EUR18905 j g ^ g TRANSPORTRESEARCH COST331 Requirements forHorizontal RoadMarking FinalReportoft heAct i on MOTORWAY ROAD MARKINGS RIGHTEDGE XROADI h>MARKING RIGHT EDGE| ROAD MARKING i???h i BffitfiirStot?miiMS8!n!&o&..epzszt, f>oiazyA, a European Co-operation in the Field of Scientific and Technical Research 0035320 COST 331 Requirements for Horizontal Road Marking FinalReportoftheAction European Commission Directorate General Transport PARLEMENTEUROPEEN LUXEMBOURG CENTRE DEDOCUMENTATION N.c.ocnP ^ ^ I / V C.L uv LEGAL NOTICE Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of the followinginformation. The views expressed in this publication do not necessarily reflect the views of the European Commission. A great deal of additional information on COST Transport is available on the World Wide Web. It can be accessed through the CORDIS server at the following address: http://www.cordis.lu/cost-transport/home.html Cataloguing data can be found at the end of this publication. Luxembourg: Office for OfficialPublications of the European Communities, 1999 ISBN 92-828-6506-1 European Communities 1999. Printed in Belgium ACKNOWLEDGMENT The COST331 ManagementCommitteewouldlike to thankall public roadadministrationswho contributed with financing and human resources to the success of this research. Morespecifically, the Management Committee would like to express their thanks to the public administrations listed belowwithoutwhosecontributionandsupporttheexecutionofthefieldexperiments,which constitutes an essential part of the Action, would not have been possible: Finland Finnish National Road Administration Portugal Junta Autnoma de Estradas Laboratrio Nacional de Engenharia Civil Cmaras Municipais de Grndola e Alccer do Sal Preveno Rodoviria Portuguesa (Delegao de Setbal) Norway Norwegian Road Directorate Sweden Swedish Road Administration and Switzerland Dpartement des travaux publics de la rpublique et canton de Neuchatel, Service des ponts et chausses. Gendarmerie cantonale neuchteloise. Commune des Ponts de Martel. Table of Contents Table of Contents CHAPTER1.INTRODUCTION7 CHAPTER 2.GENERAL DESCRIPTION OF THE ACTION9 CHAPTER 3.EXECUTIVE SUMMARY13 CHAPTER 4.STATE OF THE ART.15 4.1INTRODUCTION15 4.2RESULTS OFTHE QUESTIONNAIRE16 4.3CONCLUSIONS21 CHAPTER 5.VISIBILITY OF ROAD MARKINGS23 5.1INTRODUCTION;23 5.2CALCULATION PROCEDURE23 5.3CALCULATION OF TARGET SEE25 5.4CALCULATION OF LUMNANCE27 5.5EXAMPLES OF APPLICATION30 5.6VALIDITY OF THE CALCULATION MODEL36 CHAPTER 6.DRIVERS' NEEDS OF PREVIEW TIME37 6.1INTRODUCTION37 6.2METHODOLOGY37 6.3ANALYSIS OF THE RESULTS43 6.4CONCLUSIONS61 CHAPTER 7.DRIVER BEHAVIOUR65 7.1INTRODUCTION65 7.2METHODOLOGY66 7.3RESULTS72 7.4CONCLUSIONS79 CHAPTER 8.DESIGN OF ROAD MARKINGS81 8.1INTRODUCTION81 8.2CRITERIA FOR DESIGN81 8.3VISUAL CRITERIA FOR ROAD MARKING MAINTENANCE85 CHAPTER 9.CONCLUSIONS87 CHAPTER10.BIBLIOGRAPHY89 10.1INTERNATIONAL AGREEMENTS89 10.2ARTICLES AND OTHER DOCUMENTS89 10.3CIEREPORTS96 10.4EUROPEAN STANDARDS96 ANNEX AVISUAL DATA FOR THE CALCULATION OF VISIBILITY LEVEL97 ANNEX BDRIVING EXPERIMENT103 ANNEX CEXAMPLE OF CALCULATION OF VISIBILITY DISTANCE115 ANNEX DVISIBILITY OF LONGITUDINAL ROAD MARKINGS IN HEADLAMP ILLUMINATION117 D.lINTRODUCTION AND DISCUSSION117 D.2GEOMETRYOF LONGITUDINAL ROAD MARKINGS,118 D.3CALCULATIONS AND RESULTS122 D.4DISCUSSION OF RESULTS122 COST 331 APPENDIX1 - MEMBERS OF THE MANAGEMENT COMMITTEE129 APPENDIX 2 - MEMORANDUMOF UNDERSTANDING - TECHNICAL ANNEX133 APPENDIX 3 - COST TRANSPORT OVERVIEW137 APPENDIX 4 - COMPUTER PROGRAMME FOR THE VISIBILITY DISTANCE OF ROAD MARKINGS139 A4.1INTRODUCTION139 A4.2DRIVER,VEHICLE AND GLARE141 A4.3ROAD GEOMETRY144 A4.4HEADLAMP ILLUMINATION AND COEFFICIENTS OF RETROREFLECTED LUMINANCE RL145 A4.5DAYLIGHT/ROAD LIGHTING AND LUMINANCE COEFFICIENT IN DIFFUSE ILLUMINATION QD147 APPENDIX 5 - COST 331 CD-ROM149 INDEX151 Chapter 1 Chapter 1.Introduction Inthe15MemberStatesoftheEuropeanUnion,45,000peoplearekilledeachyearin trafficaccidents;thatmeans900everyweek,and1.6millionareinjuredresultingin0.5 millionadmissionstohospitals,ofwhich25% resultininvalidity.Roadtrafficaccounts forsome 95% of all persons killed in transportaccidents. Onthebasisofcurrentfiguresandwithoutchangesinpolicies, practicesandbehaviour1 in 80 people now living in the EU will dieas a resultofa roadaccidentandabout1 in 3 of theUnion'scitizenswillneedhospitaltreatmentintheirlifetimebecauseofinjuries sustainedinroadaccidents.Thepainandanguishcausedbytheserealitiesisobviously beyondmeasurement, but there is a huge economic price to be paid as well. The economic costs arising frommedical expenses, police and emergency services,damage topropertyandlosteconomicoutputofthekilledandinjuredpersonsamounttoabout EURO 45 billiona year,somethinglike0.75% to1% oftheGDP.Theaverageeconomic costofeachpersonkilledintrafficaccidentsamountstoEURO1million.Other, apparentlymorerealisticcalculationsestimatethecostofroadaccidentsatleastatEURO 100 billion annually and the total socio-economiccosts exceeding EURO160 billion. Thisunaffordablesituation,inmereeconomictermsnottomentionunmeasurablehuman suffering,leadsmoreandmoretogrowingacceptancethatawiderangeofstrategiesis neededtoaddressthe problem.Thetrafficsystemhastoadapttotheneeds,mistakesand vulnerabilitiesof roadusersratherthanthe otherwayaround. Itisassumedbytrafficexpertsthatroadcasualtiesarecausedbyfailuresinthetraffic systemasawhole(whichincludesroadusers'decisionsandactions,infrastructureand vehicles)andcanonlybereducedeffectivelybyadoptingasystematicapproachtothis problem.Roadsafetyisacomplexandinterdisciplinarysubjectinwhichthevarious factors(driver,car,infrastructure)playanimportantroleandinteracttoasignificant degree. Latertrendsinroadsafetylook,moreandmore,towardsLowCostroadengineering Measures(LCM), suchas minor changes injunctionoperation, road lay-out, lighting,signs andmarkingswhichcanbeimplementedquicklyandmakesignificantcontributionsto roadsafety. Withintheabove mentionedsetofLCM,those concerningroadsigningingeneral(andin particular,roadmarkings),havebeentraditionallyconsideredinterestingalternativesto improve roadsafety. Unfortunately,thispotentialbenefit-andwellproveneffectiveness-ofroadmarkingsis not sufficientlyexploitedby the relevantdecisionmakers. This is well demonstratedby the factthat, at present,most national regulationsor technicalspecificationson thissubjectlay downminimum values forthe parameterswhichdefinetheir essential characteristics(night anddaytimevisibilityandskidresistance)withoutalwaystakingintoaccountthereal visual demandsof drivers. COST 331 Therefore,there is an urgent need to establishan up-to-date scientificmethod with which, on the basis of drivers'visual needs, to determine the optimum pavement marking design in order to ensure that it isvisible, by day and by night, in all weather conditions.Only after having validated uniform criteria for the appropriate design of the road markings, will driversbeabletoenjoyaharmonisedqualityofroadmarkingscapableofpositively contributing to improve their road safety level in Europe. This level of safety throughout the European road network is a right fordriversaswell asanobligation fortheEuropeanUnionauthoritiesasstatedinSingleMarket legislation in 1987 (Article 100 a) and the Treaty of Maastricht in 1993(Article75). Chapter 2 Chapter 2.General description of the Action The final purpose of COST 331 is to establish an up-to-date scientificmethod with which, on the basis ofdrivers'visual needs, to determine the optimum pavement marking design in order to ensure that markings are visible, by day and by night, in all weather conditions. At present, most research in this area, both national and international, concentrates on: Developmentof new marking products which meet