• Title of article

    Probable cause analysis of cracks observed on vertical centrifugal pump

  • Author/Authors

    Achouri، نويسنده , , Ryma and Nouicer، نويسنده , , Omeima and Mhiri، نويسنده , , Hatem and Bournot، نويسنده , , Philippe، نويسنده ,

  • Issue Information
    روزنامه با شماره پیاپی سال 2013
  • Pages
    11
  • From page
    1
  • To page
    11
  • Abstract
    Two identical vertical centrifugal pumps are used to cool the equipment used for electricity generation by combined cycle. One of these two pumps is cracked at its elbow, unlike the second which operates under the same conditions and which displays no anomaly. These cracks whose causes remain undetermined present a serious threat to electricity production. These causes may be of a mechanical nature related to the nature of material of the pump or related to the dynamic flow characteristics. The first hypothesis was excluded since both pumps are made from the same material (ductile iron (S-NC 20-2)), which leads us to analyze the second hypothesis by studying the flow occurring in the pump for particular operating conditions. Two scenarios are considered in this work, the first considering an improper attachment of the wheel upon its replacement, while the second deals with the possibility of a sudden closing of the valve “water hammer”. We therefore propose in this work, the numerical modeling of the flow through the vertical centrifugal pump. The study was conducted using the computational code Fluent. The studied pump provides a flow of 18,100 m3/h with a total head of 10.1 m head. Fluent simulation of the pump operation for the two selected scenarios and the comparison between the results obtained and those of the normal operation allowed us to conclude that a sudden closing of the valve is the most likely cause explaining these cracks on the one of two pumps.
  • Keywords
    Failure analysis , Pump failures , cracks
  • Journal title
    Engineering Failure Analysis
  • Serial Year
    2013
  • Journal title
    Engineering Failure Analysis
  • Record number

    2339669