Research And Application On Non-probabilistic Reliability Of In-service Arch Bridge

The in-service arch bridge has been widely distributing, and share an important partin China’bridges. The arch bridge collapses had been frequently happening in recentyears, in order to ensure the highway traffic safety, it should make sure that the carryingcapacity, traffic capacity and better working conditions of the arch bridge, give the fullestexpression to the function of the in service arch bridge. For this purpose, it should makean assessment of the reliability of the arch bridge.In this paper, Firstly, discuss the development history from reliability theory toin-service structure reliability theory, which characteristic is using a large amount of datato count the distribution of the form of parameters, and the parameters of the smallerror can cause erroneous results, in this situation, scholars present the non-probabilityreliability theory, this theory use a small amount of data, and only needs to control thelimit of variables, and the compute is more simple, can effectively reduce the calculationworkload. In order to overcome the shortcomings of traditional reliability assessmentmethods, this paper present the non-probability reliability evaluation method of thecarrying capacity of in service arch bridge based on non-probability reliability theory.Non-probability reliability theory contain a variety of models, these arenon-probability model which based on interval analysis and non-probability model whichbased on the ellipsoid convex set. The interval analysis method is more available,therefore, this paper will choose non-probability reliability model theory of based oninterval analysis method to analysis the arch bridge reliability.It always require improvable algorithms supporting when model has been finished,the paper summarized the algorithm of the structure non-probability reliability index, thecurrent algorithms are different, they are mainly interval analysis and optimal search. Forthe complex arch bridge system, because of the limit state equation is not clear, thecommon method is no longer suitable for use. This paper will use the method ofnon-probability response surface and failure point optimization to analyze the arch bridge,however, through calculation, found that the calculation amount of non-probability response surface method is huge, the calculation amount of finite element of failure pointoptimization method is less, thus, this method is more efficiency, and the advantage ismore outstandingThis paper presents a flow chart of the non-probability reliability assessment ofbefore or after reinforced arch bridge, and instance an example of curved arch bridge,adopt field measurement data to obtain the interval variables of main parameters. Usefinite element software midas to establish the before or after reinforced arch bridge of thefull bridge model, then analyze the stress and resisting force of the cross section in mainarch. Use the non-probability reliability index before reinforced to assess the crosssection, and provide reasonable reinforcement program. According to correlation studies,this paper will use the method of reinforced the hogback in arch ring. The bridge floor onvault and hogback use thinnest30cm micro-expansion with C40micro expansionconcrete and pouring together with the main arch ring，the length of reinforcement is5m,in order to as far as possible without increasing the dead load of the bridge, it shouldreinforced10~15cm concrete on the hogback in arch springing.Then use the failure point optimization method to solve the non-probabilityreliability index after reinforcement found that1, the effect is obvious and thecross-section is safe and reliable after reinforce. The research findings in this paper canaffirm that use non-probability reliability theory to analyze the assessment of the archbridge reliability is feasible and effective, it has also prove the thought of assessment-reinforcement-assessment is reasonable.The non-probability reliability theory is being used widely at present. It was utilizedin various fields but still in preliminary stages in the field of the bridge, thus, it should bepay more attention on algorithm optimization in later stage.