the above mentionedrequirements for as long as possible (maximum functionallife), and Designofnewtechnologiesforthemanufactureofhigh-performanceequipmentfor assessing those requirements. Scientificallyvalidated uniformcriterionare therforeneeded for the appropriate design of road markings in order to provide the following "benefits" to differentusers: For drivers: Optimisation of the cognitive load; Achievement of visual guidance by day and night in all weather conditions For road authorities: Availability of a scientifically validated methodology for designing road markings; Maximisation of the cost-effectivenessof road markings. Thiswork hasbeenconfinedto longitudinalroadmarkings(oftwotypes:conventional andthosedesignedtomaintainnighttimevisibilityinadverseweatherconditions) includingdirectionalarrows.Thecolourofroadmarkingsisnotconsidereditselfas a variable to be included in the differentexperimental phases. The application field of the road markings covered by the Action is restricted to interurban roads(motorways,dualcarriagewaysandsinglecarriageways).Therefore,pavement markings applied in urban areas are outside the scope of this research. For its execution, the research programme was subdivided into four main sections, linked amongthemselves,dealingwiththenecessarytasksidentifiedinthepreparationof COST 331. Section1 (describedinChapter4)dealswiththe"stateoftheart", intheareafalling within the Action's scope.To that purpose, an appropriate questionnaire, to be filled in by meansofsubsequentinterviewswithrelevantdecision-makersinEurope,hasbeen prepared and circulated.This section includes a summary of the answers provided as well as the more remarkable resulting conclusions. COST 331 Section2(coveredbyChapters5and6,respectively)describesthebackgroundtothe objectiveof"designingamathematicalmodelforthecomputationofvisibility distancestoroadmarkings" andtheidentificationof"driver'svisualneeds".On the one hand, road markings supply certain visibility distances to drivers depending on the road markingsthemselvesaswellasontheconditionsand,ontheotherhand,thesupplied visibility distances may or may not be sufficientin view of the driver'sdemand related to the intended purposes of the road markings. Thefirstsubtaskrelatestobasic visual performancedata andprovidesamathematical model forthe computationof visibility distancessuppliedby roadmarkings.The model includesthegeometricallay-outandthereflectionpropertiesofroadmarkings,and conditions of illumination and observation. Thesecondsubtaskclarifiesdrivers'demands ofvisibilitydistances.Itisfocusedon visual guidance suppliedby existing pavementmarkings.The demandsare expressedas preview times from which visibility distances can be computed according to driving speed. Section 3 (described in Chapter 7) deals with field experiments intended to "evaluate the impactonroadsafetyofroadmarkings"bymonitoringthebehaviourofdifferent (selected) drivers through experimental road sections properly designed and marked. It was not practicableto evaluatethe benefitsofroadmarkings by monitoringaccidents. Theinfluencesofmarkingsonroadsafetywerethereforeinferredfrombehavioural changes. These behavioural effects were measured on two levels: adaptation of individual drivers to road markings and effectsof road markings on the speed of traffic flow. Two primarygroupsofvariableusedto describesafetyeffectsondriverbehaviourwere speed related factorsand the lateral stability of a vehicle in relation to the centre/edge line. Problemsassociatedwiththelateralpositionofavehicleanditsrelationtosafetyare treated in this section.It is assumed that the higher the speed level and variability, and the greateranysuddenshiftsinthelateralpositionofavehicle,thelesssafethedriving behaviour. Section4(describedinChapter8)provides"aguidefortheuseofroadmarking elements on roads of differentclasses".The contrast, the connection between width and reflectionand other geometrical measures are stated for differenttypes of longitudinal lines usedinseveralcountriesinEurope,onmotorways,dualcarriagewayandsingle carriageway roads. The guide shows how to combine those elements, in order to provide needed visibility, and offersuser-friendlycomputersupportwiththemethodologytocalculatethevisibility distance provided by road markings. Infigure2.1, thedifferenttasksandsub-tasksdevelopedinCOST331aswellasthe execution time programmed for each of them have been tabled. 10 Chapter2 ID 0 1 2 3 4 5 6 7 e g 10 11 12 13 14 15 16 17 16 19 Task Nome COST 331 Action'sStartrigDate Action'sClosing Date Task100 - St at eof the art Tas k 200- Surveyin the praticipatingcountries Tas k 300- Evaluationof driversv i sualneeds Task300a Task300b Tas k 400- Impacton roadsaf et yof roadmarki ng Planningof theresearch Rottest Testin Finland Testin Portugal Testin Switzerland Analysts of data Reporting Tas k 500- Impactof roadstuds Tas k 600-Guldelinesfor roadmarki ng Tas k 7 00-FinalReport 1995 Q1|02|03|Q4 +26/07 26*7m 1 1/01 1996 Q1|Q2|Q3|Q4 29/12 1997 Q102|Q3|04 | ^ | ^ _28/06 2/09g30)09 1998 Ol1Q21031Q4 1JB9^31/10 1/08_30/09 2/03g31/03 13/11_ 1999 Q1I02IQ3I04 42S/04 31/12 ^J15/04 1/12^ ^ ^ ^ ^ ^ g ^26/11 23/ 1*5 30JQ4 Figure 2.1- COST 331 GANTTchart ThemembersofCOST331werefrom15COSTcountries:Belgium,Denmark,Finland, France,Germany,Greece,Iceland,Italy,Portugal,Slovakia,Slovenia,Spain,Sweden, Switzerlandand the UnitedKingdom. EachparticipatingcountrysignedaMemorandumofUnderstandingwherebytheir Governmentsagreedto co-ordinatetheirresearchefforttowardmeetingtheaims ofCOST 331. TheexecutionofCOST331,whilesupportedbytheEuropeanCommission,hasbeen directedbyaManagementCommitteedrawnfromtheMembership-thelattercomprised governmentrepresentatives,academicsand other experts in thefield. TheworkundertakenbyCOST331isbaseduponthefruitfulcontributionofthe15 signatorycountriesmentionedaboveandshownonthemapinfigure2.2.Inaddition, Ireland,theNetherlandsandNorwayhavealsocontributedtothisworkalthoughtheydid not sign the MemorandumofUnderstanding. 11 COST 331 Figure2.2- Participatingcountries 12 Chapter 3 Chapter3.Executivesummary Atpresent,mostnationaltechnicalspecificationsforroadmarkingslaydownminimum performance levels for the parameters defining their essential characteristics (namely:night time and day timevisibility and skid resistance)without having considered sufficientlythe relationshipbetweenwhatthedriverneeds, forappropriateguidance,andwhatthe road marking is able to provide (in terms of visual information). Furthermore, research in this area, both national and international, concentrates on: developing new marking products which meet the above mentioned performancelevels for as long as possible (maximum "durability", understood as"retainedperformance") and designingnewtechnologiesforthemanufactureofhigh-performanceequipmentfor assessing those requirements. Roadmarkingsare,infact,"trafficsignals"withadecisiveimpactondriver'ssafety mainlybecausetheyarenon-verbal(theirmessagebeingexpressedthroughdesignand colour) and, in consequence, readily understood by drivers and because, in poor lighting or bad weather conditions(when the informationdrivers get fromthe environment is limited to the areas illuminated by the headlamps of the vehicle), they are one of the most relevant elements to guide drivers safely along the road. Itis thereforeindispensableto haveavailablea provenscientificbasisforansweringthe most relevant questions concerning the design and use of road markings, such as: what is the visibility distance provided by a certain type, and quality, of road marking?; what is the visual demand of drivers, as far as road markings is concerned?.... TheresultsofCOST331, by establishinganup-to-datescientificmethodwithwhichto determine the optimumroadmarkingdesign(inorder to ensure that markingsare visible bydayandbynight,inallweatherconditions),providethatnecessarybasisallowing research, industry and road trafficengineerstoimprove- where necessary -the current value of road markings for drivers. Theexecutionoftheresearchprogramme,designedandapprovedbytheCOST331 Management Committee, included: 1.Acompletereviewofthestateoftheartbymeansofaliteraturesurveyanda questionnaire answered by 15 European countries. 