A hydraulic system has two end-of-line components — the pump that pressurises fluid and the cylinder that converts pressure to force — and only the pump side is sensitive to NPSH (net positive suction head) margin [S1]. The cylinder simply reacts to the pressure the pump delivers; it has no inlet, no suction line, and no cavitation risk of its own, which is why specifying NPSH is a pump problem before it is a cylinder problem.
The split of responsibility is mechanical: the pump is hydrostatic or hydrodynamic, moves a fixed or variable displacement per revolution, and demands a minimum inlet pressure to avoid vapor bubble collapse at its eye [S1]. The cylinder converts that delivered flow into linear force on a piston rod, with bore and rod diameters setting the area ratio and therefore the extension-to-retract force balance. Conflating the two leads to a system that cavitates, overheats, and erodes — long before the cylinder has any reason to fail.
Why NPSH Margin Belongs to the Pump, Not the Cylinder
Net positive suction head available (NPSHa) is calculated as atmospheric head plus fluid head on the reservoir side minus vapour pressure minus line losses, and net positive suction head required (NPSHr) is a curve published by the pump manufacturer for each speed and displacement point [S1]. Cylinders have no equivalent curve because they are downstream of the pump and operate at discharge pressure, not at suction pressure. A 350 bar cylinder does not care whether the pump inlet is at 0.3 bar absolute or 0.8 bar absolute, but the pump very much does.
Pump types commonly used in industrial hydraulic power units fall into gear, vane, and piston (axial / radial) families, each with its own NPSHr envelope [S2]. Piston pumps generally tolerate the lowest inlet pressure because their volumetric clearances are tighter and their internal leakage paths smaller; gear pumps are the least forgiving and are the usual culprit when an installation cavitates. Cylinder selection — single-acting, double-acting, telescopic, mill-type — does not move this requirement in either direction.
Pump Selection Criteria That Drive the NPSH Decision
Four pump-side parameters decide whether an installation has enough NPSH margin, and each one is set before the cylinder is even specified. First, pump speed: doubling shaft speed roughly doubles NPSHr on a fixed-displacement piston pump, so a 1450 rpm electric-drive unit gives a fundamentally different suction envelope than a 3000 rpm engine-driven unit [S6]. Second, fluid viscosity: ISO VG 46 at 40 °C is the reference; cold-start oil at VG 100 inflates NPSHr by 30–60 % and is the most common cause of marginal systems failing on the first cold morning. Third, inlet line length, diameter, and number of fittings — every 90° elbow and every metre of undersized hose adds friction head and eats margin. Fourth, reservoir condition: a pressurised tank adds head; a return line above the fluid level subtracts it.
The decision tree runs: calculate NPSHa at worst case (coldest expected oil, highest duty cycle, filter partially blocked), subtract the manufacturer's NPSHr at the operating point, and require at least 0.5 bar of margin for a piston pump and 1.0 bar for a gear pump. If the math does not close, the fix is always on the pump inlet side — larger line, lower pump speed, charge pump on a closed-loop piston unit, or flooded suction from a tank above the pump. Oversizing the cylinder will not save a pump that is starving.
Cylinder Selection Criteria That Do Not Move the NPSH Math
Cylinder specification is a force, speed, and stroke problem. Bore area sets force at a given pressure, rod diameter sets the retract-side area, and stroke sets the machine envelope. Operating pressure (typically 160–250 bar for industrial presses, 350 bar for mobile and marine, 700 bar for clamping tools) is set by the pump, not negotiated by the cylinder. Cylinder-side standards such as classification-society type approval (DNV, ABS, LR, BV, CCS) on mill-type cylinders cover material traceability, welding, and pressure testing, not inlet suction [S7].
Where the cylinder does influence the system in a way that can feed back into pump loading — and therefore into NPSH stability — is duty cycle. A cylinder that demands full flow for 8 seconds out of every 12 sets a much higher average pump demand than one that strokes in 2 seconds and dwells for 30, and the higher average demand forces the pump closer to its displacement limit. If that limit is reached on a pressure-compensated piston pump, the pump de-strokes, inlet velocity drops, and NPSHr typically improves — but on a fixed-displacement gear pump, the regulator simply dumps the surplus over the relief, with no NPSH benefit. So cylinder duty influences pump selection, but never NPSH directly.
Side-by-Side: Pump-Side vs Cylinder-Side Decisions
Lining the two sides up against four decision criteria clarifies where engineering effort belongs. On the criterion of "who can cavitate," the pump is the only candidate — the cylinder cannot. On "what sets system pressure," the cylinder sets the force requirement and the pump delivers the matching pressure, so the answer is shared but the NPSH half of the calculation is pump-only. On "what moves when fluid is cold," the pump dominates because cold oil raises NPSHr sharply while the cylinder is largely unaffected. On "what does classification-society approval certify on the cylinder," it certifies structural and pressure-test compliance for marine and offshore service, not any inlet condition [S7].
