Vehicles, pedestrians and flood risk: a focus on the incipient motion due to the mean flow Dissertation submitted to and approved by the Department of Architecture, Civil Engineering and Environmental Sciences University of Braunschweig – Institute of Technology and the Department of Civil and Environmental Engineering University of Florence in candidacy for the degree of a Doktor-Ingenieur (Dr.-Ing.) / In Civil and Environmental Engineering by Chiara Arrighi born 09/03/1986 in Fiesole (FI), Italy Submitted on 1st March 2016 Oral examination on 9th May 2016 Professorial advisors Prof. Fabio Castelli Prof. Hocine Oumeraci 2016 i UNIVERSITYOFFLORENCE Abstract FacultyofEngineering DepartmentofCivilandEnvironmentalEngineering DoctorofPhilosophy Vehicles,pedestriansandfloodrisk: afocusonincipientmotionduetothemeanflow byChiara ARRIGHI Floodsareoverflowingofwaterontoland,whichisnormallydryandareoneofthecostliest natural hazards. According to global scale reports, in the last decade floods affected the largest numberofpeoplewithrespecttootherhazardssuchasearthquakesordroughts.Besidethedam- agestostructuresandinfrastructures, floodsalsocausemanyfatalitiesandinjuries. Ithasbeen demonstrated that the majority of fatalities occurs as a consequence of inappropriate high-risk behaviours like driving and walking in floodwaters. In fact, vehicles can lose stability also for very low water depths and may turn into deadly traps. For flood risk managers, people safety istheprimaryobjective, butalthoughvehiclesaresocrucial, verylittleisknownaboutthecrit- ical conditions in which the onset of motion occurs. Besides, the existing instability criteria for pedestriansunderwaterflowsufferfromthelargescatterofexperimentalpairsofcriticalwater depthandvelocity. Asamatteroffact,theinstabilityconditionsofbothvehiclesandpedestrians are affected not only by flood parameters (i.e. water depth and velocity), but also by geometric andphysicalpropertiesof theobject. The mainaimofthisPhDresearch projectistobetterun- derstandtheinstabilitymechanismsforpedestriansandvehicles,whichareresponsibleformost of the casualties in order to introduce new hazard criteria capable of accounting for both flood andobjectcharacteristics.Forthispurpose,acomprehensiveanalysisofthecurrentknowledgeis firstlypresented. Secondly,theforcesactingonapartlyimmersedvehicleandhumansubjectare examinedandtwodimensionlessmobilityparametersareintroduced. Theexistingexperimental data on vehicles and people instability are used to identify a dimensionless critical threshold of incipient motion. Thirdly, a 3D numerical model in the OpenFOAM framework is adopted to clarify the role of hydrodynamic forces and determine relevant dimensionless parameters and scaling numbers involved in the instability mechanisms, considering the mean flow properties. Then, the results of the numerical simulations for a selected vehicle, which reproduce a set of existingexperiments,areanalysedanddiscussed. Finally,twocasestudiesarepresentedinorder todemonstratetheapplicabilityofthemobilityparameterstothefieldscaleandtheadvantages of hazard maps implemented with the proposed method. The results show that the scatter of existingexperimentaldatacanbeovercomeusingadimensionlessapproach,whichaccountsfor bothfloodsandobjectsproperties. Sincethecriticalthresholdsaredimensionless,theinstability conditionsforpeopleandvehiclescanbeeasilycomparedinordertodevelopbehaviouralrules, managementstrategiesandsupportpeopleeducation.Thenumericalmodelhintstheimportance offlowregimesforthehydrodynamiceffectsandclarifiesthecontributionofdragandliftforces. Theapplicationofthemobilityparameterstoinundationmapsdemonstratestheapplicabilityof themethodtothefieldscale. Thequalitativecomparisonbetweenthereconstructedinundation mapsandthegeoreferencedpicturestakenafterthe2014floodeventinGenoashowsaverygood accordance,thusthemethodappearspromisingandprovidesanimprovedbasisforfloodhazard andriskmapping. iii Acknowledgements The present PhD research has been conducted thanks to a PhD scholarship given by theUniversityofFlorence(Italy)whichisgreatlyacknowledged. Iwouldliketoexpress myspecialgratitudetoProf. Dr. IngHocineOumeraciandProf. Dr. IngFabioCastellifor allowingmetoconductthePhDstudyundertheirsupervisionintheLeichtweiss-Institut Hydraulic Engineering and Water Resources (LWI), Braunschweig, Germany and in the Department of Civil and Environmental Engineering, Florence (Italy) and for encourag- ing my research with their valuable suggestions. The support from Prof. Dr. Ing. Fed- ericoDomenichini(UniversityofFlorence)andProf. Dr. Ing. AbdellatifOuhasine(UTC, Compiègne) during the PhD research is acknowledged. A special thanks to Dr. Simone Gabellani (CIMA Research Foundation, Italy) and to Ing. Patrick Chassé (CEREMA, France) for providing the inundation model datasets. The laboratory Lasis at the Uni- versity of Florence is acknowledged for having provided the 3D numerical geometry of thesimulatedcarmodel. IwouldalsoliketothankmycolleagueJuanCarlosAlcèrreca- Huertaforhisfriendship,guidanceandassistanceduringmystayinBraunschweig. v Contents Abstract i Acknowledgements iii 1 Introduction 1 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 Expectedresultsandimpacts . