Stable material temperature is critical in polyurethane processing because it sets three things at once: component viscosity, component density and reaction speed. Move any of them and the machine keeps pumping while the output quietly changes. The usual mistake is to treat temperature as a utility setting instead of a process variable. A tank controller can read 25 °C all night while the polyol at the pump inlet sits several degrees colder, because the sensor is on the jacket and the material in the tank has stratified. What matters is the size of the band the material stays in, measured where it enters the metering pump and the mixing head.
What Temperature Actually Changes in the Mixing Head
Mixing quality is decided inside a small chamber in a fraction of a second, and temperature is one of the inputs that decides it. In a low-pressure head with a rotating mixer, shear depends on rotor speed, chamber clearance and the viscosity of the two streams. Warm polyol arrives thin enough to be thrown against the chamber wall and folded into the isocyanate. Cold polyol arrives thick, slides past the rotor and leaves the head in streaks that look mixed but are not. The same RPM produces a different mixing result because the material, not the motor, sets the shear.
In an impingement head the mechanism is different and the sensitivity is similar. Two streams meet at high pressure and the energy of that collision does the mixing. Viscosity changes the pressure drop across the orifice, so a component running colder than it did at commissioning reaches the impingement point with less kinetic energy and a different flow split. Operators raise pressure to restore the reading on the gauge, which restores the number and not always the mixing.

Density is the second effect and the one most often missed. Metering pumps move volume; the recipe is written in mass. Polyol, isocyanate and curative all expand and contract with temperature, and they do not do it at the same rate. A line can hold a perfect 1:1 volume ratio while the mass ratio walks away from the number on the technical data sheet.
The third effect is reaction speed. The hydroxyl and isocyanate reaction is exothermic and faster at higher temperature, so cream time, gel time and demold time follow the material temperature rather than the mold clamp. When the material runs cold on a winter morning, the shot that filled the cavity in summer now gels late and sags at the far end. When it runs hot, the same shot gels in the runner before it reaches the last section of a long mold.
Phase state belongs in this list as well. Pure 4,4′-diphenylmethane diisocyanate is solid near 40 °C, and a drum stored in an unheated bay can carry crystals that a jacket set to 40 °C will not fully redissolve. MOCA curative has to be completely molten before it is metered; a partly solid phase meters as a rich or lean shot at random. Both problems look like a metering fault until somebody opens the temperature log.
How a Temperature Error Becomes a Ratio Error
On a volumetric machine, the ratio error produced by temperature arrives through three doors. The first is density. Take a pump pair holding a 1:1 volume ratio. If the polyol density falls by half a percent while the isocyanate temperature holds, the mass ratio moves to 0.995:1. Nothing on the machine has changed, no pump has worn, and the shot record still reads 1:1.
The second door is pump slip. A gear pump with worn or cold internals leaks a small fraction of each revolution back to the inlet. Cold material is thicker, inlet pressure drops and the pump has to pull harder; filled, high-viscosity grades used in elastomer and imitation wood work are the most sensitive to this. When one pump slips more than the other, the ratio moves in a direction that follows the temperature difference between the two components, not the accuracy label on the machine.
The third door is response time. A thermoregulator that is working perfectly still needs time to react to a cold drum charged into a half-full day tank. For the shots immediately after that change, the material at the head is colder than the sensor position, and the machine is metering an average of two temperatures. This is why a ratio check taken right after a drum change can look acceptable while the parts made twenty minutes later fail.
The practical way to separate temperature from wear is a weigh-shot test run against a temperature log. Take a shot into a tared container at the start of the shift, after a stoppage and at the end, and write the pump inlet temperatures beside the weights. If the weights drift with the temperature and return when the temperature returns, the machine is doing what it was told. If the weights drift while the temperature holds, the problem is mechanical.
Ratio data means little when it is read without the material temperature beside it. <PU Foam Injection Machine: Key Components & Calibration Best Practices> covers how the metering components and the calibration routine fit together on an injection machine, which is where the mechanical and thermal causes get separated.
Where Temperature Drifts on a Running Line
A drum standing in a loading bay follows the weather. A drum that has been in the bay overnight is not fixed by a two-hour preheat if the core is still cold, because polyol conducts heat slowly and a jacketed tank does not stir itself. The material closest to the wall reaches setpoint first, and the material in the middle of the tank may be several degrees behind it for hours.
Day tank level changes the problem from week to week. A high level gives a long residence time in the jacket and a fairly uniform bulk temperature. A low level gives a short residence time, more sensitivity to what just came in from the drum and a faster response to a cold top-up. Neither condition is visible on the machine display, which is why level and temperature need to be read together.
