Input shaft torque measurement for gearbox efficiency
Using torque measurement on the input shaft to derive gearbox mechanical efficiency is most useful when it forms part of a measured power balance, not a stand-alone calculation. The OPTIMIZE Project treats gearbox efficiency as a system property shaped by gear geometry, bearing loads, lubrication, manufacturing variation and operating condition. Design-of-experiments methods, simulation, tolerance analysis and physical testing turn that complexity into traceable engineering evidence.
A reliable efficiency result depends on simultaneous measurements of torque and rotational speed, a clearly defined gearbox boundary and a sound uncertainty budget. These elements are particularly relevant to geared aircraft engines, where changes in transmitted power can influence fuel consumption, thermal margin, component life and the performance of the wider propulsion system.
For Australian readers, the technique must also fit a disciplined maintenance and certification environment. Airworthiness requirements administered by the Civil Aviation Safety Authority, approved engineering data and controlled test records matter as much as the sensor itself. The method is therefore best treated as a transferable engineering process that can be adapted for local validation.
The power balance behind efficiency
Mechanical efficiency describes the proportion of mechanical power entering a gearbox that leaves it as useful output power. With torque expressed in newton metres and angular velocity in radians per second, input and output power are calculated from (P=T\omega). The corresponding efficiency is (\eta=P_{\text{out}}/P_{\text{in}}), giving (\eta=T_{\text{out}}\omega_{\text{out}}/(T_{\text{in}}\omega_{\text{in}})).
For a fixed speed ratio (i=\omega_{\text{in}}/\omega_{\text{out}}), this can be rearranged as (\eta=T_{\text{out}}/(iT_{\text{in}})). This form shows why input torque alone is insufficient: it establishes power entering the gearbox but does not reveal how much power survives bearing, seal, gear-mesh and lubrication losses. Output torque, an independently measured output power or a validated loss model is also required.
Professional learning is part of maintaining that sound measurement practice. General guidance on reading more technical books can help engineers broaden their understanding, while current standards, approved procedures and instrument manuals should remain the controlling sources for test decisions.
What input shaft torque can reveal
An input-side torque measurement provides the boundary between the power source and the gearbox. Engineers can compare it with an output-side measurement at each operating point, revealing efficiency changes caused by speed, load, temperature, lubricant viscosity or component condition. Repeating the comparison after a modification also provides direct evidence of whether the change reduced losses or merely shifted the operating point.
If only the input shaft is instrumented, the remaining losses must be estimated from a validated thermal or analytical model. This approach can support screening and control applications, but its confidence depends on how well the model represents churning, bearing drag, seal friction, windage and accessory power. Reporting input torque as a direct efficiency value without this evidence risks assigning gearbox losses to the wrong subsystem.
The measurement boundary must be stated carefully. A gearbox may drive a pump, generator or other accessory through a separate output, and that parasitic power may be included or excluded depending on the study. Torque should also be averaged over a consistent interval because gear-mesh ripple and torsional vibration can make an instantaneous reading unrepresentative.
Instrumenting the rotating drivetrain
Common arrangements include inline torque transducers, reaction-torque sensors mounted through the test-bench structure, and strain-gauged shafts. Each has practical limitations. Inline devices can introduce alignment and torsional-compliance effects, while reaction-torque systems depend on the stiffness of the support structure. High-speed testing also requires attention to bearing temperature, sensor overspeed limits, electromagnetic compatibility and safe guarding.
Rotational speed must be measured at a defined location and time. A useful acquisition system records torque and speed synchronously, applies calibration factors derived from traceable standards and checks zero or balance before each run. Resource material on optical torque sensing is relevant when engineers are comparing contact and contactless approaches, particularly for rotating equipment where shaft access is restricted.
Uncertainty should be estimated before the test campaign, not reconstructed afterwards. Common contributors include transducer accuracy, speed error, ratio uncertainty, parasitic-load estimates, repeatability and the treatment of dynamic torque. A narrow confidence interval based only on instrument specifications can be misleading when the operating environment contributes most of the uncertainty.
Dynamic torque and gearbox response
A torque trace contains more than the mean load. Gear-mesh harmonics, bearing excitation, shaft flexibility and torsional oscillations can produce peaks that are not visible in a low-resolution display. These effects may alter both the measured input power and the phase relationship between input and output quantities. Sampling rate, filtering and averaging must therefore preserve the mechanical events that influence the efficiency result.
Housing vibration is another important consideration. The project discussion of gearbox housing resonance modes explains how gear-mesh excitation can interact with structural modes. If a housing or sensor support resonates during testing, its response may contaminate the torque signal or indicate a durability concern beyond the immediate efficiency calculation.
