Most rotating equipment failures do not happen suddenly. Bearings deteriorate over weeks and months. Misalignment causes progressive damage before catastrophic breakdown. Imbalance worsens as operating conditions change. The gradual nature of these failure processes means there is almost always a window of opportunity for detection and intervention – if the right monitoring is in place.

Vibration analysis exploits this window. By measuring and interpreting the vibration patterns generated by rotating equipment, maintenance teams can identify developing faults while machines are still operating normally. This capability is the foundation of any serious predictive maintenance program, and it shifts maintenance from a reactive, crisis-driven activity to a planned, data-driven one.

This guide explains how to integrate professional vibration analysis into existing maintenance workflows, from establishing measurement baselines through to connecting diagnostic findings to maintenance management systems and building the analyst capability to use this approach effectively.

Vibration Analysis Fundamentals

Vibration analysis measures the mechanical oscillations generated by rotating equipment and interprets those oscillations to assess machine health. Every rotating component generates vibration as it operates. When components function normally, their vibration signatures are stable and characteristic. When they wear, misalign, or become unbalanced, their vibration signatures change in ways that are identifiable and interpretable.

Accelerometers are the primary sensing technology. Mounted on bearing housings, they convert mechanical vibration into proportional electrical signals. These signals are processed by vibration analysers into formats that reveal the nature and severity of any developing faults.

FFT (Fast Fourier Transform) analysis is the core mathematical process that makes vibration analysis effective. It decomposes complex vibration waveforms – which contain contributions from many different mechanical events happening simultaneously – into individual frequency components. The result is a frequency spectrum showing the amplitude of vibration at each frequency, making each fault’s contribution visible and measurable independently.

Key frequency ranges that machine health diagnostics relies on include:

  • Rotational speed frequencies (1X running speed) – Indicate shaft-related problems including imbalance and bend
  • Bearing defect frequencies – Calculated from bearing geometry; reveal inner race, outer race, ball, and cage defects
  • Gear mesh frequencies – Running speed multiplied by tooth count; indicate gear tooth condition
  • Sub-synchronous frequencies (below 1X) – Often indicate instability, looseness, or rub conditions
  • High-frequency bands – Detect early bearing deterioration before lower-frequency symptoms appear

Understanding these frequency relationships allows analysts to connect measured vibration patterns to specific physical causes in the equipment.

Establishing Baseline Measurements

Effective vibration analysis requires baseline measurements taken when equipment is in known good condition. Without baselines, there is no reference for assessing whether current readings represent normal operation or developing deterioration.

The ideal time to collect baseline measurements is immediately after equipment installation or following a major overhaul. At these points, the machine is known to be in the best available condition. Measurements taken then represent what normal looks like for that specific machine in its specific installation.

A complete baseline record should include:

  • Overall vibration levels in velocity (mm/s RMS) at each measurement point
  • Vibration spectra showing the full frequency content at each measurement point
  • Time waveforms capturing the pattern of vibration over time
  • Operating conditions at the time of measurement – load, speed, temperature, and process parameters

Documenting operating conditions alongside vibration measurements is important because vibration levels vary with operating conditions. A measurement taken at partial load cannot be directly compared to one taken at full load without context. Recording conditions at measurement time allows meaningful comparisons regardless of when subsequent measurements are taken.

Store all baseline data in a centralised database linked to equipment asset numbers and measurement point identifiers. This creates the foundation of the machine history that makes long-term trending possible.

Designing a Route-Based Monitoring Program

Route-based monitoring organises data collection into efficient paths through the facility. Technicians follow defined routes, visiting each measurement point in sequence and collecting data at each location. Good route design balances coverage comprehensiveness against the time required for each route.

Route structure should group measurement points geographically to minimise travel between machines. Points within the same equipment train or area should appear sequentially on the route. This reduces travel time and ensures related equipment is measured under similar operating conditions.

Measurement frequency should reflect equipment criticality. A practical approach:

  • Critical equipment (failure halts production or creates safety risk): weekly or fortnightly measurement
  • Standard production equipment: monthly measurement
  • Utility and low-priority equipment: quarterly measurement

Measurement point standardisation is essential for data consistency. Mark exact measurement locations permanently on bearing housings using paint or engraving. Document sensor type, orientation (vertical, horizontal, or axial), and mounting method (magnet, adhesive, or probe) for each point. Different technicians measuring the same point will produce comparable data only if they follow identical procedures.

Portable vibration analysers with integrated route management software guide technicians through collection sequences, display historical trend data for comparison at the measurement point, and alert technicians when readings deviate significantly from baseline values – enabling immediate follow-up investigation rather than waiting for post-route analysis.

Online Monitoring for Critical Assets

Route-based monitoring serves most equipment populations effectively. But for the highest-consequence assets, collecting periodic snapshots is not sufficient. Continuous online monitoring provides the real-time fault detection that critical assets require.

