Blog · November 21, 2026

How to Calculate Polyurethane Capacity from Cycle Time

Polyurethane production capacity from cycle time is arithmetic before it is negotiation. Take the seconds a mold is occupied, convert them into shots per hour, multiply by the cavities in each mold, then subtract the time the line spends not making good parts. Done with measured numbers, that calculation survives a customer audit or a bank review. Done with nameplate figures, it usually does not. Most capacity arguments trace back to one side counting machine output and the other counting good parts per shift, with nobody writing down which product, mold, or shift pattern the number describes.

What Counts as Capacity Before You Do Any Arithmetic

Three numbers get called capacity in the same meeting, and each answers a different question.

Rated output describes the metering unit: grams per minute at a stated ratio, or shots per hour at nominal flow. It says nothing about molds, cure, or operators.

Line capacity describes the mold loop: shots per hour multiplied by cavities, limited by the slowest station in the cycle.

Demonstrated capacity describes what actually left the plant as good parts over a defined period. Only this number survives a purchase order, a customer audit, or a serious investment decision.

A capacity figure needs four qualifiers before it means anything: the part and mold identity, the cavity count, the material system with its demold time, and the time basis in hours per shift and shifts per week. Insert loading and any post-cure step belong in the qualifiers too, because both consume time the shot interval does not show.

The time basis causes more confusion than the arithmetic. Paid hours, scheduled production hours, and actual run time are three different denominators. A line quoted at 6,000 scheduled hours a year and then measured against 8,760 calendar hours looks 30% worse without a single thing going wrong. Decide the basis first, write it into the quotation, and keep it unchanged when two options are compared.

Break the Cycle Time into Elements You Can Measure

Cycle time is not one number. It is the sum of several, or the maximum of several, and which one applies depends on the line layout.

On a single-station machine, the mold cycle is the whole story: close, pour, dwell, demold, clean, reload. On a carousel or a mold-carrier line, the mold cycle is much longer than the interval between shots, and the interval is what sets output.

Element What to record How to measure Required or optional
Pour or injection seconds and shot weight shot record plus a check weight on a scale Required
Dwell to demold seconds from end of pour to release controller timer, confirmed by part quality Required, and set by material and part thickness rather than by the machine
Demold and removal seconds, including recovery from a stuck part stopwatch across 20 consecutive cycles Required
Mold cleaning and release seconds per cycle and frequency stopwatch, noting whether it happens every cycle or every tenth Required when a release agent is used
Insert loading seconds per insert set stopwatch, timed separately from demold Required only for parts with inserts
Index or transfer seconds between successive shots controller timestamps Required on multi-station lines

a1

Measure the controller elements and the manual elements separately, because they fail differently. The controller gives clean intervals and hides nothing, provided the timestamps are exported rather than summarized by an operator. The manual elements need a stopwatch and a person standing in the cell, since demold and cleaning are where the spread between a good cycle and a bad one appears.

Twenty to forty consecutive cycles on one product is enough to see the distribution. Averages taken across a mixed shift hide exactly what you need: the interval right after a mold change, the interval when one operator loads inserts and removes parts alone, the interval when the release agent runs low.

The Capacity Formula and a Worked Example

Three equations cover most mold-based polyurethane lines.

Shots per hour = 3,600 ÷ index time in seconds.

Parts per hour = shots per hour × cavities per mold.

Good parts per year = parts per hour × scheduled hours × availability × yield.

Index time here means the interval between successive pours into different molds. On a single-station machine it equals the full mold cycle. On a loop holding N molds, the mold cycle is N times the index time, which is why a ten-mold carousel can complete a part every 42 seconds while each individual mold sits closed for seven minutes.

The inputs below are illustrative. Substitute measurements from your own line; the arithmetic does not change.

Input or result Example value Effect on the answer
Index time between shots 42 s Sets the base rate; one extra second cuts hourly output by about 2.3%
Cavities per mold 2 Doubles parts per shot and doubles shot weight
Molds in the loop 10 Fixes the mold cycle at 420 s and adds cost and floor space, but changes no rate once the interval is set
Shots per hour 86 3,600 ÷ 42
Parts per hour 172 86 × 2 cavities
Good parts per year about 890,000 172 × 6,000 scheduled hours × 0.90 availability × 0.96 yield

a3

That last row is where quotations diverge. The same line, described by the same operator, can be quoted at 1.03 million parts gross or 890,000 good parts, and both statements are arithmetically correct. Only one of them can be delivered against.

