In a polyurethane metering system, discharge pressure is not a value you dial in. It is the back-pressure the circuit presents to a positive-displacement pump at a given flow, temperature and viscosity, and the gauge simply totals every restriction between the day tank and the mixing head. That is why two machines with similar nameplates run at very different pressures on the same formulation, and why a gauge that creeps upward during a shift usually points to a changing restriction rather than a failing pump. Reading discharge pressure as a sum of resistances, rather than as a single setting, turns it into one of the most useful diagnostic signals on the line.
The Pump Sets Flow and the Circuit Sets Pressure
In a positive-displacement metering system, the pump moves a nearly fixed volume per revolution. Nothing inside it decides what pressure to produce. Pressure rises only because the material ahead of the pump resists: a check valve, a filter element, a narrow hose, a closed injection valve, the mixing chamber, the mold cavity. The pump keeps pushing until that resistance is overcome, the relief valve lifts, or the weakest part of the circuit gives way.
Read the gauge as a resistance budget. The day tank and inlet line spend part of it, the pump internals spend part, and the relief and back-pressure valves, the filters, the piping, the mixing head and everything downstream of the head spend the rest. Change any one item and the total moves, even though nobody touched the pump setting.
The same metering path appears in foam, elastomer and coating machines, with the pump, valves and mixing head arranged in much the same order. <Polyurethane Foam Machine: Types, Working Principle and Applications> covers how those machine types are built and where each one is normally applied.
The gauge port decides what the number means. A gauge at the pump outlet measures the pump, the filter and the hose between them, but nothing beyond. A gauge at the mixing-head inlet measures the restriction that actually shapes mixing, and it often reads several bar lower when a loaded filter and a long line sit in between. Comparisons between machines are only meaningful when both readings come from the same port, at the same output, temperature and material.

Which Variables Actually Move the Gauge
Most of the movement on a discharge gauge comes from six variables. Some are fixed at commissioning, others drift during production, and separating them is the core of pressure troubleshooting.
| Variable | Effect on discharge pressure | Where to verify it |
|---|---|---|
| Material viscosity at process temperature | Higher viscosity raises pressure at fixed flow; a few degrees of cooling can move the gauge visibly | Material temperature at the head inlet, not the tank |
| Output setting | Through a fixed orifice, pressure demand rises roughly with the square of flow | Recorded shot size and cycle time against the set value |
| Nozzle or mixing-chamber restriction | A small reduction in flow area raises pressure sharply | Head part number and chamber size compared between shifts |
| Filter condition | Loading adds restriction progressively, so pressure creeps upward | Differential pressure across the element, not absolute pressure |
| Hose and pipe bore, length and elevation | Long narrow runs spend pressure before the head ever sees it | Head-inlet reading against pump-outlet reading |
| Pump internal clearance and slip | Wear increases slip at pressure, so delivered flow falls while the gauge looks normal | A weighed timed shot compared with the setpoint |
The relationships behind the table explain why small changes feel large. For flow through a fixed orifice under turbulent conditions, pressure drop rises roughly with the square of flow and falls with the fourth power of the orifice diameter. Cut the nozzle area in half and the pressure demand climbs far more than the flow change suggests. In laminar flow through a long hose, the drop is closer to linear in flow and proportional to viscosity, which is why a cold drum shows up first as a high reading rather than as a mixing fault.
Viscosity is usually the largest single term in the resistance budget, and it is the one most often left out of a machine specification. What matters is the viscosity at the temperature the material reaches at the mixing head, not the value printed on the drum label at 25 °C. Filled systems add a second complication, because many filled polyols shear-thin: pressure falls as flow increases, then rises again when the line sits still and the filler settles.

