Load- Strength Order Statistics Interference Models for System Reliability Evaluation

For s-dependent systems, reliability models capable of reflecting the effect of failure dependence, i.e. the so called “common cause failure” are in demand, since the system reliability depends not only on the reliabilities of the individual components but also on the interactions among the component failures. Based on the attribute of system failure, the meaning of order statistics, as well as the mechanism of failure dependence, this paper developed s-dependent system reliability models by incorporating order statistics of component strength into load-strength interference model. The models simply characterize the fact that series system fails as long as the applied load exceeds the first (the minimum) order statistic of component strength, parallel system (with n i.i.d. components) fails when and only when load exceeds the n (the maximum) order statistic of component strength, and k-out-of-n(F) system fails if load exceeds the k order statistic of component strength. Since no assumption about s-independent failure was applied, the models are in a position to reflect the effect of failure dependence which may bring about common cause failures. To evaluate the reliability of general system comprising of non-i.i.d. components, an extended load strength order statistics interference model was presented, too.