Industrial Vibration Analysis: Techniques

What Is Industrial Vibration Analysis?

Industrial vibration analysis is the process of collecting vibration signals from rotating machinery, converting them into frequency and amplitude data, and interpreting that data to identify the mechanical condition of the asset. Every rotating machine produces a vibration signature: a combination of frequencies tied to its operating speed, component geometry, and mechanical health.

When a fault develops, the signature changes. A bearing defect creates impulses at specific frequencies determined by the bearing geometry. Imbalance amplifies vibration at the shaft running speed. Misalignment introduces harmonics at twice or three times running speed. By recognizing these patterns, engineers can identify the fault type, its severity, and how urgently it needs attention.

This diagnostic power is what sets vibration analysis apart from other condition monitoring techniques. It is not just a health indicator; it is a root cause identification tool. Combined with vibration monitoring, it forms the backbone of any serious predictive maintenance program for rotating equipment.

How Vibration Analysis Works

The analysis process starts with data collection. A vibration sensor (typically a piezoelectric accelerometer) is mounted on or near the bearing housing of the machine. The sensor converts mechanical motion into an electrical signal proportional to acceleration.

That raw signal is a time waveform: a record of vibration amplitude plotted against time. While the time waveform carries useful information, its complexity makes pattern recognition difficult when viewed alone. The signal is therefore transformed using Fast Fourier Transform (FFT), which decomposes the time signal into its constituent frequencies.

The result is a frequency spectrum, sometimes called a vibration spectrum or FFT plot. Each peak in the spectrum corresponds to a specific frequency component. By comparing peak locations and amplitudes against known fault frequencies for that machine, analysts identify which components are generating elevated vibration and why.

Modern systems automate much of this process. Continuous monitoring platforms collect data around the clock, apply automated fault detection algorithms, and alert maintenance teams when a signature changes beyond defined thresholds.

Vibration Analysis Techniques

Overall Vibration Level

The simplest measurement is overall vibration level, expressed as velocity (mm/s RMS) or acceleration (g RMS). It sums energy across all frequencies into a single number. Overall level is useful for trending: if it rises steadily over weeks or months, the machine is degrading. However, it does not identify which fault is causing the rise.

FFT Spectrum Analysis

FFT spectrum analysis is the core diagnostic tool. The spectrum is examined for peaks at fault-specific frequencies: 1x running speed for imbalance, 2x for misalignment, ball-pass frequencies for bearing defects, and gear mesh frequency for gear faults. Elevated peaks at these locations confirm which fault is present and indicate its severity by amplitude.

Time Waveform Analysis

The raw time waveform reveals patterns that the FFT can obscure. Repetitive impacts from a chipped gear tooth or bearing defect appear as regular spikes in the waveform. High crest factor values (ratio of peak to RMS amplitude) indicate impulsive faults, often in early-stage bearing damage before the FFT peak becomes obvious.

Envelope Analysis (Demodulation)

Envelope analysis isolates high-frequency impulses generated by bearing defects in their early stages. The technique bandpass-filters the signal around a resonance frequency, rectifies it, and applies a second FFT to extract bearing fault frequencies that would otherwise be buried in broadband noise. It is particularly effective for detecting incipient bearing damage.

Resonance Analysis

Every machine structure has natural frequencies at which it vibrates when excited. If an operating frequency (such as running speed or a harmonic) coincides with a natural frequency, resonance amplifies vibration dramatically. Resonance testing, using bump tests or run-up/coast-down sweeps, identifies these frequencies so structural modifications or speed changes can avoid them.

Phase Analysis

Phase measurements compare the timing of vibration at different points on a machine. Phase data distinguishes between fault types that produce similar spectra: imbalance and misalignment both produce elevated 1x and 2x peaks but show different phase relationships between measurement points. Phase analysis is especially valuable during precision balancing and alignment verification.

Common Faults Detected by Vibration Analysis

Vibration analysis can identify a wide range of faults in rotating equipment. Each fault type produces a characteristic signature that trained analysts and automated systems can recognize.

Vibration Analysis vs. Other Condition Monitoring Methods

Vibration analysis is the most versatile condition monitoring technique for mechanical faults in rotating equipment, but it works best as part of a broader monitoring program that includes complementary methods.

