Bearings rarely fail without warning. Yet, most teams can’t recognize the signs of failure until it’s too late, which is when the equipment breaks down.
The truth is, a trail of warnings precedes every failure. Long before the noise, heat, or visible wear, vibration patterns shift in ways that tell you exactly what’s going wrong, and where. However, humans lack the capacity to manually detect or sense these.
This is why vibration analysis has become a go-to strategy for high-performance maintenance teams. They understand their limitations. They deploy tools that boost their productivity and secure an efficiency advantage over other companies in places where many fail to look.
Vibration analysis offers a strategic shift in the maintenance approach, transitioning from a reactive response to proactive early defect detection. At this early stage, intervention is still low-cost and low-risk.
In this article, we’ll walk through how vibration analysis works to monitor bearing health, how to interpret fault signatures, and which tools and techniques can give you a real predictive edge.
What Is Bearing Vibration?
Bearing vibration refers to the oscillatory motion that occurs when a bearing is in operation.
Every rotating asset generates some level of vibration. That’s normal. But when there's a defect in the bearing, those vibration patterns change in very specific ways.
These changes are mechanical responses to physical issues: a cracked inner race, spalled rolling elements, misalignment, imbalance, or lubrication problems.
Each condition leaves a distinct fingerprint in the vibration signal.
By analyzing those patterns — particularly the frequency, amplitude, and phase — maintenance teams can identify what’s wrong and where it’s happening inside the bearing.
Vibration analysis decodes mechanical behavior in real-time and turns raw signals into actionable diagnostics. Pattern identification and problem detection are the heart of its value to maintenance operations.
How Vibration Analysis Contributes to Bearing Failure Prevention
By continuously tracking how vibration signals evolve over time, maintenance teams can catch subtle shifts that indicate wear, misalignment, or structural damage.
These changes show up in the data before they become audible, visible, or result in operational disruptions.
When done correctly, vibration analysis enables a comprehensive approach: early detection, accurate fault identification, progress tracking, and root cause elimination.
Let's explore how it works at each of these stages and helps build a smarter, more proactive maintenance routine.
1. Early Detection of Anomalies
Every bearing starts to “speak” long before it fails. The key is knowing how to listen. Vibration sensors pick up on subtle irregularities that occur in the earliest stages of bearing degradation.
At this point, the defect is often microscopic. There’s no heat, no noise, nothing visible during a standard inspection. However, in the vibration spectrum, the signal begins to shift, typically in the high-frequency regions first.
These early anomalies often indicate problems like surface fatigue, poor lubrication, or minor misalignment.
By identifying them early, maintenance teams can intervene while the issue is still isolated, avoiding damage to other components and reducing the cost of repairs.
2. Identification of Fault Conditions
Once an anomaly is detected, the next step is pinpointing what’s actually going wrong inside the bearing. Different faults leave different vibration signatures.
A cracked outer race produces peaks at specific frequencies (such as BPFO), while inner race damage appears differently, often with sidebands around the BPFI. Cage issues? Look for patterns tied to the FTF. Rolling element defects? That’s where BSF shows up.
By analyzing these fault-specific frequencies and their harmonics, technicians can go beyond general alerts and accurately identify which component of the bearing is compromised.
This level of diagnostic precision is what separates basic asset monitoring from true predictive maintenance. Its true form gives you the precision to plan targeted interventions instead of blanket overhauls.
3. Monitoring Wear Progression
Understanding how a fault evolves is what keeps failures off your floor. Vibration analysis enables you to monitor the progression of a defect over time. As the bearing degrades, the amplitude and frequency patterns intensify or shift.
This data creates a timeline of wear, giving you insight into how fast the issue is escalating. With continuous monitoring, you’re not just getting a snapshot of a single spike. You’re observing a trend in real time.
This real-time accuracy enables you to plan maintenance with precision, aligning work orders with the actual asset condition instead of arbitrary schedules. When you know the rate of progression, you know exactly when it’s time to act. Not too early, not too late - right on time
4. Predictive Maintenance Strategies
Vibration analysis is a core enabler of predictive maintenance. By analyzing vibration trends and fault frequencies over time, you can forecast when a component is likely to fail.
But to be effective, it needs more than just sensors and alerts. It requires structured strategies built on the data those sensors provide.
One of the most powerful applications is trend-based forecasting. By tracking vibration amplitude at key fault frequencies over time, teams can model the degradation curve of a bearing.
This helps prioritize your interventions. For instance, a bearing with a stable fault signature may not require immediate action, while one showing exponential growth is a red flag for urgent maintenance.
Another critical strategy is threshold customization. Instead of using generic alarm limits, you can define baseline conditions for each asset, then adjust alert thresholds dynamically as machine behavior changes.
5. Root Cause Analysis
Vibration analysis doesn’t just show that something is wrong. It helps explain why it’s happening. The emphasis is on cause.
