Carburized gear case-core debonding and hardness profile effects
When a turboprop or turbofan engine reduces fuel burn by a single percentage point, the ripple through route economics, emissions reporting, and payload planning is felt from Brisbane to Perth. That pressure cascades into the drivetrain, where every reduction stage must transmit torque with tighter losses than the generation before. Within these gear stages, the case-hardened flank of each tooth carries enormous cyclic and bending loads while the tougher core absorbs impact and resists fracture. The interface between those regions is where a subtle metallurgical variable can decide whether the gearbox runs a full 25,000-hour life or returns to a maintenance base in Adelaide wearing a hairline crack no ultrasonic probe had flagged before teardown.
Researchers working on the OPTIMIZE Project, an Australian-European Clean Sky 2 initiative, have turned their attention to that interface. By systematically varying hardness profiles during case hardening and recording what happens at the case-core boundary, the consortium is producing evidence to refine both design margins and process controls for next-generation power reduction gearboxes intended for propulsion applications.
The metallurgical basis of case-core integrity
A carburized gear is not a single material but a manufactured gradient. Carbon is diffused into the surface during furnace treatment, raising carbon equivalent to roughly 0.7 to 1.1 percent in the outer millimetres while the inner alloy retains its lower bulk composition. Quenching then transforms the high-carbon surface into a hard martensitic structure with retained austenite, while the lower-carbon core converts to a tougher mix of martensite, bainite, and fine ferrite. The two regions are metallurgically continuous, yet their mechanical responses diverge sharply under load.
That gradient is what the engineer exploits. The case resists contact fatigue and wear from surface-initiated micro-pitting; the core resists tooth root bending fractures that originate below the surface. Aerospace gearing relies on this division of labour to survive the 10,000 to 30,000 rpm ranges encountered in modern reduction gearboxes. Any metallurgical disruption along the transition zone weakens both functions simultaneously, allowing a sub-surface flaw to grow toward the surface as Hertzian stress cycles continue.
The OPTIMIZE research team has observed that profile deviations of just a few HRC points near the effective case depth are enough to shift the load-bearing response of a test pinion. At RMIT's Advanced Manufacturing Precinct, repeat batches have confirmed the trend with statistical confidence, suggesting that profile shape matters as much as nominal hardness values stamped on a drawing.
How hardness profiles emerge from carburizing
The shape of the hardness curve is not a single decision but the outcome of competing process variables. Atmosphere carbon potential, furnace temperature, soak time, quench severity, and subsequent tempering all leave measurable signatures. A longer soak pushes the diffusion front deeper; a more aggressive quench sharpens the surface-to-core transition; a higher carbon potential elevates the surface peak but risks excessive retained austenite that may later transform in service.
Process variation is not a defect in itself; it is the natural consequence of running batch furnaces across multi-week campaigns. Aerospace forges in Brisbane, western Sydney, and the outer suburbs of Melbourne routinely hold work for redistribution within their heat-treat queues, where each delay shifts the soak profile. Uncontrolled humidity, atmosphere composition drift, and quench bath ageing introduce further scatter that complicates repeatable outcomes.
The OPTIMIZE workflow treats these variables as inputs to a designed experiment rather than as noise to be averaged. By running full factorial and space-filling designs across candidate process windows, the team has built predictive models linking furnace set points to the resulting hardness curve at multiple depths. Those models feed directly into manufacturing tolerance conversations, which in Australia are constrained by the airworthiness framework administered by the Civil Aviation Safety Authority and, for military programmes, by the Department of Defence's airworthiness authority.
Mechanisms of case-core debonding under load
When a case-core interface begins to separate, the failure is rarely sudden. It starts as a sub-surface discontinuity, often a thin seam of brittle microstructure or a network of micro-voids coalesced along prior austenite grain boundaries. Under repeated Hertzian contact, these seams grow parallel to the running surface, then tilt toward the nearest free surface to form a spall.
Several mechanisms accelerate that propagation when the hardness profile is shaped unfavourably. Excessive retained austenite at the surface transforms to brittle martensite under operating heat and stress, generating dimensional micro-strains at the transition zone. An over-deep case allows plastic flow to push energy into softer material below, while an under-deep case places a sharp load step in the region where bending stress peaks. Residual tensile stresses from aggressive carburizing can stack with operational stresses to push the local stress state past the bond strength of the case-core boundary.
The OPTIMIZE team's published bench tests show that profile curvature in the 0.4 to 0.8 mm below-surface band is most predictive of spalling life. Where the gradient falls too steeply, sub-surface shear bands concentrate; where it falls too shallowly, the surface deforms plastically and initiates case crushing. The optimum is rarely a flat hardness line but a specific concave-up curve that aligns the iso-stress contours with tough material.
The aerospace operating environment and its pressures
Geared engines complicate everything. The reduction gearbox sits between a high-pressure turbine shaft and a slower propeller or fan shaft, absorbing angular misalignment, thermal growth, and torque spikes. The OPTIMIZE baseline configuration targets power densities above 200 kW per kilogram, which means each kilogram of mass is doing more mechanical work than in any previous generation of regional turboprop hardware.
