IEC 60605-4:2001

FOREWORD
1 Scope
2 Normative references
3 Definitions and Symbols
   3.1 Definitions
   3.2 Symbols
4 Assumptions and requirements
   4.1 Assumptions and information required for point
        estimates and confidence intervals
   4.2 Assumptions and requirements for prediction
        intervals
   4.3 Assumptions and requirements for tolerance
        intervals
5 Procedure for calculating point estimates and
   confidence intervals
   5.1 Time terminated tests
   5.2 Analytical procedure - Failure terminated tests
6 Prediction intervals for the number of failures in a
   future period
   6.1 Two-sided prediction intervals r[L2] and r[U2]
   6.2 One-sided prediction interval
7 Procedure for assigning tolerance intervals
   7.1 Upper Poisson tolerance bound
   7.2 Lower Poisson tolerance bound
Annex A (informative) Examples
      A.1 Point estimate of MTTF
      A.2 Application of lower one-sided 90 % confidence
          limit on mean time to failure (MTTF)
      A.3 Application of two-sided 90 % confidence limits
          on MTTF
      A.4 Application of a two-sided 90 % prediction
          interval
      A.5 Application of upper 90 % tolerance bound at
          95 % confidence
      A.6 Application of lower 90 % tolerance bound at
          95 % confidence
Annex B (informative) Relation between confidence,
        prediction and tolerance intervals
      B.1 Confidence intervals
      B.2 Prediction intervals
      B.3 Tolerance intervals
Annex C (normative) Computation of accumulated test
        time T*
      C.1 Case 1, one repaired item with constant failure
          intensity
      C.2 Case 2, more than one repaired item with
          identical constant failure intensities
      C.3 Case 3, non-repaired items
Annex D (normative) Tables of fractiles of the
        chi-squared distribution: X[2][alpha] (nu)
Annex E (normative) Integral of the chi-squared
        distribution and cumulative Poisson distribution
Annex F (normative) 0,95 fractiles of the F distribution
Bibliography

Provides statistical methods for evaluating point estimates, confidence intervals, prediction intervals and tolerance intervals for the failure rate of items whose time to failure follows an exponential distribution.