Use Of Vibration Data For Structural Health Monitoring Of Bridges

USE OF VIBRATION DATA FORSTRUCTURAL HEALTH MONITORINGOF BRIDGESbyNicolas A. LondonoA thesis submitted tothe Faculty of Graduate Studies and Researchin partial fulfillment of the requirementsfor the degree ofDoctor of PhilosophyDepartm ent o f Civil and Environm ental EngineeringCarleton University, OttawaSeptember, 2006T he D octoral P rogram in Civil and E nvironm ental E ngineering is a joint program w ith theUniversity o f Ottaw a, A dm inistered by the O ttaw a-C arleton Institute for Civil E ngineering Copyright 2006, Nicolas A. L ondonoR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

Library andArchives CanadaBibliotheque etArchives CanadaPublished HeritageBranchDirection duPatrimoine de I'edition395 W ellington StreetOttawa ON K1A 0N4Canada395, rue W ellingtonOttawa ON K1A 0N4CanadaYour file Votre referenceISBN: 978-0-494-18224-6Our file Notre referenceISBN: 978-0-494-18224-6NOTICE:The author has granted a non exclusive license allowing Libraryand Archives Canada to reproduce,publish, archive, preserve, conserve,communicate to the public bytelecommunication or on the Internet,loan, distribute and sell thesesworldwide, for commercial or non commercial purposes, in microform,paper, electronic and/or any otherformats.AVIS:L'auteur a accorde une licence non exclusivepermettant a la Bibliotheque et ArchivesCanada de reproduire, publier, archiver,sauvegarder, conserver, transmettre au publicpar telecommunication ou par I'lnternet, preter,distribuer et vendre des theses partout dansle monde, a des fins commerciales ou autres,sur support microforme, papier, electroniqueet/ou autres formats.The author retains copyrightownership and moral rights inthis thesis. Neither the thesisnor substantial extracts from itmay be printed or otherwisereproduced without the author'spermission.L'auteur conserve la propriete du droit d'auteuret des droits moraux qui protege cette these.Ni la these ni des extraits substantiels decelle-ci ne doivent etre imprimes ou autrementreproduits sans son autorisation.In compliance with the CanadianPrivacy Act some supportingforms may have been removedfrom this thesis.Conformement a la loi canadiennesur la protection de la vie privee,quelques formulaires secondairesont ete enleves de cette these.While these forms may be includedin the document page count,their removal does not representany loss of content from thethesis.Bien que ces formulairesaient inclus dans la pagination,il n'y aura aucun contenu manquant.i*iCanadaR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

AbstractAccurate and timely assessm ent o f m aintenance and repair requirem ents o f bridges isim portant for ensuring public safety and for efficient allocation o f lim ited resources by roadtransportation authorities. Periodic visual inspections have so far been the typical approachin assessing the condition o f bridge structures. B ut the visual inspection approach issubjective as the results can vary from operator to operator, and it m ay fail to reveal hiddendeficiencies that m ay affect the integrity o f the inspected structures.Structural healthm onitoring (SHM) based on som e form o f structural response m easurem ent has em erged asan alternative m eans to provide a quantitative and m ore accurate structural conditionassessment. T he techniques o f m onitoring vibration responses to am bient loads, such asw ind and traffic are ideally suited for this purpose since theoretically any changes in stiffnessdue to damage or deterioration o f the m onitored structure are reflected in its vibrationpatterns. Furtherm ore, the m onitoring o f vibration responses o f bridges to operational loadscan be im plem ented cost effectively in relation to its potential benefits and does n o t lead toservice disruptions.This thesis presents and elaborates u p o n the different aspects o f data m anipulationsinvolved in vibration-based SHM, including data processing, system identification, damagedetection, and the developm ent o f com puter tools to facilitate data analysis, visualization andinterpretation o f results. O ne o f the m ain challenges in SHM using vibration data in practicalapplications in the field is the variability in the data arising from uncertainties in theenvironm ental and loading operational conditions, as well as errors in m odelling,m easurem ent and com putation, all o f w hich can obscure the effects o f damage ordeterioration o f the m onitored structure. T he question o f w hether the conditions o f largeiiR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

complex civil engineering structures in the field can be realistically assessed using vibrationresponse data has n o t so far been satisfactorily answered. This thesis examines this issue indetail using the m onitoring data from the C onfederation Bridge, in Canada.T heC onfederation Bridge is a large com plex bridge located in a severe m arine environm ent. T hem onitoring project o f the C onfederation Bridge provides the unique data used in this w orkto realistically assess the implications o f the actual field observed variability o f m onitoringdata on the feasibility o f using vibration response data for structural condition assessm ent.The results obtained through a finite elem ent updating dam age detection algorithm indicatethat continuous m onitoring using m ultiple independent datasets is necessary to offset theerrors arising from variability in the data to m ake it feasible to extract m eaningful and usefulinform ation regarding the existence, location and m agnitude o f changes in stiffnessassociated w ith damage or deterioration from the m onitoring data.iiiR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

