• DocumentCode
    2209660
  • Title

    Reliability quantification of induction motors-accelerated degradation testing approach

  • Author

    Wang, Wendai ; Dragomir-Daescu, Dan

  • Author_Institution
    Gen. Electr. Corp. R&D Center, Schenectady, NY, USA
  • fYear
    2002
  • fDate
    2002
  • Firstpage
    325
  • Lastpage
    331
  • Abstract
    Characterizing life of cheap, highly reliable devices, such as small induction motors, is a challenge due to cost and time restriction. Accelerated degradation testing (ADT) provides a way to predict its life cost- and time-effectively. Having time tracking on degradation signals within relatively short test duration, the reliability can be estimated through accelerated degradation testing, even without any failures. ADT on induction motors have been conducted in the GE R&D Center to quantify the motor reliability. A new degradation index, the Orbit Area, for the induction motors is proposed, which turns out to be one of the most reliable indices for time tracking motor bearing wear. Signal processing method for extracting this degradation index from measured signals was developed. A time-to-degradation model is constructed based on the physics of failure and test observations. Maximum likelihood estimation is employed to analyze test data. The reliability of the motors is finally predicted through the obtained degradation model. Several issues regarding ADT and lessons learned are also discussed in the paper
  • Keywords
    induction motors; life testing; machine bearings; machine testing; maximum likelihood estimation; reliability; signal processing; accelerated degradation testing; induction motors; life prediction; maximum likelihood estimation; motor bearing wear; physics of failure; reliability estimation; reliability prediction; reliability quantification; signal processing method; test observations; time-to-degradation model; Costs; Data mining; Degradation; Extraterrestrial measurements; Induction motors; Life estimation; Life testing; Maximum likelihood estimation; Physics; Signal processing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Reliability and Maintainability Symposium, 2002. Proceedings. Annual
  • Conference_Location
    Seattle, WA
  • ISSN
    0149-144X
  • Print_ISBN
    0-7803-7348-0
  • Type

    conf

  • DOI
    10.1109/RAMS.2002.981662
  • Filename
    981662