Why the A and B Sides Rarely Read the Same
Impingement mixing depends on both streams arriving at the mixing point with comparable velocity. When the polyol side reads well above the isocyanate side, the faster stream penetrates further into the chamber, the impingement point shifts off-center, and the blend changes even though each pump still displaces its set volume.
A stable offset between the two sides is normal when the materials differ in viscosity or the lines differ in bore and length. A growing offset is not. It usually means one filter is loading, one pump is slipping, one side’s temperature control has drifted, or one back-pressure valve has moved from its recorded setting.
The ratio consequence is the part that costs money. Volumetric efficiency falls as pressure rises, and it falls faster on the side with the worn pump or the hotter material. The machine’s own flow or displacement readings can still show the correct ratio while the ratio actually delivered at the head has shifted, because those readings are taken upstream of the restriction that caused the change. A weighed shot taken at working pressure is the cheapest way to see it.
Pressure imbalance and ratio drift are usually discussed together, since both show up as weight variation long before anyone inspects a pump. <Polyurethane Foam Machine Types: The Scrap Cost of Inaccurate Metering> covers how metering error turns into scrap across different machine types.
Record both pressures, the output, the material temperature and the shot weight at commissioning. That baseline is what makes a later reading meaningful, because an absolute number without its conditions cannot be judged.

Recirculation, Back-Pressure Valves and What Normal Looks Like
Most metering systems circulate material through a back-pressure valve while the injection valve is closed. That valve sets a loop pressure which is present during standby, and the gauge will show it. Loop pressure is a transport and temperature-stability parameter. It says nothing about the restriction the material meets at the moment of injection, so a standby reading should not be compared with a shot reading.
Recirculation earns its place. It keeps temperature uniform through the line, prevents filled materials from settling, and keeps the head supplied with fresh material. The trade-off is additional shear and heat history, and a loop that runs at unnecessarily high pressure adds both without improving anything downstream.
Running at the relief valve setting is a different signal altogether. If the machine sits at relief during normal injection, the circuit is over-restricted or the relief is set too low for the installed head, and the pump is converting energy into heat rather than into flow.
What counts as normal also depends on the mixing principle. A low-pressure machine mixes by mechanical shear in a dynamic head, and pressure there is mainly a transport value. A high-pressure machine mixes by impingement, where pressure creates the velocity that does the mixing, so the same gauge number carries a completely different meaning.
The two principles are often compared on price when the real difference sits in what the pressure is doing. <High Pressure vs. Low Pressure Polyurethane Foam Machine: How to Choose the Right One for Your Production> covers how the two machine families differ and which production profiles suit each.
So the question “what pressure should this machine run at” has no answer until the gauge port, the flow, the material temperature and the mixing principle are named. A number carried over from another plant is not a specification.
Diagnosing a Pressure Change on a Running Machine
Work through the circuit in order instead of starting at the pump.
- Confirm the gauge and the port. Use the same calibrated gauge at the same port, and check that the port itself is not partly blocked with reacted material.
- Confirm the material state. Measure temperature at the head inlet rather than the tank, and confirm the batch and the drum condition.
- Weigh a timed shot. If shot weight fell while pressure rose, the cause is restriction or slip, not a sensor.
- Isolate the restriction. Change the filter element first, then swap the mixing head or nozzle. Pressure that returns to baseline identifies the component.
- Compare the back-pressure valve setting with the commissioning record.
- Test the pump at working pressure. A slip check at zero pressure tells you very little about behavior at 100 bar.
| Reading | What it supports | What it does not prove |
|---|---|---|
| High steady pressure on both sides with falling shot weight | A restriction upstream of the head, or pump slip | Which of the two; a two-pressure weight test separates them |
| One side rising over several hours | Progressive restriction, most often filter loading | Whether the material also changed; check the temperature log first |
| Pressure fluctuating with each pump stroke | Pulsation, or a starved inlet drawing air | Metering accuracy; a fluctuating gauge can still average out correctly |
| Normal pressure with poor mixing | The fault is downstream, in head geometry, ratio or temperature | Pump condition; that needs a shot check and head inspection |
| Normal pressure with a drifting ratio | Slip, worn check valves, or a sensor offset | Which side is at fault without a weighed comparison |
| Both pressures falling at unchanged output | Worn pump internals, or a bypass that has opened | That viscosity is unchanged; compare temperature before ordering parts |
A pressure reading describes the resistance at that point in the circuit. It does not certify the ratio delivered to the mold. Only a weighed shot or a calibrated flow measurement does that, and only for the conditions under which it was taken.