By isolating the frequency components and patterns associated with specific mechanical issues, you can determine whether the failure stems from misalignment, imbalance, looseness, improper installation, or a lubrication issue.
For example, harmonics with sidebands often indicate misalignment, while random high-frequency noise might signal lubrication breakdown.
This precision is crucial for conducting an effective root cause analysis. Instead of treating symptoms, you can trace the fault back to its origin, whether it’s a process problem, an installation error, or an upstream mechanical issue.
Addressing the root cause avoids repeat failures, saves parts and labor, and improves overall equipment reliability across the board.
6. Continuous Monitoring and Trend Analysis
Spot checks can tell you what’s happening right now, but they miss the bigger picture. Continuous monitoring reveals how machine behavior changes over time, which is where real predictive power lies.
Trend analysis converts raw vibration data into a timeline.
By tracking how key metrics—like amplitude at fault frequencies, overall RMS levels, or high-frequency energy—evolve over days or weeks, you start to see patterns that one-time inspections can’t capture.
For example, a slowly rising BPFO signal isn’t urgent today, but if it doubles in a week, that tells you the damage is accelerating. That’s the insight you need to make timing decisions with confidence.
Combined with automated thresholds and contextual alarms, trend analysis helps teams prioritize the right assets and plan interventions based on actual failure progression.
Formulas for the Calculation of the Bearing Failing Frequencies
Every bearing has its own unique vibration signature. The key to reading it lies in understanding its geometry.
When a defect develops in a bearing, it tends to excite specific frequencies tied to the inner race, outer race, rolling elements, or the cage.
These are known as Characteristic Defect Frequencies (CDFs), which can be calculated using a few essential formulas.
To calculate them, you’ll need basic bearing parameters:
- Number of rolling elements (n)
- Ball diameter (d)
- Pitch diameter (D)
- Contact angle (β)
- Shaft rotational speed (fᵣ)
Here are the four primary formulas:
- BPFO (Ball Pass Frequency Outer Race): BPFO = (n / 2) * fᵣ * (1 - (d / D) * cos β)
- BPFI (Ball Pass Frequency Inner Race): BPFI = (n / 2) * fᵣ * (1 + (d / D) * cos β)
- BSF (Ball Spin Frequency): BSF = (D / 2d) * fᵣ * (1 - ((d / D) * cos β)²)
- FTF (Fundamental Train Frequency): FTF = (1 / 2) * fᵣ * (1 - (d / D) * cos β)
These frequencies help identify where the fault is occurring. For example, if you're seeing consistent peaks at BPFO, the issue is likely on the outer race. Sidebands and harmonics can further confirm the diagnosis.
Knowing these values allows analysts to correlate specific frequencies in the vibration spectrum to physical damage in the bearing, making their diagnostics faster, clearer, and more accurate.
How To Measure Vibration Frequency Spectrum Analysis
Capturing vibration is only useful if you make sense of it. To detect bearing faults with precision, raw vibration signals must be converted into the frequency domain.
Fast Fourier Transform (FFT) breaks down a complex time-based waveform into its individual frequency components, making it easier to spot characteristic defect frequencies like BPFO, BPFI, BSF, and FTF.
Spectrum analysis gives you a visual map of where energy is concentrated. Peaks in specific frequency bands reveal the presence and type of fault.
For example, harmonics around BPFI often signal an inner race defect, especially when combined with sidebands spaced by the shaft speed.
The process usually involves:
- Signal collection via accelerometers mounted near the bearing housing.
- Digitization of the data through a data acquisition system.
- Processing with FFT to generate a spectrum.
- Interpretation based on fault frequencies, amplitude trends, and harmonics.
The frequency resolution, sampling rate, and bandwidth all impact the quality of the spectrum and your ability to spot early-stage issues. Low-frequency resolution might miss subtle signals. Too much bandwidth without filtering can bury insights in noise.
Acoustic Emission Techniques
When it comes to detecting bearing faults in their earliest stages, acoustic emission (AE) techniques go even deeper than standard vibration analysis.
AE sensors capture high-frequency stress waves emitted by microscopic events, like crack formation, material fatigue, or metal-to-metal contact.
These signals often occur in the ultrasonic range (typically above 20 kHz), well before the vibration levels become abnormal in traditional spectra.
What makes AE especially useful is its sensitivity. It’s capable of identifying surface defects at a stage where they haven't yet translated into measurable vibration.
This makes monitoring of low-speed or lightly loaded machinery ideal, as conventional vibration metrics might miss early indicators.
Used in conjunction with FFT analysis, AE provides another layer of insight.
It confirms the presence of faults and can be used to localize them with higher accuracy, especially when combined with envelope detection or high-frequency demodulation techniques.
Thermographic Analysis
Thermographic analysis uses infrared imaging to detect abnormal heat patterns on machines.
