Vibration measurement and condition monitoring
Measuring displacement, velocity and acceleration, seismic vibrometers and accelerometers, sampling and FFT, typical fault signatures and how vibration fits into condition monitoring, with worked numericals.
Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
Most rotating-machine faults, such as unbalance, misalignment, looseness, bearing wear and gear damage, show up in the vibration signature weeks or months before a breakdown. Measuring vibration and reading its spectrum is the core of predictive maintenance in plants, wind farms and fleets, and it is a natural mechatronics job: sensor selection, signal conditioning, sampling, FFT and decision rules.
Key ideas
What is measured. For harmonic motion x = X·sin ωt, velocity and acceleration have amplitudes ωX and ω²X. So at low frequency displacement is large and acceleration small, and at high frequency the reverse. Hence:
- displacement (µm) for low-frequency and shaft-motion problems (below about 10 Hz, and journal-bearing shafts);
- velocity (mm/s, usually RMS) for overall machine severity, roughly 10 Hz to 1 kHz, the basis of ISO severity charts;
- acceleration (m/s² or g) for high-frequency faults such as rolling bearings and gears. Amplitudes are quoted as peak, peak-to-peak or RMS; for a pure sinusoid RMS = peak/√2.
Seismic (inertial) instruments. A small mass on a spring and damper inside a case fixed to the vibrating body. The measured quantity is the relative displacement Z between mass and case. With r = ω/ω_n:
- Vibrometer (displacement pickup): very soft, low ω_n, used at r ≫ 1. The mass stays nearly still in space, so Z ≈ Y (the base displacement). Large and heavy.
- Accelerometer: very stiff, high ω_n, used at r ≪ 1. Then Z ≈ a/ω_n², proportional to base acceleration. With ζ ≈ 0.65–0.7 the amplitude error stays small up to about r = 0.2–0.4 with nearly linear phase.
- Modern sensors: piezoelectric accelerometers (the crystal is the spring; wide range, a few Hz to over 10 kHz; mounting method limits the usable top frequency), MEMS accelerometers (low cost, down to DC), velocity pickups (coil and magnet, self-generating), non-contact eddy-current proximity probes (shaft displacement relative to the bearing), and laser vibrometers.
Frequency-measuring instruments. Frahm reed tachometer (a set of reeds of graded natural frequency; the reed in resonance shows the frequency) and the Fullarton (variable-length reed) tachometer, classic examples of using resonance.
Signal processing. Sample above twice the highest frequency of interest (Nyquist), with an anti-aliasing filter before the ADC. The FFT of N samples at f_s gives lines spaced Δf = f_s/N. Windowing (e.g. Hanning) reduces leakage; averaging reduces noise. Envelope (demodulation) analysis extracts repetitive impacts from bearings out of high-frequency resonances. Order tracking follows the running speed in variable-speed machines.
Fault signatures (typical).
- Unbalance: high 1× (running speed) radial vibration; at each bearing the horizontal and vertical readings differ in phase by about 90°.
- Misalignment: high 2× and 1×, strong axial component, about 180° phase across the coupling.
- Mechanical looseness: many harmonics of running speed, sometimes ½× sub-harmonics.
- Rolling-element bearing defects: non-integer frequencies (BPFO, BPFI, ball spin, cage), often seen in the envelope spectrum.
- Gears: gear-mesh frequency (teeth × shaft speed) with sidebands spaced at shaft speed.
- Resonance: a high peak that changes greatly with small speed changes.
Condition monitoring programme. Baseline measurement, regular trending at fixed points and directions, alarm levels (from ISO 10816/20816 severity zones or the manufacturer; take values from the standard), diagnosis by spectrum and phase, then planned repair. Vibration is combined with oil and wear-debris analysis, infrared thermography, airborne ultrasound, motor current signature analysis and performance monitoring.
Formulas
v_peak = ω·X ; a_peak = ω²·X ; ω = 2πf (X in m, v in m/s, a in m/s²)
RMS = peak / √2 (sinusoid) ; peak-to-peak = 2 × peak
Seismic instrument, relative to base motion Y:
Z / Y = r² / √((1 − r²)² + (2ζr)²) (vibrometer reads Z ≈ Y for r ≫ 1)
Z·ω_n² / (ω²·Y) = 1 / √((1 − r²)² + (2ζr)²) (accelerometer reads Z·ω_n² ≈ a for r ≪ 1)
f_s > 2·f_max (Nyquist) ; Δf = f_s / N
Bearing outer- and inner-race defect frequencies (n balls, ball diameter d, pitch diameter D, contact angle α, shaft frequency f_r):
BPFO = (n/2)·f_r·(1 − (d/D)·cos α) ; BPFI = (n/2)·f_r·(1 + (d/D)·cos α)
Gear mesh frequency = T × f_shaft
Worked examples
Example 1 (standard): from a velocity reading to displacement, acceleration and a bearing frequency. Given: a motor runs at 1480 rpm; vibration velocity at running speed is 4.5 mm/s RMS. Its bearing has 9 balls of 7.94 mm on a 39 mm pitch diameter, α = 0°.
f_r = 1480/60 = 24.67 Hz→ω = 2π × 24.67 = 155.0 rad/s.v_peak = 4.5 × √2 = 6.36 mm/s.X_peak = v_peak/ω = 6.36/155.0 = 0.0411 mm = 41 µm(82 µm peak-to-peak).a_peak = ω·v_peak = 155.0 × 6.36 × 10⁻³ = 0.986 m/s².BPFO = (9/2) × 24.67 × (1 − 7.94/39) = 111.0 × 0.7964 = 88.4 Hz;BPFI = 111.0 × 1.2036 = 133.6 Hz. Answer: 41 µm peak, 0.99 m/s² peak; watch for 88.4 Hz (outer race) and 133.6 Hz (inner race) in the envelope spectrum.
