Thrust bearing arrangements for high-speed helical gears
Helical gears provide smooth, quiet power transmission, but their angled teeth also generate axial thrust. Every gear mesh therefore loads its shaft in the direction of the helix, and the resulting forces must be transferred safely through the gearbox structure. A suitable thrust bearing arrangement controls that load while accommodating shaft deflection, thermal growth, manufacturing variation and the high rotational speeds found in aircraft engine gearboxes.
The OPTIMIZE Project combines design-of-experiments methods, simulation, tolerance analysis and physical testing to improve the efficiency, durability, weight and power density of geared aircraft engine power-reduction systems. Its bearing work is particularly important because an apparently small axial-load problem can influence gear alignment, lubricant behaviour, vibration and fatigue life across the complete drivetrain.
Understanding the thrust load path
The axial component produced by a helical gear can be estimated from the tangential tooth force and helix angle. For an ideal gear, the relationship is approximately (F_a = F_t \tan \beta), where (F_t) is the tangential force, (F_a) is the axial thrust and (\beta) is the helix angle. The sign and direction depend on the tooth-hand arrangement and whether the gear is the driving or driven member.
In a multistage gearbox, these forces must be combined rather than considered in isolation. A driven gear can receive thrust in one direction while the shaft also carries thrust from a neighbouring stage, an output coupling or an attached compressor. Depending on the layout, the loads may reinforce one another or partly cancel. The bearing arrangement should be selected from the complete load path, including reversing torque, start-up transients, shock loads and any preload from the output train.
Axial load must also be separated from the radial and moment loads acting on each bearing. A bearing with adequate radial capacity may still be unsuitable if its axial rating, contact angle, internal clearance or mounting method cannot carry the calculated thrust. The design process should define separate envelopes for steady operation, peak torque, overspeed and thermal soak. This prevents a nominal load case from concealing a short-duration overload that controls bearing or housing fatigue.
Choosing a fixed and floating architecture
A fixed-floating arrangement is often the most practical starting point. The fixed bearing locates the shaft axially and carries the primary thrust load, while the floating bearing supports radial load and permits controlled axial movement. The floating member might be a deep groove ball bearing, a cylindrical roller bearing or another type selected according to speed, load direction and lubricant requirements. This arrangement reduces the risk of a hyperstatic condition in which several bearings and housing features compete to impose incompatible axial positions.
The bearing types then determine how the thrust is actually reacted. Paired angular contact bearings provide positive axial support, while tapered roller bearings offer high thrust capacity and are often useful in demanding gear drives. Their greater sliding contact creates additional friction and heat. Deep groove ball bearings have lower frictional losses but more limited axial capacity. Selection should therefore balance thrust rating against efficiency rather than simply choosing the bearing with the highest static load limit.
Bearing architecture options
- A fixed angular contact or tapered roller location combined with a floating radial support limits axial overconstraint.
- Two locating bearings can improve positional control where both shaft ends must resist thrust, provided thermal expansion and mounting tolerances are explicitly included.
- A floating cylindrical roller bearing can accommodate axial movement while carrying substantial radial load from the helical gear.
- Duplex angular contact bearings can provide bidirectional thrust capacity where reversing axial loads cannot be avoided.
Selection criteria
- Calculated axial and radial loads with suitable service factors for start-up, shock and maximum torque conditions.
- Required fatigue life, static safety margin, limiting speed and sensitivity to internal clearance.
- Preload, mounting fit, housing stiffness and the resulting change in shaft alignment.
- Lubricant type, oil flow, temperature range, contamination control and service intervals.
- Compatibility with inspection, replacement and the traceability expected by Australian aerospace organisations.
The final choice must be supported by load-deflection data rather than catalogue ratings alone. For early bench rigs, a laboratory-equipment supplier such as Cavendish Instruments can support mechanics experiments and measurement fixtures, while production bearings still require the appropriate aerospace quality and traceability. After comparing the alternatives, the chosen architecture should be entered into the rotor model with realistic fits, clearances and housing constraints.
