Mechanical
The Gerotor Pump: How Nested Rotors Meter Oil One Chamber at a Time
Idle your car engine at 700 rpm and the little gerotor bolted to the crankshaft nose is already flooding the bearings with roughly 5–10 L/min of oil at 300–500 kPa — a pump the size of a hockey puck, with exactly two moving parts, no valves, and no rubbing gear teeth in the conventional sense. Its secret is a mathematical curiosity: an inner rotor with N lobes spins inside an outer ring with N+1 lobes, and the epitrochoidal geometry forces every tooth of the inner to stay in contact with the outer at all times, sweeping a chain of sealed crescent chambers from inlet to outlet.
The name is a contraction of generated rotor, and the “generation” is literal — the outer lobe profile is the mathematical envelope traced by the inner rotor as it orbits. Get the conjugate geometry right and you have a compact, quiet, self-priming positive-displacement pump that dominates automotive lube systems, transmission charge pumps, and low-pressure fuel delivery. Get the tip clearances or the inlet velocity wrong and it cavitates, aerates, and eats itself.
- TypePositive-displacement internal gear (trochoidal)
- Rotor setInner N lobes, outer N+1 lobes
- FlowQ = V_d · n · η_v (typ. 1–60 L/min)
- Pressure0.3–2 MPa typical; up to ~20 MPa gerotor variants
- η_v0.85–0.95 at rated speed, warm oil
- Used inEngine lube, ATF charge, fuel & scavenge pumps
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
How it works: N inside N+1
A gerotor is an internal gear set with a one-tooth difference. The inner rotor has N lobes (teeth); the outer rotor (the ring) has N+1 lobes. Their centres are offset by a small distance e, the eccentricity. The inner rotor is driven by the shaft; because the outer has one more lobe, the two turn at different rates — the outer runs at N/(N+1) of the inner rotor speed. As the inner rotor rotates, each of its lobes stays in sliding contact with the outer profile, so at any instant the pair form a ring of N sealed chambers between them.
The single-tooth difference is what makes it a pump. As the inner rotor advances, the volume of each chamber grows on the inlet side (expanding, drawing oil in) and shrinks on the outlet side (collapsing, pushing oil out). There are no check valves: porting plates on the front and rear faces expose the expanding chambers to the low-pressure inlet kidney port and the shrinking chambers to the high-pressure outlet kidney port. One full inner-rotor revolution displaces N chambers, each cycling once — the whole rotor set breathes like a ring of tiny cylinders.
- Displacement per revolution: V_d = N · (V_max − V_min) per chamber, where V_max and V_min are the maximum and minimum chamber volumes over one cycle.
- Ideal flow: Q_ideal = V_d · n, with n the shaft speed. A common 4/5-lobe automotive rotor with V_d ≈ 10 cm³/rev at 3000 rpm gives Q_ideal ≈ 30 L/min.
- Actual flow: Q = V_d · n · η_v, where η_v is volumetric efficiency (0.85–0.95 warm, higher cold).
The geometry: epitrochoids and their conjugate envelope
The elegance of the gerotor is that the two profiles are mathematically conjugate — one is generated from the other. In the classic construction the inner rotor is an epitrochoid (or an epitrochoid whose points are replaced by circular arcs) and the outer rotor lobes are circular arcs of radius r positioned on a pitch circle. The outer profile is precisely the envelope swept by the inner rotor as it performs its planetary orbit: inner rotor spinning about its own centre while that centre orbits the outer centre at radius e.
Parametrically, an epitrochoid point on the inner rotor traces x(θ) = (R + r)cos θ − d·cos((R+r)/r · θ), y(θ) = (R + r)sin θ − d·sin((R+r)/r · θ), where R is the fixed (base) circle radius, r the rolling circle radius, and d the tracing-point offset. The design levers are the lobe count N, the eccentricity e, the outer lobe radius r_c, and the rotor thickness (axial length) L.
- Eccentricity e sets chamber travel: bigger e means larger V_max − V_min per chamber, hence more displacement, but larger sliding velocities and higher contact stress at the lobe tips.
- Lobe count N trades displacement-per-lobe against flow smoothness. Fewer lobes (2/3, 3/4) give big chambers and high displacement but chunkier flow ripple; more lobes (7/8, 9/10) give many small chambers, low ripple, quieter running, but tighter tolerances.
- Contact geometry: the theoretical profile has continuous line contact, but a small tip clearance (typ. 20–60 µm) is machined in so the parts don't seize — this clearance is the dominant leakage path.
