Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 110
PROCEEDINGS OF THE THIRD INTERNATIONAL CONFERENCE ON RAILWAY TECHNOLOGY: RESEARCH, DEVELOPMENT AND MAINTENANCE
Edited by: J. Pombo
Paper 141

Selection of Sensitive Features for Damage Identification in a Railway Bridge

A. Meixedo1, V. Alves2, D. Ribeiro3 and R. Calçada1

1Faculty of Engineering, University of Porto, Portugal
2Post-Graduate Program in Civil Engineering, Federal University of Ouro Preto, Brazil
3School of Engineering, Polytechnic of Porto, Portugal

Full Bibliographic Reference for this paper
A. Meixedo, V. Alves, D. Ribeiro, R. Calçada, "Selection of Sensitive Features for Damage Identification in a Railway Bridge", in J. Pombo, (Editor), "Proceedings of the Third International Conference on Railway Technology: Research, Development and Maintenance", Civil-Comp Press, Stirlingshire, UK, Paper 141, 2016. doi:10.4203/ccp.110.141
Keywords: damage detection, features selection, railway bridge, moving loads, genetic algorithm.

Summary
This paper aims selects sensitive features for damage identification in a railway bridge. For this purpose, a damage identification methodology based on a genetic algorithm and its application to the numerical model of a railway bridge is proposed. The identification method is based on an iterative process that in each iteration compares the values of the dynamic responses of a bridge for a test damage scenario with the response values for a reference damage scenario. The introduction of damage in the test numerical model is based on the modification of preselected numerical parameters values. The genetic algorithm allows the identification of damage by minimizing the residue of an objective function constructed from the responses of the test and reference model, and based on damage indicators. In this study damage indicators, based on the deck accelerations of the bridge subject to the action of the Alfa Pendular train defined as a set of moving loads, are defined and tested. The proposed methodology for identifying damage has been tested using a two dimensional numerical finite element model of a railway bridge including the railway track. The results enabled the evaluation of the efficiency and reliability of the methodology and of the damage indicators which, in most situations, showed potential for the detection, location and quantification of the severity of the damage.

purchase the full-text of this paper (price £22)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description