Good candidates for online monitoring include equipment whose sudden failure would immediately halt production, assets in hazardous locations where frequent manual inspection creates unacceptable risk, machines with rapid fault progression characteristics where monthly measurements may miss developing faults entirely, and equipment whose failure creates significant safety risks.

Online condition monitoring systems use permanently mounted sensors to collect data continuously, feeding analysis software that runs FFT processing on incoming data in real time. When parameters cross defined thresholds, alerts are generated immediately.

Configure alarm thresholds in a three-tier structure that balances sensitivity against false alarm rate:

Alert level – Parameter trending toward the alarm zone. No immediate action required, but monitoring frequency should increase and a work order created for investigation at the next planned opportunity.

Alarm level – Parameter clearly elevated above normal. Schedule corrective maintenance within two to four weeks. Initiate parts planning.

Danger level – Parameter indicates imminent failure risk. Arrange urgent intervention or a controlled, planned shutdown before catastrophic failure occurs.

Setting initial thresholds based on ISO 20816 severity guidelines provides a defensible starting point. Refining thresholds based on individual machine baseline data and operational experience improves specificity over time.

Frequency Analysis and Fault Identification

Bearing Defect Frequencies

Rolling element bearings generate vibration at specific frequencies when surface defects develop on their races or rolling elements. These frequencies depend on bearing geometry (ball diameter, pitch circle diameter, number of rolling elements, and contact angle) and shaft speed.

Outer race defects produce vibration at Ball Pass Frequency Outer (BPFO) – typically three to five times shaft speed. Inner race defects generate Ball Pass Frequency Inner (BPFI), which is higher than BPFO and produces characteristic sidebands spaced at shaft speed. Ball defects appear at Ball Spin Frequency (BSF), with associated harmonics. Cage defects generate Fundamental Train Frequency (FTF), which is typically below one times shaft speed.

Modern vibration analysers calculate these frequencies automatically from bearing dimensions entered by the analyst. Comparing measured frequency peaks against calculated defect frequencies confirms bearing fault identification and allows severity progression to be tracked over time.

Misalignment Signatures

Shaft misalignment generates characteristic vibration patterns that distinguish it from other fault types. Parallel misalignment produces elevated vibration at twice running speed (2X) in radial directions. Angular misalignment creates high axial vibration at one and two times running speed.

The ratio of axial to radial vibration helps distinguish misalignment from other conditions. When axial vibration approaches or exceeds radial vibration amplitude, misalignment should be investigated. Phase measurements between bearing housings at the coupling can further confirm misalignment diagnosis and distinguish angular from parallel offset conditions.

Rotor Imbalance

Imbalance generates strong vibration at exactly one times running speed (1X) in radial directions. The characteristic property of imbalance is its response to speed change: imbalance-related vibration amplitude increases with the square of speed. Doubling shaft speed quadruples imbalance vibration.

This speed-sensitivity distinguishes imbalance from other faults generating 1X vibration. Phase measurement confirms imbalance diagnosis – imbalance produces consistent phase angles around the shaft that change predictably with speed.

Mechanical Looseness

Mechanical looseness – in bearing fits, base plate bolting, or structural connections – generates a pattern of multiple harmonics at running speed. Peaks at 1X, 2X, 3X, and 4X running speed appearing together are characteristic. The multi-peak pattern distinguishes looseness from single-frequency faults like imbalance.

Looseness in specific components generates slightly different patterns. Bearing housing looseness typically shows strong sub-harmonics at 0.5X running speed alongside the harmonic series. Foundation or base plate looseness often produces truncated waveforms in the time domain alongside the harmonic frequency pattern.

Vibration Severity Assessment and ISO Standards

Comparing measured vibration levels against established standards provides objective criteria for maintenance decision-making.

ISO 20816 (which supersedes ISO 10816) classifies machines into groups based on rated power and mounting configuration:

  • Group 1 – Small machines up to 15 kW
  • Group 2 – Medium machines 15-75 kW on rigid foundations
  • Group 3 – Large machines above 75 kW on rigid foundations
  • Group 4 – Large machines on flexible foundations

Each group is assessed against four severity zones:

  • Zone A – Condition typical of a newly commissioned machine
  • Zone B – Condition acceptable for unrestricted long-term operation
  • Zone C – Condition considered unsatisfactory for long-term operation; corrective action should be scheduled
  • Zone D – Condition sufficiently severe that damage to the machine is likely; immediate action required

For a standard industrial pump in Group 2, Zone B extends to 2.8 mm/s RMS. Zone C covers 2.8 to 7.1 mm/s RMS. Zone D begins above 7.1 mm/s RMS.

These thresholds provide consistent, defensible decision criteria. They prevent the subjectivity of “this machine always vibrates a bit” from deferring maintenance on genuinely deteriorating equipment.

