Superyacht Machinery Vibration Measurement & Frequency Analysis: Accelerometers, FFT, Orders & Trending

Rotating and reciprocating machinery produces changing forces that cause bearings, casings, foundations and surrounding structure to vibrate. HBK describes vibration measurement as a valuable machine condition indicator because changes in the response can reveal developing mechanical problems. On a yacht, the measured response may contain contributions from the engine or motor itself, coupled machinery, shafting, gears, propellers, pipework and the structure supporting them, so diagnosis requires more than one amplitude reading.

A vibration result has meaning only when the conditions under which it was collected are understood. Machinery speed, load, temperature, clutch state, generator load and other operating variables can change the forcing functions acting on the machine. Measurements intended for trending should therefore be collected at repeatable operating points wherever practical. A change caused solely by testing at a different RPM should not be mistaken automatically for deterioration of the machinery.

Piezoelectric accelerometers are widely used for machinery vibration measurement. HBK describes them as sensors whose output is proportional to acceleration and which contain no moving components that wear in normal measurement service. Sensor selection still has to suit the expected frequency range, amplitude and environment. The measurement is the motion at the sensor location, so a well-chosen transducer cannot compensate for an unsuitable or inconsistent measurement point.

A measured acceleration signal can be processed to obtain velocity or displacement information where the measurement chain and frequency range permit. HBK notes that accelerometer output can be integrated for velocity and displacement representations. Each parameter emphasises vibration differently across frequency, so the selected quantity should match the diagnostic purpose and any applicable manufacturer or measurement criterion rather than being chosen simply because one produces the largest numerical value.

HBK describes a single wide-frequency-band vibration measurement as a useful quick-look indicator for general machine condition. Such an overall value is valuable for trending because it can show that the machine response has changed. It does not, however, identify which frequency component caused that change. A rising overall level should therefore trigger further investigation rather than an immediate diagnosis based on amplitude alone.

For diagnosis, HBK emphasises frequency analysis because a complex vibration signal can contain many simultaneous forcing components. FFT-based analysis represents the measured response as amplitude versus frequency, allowing significant peaks to be examined individually. Frequencies associated with rotating shafts, gear meshing and other periodic machinery events can then be compared with the measured spectrum. This transforms vibration from a general severity indicator into evidence that can help localise the source.

For rotating equipment, shaft speed establishes a fundamental rotational frequency. A machine running at a known RPM therefore provides a reference against which spectral components can be compared. HBK identifies shaft rotation speed as one of the forcing functions that can appear directly in a vibration spectrum. A peak associated with rotational frequency is useful evidence, but its presence alone does not prove a particular fault because several mechanical conditions can produce response at the same order.

HBK notes that machinery spectra commonly contain harmonics at multiples of fundamental frequencies, particularly where the original motion is not purely sinusoidal. The number, amplitude and change in these harmonics can provide useful diagnostic information. They should be interpreted together with machine construction, operating condition and historical data rather than by assigning a universal fault to every second, third or higher harmonic seen in a spectrum.

Meshing gear teeth create periodic excitation related to tooth engagement. HBK identifies gear-tooth mesh frequency and its harmonics as common spectral components and explains that sidebands can appear around them because of periodic variation such as eccentricity. Gear wear, loading and geometry can influence these components. Trending the gear-mesh region against a healthy reference can therefore reveal developing changes that may not be obvious from overall vibration alone.

Frequency in hertz changes when rotating machinery changes RPM, which can make run-up and run-down data difficult to compare directly. HBK describes order analysis as relating vibration components to the fundamental rotational rate, using a tachometer or other speed reference. A first order remains associated with one event per revolution while higher orders follow multiples of that rate. This allows speed-related machinery excitation to be distinguished from frequencies that remain fixed as RPM changes.

Every machine element and supporting structure has dynamic characteristics that can include natural frequencies. HBK notes that vibration can become much larger when an excitation component coincides with a structural resonance, potentially increasing transmitted force and noise. The underlying forcing source may therefore be modest while the measured response is severe. Persistent vibration confined to a narrow speed range should prompt consideration of resonance as well as rotating-component condition.

HBK describes the spectrum of machinery in normal running condition as a reference signature against which future measurements can be compared. This is especially valuable on a yacht because identical machines can exhibit different installed responses due to foundations, connected pipework and local structure. Establishing a baseline after commissioning or a verified repair allows later changes in individual frequencies and overall vibration to be judged against the actual installed machinery rather than memory.

Vibration is directional and can vary substantially across a bearing housing or structure. A trend is therefore weakened if successive measurements are collected from different locations, axes or mounting conditions. Record the measurement point, orientation, sensor and operating condition so that future data can be reproduced. When comparing port and starboard machinery, use equivalent points where possible, but do not assume their numerical values must be identical when installation geometry or loading differs.

HBK's order-analysis guidance shows how vibration can be examined while machinery speed changes. Components linked to rotation move with RPM, whereas structural resonances remain at approximately fixed frequency. A controlled run-up or run-down can therefore expose critical-speed regions and help separate a rotating forcing function from the structure that amplifies it. Such testing should remain within the machinery manufacturer's authorised operating envelope and should not require deliberately crossing prohibited or known damaging speed ranges.

Begin with the symptom and the exact operating condition at which it occurs. Confirm machinery speed, load and recent maintenance history, then obtain repeatable measurements at defined bearing, casing or structural points. Compare overall vibration with the established healthy baseline and use frequency analysis to identify which components have changed. Relate significant frequencies to shaft rotation, gear mesh and other known periodic events without diagnosing a fault from one spectral line alone. Where speed varies, use RPM-referenced order or run-up analysis to separate rotating excitation from fixed structural resonance. Inspect alignment, bearings, gears, foundations and connected structure according to the evidence obtained. Correct only the confirmed fault, then repeat the same measurements under the same operating condition and retain the resulting spectrum and overall values as the new verified baseline.