When a bearing starts to fail, friction increases. That friction generates heat, especially in cases of overloading, misalignment, or poor lubrication. A thermal camera can capture these changes in surface temperature long before they're visible to the naked eye or felt during a routine check.
Thermal imaging is a non-invasive and fast method, making it an excellent tool for screening multiple assets in a short period. It’s especially effective in detecting lubrication issues, one of the most common root causes of premature bearing failure.
While it doesn’t provide the fault-specific detail of vibration analysis, thermography adds context. For example, a bearing showing elevated temperature alongside rising vibration amplitude is a clear indicator of progressing damage.
Lubricant Analysis
Lubricant analysis involves testing oil or grease samples for signs of contamination, degradation, or metal wear particles.
These indicators can reveal early signs of bearing issues that aren’t yet detectable via vibration or thermal analysis. For example, an increase in ferrous debris suggests metal-on-metal contact inside the bearing, possibly due to spalling or surface fatigue.
A high moisture content may indicate seal failure, which compromises lubrication effectiveness and accelerates wear.
Modern analysis goes beyond visual inspection. Techniques like spectrometric analysis, ferrography, and viscosity testing offer detailed insights into both the lubricant’s condition and the internal state of the bearing.
When used together with vibration data, lubricant analysis provides a powerful confirmation tool. If a vibration spike aligns with a rise in wear particles, it reinforces the diagnosis and helps teams act faster and with greater confidence.
Periodic Inspection and Monitoring
Even with the rise of real-time monitoring technologies, manual inspection still has a place in a well-rounded maintenance strategy, especially for facilities without full online monitoring coverage.
Periodic inspections involve regularly scheduled checks using handheld vibration analyzers, temperature guns, or stethoscopes to capture baseline data and identify changes in machine behavior.
When logged consistently, these snapshots help build trend histories over time.
Inspections can also catch things that sensors might miss, such as loose mounting bolts, worn belts, improper lubrication application, or visible misalignment. These issues often contribute to bearing failures but don’t always produce strong vibration signals at early stages.
The key is consistency. When conducted on a fixed schedule and aligned with data from other monitoring tools, periodic inspections provide teams with an additional layer of insight, particularly for assets not yet covered by permanent sensors.
They aren’t a replacement for continuous monitoring, but they serve as a cost-effective bridge, keeping an eye on equipment that might otherwise go overlooked.
Vibration Analysis
Vibration analysis remains the most direct and scalable method for diagnosing bearing health.
It works by capturing the dynamic motion of a bearing in operation and converting that motion into measurable signals. These signals are then analyzed for amplitude, frequency content, harmonics, and modulation, each revealing specific mechanical behaviors or faults.
What sets vibration analysis apart is its diagnostic precision. Unlike temperature or oil changes, vibration shifts are tied to exact mechanical causes.
A consistent frequency peak at BPFI? That’s likely inner race damage. Sidebands around harmonics? It could be looseness or imbalance. Rising broadband energy? Potential lubrication failure.
The data can be collected using both online sensors (for continuous monitoring) and portable analyzers (for scheduled routes), giving flexibility across asset criticalities.
Vibration analysis tells you what’s wrong, where, and how bad. When paired with other techniques, such as thermography and oil analysis, it forms a comprehensive and detailed view of machine health.
Tools for Bearing Vibration Analysis
When it comes to bearing vibration analysis, precision matters. You need tools that can detect subtle signals, handle real-time data, and process it in a way that reveals actual failure modes.
A well-built vibration monitoring setup typically includes four essential components:
- Vibration sensors that detect motion in multiple axes.
- Data acquisition systems that capture and digitize those signals.
- Portable analyzers for route-based monitoring or spot checks.
- Diagnostic software that processes the data and helps identify fault signatures.
These tools work together to deliver alerts and provide context. That context should tell you whether a spike in amplitude is a sign of failure, or just a transient event.
See below each of these tools and what makes them effective or limiting in the field:
Vibration Sensors
Vibration sensors, typically accelerometers, are the front line of bearing analysis. They’re responsible for detecting the physical motion of a machine and translating it into a digital signal that can be analyzed.
Accelerometers come in various forms, but industrial-grade sensors typically offer high-frequency bandwidth, multi-axis measurement capabilities, and robust shielding for use in harsh environments.
These specs are crucial for detecting early-stage bearing faults, which often manifest in high-frequency ranges.
Sensor placement also matters. Mounting them as close as possible to the bearing housing, in the direction of the expected force, ensures accurate readings and improved sensitivity to internal defects.
Some sensors include built-in signal processing, filtering, or even onboard FFT capabilities. Others are designed to transmit raw data wirelessly to a centralized system.
The right choice depends on your monitoring strategy: permanent installation for critical assets, or portable setups for route-based inspections.
