TY - BOOK

T1 - Condition Monitoring and Damage Prediction in Railway Applications

AU - Velic, Dino

N1 - no embargo

PY - 2022

Y1 - 2022

N2 - In this thesis a hybrid modelling approach is used to develop monitoring and prediction tools for railway components. The developed models shall facilitate a condition based and predictive maintenance strategy of switches and crossings with the ultimate goal of reducing the component life cycle costs. This work is subdivided into two sub tasks, which are defined as: •The development of a condition monitoring system for railway crossings •The development of a lifetime assessment and damage prediction tool for switch rails Dynamic FE models are used as digital twins of the analysed components. The structural response during the loading and the degradation mechanisms leading to component failure are thereby analysed. In addition, data-driven methods are used to improve the accuracy and efficiency of the developed FE models, as well as for statistical evaluation and advanced signal processing. The condition monitoring system for crossings is based on a virtual sensor concept, which is used for the detection of the transition zone (point) of the wheel during a passage through the crossing. A sensor concept based on strain and acceleration measurements is introduced and two structurally identical demonstrators with different lifecycle phases are instrumented. The instrumented demonstrators are used to test the proposed virtual sensor concept and to provide a Big-Data set of vehicle passages, which is used for subsequent data analytics. The switch rail damage prediction tool is implemented by means of cyclic dynamic FE calculations of a wheel passing a switch in the facing diverging direction, with an intermediate extrapolation algorithm to reduce the computational costs. By this approach the calculation time for 2 MGT of traffic is reduced by 96 %, which enables the execution of parameter studies with various switch rail geometries and materials. The results from the parameter study are used to enable a comparison of the expected service lifetime of the analysed switch rails under given loading conditions. The developed tools shall support track operators and component manufacturers during the designing and operating phase of vital infrastructure assets such as switches and crossings.

AB - In this thesis a hybrid modelling approach is used to develop monitoring and prediction tools for railway components. The developed models shall facilitate a condition based and predictive maintenance strategy of switches and crossings with the ultimate goal of reducing the component life cycle costs. This work is subdivided into two sub tasks, which are defined as: •The development of a condition monitoring system for railway crossings •The development of a lifetime assessment and damage prediction tool for switch rails Dynamic FE models are used as digital twins of the analysed components. The structural response during the loading and the degradation mechanisms leading to component failure are thereby analysed. In addition, data-driven methods are used to improve the accuracy and efficiency of the developed FE models, as well as for statistical evaluation and advanced signal processing. The condition monitoring system for crossings is based on a virtual sensor concept, which is used for the detection of the transition zone (point) of the wheel during a passage through the crossing. A sensor concept based on strain and acceleration measurements is introduced and two structurally identical demonstrators with different lifecycle phases are instrumented. The instrumented demonstrators are used to test the proposed virtual sensor concept and to provide a Big-Data set of vehicle passages, which is used for subsequent data analytics. The switch rail damage prediction tool is implemented by means of cyclic dynamic FE calculations of a wheel passing a switch in the facing diverging direction, with an intermediate extrapolation algorithm to reduce the computational costs. By this approach the calculation time for 2 MGT of traffic is reduced by 96 %, which enables the execution of parameter studies with various switch rail geometries and materials. The results from the parameter study are used to enable a comparison of the expected service lifetime of the analysed switch rails under given loading conditions. The developed tools shall support track operators and component manufacturers during the designing and operating phase of vital infrastructure assets such as switches and crossings.

KW - Condition monitoring

KW - Switches and Crossings

KW - Finite element modelling

KW - Wear

KW - Plastic deformation

KW - Wheel

KW - rail interaction

KW - Zustandsüberwachung

KW - Weichen

KW - Finite elemente

KW - Verschleiß

KW - Plastische Deformation

KW - Rad/Schiene Kontakt

M3 - Doctoral Thesis

ER -