Quick answer. Confirm a two-component PU machine off-ratio condition with separate, timed A-side and B-side catch tests; calculate the delivered mass ratio instead of judging foam color or hardness alone. If the ratio is wrong, check material temperature and density first, then pump displacement, leakage, flowmeter scaling, pressure stability, valves, and the mixing head in that order. Do not correct a mechanical error by changing the chemical supplier’s specified formulation ratio.
An incorrect polyurethane ratio can produce brittle rigid foam, tacky elastomer, collapsed cells, soft spots, uncured adhesive, or carbon-dioxide bubbles caused by moisture. Those symptoms overlap with poor mixing, contaminated chemicals, incorrect temperature, and wrong catalyst level, so appearance alone cannot identify the failed component.
The fastest investigation creates three sets of evidence: actual component mass over a measured time, pressure traces during recirculation and dispensing, and cured-part test results from the same batch. This method prevents an operator from replacing a pump when the true fault is a blocked nozzle, drifting flowmeter K-factor, leaking changeover valve, or cold high-viscosity polyol.
Prove the Ratio Error Before Adjusting the Machine
Start with the formulation supplier’s target expressed on one basis: parts by weight, mass flow, or volume at a stated temperature. A specification of 100:50 by weight is not interchangeable with 1:1 by volume unless both components have the same density at the operating temperature.
For example, 1.00 L of isocyanate at 1.22 kg/L and 1.00 L of resin at 1.03 kg/L deliver a mass ratio of 1.22:1.03, or 100:84.4 by weight. A controller displaying 1:1 volumetric flow can therefore be chemically far from a formulation requiring 100:100 by weight.
Use clean, pre-weighed containers and a calibrated balance with resolution suited to the sample. A 1 g resolution is adequate for a 1,000 g catch but creates up to 1% quantization uncertainty around a 100 g sample; increasing the catch mass reduces that error.
Where the machine provides dedicated calibration outlets, collect each component for the same commanded interval, commonly 10 to 30 seconds. On fast-reacting high-pressure equipment, never improvise by disconnecting lines at the mixing head; use the manufacturer’s calibration, recirculation, or test-shot procedure.
Subtract the container tare and calculate delivered ratio as A mass divided by B mass. If a test produces 1,000 g of A and 520 g of B against a 100:50 target, the result is 100:52: the B-side delivery is 4% above its 500 g target, not merely “two points wrong.”
Repeat at least three catches at the production setpoint. Record each mass, collection time, component temperature, tank level, pump speed, displayed flow, and pressure. A stable 4% error suggests calibration or displacement; results that alternate between correct and wrong suggest air ingestion, cavitation, valve leakage, or intermittent restriction.
Do not create a universal acceptance limit. Foam systems, casting elastomers, FIPFG gaskets, adhesives, and polyurea have different chemical tolerances. Set the release limit from the material supplier’s technical data sheet and the finished-product control plan; procurement documents should state that limit explicitly.
Read the Evidence: Ratio, Pressure, and Part Symptoms
| Observed evidence | Likely mechanism | Discriminating check | Corrective action |
|---|---|---|---|
| Both catch masses are low by a similar percentage; ratio remains correct | Common speed-command error, underspeed drive, or incorrect shot-time setting | Compare measured shaft rpm and actual dispense time with the PLC command | Correct drive scaling, encoder feedback, or timing before touching individual calibration factors |
| One side is consistently 3-5% low at every output | Wrong pump displacement setting, worn metering element, or incorrect flowmeter K-factor | Compare gravimetric mass with controller total and pump revolutions | Calibrate the measurement channel; inspect the pump if mass per revolution is also low |
| Ratio is correct at low output but drifts at high output | Suction starvation, excessive viscosity, undersized filter, restricted hose, or drive torque limit | Trend inlet condition, motor current, temperature, and discharge pressure across at least three setpoints | Remove the restriction, restore material temperature, or resize the feed path |
| Pressure rises while flow falls on one side | Downstream blockage, partially closed valve, crystallized isocyanate, obstructed injector, or kinked hose | Measure pressure before and after the suspected restriction where approved test ports exist | Depressurize under the lockout procedure, then clean or replace the restricted component |
| Pressure falls and flow falls together | Tank depletion, air ingress, cavitation, slipping coupling, bypass leakage, or worn pump | Inspect suction bubbles and tank level; compare commanded rpm with actual rpm and mass per revolution | Seal the suction path, prime the circuit, repair the drive, or rebuild the metering element |
| Separate catches pass, but the cured part has streaks or soft zones | Ratio is reaching the head, but mixing energy or distribution is inadequate | Inspect injector condition, chamber deposits, rotor speed where fitted, and shot start synchronization | Service and align the mixing head; do not offset the ratio to hide poor mixing |
| First shot is wrong after a long pause; later shots pass | Valve seepage, unequal recirculation heating, dead-leg material, or delayed valve timing | Compare the first three shots after a controlled dwell and trend each component’s temperature | Repair sealing or timing and define a validated purge quantity |
Product testing should support the machine evidence. For cellular products, apparent density can be measured using ISO 845:2006 or ASTM D1622, subject to the applicable material specification and specimen-conditioning rules.