Data Acquisition Systems
Once vibration data is captured by sensors, it needs to be accurately recorded. That’s where data acquisition systems (DAQs) come in.
A DAQ system converts the analog signal from the sensor into digital data, filters out noise, and conditions the signal for analysis.
This step is critical: poor signal integrity or low-resolution sampling can distort frequency readings and lead to missed or misinterpreted faults.
Key capabilities to look for include:
- High sampling rate to capture high-frequency bearing fault signals.
- Multi-channel input for analyzing multiple bearings or axes simultaneously.
- Real-time streaming to enable continuous monitoring and instant alerts.
Modern DAQ systems are often paired with edge computing capabilities, processing the data locally before sending it to the cloud.
This reduces latency and improves responsiveness, which is essential for facilities that can’t afford delayed decisions.
Portable Analyzers
Portable vibration analyzers are essential for facilities that rely on route-based maintenance or need flexible, on-the-spot diagnostics. These handheld tools integrate vibration sensors, signal processing, and visualization into a single, compact device.
Technicians can walk up to a machine, take a reading, and instantly view the spectrum to identify anomalies without needing to install permanent equipment.
Advanced models feature built-in FFT capabilities, fault frequency libraries, and trend tracking capabilities. Some even sync with CMMS platforms to log data and automatically trigger work orders.
Portable analyzers are especially valuable in facilities where full online monitoring isn’t feasible. They allow teams to prioritize inspections, validate alerts, and make quick maintenance calls with confidence.
While they can’t replace the consistency of continuous monitoring, they’re a powerful tool for extending diagnostic coverage across your operation.
Diagnostic Software
By interpreting it all, diagnostic software ensures that the collected data is actually valuable. These platforms process vibration signals, apply frequency-domain analysis, and highlight patterns linked to specific bearing faults.
Whether you’re looking for harmonics, sidebands, or trends over time, software makes sense of the noise and surfaces what matters.
Modern diagnostic tools go beyond basic FFT. They use machine learning algorithms to detect anomalies, correlate signals with asset history, and even recommend corrective actions.
The best ones integrate directly with CMMS systems, giving you a clear path from detection to resolution.
Key features to look for:
- Automatic fault frequency detection
- Baseline tracking and trend visualization
- Integrated alert management
- Custom threshold configuration per asset
With the right software, your team doesn’t have to be vibration experts. The insights come pre-analyzed, contextualized, and ready for action.
Bearing Characteristic Defect Frequencies
Every bearing has a set of fault frequencies tied to its internal geometry, and these frequencies serve as a roadmap for diagnosing exactly where a problem is developing. These are referred to as Characteristic Defect Frequencies (CDFs).
When a bearing component begins to fail, it produces vibrations at specific frequencies based on its design and the machine’s operating speed.
Identifying these frequencies in your vibration spectrum helps pinpoint the exact source of the problem.
Here are the four primary CDFs every analyst should know:
- BPFO (Ball Pass Frequency Outer Race): Indicates faults on the outer race.
- BPFI (Ball Pass Frequency Inner Race): Tied to inner race damage.
- BSF (Ball Spin Frequency): Signals defects on the rolling elements.
- FTF (Fundamental Train Frequency): Related to cage damage or instability.
These frequencies depend on variables like shaft speed, number of rolling elements, ball diameter, pitch diameter, and contact angle. Because they’re not simple multiples of the shaft rotation speed, they stand out clearly in the spectrum, even when the bearing type isn’t known.
Tracking harmonics and sidebands of these frequencies gives a fuller picture. For example, if you see multiple harmonics of BPFI with sidebands spaced at 1x RPM, you’re likely dealing with inner race damage that’s actively progressing.
Knowing and recognizing these frequencies is essential for rapid, confident diagnostics and for taking action before failure spreads.
How Tractian’s Predictive Maintenance Solution Can Help Prevent Failures
Diagnosing a fault is one thing. Intervening before it turns into downtime is another. Tractian’s condition monitoring solution bridges the gap between diagnostics and intervention.
By continuously collecting and analyzing high-frequency vibration and temperature data, our technology gives your team real-time visibility into asset health. But Tractian goes beyond detection. When an anomaly is identified, our system doesn’t just trigger an alert- it initiates action.
Each warning is automatically linked to a work order within Tractian’s CMMS, ensuring that inspections, repairs, or part replacements are planned and prioritized based on the actual machine condition, rather than guesswork or outdated PM calendars.
This closes a critical gap between insight and execution for maintenance teams. With Tractian, all diagnostics are traceable. Every alert, every intervention, every fix. They're all documented and centralized for root cause analysis and continuous improvement.
And because the platform is fast to implement and operationally focused, your team moves from reactive firefighting to structured, data-driven action in weeks.
Do you want to stop guessing about your assets' health? Discover how Tractian's condition monitoring solution can predict failures before they happen.