2.Aninvestigationofthevisibilitydistanceofroadmarkings,inadrivingexperiment involvinga numberoftestpersonsandvariableconditions(concerningroadmarking pattern and reflectivity, and headlamp intensity). 3.Aninvestigationofthedriverneedforvisibilitydistance,carriedoutinadriving simulator,involvinganumberoftestpersonsandvariableconditionsconcerning driving speed, visibility distance and road curvature. 4.Monitoring driver behaviour in real trafficconditions throughout differentroad sections in Finland, Portugal and Switzerland(built up with differentdesigns and quality of road markings),byusinganunobtrusiveinstrumentedcarandinvolvingastatistically selected number of test persons. 13 COST 331 From the analysis of the results of the questionnaire, it can be statedthatall countries do havenationalregulationsortechnicalrecommendationswhichspecifygeometry,design and colourof road markings. However,and probably resulting fromthe lack ofscientific background, the design of road markings vary fromcountry to country. In some countries, additionally,theuseofa typeofroadmarkingintendedtoimprovenighttimevisibility under wetness or rain is also prescribed. The lack of a proper scientific background referredto above is also reflectedin the criteria forthe use of colours. While there seems to bea generalconsensuson the colourof the permanentroadmarkings(forthispurpose,mostcountriesinEuropeusewhite), agreementona standardcolourfortemporaryroadmarkingsstill hastocome(although most countries use yellow, white and orange are currently used as well). Theinvestigationsofthevisibilitydistanceofroadmarkingsandofthedriverneedfor visibility led to the development of a computer programme which allows the calculation for a given type of vehicle, driver, environment, headlamp and quality and designof the road marking,ofthevisibilitydistanceandpreviewtimeoflongitudinalroadmarkings. This computerprogrammeisusefulforresearchandeducationandasatoolforderivinga national policy on road markings. The previously mentioned researches, when combined with the analysis of driver behaviour (monitoredinthethreecitedfieldexperiments)andthe existingliteratureonthis topic, also suggest that road markings should be able to provide a preview time of 3 to 5 seconds to achieve comfort as well as safety1. Finally, the results of COST 331 were able to demonstrate the influenceof road markings onthebehaviourofdriverswhoadapttheirdrivingattitude(basically,speed)tothe information(i.e.visibilitydistance)derivingfromtheroadmarkings.However,more research needs to be done beforeany direct link can be established between such changes in driving behaviour and road safety. This is all very well, but it is not technically possible to provide markingsgivinga night time previewtime of 5 seconds in all conditions; this could require visibility distances exceeding 150 metres. The conclusions show that there is a need to establish a national policy (taking account of driverage,headlampintensityandglarefromopposingtrafficandclimate)forroad marking design, due to their influenceon road safety. To do that, the scientificbasis - and evidence -provided in COST 331 can fruitfullybe used. Nevertheless, COST 331 does not provide answers to all the questions which may be asked in connection with road markings - further research is needed to achieve that - but has taken a big step forward in establishing better knowledge of the driver's visual needs and the capability of road markings to provide information.COST 331, inthissense, providesanoutstandingscientificbackgroundfor future research in thisfield. Previewtimeisdefinedasthenumberofsecondstakentodriveadistanceequaltothevisibility distance. Inchapter6, it is concludedthat1,8seconds should beconsideredasthe minimum preview time needed just for safe driving. Later on, the results of the field experiments (chapter 7) suggest that a mean preview time of 2,2 seconds is even too short for driving confort. 14 Chapter 4 Chapter 4.State of the art 4.1Introduction Today'sexperienceleadstoaquitesimpleconclusion:roadmarkingsaresignalswitha verysimplemessage(whenretroreflectingroadstudsareplacedontheroad,evenhaving no particularcodedmessage in the legalsense, they clearly mark- forinstance- the centre ofthecarriageway).TheirinfluenceontrafficsafetyisdulynotedintheOECDreport (1975) "Roadmarking and delineation" in which it isstated: "Aclearandprecisesystemofhorizontalsigninginthe formofroadmarkingstogether withlateraldelineators(posts)servesaspecialpurposeby facilitatingdriverguidance, thusimprovingtrafficflowandcontributingto drivingcomfortandsafety. " Surprisingly,thisindispensableroadequipment,inspite ofthe factthatitservesthesame intendeduse(upgradingroadsafety)regardlessofthecountry,issubjectedtodifferent regulationsand technicalspecificationsin Europe.Thisshould not be a problemas long as thesamelevelofroadsafetyisachieved.However,experiencesuggestsandaliterature researchconfirmsthat littlescientificresearch hadbeen carriedout by roadauthoritiesand authorizedthirdpartiestodesignappropriateroadmarkingsystems(definitionof performance,colour,shape,dimensions,applicationcriteria,etc.)ensuringaminimum trafficsafetylevel.Inconsequence,establishingthe"stateoftheart"inthefieldofroad markingsis a necessary firststep in the preparationofstrategiestoachieveconvergencein the current differencesin roadsafety in differentEuropean countries. This phase of the work, therefore,deals with the "state of the art", in the area fallingwithin the Action'sscope. Tothatpurpose,anappropriatequestionnaire,to befilledin bymeans of interviewswith relevantdecision makers in Europe, was preparedand circulated(forthe textof the'questionnaire'referto theCD-ROM). This questionnaire is divided into three parts: Par t i :Road markings. Part 2:Retroreflectiveroad studs. Part3:Additionalinformation. Theparticipantcountries(14)completingthequestionnairewere:Belgium(B), Switzerland(CH),Germany(D),Denmark(DK),Spain(E),France(F),Finland(FTN), Greece(G),Iceland(ICE),Ireland(IRL)theNetherlands(NL),Sweden(S),Slovenia (SLO) and the UnitedKingdom (UK). Throughtheanalysisoftheanswersprovided,informationaboutWHY roadmarkingsand retroreflectiveroadstuds(RRS)areused,WHICHcriteriaareusedasbackgroundfor currentregulationsandtechnicalspecifications,WHICHcriteriaarepresentlyusedfor determiningmaintenanceandrelevantBIBLIOGRAPHYcanbefound(forthecomplete reporton the answerscollatedduringthe survey,referto theCD-ROM). 