What are bearing defect frequencies (BPFO, BPFI)?

These are characteristic vibration frequencies that appear in the spectrum when rolling element bearings develop faults. BPFO (Ball Pass Frequency Outer race) is calculated as N_b×f_s/2×(1−Bd/Pd×cosα). These frequencies depend on bearing geometry and shaft speed, enabling specific fault location identification.

What is envelope analysis in vibration diagnostics?

Bearing defects generate high-frequency impact signals. Envelope analysis extracts the signal's amplitude modulation (envelope) and applies FFT to it, revealing bearing defect frequency components buried in broadband noise. It is one of the most effective techniques for early bearing fault detection.

How do imbalance and misalignment differ in vibration signature?

Imbalance predominantly shows energy at the fundamental rotating frequency (1×). Misalignment generates harmonics at 2× and 3× the rotational frequency. Bearing defects produce characteristic non-integer frequencies (BPFO, BPFI, etc.). FFT analysis distinguishes these patterns for fault identification.

What is crest factor and why is it important?

Crest factor is peak amplitude divided by RMS value. Normal rotating machinery typically shows values between 2–4. Early-stage bearing damage raises this to 4–10 as the defect generates impact peaks. It provides a simple, sensitive indicator of developing bearing faults without requiring spectral analysis.

› Vibration Signal Analyzer Back

Vibration Engineering & Machinery Diagnostics

Vibration Signal Analyzer (Machinery Diagnostics)

Simulate vibration signals from bearing defects, imbalance, and misalignment. Experience machine health monitoring principles through FFT and envelope spectrum analysis. Bearing defect frequencies calculated automatically.

Rotation Conditions

Shaft Speed [RPM]1500

Add Fault Components

Imbalance (1×)

Amplitude1.00

Misalignment (2×, 3×)

Amplitude0.50

Bearing Defect (BPFO)

Amplitude0.30

Bearing Geometry

No. of rolling elements N_b9

Ball/Pitch diameter ratio Bd/Pd0.40

Contact angle α [°]0°

Noise level0.10

Bearing Defect Frequency

$\text{BPFO}=\frac{N_b f_s}{2}\left(1-\frac{B_d}{P_d}\cos\alpha\right)$

—

RMS

—

Peak

—

Crest Factor

Time waveform x(t)

FFT Spectrum (amplitude vs. frequency)

Envelope Spectrum (bearing diagnostics)

What is Vibration Signal Analysis?

🧑‍🎓

What exactly is a vibration signal telling us about a machine's health? It just looks like a wiggly line.

🎓

Basically, that "wiggly line" is a treasure map of information. Every bump, spike, and repeating pattern corresponds to a physical event inside the machine. For instance, a tiny, regular spike might mean a single chipped tooth on a gear. In this simulator, the main signal you see is the raw vibration from a bearing, which is a mix of all its defects and noise.

🧑‍🎓

Wait, really? So how do we find that tiny gear chip signal in all the noise? The raw signal looks so messy!

🎓

That's the key challenge! We use a mathematical tool called the Fast Fourier Transform (FFT). It converts the messy signal from the "time domain" (vibration vs. time) into the "frequency domain" (amplitude vs. frequency). Try it: click the "FFT Spectrum" tab above. Suddenly, hidden repeating patterns become clear peaks at specific frequencies, which we can link directly to physical causes like imbalance or bearing defects.

🧑‍🎓

Okay, I see peaks in the FFT. But what's "envelope analysis" for? The FFT seems to show everything already.

🎓

Great question! The FFT can miss early-stage bearing faults. These faults create very short, sharp impacts that get buried. Envelope analysis is like a detective's magnifying glass. It isolates and amplifies those impact patterns. A common case is a tiny pit in a bearing raceway. Try switching to the "Envelope Spectrum" tab after adding some noise with the slider—you'll see how it can reveal the defect frequency even when the normal FFT looks fuzzy.

Physical Model & Key Equations

The core of predictive maintenance is linking vibration frequencies to specific mechanical failures. For rolling element bearings, defects on the outer race create impacts at a predictable Ball Pass Frequency of the Outer race (BPFO).

$$ \text{BPFO}= \frac{N \cdot f_s}{2}\left( 1 - \frac{B_d}{P_d}\cos \alpha \right) $$

Where:
$N$ = Number of rolling elements (balls/rollers)
$f_s$ = Shaft rotational frequency (Hz)
$B_d/P_d$ = Ball (roller) diameter to Pitch diameter ratio
$\alpha$ = Contact angle (degrees)
This equation shows why the simulator's parameters matter: changing the Contact angle or the Ball/Pitch diameter ratio directly shifts the defect frequency you're trying to detect.

Another critical fault is mass imbalance, which is one of the most common issues in rotating machinery. It appears as a dominant vibration at the fundamental shaft rotation frequency.

$$ f_{\text{imbalance}} = f_s $$

Where:
$f_s$ = Shaft rotational frequency.
The physical meaning is straightforward: an uneven mass distribution causes a force that pushes the shaft outward once per revolution. In the simulator, you can create this by increasing the Amplitude for "Imbalance" and watch a dominant peak appear at 1× RPM in the FFT spectrum.

