722 MECHANICAL DESIGN AND PACKAGING 



erable difficulty will be experienced in meeting the system reliability 

 requirement. 



Several possible corrective approaches exist. First, the radar design 

 may be simplified by sacrificing features that seem desirable but are not 

 absolutely essential to overall mission accomplishment. This is always a 

 worthwhile objective; a reliability deficiency provides a sharp spur to its 

 eventual achievement. 



A second approach is to examine the possible use of more reliable compo- 

 nents such as diodes and transistors to replace vacuum tubes. 



A third approach is to provide redundancy, i.e. to duplicate portions of 

 the system where a failure is more likely. 



An approach of a somewhat different character is to examine the original 

 requirement to determine whether it represents the only possible answer 

 to the tactical problem. For example, the three-hour operating requirement 

 of the hypothetical AI radar and fire-control system was based on the 

 mission time of the CAP aircraft. Actually, only 12 of the CAP aircraft are 

 aloft at any given time, and only 6 of these are used to engage the specified 

 raid. The remaining 36 aircraft used to combat the specified attack being 

 deck-launched have a required operating time of less than one-half hour. 

 Eighteen aircraft are unavailable for combat because of previous failures 

 which have not been corrected. 



Let us assume a system failure rate of 0.1 per hour and one-half hour of 

 effective operating time for the equipment in the deck-ready interceptors. 

 The expected number of failures that would occur among the 42 aircraft 

 used to engage the raid would be 



(36) (0.1) (0.50) + (6) (0.1) (3) = 3.6 = 4 failures 



This represents slightly less than 10 per cent of the total number. Thus 

 the 90 per cent reliability requirement can be met by a system with a 

 10-hour mean time-to-failure provided that this increased failure rate 

 does not overload maintenance facilities to the point where aircraft avail- 

 ability is reduced below the assumed value of 48 out of 66. 



As this analysis shows — and as has been pointed out in several previous 

 cases — the original radar requirements are not sacred. They merely 

 represent a first approach to the translation of tactical requirements into 

 technical requirements. As more becomes known about the problem and 

 the factors that limit the solution, it is often necessary to modify this first 

 approach through a more intensive appraisal of the original tactical 

 problem. This is the essence of the system approach to radar design. 



