A novel risk assessment for complex structural systems

Risk management is an essential tool for safe, economical, and efficient design, operation, and maintenance of complex engineering systems. Seismic risk assessment of structures, particularly in nuclear power plants, needs special attention from the reliability community. Available risk assessment methods may not be sufficient to estimate the risk of complex systems made with different materials, numerous ways elements are connected to each other, and excited by dynamic loadings including seismic loading applied in the time domain. A hybrid risk assessment approach is proposed by intelligently integrating the stochastic finite element method, and the response surface method. It is capable of estimating the probability just before failure considering all major sources of nonlinearity and uncertainty, eliminating the deficiencies of the currently available reliability methods. With the help of the illustrative examples, it is shown that the method is robust, accurate, and efficient in estimating risk of complex systems excited by dynamic loadings applied in the time domain.