Connecting Analysis Results to Maintenance Actions

Vibration analysis delivers its full value only when findings translate directly into maintenance decisions. The workflow from fault detection to corrective work must be clear, consistent, and fast.

Effective work orders generated from vibration analysis findings should include the equipment identification and location, a description of the detected fault with supporting frequency data, the severity classification based on ISO zones or site-specific thresholds, recommended corrective action with specific repair guidance, and the priority level based on severity and equipment criticality.

Link work orders to the vibration analysis reports from which they originate. This gives maintenance technicians full diagnostic context when planning repairs – not just a notification that a bearing needs changing, but the spectral data showing what stage of deterioration the bearing is at and whether any secondary faults are developing.

Aquip provides vibration analysis reports formatted for direct integration with common CMMS platforms, removing the manual data entry step that delays the translation of monitoring findings into maintenance actions.

Coordinating Vibration Analysis with Other Monitoring Techniques

Vibration analysis is the most comprehensive diagnostic tool for rotating equipment, but it works best as part of a multi-technology monitoring approach.

Thermography detects electrical problems, lubrication deficiencies, and overloading that may not produce early vibration symptoms. Thermal cameras identify hot spots in motor windings and electrical cabinets that vibration accelerometers cannot sense. Combining thermographic surveys with vibration routes provides broader fault type coverage.

Oil analysis detects wear particle generation in gearboxes and lubricated bearings before vibration spectra show significant changes. For large slow-speed gearboxes where vibration energy is low, oil analysis often provides the earliest warning of developing deterioration.

Ultrasonic monitoring detects bearing lubrication failure and early-stage bearing defects using high-frequency acoustic emission. On slow-speed equipment where vibration defect frequencies are below the useful detection threshold, ultrasonic contact probe testing identifies friction-based deterioration that vibration analysis misses.

For critical equipment, schedule all monitoring technologies with coordinated frequency. When any technology flags a developing problem, increase monitoring frequency across all methods to track fault progression more closely and determine the optimal timing for intervention.

Building Analyst Capability

The effectiveness of a vibration analysis program depends directly on the competence of the analysts interpreting the data. Technology provides the measurements; human expertise turns those measurements into correct diagnoses.

ISO 18436 provides the internationally recognised framework for vibration analyst certification:

Category I – Basic data collection and route monitoring. Analysts at this level collect measurements correctly and identify obvious deviations from normal. Appropriate for technicians responsible for route collection.

Category II – Fault diagnosis and severity assessment. Analysts calculate bearing frequencies, identify fault signatures, assess severity against standards, and recommend specific corrective actions. Appropriate for analysts responsible for program data interpretation.

Category III – Advanced analysis and program management. Analysts handle complex machinery diagnostics, design monitoring programs, establish acceptance criteria, and manage the technical direction of the reliability program.

Technical training courses that combine theoretical instruction with hands-on equipment experience build genuine practical capability. Reading about bearing frequencies in a textbook is not the same as identifying them in real machine spectra and making correct maintenance recommendations based on them.

Structured mentoring of developing analysts by experienced practitioners accelerates the development of diagnostic judgement – the ability to distinguish real faults from noise, understand how multiple fault types interact, and make nuanced recommendations that account for equipment criticality and operational context.

Cost Justification and Program ROI

Quantifying the return on vibration analysis investment requires systematically tracking what the program prevents, not just what it costs.

Document every instance where analysis detects a developing fault before failure. Estimate the cost of the failure that was prevented: the direct repair cost, the associated production downtime, and any consequential damage to surrounding equipment. Accumulate these estimates over a twelve-month period.

For a main process pump, the prevented failure cost might include:

  • Emergency repair: $20,000-$100,000
  • Production loss during unplanned downtime: $50,000-$200,000 per day
  • Consequential damage to connected equipment: variable but potentially substantial

A single prevented failure of this type typically justifies an entire year of vibration monitoring program cost many times over. Programs that prevent three to five such failures per year routinely deliver 5:1 to 10:1 returns on the monitoring investment.

Aquip provides condition monitoring services with reporting structured around prevented failure documentation, helping maintenance managers build the evidence base needed to justify program continuation and expansion to facility management and financial decision-makers.

Conclusion

Professional vibration analysis is the most effective tool available for detecting developing rotating equipment faults before they cause failures. Integrated into a structured workflow with baseline measurements, route-based collection, CMMS connectivity, and analyst capability development, it transforms maintenance from reactive crisis management into proactive reliability program management.

The path to a mature program takes twelve to twenty-four months, but benefits begin with the first fault detected before failure. Explore condition monitoring products for vibration analysis equipment suited to your program requirements. Review condition monitoring services for expert diagnostic analysis support that supplements internal capability.

To discuss how professional vibration analysis can be integrated into your maintenance workflow, contact us and a specialist will help design an approach matched to your equipment, team capability, and program objectives.