System Reliability Estimation With Input Data From Deterministic Simulations
The possibility of estimating reliability of hardware, both for components and systems, is important in engineering design, since many failures result in substantial impact on safety or functional requirements. Existing reliability estimation methods require measured or estimated input data which can be difficult to retrieve. The objective of this paper is therefore to derive a simulation-driven method, including variation management, for combining deterministic simulations with Fault Tree Analysis, to estimate system reliability when measured data is not available. The research work started with a literature survey followed by description of a typical as-is situation and definition of a to-be scenario. Then, a simulation-driven method was derived and verified by a case study. In particular, the system used for the case study was modeled and simulated as a transient dynamical system to derive information about loads on components. It was found that deterministic simulations can be used to produce relevant input data for fault tree analysis. The derived simulation-driven system reliability estimation method includes variation management and can be used for evaluation of concepts in the early stages of product development when limited measurement data is available.