TABLE OF CONTENTS Three possible methods of determining the target reliability that a system should be designed to are the following:
Calculate the reliability of several different configurations of the system, and some with identical configurations but utilizing different state of the art components. These should be those configurations which will perform the specified functions and will attain the desired mission objectives. Next, calculate the total cost to the user of owning, operating, and maintaining this system in its different configurations. Now find the optimum level of system reliability at which the total cost is a minimum. This optimum reliability may be selected to be the target reliability, which is that of configuration C 4 shown in Fig. 12.1.
Figure 12.1: Determination of the optimum configuration yielding the minimum total cost reliability. C i stands for the ith configuration. Configuration C 4 is the optimum configuration.
Obtain a multivalued prediction of the chosen system configuration using such failure rates as published by GIDEP, MIL-HDBK-217, RADC, and other sources. By multivalued is meant that for the same component use the lowest, the most likely, and the highest expected failure rates that component will demonstrate in its actual use environment. Based on the state of the art of the components and subsystems to be actually used in the system, properly modify these failure rates, determine the system's overall failure rate (three of them based on the lowest, the most likely, and the highest failure rates), and select the target...
