How Case Carburizing Depth Variation Changes Tooth Fatigue Scatter
For a geared aircraft engine, tooth bending fatigue is governed by more than nominal geometry and rated torque. The case-hardening profile creates a load-bearing gradient through each tooth, and small differences in carburizing depth can change where cracks start, how quickly they grow, and how widely test results are distributed. The question of how case carburizing depth variation affects tooth bending fatigue scatter therefore belongs in the design process, not only in final quality inspection.
This matters particularly in reduction gearboxes, where high rotational speed, compact dimensions, repeated load reversals and strict mass targets leave little room for unmeasured variation. A design that performs well at the average material condition may still produce an unacceptable lower tail of fatigue life when manufacturing tolerances, heat treatment, lubrication and installation effects interact.
The Role Of The Carburized Case
Carburizing enriches the steel surface with carbon before hardening, producing a hard martensitic case over a tougher, more ductile core. The case resists wear and supports high contact stress, while the core helps the tooth absorb bending loads without behaving as a brittle solid component. This transition is gradual rather than abrupt, and its depth is commonly assessed using effective case depth, total case depth and hardness profiles.
At the tooth root, bending stress is concentrated around the fillet. A sufficiently deep case can keep the highest cyclic stress within strong, hardened material. If the hardened layer is shallow, the critical stress zone may extend into material with lower hardness and different residual stresses. The result can be earlier crack initiation, reduced permissible stress, or a shift in the failure mechanism from surface-dominated damage to subsurface or root cracking.
The useful depth is also influenced by tooth size, profile modification, root radius, shot peening and the applied load spectrum. A case depth that is suitable for a large industrial gear may be excessive or insufficient for a lightweight aerospace gear. The specification must relate the heat-treatment profile to the actual bending stress field rather than treating case depth as an isolated number.
Why Small Variations Produce Wide Fatigue Results
A batch can meet its average carburizing specification while individual teeth or components carry meaningfully different fatigue strength. Furnace temperature distribution, carbon potential, atmosphere flow, loading arrangement, quench delay and part orientation can all affect the final profile. Grinding stock removal adds another source of variation because it can reduce the remaining hardened layer at the tooth flank or root.
Fatigue life is especially sensitive near a threshold. When the applied root stress approaches the local endurance capability, a modest reduction in effective case depth can cause a large change in cycles to crack initiation. This non-linear response explains why fatigue-test results often show a broad scatter band rather than clustering tightly around the mean value. A small population of weak parts can control the design decision.
Surface finish, grinding burns, inclusions, retained austenite and residual compression also interact with case depth. A deep case does not automatically guarantee high bending strength if the tooth root has tensile residual stress or if grinding has damaged the hardened layer. Conversely, a slightly lower measured depth may be acceptable when beneficial compressive stress and a clean root geometry provide additional protection.
The gearbox environment adds another layer. Thermal gradients, lubricant temperature and transient torque can alter tooth load sharing during operation. In a hyperstatic arrangement, small alignment errors can distribute load unevenly across several gear meshes. The OPTIMIZE discussion of gearbox misalignment effects is relevant because a locally overloaded tooth can expose weaknesses that a uniform laboratory load would not reveal.
Connecting Material Profiles With Tooth Stress
A robust assessment begins with a detailed stress map. Finite element analysis or validated gear calculation methods can identify the position and depth of maximum principal stress at the root, including the effects of fillet geometry, rim thickness, load distribution and tooth modifications. The carburized hardness profile can then be mapped onto that stress field to estimate the strength available at each critical location.
This approach is more informative than comparing one effective case-depth value with one catalogue recommendation. Two gears with the same effective depth can have different hardness gradients, retained-austenite levels and residual-stress distributions. Their bending fatigue performance may therefore differ even when both pass the same basic inspection.
Design-of-experiments methods help separate the influence of these variables. A study can vary case depth, root radius, torque, misalignment, material cleanliness and shot-peening intensity in a controlled simulation campaign. Response measures may include maximum root stress, crack-initiation probability, predicted fatigue life and the width of the lower confidence bound. This produces a clearer view of interactions than changing one factor at a time.
The OPTIMIZE Project’s project objectives reflect this systems approach, linking simulation, tolerance analysis and physical testing to improved efficiency, durability, weight and power density. For a geared aircraft engine, the aim is to understand how manufacturing variation changes gearbox behaviour before committing to expensive hardware trials.
Measuring The Variation That Matters
Quality control should capture the complete hardening profile rather than relying on a single microhardness traverse from a convenient coupon. Measurements may be taken at representative tooth roots, flanks and transverse sections, with attention to the areas affected by grinding. The inspection plan should also establish how sample location, orientation and measurement spacing influence the recorded result.
