Gear Load Sharing and Bearing Life in Starved Lubrication
A geared aircraft engine depends on a compact transmission system operating under severe mechanical and thermal conditions. The gearbox must transfer high torque at speed while keeping mass, vibration, friction and heat within tightly controlled limits. Bearings are central to that task, and their service life can change significantly when lubrication is insufficient.
One important variable is the gear load sharing factor. It describes how evenly a transmitted load is distributed across mating gear teeth, planets, bearings or parallel power paths. A study of this factor under starved lubrication helps connect gearbox efficiency with bearing reliability, manufacturing tolerance, oil supply and real operating conditions. These relationships are central to the OPTIMIZE project’s work on power reduction gearboxes for geared aircraft engines.
Why Load Sharing Matters
In an ideal gearbox, every load-carrying path would receive its intended share of the torque. Real hardware behaves differently. Gear misalignment, shaft deflection, housing distortion, tooth spacing errors and bearing clearance can force one path to carry more load than another. The load sharing factor captures this uneven distribution through a multiplier applied to the nominal force.
If the factor rises from 1.0 to 1.2, a bearing or gear path may experience roughly 20 per cent greater local loading than the simplified calculation suggests. Bearing life does not decline in a linear manner. For many rolling-element bearings, fatigue life is approximately proportional to the cube of dynamic load rating divided by equivalent load. A moderate increase in load can therefore produce a much larger reduction in calculated life.
This effect becomes especially important in planetary gear trains. Several planets can share torque, yet small differences in stiffness and position may cause one planet-bearing assembly to take a disproportionate load. The resulting contact stress, sliding, cage forces and lubricant demand can rise together.
Starved Lubrication Mechanisms
Starved lubrication occurs when the rolling contacts receive less oil than required to form a fully replenished elastohydrodynamic film. The issue is different from total oil starvation. A bearing may still receive some lubricant, but the available volume, replenishment rate or inlet supply is too low for stable film formation across all operating conditions.
Several mechanisms can create this condition. High rotational speed may fling oil away from the raceway inlet. Seals and shields can restrict replenishment, while churning losses can disturb the flow field. Oil may also drain rapidly from a hot bearing, become aerated, or arrive with a temperature and viscosity that are unsuitable for the actual contact conditions.
A thin film increases the risk of asperity interaction between the raceway and rolling element. Frictional heat can then increase, causing viscosity to fall further. In a gearbox designed for low power loss, the trade-off is demanding: less oil may reduce churning drag, but inadequate supply can accelerate wear, smearing, micropitting and fatigue damage.
The effect of load sharing is coupled to this process. An overloaded path produces a smaller film thickness because contact pressure rises, while the same path may also receive less lubricant due to local flow restrictions. This means a bearing can become the critical component even when the average gearbox oil flow appears adequate.
Modelling Bearing Life
A useful analysis begins with a baseline bearing-life calculation using the nominal radial and axial loads, speed, temperature and lubricant viscosity. The load sharing factor is then introduced as a variable rather than treated as a fixed design constant. This allows engineers to examine the sensitivity of life to load imbalance.
For a rolling bearing, the equivalent dynamic load can be adjusted to reflect the extra force on the most heavily loaded path. The calculation should also account for reliability target, contamination, surface finish, material quality and lubrication condition. A modified rating-life approach is more realistic than relying solely on the basic L10 value, particularly for an aircraft gearbox where reliability requirements are high.
Starved lubrication should be represented through a film parameter such as the ratio of lubricant film thickness to composite surface roughness. This parameter, often written as lambda, helps distinguish full-film, mixed and boundary-dominated regimes. It should be evaluated over the complete duty cycle rather than at a single rated point.
The critical outputs include predicted bearing life, peak contact stress, temperature rise, friction torque and the probability of entering a damaging lubrication regime. A time-weighted duty cycle can combine take-off, climb, cruise, descent and transient operation. For Australian aviation applications, long distances and remote operating environments make this approach valuable because a gearbox may spend extended periods at steady high load far from a major maintenance base.
Design Of Experiments And Tolerance Effects
Design-of-experiments methods provide an efficient way to identify which variables control the response. Candidate factors may include gear load sharing factor, oil flow rate, oil inlet temperature, speed, bearing clearance, shaft misalignment, housing stiffness and surface roughness. A screening design can identify dominant effects before a response-surface model explores interactions.
The interaction between load sharing and oil supply is often more important than either factor in isolation. At a moderate load imbalance, the bearing may retain a safe film thickness. Under reduced flow, however, the same imbalance may push the contact into mixed lubrication. A statistical model can reveal this threshold and indicate where design margins are being consumed.
Manufacturing variation should be included explicitly. Tooth thickness, planet pin position, bearing internal clearance and carrier geometry all influence how torque is divided. Tolerance analysis can show whether a nominally efficient gearbox becomes unreliable when several adverse deviations occur together.
This approach aligns with the broader OPTIMIZE methodology, which combines simulation, tolerance analysis and physical testing to improve power density and durability. Collaboration between universities, component manufacturers and propulsion specialists is particularly important; the project’s network of engineering partners reflects the multidisciplinary nature of this work.
Simulation And Physical Testing
A multibody or finite-element model can estimate gear mesh forces, shaft deformation and bearing reactions. Computational fluid dynamics or specialised lubrication models can then examine oil distribution and the likelihood of an under-supplied contact. These models need careful calibration because small assumptions about inlet conditions can have a large effect on predicted film thickness.
