Probabilistic modelling of the life time of structures submitted to thermal fatigue
Abstract
Material behaviour under thermal fatigue is extremely affected by numerous uncertainties. These uncertainties are due to random nature of input variables and lack of knowledge on thermal fatigue. This phenomenon concerns in particular the nuclear industry, where the structural components are subjected to thermal loading. The current nuclear design code for thermal fatigue takes into account the uncertainties by applying empirical safety margins in order to make the design conservative. This approach does not allow for a quantification of the associated risk yet. It is therefore interesting to introduce a global probabilistic framework as an alternative to the current thermal fatigue design code. This requires:
- to define the probability density function of the random variables involved in thermal fatigue model,
- to incorporate these random variables in the design procedure to compute the random life time of the structure under consideration.
The objective is to compute the lifetime probability and the effects of each uncertainty on the reliability of the structure. The proposed probabilistic approach of thermal fatigue design is based on the theory and methods of structural reliability. Two kinds of approaches are defined:
- a reliability analysis in the time domain, where the loading is described by a thermal time history,
- a reliability analysis in the frequency domain, where the loading is described by its power spectral density function.
The above methods are applied on an example of pipe submitted to thermal loading due to water flow. The results show that it is possible to perform a complete reliability analysis to assess failure probability. They show also that scatter of fatigue data and the heat-transfer coefficient are the most important variables in thermal fatigue reliability.
Keywords
Probabilistic fatigue models - structural reliability - thermal fatigue - random processes - sensitivity analysis
BibTeX cite
@PHDTHESIS{GuedeThesis,
author = {Gu\'ed\'e, Z.},
title = {Approche probabiliste de la dur\'ee de vie des structures soumises
\`a la fatigue thermique},
school = {Universit\'e Blaise Pascal, Clermont-Ferrand, France},
year = {2005}
}