15 COST 331 4.2Results of the Questionnaire As has been described in the previoussection, the "state of the art" was identifiedby means ofinterviewswithrelevantdecisionmakersinEurope.Theanalysisoftheanswers resultingfromthesurveyconstituterealbackgroundinformationenablingvaluable conclusions to be reached about the road marking systems used in Europe. To make that analysis more comprehensivefor non-expert readers, the resultsare presented inthissectionasappropriateanswerstothemostrelevantquestionsincludedineach section of the questionnaire. Specialconsiderationshouldbepaidtotheuseofretroreflectiveroadstuds.Forthese devices,inadditiontothetreatmentoftheinformationgatheredfromthesurvey,an analysisoftheexistingliteratureinEnglish,FrenchandGermanhasbeenmade{forthe completereportonthisanalysisrefertotheCD-ROM).Unfortunately,theusable-informationisratherlimitedandnotcloselylinkedtothemaingoalofthispartofCOST 331.Howeveritisclearthattheuseofretroreflectiveroadstudsisnotwidespreadin Europeandtheiruseisinconsistent.Thisisratherdismissive.Studshaveconsiderable visibility advantagesover markings, in wet or foggyweather. Part 1: Road Markings Question1-Everybodyisusedtoseepavementmarkingsonroadsbuthasanyone wonderedwhethertheir applicationis compulsorybylegislation? Within the consultedcountries, in only three of them: FIN, UK and IRL theapplication of road markings is not compulsory. However, there are criteria fortheir use. Itshouldbenotedthatinthe11 countrieswheretheapplicationofroadmarkingsis compulsorybylegislation,itisalmostimpossibletoextendtheuseofroadmarkings (e.g. to use centre Unes on roads where only edge lines are prescribed). Question2-Whatcriteriaareusedforrecommendingtheapplicationofroad markings? Regardlessofthe factof havingor not having regulationsforroadmarkings,theroad widthincombinationwiththe AverageDailyTraffic(ADT)andaccident frequencyis the most commonly usedcriteria. Surprisingly,inspiteofthe factthat road markingsare intendedforthe samepurpose (roadsafety),theirdesignvariesfromcountrytocountry.Thatmightbeduetothe factthatnoneofthecountriesappearstohavecarriedoutscientificresearchto supporttheir technicalspecificationsorregulations. Therefore: Onmotorways:OnlyIRL usesbrokenlinesforleftorrightedgelines.EandF use broken lines for right edge Unes only (all other countriesuse continuouslines). The width aedge linesvaries from0.15mto 0.30m,while forlanemarkings,the width varies from0.10m to 0.20mand length from2.0m to6.0m. On interurban dualcarriagewayroads: Only IRL andS use brokenlines forleft and right edge Unes.E and F use them forrightedge Unes only(all othercountries use continuousUnes).The widthofedge Unes variesfrom0.10mto0.30m,while 16 Chapter 4 forlanemarkings,thewidthvariesfrom0.10mto0.20andlengthfrom2.0mto 6.0m. Oninterurbansinglecarriagewayroads:OnlyF,ICE,IRLandSusebroken Unes as edge lines.Theirwidthvariesfrom0.10mto0.30m.The widthofcentre lines varies from0.10mto 0.15mand the length from1.0mto.Om. Finally, it should be notedthat in mostcasesinternationalagreements(e.g.Agreement oftheEconomicCommissionofthe UN,Vienna,1968andtheEuropeanProtocolon RoadMarkings,Geneva,1971)aswellasnationalregulationsofothercountries (generallyspeaking,adapted-totallyorpartially-tonationaltrafficandroad conditions)havebeenthebasisusedasbackgroundfornationalspecificationsor regulationson road markings. Question3-Isthereanyconsensusaboutthe colourto be usedin pavementmarkings? Forpermanentroadmarkings,theconsensusthroughoutEuropeseemstobeevident: WHITE.InIRLandSLOonlyedgeUnesare yellow.In FIN,bothwhiteandyeUow are used forlongitudinalUnes, chevrons and hatch markings. Thesituationisnotsoclearfortemporaryroadmarkingsusedatroadworkswhere althoughmost countries use YELLOW, FTN, ICE, SLO and the UK use white and in G andIRL bothcolours(whiteandyellow)areused.InCHandS onlyorangeisused. InBonlyorangeretroreflectingroadstudsareused,whileintheUKyellow temporary road studs are used at road works. Question4-Theuseoftype2roadmarkings(thosedesignedtoimprovenighttime visibilityunderdifficultweatherandtrafficconditions)isbecomingincreasingly widespread.Butis theiruse regulatedinEurope? Somecountriesdonotusetype2roadmarkings(i.e.thosecapableofmaintaining nighttimevisibiUtyinadverseweatherconditions):FTN, G,ICE,IRLandSLO.In spite ofthe factthat in olher countries theirappUcationis recommendedinsome cases (B,CH,EandNL),onlyD,DK,F,SandUKsetnationalregulationsortechnical recommendationsprescribing the use of these markings: Profiledis the most commondesignoftype 2 roadmarkings.Systems basedonlarge glassbeadsarealsousedinB,CH,D,FandUK.Thereasonforthisisthatnoise ("wakeupmarkings")isincreasinglyconsideredbydecisionmakerstobea characteristicoftype2roadmarkingsasimportantasimprovednighttimevisibiUty. Itseemstobeobviousthatatype2roadmarkingshouldincludethesetwoessential characteristics:improvednighttimevisibilityandmechanical-acousticeffect. Finally,theapplicationfieldoftype2 roadmarkingsis restrictedtolongitudinallines (mostlyonedgelines)andonroadsoutsideurbanareas.However,type2road markingsarealsousedonhatchmarkingsinDandF,onmotorways.Inaddition,F usesthosemarkingsforchevronsanddirectionalarrowsas weU as ondualandsingle carriageway roads. Thereforeit canbestatedinthisfieldalso,thattherei^^&Tk^Fn>sstigation.Only D,DK,S,FandtheUKhavecarriedoutsomerpa$rchontheus | ^{type2road markings. 17 COST 331 Question5-Themostimportantwayofimprovingroadsafetybymeansofroad markingsis by implementingappropriateregulationsor technicalspecifications,but: Question5.1-Doroadauthoritiesconsiderthatthoseregulationssatisfactorilymeet currenttrafficneeds,in theircountries? Fordifferentreasons,only5countries(DK,E,FIN,GandSLO)recognisethat theircurrentstandardsdonotmeetpresenttrafficneeds.However,CH,DK,F, FIN, G, S and SLO are presently considering improvementsor changes innational regulations. Question5.2- Aretheenvironmentalprotectionandthetrafficdisturbance(application speed, dryingtime,etc.) characteristicsconsideredessentialby roadauthorities? Ingeneral,itcanbestatedYES.Thisisreflectedinthenationalregulationsor technicalspecificationsofseveralcountries.Theabsenceofaromaticororganic solventsandleadin combinationwith fastdryingmaterialsare the mostcommon requirementstoachieveenvironmentalandtrafficprotectionduringandafterthe applicationof road marking products. Question5.3-Specifyingappropriateguaranteesforroadmarkingperformance(dayand nighttimevisibilityandskidresistance),isthelastrequirementtobeprescribedin technicalspecificationand regulations.IsthatthesituationinEuropetoday? NO.Thevastmajorityofcountriesprescribetheuseofcertifiedproductsalong withaminimumguaranteefortheroadmarkings.Thoseguaranteesvary considerablyfromcountry to country:e.g.inIRLandS,2 yearsregardlessofthe type of product. In D,1 year for paints, 2 years for cold plastics andthermoplastics and 4 years for inlay thermoplasticsare required. Consideringtheminimumandmaximumperiodscurrentlyspecified,itcanbe statedthatthe guaranteesin EuropevarybetweenIto4years. Question6 - Special attentionshouldbe paidto road markingMAINTENANCE. Question6.1 -What sortof criteria determineroadmarkingmaintenanceinEurope? Budgetand minimum road marking performanceare the most importantcriteria. Question6.2- AreEuropeanroadauthoritieshappywiththeircriteria fordetermining maintenance programmes ? Althoughbudgetconstrictions,togetherwithlimitedabilitytomonitorroad markingquality,are strong influencesonmaintenanceexpenditure,roadauthorities wouldliketoimprovethepresentsituation.Withinthepossiblealternatives,it seemsthattheywouldliketobeabletoplanroadmarkingmaintenance programmestorestoreperformancelevelswheneverthesedeclinetotheminimum valuesspecified. 