A practical test programme may begin with a bearing or gear-bulk rig before progressing to a representative gearbox. Test variables should include torque, speed, oil temperature, flow rate and imposed misalignment. Instrumentation can include thermocouples, accelerometers, torque meters, oil-debris sensors and vibration monitoring.
Bearing temperature is a useful indicator, but it should not be treated as the only health measure. A bearing may experience surface distress before the bulk temperature reaches an obvious alarm level. Post-test inspection, lubricant analysis and microscopy can identify smearing, micropitting, discolouration and wear particles that validate the life model.
Digital access to project information also matters when teams operate across laboratories and time zones. Technical groups may review test results outside normal workshop hours, alongside public-facing resources such as 24/7 slots, although engineering decisions should remain governed by controlled records, approved procedures and traceable data.
Australian Operating Context
Australian aerospace engineering often has to account for long travel distances between design offices, test facilities and operating locations. Work connected with Melbourne, Sydney, Brisbane, Adelaide and Perth may involve different supplier networks and transport arrangements. A gearbox that performs acceptably in a controlled laboratory still needs robust logistics for oil samples, replacement hardware and inspection equipment.
Environmental exposure can also affect lubrication performance. High ambient temperatures in inland regions reduce oil viscosity and may reduce film thickness at a given speed and load. Dust and contamination risks are relevant to aircraft operating from regional or remote locations, while coastal facilities may need to consider salt-laden air during storage and ground operation.
Civil aviation designs and maintenance practices must fit the Australian regulatory framework administered by the Civil Aviation Safety Authority. Depending on the aircraft, modification and approval pathway, evidence may need to support requirements under the Civil Aviation Safety Regulations and related airworthiness processes. This makes documented load cases, material traceability, configuration control and repeatable test evidence essential.
The local market also values efficient propulsion because fuel, freight and maintenance costs are significant across long domestic sectors. Operators serving mining, regional communities and remote infrastructure may place a high value on reliability and predictable servicing. A gearbox that saves power through reduced oil churning but suffers premature bearing damage would create an unacceptable whole-of-life cost.
Comparing Design Scenarios
The following comparison illustrates how the load sharing factor and lubricant supply can shift the likely bearing-life outcome. The values are conceptual rather than certification data; an actual programme would use measured geometry, approved lubricant properties and validated bearing models.
| Scenario | Load sharing factor | Lubricant condition | Expected film behaviour | Bearing-life tendency | Main engineering response |
|---|---|---|---|---|---|
| Balanced nominal operation | 1.00 | Full replenishment | Stable elastohydrodynamic film | Highest calculated life | Maintain alignment and oil delivery |
| Mild load imbalance | 1.10 | Full replenishment | Film margin reduced locally | Moderate life reduction | Improve stiffness and tolerance control |
| Mild imbalance with reduced flow | 1.10 | Starved inlet | Intermittent film replenishment | Significant life reduction | Optimise jets, drains and flow rate |
| High load imbalance | 1.20 | Full replenishment | High pressure, thinner local film | Sharp life reduction | Correct load distribution and geometry |
| High imbalance with hot oil | 1.20 | Starved and low viscosity | Mixed lubrication likely | Highest damage risk | Control temperature, supply and clearance |
A robust design should target the combined condition rather than optimising each parameter separately. For example, increasing oil flow may improve film thickness, but it can also increase churning power and gearbox temperature if the flow path is poorly designed. Similarly, tightening manufacturing tolerances may improve load distribution while increasing cost or reducing assembly flexibility.
Sensitivity results should be expressed with confidence intervals where possible. A mean predicted life can conceal a wide spread caused by tolerance variation. Reliability-based design is better served by identifying the lower tail of performance and establishing inspection or monitoring strategies around credible degradation mechanisms.
Recommendations For Gearbox Development
The study can support decisions from concept design through certification testing. The most useful recommendations are:
- Model load sharing as a distribution, with manufacturing tolerances and structural deflection included.
- Couple bearing-life calculations with oil-flow, viscosity and temperature predictions.
- Test starved-lubrication conditions at realistic speed, torque and transient duty-cycle points.
- Use vibration, temperature, oil debris and post-test inspection as complementary evidence.
- Set design limits for load imbalance, inlet flow and lubricant temperature before hardware testing.
- Maintain traceable configuration records for geometry, material batches, lubricant specification and test data.
These actions help prevent an efficient gearbox from becoming fragile in service. They also support a clearer trade-off between power loss, mass, bearing life and manufacturing cost. In an aircraft propulsion system, the best design is the one that retains margin across realistic variations rather than the one that delivers the strongest result at a single nominal condition.
A focused study of load sharing under starved lubrication can therefore improve more than a bearing calculation. It can reveal weaknesses in carrier stiffness, oil distribution, tolerance allocation and monitoring strategy. When combined with design-of-experiments methods and physical validation, the result is a more dependable basis for developing compact geared engines.
For Australian aerospace organisations, this evidence can strengthen collaboration between local manufacturers, research institutions, maintenance providers and regulators. It can also support propulsion systems suited to long sectors, high ambient temperatures and demanding regional service. The next step is to turn the model into a controlled test matrix, verify the critical interactions and use the findings to refine the gearbox before production commitments are made.