Downloaded from https://www.everyspec.com AMCP We-i7f L Made up o' IFR elements, are them- 2. R. E. Barlow and A. W. Marshall, selves IFR. "Bounds for Distibution with Mono- 2. Made up of DFR elements, are them- tone Hazard Rate-, DI-82-0247, Boeing Scientific Research Ladorr~i, 1963. selves DFR. Ra. 8 shows how to test a sam- ple to see if it come3 from a disribution with 3. R. E. Barlow and A. W. Marshall, a monotonic failure rate, and if so, whether it "Tables of Bounds for Distributions with is IFR or DFR. Monotone Hazard Rate", D1-8720249, Eventoughtis orBoeing Scientific Research Laboratories, Even though this mathematical material is 1963.
available in the literature, it is not clear how 4. R. E. Barow and F. Proschan, "Compar- valuable it can be to the reliability engineer. ison of Replacement Poies, and Re- An experienced statistician ought to be con- newal Theory Implications", suited before applying any of the results. The D1-82-0237, Boeing Scientific Research reliability engineer must also use his judgment Laboratories, 1963. in deciding how much les stringent the re- 5. R. E. Barlow and F. Prochan, Mathema- strictions for this theory really are, than just tical Theory of Reliability, John Wiley to blithely assume one of the conventional and Sons, New York, 1964. distributions. 6. J. D. Esary and F.
Proschan, "Relation- Generally speaking, the decisions about ship Between System Failure and Coin- hardware will not be radically different re- ponent Failure Rates", Technometrics 5, gardless of which of several distributions is No. 2,183-9 (1963). chosen to represent the life of the units. If 7. R. E. Barlow, A. W. Marshall, and F. that conclusion is not true, then the engineer Proechan, "Properties of Probability Dis- is in serious trouble because he needs more tributions with Monotone Hazard Rate", information than he has. Annals of Mathematical Statistics 34, - No. 2, 375-89 (1963). 8. F. Proschan and R. Pyke, Tests for REFERENCES Monotone Failure Rate. 1. F.
Proschan, "Theoretical Explanation of Observed Decreasing Failure Rate", Technometrics 5, No. 3, 375-83 (1963). 4- 4.4