18 Chapter 4 Part 2: RetroreflectiveRoad Studs In general, the use of retroreflectingroadstuds(RRS) is not widespreadenoughto reach reliableconclusionsfromtheanswersoftheparticipatingcountries.However,these provisional results offersufficientinformationto understand a bit better the contribution of retroreflective road studs to improving road safety. Ingeneral,retroreflectiveroadstudsareconsideredasa horizontalroadmarkingwhich presentssome particularadvantages for trafficsafetyin respect of visibilityat night or in adverse weather conditions (fog, rain, etc.).The general consensus is that the use of RRS in highhazardouslocationsdoesenhancedelineationandimprovestheoverallsafetyof those road sections (that is the main reason why they are used to supplement conventional road markings or even to replace them). Ontheotherhand,andbasedontheavailableliterature,itisdifficulttoassess quantitatively the effectivenessof RRS. Studies carried out on this topic conclude that the use of RRS provides a valuable guidance system, but none of them quantifythat added value. Question 1 -Isthe application of RRS compulsory by legislation? Only few countries have legislated the use of RRS: E, the NL, S and the UK. TheuseofRRSvariesfromcountrytocountrynotonlywithinthesametypeof application(permanent or temporary) but in their field ofapplication(e.g. in G, RRS areusedexclusivelyinconstructionworkzonesmeanwhile,inSweden,sometimes they are used in these zones and very seldom for permanent applications). Question 2 - What criteria are used for recommending the application of RRS? In generalthere is no singlecriterionin Europeto determinethe applicationof RRS. The reason may partly be the lack of consistency in the use of these devices.Improved visualperformanceinadverseweatherconditionsseemstobeoneofthemost common justificationsfor their use. It should be pointed out that all countries which use RRS for permanentapplications have national regulations - or technical recommendations - to specifytheir installation criteria.The exceptionisSLO where installationcriteriaare fixedtender by tender. However, in spite of finding differencesin the installation criteria among the countries using RRS in permanent applications, it can be stated that their intended use is similar: to substitute for or to supplement longitudinal lines. Question 3 - Appropriate regulations and technical specifications for retroreflective road studs are indispensable to achieve the efficiency of these devices, but: Question 3.1-Present regulationsandtechnical specificationsdonotmeetcurrent traffic needs: Only two countries (IRL and the UK), among the users of RRS, consider that their applicable regulations do not need to be improved. 19 COST 331 Themostwidespreadopinionabouthowtoimproveregulationsandtechnical specificationsofRRSisbymeansofspecifyingtheirapplicationareas (EN1463-1 shall be usedas initial performancestandardonly). Question3.2-Therearenoinstallationcriteriaspecifiedinrelationshipwith environmentalprotectionand trafficdisturbance: Evenbeingusedforpermanentapplications,onlytheUKhastechnical requirements on these topics: - avoiding hazardousmaterials; - high applicationspeedand long durability(functionallife) Question 3.3 - There are no singlecriteria to specifyguarantees: Insomecountries(e.g.GandE),itisenoughtogetappliedRRSmanufactured accordingtoinitialperformancespecifications.Others(e.g. DKandtheUK),in additionspecifyfunctionallives: from1 to 4years, inpermanentapplicationsand just3 months intemporary. Question3.4-Budgetandminimumperformancearethemostimportantcriteriato determineRRSmaintenance,however: Roadauthoritieswouldlike toimprovethepresentsituation.Withinthepossible alternatives,itseemsthattheywouldliketobeabletoplanroadstuds maintenanceprogrammesto restoreperformancelevelswheneverthesedeclineto the minimumvaluesspecified. Part 3: AdditionalInformation Withregardtotheroadsafetybenefitsofusingroadstudsorspecialtypes(profiled roadmarkings,etc.)ofroadmarkings,onlyDKandEknowofrelatedresearchthat could help roadauthoritiesin making appropriatedecisions. Withregardtotheuseof"maskingmaterials"(usedtotemporarilymaskpermanent roadmarkingsinconstructionworkareas),onlyDK,NLandtheUKhavenational regulations or technical recommendationson the visual performanceof those products. 20 Chapter 4 4.3Conclusions Regardlessofthefactofhaving,ornothaving,regulations,theroadwidthin combinationwiththeAverageDailyTraffic(ADT)andaccidentfrequencyarethe usual criteria in recommendingthe applicationof roadmarkings. Surprisingly,in spite of the factthat road markingsare intendedforthesamepurpose (roadsafety),theirdesignvarieswidelyfromcountrytocountry.Thatmightbedue tothefactthatnoneofthecountriesappearstohavecarriedoutscientificresearch to supporttechnicalspecificationsorregulations. Forpermanentroadmarkings,theconsensusaboutthe"colour"throughoutEurope seemstobeevident:mostcountriesuseonlyWHITE.InIRLandSLOedgelines can be yellow.In FIN, white andyellowcan be usedforlongitudinallines,chevrons and hatchmarkings. However,thesituationisnotsoclearfortemporaryroadmarkings.Althoughmost countriesuseYELLOW,FIN, ICE, UKandSLO usewhitewhile inG andIRLboth colours(whiteandyellow)areusedinCHandSonlyorangeisstillused.InB orangeretroreflectingroad studsare used. Thereare countriesthatdonotusetype2 roadmarkings(thosedesignedtomaintain nighttimevisibilityinadverseweatherconditions):FIN,G,ICE,IRLandSLO.In spiteofthefactthatinothercountriestheirapplicationisrecommendedinsome cases(B,CH,EandNL),onlyD,DK,F,S,andUKhavenationalregulations,or technicalrecommendations,prescribingtheuseofthesemarkings(restrictedto longitudinallinesandon roadsectionsoutside the urbanareas).Profiledisthemost commondesignof that type of roadmarkings. For differentreasons, only5 countries(DK, E, FIN,G andSLO)officiallyrecognise thattheircurrentstandardsdonotmeetpresenttrafficneeds.However,CH,DK,F, FIN,G,SandSLOarepresentlyconsideringimprovementsorchangestonational regulations. Althoughbudgetconstrictions,togetherwithlimitedabilityto monitorroadmarking quality,arestronginfluencesonmaintenanceexpenditure,roadauthoritieswould liketoimprovethepresentsituation.Withinthepossiblealternatives,itseemsthat theywouldliketobeabletoplanroadmarkingmaintenanceprogrammestorestore performancelevelswheneverthese declineto the minimumvaluesspecified. Baseduponthelimitedamountofavailableliterature,itispossibletoassertthat retroreflectiveroadstuds(RRS)-usedforsubstitutingorsupplementing conventionalroadmarkings(especiallylongitudinallines)- are an effectivemeans of improvingguidancetodriversparticularlyinadverseweatherconditions.However, thecurrentlackofresearchinthisfieldmakesitimpossibletostateanyfigure concerning the "effectiveness"of RRS in roadsafety. 21 Chapter 5 Chapter5.Visibility of RoadMarkings 5.1Introduction This chapter describes the methodology used to developa model forthe calculationof the visual information(called "visibility level" -VL) provided by road markings The preparation of a visibility model, capable of calculating the visibility distance provided byroadmarkingsunderdifferenttrafficconditions(cars,ageofdriver,weather,...),is required to deal with, and to calculate, the supply of visibility. The starting point was to use a general model for the visibility of targets on backgrounds. Themodelusedreferstoalaboratorysituationandcannotreadilybeappliedforthe complex road situation. A trial error and procedure may be used to determine the visibility distance, definedas the distancewherethe visibilitylevel has a selected value. The model developedin COST 331 may thereforebe considereda replacementforthe methodology given, for the same purpose, in CIE report N 73 which has a smaller range of applicability and lack of support from driving experiments. 5.2Calculationprocedure The calculationprocedure, forthe "visibilitylevel" (VL) ofa road marking, involves the following steps: a)Calculate the equivalent target size of the road marking. b)Calculate the luminances of the road marking and the road surface. c)Calculate the visibility level (VL) according to Equation 1. d)Evaluate the calculatedvisibility level. A minimum level of10 for practical driving situations is recommended. If the visibility distance is to be computed, a further step is added: e)Ifthecalculatedvisibilitylevelishigher/lowerthantheselectedvalue,then increase/decrease the distance to the road marking and repeat steps a) to d), otherwise accept this distance as the visibility distance. Theequivalenttargetsizeofaroadmarking,tobecomputedinstepa), isthesizeofa circular target of the same solid angle as obtained by integration over the surface of the road marking. The basic equation used for the VL of a target was (equation 1): VL = AL x (^/(A + B x a)2(Equation 1) where ais the target size in minutes of arc. ALis the luminance differencein (cd/m2). A and Bare functionsof the background luminance (Lb) in (cd/m2). VLis the visibility level. 