The same comparison exposes the most common engineering mistake: sizing a cylinder to "give the pump an easier life." Cylinders that are too large for the job cause low-pressure cavitation on the return side of the pump (because the pump chases a low back-pressure loop and runs near maximum displacement), while cylinders that are too small force the pump into the relief at every stroke. Neither problem is solved by re-choosing the cylinder; both are solved by re-choosing the pump or its inlet plumbing.
Real-World Failure Modes From Misattributing NPSH to the Cylinder
Three recurring field failures trace directly to NPSH being treated as a system-wide number rather than a pump-only number. First, cold-start cavitation damage on gear pumps in mobile equipment, where the installer assumed the cylinder's duty was the limiting factor and did not account for ISO VG 100 oil at −10 °C inflating NPSHr past NPSHa. Second, suction-line erosion on piston pumps in marine power packs, where the long run from a below-deck reservoir to an above-deck pump added friction losses that were never added to the NPSH calculation, and classification-society cylinder approval papers (DNV, ABS) gave a false sense that the system as a whole had been independently reviewed [S7]. Third, charge-pump starvation on closed-loop piston units, where the cylinder stroke length and frequency demanded more charge flow than the auxiliary pump could supply through its own small inlet.
In each case the field fix was on the pump side: switch to a flooded-suction reservoir, add a larger suction line, install a charge pump with a positive-displacement boost, or drop the drive rpm. Replacing the cylinder — different bore, different stroke, different manufacturer — never resolved the failure, because the cylinder was never the constraint.
Limits of the Pump-vs-Cylinder Split
The clean separation breaks down in one place: the return line. On a double-acting cylinder, oil leaving the rod end returns to the tank at near-system pressure, and if that return line is plumbed back into the pump inlet without adequate conditioning (cooling, filtering, de-aeration), it can raise the local vapour pressure of the oil and effectively shrink NPSHa. This is a hydraulic-system-level interaction, not a cylinder problem, but it is the only scenario in which cylinder-side plumbing touches the NPSH balance. The fix is still pump-side: a properly sized return-line filter, a baffle in the reservoir, and a sub-atmospheric return path that does not feed bubbles directly into the suction line. [S1]
Outside of that return-line edge case, the rule holds: a hydraulic pressure transmitter on the pump discharge and a flow meter on the pump outlet can together confirm that the pump is delivering the pressure and flow the cylinder demands, and that the pump itself is not the failing component. Likewise, a properly specified industrial valve in the suction line — a y-strainer with a known pressure drop, a full-bore ball isolation valve — protects the pump inlet from the kind of restriction that eats NPSH margin silently over months.
Sourcing and Standards for NPSH and Cylinder Design
NPSH calculation is a manufacturer-curve exercise, not a standards exercise, and the authoritative numbers come from the pump maker's NPSHr curve for the specific model, speed, and displacement [S1]. General hydraulic-pump type and operating principle references (hydrostatic vs hydrodynamic, fixed vs variable displacement) are documented openly by industry suppliers and integrators [S1][S2]. Cylinder-side material, welding, and pressure-test acceptance is governed by classification-society rules on a per-vessel basis (DNV, ABS, LR, BV, CCS, GL, KR, NK, RINA type approval on mill-type cylinders), and these rules do not certify any pump inlet condition [S7].
The practical sourcing checklist for any hydraulic system that has ever shown cavitation symptoms: (1) request the actual NPSHr curve from the pump maker for the installed speed and displacement, not a generic catalogue line; (2) measure cold-start oil temperature and re-derive NPSHa for the worst case; (3) re-derive suction-line pressure drop with the actual fittings and the actual fluid, not the clean-water numbers from the fitting catalogue; (4) confirm the reservoir condition (vented or pressurised, fluid level relative to pump inlet) under the installed attitude. Each of these four steps lives entirely on the pump side of the system diagram.
Track for the next quarter: whether the two major Chinese hydraulic-component platforms (Haite for cylinders, Poocca for pumps) publish updated NPSH curves and cylinder type-approval certificates for 2026 builds, and whether the 2025–2026 engine-driven hydraulic pump category sees new low-speed, high-NPSH-margin piston units aimed at mobile cold-climate service [S2][S6][S7]. A change in either direction will tell the market whether pump makers are responding to the cold-start cavitation problem or whether integrators are still being asked to fix it on the cylinder side.