Vibration-based condition monitoring is a powerful tool for assessing machinery health. By analyzing vibration patterns, engineers can detect faults and deterioration in rotating equipment before they lead to costly failures. This technique uses sensors to measure machine vibrations and interprets key parameters like amplitude and frequency.

Data analysis in multiple domains provides unique insights into machine condition. Techniques like Fourier transforms and envelope analysis help identify specific fault signatures. By establishing baselines and trending data over time, engineers can predict potential failures and implement proactive maintenance strategies.

Principles of Vibration-Based Monitoring

Fundamentals of Vibration Monitoring

• Vibration-based condition monitoring assesses rotating machinery health by analyzing vibration patterns
• Machine faults or deterioration cause detectable changes in vibration signatures
• Vibration sensors measure machine vibrations at specific locations
  • Accelerometers
  • Velocity transducers
  • Displacement probes
• Key vibration parameters provide information about machine faults
  • Amplitude
  • Frequency
  • Phase

Data Analysis and Trending

• Trending vibration data over time detects gradual changes in machine condition
• Vibration analysis performed in multiple domains offers unique insights
  • Time domain
  • Frequency domain
  • Time-frequency domain
• Establishing baseline vibration levels and alarm thresholds enables early fault detection
• Continuous monitoring allows prediction of potential failures before they occur

Signal Processing for Vibration Data

Frequency Domain Techniques

• Fourier Transform and Fast Fourier Transform (FFT) convert time-domain signals to frequency-domain spectra
  • Reveals dominant frequencies and harmonics in vibration data
  • Allows identification of specific fault frequencies
• Cepstrum analysis detects harmonics and sidebands in vibration spectra
  • Particularly useful for gear fault diagnosis
  • Highlights repeating patterns in frequency spectra
• Order tracking analyzes vibration data from variable speed machines
  • Enables consistent frequency analysis across different operating speeds
  • Compensates for speed fluctuations during data collection

Time Domain and Filtering Techniques

• Time synchronous averaging (TSA) enhances periodic components and suppresses random noise
  • Improves signal-to-noise ratio for repetitive events
  • Useful for isolating gear mesh frequencies
• Digital filters isolate specific frequency ranges of interest
  • Low-pass filters: Remove high-frequency noise
  • High-pass filters: Eliminate low-frequency vibrations (building vibrations)
  • Band-pass filters: Focus on specific frequency bands (bearing fault frequencies)

Advanced Signal Processing Methods

• Envelope analysis (demodulation) detects high-frequency repetitive impacts
  • Extracts modulating signal from carrier frequency
  • Particularly effective for bearing fault diagnosis
• Wavelet transform provides time-frequency analysis of non-stationary signals
  • Offers simultaneous time and frequency information
  • Useful for analyzing transient events and varying frequency content

Fault Signatures in Vibration Signals

Rotational Faults

• Imbalance manifests as dominant peak at 1x running speed in frequency spectrum
  • Amplitude proportional to square of rotational speed
  • Typically highest in radial direction
• Misalignment produces strong harmonics at 1x and 2x running speed
  • Axial vibrations sometimes higher than radial
  • May also generate higher harmonics (3x, 4x)
• Looseness generates multiple harmonics of running speed
  • Often produces truncated time waveform
  • Presence of half-order and interharmonic components

Bearing and Gear Faults

• Bearing faults produce characteristic frequencies related to bearing geometry
  • Ball Pass Frequency Outer race (BPFO)
  • Ball Pass Frequency Inner race (BPFI)
  • Ball Spin Frequency (BSF)
  • Fundamental Train Frequency (FTF)
• Gear faults generate sidebands around gear mesh frequency and harmonics
  • Wear: Increased amplitude of mesh frequency and harmonics
  • Eccentricity: Modulation of mesh frequency at shaft rotation rate
  • Tooth damage: Impulsive excitation at tooth mesh rate

Other Common Fault Signatures

• Resonance conditions identified by significant vibration amplification at specific frequencies
  • Often accompanied by 90-degree phase shift
  • Can occur at natural frequencies of machine components
• Electrical faults in motors produce vibrations at line frequency and harmonics
  • 60 Hz (or 50 Hz) and multiples in frequency spectrum
  • Sidebands around these frequencies indicate rotor bar issues
• Cavitation in pumps generates broadband high-frequency noise
  • Often accompanied by erosion and component damage
  • Can be identified by characteristic "hissing" sound

Vibration-Based Fault Diagnosis Strategies

Data Collection and Management

• Establish comprehensive database of machine specifications
  • Bearing and gear frequencies
  • Critical speeds
  • Normal operating parameters
• Implement systematic approach to data collection
  • Consistent measurement locations
  • Standardized sensor types
  • Uniform data acquisition settings across machines
• Develop system for tracking and trending vibration data over time
  • Enables detection of gradual changes in machine condition
  • Facilitates prediction of potential failures

Analysis and Diagnostic Processes

• Develop multi-stage diagnostic process
  • Initial screening: Quick assessment of overall machine health
  • Detailed analysis: In-depth investigation of potential faults
  • Fault severity assessment: Evaluation of fault progression and urgency
• Utilize multiple analysis techniques in conjunction
  • Time waveform analysis: Reveals impulsive events and overall vibration levels
  • FFT spectrum analysis: Identifies specific fault frequencies
  • Envelope analysis: Detects high-frequency repetitive impacts
• Implement automated fault detection algorithms
  • Statistical pattern recognition: Identifies deviations from normal patterns
  • Machine learning models: Classifies fault types based on vibration signatures

Integration and Advanced Techniques

• Integrate vibration analysis with other condition monitoring techniques
  • Oil analysis: Detects wear particles and lubricant degradation
  • Thermography: Identifies hot spots and abnormal temperature patterns
  • Performance data: Correlates vibration changes with efficiency losses
• Establish cross-validation procedures for fault diagnoses
  • Compare results from multiple analysis methods
  • Verify findings with visual inspections when possible
• Implement continuous monitoring systems for critical machinery
  • Real-time data acquisition and analysis
  • Automated alerts for abnormal vibration levels or fault signatures