• Title of article

    A reliability analysis method for fatigue design

  • Author/Authors

    B. Echard، نويسنده , , N. Gayton، نويسنده , , A. Bignonnet، نويسنده ,

  • Issue Information
    روزنامه با شماره پیاپی سال 2014
  • Pages
    9
  • From page
    292
  • To page
    300
  • Abstract
    Structural design against fatigue is a complex task due to the significant number of uncertainties that are inherent to the fatigue phenomenon. In this context, the safety margin and the influence of each design parameter on structural reliability are extremely valuable knowledge for the robust design of structures subjected to fatigue loadings. The Stress–stRength approach is a well established probabilistic method for assessing the failure probability of already designed structures using a load-time history whatever mechanical behavior. It consists of the comparison of two Probability Density Functions (PDFs), the Stress (S) and the stRength (R) of the structure.This widely used engineering approach is very convenient to use but presents some weaknesses that are underlined in this paper through an illustrative case study. First, the failure probability is very sensitive to the PDFs selected for S and R. Second, the influence of each random variable on reliability cannot be determined since the uncertain parameters characterizing geometry, material properties and loads are gathered in the Stress PDF. This paper proposes a more general and robust approach that is able to accurately assess the failure probability and determine importance factors of each random variable for potential time-demanding mechanical models, such as those encountered in industry. An application provided by the engine manufacturer Snecma of SAFRAN Group shows the applicability in an industrial context.
  • Keywords
    Failure probability , Fatigue design , reliability , Kriging , sensitivity
  • Journal title
    INTERNATIONAL JOURNAL OF FATIGUE
  • Serial Year
    2014
  • Journal title
    INTERNATIONAL JOURNAL OF FATIGUE
  • Record number

    1162875