Identification of the design point based on Monte Carlo simulation

This paper proposes a design point identification method based on Monte Carlo simulation (MCS), and it ad-dresses the issue that design point and parametric sensitivity cannot be obtained by random simulation. Based on the random samples generated in MCS, the proposed method takes the failure sample with the maximum value of probability density function on the limit state surface (LSS) as the design point approximately. The accuracy of the design point identification method depends on the number of failure samples generated in MCS, particularly relying on the number of failure samples close to LSS. With improved MCS algorithms, the design points can be identified more efficiently and accurately. The proposed method was illustrated using three examples. Results show that the proposed method can identify design point accurately and effectively. When random field modeling was applied to modeling the spatial variability, the proposed method based on the most probable failure realization of random field is feasible and provides an effective way for the calculation of design points considering the spatial variability of soils. Thus, it can provide helpful guidance for the calibration of the partial factors in the semi-probability RBD method.

Automated summary

This paper proposes a design point identification method based on Monte Carlo simulation (MCS) to address the issue of random simulation not being able to obtain design points and parametric sensitivity. The proposed method approximates the design point by selecting the failure sample with the maximum value of probability density function on the limit state surface (LSS) from the random samples generated in MCS. The accuracy of the method depends on the number of failure samples generated in MCS, particularly those close to the LSS, and can be improved with advanced MCS algorithms.