Finding the Element That Limits the Line

The interval equals the longest station, not the sum of the stations. Four candidates account for most of the limits on a polyurethane line: pour time, dwell, demold and handling, and cleaning.

Pour time is shot weight divided by flow rate. A 300 g shot at 60 g/s takes five seconds. The same shot at 10 g/s takes thirty. Flow capacity therefore becomes a production variable whenever shots are heavy, and it stops mattering the moment dwell runs longer than the pour.

Dwell is set by chemistry and part thickness. Mold and material temperature shift it, but only inside the range that still delivers a part strong enough to demold without tearing, sinking, or internal voids.

Handling limits appear when one person must load inserts, demold, and clean at the same station. The measured interval then depends on the crew, and the capacity figure quietly becomes a staffing calculation.

Cleaning limits are the easiest to overlook because they repeat every cycle or every tenth cycle and rarely appear in a controller log.

The response differs by limit. If pouring is longest, raise flow or split the shot. If dwell is longest, add molds to the loop or reduce part thickness. If handling is longest, separate the tasks or automate extraction. Applying the expensive fix to the wrong element is the common error, and it usually shows up as a new machine with the same output.

Machine specification matters only in the element it touches. Haifeng Polyurethane Machinery specifies a flow range of 50 to 5000 g/min on its robotic pouring system, adjusted automatically to the robot’s travel speed, which lets pour time be matched to shot weight instead of accepted as fixed. The same system is specified with a color change under three minutes, which shortens a changeover that otherwise disappears from the capacity plan entirely.

An interval that sits above the pour time points to handling or curing, not to pump capacity. <How to Increase Output of PU Shoe Machinery> covers the operational levers behind output gains on PU shoe lines, including where those gains stop.

Adjusting Output for Downtime Scrap and Changeovers

Gross capacity is what the arithmetic produces. Net capacity is what a schedule can rely on. The gap comes from losses that should be recorded, not estimated from memory.

Loss category Where to record it Whether it belongs in the capacity figure
Breakdown and repair maintenance log with reason codes Yes, as availability
Mold change and cleaning changeover log with start and end times Yes, as scheduled loss or as availability
Material waiting and drum changes operator log Yes, as availability
Micro-stops and running below set speed controller interval trend Yes, as performance
Scrap and rework quality records per shift Yes, as yield
Breaks and no-order time shift schedule No, unless the quotation basis includes them

Availability is run time divided by scheduled time, and it captures everything that stopped the line. Performance captures the difference between the interval you planned and the interval you got while the line was nominally running. Yield captures parts that were made and then rejected. Keeping the three separate tells you which one to attack; blending them into a single percentage tells you only that something is wrong.

a9

Scrap deserves its own measurement rather than a line in the availability log. A metering ratio that drifts changes density and hardness, and the result appears as demold rejects, weight variation, or a customer complaint weeks later. Weigh parts from consecutive shots and plot the spread. A tight mean with a wide spread points to mixing or dosing instability, while a shifted mean points to calibration or material lot variation.

Changeover time behaves the same way. A changeover that takes longer than planned is a capacity problem even when every station is technically capable, and the fix is usually a cleaning sequence, a color-flush routine, or a batching decision rather than new hardware.

Scrap and changeover losses usually trace back to a short list of repeatable causes rather than to equipment age. <Common Problems in PU Shoe Production Lines> covers the recurring faults on PU shoe lines and how each one shows up in output data.

How Many Molds and Stations the Target Output Needs

The arithmetic also runs backwards, and this direction is the one that decides capital spending.

Start with the annual good parts the customer requires. Divide by scheduled hours, availability, and yield to get the required parts per hour. Divide by cavities to get the required shots per hour, then convert to an index time with 3,600 ÷ shots per hour.

Take a target of 1,000,000 good parts a year on 6,000 scheduled hours, with 0.90 availability and 0.96 yield. The line must average about 193 parts per hour, which with two cavities means about 97 shots per hour, or a 37-second index. If the material system needs 330 seconds before the part can be demolded, the loop needs 330 ÷ 37 = 8.9 molds. Nine is the minimum and ten gives curing headroom for a slow shot or a cold mold.

a6

The tradeoff sits between molds and cure time. Extra molds cost money, floor space, and clamp capacity, and past a point they add maintenance without adding rate. Shortening cure by raising mold temperature is cheaper to test, but it trades against surface quality, internal voids, shrinkage, and demold tearing, and the acceptable limit differs by part thickness and material system. On most projects the cheapest useful experiment is a temperature and release-agent trial on the existing mold before any station is ordered.