Vibration analysis detects mechanical faults earlier and with more specificity than most alternatives. Thermography excels at electrical faults that produce no significant vibration change. Oil analysis catches wear that occurs too slowly or at too low a frequency to appear in the vibration spectrum. The strongest maintenance programs use all four methods together, each covering the blind spots of the others.

How to Implement Vibration Analysis in Your Plant

A successful implementation follows a structured process. Rushing any step reduces the quality of the data and the reliability of the diagnoses.

Step 1: Identify Critical Rotating Equipment

Start with a criticality ranking of all rotating assets: motors, pumps, fans, compressors, and gearboxes. Focus initial monitoring resources on assets where failure causes the most production loss, safety risk, or repair cost. A criticality matrix typically considers failure consequence, failure frequency, and replacement lead time.

Step 2: Select the Right Measurement Approach

Choose between continuous monitoring with permanently mounted sensors and periodic route-based measurement with portable analyzers. Critical assets that run continuously benefit most from always-on monitoring. Less critical or intermittently operating assets are well-served by monthly or quarterly routes.

Step 3: Define Measurement Points and Parameters

Establish standard measurement points on each machine: typically the drive-end and non-drive-end bearing housings in horizontal and vertical (radial) directions, plus axial measurements. Document speed, load, and operating conditions for each measurement so that spectra can be compared under consistent conditions.

Step 4: Establish Baselines

Collect baseline measurements on healthy, newly commissioned machines. These baselines define the normal vibration signature for each asset and serve as the reference point for all future comparisons. Without baselines, trending is unreliable and fault identification is more difficult.

Step 5: Set Alert and Alarm Thresholds

Configure overall-level thresholds using published severity standards (ISO 10816 or ISO 20816) and band-specific thresholds for individual fault frequencies. Alert levels trigger investigation; alarm levels trigger urgent action. Over-sensitive thresholds generate nuisance alerts; overly lenient thresholds allow faults to progress.

Step 6: Analyze, Diagnose, and Act

When a threshold is exceeded or a trend deviates from baseline, a qualified analyst reviews the spectrum and waveform data to confirm the fault type and estimate severity. The diagnosis feeds into the maintenance planning process, generating a work order for the appropriate repair action at the optimal time.

Vibration Analysis and Predictive Maintenance

Vibration analysis is the primary enabler of predictive maintenance for rotating equipment. The P-F curve concept underpins the entire approach: there is a detectable period between the point where a fault becomes measurable (P, the potential failure point) and the point where functional failure occurs (F). Vibration analysis detects faults in that window.

A bearing defect, for example, may first appear in the vibration spectrum months before it causes a failure. An imbalance condition may be introduced after a repair and detected immediately on the next measurement route. In both cases, the maintenance team gains the lead time they need to plan the repair during a scheduled production window instead of responding to an emergency at 2 a.m.

This capability directly improves mean time between failure. Faults caught early are repaired before they cascade: a failing bearing that is replaced promptly does not damage the shaft, housing, or adjacent components. A bearing allowed to run to destruction can require a full machine rebuild costing five to ten times the original repair.

Condition-based maintenance decisions driven by vibration data also reduce unnecessary preventive maintenance. Instead of replacing bearings on a calendar schedule regardless of their condition, teams replace them when the data shows they need it. This extends component life, reduces parts consumption, and frees maintenance labor for higher-value work.

In the manufacturing industry, where rotating equipment downtime halts production lines, vibration analysis is one of the fastest-payback maintenance investments available. Plants that implement continuous vibration monitoring on critical assets typically achieve significant reductions in unplanned downtime within the first year of deployment.

The Bottom Line

Industrial vibration analysis is one of the highest-value condition monitoring techniques available to maintenance teams because rotating equipment is both ubiquitous in industrial facilities and predictably progressive in its failure signatures. Vibration patterns provide early, specific warning of bearing wear, misalignment, imbalance, and structural looseness — typically weeks or months before functional failure, when the cost of planned repair is a fraction of the cost of an unplanned breakdown.

The transition from periodic manual vibration surveys to continuous sensor-based monitoring significantly extends the warning window and eliminates the risk of a fault developing between survey visits. Facilities that deploy continuous vibration monitoring on critical rotating equipment can detect and act on developing faults regardless of when they occur, converting the reactive urgency of unexpected failures into the controlled efficiency of planned maintenance interventions.

Frequently Asked Questions