Downloaded from https://www.everyspec.com AMCP 706-IN a full-factorial design. If variations in input with Repair, arid Availability, are specified power, frequency, and drnilar factors had to as system measures of effoctiveness. Usually, be generated, the total number of combina- the system specification states that the mys- tions would be atronomically high. If tern must meet some minimum level of a historical information is avalable on tests -parameter at a specified s-confidence level. made on similar equipments, It may be pos- For example, the specification might state sible to eliminate some environmental com- that the system must meet a 1200 hr mini- binations from the test.
A simple example mum MTBF with'a 00% a-confidence level, of this follows. If the Interacston of tern- Various measures can be selected, depending perature and vibration has been shown to on the system design, performance charao- be a critical factor only for high tempera- teristica, and mission. tures, then combinations of temperature and vibration for low temperatures can be is. There are no formal rules for selecting the noted. measures to be tested.
Engineers and con- tracting officers must use their best judgment Since new equipment designs usually make based on their experience when specifying use of many standard parts, circuits, and test parameters, Considerable care must be assemblies, it may be possible to use pub- devoted to the selection of parameters to be Wished Information on their performance and measured on the test because the types of reliability characteristics to reduce the test rils, test equipment, and procedures to amount of reliability testing required. These be used depend on the parameters tested. If data must be evaluated carefully. A check.
the proposed test is too severe and expensive, list of factors to consider in evaluating such data is given in Table 3-2, This list is to be used together with the engineer's knowledge of the system to determine how the existing TANLE 3,2 data can be used. But remember, engineers tend to be overly optimistic about their BSIC CH•CKLIST FOR DATA RIVIEW desilns 1. What Is the source of the data? The number ol' environmental combine. 2. When and where were the date obtained? 3. Whet was the purpo•e of the experiment? tions which must be appliod to a syatem 4, Concernng which population can conclusions during reliability or to Its componentu dur. be draw?
ing qualification testing depends on how 5. What was the experimental design? much information can be derived from mli- 8. How were sample items •i•eted? ability tests performed during research and 7. Met were the operating and environmental development. Thesn tests provide data that conditlons? substantially reduce the number of com- 8. If failure rates are given, how was fallurn binations of environmental and other factors defind? which must be considered In later testing. 9. If repair rates are given, under what conditions were repairs mdde? 3.3.2 SELECTION OF TEST PARAMETERS 10. Is there any messuve of experimental error? I11. Whet was the basis for oomp•.ting time?
The reliability parameters which am 12, Wht type of test Imtrumntation wa ea ployed? measured by a test are usually those which 13, What arete m~ordience beten Items have been specified in the original contract tested W and those undenconsidaration? and specifications (Refs, I and 2). Param- 14. How do the results compare with those of eters such as Mean Life, MTBF, Reliability similar Investlgations? I. ~ IV t