Study On System Reliability Of Hydraulic Structure In The Middle Route Project Of South-to-North Water Transfer

For some kind of reasons, the structural design of hydraulic structure in the Middle Route Project (MRP) of South-to-North Water Transfer followed the traditional method with single safety factor 'K'. Because the method can not take the uncertainty of various variables into account, it is been replacing by the new method based on structural reliablity gradually. As a part of the scientific project from national Water Conservancy Bureau named 'research on key technology of system risk and safety of cross-valley long distance water transfer project', this study conducts the analyses of system reliability for individual dam and beam-type aqueduct. The paper also investigates the direct numerical integration method to calculate the reliability of struacture and develops a computer program. The research of the paper is valuable for engineering practice.The main contents researched in this paper are as follows:The reliabilities of various failure modes are calculated by JC method and the system reliability based on the failure modes is calculared by PNET method for a part of the concrete dam of Danjiangkou Dam. The results show that the PNET method is suitable to calculate the system reliablity of hydraulic structure.With using the design data of Hexigou aqueduct, the seven failure modes are defined, and their reliability indices and system reliability index are analyzed. The failure modes discussed in the paper represent the main failure modes of a beam-type aqueduct, and analysis can be a typical example for reliability analysis of the beam-type aqueduct structures.Direct numerical integration is a good method to calculate the reliability of a complex structure. This paper investigates the several key problems in calculating reliablity with direct numerical integration and completes a direct numerical integration(DI) program, and conducts some parameter analyses. The results of parameter analyses show that, the direct numerical integration method is much more accurate than JC method when the variables are nonlinear and have large variability.