Gear-noise: diagnosis & control

A: Chasing the wrong lever here burns time and risks louder part-load whine that customers actually hear. Reducing crowning can spike mesh stiffness variation at light loads, pushing static TE up even if full-load contact looks pretty. Rolling back some crowning and tuning at the dominant customer torque avoids amplifying the excitation instead of trying to bury it with bearings or housing hacks.

A: Picking the wrong mechanism sends you after filtration or washing fixes while the noise keeps climbing. Humid service with acceptable oil points to micro-pitting on the flanks when case depth or RA isn’t right, which roughens the surface and raises TE over mileage rather than causing sudden failures.

A: Skipping the signal chain or order resolution fails the test and guarantees arguments later. You need correlated order data under load, which means tach quality first, controlled torque second, and only then any subjective drive to anchor perception to physics.

A: Wrong choices here lock in noise you can’t tune out later or blow packaging. Raising overlap and matching microgeometry cuts excitation at the source, while grade alone or blunt stiffness changes don’t guarantee lower TE in the customer band.