Spectrum Analysis
Spectrum analysis is a vibration-based condition monitoring technique that converts raw time-domain vibration data into the frequency domain using the Fast Fourier Transform (FFT) algorithm to identify specific fault signatures by amplitude at discrete frequencies. Each mechanical defect, such as unbalance, misalignment, or bearing wear, generates vibration at a unique defect frequency proportional to machine speed and component geometry, creating a distinct spectral signature.
On the shop floor, analysts collect vibration data via sensors, typically accelerometers, mounted on horizontal, vertical, and axial axes. They use CMMS-integrated spectrum analysis tools to compare current peaks against baseline measurements and historical trends. This technique diagnoses rotating equipment faults by correlating dominant frequency peaks to physical failure mechanisms, such as identifying 1× RPM peaks for unbalance or 2× RPM for misalignment, enabling early intervention before faults become audible or catastrophic.
How does FFT enable fault differentiation in spectrum analysis?
FFT decomposes the time-domain waveform into discrete frequency components, isolating energy at defect frequencies such as ball pass frequency for bearings, which time-domain analysis cannot resolve.
Why is log-scale amplitude preferred in spectrum plots?
A log scale improves dynamic range, allowing detection of low-amplitude defects like early bearing cracks alongside high-amplitude signals without compressing critical data.
What confirms a peak represents a true fault vs. noise?
Verification requires correlating the peak frequency to machine parameters like RPM or gear ratio, checking for harmonics and sidebands, and comparing against baseline trends across multiple measurement directions.