Flow ripple, kinematics, and why it's quiet
Because several chambers deliver simultaneously and their volume rates overlap, the instantaneous outlet flow is nearly steady. The flow ripple — the peak-to-mean variation in delivery — falls sharply with lobe count. A 4/5 set has an irregularity on the order of 1–3%; a 9/10 set can be well under 1%. The fundamental pumping frequency is f = N · n / 60 (Hz) for an inner rotor of N lobes at n rpm, plus its harmonics. A 4-lobe rotor at 3000 rpm pumps at 200 Hz — right in the audible band, which is why lobe-passing tones and porting geometry matter for NVH.
Two kinematic facts drive the acoustics and durability:
- Trapped-volume compression: as a chamber crosses from inlet to outlet porting, a small pocket of oil can be momentarily sealed off. If it's connected to neither port at minimum volume, pressure spikes; the porting timing grooves (relief notches at the kidney-port ends) are tuned to bleed this pocket smoothly and kill the pressure pulse that would otherwise ring the housing.
- Sliding at the tips: the relative velocity between inner tip and outer flank is highest near the sealing contact. This is a boundary-to-mixed lubrication regime; the oil film (a few µm) separates the surfaces, and the tribology here — plus the porting design — sets both efficiency and wear life. See tribology and journal-bearing lubrication for the underlying film physics.
Sizing a gerotor: displacement, power, and torque
Design starts from the required flow at a design speed. Rearranging Q = V_d · n · η_v gives the target displacement V_d = Q / (n · η_v). Say you need Q = 12 L/min at engine idle-plus, n = 2000 rpm, η_v ≈ 0.9: V_d = 12000 cm³/min ÷ (2000 rev/min × 0.9) ≈ 6.7 cm³/rev. With that displacement fixed you choose N, then solve the geometry (e, R, r_c, L) so that N·(V_max − V_min) hits 6.7 cm³.
The hydraulic and shaft numbers follow directly:
- Hydraulic power: P_hyd = Δp · Q. At Δp = 500 kPa and Q = 12 L/min = 2.0×10⁻⁴ m³/s, P_hyd = 5×10⁵ × 2.0×10⁻⁴ ≈ 100 W.
- Shaft power: P_shaft = P_hyd / η_o, with overall efficiency η_o = η_v · η_m ≈ 0.75–0.85; here ≈ 120–130 W.
- Drive torque: T = P_shaft / ω. At 2000 rpm, ω = 2π·2000/60 ≈ 209 rad/s, so T ≈ 130/209 ≈ 0.6 N·m. Equivalently T ≈ (Δp · V_d)/(2π · η_m).
Key scale check: displacement scales with rotor thickness L linearly and with e and R roughly as e·R·L, so a modest bump in eccentricity or length buys flow cheaply — until tip stress, sliding speed, and inlet velocity push you into cavitation.
Real hardware: where gerotors live
The gerotor is the default engine oil pump in the overwhelming majority of modern passenger cars and motorcycles — often crank-nose-driven, so it runs at engine speed. Because lube demand doesn't scale with engine speed, high-rpm oversupply is dumped over a spring-loaded relief valve (a parasitic loss), which is why variable-displacement gerotor pumps — where a control ring shifts eccentricity to modulate V_d — have become common for fuel-economy gains, actively regulating gallery pressure to ~150–350 kPa instead of dumping.
- Automatic transmissions: a gerotor charge pump pressurizes the ATF circuit (clutch apply, torque-converter charge, cooling) at 0.5–2 MPa.
- Fuel systems: low-pressure lift/transfer pumps and diesel priming pumps use gerotors; high-pressure common-rail generation is left to other topologies.
- Dry-sump & aerospace: multi-stage gerotor scavenge pumps pull oil-air froth from the sump; here the pump must tolerate high entrained-air fractions.
- Materials: rotors are commonly sintered powder-metallurgy steel (net-shape, cheap, self-lubricating porosity) or hardened tool steel for high pressure; housings in aluminum or cast iron. Powder-metallurgy net-shape forming is what makes the complex trochoidal profile economical at scale.