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 Currentstateofknowledgeandmodelling 5 2.1 Floodriskinurbanareas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Impactsoffloods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.1.2 FloodriskdefinitionandEuropeanlegislation . . . . . . . . . . . . 6 2.1.3 Hazardmodelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.1.4 Damageassessment . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.1.5 Riskassessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.1.6 Peoplefatalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2 Experimentaldataonvehiclesandpeople . . . . . . . . . . . . . . . . . . . 14 2.2.1 Experimentsonincipientmotionofsmallscalevehiclemodels . . 14 2.2.2 Experimentsonpeopleinstabilityinfloodflows . . . . . . . . . . . 19 2.3 Numericalmodelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.3.1 Effectsofanobstacleinafluidflow . . . . . . . . . . . . . . . . . . 23 2.3.2 CFDnumericalmodelling . . . . . . . . . . . . . . . . . . . . . . . 24 2.3.3 CFDcodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4 Summaryandimplicationsfortheresearch . . . . . . . . . . . . . . . . . . 27 2.5 Specificationofobjectivesandmethodology . . . . . . . . . . . . . . . . . 27 2.5.1 Specificationofobjectives . . . . . . . . . . . . . . . . . . . . . . . . 28 2.5.2 Specificationofmethodology . . . . . . . . . . . . . . . . . . . . . . 28 3 Incipientmotionofparkedcarsandpedestriansunderwaterflow 33 3.1 Incipientmotionofparkedcars . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.1.1 Mobilityparameterofparkedcars . . . . . . . . . . . . . . . . . . . 33 3.1.2 Application of the mobility parameter to available experimental data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.1.3 Effectoffloworientationonthemobilityparameter . . . . . . . . . 38 3.1.4 Sensitivityanalysisofθ touncertaintiesinH andU . . . . . . . . 42 V 3.2 Incipientmotionofpedestriansstandinginfloodwaters . . . . . . . . . . . 43 3.2.1 Mobilityparameterofpedestrians . . . . . . . . . . . . . . . . . . . 43 3.2.2 Applicationofthemobilityparametertoexistingexperimentaldata 47 3.3 Comparisonbetweenmobilityparametersforvehiclesandpeople . . . . 48 3.4 Summaryofkeyresultsandimplications . . . . . . . . . . . . . . . . . . . 49 vi 4 Numericalmodelset-upinOpenFOAM 53 4.1 TheOpenFOAMframework . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.2 GoverningequationsoftheCFDmodel . . . . . . . . . . . . . . . . . . . . 55 4.3 Numericalmodelsetup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.3.1 StructureofasimulationinOpenFOAM . . . . . . . . . . . . . . . 56 4.3.2 Solver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.3.3 Meshgeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 4.3.4 Relaxationzones,boundaryandinitialconditions . . . . . . . . . . 61 4.3.5 Controlofthesimulationandforcescalculation . . . . . . . . . . . 61 4.4 Sensitivityanalysistotheappliedmeshsizeandturbulencemodel . . . . 62 4.4.1 Sensitivityoftheresultstothemeshsize . . . . . . . . . . . . . . . 62 4.4.2 Sensitivityoftheresultstotheturbulencemodel . . . . . . . . . . . 64 4.5 Testprogrammefortheparameterstudy . . . . . . . . . . . . . . . . . . . 65 4.5.1 TestprogrammefortheFordFocus . . . . . . . . . . . . . . . . . . 65 4.5.2 Testprogrammeforhumansubjects . . . . . . . . . . . . . . . . . . 65 4.6 ApplicabilityandlimitationsoftheCFDsimulations . . . . . . . . . . . . 69 4.7 Summaryofthechapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 5 Numericalresultsanddataanalysis 73 5.1 NumericalresultsofthesimulationsoftheFordFocuscar . . . . . . . . . 73 5.1.1 FlowaroundtheFordFocus. . . . . . . . . . . . . . . . . . . . . . . 73 5.1.2 ForcesandforcecoefficientsforscaleandprototypeFordFocusfor anglesofflowincidenceβ = 0◦ andβ = 180◦ . . . . . . . . . . . . . 75 5.1.3 Comparisonofnumericalresultswithexperiments . . . . . . . . . 82 5.1.4 Discussionoftheassumptions"rigidbodyandconstantfrictionco- efficient" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 5.1.5 Effectoftheangleofflowincidence . . . . . . . . . . . . . . . . . . 85 5.1.6 Forcesandforcescoefficientsfordifferentanglesofflowincidence 86 5.1.7 SummaryofthenumericalresultsoftheFordFocus . . . . . . . . . 89 5.2 Numericalresultsofthesimulationsofpedestrians . . . . . . . . . . . . . 