23 COST331 When the luminance is uniform, for example for a transverse road marking, the solid angle is theapparentareaoftheroadmarkingdividedbythedistancesquared.Therefore,a road marking of area A which is seen at an angle of view v will have an apparent area: A' = Axcosv(Equation 2) with a solid angle of (AL ), being D the distance to the road marking. D2 When the luminance varies, for example along a longitudinal road marking, a weight of L/Lo isappliedtoeachelementofthesurfacearea beforeintegratingthesolidangle.L isthe luminance at the location of the element and LQ is the luminance at the frontend of the road marking. Section 5.3 gives more precise instructions on how to compute the equivalent target size. Luminance, to be computed in step b), is found as the product of the illuminance produced by alightsourceandaluminancecoefficientrepresentingthetypeofillumination,the geometricalsituationandthesurface.Whenmorethanonelightsourceispresent, luminances produced by each are added to provide the total luminance. When the luminance is uniform, only two luminance values need to be calculated, one for the road marking and one for the road surface. When the luminance varies, the luminance of the road marking at the front end Lo is selected to representthe roadmarking.Theluminanceoftheroadsurfacenexttothislocationis selectedto representthe roadsurface(backgroundluminance Lb). Luminancesofthe road marking at other locations are needed for the calculation of the equivalent target size, see the discussion above. Section5.4givesmore preciseinstructionsonhowtocomputeluminancesforheadlamp illumination and for daylight/road lighting. Priortothecalculationinstepc), substitutionsmay becarriedoutinordertotakeglare and/or age of the driver into considerationaccording to sectionsA.3 and A.4, ofannex A, respectively. As a practical expressionof the model, a computer programme using the model has been developed and included in the CD-ROM attached to this final report. The programme may be useful during revision of national regulations and technicalspecifications. Thementionedprogrammeoperatesinasinglepagewhichhasenoughroomforthe following input: driver, vehicle and glare; road marking geometry and location; road geometry; headlamp illumination and coefficientsof retroreflectedluminance(RL); daylight/road lighting and luminance coefficientin diffuseillumination (Q ca o o o co cc Radius (m) Figure 6.7-Lateral position in every curve at the condition of free choice of speed at the four levels of sight distance. The lateral position of each type of curve can be foundin figure 6.8 with the sight distance on the x-axis. The centre line is at lateral position 0 m and the right edge of the edge line is at 3,5 m. The radii marked -are left curves. The radii marked C are with clothoid. Radius -1.0 -1.2 c o * 'w o Q. "( 0) *- C ^ ^ - r * " _*, >S-1.6 2jnss2^"" -1.8 -2.0 -2.2 -200C -300 -300C I -450 -4S0C -675 "200C I 300 >i^^^ ;^r;r--. --r;rJ. "ji '1'. - * - ' . . i _u-i -#-i -i -i -n"i "i " 3045 Sightdistance(m) 67100 Radius Figure6.8 -Lateral position for each type of curve at the condition of free choice of speed. 48 Chapter 6 It is not clear fromthe results of the statistical analysis whether transition curves have any effecton lateral position. Because of the non-symmetry, the possibility cannot be excluded that the effectof a transition curve is combined with an effectof radius. Cruise control at 90 km/h Forthe condition90 km/hthelateralpositionsonstraightstretchesdifferverylittlefor differentsight distances (figure6.9). Lateral position is the distance between the centre of car and the centre of road. 0 m lateral position is the centre Une and 3,5 m is the outer edge of the right edge line (group mean values). c o v 'w o Q. l (D t 0 0.5 1 1.5 2 2.5 3 3.5 centreline * *^^^^^^^^^^^^^ edgeline -1i Rightcurves Straightstreches -Leftcurves 203045 Sightdistance(m) 67 Figure 6.9 -Mean lateral position for right and lefi curves and straight stretches at the condition of cruise control For right curves, the longer the sight distance the furtherto the right is the driver's lateral position. In left curves, the lateral position is furtherto the left the longer the sight distance. At the two longer sight distances (45 and 67 m) the driver has a mean lateral position to the right of the positiononthestraightstretchandtheotherway around inleftcurves. This behaviour is similar to the condition free choice of speed and might be because the driver is cutting the curves. At the shorter sight distances (20 and 30 m) the lateral position inright curves is to the leftof that on straight stretches. In leftcurves the lateral positionis to the right comparedtostraightstretches.Thiscouldbebecausethedriveris"missing"the curve.TheAnalysisofVarianceshowsthattheinteractionbetweensight distanceand right or lefi curve has a significant effect on lateral position (F(3;1465) = 376,0; p 0,001) and for leftcurves (F(l;353) = 9,0; p< 0,005). There is no significant interaction with sight distance. Cruise control at 90 km/h For the condition cruise control also, it is mainly the extreme lateral position of the outside wheel in the curve that is interesting. There is a tendency at longer sight distances for the extremelateral position ofthe outer wheeltooccur furtherintothe curve. It is probable thatthedriveris"cutting"thecurveorallowinghimselfto"stagger".Thisissimilarto what happened at the condition free choice of speed. For the shortersight distancesthe situation is the reverse, theextremelateral positionof the outer wheel moves out in the curve. This could be because the driver is "staggering" or "missing"thecurve. Thistendencyisstrongerthesmallertheradius is. Theinteraction betweensightdistanceandradiusis,forallextremepositionsoftheouterwheel, significant(F(6;349)= in the range of2,9to 4,6; p>0,01). Theextremelateral positions (group of mean values) of the wheels are shown in figure 6.18 for curves without clothoids andin figure 6.19withclothoids. Thecentreline hasthelateralposition0 mandouter edge of the right edge line 3,5m. There is one line for each radius. The lines with squares are right curves and the lines without are left curves. 55 COST 331 centreline 20 leftwheelleftcurve left wheelrightcurve right wheelleftcurve right wheelrightcurve 3045 Sight distance (m) Figure 6.18 Extreme lateral position of the outside wheel in curveswithout clothoids at the condition of cruise control c o '55 o Q. "jo CD OJCD E CDi x LII left wheelinleftcurve left wheelinrightcurve j- ,right wheelin leftcurve -" |right wheelinrightcurve 3045 Sight distance(m) -450C ---300C 200C -450C 300C 200C Figure 6.19 -Extreme lateral position of the wheels in curves with clothoid at the condition of cruise control. 56 Chapter 6 The evolution of the extreme lateral position (group of mean values) of the outside wheels, but for right and leftcurves separated, are shown in figures 6.20 to 6.23. In all the figures, the centre Une is at the side position 0 m and right edge line is at 3,5m. 0i 203045 Sight distance(m) Figure 6.20 -Extreme lateral position of the wheels in right curve without clothoid at 90 km/h. 20 3045 Sight distance(m) Figure 6.21 -Extreme lateral position of the wheels in lefi curve without clothoid at 90 km/h. 57 COST 331 c g 'to o Q. l o i (D *' ro a) E a) LU 3045 Sight distance(m) Figure 6.22 -Extreme lateral position of the wheels in right curve with clothoid at 90 km/h. c o '55 o Q. ro i B ro a> E 0,6:{ II B=0,09588xLba466 I If logA=0,2355+0,173xlogLb U!