Ground Motion Truncation And Anti-dislocation Method Of Cross-fault Tunnel

For structures such as super high-rise building structure and tunnel in the seismic area,nonlinear time history analysis is one of the common methods for evaluating the seismic performance of structures,and it is also considered to be a more accurate method.However,the nonlinear time history analysis of the structure is time-consuming because of the complexity of structure and long ground motion duration.Therefore,it is important to reduce the calculation cost and improve the efficiency of nonlinear time history analysis.And the tunnel probably needs to cross the active fault.For tunnels in the anti-dislocation area,selecting a reasonable anti-dislocation method can reduce the damage of the tunnel lining.Therefore,the seismic methods and anti-dislocation methods has been studied in this paper.For complex structures in the seismic area,a deep-learning neural network for predicting ground motion duration is proposed for the purpose of improving the efficiency of nonlinear time history analysis.For tunnels in the anti-dislocation area,the numerical verification method is used to analyze the anti-dislocation effect of composite seismic joints for the purpose of proposing a reasonable anti-dislocation method.The main research contents are as follows:(1)In this study,a deep-learning neural network for determining ground motion duration is proposed.This method takes the criterion that the maximum displacement of structure before and after the ground motion truncation is not change,and considers the influence of the period elongation,the influence of high order modes and the uncertainty in estimating the structural yield strength.The deep-learning method can give a structural period dependent prediction result of ground motion duration.This deep-learning model uses 80280 samples for training and prediction,then the model is applied to analyze the errors of maximum story drift ratios of the 4-story and 16-storey structures respectively,and compared with the errors obtained from the widely used 95% significant duration and 75%significant duration as input ground motions.The results show that the prediction method of ground motion duration based on deep learning is simple and easy to calculate.And the method is applied to pulse-like ground motion to study the generalization ability,the result show that the deep-learning method also has good application to pulse-like ground motion.(2)In order to study the effectiveness of the deep-learning-based ground motion duration determination method,extend the application of this method to super high-rise structures and tunnels through active fault.As a result,compared with the original record,taking the truncated record as an input can improve the efficiency of nonlinear time history analysis and the maximum displacement of structure before and after the ground motion truncation is not change.The calculation efficiency of the super high-rise structure is lower due to its long basic period,and the maximum truncation rate is 79.12%,the minimum truncation rate is12.06%.The calculation efficiency of tunnel is improved due to its shorter basic period,and the truncation rate is about 80%.(3)The anti-dislocation method of tunnel which through the active fault is studied in this paper.In order to verify the seismic joint setting method of tunnel,the deformation of seismic joint,the equivalent plastic strain of tunnel lining and the damage factor are researched.The result show that combined seismic joint is generally applicable according to the fault information and the tunnel information,and can improve the dislocation resistance performance of the tunnel.