An internal gear pump flushing with thin solvent may show low or unstable flow, rising housing temperature, noise, vibration, or a falling suction-pressure indication. Start with absolute suction pressure and pump housing temperature. At the stated 0.4 cP acetone viscosity, internal leakage is likely to rise, but slip alone does not decide whether the solvent vaporizes; local absolute pressure versus solvent vapor pressure does.
Read the symptoms before changing the pump
The installation reports an internal gear pump operating at about 3.5 bar differential pressure. That differential does not reveal whether acetone at 110 °C, or toluene near 112 °C, remains liquid at the inlet. You need absolute suction pressure, solvent temperature at the pump, and the applicable vapor-pressure data.
| Observed symptom | Probable mechanism | First check |
|---|---|---|
| Flow falls as solvent enters | Internal slip increases when viscosity drops from polymer service to thin-solvent service | Compare measured flow with the manufacturer’s low-viscosity performance data at the actual speed and differential pressure |
| Housing temperature rises | Internal recirculation, mechanical friction, inadequate liquid cooling, or loss of lubricating film | Trend housing and liquid temperatures; obtain the manufacturer’s alarm and trip limits |
| Crackling, gravel-like noise, vibration, or erratic discharge | Solvent flashes in low-pressure cavities and bubbles subsequently collapse or pass downstream | Measure absolute suction pressure and solvent temperature at the pump inlet |
| Suction pressure is stable but discharge flow remains low | Excessive clearance leakage, bypass flow, wear, or speed too low for the required delivery | Check bypass position, speed, differential pressure, and pump condition |
| Pressure rises rapidly with little flow | Polymer has gelled, cured, or formed a blocked flow path | Stop the pump and establish whether an open path exists before adding more solvent |
| Solvent exits without removing residue | Wrong solvent for the polymer, cured material, poor contact, or channeling through the deposit | Confirm resin-solvent compatibility and inspect the returned flush |
Do not treat 3.5 bar differential pressure as 3.5 bar inlet pressure. Record suction pressure on an absolute scale; a gauge reading alone can be misread when evaluating boiling.
Separate slip from flashing
An internal gear pump carries liquid in spaces between the gear teeth and casing. Clearance paths allow some liquid to leak from the discharge side back toward the suction side. Lower viscosity reduces resistance through those paths, so switching from 18,000 cP polymer to nominally 0.4 cP acetone can reduce volumetric efficiency sharply.
Calculate measured slip only when you know the theoretical displacement:
Q_slip = Q_theoretical - Q_actual
For a fixed-displacement pump, Q_theoretical comes from displacement per revolution multiplied by measured shaft speed. Use the manufacturer’s displacement value and apply its stated corrections. A dependable prediction also needs internal clearances, wear condition, speed, differential pressure, temperature, and fluid viscosity. Viscosity and 3.5 bar differential pressure alone are not enough.
A useful heat estimate, with all internal leakage assumed to dissipate its pressure energy as heat, is:
P_slip ≈ Δp × Q_slip
Use pascals for Δp and cubic metres per second for Q_slip to obtain watts. This estimate excludes bearing, seal, and gear friction, and it does not predict housing temperature. The steady temperature also depends on solvent flow, casing heat transfer, and the duration of the flush.
Flashing follows a different rule. Liquid starts to boil locally when its absolute pressure falls to or below its vapor pressure at the local temperature. Acetone boils at around 55 °C at atmospheric pressure. At 110 °C, it requires elevated absolute pressure to remain liquid; read the required pressure from approved vapor-pressure data for the actual solvent composition.
Define every operating state
The process spans operating points that one pump selection must cover. The described batch can begin around 30 cP, run near 120 °C, and finish at about 20,000 cP for Polymer A or 60,000 cP for Polymer B. The flush then introduces solvent described as 0.4 cP acetone or 0.2 cP toluene. Treat those solvent viscosities as stated operating assumptions until the property data are checked at the actual temperature.
Place the heat exchanger by evaluating both process duty and pump inlet conditions. Cooling before the pump raises polymer viscosity and can increase suction-line loss. Cooling after the pump lets the pump receive hotter, thinner polymer, which can increase slip and challenge its temperature limits. Check both cases against required inlet pressure, pump torque, differential pressure, and the manufacturer’s operating envelope.
Do not assume one exchanger pass changes the batch directly from hot low-viscosity liquid to its final viscosity. Final viscosity can depend on reaction progress, temperature history, shear, concentration, exchanger duty, and circulation rate. Sample or measure viscosity through the batch and use the process endpoint, not an assumed number of passes.
Select the flush before starting it
Choose acetone or toluene from polymer solubility, cure state, temperature, material compatibility, and site flammable-liquid controls. Toluene is not automatically better than acetone. Test the actual uncured resin-solvent pair or obtain compatibility data from the polymer supplier.
Acetone flushing is used with some uncured thermoset residues, but acetone does not dissolve most thermoplastics. Once polymer cures inside a pipe, circulating more solvent normally creates a channel around accessible surfaces rather than opening a solid plug. A cured obstruction may require pipe replacement or a separately engineered, vendor-approved thermal or mechanical cleaning method.
