per-branch condensate traps at every low point; no undrained dead legs
Fill lines of matched length from a common manifold; trimmable orifices at each gun
Vent pattern mapped and placed at each cavity's true last-to-fill point
Layout with mirror symmetry; realistic webs (typically 50–150 mm) between cavities; insulated plate edges
Cavity numbers engraved on core inserts
Peak steam demand calculated against boiler and machine capacity before cavity count is fixed
Same part in all cavities (or a documented family-tool exception)
Cavity weight study defined: 25–30 conditioned parts per cavity, acceptance window agreed before T1
Re-verification scheduled after any steam, water, vent or fill maintenance — plus a periodic light-weight sample

Frequently asked questions

How much weight difference between cavities is acceptable?

It depends on the programme, so it should be agreed in the mould specification rather than assumed. Common practice on well-balanced tools is per-cavity mean weights held within a few percent of nominal, with tighter windows as the programme matures. What matters most is that the window is agreed before T1 and measured on conditioned parts — and that the study separates cavity-to-cavity differences from ordinary shot-to-shot variation.

Can I fix an unbalanced mould with per-cavity machine recipes?

Partially, and usually not for long. Modern machines allow some per-cavity adjustment, and trimming shot weight with orifice inserts is a legitimate tool-side fix. But recipes cannot put steam into a branch that arrives late or drain a low point that holds condensate — those need the tool. As a rule: equalise what the tool can equalise, then trim the small remainder with inserts, and treat per-cavity recipes as a last resort that will drift as the tool ages.

Why does one cavity always make weak parts while the others are fine?

Follow that cavity's utilities one at a time: its steam branch (path length, restriction, and whether it has a low point without a trap), its vents (are they at this cavity's last-to-fill point?), its fill line (length and orifice), and its cooling circuit (flow comparable to the others?). A single consistently bad cavity almost always has a single consistent plumbing difference — and the pressure survey and thermal mapping steps usually find it in a morning.

Does more cavities always mean better economics?

No. Cavity count must be paid for three times: in tooling cost, in machine and utility capacity, and in maintenance complexity. If the boiler cannot deliver the peak steam demand of the added cavities, the steam phase stretches and the promised output never materialises. Choose the cavity count that meets real demand with the utilities you have, and revisit as the programme grows.

What is a family tool and why is it a balance problem?

A family tool moulds different parts in different cavities of one mould. It saves tooling cost but forces one recipe to suit several geometries, which differ in wall thickness, fill behaviour and venting needs. Balance as described in this article becomes impossible by construction. If a family tool is the right commercial choice, agree per-part acceptance windows and a validation plan for every cavity at the outset.

How often should a multi-cavity mould be re-balanced or re-verified?

Formally after any maintenance that touches steam, water, venting or fill plumbing — a weight study per the commissioning procedure. Informally, a light monthly sample (around ten parts per cavity) catches slow drift early. Tools running aggressive water or poor steam quality need the frequent version of this discipline, because scale and fouling do not respect symmetry.

Do multi-cavity tools need more steam pressure, or just more steam?

More flow, not higher pressure. The steam phase needs mass delivered fast, which means adequate branch sizing and a supply that holds pressure under peak draw — raising the header pressure instead usually just moves the problem and adds superheat risk at the cavities. Size the pipework and supply for the peak flow, as in the worked example above.

Is it better to buy two single-cavity moulds or one two-cavity mould?

A two-cavity tool halves the number of machines and operators per part and, designed symmetrically, holds balance well. Two single-cavity tools give scheduling flexibility and decouple maintenance. For one stable, high-volume part on adequate utilities, the multi-cavity tool usually wins; for mixed volumes or an immature programme, two tools reduce risk. The deciding inputs are peak steam capacity and real, sustained demand — both of which are known before the quote.

What should we ask a mould supplier about balance before ordering?

Ask to see the utility routing on the mould drawings: the steam manifold arrangement and branch path lengths, the condensate drainage scheme per branch, the fill manifold with line lengths per cavity, and the vent map per cavity. Then ask how the supplier will demonstrate balance — a defined cavity weight study with agreed acceptance windows at T1 is the short answer that covers all of it.