92 5.2.1 Flowcharacteristicsaroundanuprightstandinghumanbody . . . 92 5.2.2 Forcesandforcecoefficients . . . . . . . . . . . . . . . . . . . . . . . 93 5.2.3 Identificationofmotionmechanisms . . . . . . . . . . . . . . . . . . 97 5.2.4 Comparisonwithexperimentaldataanddiscussion . . . . . . . . . 99 5.2.5 Summary and discussion of the results of the numerical simula- tionsofpeople . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 5.3 Summaryofthechapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 6 Applicationtotwocasestudies 105 6.1 Definitionofhazardcriteria . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 6.1.1 Hazardcriteriaforvehicles . . . . . . . . . . . . . . . . . . . . . . . 105 6.1.2 Hazardcriteriaforpedestrians . . . . . . . . . . . . . . . . . . . . . 106 6.2 TheTELEMAC-MASCARETsuite . . . . . . . . . . . . . . . . . . . . . . . 107 6.3 ThecasestudyofAjaccioinCorsica(France) . . . . . . . . . . . . . . . . . 108 6.3.1 Casestudydescription . . . . . . . . . . . . . . . . . . . . . . . . . . 108 6.3.2 Inundationcharacteristicsandhazardmapsforvehiclesandpedes- trians . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 6.4 ThecasestudyofGenova(Italy) . . . . . . . . . . . . . . . . . . . . . . . . 113 6.4.1 Casestudydescription . . . . . . . . . . . . . . . . . . . . . . . . . . 113 6.4.2 Inundationcharacteristicsandhazardmapsforvehiclesandpedes- trians . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 vii 6.4.3 Comparison of proposed hazard criteria with photos of the flood event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 7 Summary,conclusionsandoutlook 123 7.1 Generalsummary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 7.2 Mobilityparametersandnumericalmodelling . . . . . . . . . . . . . . . . 123 7.3 Limitationsoftheresults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 7.4 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 ix List of Figures 1.1 AnepisodeofflashfloodinAthens(leftpanel,source: www.timesofmalta.com) andtheaftermathofthe2011floodinGenova(rightpanel,source: www.adnkronos.com). 1 2.1 Humanimpactsbydisastertypeinthedecade2002-2012(EM-DAT,2012) 6 2.2 Hazard,vulnerabilityandexposure(fromMerzetal.,2004). . . . . . . . . 7 2.3 Standard flow chart for flood risk assessment based on flood depth maps andsocio-economicdata(fromArrighietal.,2013). . . . . . . . . . . . . . 8 2.4 The2013flashfloodinAthens(source: http://www.timesofmalta.com). . 14 2.5 The aftermath of Genoa 2011 flood event (source: http://www.italia45- 45.it/alluvioni). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.6 Experiments on fully (a) and partially (b) submerged Mitsubishi Pajero (1:18)(fromXiaetal.,2011). . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.7 Criticalthresholdvaluesofhydraulicinstabilityforvehicles(Teoetal. 2012) 19 2.8 Experiments on human instability: water depth 1.07 m and velocity 1.0 m/s(fromKarvonenetal.,2000) . . . . . . . . . . . . . . . . . . . . . . . . 21 2.9 Humanmodel(scale1:5.54)testedbyXiaetal. 2014 . . . . . . . . . . . . . 22 2.10 Proposedhazardregimesforpeople(fromRussoetal.,2013) . . . . . . . . 22 2.11 Summaryofthetopicsdescribedinthereviewofthecurrentknowledge. 30 2.12 MethodologyoftheresearchandstructureofthePhDthesis. . . . . . . . . 31 3.1 Forcesactingonaparkedvehicleunderwaterflow. . . . . . . . . . . . . . 33 3.2 Definitionofgeometricparametersofthevehicleandwaterflowparameters. 34 3.3 Mobilityparameterθ versusFroudenumberFr ofundisturbedflowus- V ingtheexperimentaldataofXiaetal. (2011)andShuetal. (2011). . . . . . 36 3.4 Experimental data by Shu et al. 2011 (left panel) and by Xia et al. 2011 (rightpanel)scaledtoprototypevalues. . . . . . . . . . . . . . . . . . . . . 37 3.5 Mobilityparameterθ versusFroudenumberFr ofthevehicleusingthe V v experimentaldataofXiaetal. (2011)andShuetal. (2011). . . . . . . . . . 38 3.6 Plan view of the surfaces affected by drag force according to the angle of flowincidence.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.7 Geometricschemeforavehiclewithangleofflowincidenceβ.. . . . . . . 39 3.8 Mobility parameter accounting for the angle of flow incidence (experi- mentswith90◦ anglerepresentedwithdiamonds,Xiaetal.,2013) . . . . . 41 3.9 Effectofadditionofasmallrandomnoisetothepairs(H,V)onthemobil- ityparameter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.10 Effectofadditionofalargerandomnoisetothepairs(H,V)onthemobility parameter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.11 Mobilising and resisting forces on a human body in floodwater (a) lateral view,frontalview(b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.12 Mobilityparameterθ versusFroudenumberFrfortheexperimentsfrom P publishedselectedstudies. . . . . . . . . . . . . . . . . . . . . . . . . . . . 47