The Sensitivity Analysis Of Reliability Related To Uncertain Factors To Bridges

The probability of the structure completing fixed process under the certain condition and the certain time, called the reliability of the structure. The thesis completes the analysis of the new building bridge' reliability, and also completes time-variant reliability of bridge under its life. The analysis could get the probability of the structure completing fixed process, and the result could be guidance to the design, building and maintenance. In the past, because of the limit of the computation speed, we could only focus on the reliability of partial structure. But now, the computation speed has been great enhanced, it is not difficulty to process the analysis of stable system reliability of the large structure with several failure models. To the point, the thesis does the analysis of the reliability of the new building bridge in the whole life, with the Monte Carlo Method and Response Surface Method. The thesis also does the sensitivity analysis related to the input parameters for reliability.Qingxi River Bridge is a down-support Arch Steel Tube Bridge with rigid tie bar, its main span is 90 m. Firstly, the thesis analyses the bridge under dead loads, and gets the stress distributing of the bridge. Doing that, we could confirm the largest probability of failure models. After that, we could lay the vehicle loads basis on the most dangerous of the bridge. So that, we could confirm the ultimate bear the weight of the Bridge, and could do reliability analysis easily. Basis on the stable analysis of the bridge structure, we could do the sensitivity analysis related to stochastic variable for structure by the zone method.After confirming the failure models of the structure and the probability distributing of the stochastic variable, we could use Monte Carlo Method and Response Surface Method doing reliability analysis of the bridge. The result from the two methods, it is very close and good for the bridge. We also do correlation to the stochastic variable with the stress of the failure models. Doing that, we could know which variables id sensitive to reliability. Considering the real bridge' resistance would be reducing under the nature environments, time-varying loads and material deteriorating. So the time-variant system reliability of the bridge subjected totime-dependent load and resistance is investigated. The approach is suitable for predicting the reliability of the bridge. The effects of time-varying variable could be getting from doing correlation analysis.