Random Vibration Fatigue
Random Vibration Fatigue: Theoretical Foundations
What is Random Vibration Fatigue?
Professor, does fatigue occur under random vibration?
Random vibration is the repetition of irregular stress. The stress range fluctuates probabilistically, but damage accumulates cumulatively, leading to fatigue failure.
Fatigue Evaluation in the Frequency Domain
Instead of time-domain fatigue (Rainflow method + Miner's Rule), a method to directly estimate fatigue life from PSD:
Dirlik Method (1985)
Estimates the probability density function (PDF) of stress ranges from the spectral moments of the stress PSD $S_{\sigma}(f)$, and calculates fatigue life using Miner's Rule.
Constructs Dirlik's PDF from spectral moments $m_0, m_1, m_2, m_4$ and calculates the expected fatigue damage rate.
So you can know the fatigue life without converting the PSD back to a time history!
The Dirlik method is reported to have good agreement with the time-domain Rainflow method. It handles broadband random stress and is a standard method for vibration fatigue evaluation.
Narrowband and Broadband Methods
| Method | Assumption | Accuracy |
|---|---|---|
| Narrowband | Stress is narrowband (dominated by a single resonance) | Conservative (overestimates for broadband) |
| Dirlik | Broadband compatible | High (practical standard) |
| Benasciutti-Tovo | Broadband compatible | Comparable to Dirlik |
| Zhao-Baker | Broadband compatible | Comparable to Dirlik |
Summary
Key Points:
- Directly estimate fatigue life from PSD โ No need to convert back to time domain
- Dirlik method is the practical standard โ Handles broadband random stress
- Spectral moments $m_0, m_1, m_2, m_4$ โ Calculated from PSD integration
- Vibration fatigue is an interdisciplinary field of NVH and fatigue โ PSD analysis + fatigue evaluation
Palmgren-Miner Rule and Random Fatigue
The foundation of random fatigue life prediction is the linear cumulative damage rule proposed by Palmgren (1924) and Miner (1945). Failure is judged when the ratio ฮฃ(ni/Ni) of the number of cycles ni at each stress amplitude Si to the fatigue life Ni at that Si (read from the S-N curve) reaches 1.0. However, the critical cumulative damage value for Miner's rule has a large experimental scatter of 0.3 to 3.0, and a 2009 survey by Cten reported an average of 0.7 (standard deviation 0.4) for carbon steel welded joints.
Computational Methods for Random Vibration Fatigue
Random Fatigue Calculation Procedure
1. FEM PSD Analysis โ Calculate stress PSD $S_{\sigma}(f)$ at all nodes
2. Calculate Spectral Moments โ $m_0, m_1, m_2, m_4$
3. Estimate PDF using Dirlik Method โ Probability density function of stress ranges
4. Fatigue Damage using Miner's Rule โ $D = \sum n_i / N_i$
5. Fatigue Life โ $T = T_{test} / D$
Solver/Tools
So dedicated fatigue software is necessary.
Summary
Rainflow Counting Method Implementation and Standards
The Rainflow counting method is a stress amplitude counting algorithm jointly published in 1968 by Matsumoto Hiroshi (Kyoto University) and Yamada Michio, with the name inspired by the image of rain flowing down a roof. It is now standardized as ASTM E1049-85 (revised 1997). In Python, it can be implemented with the rainflow package (pip install rainflow), and counting for 10,000 points of time-history stress data completes in under 0.1 seconds. The counting result matrix display (From-To Matrix) is also a standard output in the MATLAB Fatigue Toolbox.
Random Vibration Fatigue in Practice
Random Fatigue in Practice
Random vibration fatigue is a problem in automotive exhaust systems (mufflers, catalytic converters), aircraft structures, and electronic device PCBs.
Practical Checklist
Do you use the same S-N curve as for regular fatigue?
Related Topics
Experience the theory firsthand with the interactive simulator for this field
All Simulators