Estimation For Critical Flutter Velocity Of The Bridge With Flat Box Girder

Flutter of bridges plays a crucial role in construction of long span bridges as one kind of the wind-induced vibrations,many scholars have done much research in this area,however,the methods used to estimate flutter of bridges are not good in terms of precision and applicability.Based on the previous theories and researches,this paper puts forward two methods to help engineers calculate critical flutter wind velocity of flat box girder combined with database of wind tunnel experiments.To get those two methods,the work is as following:To begin with,methods which can be used to judge bridge flutter stability and the process of how scholars did some research on this aspect are introduced in brief.Meanwhile,this paper has listed some reasons why it is important to do some research to solve this problem.In order to make paper logic and intelligible,some tools and basic theories related to this paper are also introduced.Furthermore,this paper has explained advantages of using the methods to analyze flutter of bridges.All of them discussed above are in the first two chapters.Besides,this paper has analyzes key factors influencing bridge flutter stability and explained reasons why formula consists of four parameters,then selecting the mathematical model required for fitting out formula.During the process of fitting out the formula,two coefficients are solved firstly by the relationship between bend-torsion frequency and reduced wind speed,then according to least square principle,last three coefficients are gotten.Given that effects of width/height ratio and web slopes of box girder on flutter,the precision of this formula has been improved with use of CFD.Last,applying this formula to specific cases,to some extent,precision of formula is good.Finally,it has demonstrated that aerodynamic derivatives has an influence on bridge flutter stability and that four factors,like H₃*,A₁*,A₂*and A₃*,influence flutter in a large way.Database of these specific bridge cases are included in this chapter.What's more,parameters of BP neural network are set,therefore applying trained BP neural network to Qipanzhou bridge,the result is good.This chapter verifies that engineers can use BP neural network to predict flutter critical wind speed.