• DocumentCode
    3395654
  • Title

    Evaluation of losses in DC machines-a statistical approach

  • Author

    Safiuddin, Mohammed

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Buffalo Univ., Amherst, NY, USA
  • fYear
    1988
  • fDate
    2-7 Oct. 1988
  • Firstpage
    551
  • Abstract
    The power ratings of the drive systems used cyclic duty applications are dictated by the thermal considerations associated with the losses produced by the machine during an operating cycle and the specified production rate. A study was conducted to determine an expression that would describe these losses as a function of its operating speed and torque. Multiple regression was applied to the data collected from a set of sample machines for this purpose. Starting with the analysis of individual expressions for each of the components of losses in a DC machine, the author presents two regression models. The first one, based on a sample size of ten machines, provides an expression for determination of operating point losses at any specified speed and torque per unit of rated machine losses (at rated speed and torque). The second model provides an expression for the determination of rated losses in percent of machine power rating as a function of rated motor speed, armature volume, and percent armature droop. The design of the enclosure influences the power rating and hence the rated losses for a given armature volume. Mean estimates are presented for the determination of percent droop and armature volume for a range of power ratings from 50 to 800 hp.<>
  • Keywords
    DC machines; losses; 50 to 800 hp; DC machines; armature volume; losses; multiple regression; operating point losses; operating speed; percent armature droop; power rating; rated losses; regression models; statistical approach; torque; DC machines; Production systems; Torque;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Industry Applications Society Annual Meeting, 1988., Conference Record of the 1988 IEEE
  • Conference_Location
    Pittsburgh, PA, USA
  • Type

    conf

  • DOI
    10.1109/IAS.1988.25114
  • Filename
    25114