Density is not a standalone ratio meter. A change from 40 kg/m³ to 36 kg/m³ may come from ratio, moisture, blowing-agent level, mold temperature, venting, or shot mass. Pair density with catch-test results, cream time, gel time, tack-free time, and a retained reference sample.
Calibrate the Metering Pump and Flow Measurement Chain
Metering pump checks
For a positive-displacement pump, calculate actual delivery per revolution: collected mass divided by material density and pump revolutions. Compare the resulting cm³/rev with the machine’s commissioned value at the same temperature and operating condition.
A falling cm³/rev result can indicate internal slip, worn gears, damaged seals, a leaking bypass, or incomplete filling at the inlet. Internal slip often increases as discharge pressure rises, while inlet starvation usually becomes worse as speed or viscosity increases.
Check mechanical transmission before condemning the pump. A loose coupling, incorrect gearbox ratio, slipping belt, damaged key, or encoder scaling error can make the controller report 100% command while the pump shaft turns slower than expected.
Perform PU metering pump calibration at several production-relevant points rather than one convenient speed. A practical commissioning series might use 25%, 50%, 75%, and 100% of the intended operating output, with three repeat catches per point; these percentages are test points, not universal acceptance limits.
Flowmeter checks
A gear flowmeter measures displaced volume through pulse frequency. Its controller converts pulses through a K-factor such as pulses per litre; changing the K-factor by 5% changes displayed volume by 5% even though the physical flow has not changed.
Compare the flowmeter total with gravimetric truth. Convert mass to volume using density measured or documented at the test temperature. Using density at 20°C while the component runs at 45°C introduces a systematic conversion error, especially with filled or formulated polyol blends.
If gravimetric delivery is correct but the HMI is wrong, investigate the K-factor, unit conversion, pulse input, shielding, grounding, and missed pulses. If both the display and collected mass are low, the fault lies in actual delivery or in a control loop that is incorrectly reacting to the meter signal.
A meter can repeat a wrong number. Deposits, abrasive filler, magnet damage, bearing drag, or viscosity outside the meter’s intended range can shift its calibration without producing unstable readings. The purchase specification should state fluid viscosity range, filler content, expected flow range, allowable pressure drop, signal type, and required calibration record.
Temperature and material condition
Viscosity affects pump filling, leakage, filter pressure drop, atomization, and low-pressure mixing. Compare actual component temperatures at the tank, pump inlet, conditioned hose, and head; a single tank sensor cannot prove that both components reach the mixing head at the set temperature.
Inspect the resin side for settled fillers and the isocyanate side for crystals or reaction products caused by moisture. Agitation can restore a supplier-approved filled blend, but it cannot reverse chemical contamination or expired material.
Keep water away from the isocyanate circuit. Isocyanate reacts with moisture and can create carbon dioxide, pressure, solids, and blocked passages. OSHA identifies respiratory and dermal hazards associated with isocyanates, so maintenance requires the site’s chemical PPE, ventilation, spill-control, depressurization, and lockout procedures; see the OSHA isocyanates guidance.
Use Pressure as a Diagnostic Signal, Not a Ratio Setting
Pressure does not directly prove mass ratio. Two circuits can show equal pressure while delivering unequal mass because pump displacement, fluid density, viscosity, injector area, and leakage differ between A and B.
High-pressure impingement equipment commonly operates at pressures measured in tens of MPa, while low-pressure dynamic-mixing machines use much lower circuit pressure. The correct value is the approved process window on the machine documentation, not a generic internet setpoint.
Trend pressure rather than recording one gauge reading. Capture recirculation, valve opening, shot plateau, valve closing, and return to recirculation; a 100 ms transient can be invisible on an analogue gauge yet large enough to disturb the leading edge of a short shot.
A rising differential across a filter points toward restriction. A pulsating trace synchronized with pump rotation can indicate entrained gas, damaged pumping elements, or a pressure-control valve hunting. A slow decay during a closed-valve hold can reveal leakage, provided the machine manufacturer permits that test.