Failure modes, cavitation, and design limits
The gerotor's Achilles' heel is the inlet. As a chamber expands it must fill in the time available; if oil can't rush in fast enough, local pressure drops below the vapour/gas-release pressure and the chamber only partially fills. The metrics that govern this:
- Cavitation / incomplete filling: at high speed and cold, high-viscosity oil, filling lags. Symptoms are a flow curve that falls off above a threshold rpm, noise, and vapour-bubble collapse that erodes rotor tips and porting. Fixes: larger inlet kidney port, lower inlet velocity (<~5–8 m/s), adequate NPSH, and limiting shaft speed. See cavitation and NPSH.
- Aeration: entrained air (foamy scavenge oil, low sump level) makes the fluid compressible, collapsing η_v and delivery pressure — the pump can't build head on a spongy, gas-rich charge.
- Wear & leakage: tip and face clearances open with wear; because internal leakage Q_leak ∝ Δp·h³/µ (a Poiseuille gap flow, cubic in clearance h), a clearance that doubles from 25 to 50 µm can raise leakage ~8×, tanking η_v — most acute hot and at low speed.
- Cold-start torque: at −20 °C, oil viscosity can be 100–1000× warm value; drive torque and housing stress spike, and the relief valve may pin open. Rotors and drives are sized for this transient.
Best practice: keep tip clearance in the 20–60 µm band, generous inlet porting, add porting relief grooves to control trapped-volume pressure spikes, choose N≥5 where quiet operation matters, and use a pressure-relief or variable-displacement scheme rather than deadheading the outlet. Well-designed automotive gerotors routinely exceed 300,000 km of service.
| Attribute | Gerotor (trochoidal) | External gear pump | Crescent internal gear |
|---|---|---|---|
| Moving parts | 2 (inner + outer rotor) | 2 meshing gears | 3 (gear, ring, crescent) |
| Contact points / chamber seal | Continuous multi-tooth line contact | Single mesh line | Gear-ring mesh + crescent seal |
| Typical pressure | 0.3–2 MPa (spec. to 20 MPa) | up to 25–30 MPa | up to ~17 MPa |
| Speed range | high (to 6000+ rpm) | high (to 3500 rpm) | moderate |
| Flow ripple / noise | Low ripple, quiet | Higher ripple, whine | Low ripple |
| Cost / compactness | Very low, very compact | Low | Moderate, longer axially |
Frequently asked questions
Why use a gerotor instead of an external gear pump?
Gerotors have only two moving parts, are extremely compact and cheap (net-shape sintered rotors), run quietly with low flow ripple, and self-prime well — ideal for engine lube at 0.3–2 MPa. External gear pumps handle higher pressures (25–30 MPa) but whine more and have a single, harder-loaded mesh line. When you need low-to-medium pressure in a tiny package, the gerotor wins.
How do you size a gerotor pump?
Start from required flow: V_d = Q/(n·η_v). For 12 L/min at 2000 rpm and η_v ≈ 0.9, V_d ≈ 6.7 cm³/rev. Then pick lobe count N and solve the trochoidal geometry (eccentricity e, radii, thickness L) so N·(V_max − V_min) equals V_d. Check drive torque T ≈ Δp·V_d/(2π·η_m) and verify inlet velocity stays low enough to avoid cavitation.
Why does the outer rotor have one more lobe than the inner?
The single-tooth difference is the entire pumping principle. It ensures each inner lobe stays in contact with the outer profile through the whole rotation, forming a ring of N sealed chambers whose volumes grow (intake) on one side and shrink (discharge) on the other. Equal lobe counts would just be a coupling, not a pump.
What limits gerotor pump speed and pressure?
Speed is limited by inlet filling — above a threshold rpm the expanding chambers can't fill fast enough and the pump cavitates, so flow plateaus and tips erode. Pressure is limited by internal leakage (which scales as the clearance cubed) and by tip contact stress. Standard automotive gerotors run 0.3–2 MPa; hardened high-pressure variants reach ~20 MPa.
What is volumetric efficiency and why does it drop?
η_v is actual flow divided by ideal displacement flow, typically 0.85–0.95 warm. It falls because oil leaks back across the tip and face clearances under pressure; leakage rises with Δp and with clearance cubed, and worsens as viscosity drops with heat. So η_v is lowest hot, at low speed, and high pressure — and gets worse as the pump wears.
How is a variable-displacement gerotor different?
A conventional gerotor's displacement is fixed, so at high engine rpm it over-pumps and dumps excess oil over a relief valve — wasted power. A variable-displacement gerotor mounts the outer rotor in a movable control ring; shifting the ring changes the eccentricity e (or effective port timing), reducing V_d to match demand and holding gallery pressure at a target, saving several hundred watts and improving fuel economy.