AcknowledgementsI wish to express my gratitude to the people and institutions w ho m ade this thesispossible. A bove all, to the D epartm ent o f Civil Engineering at Carleton University for thefacilities it offers for research, in particular to the Faculty o f G raduate Studies for theScholarship and T eaching A ssistantship w hich provided m e w ith the necessary funding topursue graduate studies and for awarding the Jag M ohan G raduate Student Fellowship forCivil and E nvironm ental Engineering for 2004-2005 and the Jo h n Adjeleian G raduateScholarship for 2005-2006.I owe special thanks to my thesis supervisor, P rofessor D avid T. Lau, w ho gave m ethe opportunity o f participating in the C onfederation Bridge Research Project as a ResearchAssistant. I profited from doctor Lau’s careful and thoughtful observations, and from hisconstant support and advice at all stages o f my work, b o th in Canada and abroad w henattending various international conference presentations.I w ould like to thank Mr. D onald M cG inn and the staff o f Strait Crossing BridgeLimited for their assistance during various trips to the C onfederation Bridge, as well as Jo h nEgan and Alain Solari o f Public W orks and G overnm ent Services o f Canada for theirtechnical assistance, b o th w ith rem ote data collection and w ith in situ guidance at them entioned bridge. I also w ant to acknowledge the help o f my colleague Serge L. D esjardins,w ho in 2004 devoted him self fully to w ork w ith m e on the developm ent o f the softwareplatform for the processing and analysis o f the m onitoring data. I appreciated his family’shospitality at M oncton during one o f my trips to the bridge.ivR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

I w ould like to acknowledge the assistance by Professor G uido D e Roeck from theCatholic University o f Leuven, Belgium, w hose detailed observations o n the finite elem entupdating m ethod were instrum ental for my research in that area.My deepest gratitude goes to M arcela D ereix for her com pany and support while welived in O ttaw a; to my m other and relatives in Colom bia, for their unconditionalencouragem ent and support; and to my life-long friends D ario Jaram illo, Juan Camilo Sierra,Juan D avid Ram irez and Federico Carrillo for their generosity and for always being there.Finally, I w ant to m ention my friends in O ttaw a, especially Ali and Julie, Stelios andKaterina, Jose and Silvia, K ate, Tuan, Ryan, Viet and Linh, Freddy and Claudia. They m ademy years in Canada a rewarding and unforgettable experience.vR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

ContentsAbstract. iiAcknowledgements.ivContents.viList of tables. xiList of figures. xiiChapter 1 . Introduction. 11.1B ackground.11.2Problem sta tem en t. 21.3Objectives and thesis o u tlin e. 4Chapter 2 . Characteristics of dynamic monitoring data and observedbehaviour of the Confederation Bridge.72.1Introduction.72.2Continuous m onitoring system . 102.3Monitoring data. 112.4Typical vibration beh aviou r.122.4.1High-w ind scenarios.142.4.2 Traffic triggered scenarios. 142.4.3 A m bient scenarios. 152.4.4 A m bient w ith ice scenarios. 152.5D ata analysis.162.5.1D ata p ro cessin g . 16viR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

2.5.22.62.72.8System identification. 17Finite elem ent m o d e ls. 222.6.1Shell m o d e l. 242.6.2Beam m odel. 24Analysis results from m onitoring d a ta . 252.7.1M odal frequencies. 252.7.2M ode shapes. 282.7.3M odal dam ping ra tio s . 31C on clu sion s.332.9 A ck n ow led gem en ts. 35Chapter 3 . Variability of dynamic properties from field data andimplications for health monitoring. 503.1 Introduction.503.1.1C ontinuous m onitoring system .543.1.2System identification.553.2 Potential sources o f variability. 563.3 Baseline variability.623.3.1D a ta . 633.3.2D ata p ro ce ssin g . 653.3.3System identification.663.3.4R esults. 673.4 Variability under environmental & loading fluctuations. 693.4.1D a ta . 703.4.2D ata processing and system identification. 71viiR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.

3.4.33.53.6R esults. 71Variability o f m odal parameters vs. sensitivity to d a m a g e. 763.5.1Results —pier stiffness degradation scenarios. 773.5.2Results —damage at continuous drop-in span joints.79Variability m o d el.803.6.1A m plitude dependent m ode shape variability m o d el.803.6.2N oise sim ulation for dam age d e te c tio n . 813.7 C o n clu sio n s. 863.8 A ck now ledgem en ts.90Chapter 4 . Application platform for processing, analysis and visualizationof Confederation Bridge monitoring data. 1234.1 Introduction. 1234.2 Confederation Bridge m onitoring d ata. 1274.3 Data p rocessin g. 1304.4 Data visualization. 1344.4.1Real-time visualization. 1354.5 D ata analysis. 1364.5.1Spectral analysis. 1364.5.2Stochastic subspace identification application to o l.1374.6 Research applications. 1384.7 C on clu sion s. 138Chapter 5 . Damage detection techniques. 1455.1 Introduction. 1455.2 D am age detection via finite elem ent u p dating.147viiiR eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.