< 0,00418:i II logB =-0,6835+0,5275xlogLb+0,0227(logLb)2 I I logA =0,1355+0,3372xlogLb+0,0866(logLb)2 I in between:{ {logB=-l,0485+0,3190xlogLb NOTE:The equation is rearrangedas compared to the paper by Dr.-Ing. Werner Adrian. The functionsA and B equal (kxO)H and (kxL)^ of that paper. 97 COST 331 A.3Influenceof disabilityglare The veiling luminance caused by disability glare is expressed by: Lv = kLE/e2(Equation A3) whereLvis the veiling luminance in cd-m"2. kis a constant of 9,2. Eistheilluminanceinlxontheeyeoftheobserverfromaglare source. 8is the glare angle in degreesmeasuredfromtheline ofsightto the direction towards the glare source, andEmeans summation for all glare sources. The influence of disability glare on the threshold target size is obtained by substituting Lb by Lb + Ly in equation A.2. A.4Influenceof age Withage, theoculartransmissionandtheopticalclarityoftheeyebothdecrease. These effectsare highly individual, data given below are averages for a number of test persons. The first-mentioned effectofage is describedby substitutingAL withAL/AFIin equation A. 1and Lb with Lb/AFl in equation A.2. AFI is an age factor with values given by: Age from 20 to 44: 0,0100xAge + 0,8 AFI =!Age from 44 to 64: 0,0282xAge(Equation A4) {Age from 64 to 80: 0,1876xAge -10,2 The last-mentioned effectis described by multiplying the veiling luminance Ly as derived by equation A.3 with a factor AF2 before making the substitutions defined in section A.3. AF2 is given by: Age from 20 to 25: unity AF2={(Equation A. 5) {Age from 25 to 80: 1 + (0,0248xAge - 0,62)2 A.5Comments on the basic equation In laboratorysituations, when observers knowwhatto expectand haveunlimitedtimefor observation (2 seconds or more), a visibility level VL of unity is sufficientto ensure detection of the target with a high probability. Intrafficsituations, ontheotherhand,thetimeforobservationofeachpossibletargetis limited, and targets may have to be searched for. Dr.-Ing. Werner Adrian [1989] recommends the use of VL = 10 for certain detection. EquationA.1hasanextremecasewhenthetermBxaislargecomparedtothetermA, making VL roughlyindependentofthetargetsize a(VL = AL/B2). Thisoccurswhenthe 98 Annex A contrastC=AL/Lb issmallanddefinesthedomainofWeber'slaw.Sometypicallimiting values of the contrast are given in table A.l. NOTEI:Inconditionsofroadlightinganddaylight,someroadmarkingshavesuch small contrast to the road surfacethat their visibility is governed by Weber's law. This meansthatthevisibilitydistancemaychangefromshorttolongwithonlysmall improvements in contrast. EquationA.lhas another extreme case whenthe term Bxais smallcomparedto the termA, givingthe targetsizeaitsfullinfluenceontheVL(VL = ALxa2/A2). Thisoccurswhenthe contrastC = AL/Uis largeanddefinesthe domainofRicco'slaw. TableA.lcontainssome typical limiting values forthis case also. NOTE2: In conditionsof headlamp illumination,the contrastof roadmarkings tothe road surface is sometimes so high that their visibility is governed by Ricco's law. This meansthat thevisibilitydependsstrongly on the distance, andon thesize(widthand pattern) of road markings. NOTE3: For verysmalltargetsof highluminance,the termALxa2is proportionalto the illuminance at the eye, which is the stimulus to the eye in such situations. Table A.1-Typicallimitingvaluesofthecontrast,wherethelawsofWeber andRicco apply.Valuesare forVL= 10. Typical limiting values of the contrast AL/Lb at backgroundluminance I^,: 0,001 headlamp lighting\0,01 I0,1 road lighting"i1 fI10 daylight"!100 I1.000 cd-rn"2 Weber's law close to: 0,76 0,43 0,19 0,09 0,08 0,07 0,06 Ricco's above: 25,0 15,0 6,8 3,4 2,8 2,4 2,2 law (12') (6,3') 0,6') (2,6') (1,7') (1,4') (1,4') Inbetweenthetwoabovementionedextremecases,thesizeofthetargetdoeshavean influenceon visibility, but not as strong as in the domain of Ricco's law. In all cases, the contrast sensitivity increases with increasing backgroundluminance,meaning that targets may be visible even when contrast or target size are smaller. This works up to the high levels of daylight conditions, where the contrast sensitivity saturates. 99 COST 331 Table A.2 shows some values of the thresholdtarget size aforsome values of the contrast AL/Lb and the background luminance Lb. Table A.2 -Values of the threshold target size a forVL = 10. Lb headlamp lighting road lightingi I daylight{ I f i { 0,001 0,01 0,1 1 10 100 1.000 cd-m"2 AL/U0,1 64,2 30,8 22,7 1 58,0 13,4 5,8 3,2 2,4 2,3 10 22,7 8,1 2,8 1,4 0,8 0,6 0,6 100 5,7 2,2 0,8 0,4 A.6Comments on disability glare In conditions of headlamp illumination and road lighting, the most severe sources of glare are headlamps of oncoming vehicles. Consider a simple situation where: the observer drives his car on a straight road, looking straight ahead the observer meets a motorcycle with one headlamp using the dipped beam the luminousintensityI ofthe headlampindirectionstowardstheobserverduring the meeting is constant the lateral separation S of the driver and the motorcycle is constant during the meeting At a point during the meeting, the distance between the two vehicles, as measured along the road, is D. The illuminance E is I/(D2+S2) and the angle 0 is arctan(S/D). As long as the distance D is much larger than S, this can be approximated by E = /D2 and 0 = 57,3xS/D. Accordingly, the veiling luminance becomes Ly = kxEx"2 = 0,0028xl/S2. The veiling luminance isseen to be constantin thissimple meeting situationona straight road. In practice, glare is less when the distance D is very large, among other reasons because ofabsorptionin theatmosphere. Furthermore, glaredoes decreaseinthe last phase ofthe meeting, partly because the luminous intensity I decreases at wide angles to the road. However,ata meetingonastraightroad,glaredoesemergeatlargedistances,normally hundreds of metres, and does not decrease significantlyuntil shortly beforethe two vehicles pass each other. The level can be evaluated using a value of 200 cd for I, being typical, and a value of 3,5 m forS,correspondingtoameetingonatwolaneroad,givingaveilingluminanceof 0,046 cd-m"2. 100 Annex A Whenmeetingvehicleswithmorethanoneheadlamp,perhapswithahigherluminous intensity, and perhaps several vehicles at a time, the veiling luminance may become as high as 1 cdm"2 in some cases. In the case of headlamp illumination, the background luminance of the road surface is mostly only a small fractionof1 cdm"2. So that glare fromoncoming vehiclesmay besevere. In some cases, glare may shiftthe situationfromthe domainof Ricco'slaw to the domain of Weber's law, making it doubtful if road markings can be seen at any distance. Even in the case of road lighting, where the luminance of the roadsurfaceis mostly in the range from 0,5 to 2 cdm"2, glare from oncoming vehicles may be quite severe. The levels of road lighting have probably been set so as to be adequate to compete with glare. The observer can reduce glare by looking away from the glare source, such as fixing his gaze on the edge line at some distance ahead, instead of looking straight ahead. Furthermore, glare is strongly reduced by an increase of the lateral distance S. Accordingly, glare is much less, when driving in the near side lane on a four lane road, and is much less on a road with a central reserve. For a realistic calculation of glare, all details of the geometry should be considered. However, the values of the veiling luminance given in table A.3 may be used for simplified evaluations. Table A.3 -Values of the veiling luminance Lv in cdm2by glare from oncoming cars. number of on-coming cars: 1 2 3 4 5 lateral separation to oncoming cars: 3,5 m7,0 m10,5 m 0,0980,0240,011 0,1960,0490,022 0,2940,0730,033 0,3920,0980,044 0,4900,1220,054 14,0 m 0,006 0,012 0,018 0,024 0,031 17,5 m 0,004 0,008 0,012 0,016 0,020 101 COST 331 A.7Comments on the influenceof age Some values of the factors AFI and AF2 defined in section A.4 are given in table A.4. Table A.4 -Values of the factors AFI and AF2for the influence of age. Age AFI AF2 20 1,00 1,00 30 1,10 1,02 40 1,20 1,14-50 1,41 1,38 60 1,69 1,75 70 2,93 2,25 80 8,81 2,86 The effectof age on the transmission of the eye (AFI) is like driving with sunglasses or with a tinted wind screen. The effecton the clarity of the eye (AF2) is like driving with a dirty or worn wind screen. Judged by the values of table 4, the influenceon visibility is perhaps not bad forthefirst-mentioned effectalone, but in combination with glare the influence must be considerable. 