Docstoc

Reliability Model for Compressor Failure

Document Sample
Reliability Model for Compressor Failure Powered By Docstoc
					Reliability Model for
Compressor Failure
  SMRE Term Project
    Paul Zamjohn
    August 2008
                                           Proposal
                   Compressor Failure Data: Case 2.16 of Blischke-DATA
      Data on “large air compressors” for a military base near the seacoast will be
analyzed to determine the probabilistic failure structure. Air compressors require
“bleeding” prior to operation to function properly, the data below represents failure
due to binding in the bleed system. Salt air due to proximity to the ocean is believed
to be a major contributor, nothing is known about other variables and their impact to
reliability.

      Analysis will include:

•    Generating the descriptive statistics
•    Selecting the distribution that best describes the data and the distribution
     parameters
•    Calculating the failure probability density function (f)
•    Calculate the cumulative distribution function (F)
•    Calculating the survival probability function (R)
•    Calculating the hazard function (z)
•    Determining the MTTF
•    Perform Monte Carlo simulation to model and assess reliability
Compressor Failure Data

  Operating time for 202 compressors (failed and unfailed units)
      operating hours                         frequency
           0-200                                   0
          201-300                                  2
          301-400                                  0
          401-500                                  0
          501-600                                  2
          601-700                                  2
          701-800                                 10
          801-900                                 26
         901-1000                                 27
        1001-1100                                 22
        1101-1200                                 24
        1201-1300                                 24
        1301-1400                                 11
        1401-1500                                 11
        1501-1600                                 20
        1601-1700                                  8
        1701-1800                                  4
        1801-1900                                  2
        1901-2000                                  3
        2001-2100                                  3
        2101-2200                                  1
                                                     Probability Plot for Start
                                                LSXY Estimates-Arbitrary Censoring
                                                                                                                     C orrelation C oefficient
                             Weibull                                                    Lognormal
                                                                                                                              Weibull
            99.9                                                          99.9
                                                                                                                               0.971
             90                                                            99                                               Lognormal
             50                                                            90                                                  0.952
                                                                                                                           E xponential
P er cent




                                                              P er cent
             10                                                            50                                                    *
                                                                                                                            Loglogistic
              1                                                            10                                                  0.954

                                                                            1
             0.1                                                           0.1
                       500            1000          2000                         500        1000      2000
                             Star t                                                       Star t

                        E xponential                                                    Loglogistic
            99.9                                                          99.9
             90
                                                                           99
             50
                                                                           90
P er cent




                                                              P er cent




             10                                                            50
                                                                           10
              1
                                                                            1

             0.1                                                           0.1
                   1   10       100          1000     10000                            1000                  10000
                             Star t                                                       Star t
Failure vs. Reliability Function            Probability Distribution Function




  Hazard (failure) Rate              Monte Carlo Simulation vs. Equation
                                                          Comparison of Monte Carlo vs.Equation

                                   1.2000




                                   1.0000




                                   0.8000




                                                                                                                 MC
                                   0.6000
                                                                                                                 EQ




                                   0.4000




                                   0.2000




                                   0.0000
                                             250   500   750        1000             1250   1500   1750   2000
                                                                           (hours)

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:4/20/2012
language:English
pages:5