Stochastic Analysis About Creep And Shrinkage Effects Of Long-Span Concrete Arch Bridge With Cfst Stiffening Frame

There exist a lot of random factors that influence shrinkage and creep of concrete. For large-scale concrete stuctures, such as tall buildings and long-span bridges, determinstic analysis can not meet the requirement of engineering practice, and stochastic analysis is strongly recommended. In this thesis, with Beipanjiang Bridge as researching object, the long-term creep and shrinkage behaviours of long-span concrete arch bridge with stiffening frame are investigated by the stochastic analysis method. The following studies are mainly involved in the present thesis.First, based on those previous studies, the current prediction models and analysis methods of creep and shrinkage are summarized, and the current research status about stochastic analysis of creep and shrinkage effect is discussed. Meanwhile, the theory of degenerated beam element and the finite element software (CSBNLA) are introduced.Then, the major random variables influencing the creep and shrinkage effects are introduced. The basic theory of stochastic analysis of creep and shrinkage effect is summarized. Moreover, a hybrid approach, which combines the Response Surface Method (RSM) and Monte Carlo samping (MC), is introduced into random analysis of creep and shrinkage.Based on those aboved mentioned studies, the probabilistic-meaning prediction about long-term deflection and structural stress of Beipangjiang Bridge is carried. It can be fround that the numercal results obtained from RSM and MC agree with those obtained from Latin Hypercube Sampling (LHS) well, which demonstrates the accuracy of the presented approach. It can be concluded from the research results that the randomness of creep and shrinkage effect can not be ignored. If deteterminstic apparoach is used, no results with probabilistic meaning can be drawn. Therefore, it is suggested that the randomness of input variables should be concerned in the long-term effect analysis.Finally, with response surface function, a new sensitivity index of random variables is defined. It can be found that the deflection is sensitive to the randomness of prediction model of creep and shrinkage and ambient humidity. The major random contributions to structural stress are the uncertainties of prediction model of creep and elastic modulus, meanwhile, the influence of magnitude of sustained load cannot be ignored also.