Publication detail

Probabilistic modelling of fatigue crack - Some observations about conditional probability

KALA, Z.

Original Title

Probabilistic modelling of fatigue crack - Some observations about conditional probability

Type

journal article in Scopus

Language

English

Original Abstract

The article deals with time-dependent analysis of the failure probability of a load bearing element of a steel bridge. The theoretical model of fatigue crack propagation is based on linear fracture mechanics. The Latin Hypercube Sampling statistical method is used to calculate the unconditional and conditional probabilities. The conditional probability is calculated by excluding those random observations that have a detectable crack length observed during the inspection. The effect of included and excluded observations of input random quantities on the conditional failure probability is mapped and shown graphically. It is proven that the number of stress peaks per year can be considered as one random variable for any time interval of the life span of the bridge.

Keywords

Fatigue, steel, bridge, probability, stochastic, fracture mechanics, random sampling

Authors

KALA, Z.

Released

15. 11. 2018

ISBN

1998-4448

Periodical

International Journal of Mechanics

Year of study

2018

Number

12

State

United States of America

Pages from

121

Pages to

130

Pages count

10

BibTex

@article{BUT151663,
  author="Zdeněk {Kala}",
  title="Probabilistic modelling of fatigue crack - Some observations about conditional probability",
  journal="International Journal of Mechanics",
  year="2018",
  volume="2018",
  number="12",
  pages="121--130",
  issn="1998-4448"
}