Greased bearings for a geared turbofan gearbox
A geared turbofan uses a reduction gearbox to connect the core engine’s high-speed, high-temperature rotating system to a slower, high-torque output. The gearbox concentrates loads in compact rolling-element arrangements where speed, heat, lubrication and manufacturing variation interact. Greased bearings could replace oil-fed supports in selected, cooler locations, but only where a documented thermal and speed case supports the choice.
The engineering case is not simply “grease instead of oil”. It is a system decision: which shaft can be supported by grease, how the lubricant reaches the contacts, how preload survives assembly tolerances, and what evidence is needed before an Australian operator can rely on it. Simulation, tolerance analysis and physical testing must answer those questions together, with bearing fatigue, gearbox efficiency and drivetrain performance treated as linked requirements.
Why a grease option is worth testing
Grease is attractive because it can remain where it is placed, reducing the need for scavenge flow, drain lines and oil-level checks at isolated supports. In an enclosed housing it can simplify the lubricant path and retain a protective film during stops, when an oil-fed bearing may otherwise drain down. These benefits are credible at selected low-speed or intermediate duty locations. They do not extend automatically to the high-speed core or to gear meshes that require a controlled fluid supply.
The hardest case is a heavily loaded, hyperstatic support, where assembly constraints maintain high contact preload. Preload closes the lubricant film, so boundary contact, friction and heat generation become more important; a grease that looks suitable in a bucket test can still fail when retained at the raceway. Conversely, grease can remain effective where low speed would starve conventional splash lubrication. The useful question is where grease creates a stable operating window, not whether it can outperform oil everywhere.
The OPTIMIZE Project’s wider work on power-reduction gearboxes provides a sensible setting for this comparison. Its objectives and methods combine design-of-experiments, simulation, tolerance analysis and physical testing rather than treating a bearing change as an isolated calculation. The project documentation is the appropriate place to examine the supporting research context before selecting a candidate architecture.
A practical case must acknowledge what grease cannot remove. It still obeys speed, temperature and load limits; excessive fill can cause churning and heating, while centrifugal redistribution can deplete one raceway and load another. Seals reduce contamination but may add drag, and grease remains combustible. The opportunity is targeted simplification—fewer local fluid interfaces, retained lubrication during stops and potentially lower local mass—not a blanket claim of higher efficiency, longer life or greater power density.
From grease chemistry to bearing selection
Lubricant choice starts with the operating envelope: temperature, speed, radial and axial load, shock loads, duty cycle and expected contamination. A synthetic complex grease may suit high-speed rolling contacts, while a higher-viscosity or extreme-pressure grade may be needed for heavy loads; neither label guarantees compatibility with seals, adjacent lubricants or the gearbox’s other materials. The selected grade should be supported by viscosity-temperature data, shear stability, grease life testing and compatibility checks, including the effect of any oil carry-over.
Bearing geometry then follows the load path. Cylindrical roller bearings can carry high radial loads with low relative speed, deep-groove ball bearings suit combined loads at moderate speeds, and spherical rollers tolerate misalignment at the cost of higher friction and sensitivity to lubrication. Selection should use the recognised dynamic-rating framework with application-specific preload, contamination, temperature and grease-lubrication factors. Catalogue oil ratings should not be transferred unchanged to a grease-fed design.
Preliminary calculations should also separate bearing failure from drivetrain failure. A shaft may pass bearing fatigue checks yet produce unacceptable torque drag, or cause gear excitation that invalidates the lubrication model. Useful outputs include calculated life, friction power loss, thermal margin, required fill quantity and a sensitivity ranking for every uncertain input. These assumptions should be cross-checked against an external bearing engineering reference and against the actual gearbox interface.
Grease delivery needs its own architecture. A sealed, grease-filled cavity depends on initial fill and redistribution; a reservoir or externally fed arrangement adds pumping and return paths; and a grease slinger can improve circulation at higher speeds while introducing another moving part. Each layout also needs a pressure-control strategy so trapped air and thermal expansion do not load the seals or displace lubricant from the rolling contacts. The simplest arrangement is attractive only if rig evidence shows stable temperature, film thickness and retention; otherwise, a local oil feed may remain preferable.
Predicting mixed lubrication and contact behaviour
A useful simulation links lubricant-film formation, friction and heat rather than assigning a fixed grease temperature. Elastohydrodynamic and mixed-lubrication models can estimate film thickness and asperity contact at selected points, while a thermal model traces heat from those contacts into the housing and adjacent structure. The shaft must also be allowed to deflect under gear forces and bearing preload. Otherwise, a model may show adequate film thickness at nominal geometry while predicting starvation at an assembly extreme.
