Abstract :
In order to achieve a high probability of mission success (reliability), prolonged manned space missions require low failure rates for critical subsystems or components whose failure can be corrected before a mission abort or mission degradation, i.e., in-flight maintenance is one method of increasing mission success probability. In-flight repair or replacement of subsystems, subject to random times to failure, must be made before the maximum subsystem downtime is exceeded. Also, in-flight maintenance must take into account such factors as crew availability, time to repair, redundancy, failure rates, maximum allowable downtimes, and distributions of the preceding factors. The deterministic approach to obtain mission success probability is often too difficult to be applied within budget constraints. Therefore, a computer program was developed to estimate the reliability through simulation. Failures and repairs within a space vehicle were simulated, assuming a specific number of crewmen initially available for repair, constant failure rate, and lognormal repair times. Various parametric and sensitivity analyses of the important parameters were performed to determine their effects on mission success. For example, the effect of a subsystem´s mean time between failures on mission success for selected crew sizes (keeping other variables constant) can be shown. An important realistic feature of the program is the associating of a repair priority with each subsystem. A higher priority subsystem can displace a lower priority subsystem already under repair and can be placed ahead of a lower priority subsystem in a waiting queue.