Transfer lines are the next source of loss. An insulated line holding material at temperature and an uninsulated line sitting idle between shifts behave very differently. When recirculation is switched off at the end of the shift, the material in the run cools and the first shots of the next shift push that cold volume into the head. Dead legs and long runs to a robot station make the effect larger.
Mechanical energy works in the opposite direction. Pumps and mixing heads add heat to the material, so after several hours of high-pressure running the head can be warmer than the tank it is fed from. A line that started the shift perfectly balanced can end it running hot at the head and cold at the tank, with the average hiding both.

Choosing the Right Temperature Control Point
No single device fixes material temperature, because the material passes through several places where heat is gained or lost. The useful question is which point in the chain needs control, and what that point can and cannot correct.
| Control point | What it stabilises | Fits when | Limitation |
|---|---|---|---|
| Tank jacket with thermoregulator | Bulk temperature in the day tank | Tank volume is large and residence time is long | Slow to respond, and the tank stratifies without agitation |
| Inline heat exchanger in the recirculation loop | Temperature at the pump inlet and head | Lines are long or shot rate varies through the shift | Only works while recirculation is running |
| Container preheating oven | Incoming material before it is charged | Drums arrive cold or viscosity is high | Batch treatment, not a closed control loop |
| Trace heating on transfer lines | Heat loss along the run | Long distance between storage and machine | Risk of local overheating near the pump if not thermostatted |
| Mixing head chamber control | Temperature at the point where reaction starts | Short shots and fast reactivity make start conditions decisive | Cannot correct drift that happened upstream |
Each row in that table solves a different problem. A tank jacket controls bulk temperature but says nothing about what reaches the head at the end of a long line. An inline exchanger controls the material where it matters but only while the loop is actually circulating. Choosing one and expecting it to cover the others is the most common specification gap I see on foam and elastomer lines.
Matching a machine type to a process requirement is easier when the temperature chain is mapped first. <Polyurethane Machinery 101: A Comprehensive Guide to Equipment Types and Engineering Selection> covers how equipment categories line up with different process requirements and where the engineering decisions sit.
Component-level control is one way to close the last gap in the chain. A mixing head with a temperature-controlled chamber holds conditions at the point of reaction rather than at the tank, and Haifeng Polyurethane Machinery lists a chamber stability of ±0.5 °C with a mixing uniformity of ≥99.5% for its E-type head, alongside a drum preheating oven that covers room temperature to 90 °C for 200 kg containers on the incoming-material side. The number itself matters less than the location: chamber temperature is controlled where the reaction starts, separately from whatever the tanks are doing.
What to Measure and How Often
Temperature control is only as good as the measurement point behind it. Five checks cover most of what goes wrong.
- Measure at the pump inlet, not at the jacket. The jacket sensor tells you what the heater is doing. The pump inlet tells you what the machine is metering.
- Read both components at the same moment. A polyol temperature logged at 09:00 and an isocyanate temperature logged at 11:00 describe two different machines.
- Track the gap between tank and head. A gap that stays constant is normal heat loss. A gap that widens through the shift points to recirculation, line insulation or a failing exchanger.
- Run weigh-shots on temperature events, not only on clock time. After a drum change, after a stoppage, after a shift handover and at the end of the run.
- Set the alarm band from the ratio tolerance, working backwards. Find the temperature change that moves the weigh-shot result outside tolerance, then place the alarm inside that figure rather than at whatever the controller can display.
- Check probes against a calibrated reference quarterly. A probe reading one degree high at working temperature is a systematic bias in every record the line produces.
An alarm nobody acts on is only a data point. A deviation alarm with a short delay, combined with an interlock that holds the pour or stops the shot when the deviation persists across a set number of shots, turns temperature control into something the process depends on rather than something the operator watches.
Matching the Symptom to the Temperature Cause
Most defects blamed on temperature are blamed correctly, but not always for the stated reason. The table below separates the mechanism from the symptom so the first check lands in the right place.
| Symptom | Temperature mechanism | First check |
|---|---|---|
| Short shots toward the end of a run | Component cools, viscosity rises and pump output falls | Log pump inlet temperature against shot weight |
| Density drifting between batches | Volumetric pumps deliver a changed mass when component density changes | Weigh-shot at the start and end of the batch |
| Surface bubbles and pinholes | Dissolved gas comes out of solution as temperature rises or pressure drops | Compare component temperature with the day the drum was charged |
| Hardness variation in elastomer parts | Prepolymer and curative temperatures diverge and the ratio shifts | Read both component temperatures at the head |
| Color streaks after a changeover | Cold color paste and cold polyol disperse at different rates | Confirm the paste is conditioned to the polyol temperature |
| Gel time moving through the shift | Reaction rate follows material temperature, not the mold | Trend gel time beside head inlet temperature |
Temperature is one cause among several. A cold component and a worn pump can both produce a short shot on the same morning, and raising the setpoint to fix the second one only adds a new variable. Confirm the mechanism before touching a setpoint.

