| The pipeline system plays a key role in transporting reactor coolant in the nuclear power system.China's geological disasters occur frequently.Excessive deformation of pipelines caused by earthquakes leading to coolant leakage and other serious threats to the safe operation of nuclear power plants.Therefore,the reliability evaluation of the pipeline system under seismic conditions is of vital importance.The reliability assessment of pipeline systems under earthquake conditions is mainly obtained by simulation using finite element models.The commonly used method is to simulate the vibration response of a particular pipe system under a certain seismic condition.In the reliability evaluation of the pipeline system,key data such as failure probability cannot be obtained through simulation by a single model.At the same time,the boundary conditions need to be set in the process of building the finite element model of the structure.The setting of boundary conditions is accomplished by setting certain key parameter values,such as the fixed rotational stiffness of the two pipes.Such parameters can often not be accurately obtained through on-site measurement and other means,and these structural parameters have a large impact on the structural response.The current method of establishing a finite element model is to idealize such key parameters,such as the rotational stiffness of the two fixed ends as infinity.However,the rotational rigidity of the two fixed ends of the actual structure of the pipe structure is not infinite,which causes the deviation between the finite element model and the actual structure.Aiming at these problems,this paper proposes a reliability evaluation method for pipeline systems based on the identification of key parameters of the finite element model under seismic conditions.The specific research content is as follows:(1)The background and significance of the reliability assessment of the pipeline system under seismic conditions and the parameter identification methods for the finite element models at home and abroad are introduced.At the same time,it researches and compares the commonly used reliability evaluation methods and their advantages and disadvantages.(2)In view of the shortcomings of current parameter identification methods,try to directly use time-domain vibration acceleration data for parameter identification.This paper proposes a method that can directly use the time-domain vibration acceleration data for parameter identification.Among them,with regard to the problem that the time-domain vibration acceleration data is too large to be used directly,this method uses the principal component analysis(PCA)dimension reduction method to effectively reduce the dimension of the data.When constructing the mathematical model of the identification parameters,after comparing the problems and deficiencies in the BP and GA-BP neural networks commonly used in the current research,the MEA-BP neural network was selected to fit the nonlinear relationship in the mathematical model.The validity of the mathematical model is verified by numerical simulation.After identifying the structural parameters,this parameter is used to build a high-precision finite element model.Based on this finite element model,a reliability evaluation method considering structure size,working conditions or randomness of materials is proposed.For the random sampling problem of different parameters,this method uses Monte Carlo sampling method to solve.(3)Establish a general pipeline representative structure,and perform seismic station tests on it to obtain its vibration response data under different seismic conditions.Using the proposed method to directly identify key parameters using time-domain vibration acceleration data to identify the elastic modulus of the pipe joint and the spring stiffness of the two fixed ends in the finite element model of the structure,a high-precision finite element of the pipeline structure was established based on model.In order to verify the accuracy of the finite element model,the study simulated and experimented the structure under the same working conditions.The simulation data and experimental data are compared to verify the reliability and effectiveness of the proposed parameter identification method in practical structural applications.At the same time,it is compared with other parameter identification methods in the current research to verify the advantages of the proposed method.The research carried out simulation and experiment by adding load to the structure and choosing another seismic condition.To further verify whether the identified parameters can still meet the accuracy requirements when the structural mass distribution changes and the seismic conditions change.The comparison between simulation data and experimental data shows that the structural parameters identified by the proposed method are reliable and accurate.(4)Based on the results of the elastic modulus and spring stiffness of the two ends of the pipe joint identified above,a finite element model of the general structure of the pipe is established.The accuracy of the finite element model has been verified.Next,through the finite element model,the Six sigma module of Workbench software is used to analyze the reliability of the pipeline based on the randomness of the pipeline structure size,material and load.The comparison and analysis of pipeline reliability under different earthquake conditions and different load quality conditions prove that the structure cannot be reliably operated under rare earthquake conditions of magnitude 8 and above.This research proposes a method that can directly use time-domain vibration data to identify pipeline structural parameters and perform reliability assessment.The simulation verification and the experimental verification of the real structure show that the proposed method has good feasibility and reliability.It is of great significance to improve the accuracy of the reliability assessment of pipeline structures in practical engineering applications. |
PDF Full Text Request