Research On The Performance Safety Of Costal HSR Bridge Under Wind Environment

Coastal railway is a key component of designed high-speed railway net of China.However,compared to other aeras,the southeast coast and island,where the coastal railway located,are prone to typhoons and strong short-term convection activity.The harsh wind environment brings great challenges to the operational safety of high-speed trains,structural stability,and safety of HSR sea-crossing bridges.The high proportion of bridges in high-speed railway lines indicates the extension and growth of train operation time and frequency on the bridge.The safety and reliability of train-bridge interaction system under wind attack is critical to the overall operation safety of high-speed railway lines.On the other hand,the trend of lightweight design and high operating speed of high-speed trains increases the concern on high-speed railway operation safety.As a result,this dissertation constructs the probabilistic models of parameters reflecting wind property at the site of Quanzhou Bay HSR seacrossing bridge first based on long-term field collected wind data.Then,the main bridge of the sea-crossing Bridge is taken as the research object,and the short-term and long-term extreme buffeting responses of the bridge are analyzed.A train-bridge-wind interaction spatial stochastic system is also established to study the uncertainty of train running safety curve caused by the dual-randomness of turbulence wind and track irregularity.Finally,the efficient prediction approach for evaluating the extreme response of train-bridge-(wind)interaction system is discussed with taking a novel 3×70 m integral bridge as an example.The main contents and conclusions of this dissertation are shown as follows:(1)Probabilistically modelling of wind characteristics parametersThe meteorological observation mast located near the site of the seacrossing Bridge is introduced first.Then,the probabilistic models of the parameters reflecting mean wind and turbulence wind environment are constructed.The results show that the PDF of mean wind speed at the elevation of bridge deck is unimodal with a peak of approximate 4 m/s,whereas the wind direction show a multimodal distribution where the first and second main wind directions are 40.68 ° and 231.48 ° respectively.The mean vlues of the parameters in turbulence spectral,i.e.,A,b,B and e,are suggested as 7.19,0.78,26.07 and 1.06 respectively.(2)Extreme buffeting response of the main bridge of Quanzhou Bay HSR Bridge considering the non-Gaussianity of Typhoon BailuA hybrid method evaluating bridge buffeting extreme response under multi-dimensional environmental variables including turbulence stochastic high-order statistics is proposed first.Then the extreme buffeting response distribution of the main bridge under typhoon Bailu is predicted adopting the hybrid approach.The results indicate that: the non-Gaussianity of typhoon turbulence reduces the lateral extreme displacement at the mid main span(the mean value decreases by 25%),and increases the mean extreme vertical response by 20%.Finally,the sensitivity of non-Gaussian intensity and extreme value of bridge buffeting response on non-Gaussian intensity of wind turbulence are discussed.(3)Effect of non-Gaussian turbulence on long-term extreme buffeting response of the main bridge of Quanzhou Bay HSR BridgeA hybrid method combining artificial neural network(ANN)is proposed to evaluate the bridge long-term extreme buffeting response,and FLM(Full long-term method)is used to verify the accuracy and efficiency of the algorithm.Then,with the adoptions of the hybrid method and probability models of turbulence high-order statistics,the impact of turbulence skewness(kurtosis)on the long-term extreme buffeting response of the main bridge is analyzed.Results show the impact of turbulence skewness at the bridge site on the long-term extreme response is adverse but limited.In addition,compared to the case only considering the effect of turbulence skewness,the joint effect of skewness and kurtosis of component u or w on the long-term EVD(extreme value distribution)is also insignificant,where the increments of lateral and torsional buffeting responses are only 3.3% and 2.9% with MRI = 100 year.(4)Estimation of probabilistic train running safety curve over the main bridge under wind environmentA train-bridge-wind spatial interaction(TBWI)system considering the randomness of turbulent wind and track irregularity is established.Based on VPM(Virtual Process Method),the EVD of various system indexes under different wind speeds and vehicle running speeds are predicted.The result shows that Weibull probability model evaluates the EVD of TBWI interaction system well;additionally,the probability surface reflecting the train running safety over the Quanzhou Bay HSR bridge is analyzed,consequently the critical running speed corresponding to any probability can be obtained with considering different constraint criterions.The sensitivities of train and bridge EVDs on wind speed and running speed are also discussed.(5)Pilot research of efficient approach using for extreme response prediction of train-bridge wind interaction systemThe novel 3×70 m integral bridge adopted in new Fujian-Xiamen electrified railway is taken as the research object,the efficiency and accuracy of GMM(Global maximum method),VPM and ACERM(Average conditional exceedance rate method)in the extreme value response prediction of TBWI system are evaluated.The results show that both of VPM and ACERM are effective enough in the evaluation of bridge buffeting and TBI system extreme responses,and ACERM can be considered as the most efficient among the three algorithms.Finally,the comparisons of EVDs among three different bridge types show that the mean extreme responses of vertical displacements at the mid side span and mid middle span decrease by 25.2% and 42.5% respectively compared the integral bridge with continuous bridge and rigid-frame bridge.89 figures,27 tables and 243 references in total.