Abstract:

Estimation of system reliability by classical system reliability methods generally assumes that the components are statistically independent, thus limiting its applicability in many practical situations. A method is proposed for estimation of the system reliability with dependent components, where the leading failure mechanism(s) is described by physics of failure model(s). The proposed method is based on structural reliability techniques and accounts for both statistical and failure effect correlations. It is assumed that failure of any component is due to increasing damage (fatigue phenomena) and the component lifetimes follow some continuous and non-negative cumulative distribution functions. An illustrative example utilizing the proposed method is provided, where damage is modeled by a fracture mechanics approach with correlated components and a failure assessment diagram is applied for failure identification. Application of the proposed method can be found in many real world systems.