| Due to the uneven spatial and temporal distribution of water resources and the imperfect construction of water conservancy facilities,water security problems are prevalent in the southwest of China,and in order to solve the problem large-scale construction of water conservancy facilities must be built.However,most of southwest China is a high-intensity earthquake prone area.The safety of hydraulic buildings under the effect of strong earthquakes needs to be guaranteed.Under the second generation of performance-based seismic design concept,an effective way to study its seismic performance is to analyze the seismic vulnerability of various hydraulic structures.The seismic susceptibility studies on underground inverted siphon structures are less compared to aboveground hydraulic structures such as ferries and dams,which is very detrimental to the construction and development of underground inverted siphon structures.Therefore,it is necessary to carry out an in-depth seismic vulnerability analysis of underground inverted siphon structures.In this paper,Xiazhuang subsurface inverted siphon structure in Yunnan is studied,and a nonlinear dynamic analysis based on vector ground vibration intensity parameters is carried out using the IDA method,and then an analysis of the seismic vulnerability of the subsurface inverted siphon is carried out.The specific work is as follows:1.16 scalar ground vibration intensity measures(IM)were selected to carry out soil-structure dynamic nonlinear finite element calculations on the underground inverted siphon structure using the IDA method.Based on the calculated results,a log-linear fitted regression was performed for each scalar IM with the subsurface inverted siphon structure damage measures(DM).The four parameters of validity,usefulness,effectiveness and relevance are used as evaluation indicators,and the fitting results are used to evaluate each scalar IM.It was found that peak acceleration(PGA)is the optimal scalar IM parameter and the next best scalar IM parameter is acceleration spectral intensity(ASI)when using the maximum diameter deformation rate of the tube body of the underground inverted siphon structure as the structural damage indicator DM;the peak velocity(PGV)is the optimal scalar IM parameter and the next best scalar IM parameter is the characteristic strength(I_C)with the maximum inter story displacement angle of the underground inverted siphon structure as the structural damage indicator DM.2.Based on the preferred results,a subsurface invert siphon seismic susceptibility curve based on the optimal scalar IM parameter PGA was developed.The two scalar parameters,PGA and ASI,were obtained by prioritizing the ground shaking intensity parameters,and the vector parameter IMs was obtained by combining the two,based on which the subsurface inversion siphon seismic susceptibility surface was established.The results of the comparative analysis show that the susceptibility surfaces established based on vector parameters IMs contain more information on ground vibration characteristics,and the vector parameters can better evaluate the seismic performance of underground inverted siphon structures.3.GA-BP neural network is built on the basis of back propagation(BP)neural network optimized using genetic algorithm(GA)method.The preferred vector parameters IMs were used as the input values of the GA-BP neural network,and 30%of the finite element calculation results were randomly selected as the training sample set to train the GA-BP neural network.The predicted values with correlation coefficients above 0.9 with the finite element calculation results were obtained,and the susceptibility surfaces were established.The results show that using the preferred vector parameters IMs as the input values of GA-BP neural network for seismic response prediction can effectively reduce the computational cost of IDA-based subsurface inversion siphon seismic susceptibility analysis. |
PDF Full Text Request