Statistical process control can reveal whether variation is random, drifting or associated with a particular furnace load. Useful data include effective case depth, surface hardness, core hardness, retained austenite, residual stress and dimensional change. Linking these records to the individual gear or batch makes it possible to identify whether fatigue outliers arise from heat treatment, machining, material quality or assembly.
Tolerance analysis should use distributions rather than nominal values alone. For example, a Monte Carlo model can combine case-depth variation with root-radius tolerance, tooth thickness, runout, torque fluctuation and bearing displacement. The output is a probability distribution for root stress or fatigue life. Engineers can then set limits based on an agreed reliability target instead of selecting a tolerance simply because it is familiar.
Australian aerospace supply chains make this traceability especially valuable. A program may involve heat treatment in Melbourne, gear finishing in another state and final gearbox assembly near Brisbane or Adelaide. Transport, subcontracting and limited local batch volumes can make rapid root-cause investigations difficult. A consistent digital record reduces the risk of treating every unexpected test result as an isolated event.
Designing Tests That Expose The Lower Tail
Physical testing should represent the loads and conditions that make case-depth variation consequential. Constant-amplitude tooth-bending tests are useful for comparing material conditions, but they may not capture the effect of torque transients, load spectrum, thermal cycling or misalignment. Component tests and gearbox rigs can show whether local load concentration amplifies the influence of a shallow or uneven case.
A practical test matrix includes gears from different positions within a heat-treatment load, not only parts selected near the process average. Specimens should retain their measured manufacturing data so fatigue performance can be correlated with actual hardness and depth profiles. Testing several nominally identical parts helps estimate scatter, while deliberately including near-limit conditions helps validate the lower-bound model.
Fractography provides another important link between process and failure. Crack origin location, intergranular features, inclusions, grinding damage and the depth of the hardened region can show whether the predicted mechanism is correct. If cracks repeatedly initiate below the expected case boundary, the design model may be placing too much confidence in surface hardness or using an unsuitable effective-depth definition.
Operating conditions in Australia can make environmental realism important. Aircraft and engine systems may support remote operations across the Northern Territory, Western Australia or Queensland, where long distances increase the cost of maintenance and spares. High ambient temperatures, dust exposure and extended duty cycles can intensify lubricant and thermal demands. A fatigue test that ignores these conditions may understate the value of controlling case-depth variation.
Turning Scatter Into Engineering Controls
The most effective response is not always to specify an extremely deep case. Excessive case depth can increase distortion, grinding allowance, retained austenite and processing cost. It may also reduce the toughness margin if the hardened region leaves too little ductile support beneath a highly stressed tooth root. The target should be a controlled profile that matches the stress gradient and the intended duty cycle.
Engineers can reduce risk by tightening furnace uniformity, validating carbon-potential control, improving quench repeatability and separating gears by appropriate load configuration. Root finishing, shot peening and inspection methods should be qualified together with the carburizing process. Supplier capability studies should demonstrate both process centring and variation control, rather than showing only that a small sample met the drawing limits.
A useful specification includes a minimum effective case depth, a maximum depth where toughness or distortion could become limiting, surface and core hardness ranges, and acceptance criteria for residual stress or grinding damage where appropriate. It should also define sampling frequency and the action required when a result falls outside the preferred process window.
For teams developing aircraft gearboxes, the following controls provide a practical starting point:
- Map the measured hardness and case-depth profile onto the calculated tooth-root stress field.
- Use heat-treatment batch data in fatigue-life and reliability models instead of nominal material values.
- Include misalignment, torque variation and thermal conditions in gearbox-level tolerance studies.
- Test gears from process extremes as well as from the centre of the manufacturing distribution.
- Inspect root geometry, grinding condition, retained austenite and residual stress alongside case depth.
- Set acceptance limits using lower-bound fatigue performance and aircraft duty-cycle requirements.
- Review suppliers through capability data, traceability records and repeatability demonstrations.
These measures support a balanced design philosophy: control the variation that drives failure probability, while avoiding unnecessary material or processing margins that add weight and cost. That balance is central to power reduction gearboxes, where every additional gram and every efficiency loss must be justified against durability requirements.
When the hardened case is treated as a statistical design variable, tooth-bending fatigue becomes easier to predict and manage. Simulation can identify sensitive combinations, physical tests can verify the failure mechanism, and production data can confirm whether the process remains centred over time. This closes the gap between a promising prototype and a gearbox capable of dependable service.
Apply the same discipline to current gearbox development by recording case-depth distributions, linking them to root-stress calculations and prioritising tests that reveal the lower tail of performance. A controlled, evidence-based heat-treatment strategy can improve fatigue reliability while preserving the weight, efficiency and power-density gains expected from advanced geared aircraft engines.