| Recent years have seen a steady growth in the size and running speed of China’s high-speed rail network.The bogie frame,as a critical component of a high-speed train,operates against complex forces exerted by car body,rail-wheel contact and other components.Therefore,field tests and simulation modelling investigating the loads of bogie frame mushroom,and large amount of data ensue.This situation makes it even more challenging to organise,synergise,and analyse the data.Hence,a load database and its corresponding software are required if one is to pool massive data and extract multiple load features out of them.The bogie frames of Fuxing trains were identified as the topic of this dissertation.To carry out the research on the characteristics of bogie frame loads,data processing code was compiled,and both field tests and dynamics simulation were conducted.Moreover,a bogie frame load database was established based on how data is collected and analysed.Finally,an integrated platform for the management and analysis of load data was built.Main contents of this dissertation are as follows:(1)Based on Matlab environment and load processing and analysing routine,multiple algorithms were programmed,and their functions cover file reading,filters,frequency spectrum,zero drift removal,filtering of the facility-oriented disturbance,spike elimination,histograms describing rainflow or peak counting results,curve fitting of distribution functions,measures of goodness of fit,and kernel estimation.Besides,a desktop GUI was designed to visualise data processing.With these algorithms,a case study focusing on CR400 AF and CR400 BF was carried out.To be more specific,a unified taxonomy was employed to compare and analyse the time-domain,frequency-domain,and statistical characteristics of 14 types of bogie frame loads of the two EMUs.(2)Using My SQL,a bogie frame database was established after surveying the data collecting and simulation procedures.The database encompasses information on the project,tested objects,test detail,data files,and analytic files,etc.Then,an interactive software for the database was created with Windows Forms.NET,so that users can access and process well-grouped files from either a local or a remote server.By synergising those works with the desktop GUI for load analysis,an integrated platform for load data management and analysis of was completed.(3)A dynamics model of CR400 BF,the topic EMU of this dissertation,was built in Simpack software.Whether this model could yield bogie frame loads in line with the experimental data in terms of index values and index-velocity relationship were verified with rainflow counting and peak counting results.To evaluate how loads behave when the train passes on a curve with a certain speed,5 bogie frame loads were selected since they are relatively more vulnerable to curve negotiation.The loads were simulated with the model,and then broken down to low-frequency trend and high-frequency vibration in an effort to calculate the load-governed optimal superelevation.Furthermore,the model was run under different speed band,and normalised continuous curves demonstrating the relationship between the equivalent load and velocity were obtained by calculating the3.5th-order moments of the rainflow counting range and conducting polynomial curve fitting.By adapting these curves to data collected along Beijing–Shanghai railway,a prediction of the equivalent load at 400 kph speed band was made for the corresponding stretch.The integral data management and analysis platform presented in this dissertation can likewise be applied to a wide spectrum of railway vehicle data,for instance dynamic stress and vibration data.Apart from that,the proposed method of the equivalent load prediction and the analysis of the load characteristics when negotiating curves can shed some light on the endevour to further raise the speed of EMUs.There are 58 figures,11 tables,and 84 references. |
PDF Full Text Request