Bearing preload, thermal growth and internal clearance
In a geared aircraft engine, a bearing rarely operates at the clearance or preload set on the assembly line. Shaft speed, transmitted torque, lubricant temperature, housing distortion and manufacturing variation all alter the relationship between the rolling elements and their raceways. The impact of bearing preload method on thermal growth and internal clearances therefore becomes a central design issue rather than a narrow assembly detail.
This relationship is especially important in power reduction gearboxes, where compact layouts must carry high loads at high speed. A small change in bearing contact force can influence friction, heat generation, fatigue life, gear alignment and power density. The OPTIMIZE Project examines these interactions through simulation, design of experiments, tolerance analysis and physical validation, offering a useful framework for engineers working across the aerospace supply chain.
Why preload changes during operation
Bearing preload is an intentional axial or radial force applied before the gearbox reaches its operating condition. It removes unwanted free play, improves stiffness and can support accurate gear positioning. In angular-contact and tapered roller bearings, preload also controls the position of the rolling elements under combined radial and axial loads. Too little preload can permit skidding, vibration and shaft displacement, while too much creates excessive friction and an avoidable thermal burden.
The assembled value is only the starting point. During operation, the inner ring, outer ring, shaft and housing expand at different rates. A steel shaft carrying a hot gear train may grow axially or radially more than a surrounding housing, increasing contact force. If the housing expands more rapidly, the original preload can be reduced and the bearing may move towards a loose or lightly loaded state. The resulting operating condition is often called the effective clearance or running clearance.
The bearing arrangement determines how strongly these changes are transferred into contact force. A fixed-and-floating arrangement can allow one bearing to accommodate axial growth, whereas a locating pair may transmit thermal expansion directly through spacers and shoulders. In a hyperstatic gearbox, several bearings may constrain the same shaft in ways that make the final load distribution sensitive to small dimensional deviations.
Comparing preload methods in a gearbox
Common preload methods include ground spacers, selective shims, spring elements, controlled endplay adjustment and direct axial displacement during assembly. A rigid spacer arrangement can deliver strong stiffness and repeatability when component dimensions are tightly controlled. Its weakness is that thermal expansion can rapidly alter the internal force. The initial setting must therefore be chosen with the full temperature range and load spectrum in mind.
Spring preload, using a disc spring or another compliant element, gives the system more ability to absorb thermal movement. As the shaft and housing change length, the spring deflects instead of transferring the entire dimensional change into bearing force. This can protect against excessive preload at elevated temperature, although the spring adds its own stiffness, fatigue and packaging considerations. At high speed, designers must also account for centrifugal effects, spring relaxation and changes in lubrication drag.
Selective fitting and shim adjustment can achieve excellent results in low-volume aerospace production, provided inspection and traceability are strong. The method may be less attractive when a gearbox must be produced at a higher rate or serviced in remote locations. A controlled preload system based on measured axial displacement can improve repeatability, but it requires reliable tooling, calibration and a clear relationship between displacement, temperature and bearing load.
The correct choice is consequently a system decision. It depends on bearing type, shaft length, housing material, gear mesh stiffness, oil viscosity, assembly capability and maintenance practice. Engineers should compare methods using the operating load envelope rather than selecting a preload value from a catalogue in isolation.
Linking thermal growth with internal clearance
Internal clearance describes the movement available within a bearing before external loads and thermal effects establish the operating contact condition. Radial clearance is particularly important in cylindrical roller bearings, while axial endplay and contact angle are central to tapered roller and angular-contact arrangements. Preload is often described as negative clearance, but the two concepts should be considered together because both define how the bearing responds to load and temperature.
A useful thermal model tracks the growth of every relevant component. This includes the shaft, inner and outer rings, housing, spacers, gears and locating shoulders. Material coefficients, temperature gradients and contact resistances all matter. A gearbox may have a relatively modest average temperature while still developing local differences around a bearing seat or near a gear mesh. Treating the entire assembly as a single uniform temperature can hide the condition that causes peak preload.
Lubricant behaviour adds another layer. As oil becomes hotter, viscosity usually falls, reducing churning losses and some viscous drag. At the same time, higher bearing contact force can increase sliding, rolling resistance and heat generation. Inadequate flow can produce a local temperature rise that changes clearance further, creating a feedback loop. The thermal balance must include frictional torque, oil delivery, heat transfer through the housing and heat carried away by the lubrication system.