Compare both sides on the same time axis. If A reaches its stable flow 200 ms before B during a 2.0-second shot, 10% of the shot duration is exposed to start-up imbalance even if the middle of the shot is correct.
Verify pressure instruments against a calibrated reference suitable for the circuit pressure and chemical service. A transmitter with a 0-25 MPa range cannot establish a 0.05 MPa difference with useful confidence if its stated accuracy and installation error exceed that difference.
Inspect the Mixing Head and Specify a Machine That Can Be Verified
After separate delivery passes, inspect the final common point. In a high-pressure impingement head, unequal injector wear, blocked orifices, poor nozzle alignment, weak hydraulic actuation, or mistimed cleaning piston movement can reduce collision energy and leave streaks despite correct upstream flow.
In a low-pressure head, check mixer rpm under load, shaft coupling, blade condition, chamber clearance, dead zones, and cleaning sequence. A mixer that reaches its nominal rpm in air may slow in a 3,000 mPa·s formulation if the motor, gearbox, or drive current limit is inadequate.
Check valve synchronization at the physical valve, not only at PLC outputs. Solenoid command, hydraulic response, mechanical travel, and material arrival are different events; high-speed video, position sensors, or controller traces can expose a delayed component.
For short shots, collect whole-shot mass and verify individual calibration through the machine’s designed outlets. A steady-state flow test can miss the opening and closing errors that dominate a 0.5-second gasket dot or small elastomer casting.
Procurement should require an acceptance protocol using the buyer’s actual resin and isocyanate or fluids with documented equivalent density and viscosity. The protocol should identify target ratio, output points, shot sizes, material temperatures, number of repeats, allowable deviation, pressure records, and retained test data.
A manufacturer-direct project also allows the pump size, drive ratio, tank conditioning, filters, flowmeters, hoses, injectors, and mixing head to be selected as one metering chain. Buyers can review configurable PU foaming machine systems and request a design matched to their formulation, mold cycle, shot mass, and plant utilities.
Ask for component certificates and machine conformity documents required by the destination market, but do not accept a logo as proof of metering performance. Ratio acceptance needs serial-numbered instrument records, catch-test data, electrical drawings, pressure ratings, safety interlocks, and a witnessed factory acceptance test.
Specify maintainability as part of the purchase order: safe calibration outlets, pressure sampling points, individually trended temperatures, exportable alarm history, pump-revolution data, flowmeter totals, and access to injector and seal wear parts. These features shorten diagnosis from days of trial pouring to a measured comparison completed within one production shift.
For a custom line, provide the machine builder with each component’s safety data sheet, density-versus-temperature data, viscosity-versus-temperature curve, filler type and loading, target mass ratio, minimum and maximum output, smallest shot, largest shot, allowable cycle time, and finished-part test method. Missing fluid data forces pump and heater selection to rest on assumptions.
FAQ
Q: What is the first test for a two-component PU machine off-ratio problem?
Run separate, equal-time gravimetric catches through the machine’s approved calibration outlets. Record net A and B masses, actual time, temperatures, pressures, pump speed, and displayed flow for at least three repeats.
Q: Can equal A-side and B-side pressure prove a 1:1 ratio?
No. Equal pressure only describes hydraulic resistance at the measurement points. Different densities, pump displacements, viscosities, injector areas, or leakage rates can produce unequal mass flow at the same pressure.
Q: How often should PU metering pump calibration be performed?
Set the interval from drift history, chemical abrasiveness, duty cycle, and the product control plan. Recheck after pump or flowmeter service, seal replacement, formulation change, abnormal pressure, prolonged shutdown, or any finished-part trend outside its approved limits.
Q: Why does the ratio pass during calibration but fail in production?
The calibration may test steady flow while production uses short shots. Valve delay, pressure transients, hose temperature differences, first-shot dead material, or mixing-head restriction can distort a 0.5-2.0 second dispense without appearing in a 30-second catch.
Q: Should operators change the ratio to correct soft or brittle foam?
Only after confirming the formulation supplier authorizes the change. First verify component identity, batch condition, temperature, mass delivery, mixing, shot mass, mold temperature, moisture, and cure schedule; ratio changes can hide a machine fault and move the isocyanate index outside the validated formulation.
Q: Which records should a buyer request during factory acceptance?
Request raw catch masses, calculation sheets, test-fluid density and temperature, pump revolutions, flowmeter totals, pressure and temperature traces, repeatability results at multiple outputs, instrument calibration status, alarm tests, and the signed acceptance criteria. Finished samples alone cannot show how the ratio was achieved.