Random Vibration Fatigue

Category: Structural Analysis | Integrated 2026-04-06
CAE visualization for random fatigue theory - technical simulation diagram
Random Vibration Fatigue

Random Vibration Fatigue: Theoretical Foundations

What is Random Vibration Fatigue?

๐Ÿง‘โ€๐ŸŽ“

Professor, does fatigue occur under random vibration?


๐ŸŽ“

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

๐ŸŽ“

Instead of time-domain fatigue (Rainflow method + Miner's Rule), a method to directly estimate fatigue life from PSD:


Dirlik Method (1985)

๐ŸŽ“

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.


$$ m_k = \int_0^\infty f^k S_{\sigma}(f) df \quad (k = 0, 1, 2, 4) $$

Constructs Dirlik's PDF from spectral moments $m_0, m_1, m_2, m_4$ and calculates the expected fatigue damage rate.


๐Ÿง‘โ€๐ŸŽ“

So you can know the fatigue life without converting the PSD back to a time history!


๐ŸŽ“

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

๐ŸŽ“
MethodAssumptionAccuracy
NarrowbandStress is narrowband (dominated by a single resonance)Conservative (overestimates for broadband)
DirlikBroadband compatibleHigh (practical standard)
Benasciutti-TovoBroadband compatibleComparable to Dirlik
Zhao-BakerBroadband compatibleComparable to Dirlik

Summary

๐ŸŽ“

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

Coffee Break Yomoyama Talk

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

๐ŸŽ“

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

๐ŸŽ“
ToolRandom Fatigue
nCode DesignLifeSupports Dirlik method. Directly inputs FEM results (PSD stress)
fe-safeDirlik method. Reads results from Abaqus/Nastran
FEMFATDirlik method. Automotive vibration fatigue standard
Nastran + PARAM,FATIGUELimited random fatigue capability
๐Ÿง‘โ€๐ŸŽ“

So dedicated fatigue software is necessary.


๐ŸŽ“

FEM solvers themselves have limited random fatigue functionality. The standard workflow is to calculate PSD stress with FEM and evaluate life with dedicated fatigue software (nCode, fe-safe, FEMFAT).


Summary

๐ŸŽ“
  • FEM PSD analysis โ†’ Dirlik method in fatigue software โ€” Standard workflow
  • nCode DesignLife / fe-safe / FEMFAT โ€” Dedicated fatigue software
  • Estimate PDF from spectral moments โ€” Damage using Miner's rule

  • Coffee Break Yomoyama Talk

    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

    ๐ŸŽ“

    Random vibration fatigue is a problem in automotive exhaust systems (mufflers, catalytic converters), aircraft structures, and electronic device PCBs.


    Practical Checklist

    ๐ŸŽ“
    • [ ] Is the stress PSD calculated correctly? (FEM PSD analysis results)
    • [ ] Is the S-N curve based on material test data?
    • [ ] Is the Dirlik method (or an appropriate broadband method) used?
    • [ ] Is the fatigue damage degree $D < 1$?
    • [ ] Is the stress concentration factor $K_t$ considered?
    • [ ] Is the effect of mean stress (Goodman, etc.) considered?

    • ๐Ÿง‘โ€๐ŸŽ“
      Related Simulators

      Experience the theory firsthand with the interactive simulator for this field

      All Simulators

      Related fields

      Rate this article
      Thank you for your feedback!
      Helpful
      More details
      Report error
      Helpful
      0
      More details
      0
      Report error
      0
      Written by NovaSolver Contributors
      Anonymous Engineers & AI โ€” Sitemap