Real-World Applications

Wind Turbine Gearbox Monitoring: Gearboxes in wind turbines are expensive and hard to access. Vibration sensors on the gearbox housing stream data to analysts who use FFT and envelope analysis to detect early bearing spalling or gear pitting, scheduling maintenance before a catastrophic failure causes weeks of downtime.

Predictive Maintenance in Manufacturing: On a factory floor, critical pumps and motors are fitted with vibration sensors. By tracking the growth of peaks at specific defect frequencies (like BPFO), maintenance teams can move from a fixed schedule to a condition-based approach, replacing parts only when needed and avoiding unexpected production stops.

Aircraft Engine Health Monitoring (EHM): Jet engines are instrumented with accelerometers. Between flights, vibration data is analyzed for shifts in imbalance frequencies (indicating possible blade damage) or bearing tones, providing crucial data for "on-condition" maintenance that ensures safety and optimizes engine life.

Paper Mill Roll & Bearing Analysis: Large rolling drums in paper mills operate under high load. Envelope analysis is particularly useful here to detect the early-stage bearing faults that manifest as subtle impacts, masked by the high noise levels from the rolling process itself, allowing for planned bearing replacement during a scheduled line shutdown.

Common Misunderstandings and Points to Note

While experimenting with this tool, you might encounter a few points that are easy to misunderstand. First, you might tend to think "a lower natural frequency (heavier mass / softer spring) is more dangerous," but that depends on the situation. While lower frequencies do occur more easily in daily life, the real issue is how close the "excitation force frequency" is to the "natural frequency." For example, a chassis with a natural frequency of 2Hz is hardly affected by a high-speed fan (excitation frequency 100Hz). Conversely, engine idling vibration (20Hz) and a mount component with a natural frequency of 20Hz will resonate, even though the frequency itself is higher.

Next, note that this tool does not consider "damping." Real-world structures always have damping (an effect that converts vibration energy into heat, etc., to dissipate it). With significant damping, the response amplitude at the resonance peak is much lower than the theoretical value. This simulator's frequency response graph shows the ideal result for a "non-damped system," so in practice, the next step is to re-evaluate using a model that includes a damping term.

Finally, a pitfall in parameter settings. When setting multiple masses or spring constants in "series" or "parallel," it's common to miscalculate their combined values. For instance, if two springs are in "series" between masses, the combined spring constant k is given by $1/k = 1/k_1 + 1/k_2$. When adjusting k1 and k2 separately in the tool, if you don't keep this relationship in mind, you might end up with an unintended stiffness distribution.

Related Engineering Fields

The concepts of this multi-degree-of-freedom vibration analysis actually form the foundation of various, often invisible, engineering fields. In acoustical engineering, which deals with "sound" as air vibration, the speaker diaphragm or instrument body is a continuous medium. However, understanding their vibration modes (segmented vibration) is aided by the mental image of "multi-degree-of-freedom system modes" you learn with this tool. Particularly, the "nodes" (points with almost zero vibration) that appear in the second and higher modes directly correspond to the harmonics of a guitar string or the segmented vibration of a speaker cone.

Another field is control engineering. In the positioning control of robot arms or precision stages, the natural vibrations of the structure can significantly compromise the control system's stability. If the frequency of the command signal from the controller is close to the mechanical structure's natural frequency, excessive vibration beyond the command (ringing) occurs, reducing accuracy. To prevent this, "eigenvalue analysis" is performed as part of "model-based design" during control system design to safely set the control bandwidth.

Furthermore, it connects to material mechanics and fracture mechanics. Repeated stress from vibration is a primary cause of "fatigue failure." In a resonant state, even a small excitation force can generate large stresses, easily exceeding the material's endurance limit. Combining "stress analysis," which predicts where maximum stress occurs, with the "modal analysis" you learn with this tool, enables "vibration fatigue analysis" to predict fatigue life.

For Further Learning

Once you're comfortable with this tool, next consider the vibrations of a "continuous medium." Objects like beams or plates where mass and stiffness are distributed continuously. Here, an infinite number of natural frequencies exist, but the fundamental concepts are the same. For example, the first natural frequency $f_1$ of a cantilever beam is expressed by a formula like $$f_1 \approx \frac{1}{2\pi} \cdot \frac{1.875^2}{L^2} \sqrt{\frac{EI}{\rho A}}$$. Here, L is length, EI is bending stiffness, and ρA is mass per unit length. Comparing how parameters equivalent to the spring constant k and mass m appear in this formula can deepen your understanding.

Mathematically, rewriting the multi-degree-of-freedom equations of motion into state-space representation increases compatibility with control theory and numerical simulation. Also, the modal analysis performed by the Finite Element Method (FEM) used in practice can be thought of as scaling up this tool's calculations to thousands or millions of degrees of freedom. A recommended learning path is: 1. Formulating equations of motion for multi-degree-of-freedom systems, 2. Numerical solutions for matrix eigenvalue problems (Jacobi method, QR method, etc.), 3. Generalized eigenvalue problems including damping, 4. Fundamentals of the Finite Element Method. First, extensively experimenting with parameter changes in this tool and observing the resulting behavioral changes with your own hands forms the foundation for everything.