Surface condition is part of that model, not a finishing detail. Although the work addresses gear teeth rather than bearing raceways, the project’s gear flank surface roughness study explains why roughness peaks, lambda ratios and local contact conditions can control micropitting in mixed lubrication. Bearing and gear materials should therefore be characterised with suitable metrology, and the fatigue assessment should distinguish rolling-contact damage from the wear mechanisms observed at the gear flanks.
Tolerance analysis then tests whether the design still works after real components are made. Housing bore roundness, fit, spacer thickness, raceway parallelism and shaft deflection can redistribute load among rollers, alter contact stress and change grease movement. A Monte Carlo or sensitivity study can reveal whether nominal clearance is actually robust, while a worst-case build can expose an overconstrained arrangement in which friction, temperature and fatigue rise together. The result should be an adjustable preload strategy, not a nominal number carried forward without evidence.
The main architecture choices have different consequences:
| Architecture | Lubrication and heat rejection | Principal advantages | Main constraints |
|---|---|---|---|
| Recirculating oil | Oil is pumped through the bearing and returned to a reservoir. | Strong heat transport, controlled flow and established monitoring methods. | Pumps, lines, seals and scavenge losses add mass and maintenance. |
| Grease-filled cavity | Grease is retained around the rolling elements. | Fewer local fluid paths and less drain-down during stops. | Heat removal and film replenishment depend on geometry, fill and speed. |
| Hybrid arrangement | Grease supports a selected bearing while another element remains oil-fed. | Allows a focused trial without redesigning the entire lubrication system. | More interfaces to justify, qualify and maintain. |
Building a test and tolerance programme
Design-of-experiments can separate the variables that matter from those that merely consume test time. Candidate factors might include base-oil viscosity, grease consistency, fill quantity, internal clearance, preload, raceway finish, temperature and contamination level. Responses should include friction torque, bulk and raceway temperature, vibration, lubricant condition, wear and fatigue after operation. A fractional factorial screen can identify interactions before a smaller set of confirmation tests establishes acceptance limits.
The same methods discipline appears in the project’s design of experiments methods for spline-fit selection and fretting fatigue. Although spline fretting is not bearing rolling fatigue, its treatment of fit, repeated micro-displacement and test repeatability is directly relevant to a hyperstatic drivetrain. The bearing programme should borrow that logic: randomised test order, repeated specimens, declared exclusion rules and measurement uncertainty stated before results are compared.
Physical testing should progress from grease rigs to a representative gearbox. Bench rigs can isolate bearing temperature, preload and lubricant condition; a gearbox rig can then measure torque, gear-mesh interaction, chattering and heat rejection under service-like schedules. Start-stop cycles and vibration deserve particular attention because grease distribution can change between stationary and high-speed states. Post-test teardown, lubricant analysis and metrology should be planned as part of the test, not added after unexpected damage appears.
Manufactured test hardware should capture the tolerances seen in service, including housing fit, internal clearance and surface finish. Each specimen needs a traceable grease batch, fill record and assembly history. Torque and temperature instrumentation should be calibrated, and the measurement plan should define how distress, lubricant degradation and abnormal vibration will be detected. Endurance testing can then separate normal wear from an early thermal or starvation failure, providing evidence for inspection intervals and the durability case.
A practical route to Australian qualification
For future installations or engine modifications, Australian operations provide a demanding reference environment. Routes such as Brisbane–Cairns, Melbourne–Launceston and Perth–Karratha expose aircraft to high ambient temperatures, dust and long sectors. A bearing concept intended for this environment must retain lubricant during extended operation, remain inspectable and avoid creating a new heat or contamination problem near the installation.
Regional operators already rely on ageing turbine fleets, and approved maintenance organisations in Melbourne, Brisbane and Perth support them with established inspection routines. Engine oil analysis and filter inspection are familiar tasks; a grease-lubricated support would need equivalent evidence that its condition can be checked without opening the gearbox more often. Access to qualified workshops, controlled assembly areas and specialist balancing equipment is as important as the grease itself.
Any engine or installation change must fit the applicable Australian approval framework. Depending on the project, that means design and production approvals under CAS Part 21, compatibility with the operator’s Part 121 continuing-airworthiness system, and controlled changes to maintenance instructions. Supplier quality systems commonly follow AS9100, while parts and repair data require the relevant CASA pathways. Grease type, fill mass, sealing, inspection limits and reuse or disposal decisions must all appear in traceable approved data.
A sensible programme would move from candidate selection to component rigs, then to a representative gearbox and eventually to an installation-level case. Reviews should track mass, efficiency, thermal margin, life scatter, inspection burden and certification risk, with a clear decision to retain, modify or stop the option at each stage. Explore the OPTIMIZE Project’s objectives and test videos, then contact its research team to scope a joint greased-bearing feasibility study with Australian operating conditions included from the outset.