That weight pressure is felt acutely in Australian fleet economics, where operators shuttle between Sydney, Cairns, and remote communities through the Torres Strait and across the outback. A heavier gearbox burns more fuel per sector, eats into payload on bush runs, and increases maintenance intervals that already feel the strain of fly-in fly-out workforce logistics at regional bases such as those serving the Pilbara mining region. Lighter, longer-lasting gearboxes directly improve the business case for direct air links that maintain social and medical access to remote communities.
Reliability expectations are equally unforgiving. Australian defence operates a growing F-35A fleet out of RAAF Williamtown near Newcastle, where engine availability directly affects pilot training throughput. The Australian Defence Innovation Hub has funded several adjacent propulsion and driveline projects, recognising that reducing unscheduled maintenance hours translates into fleet readiness rather than hangar time. Temperature gradients caused by uneven oil distribution across a planetary stage can shift the effective case depth during a single test cycle, a relationship documented in the planetary temperature gradients study from the wider project team.
Validation through static and dynamic testing
A hardness profile model is only credible if the engineering community can replicate the test outcome with independent hardware. The OPTIMIZE programme addresses that challenge through a structured testing pyramid: coupon-level specimens for material characterisation, sub-scale gear pairs for contact pattern validation, and full bench drivelines representative of the final architecture.
The geometry problem is the hardest to control. A test rig that loads a real gear differently from the design intent will produce a misleading dataset. The team's open static test design methodology shows how to lay out a static load case that reproduces the predicted contact patch without imposing parasitic bending, offering one of the more pragmatic tutorials now available for cross-checking bench tests against finite element predictions.
Dynamic testing then layers in lubrication, speed, and thermal effects that no static rig can replicate. Planetary test rigs in partner laboratories have already produced oil-out temperature gradients large enough to shift the effective case depth during a single test cycle, with consequences that the team's models now capture in their thermal-mechanical coupling.
Manufacturing tolerance analysis and process control
Once the design experiment identifies the optimum profile, the supply chain has to deliver it consistently. This is where tolerance analysis becomes the bridge between laboratory ideal and production reality.
Drivers of heat-treat variability
The challenges fall into a small set of recurring categories:
- Furnace atmosphere variability, including carbon potential drift over campaign life, moisture ingress, and probe calibration cycles.
- Quench uniformity, where bath temperature gradients and agitation flow paths introduce scatter between identically loaded fixtures.
- Material variability in incoming steel bar, including residual alloying element deviation and prior austenite grain size.
- Heat-treat fixture design, where shielding and stacking geometry alter local convection and radiation balance.
- Post-process handling, including straightening and tempering rates that introduce residual stress shifts.
Each of these variables can be captured in a statistical tolerance budget. The team has used Monte Carlo propagation from the hardness profile model through to predicted spalling life, generating probability distributions that maintenance planners can use to set realistic inspection intervals. Where the budget is too loose, manufacturing invests in tighter control; where it is tighter than necessary, cost is recovered without performance penalty.
From design data to industrial decision-making
Engine programmes do not run on metallurgy alone. They run on airworthiness compliance, supply chain readiness, and the willingness of certifying engineers to sign off on hardware that meets a novel specification.
Practical considerations for industrial adoption
Several practical considerations shape how the OPTIMIZE findings are likely to translate into industrial practice:
- Engagement with CASA airworthiness staff early in the design freeze, so that the chosen hardness profile window is documented with enough margin to satisfy Part 33 fatigue and wear provisions.
- Collaboration with Australian forging and heat-treatment partners around Melbourne, Adelaide, and Brisbane to qualify secondary suppliers, building resilience into a domestic gear manufacturing base.
- Investment in non-destructive evaluation techniques, such as advanced ultrasonic and eddy-current methods suitable for case-depth verification on every shipped component.
- Inclusion of profile acceptance criteria as a contractual deliverable rather than a statistical hope, with cost-of-quality reviews scheduled into supplier scorecards.
- A pipeline of trained metallurgists from institutions such as the University of Melbourne, RMIT, and the University of Adelaide, all of which run targeted aerospace materials programmes with industry placements.
These steps will determine whether the laboratory gains become embedded flight hardware or remain as academic publications. The narrower technical takeaway is that hardness profile shape, not just case depth, must be controlled as a first-class design variable. The wider implication is that designing a power reduction gearbox benefits from the same statistical mindset that has reshaped manufacturing elsewhere: define the response surface, control the inputs, and only then trust the certification argument.
The link between hardness profile and case-core debonding is one of those engineering realities that hides in plain sight, visible to anyone studying a failure yet often under-treated in the hurry to ship a new architecture. As the OPTIMIZE Project builds out its full dataset, the team is making material available to qualified partners through the https://optimizecleanskyproject.com/members-area portal, with summaries appearing on the project's public blog channels. Aerospace propulsion teams across Australia and beyond have an opportunity now, while the programme is still forming its final recommendations, to subscribe to those updates, request review copies of the technical reports, and bring their own field experience into the conversation before the specification freezes. Engaging early will shape the gearboxes of the next decade long before a single test article is cut.