Development of a damage detection framework for railway bridges using operational response

• Author / Creator

  Azim, Md Riasat

• Railway bridges are critical components of the railway infrastructure system. These bridges are subjected to various potential hazards resulting in different levels of damage. Therefore, developing operational response-based robust damage investigation strategies specifically tailored to railroad bridges is the goal of this doctoral research. In this research, non-parametric damage detection methods are proposed using operational acceleration and strain data. These methods only require basic information about the bridge (e.g. the type of the bridge, such as girder bridge or truss bridge, and its geometric configuration) during the setting up of the instrumentation/data collection plan. Once the system of operational response data acquisition is installed, the methods do not require further information from a numerical model or information about the design of the bridge. These methods then use the operational response to provide information about the structural condition of the bridge. The methods also incorporate operational variability in terms of train speed and loading.
  The research contributions of this thesis are discussed in three parts. In the first part, a new damage detection method using acceleration data is proposed for railway bridges. This method applies a sensor cluster approach to the time-series analysis of operational acceleration data recorded during and after the passage of trains. The damage feature is investigated based on the difference of fit ratio of the time series models fit to the measured free acceleration response of the baseline bridge and unknown-state bridge. It is shown that the proposed framework provides useful information on the existence, location, and relative severity of the potential damage in terms of nodal damage features.
  In the second part, damage detection methods are developed for railway bridges that utilize statistical analysis techniques such as coefficient of variation and principal component analysis employing operational strain response that extracts damage features from structural elements. The damage features are investigated as the difference of coefficient of variations and the difference of geometric distances in the principal component space respectively. The numerical results demonstrate that the proposed methods are effective in identifying, locating, and relatively assessing the severity of damage in instrumented truss elements. The damage features extracted from elements using strain-based methods and those extracted from the nodes using acceleration-based methods could be utilized as independent damage assessment tools as well as complementary tools to each other.
  In the third and final part, laboratory experiments are carried out on a steel deck type bridge and a timber truss bridge to validate the acceleration response-based damage detection method proposed in the first part of the thesis. The experimental results show that the presented method indeed has the potential for implementation in real-life bridges.
  Finally, the thesis discusses the limitations of the present research and recommendations for future research.

• Subjects / Keywords

  • Structural health monitoring
  • Damage detection
  • Operational response
  • Railway bridges

• Graduation date

  Fall 2020

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-fcmx-1q69

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.