• DocumentCode
    2179774
  • Title

    Competing failure modes modeling with limited wearout failures

  • Author

    Peng Liu ; Peng Wang

  • Author_Institution
    JMP Div., SAS Inst. Inc., Cary, NC, USA
  • fYear
    2013
  • fDate
    28-31 Jan. 2013
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    This paper proposes a new modeling approach which can be applied to solve some practical reliability engineering problems when competing failure modes present, but only one or limited failures have occurred. This type of failure data cannot be modeled by traditional time-to-failure distributions. The new approach is derived from Weibayes method and provides a viable solution when reliability inference is needed during early product deployment phase. The most important value of the derivation is: it provides a distributional interpretation about the Weibayes result, and the assumptions that are made. The derivation not only reproduces the important result in the literature, but also gives insight into the sampling distribution of the parameter estimate. A parametric bootstrap method is used to illustrate how to incorporate the result in competing failure mode modeling.
  • Keywords
    Bayes methods; failure analysis; parameter estimation; reliability; sampling methods; statistical distributions; Weibayes method; competing failure modes modeling; failure data; limited wearout failure; parameter estimate; parametric bootstrap method; product deployment phase; reliability engineering problem; reliability inference; sampling distribution; Bayes methods; Data models; Maximum likelihood estimation; Reliability engineering; Suspensions; Synthetic aperture sonar; Bayesian; Failure Mode; Reliability; Wearout; Weibayes;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Reliability and Maintainability Symposium (RAMS), 2013 Proceedings - Annual
  • Conference_Location
    Orlando, FL
  • ISSN
    0149-144X
  • Print_ISBN
    978-1-4673-4709-9
  • Type

    conf

  • DOI
    10.1109/RAMS.2013.6517738
  • Filename
    6517738