Controlling preload, lubrication and temperature
Preload removes internal clearance and helps prevent gear walk, vibration and impact during transients. It does not increase the external load capacity indefinitely. Excessive preload raises rolling-element and contact stress, consumes lubricant through greater churning and sliding, and can reduce bearing life. The appropriate value depends on bearing geometry, internal clearance, mount stiffness, operating speed and temperature. Light, repeatable preload is usually more effective than a large force created by imprecise shimming.
Temperature changes the relationship between preload, internal clearance and housing fit. A shaft that expands axially can create loads that did not appear in the cold model, while a warm lubricant changes viscosity and the hydrodynamic film supporting the rolling elements. The arrangement should therefore be checked over the full operating envelope. Suitable methods include oil-bath lubrication, directed jets, circulating oil or supplier-recommended combinations, with seals and drains positioned to prevent aeration or contamination.
Australian procurement adds practical constraints to this engineering work. Many precision bearings, machine tools and specialty lubricants are imported, and Australian Customs and Border Force requirements include correct tariff classification, import declarations, GST treatment and agreed Incoterms. Lubricants may also require dangerous-goods documentation. Melbourne’s aerospace and precision-engineering cluster can support local machining of spacers and adaptors, while Brisbane’s universities and advanced-manufacturing businesses offer useful capability for instrumentation and rotating-equipment research.
Import lead times make it especially important to qualify alternatives before a rig is ready to run. General web results can contain stale or misleading supplier information, so any external supplier listing should be checked against manufacturer records, test certificates and approved supplier lists before a component is purchased. Local distributors may hold standard sizes, but an aerospace project should confirm provenance rather than assuming that apparently identical bearings are equivalent.
Verifying dynamics, tolerances and test evidence
Critical speed and rotor stability depend strongly on the span and stiffness of the shaft bearings. Moving one support can change the first flexible mode, the position of an operating speed on a Campbell diagram and the response to gear-mesh excitation. A detailed OPTIMIZE discussion of shaft bearing span explains why this geometric decision belongs in the early dynamic model, rather than being treated as a late packaging adjustment.
Tolerance analysis should connect bearing selection to the variation expected in the gearbox. Relevant parameters include helix angle, tooth lead, profile error, face width, shaft runout, bearing clearance, fit, housing roundness and coaxiality. Monte Carlo analysis can reveal whether an occasional combination of dimensions produces edge loading, excessive preload or unstable support stiffness. Design-of-experiments methods can then identify which variables have the strongest influence on thrust load, vibration, friction loss and bearing temperature.
The simulation should include both nominal and off-nominal conditions. It should also represent the floating clearance rather than locking every node to the housing. This is essential in a hyperstatic arrangement, where small manufacturing differences can redistribute large internal loads without a corresponding change in external thrust. Validated models should predict axial reaction, shaft deflection, bearing life, lubricant temperature and vibration, giving the test team measurable acceptance criteria rather than a single pass-or-fail spin result.
Moving from design to an Australian test programme
Physical testing should progress from low-speed loading through speed sweeps, thermal equilibrium and controlled overload cases. Useful measurements include axial force, torque, bearing-metal temperature, housing vibration, oil pressure and flow. Sensors should be positioned close enough to distinguish thrust reaction from housing distortion. After testing, the gearbox can be inspected for heat marks, lubricant degradation, fretting, brinelling and evidence of uneven contact.
Local operating conditions should shape the test plan. Melbourne can experience substantial daily temperature variation, while coastal operations around Sydney, Brisbane and Darwin expose components to salt-laden air. Australian charter, regional airline, aerial-survey and maintenance organisations also place a premium on reliability, repairability and predictable life because grounded aircraft and long component lead times are costly. A rig located in an Australian facility should therefore use locally representative lubricants, ambient conditions and contamination controls.
The bearing arrangement should be judged as part of the whole gearbox rather than as an isolated component. A lower-drag support may reduce heat and improve power density, yet excessive movement can increase gear misalignment. Conversely, rigid supports can improve alignment while creating thermal overload. The most credible design is the one that remains stable across expected tolerances and demonstrates repeatable performance through simulation and physical evidence.
Researchers and aerospace suppliers interested in contributing test data, developing a bearing rig or aligning their work with the OPTIMIZE objectives can register through the project contact page. A defined load case, proposed instrumentation and shared validation data will help turn the thrust bearing arrangement into a practical improvement in gearbox efficiency, durability, weight and power density.