Stochastic Fatigue Crack Growth in Steel Structures Subjected to Random Loading

Abstract

Fatigue crack growth can be a significant problem in steel structures that are subjected to a large number of repeated load cycles during their service lives. Uncertainty in fatigue behavior arises from the random nature of the service load, environmental conditions, material properties and other factors. Stochastic approaches to model this uncertainty can lead to improved fatigue-resistant design or in-service inspection and maintenance policies. This report generalizes the stochastic model of uncertainty in crack growth under constant amplitude loading by introducing a random noise process with arbitrary (generally non-Gaussian) marginal distribution and correlation structure. A computationally efficient method based on the rainflow method of stress identification is proposed for handling broad-band stress processes in analyzing stochastic crack propagation. The impact of uncertainty in flaw detection and measurement capabilities on in-service structural condition assessment and reliability-based service life prediction also is investigated. The proposed method is applied in a time-dependent reliability analysis of a steel miter gate at the Emsworth Lock and Dam that suffered severe deterioration from corrosion-fatigue. The predicted fatigue behavior is consistent with observations of damage over a sixty-year service life, providing some confirmation of the use of stochastic fatigue analysis in structural condition assessment and service life prediction.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1998
Accession Number
ADA354172

Entities

People

  • Bruce R. Ellingwood
  • Ruohua Zheng

Organizations

  • Johns Hopkins University

Tags

Communities of Interest

  • Energy and Power Technologies
  • Ground and Sea Platforms

DTIC Thesaurus Topics

  • Civil Engineering
  • Computational Science
  • Crack Propagation
  • Data Science
  • Detection
  • Differential Equations
  • Fokker Planck Equations
  • Gaussian Processes
  • Information Science
  • Materials
  • Measurement
  • Mechanics
  • Probabilistic Models
  • Random Variables
  • Regression Analysis
  • Reliability
  • Stochastic Processes

Fields of Study

  • Engineering

Readers

  • Computational Modeling and Simulation
  • Statistical inference.
  • Structural Health Monitoring of Composite Structures.