Check the pump’s casing, gears, bushings, bearings, shaft seal, gaskets, and elastomers against the selected solvent and temperature. Also ask the pump manufacturer whether the bearing and gear design may run on such low-viscosity liquid. Chemical resistance alone does not prove that the pump retains the required lubricating film.
Run the flush in a controlled sequence
- Complete the batch transfer before residue gels or cures. Record the final polymer temperature, viscosity indication, pump speed, suction pressure, discharge pressure, and motor load.
- Confirm an open flow path to the receiving system. Do not use the PD pump to force against a suspected solid plug; pressure can rise rapidly when discharge is restricted.
- Remove the bulk polymer before the solvent flush where the approved procedure allows it. Compressed air can reduce solvent consumption, but it can spray polymer from the outlet and introduces an ignition and exposure hazard. Use it only under a reviewed procedure with a controlled receiver and appropriate flammable-atmosphere controls.
- Lower the system to the approved solvent-flush temperature. A room-temperature flush may dissolve uncured solid polymer and reduces vapor-pressure risk compared with a
110–112 °Cflush. - Verify that the work area and equipment are electrically rated for the flammable atmosphere. Confirm bonding, grounding, ventilation, containment, and the approved discharge destination through the site procedure.
- Line up the solvent supply with enough positive absolute pressure at the pump inlet. Include vessel pressure, static head, inlet-line loss, filter loss, and local temperature when checking the inlet margin over vapor pressure.
- Start at the manufacturer-approved speed and differential-pressure range for low-viscosity service. Never infer a safe speed from the polymer operating point.
- Trend suction pressure, discharge pressure, differential pressure, delivered flow, housing temperature, solvent temperature, vibration, noise, and motor load. Stop on loss of stable flow, rapid temperature rise, abnormal noise, or a pressure increase indicating blockage.
- Continue only while the return stream shows progressive removal of uncured polymer. Follow with the approved draining, recovery, or drying step before introducing the next batch.
Verify liquid delivery and thermal control
Prove the flush with measurements, not discharge pressure alone. A PD pump can build pressure while much of its displacement leaks internally. Record actual solvent flow or collected volume over a measured interval, then compare it with theoretical displacement at the measured shaft speed.
Mount one or two temperature detectors on the housing where the pump manufacturer specifies. Obtain its alarm and shutdown setpoints; the source information gives no safe universal temperature. If temperature keeps climbing at stable inlet conditions, stop and check flow, clearances, lubrication requirements, bypass position, and mechanical condition.
Confirm completion by inspecting or testing the return solvent under the plant’s approved method. A stable flow, stable housing temperature, clean return, open flow path, and repeatable pressure establish that the solvent reached the wetted system. A brief pressure pulse does not.
If bubbles appear in the return, determine whether they originate from inlet flashing, dissolved gas release, air ingress, or intentional gas displacement. Correlate bubbles with absolute suction pressure, liquid temperature, noise, and vibration. Raising discharge pressure does not repair inadequate inlet pressure.
Avoid the fixes that waste time
- Do not rely on residual polymer to thicken the solvent. Mixture viscosity changes through the flush and cannot provide a controlled low-viscosity operating point.
- Do not keep circulating solvent through a cured plug. Stop and establish a flow channel or replace the blocked section.
- Do not calculate vaporization from differential pressure. Use local absolute pressure and vapor pressure at actual temperature.
- Do not assume slip converts the entire solvent charge to vapor. Quantify flow loss and heat input separately, then measure housing temperature.
- Do not select acetone or toluene by viscosity alone. Polymer solubility, cure state, pump materials, seal compatibility, and flammability govern the choice.
- Do not copy the polymer speed, temperature, or pressure settings into solvent service. The pump may cross from a well-lubricated high-viscosity duty into a clearance-leakage and bearing-lubrication limit.
- Do not wait until the next batch to discover residue. Define an objective return-solvent acceptance check and inspect the flow path before restoring polymer service.
FAQ
How do I calculate internal gear pump slip with acetone?
Use Q_slip = Q_theoretical - Q_actual. Obtain displacement per revolution from the pump manufacturer, multiply by measured speed, and measure actual acetone flow at the real viscosity, temperature, and 3.5 bar differential pressure.
How do I tell whether acetone will boil in the pump?
Measure absolute suction pressure and acetone temperature at the inlet, then compare that pressure with approved vapor-pressure data. Acetone boils around 55 °C at atmospheric pressure, so operation at 110 °C requires elevated absolute pressure to keep it liquid.
How do I prevent an internal gear pump from overheating?
Maintain verified solvent flow, operate inside the manufacturer’s low-viscosity envelope, and trend one or two housing temperature detectors. Get the alarm and trip setpoints from the pump manufacturer rather than inventing a universal limit.
How do I choose between acetone and toluene for flushing?
Match the solvent to the actual polymer and cure state, then verify pump, seal, gasket, and elastomer compatibility at flush temperature. The stated 0.4 cP acetone and 0.2 cP toluene viscosities do not determine cleaning performance.
How do I know when to stop and call pump support?
Stop if flow becomes unstable, housing temperature rises rapidly, abnormal noise or vibration starts, or a blocked line drives pressure upward. Do not restart until the flow path and inlet pressure are verified. Escalate to official pump-manufacturer support when low-viscosity limits, material compatibility, allowable speed, or temperature setpoints are missing.