What to Put in an RFQ and an Acceptance Test
A pressure figure becomes useful only when it is tied to a point, a flow and a material state. These are the fields worth fixing before a purchase order.
| RFQ field | What to state | Why it changes the answer |
|---|---|---|
| Material and viscosity | Resin grade plus viscosity at the temperature you actually run | Viscosity is the largest single term in the resistance budget |
| Output range and ratio | Minimum and maximum continuous output, ratio with tolerance | Pressure rises with the square of flow, so the top of the range sets the design point |
| Mixing head or nozzle | Head type, chamber size, nozzle bore, and whether the head will be changed for other products | Head restriction fixes the pressure at the point that shapes mixing |
| Circuit layout | Hose bore and length, elevation, filters and valves between pump and head | Long narrow runs consume pressure the head never sees |
| Protection settings | Relief valve setting, filter differential alarm point, burst protection | Stops an operating pressure from quietly becoming a relief pressure |
| Measurement basis | Gauge port, gauge class, calibration method, and whether A and B are recorded separately | A pressure number without a location cannot be checked or repeated |
Reject this RFQ line: “Metering pressure: maximum 25 bar.” Replace it with: “Pressure at the mixing-head inlet, measured separately on the A and B lines with a calibrated gauge, remains within the band agreed at acceptance across the specified output range, at the agreed material temperature and ratio.”
As an example of how a specification holds together, the LJJ-series low-pressure machine from Haifeng Polyurethane Machinery is quoted at ±0.5% metering accuracy. That number means something only when the test states the output, the ratio, the material temperature and the gauge port used during the check. Without those conditions, a buyer cannot reproduce the figure on their own floor.
Agree a short sequence rather than a single number. Run at the bottom, middle and top of the output range at a fixed material temperature, record A and B pressures at the head inlet, weigh a timed shot at each point, and repeat the sequence once the machine is warm. A machine that holds pressure and shot weight across that sequence is behaving. One that holds pressure at a single point proves much less.

Send the Process Data Before the Pressure Specification Is Fixed
Pressure questions are usually answered badly because the missing information arrives late. Send the material TDS with viscosity at your process temperature, the required output range, the mixing ratio, the mixing head or nozzle you intend to run, hose bore and length, elevation, and the temperature setpoints from tank to head. With those, the resistance budget can be worked through before a machine is specified or rebuilt, and the gauge port can be agreed while the piping is still being drawn.
Email info@haifeng-automation.cn or message WhatsApp 86 13566296633 with those details, and we will go through the circuit item by item with you.
FAQ
Does higher discharge pressure mean better mixing?
No. In an impingement head, mixing quality depends on the velocity of both streams at the moment they meet and on the chamber geometry, not on the largest number the gauge can show. Above the point where the streams meet with enough energy, extra pressure adds heat, wear on the head and stress on the hoses without improving the blend. In a low-pressure dynamic head, pressure is mostly a transport value and mixing comes from mechanical shear, so raising pressure changes very little about the mix.
Why does pressure creep up during a shift?
The usual order is filter loading first, then temperature drift, then lot-to-lot viscosity variation, then a partly blocked head or a back-pressure valve that has moved. Filled systems add settling as a cause. Check the differential pressure across the filter element and the head-inlet temperature before touching the pump, because both are quick to read and account for most of the drift you will meet. A steady rise that survives a filter change deserves a head inspection.
Why do my A and B pressures differ?
Different viscosities, different hose bores and lengths, different filter condition and different pump wear all produce a legitimate offset. What matters is whether the offset is stable. If it is, record it as part of the baseline and leave it alone. If it grows, one side is changing: a loading filter, a slipping pump, a drifting temperature controller or a moved back-pressure valve. Confirm with a weighed shot before ordering parts.
Should I specify maximum pressure or working pressure when buying a machine?
Specify working pressure at a defined operating point, and treat the maximum as a limit rather than a running value. Give the supplier your output range, ratio, material viscosity at process temperature, head type and circuit layout, and ask for the expected pressure at the head inlet across that range. A machine that only has a maximum pressure on its data sheet leaves the buyer to discover the real operating window after installation.
If you’re interested, check out these related articles:
Polyurethane Foaming Machine: How It Works, Components & Maintenance Tips
PU Low Pressure Foaming Machine: Maintenance Checklist & Troubleshooting Guide
Phenolic Sandwich Panel Production Line vs PU/PIR Panel Line: Key Differences for Factory Investors
How Polyurethane Foam Machines Work: Types and Applications