Understanding Differentiation in Vibration Analysis

Devices

Portable balancer & Vibration analyzer Balanset-1A

$2,353.83

Parts

Vibration sensor

$107.26

Parts

Optical Sensor (Laser Tachometer)

$147.78

Devices

Balanset-4

$8,107.90

Parts

Magnetic Stand Insize-60-kgf

$54.82

Parts

Reflective tape

$11.92

Devices

Dynamic balancer “Balanset-1A” OEM

$2,067.80

Definition: What is Differentiation?

Differentiation in vibration analysis is the mathematical process of converting vibration measurements from one parameter to another by taking the derivative in the time domain or multiplying by frequency in the frequency domain. Differentiation converts displacement to velocity, or velocity to acceleration. It is the inverse operation of integration, and while less commonly performed than integration (most sensors are accelerometers), differentiation is sometimes needed when displacement measurements from proximity probes must be compared to velocity standards or analyzed for high-frequency content.

Differentiation is a frequency-weighting process that emphasizes high-frequency components while de-emphasizing low frequencies—the opposite effect of integration. This makes differentiation useful for enhancing high-frequency diagnostic information but also amplifies high-frequency noise, requiring careful application.

Mathematical Relationships

Time Domain Differentiation

• Velocity from Displacement: v(t) = d/dt [x(t)]
• Acceleration from Velocity: a(t) = d/dt [v(t)]
• Acceleration from Displacement: a(t) = d²/dt² [x(t)] (second derivative)

Frequency Domain Differentiation

Simpler in frequency domain:

• Velocity from Displacement: V(f) = D(f) × 2πf
• Acceleration from Velocity: A(f) = V(f) × 2πf
• Result: Multiplying by frequency, so high frequencies amplified, low frequencies reduced

Why Differentiation is Used

Proximity Probe Applications

• Proximity probes measure shaft displacement directly
• Standards often specify velocity limits
• Differentiate displacement to velocity for comparison
• Enables standards compliance with displacement sensors

Emphasizing High Frequencies

• Differentiation amplifies high-frequency components
• Can reveal high-frequency defects in displacement data
• Converts low-speed displacement to more analysis-friendly acceleration

Sensor Comparison

• Compare displacement sensors to accelerometers
• Convert both to same parameter (usually velocity)
• Verify measurement consistency

Differentiation Challenges

Noise Amplification

The primary differentiation problem:

• Differentiation multiplies by frequency (high frequencies amplified)
• High-frequency noise amplified more than signal
• Signal-to-noise ratio degraded
• Example: 1% noise at 10 kHz amplified 100× relative to signal at 100 Hz
• Solution: Low-pass filter before differentiation

Sensor Noise

• Displacement sensors have noise (electrical, quantization)
• Differentiation to acceleration amplifies this noise dramatically
• Double differentiation (displacement → acceleration) compounds problem
• Generally avoid double differentiation if possible

Numerical Differentiation Errors

• Time-domain differentiation amplifies digitization errors
• Sensitive to sampling artifacts
• Frequency-domain method preferred for accuracy

Proper Differentiation Procedure

Single Differentiation (Displacement to Velocity)

1. Low-Pass Filter: Remove high-frequency noise (cutoff at 2-5× highest frequency of interest)
2. Verify Signal Quality: Check for noise, artifacts
3. Differentiate: Multiply by 2πf in frequency domain
4. Verify Result: Check reasonableness, compare to expected values

Double Differentiation (Displacement to Acceleration)

• Generally Avoid: Rarely gives good results
• If Necessary: Aggressive low-pass filtering (cutoff at highest frequency of interest)
• Limited Bandwidth: Accept that high-frequency content will be noise-limited
• Alternative: Use accelerometer if acceleration needed

Frequency Domain Implementation

Procedure

1. Compute FFT of displacement or velocity signal
2. Multiply each frequency bin by 2πf (or (2πf)² for double differentiation)
3. Apply low-pass filter in frequency domain if needed
4. Result is spectrum in differentiated parameter
5. Can compute inverse FFT for time waveform if needed

Advantages

• No cumulative errors
• Easy to apply filtering
• Computationally efficient
• Standard approach in modern analyzers

When to Use Differentiation

Appropriate Uses

• Converting proximity probe displacement to velocity for ISO standards
• Enhancing high-frequency content in low-speed displacement measurements
• Comparing different sensor types on same basis
• When proper filtering can be applied

When to Avoid

• Noisy displacement signals
• Double differentiation unless absolutely necessary
• When accelerometer available (measure acceleration directly)
• High-frequency analysis from displacement (use accelerometer instead)

Differentiation vs. Integration Comparison

Aspect	Integration	Differentiation
Frequency Effect	Amplifies low frequencies	Amplifies high frequencies
Common Use	Acceleration → Velocity, Velocity → Displacement	Displacement → Velocity
Problem	Low-frequency drift	High-frequency noise amplification
Required Filter	High-pass before integration	Low-pass before differentiation
Frequency	Very common	Less common

Modern Instrumentation

Automatic Conversion

• Modern analyzers automatically convert between parameters
• User selects desired parameter, instrument handles filtering and conversion
• Proper filters applied automatically
• Reduces user error

Multi-Parameter Display

• Show acceleration, velocity, and displacement simultaneously
• Each emphasizes different frequency ranges
• Comprehensive view of vibration characteristics

Differentiation, while less common than integration in vibration analysis, is a valuable tool for converting displacement measurements to velocity or acceleration, enabling standards compliance and multi-parameter analysis. Understanding differentiation’s noise amplification characteristics and proper filtering requirements ensures accurate parameter conversion when differentiating vibration signals.

---

← Back to Main Index