Vibration Analysis for Predictive Maintenance: An Industrial Guide

Why Do Machines Vibrate? #

Every machine vibrates. But when that vibration changes, the machine is "talking" to you. The problem is that most plants run in run-to-failure mode or do blind preventive maintenance (swapping parts "just in case").

Vibration analysis is the cornerstone of predictive maintenance (PdM). It lets you detect:

• Shaft and pulley misalignment.
• Mass imbalance.
• Bearing faults (inner race, outer race, cage).
• Mechanical looseness (soft foot).

Level 1: Overall Monitoring (ISO 10816) #

This is the machine's "blood pressure check." You measure total vibration (RMS) in mm/s (velocity).

The ISO 10816 standard defines clear limits:

Level 2: Spectral Diagnosis (FFT) #

Knowing THAT something vibrates is not enough. You need to know WHY. This is where the Fast Fourier Transform (FFT) comes in: it breaks the complex waveform down into its component frequencies.

The Fault Signature:

• 1x RPM: the fundamental frequency. Usually points to imbalance.
• 2x RPM: usually points to misalignment.
• High frequencies (non-synchronous): point to a bearing fault (a ball rolling over a crack).

Sensors: MEMS or Piezoelectric? #

Ten years ago a vibration sensor was expensive (an industrial piezoelectric unit). Today, thanks to smartphones, there are MEMS (Micro-Electro-Mechanical Systems) sensors that cost a fraction of that.

Are they the same? No. MEMS sensors have a higher noise floor and lower bandwidth (up to 2-4 kHz). But for 90% of industrial pumps and fans, a MEMS sensor tracking ISO 10816 is infinitely better than "putting your hand on it" or "listening with a screwdriver."