For high-speed tapered roller bearings, friction torque under combined load is a valuable input to that calculation. The bearing friction torque study provides relevant context for understanding how speed, load and contact conditions affect heat generation. Such data can help distinguish a safe preload window from a setting that appears acceptable at assembly but becomes unstable during flight.
Managing manufacturing variation and clearances
Even a carefully designed preload method is affected by tolerances. Bearing bore and outside diameter, shaft seat size, housing geometry, spacer length, shoulder runout and assembly temperature all contribute to the final condition. In a geared engine, gear tooth errors and shaft bending can redistribute forces between bearings, so the measured preload may not represent the load experienced by each rolling element during operation.
Tolerance analysis should therefore consider distributions rather than a single nominal build. Monte Carlo studies, sensitivity analysis and design-of-experiments methods can reveal which dimensions have the greatest influence on running clearance. A spacer length may appear critical during a basic calculation, while housing ovality or shaft shoulder squareness proves more influential once realistic loads and temperature fields are included.
The target should be an acceptable operating band across the complete production population. A narrow nominal value with poor tolerance robustness can create rejected parts, inconsistent service life and difficult maintenance decisions. A slightly wider but stable preload range may provide better total performance, particularly when components are manufactured or repaired at different sites.
Australian operating conditions make this robustness especially practical. Gearboxes supporting aircraft in Darwin or northern Queensland can encounter high ambient temperatures, while aircraft operating from Adelaide or Melbourne may see much cooler ground conditions and different turnaround patterns. Components may also move through a local aerospace supply chain before final assembly, making clear measurement procedures and digital records important for consistent fit.
Validating the design through test and simulation
Simulation should be connected to physical evidence. A coupled thermal-structural model can predict ring growth, shaft displacement, bearing contact force and operating clearance across speed and load points. Computational models may include elastic deformation, lubricant properties and gear mesh forces, but their value depends on accurate boundary conditions and measured material behaviour.
Test rigs can then establish whether the predicted preload and temperature trends are credible. Useful measurements include bearing outer-ring temperature, oil inlet and outlet temperature, shaft displacement, housing strain, torque and vibration. Testing should cover cold start, steady-state operation, transient acceleration and shutdown, because the most severe condition may occur during a transition rather than at maximum continuous speed.
A carefully designed test programme can separate the effects of preload, lubricant flow, speed and external load. This is where design of experiments is more efficient than changing one variable at a time across a large number of runs. It can identify interactions, such as a preload method that performs well at moderate speed but creates excessive heat when oil temperature and axial load rise together.
The OPTIMIZE research programme demonstrates how these engineering activities can be integrated around power reduction gearboxes for geared aircraft engines. For an Australian aerospace manufacturer or maintenance organisation, the same approach supports decisions about inspection limits, replacement parts and field servicing. It is particularly useful where access to specialised test equipment is limited and each test campaign must produce dependable evidence.
A well-instrumented rig can also help establish practical acceptance criteria. For example, a temperature rise limit may be paired with a vibration threshold and a measured endplay range. This creates a more meaningful health assessment than relying on a single static preload measurement. The results can feed back into assembly instructions, bearing selection, lubrication schedules and future gearbox revisions.
The final design should state how preload is created, how it is checked and how thermal growth is accommodated. It should define the allowable clearance range at assembly and the expected range at operating temperature. It should also explain what happens after bearing replacement, spacer wear, housing repair or lubricant changes. Clear documentation reduces the risk that a correct design is undermined by inconsistent production or maintenance practice.
For Australia’s aviation market, this clarity has operational value. Aircraft and propulsion systems may support long routes, regional services, defence activity or remote logistics, where an unexpected gearbox inspection can be costly. Reliable preload control can extend bearing life, reduce parasitic power loss and improve confidence in high-speed operation. It also supports local engineering, testing and maintenance capability rather than treating the gearbox as an isolated imported component.
Engineers developing geared aircraft engines can use preload studies to balance stiffness, efficiency, durability and manufacturability from the earliest design stage. Explore the OPTIMIZE Project’s research and engineering resources to examine how thermal modelling, tolerance analysis and testing can turn bearing clearance into a controlled performance variable.