Engineers should compare time-domain traces, filtered mean torque, frequency content and test-bench accelerometer data. This combined view separates genuine load-dependent losses from artefacts. It also prevents the convenient practice of filtering away a mechanical problem merely because a narrower frequency band produces a more stable efficiency estimate.
Separating physical losses from uncertainty
A gearbox efficiency map should ideally separate controllable gearbox losses from accessory loads and measurement uncertainty. Gear-mesh loss varies with tooth geometry, contact pattern, sliding, surface finish and lubricant film formation. Bearing and seal losses depend on preload, fit, temperature, lubrication regime and shaft speed. Churning and windage can become significant as rotational speed rises, while heat generated by earlier losses can change the lubricant condition encountered later in the power path.
Hyperstatic bearings add another layer of sensitivity. When there is no relative movement between bearing surfaces, friction and temperature depend strongly on preload, fit and fit changes caused by thermal expansion or structural load. A simulation that assumes a stable hydrodynamic film may therefore misrepresent a real installation, particularly when manufacturing tolerances alter bearing preload.
A useful uncertainty budget reports each contribution separately and states whether errors are systematic or random. Calibration offsets can be corrected, whereas run-to-run variation should be represented in the final confidence limits. Comparing predicted and measured losses then helps engineers decide whether additional instrumentation or a more representative test article would improve the result most.
From test data to an efficiency map
Design-of-experiments methods help teams explore speed, torque, temperature, lubricant viscosity, bearing preload and geometric variables without relying on an impractically large test matrix. Simulation supplies a structured starting point, while physical tests anchor the model to reality. The most informative campaigns include repeated centre points and deliberately chosen conditions where expected loss mechanisms differ.
Results can be presented as efficiency against input speed, output torque and temperature, accompanied by uncertainty bands and information about lubricant and hardware configuration. The OPTIMIZE Project applies this type of integrated methodology to improve efficiency, durability, weight and power density. A complementary view of its wider communication work is available through a Dutch engineering discussion.
Data management is as important as the instrumentation. Each run should identify the gearbox serial number, tooth and bearing configuration, lubricant quantity and condition, ambient temperature, warm-up procedure, alignment state and calibration reference. Re-running a small number of baseline points during the campaign provides a practical check on drift.
Manufacturing variation and repeatable evidence
Production tolerances can move a gearbox away from its nominal efficiency curve even when the design is sound. Tooth errors alter contact and mesh loss; backlash changes impact and transmission error; bearing fits and housing bores change preload; and surface finish affects lubricant film and sliding. Tolerance analysis should therefore connect geometric variation to loss-sensitive variables rather than stopping at fit limits.
Monte Carlo studies, sensitivity analysis and targeted physical tests can reveal which combinations deserve attention. The most influential variables can then inform inspection requirements, tolerance allocation or manufacturing process controls. This converts efficiency testing from a one-off laboratory result into evidence that supports production decisions.
Collaboration is valuable because drivetrain dynamics, lubrication, structural response, metrology and manufacturing variation overlap. The project partner information provides a starting point for understanding the disciplines and organisations involved. A shared test protocol reduces disputes over boundary conditions and makes comparisons between models, prototypes and hardware more credible.
Validation for Australian operations
Australian operating conditions make local validation valuable. Brisbane is an important hub for regional aircraft maintenance and charter operations, while Melbourne has a dense aviation engineering, manufacturing and supplier network. Teams can draw on that local capability when adapting an efficiency method to Australian aircraft, maintenance organisations and test facilities.
Long distances between maintenance centres and the use of fly-in, fly-out personnel mean measurement procedures must be portable, repeatable and resistant to environmental variation. Hot inland conditions, dust and coastal salt exposure can affect seals, lubricants and sensor mounting even when the gearbox is operating within its certified temperature range. Test planning should account for warm-up, oil condition and the time available for repeat measurements.
CASA oversight adds another layer of discipline. An efficiency investigation may support design, maintenance or performance monitoring, but it does not replace approved procedures or authorised engineering decisions. Results should be traceable, uncertainty should be understood, and changes that affect aircraft configuration or continued airworthiness must follow the applicable approval pathway.
The practical value of input shaft torque measurement comes from combining it with speed, output evidence and a clear model of physical losses. It can then reveal where efficiency is lost, whether a design change works and whether manufacturing variation threatens the expected performance.
Review the OPTIMIZE Project videos, objectives and methodology, then establish a baseline efficiency map with calibrated torque and speed instrumentation. The next action is to define the gearbox boundary, agree on the operating points and test uncertainty, and connect the measured results directly to the next design decision.