There is a second decision hiding in the same arithmetic. A second line duplicates metering, controls, and utilities, yet it decouples products so a color change on one does not stop the other. When the product mix is wide, that decoupling is often worth more than the apparent efficiency of one large loop.

Station count, carrier layout, and mold handling decide how much of the calculated rate survives contact with a real shift. <Fully Automatic PU Shoe Production Line Explained> covers how a complete line is configured from metering through mold carriers to demold.

Verifying the Number with a Production Trial

A calculated capacity is a hypothesis until a trial confirms it. The trial does not need to be long, but it does need to be specific.

  1. Fix the product, mold, and material lot before the run starts (required).
  2. Run at the target index until every station has completed at least one full mold cycle, then continue for 30 consecutive shots (required).
  3. Export every interval from the controller rather than the shift average (required).
  4. Weigh twenty consecutive shots and record the spread, not only the mean (required).
  5. Log scrap with reason codes at demold (required).
  6. Repeat a short run after a mold clean and after a drum change (optional, and usually revealing).

Demonstrated capacity is good parts divided by elapsed run hours. Compare it with the model. A gap of a few percent is normal and belongs in the availability assumption. A larger gap needs a named cause before the number goes into a quotation, because an unexplained gap tends to grow once production pressure arrives.

State the scope of the result as carefully as the result itself. A trial establishes performance for that material lot, that mold, that part, that crew, and that shift pattern. It does not establish performance after six months of wear, with a different lot, with a new operator, or with a heavier part. Writing that sentence into the acceptance record prevents the number from being quoted later in conditions it never covered.

a14

Send Your Cycle Data for a Capacity Check

If the arithmetic needs a second opinion, send the part drawing with shot weight, the cavity count, the index time you measure today, the demold time your material supplier specifies, the number of molds in the loop, and the annual volume you need. Those six inputs are enough to show where hours are lost and whether the limit is cure, pour, or handling, without treating one interval as the whole specification. Haifeng Polyurethane Machinery, info@haifeng-automation.cn, WhatsApp 86 13566296633.

FAQ

Should capacity be calculated per hour, per shift, or per year?

All three describe the same line, so calculate per hour first and convert. Hourly output comes from the index time and cavity count and needs no assumptions. Shifts and years add assumptions about availability and yield, and that is where departments usually diverge. State the basis in the same sentence as the number: 172 parts per hour at a 42-second index, or about 890,000 good parts a year at 6,000 scheduled hours, 90% availability, and 96% yield. A capacity figure without its basis invites an argument nobody can settle.

Does a higher-output machine raise line capacity?

Only when pouring is the constraint. A machine with more flow shortens the pour element, and if the pour is the longest station the interval falls with it. When dwell, demold, or cleaning sets the interval, extra pump capacity changes nothing except the energy bill and the purchase price. Work through the element list before buying output: measure the interval, identify the longest station, and only then decide whether the metering unit is the limit.

How do I account for mold changes and color changes?

Treat them as scheduled loss and keep them visible. If a changeover takes 20 minutes and happens twice a shift, that is 40 minutes out of roughly 480, and the annual figure has to reflect it. Batch similar products to reduce the count, shorten cleaning with a better flushing sequence, and record the actual duration rather than the planned one. A changeover that routinely overruns is a capacity problem even when no machine has failed.

What cycle time should go into a quotation?

Use a demonstrated interval with its conditions attached: this part, this mold, this material lot, this availability. Nameplate numbers describe a machine running under ideal conditions and are not a delivery promise. If the quotation must be issued before a trial, build the figure from measured elements on a comparable part and label the assumption clearly. Confirm it in the acceptance run and adjust the schedule if the gap turns out to be material.

Can one line run several products without losing capacity?

It can run them, but each product carries its own capacity. Blended output is the sum of each product’s hours divided by its own interval, minus the changeover time between them, so a mixed schedule always produces less than the best single product on the same line. Model the product mix explicitly. When the mix is wide and volumes are small, a second smaller machine or a dedicated low-pressure cell often beats stretching one line across every recipe.

If you’re interested, check out these related articles:

PU Safety Shoe Production Line Equipment List
What Is Dual Density PU Shoe Technology
Polyurethane Turnkey Plants: Guide to Polyurethane Bumper Stopper Manufacturing

Related Posts