Writing Temperature Requirements Into the Machine Specification
Most machine specifications say “temperature control” without a band, a measurement location or an ambient range. That sentence cannot be tested, which means it cannot be accepted or rejected at handover.
| Requirement | What to write | How to verify |
|---|---|---|
| Control band | The temperature band at the mixing head inlet across the site ambient range | Log the trend over a full production shift |
| Metering accuracy | Ratio accuracy at steady temperature, plus the band inside which it holds | Weigh-shot at three points, with temperatures recorded |
| Recovery time | Time to return inside the band after a drum change or a cold start | Timed trial starting from a cold drum |
| Alarm and interlock | Deviation threshold, alarm delay and the interlock action | Simulate the fault and record the response |
| Utility envelope | Heating and cooling capacity, flow and pressure needed to hold the band | Compare the utility schedule against the site supply |
| Records | Temperature trend export, calibration certificates and alarm history | Review the data files during acceptance |
Acceptance testing has to include the conditions that cause drift, not only the conditions that flatter the machine. A run at steady production rate in a warm room proves the machine can hold a band. A run starting from a cold drum in a cold bay proves it can recover, which is what the plant will actually ask of it every winter morning.
One more boundary is worth writing down. A band quoted at 25 °C ambient is not the same promise at 5 °C ambient, and a machine rated for a maximum flow may hold its temperature band only below that flow. Nominal ratings and continuous ratings are different commitments, and the temperature specification is where that difference shows up first.
Machine selection gets more reliable when the specification lists measurable conditions rather than categories. <Polyurethane Foam Machine Selection: A 2026 Buyer’s Guide> covers how selection criteria are weighed across foam machine types and where a stated band should be checked against real operating conditions.
Send the Material Data Before You Compare Machines
Temperature requirements cannot be quoted from a product list. They come from the material system, the site and the tolerance you can live with. Send the technical data sheets for both components, the target output and shot size, the ambient range the plant actually sees in winter and summer, the ratio tolerance that matters to your product, and the point where you currently measure temperature. With those five items, a machine configuration can be checked against your process instead of against a brochure.
Haifeng Polyurethane Machinery
Email: info@haifeng-automation.cn
WhatsApp: +86 13566296633
FAQ
What is the correct material temperature for polyurethane processing?
There is no universal number. Most foam systems are conditioned close to the value the technical data sheet gives for the polyol and the isocyanate, commonly somewhere in the low twenties Celsius, while elastomer prepolymers are held much hotter to keep viscosity workable for metering. The useful target is the band the sheet specifies at the point where the material enters the mixing head, not the setting on the tank jacket. Confirm the number with the material supplier and then verify it with a weigh-shot at that temperature.
How much temperature variation is acceptable?
Acceptable variation is the band that keeps your weigh-shot result inside your ratio tolerance. Deriving it backwards from the product is more useful than adopting a figure from a machine brochure. A quoted stability of ±0.5 °C at the mixing chamber means little if the tank is 6 °C colder and nothing recirculates between shots. Measure the actual band at the pump inlet and head over a full shift, then decide whether it is tight enough.
Does temperature change the mixing ratio or only the viscosity?
Both, through separate mechanisms. Viscosity changes pump slip and the pressure drop through the mixing head, which shifts the delivered volume. Density changes the mass that a fixed volume carries, so a stable volume ratio can still deliver a drifting mass ratio when the two components sit at different temperatures. The viscosity effect usually shows up as a flow change and the density effect as a slow ratio drift, which is why one can be mistaken for the other.
Why are the first shots after a break always wrong?
Because the material sitting in the lines cooled while the machine was stopped and the jacket was holding a tank the head is no longer connected to thermally. Recirculating before production and purging the cold volume solves most of it. If the problem persists after a proper purge, check whether the day tank stratified during the stop, and whether the drum charged at restart was conditioned to the same temperature as the material it replaced.
Can a thermoregulator alone fix ratio drift?
Only the part of the drift that temperature caused. A thermoregulator cannot correct a worn pump, a miscalibrated flow meter, a drifting sensor or a recipe entered at the wrong density. The way to know which cause you have is the paired record: shot weight and pump inlet temperature taken at the same moment. When the weight tracks the temperature, thermal control is the fix. When it does not, the mechanical side needs attention first.
If you’re interested, check out these related articles:
Best Equipment for Foam Filling Mining Tires
What Is a Polyurethane DIP Machine for Safety Shoes