Example 2 (GATE level): seismic instrument errors. (a) A vibrometer with f_n = 4 Hz and ζ = 0.2 reads a relative amplitude of 1.0 mm on a machine vibrating at 20 Hz. True amplitude?
- r = 20/4 = 5.
Z/Y = 25 / √((1 − 25)² + (2 × 0.2 × 5)²) = 25 / √(576 + 4) = 25 / 24.08 = 1.038Y = 1.0 / 1.038 = 0.963 mm. (b) An accelerometer with f_n = 2 kHz and ζ = 0.7 measures a 200 Hz vibration. Amplitude error?- r = 0.1 →
1 / √((1 − 0.01)² + (0.14)²) = 1 / √(0.9801 + 0.0196) = 1.00015. Answer: (a) true amplitude ≈ 0.963 mm (the reading is 3.8% high); (b) error ≈ +0.015%, negligible.
Common mistakes
- Mixing peak, peak-to-peak and RMS values when comparing with alarm limits.
- Using a vibrometer below its natural frequency or an accelerometer near its resonance.
- Forgetting ω = 2πf when converting between displacement, velocity and acceleration.
- Sampling too slowly, so high-frequency content aliases into false low-frequency peaks.
- Diagnosing unbalance from a 1× peak alone without checking phase; misalignment and resonance can also give 1×.
- Mounting an accelerometer on a magnet or thin bracket and trusting readings near the mount resonance.
For GATE ME
Questions here are usually about seismic instruments: the frequency-ratio condition for vibrometers and accelerometers, the amplitude ratio formula, and simple conversions between displacement, velocity and acceleration amplitudes. Practise the Z/Y expression for r ≫ 1 and r ≪ 1 and the role of ζ ≈ 0.7.
Quick check
- A vibration of 10 µm peak at 100 Hz. Peak velocity?
- Which instrument needs a natural frequency well above the measured frequency?
- RMS of a sinusoid with 2 mm/s peak?
- Signals up to 5 kHz are needed. Minimum sampling rate?
- A 40-tooth gear on a 25 Hz shaft. Gear mesh frequency?
Answers: 1. 2π × 100 × 10 × 10⁻⁶ = 6.28 mm/s. 2. Accelerometer. 3. 1.41 mm/s. 4. Above 10 kHz (in practice about 2.56 × 5 kHz = 12.8 kHz). 5. 1000 Hz.
Interview questions
All Theory of Machines and Vibrations interview questionsTry answering each one aloud before you open it.
1.What is vibration measurement in the context of mechanical systems?Concept
Vibration measurement refers to the process of quantifying the oscillations or movements of a mechanical system. It involves using sensors and instruments to capture data on the amplitude, frequency, and phase of vibrations. This data helps in understanding the dynamic behavior of the system and is crucial for diagnosing potential issues.
2.Explain the importance of condition monitoring in mechanical systems.Concept
Condition monitoring is essential for maintaining the health and efficiency of mechanical systems. It involves regularly checking the system's performance and identifying any deviations from normal operation. This proactive approach helps in early detection of faults, reducing downtime, and preventing catastrophic failures, ultimately saving costs and improving safety.
3.What are the common methods used for vibration measurement?Concept
The usual transducers are accelerometers (piezoelectric or MEMS), velocity pickups and displacement sensors such as non-contact eddy-current proximity probes or laser vibrometers. Because velocity is ω times displacement and acceleration ω² times displacement, displacement suits low frequencies and shaft motion in journal bearings, velocity suits overall machine severity from about 10 Hz to 1 kHz, and acceleration suits high-frequency faults in rolling bearings and gears. Piezoelectric accelerometers are the general workhorse because they are small, rugged and cover a few Hz to over 10 kHz.
4.Why are accelerometers widely used in vibration measurement?Application
Accelerometers are widely used because they are versatile and can measure a wide range of frequencies with high accuracy. They are compact, easy to install, and can provide real-time data, making them suitable for various applications in condition monitoring and fault diagnosis.
5.What happens if vibration levels in a machine exceed the recommended limits?Application
If vibration levels exceed recommended limits, it can lead to increased wear and tear, reduced efficiency, and potential failure of machine components. Excessive vibrations can cause misalignment, loosen fasteners, and lead to fatigue in materials, ultimately resulting in costly repairs and downtime.
6.How does condition monitoring contribute to predictive maintenance?Application
Condition monitoring provides real-time data on the health of a machine, allowing for the prediction of potential failures before they occur. By analyzing trends and patterns in the data, maintenance can be scheduled proactively, reducing unexpected breakdowns and optimizing maintenance resources.
7.Explain the role of frequency analysis in vibration measurement.Concept
Frequency analysis involves breaking down vibration signals into their constituent frequencies using techniques like Fast Fourier Transform (FFT). This helps in identifying specific vibration patterns associated with different faults, such as imbalance, misalignment, or bearing defects, enabling targeted maintenance actions.
8.Calculate the natural frequency of a simple pendulum with a length of 2 meters.Numerical
The natural frequency (f) of a simple pendulum is given by the formula f = (1/2π) * √(g/L), where g is the acceleration due to gravity (9.81 m/s²) and L is the length of the pendulum. Substituting the values, f = (1/2π) * √(9.81/2) ≈ 0.35 Hz.
9.A machine exhibits a vibration amplitude of 0.01 m at a frequency of 50 Hz. Calculate the velocity of the vibration.Numerical
The velocity (v) of vibration can be calculated using the formula v = 2πfA, where f is the frequency and A is the amplitude. Substituting the values, v = 2π * 50 * 0.01 = 3.14 m/s.
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?