Every gearbox has a heartbeat: the frequency at which its teeth mesh, one tooth engagement after another as the gears turn. Gear mesh frequency analysis reads that heartbeat and its surrounding structure in the vibration spectrum to judge the health of the gear set. The mesh frequency itself, its harmonics, and the sidebands spaced around it carry specific information, whether the gears are worn overall, whether one is eccentric, or whether a single tooth is cracked. It is the primary spectral method for diagnosing gearboxes before a tooth lets go.
Gear mesh frequency in one line: Gear mesh frequency analysis examines the vibration frequency at which gear teeth engage, which equals the number of teeth on a gear multiplied by that gear's shaft speed. The mesh frequency and its harmonics indicate general gear condition, while sidebands spaced at shaft speed around the mesh frequency reveal eccentricity, a cracked tooth, or other localized faults.
The gear mesh frequency, or GMF, is the rate at which teeth come into contact, and it is calculated as the number of teeth on a gear times the rotational speed of that gear's shaft. Because both meshing gears share the same contact events, the GMF is identical whether you compute it from the pinion's tooth count and speed or from the gear's, the higher tooth count on the slower shaft equals the lower tooth count on the faster shaft. That single frequency, typically well above running speed, is the dominant gear-related feature in the vibration spectrum and the anchor for the whole analysis.
The GMF is almost always present to some degree, even on a healthy gear set, because real teeth are never perfectly formed and some vibration at the mesh rate is normal. What matters diagnostically is how its amplitude changes over time and how many harmonics appear. As gears wear across all their teeth, the GMF amplitude tends to rise and its harmonics, two and three times the mesh frequency, grow more prominent. A gear set that develops strong second and third harmonics of the mesh frequency is showing general wear or a deteriorating tooth contact pattern, and trending the amplitude at GMF and these harmonics is the baseline of gearbox condition monitoring.
A subtle related feature is the hunting tooth frequency, a very low frequency that reflects how often a specific pinion tooth re-engages the same gear tooth. It is set by the tooth counts and their common factors and is important because a fault involving one tooth on each gear repeats only at that slow rate. It rarely appears strongly unless there is damage on mating teeth, but when it does it points to a specific wear or fault condition on paired teeth. The main analysis, though, centers on the GMF, its harmonics, and the sidebands around them.
The most informative part of a gear spectrum is often not the mesh frequency itself but the sidebands, the smaller peaks spaced evenly on either side of the GMF. Sidebands arise from modulation: something is varying the mesh vibration once per revolution of one of the shafts. The spacing between the sidebands identifies which gear is responsible, because the spacing equals that gear's shaft speed. Sidebands spaced at the pinion shaft speed implicate the pinion; sidebands spaced at the gear shaft speed implicate the gear. Reading the sideband spacing is how an analyst assigns a fault to a specific gear in the set.
The nature of the sidebands separates the faults. Eccentricity, a gear mounted off-center or with a bent shaft, modulates the mesh smoothly once per revolution and produces a symmetric family of sidebands around the GMF. A single localized defect, most importantly a cracked or chipped tooth, produces a sharp impact once every time that tooth engages, which spreads energy into many sidebands and often shows more strongly in the time waveform as a periodic spike at the shaft rate. The distinction matters because a cracked tooth can fail suddenly and catastrophically, while distributed eccentricity is a wear condition that can be planned around.
Growth in sideband amplitude, and the appearance of a wider spread of sidebands, is one of the clearer indicators that a localized gear fault is developing rather than general wear. A gearbox that previously showed only the mesh frequency and now shows a growing skirt of sidebands, or a periodic impact in the waveform at one shaft's rate, is signaling that a single tooth is failing. Catching that pattern early is the whole point, because it is the difference between replacing a gear on schedule and dealing with a gearbox that has shed a tooth and destroyed itself.
Gear mesh analysis lends itself to automated trending because its key features live at predictable frequencies. Once the tooth counts and shaft speeds are known, the GMF and its harmonics are fixed multiples that a monitoring system can compute and track, and the sideband regions around them can be watched for growth. A monitoring platform can define bands at the mesh frequency and its harmonics and trend their amplitudes, flagging the rising GMF and harmonic energy that signal wear, and the expanding sidebands that signal a localized tooth fault.
Bringing these values into a SCADA platform puts gearbox condition on the same dashboard as the rest of the machine, which matters for the pumping-unit reducers, compressor gearboxes, and drive trains scattered across oil and gas sites. Merobix can trend the mesh-frequency and harmonic amplitudes for each monitored gearbox alongside overall vibration and process load, so an operator sees not just that a gearbox is vibrating more but that the increase is concentrated at the mesh frequency and its sidebands, pointing straight at gear wear or a tooth defect. For remote and unmanned locations, that remote gear diagnosis is what makes it feasible to watch gearboxes that no analyst visits regularly.
The historized trend is what converts a spectrum into a maintenance decision. A gearbox's mesh-frequency amplitude and sideband content have a normal baseline, and a slow climb in the GMF or a widening sideband family over successive readings is the signature of progressing wear or a developing cracked tooth. By trending those features over weeks and months and correlating them with load, the dashboard distinguishes a genuine developing fault from load-driven variation and gives maintenance the lead time to schedule a gear change. Because a shed tooth can wreck a gearbox quickly, that early trend-based warning is especially valuable for gear sets.
Gear mesh frequency equals the number of teeth on a gear multiplied by that gear's shaft rotational speed. Because both meshing gears share the same tooth-contact events, you get the same value from either gear, since the higher tooth count on the slower shaft matches the lower tooth count on the faster shaft. It is the dominant gear-related frequency in the vibration spectrum and the anchor for gear analysis.
Sidebands are smaller peaks spaced evenly around the mesh frequency, caused by something modulating the mesh vibration once per shaft revolution. Their spacing equals a gear's shaft speed, which identifies the responsible gear. Symmetric sidebands point to eccentricity or a bent shaft, while a wide spread of sidebands, often with a periodic impact in the time waveform, points to a localized fault such as a cracked tooth.
A cracked or chipped tooth produces a sharp impact each time it engages, spreading energy into many sidebands around the mesh frequency and often showing a periodic spike in the time waveform at the shaft rate. Trending the mesh frequency and its sidebands reveals this pattern emerging and growing before the tooth fails. Because a shed tooth can destroy a gearbox quickly, that early spectral warning provides the lead time to plan a gear change.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.