102 Annex B AnnexBDrivingexperi ment B.lIntroduction Annex A defines a basic equation A.1 for the visibility level VL using parameters definedfor a laboratorysituationwitha circular target presentedona background,bothwithuniform luminance. In terms of visibility of road markings in driving situations, a road marking is the object and the road surfaceis the background. However, the equation is not readily applicable, as road markingsaregenerallynotseenascircularobjectsandas theroadmarkingandthe road surface can have non-uniform luminance. A translation is obviously required. The translation must include methods for substituting the road marking by a circular target of uniformluminance, and fordetermining the target and background luminance. The translation must be logical, and must provide correct results, at least to an acceptable approximation. The scope of the driving experiment is to establish such a translation, and to supply data for testing it. A further aim is to determine the visibility level VL needed in driving situations. The experiment was carried out by the Swedish Road and Transport Institute in the autumn of 1996. For practical reasons, the experiment included only situations of headlamp illumination and only longitudinal markings. The experimentdid however providea fairlysoundtest ofthe proposedtranslationas the Unesincludedbothbrokenandcontinuouslinesandastheconditionswerevaried considerably. TheexperimentalconditionsareexplainedinsectionB.2,whiletheresultingvisibility distances are given in section B.3. A factoranalysis of the results given in section B.4 shows that the parameters definingthe experimental conditions influence the results strongly, leaving only a small residual variation. It is pointed out that conditions are in the domain of Ricco's law. In section B.5, a translationis introducedand used to give calculated visibility distances in close agreement with those of the experiment. B.2Experimentalconditions Table B.lsummarizes the experimental conditions. The experiments were conducted on part of road 1050 in the central part of Sweden. The road is straight and flat, having been built on an old railway embankment. It was selected to avoid situations where road markings may be hidden by horizontal curves, or their visibility affected by vertical curves. The road markings were in thermoplastic material, placed in a line in the centre of a driving lane,appliedfortheexperimentandremovedafterwards.Thelengthwasapproximately 75 m, selected to be sufficientto simulate very long road markings. The spacing of driving lengths was at least 200 m in order to secure independent observations. 103 COST 331 The road markings had four differentpatterns and two levels of retroreflection. Each of these eight types was repeatedsix times. Accordingly,a total of 6x8 = 48 sections of 75 m long road markings were applied to the road. The aim forthe two levelsof retroreflectionwas values ofthe coefficientofretroreflected luminance RL of 100 and 400 mcdm^lx"1 respectively. The low level showed some variation forthe fourpatterns, whilethehighlevelwasestablishedtoagoodapproximationinall cases. The test vehicle and its recording equipment is equipped forexperimentsof this nature and has been usedon several previous occasions. For this particular experiment,threedifferent lighting systems were used, providing clearly differentlevels of total luminous intensity. The aim was to provide uniform intensities towards points along the road markings, so as to avoid the disturbing influence of the cut-off of the low beam and other variations. The experiment involved nine observers, all in their twenties and with normal vision. The method was to bring three observers through the road at a time, each with his own push button for recording detection of a road marking. Observers took part in three tours with the different lighting systems. The driving speed was kept at 90 km/h. The retroreflectionlevels of the road markings and the road surface, i.e.RL, were measured once and checked before each tour using an LTL2000 retrometer with the standard geometry defined in the European standard EN 1436. The total luminous intensity of the headlamps was also measured once and tested before each tour. Measurementswere madeby the use ofa photometerplacedattheroadand turned towards the headlamps. The weather conditions were good with clear nights and dry road surface. If an opposing car was encountered, which happened rarely, the test vehicle was stopped at the road side for a momentbeforestartingoffagain.Ascoutinavehicleseveralkilometresaheadgave warnings. 104 AnnexB Figure B.l- Broken line: 3 + 9m;width: 10 cm * " ! % *$-" Figure B.2 - Broken line: 3 + 3 m; width: 10 cm 105 COST 331 wS"jp, FigureB.3- Continuousline;width:10cm es*.* FigureB.4- Continuousline;width:30cm 106 AnnexB Table B.l-Experimental conditions. The road is a two-lane road in rural conditions, it is straight andflat. The RL value, as an average along the road and measured in the geometry of EN 1436 is 7,9 mcd-m'^lx"1. The road markings were applied for the experiment and removed afterwards. They were of a length of appr. 75 m, spaced at least 200 m apart and placed in a line in the centre of a driving lane. The pattern of road markings: P(l) broken Une 3+9 m, 10 cm P(2) broken line 3+3 m, 10 cm P(3) continuous Une, 10 cm P(4) continuous line, 30 cm The retroreflectionRL of road markings: RL(1) approximately 100mcdm^lx1 *) RL(2) approximately400mcdm^lx"1 **) *)the following values apply for RL(1): P(l):77 mcdm^lx1 P(2):83 mcdm^lx"1 P(3):96mcdm2be1 P(4):118mcdm-2-lx"1 **) 400 mcdm^lx"1 applies for all patterns The test vehicle is a Volvo 245 Station Wagon with a recording system for the position and equipped with indicators for three test persons at a time. The observer eye height is 1,2 m and the height of the headlamps 0,65 m. The headlamps of the test vehicle: 1(1):two normal dipped beam headlamps aimed low (-2,5%) so that the sharp gradient from the cut-off is projected onto the road surface between 25 and 30 m in front of the car 1(2):1(1) plus four fog lamps aimed parallel to the road surface and dimmed 1(3):1(2): but with the fog lamps at full power total luminous intensity of headlamps, measured in positions along the centre of the driving lane: Distance4060100150200m 1(1)16001270117011001030cd 1(2)57006500720074007400cd 1(3)2030025000280002870029000cd 107 COST 331 B.3Experimental visibility distances The experimental visibility distances are shown in table B.2. It may be noted that observers have individual levels of visibility distance, as is always the case in experiments of this nature. This may be due to individual eye vision and/or criteria for detection. However, the visibility distances of the individual observers depend on the conditions in the same manner. The averages, also shown in table B.2,are used in the following. Table B.2 -Experimental visibility distances. conditions (see table B.l) I(1)RL(1)P(1) P(2) P(3) P(4) R L ( 2) P ( 1 ) P(2) P(3) P(4) I(2)RL(1)P(1) P(2) P(3) P(4) R L ( 2) P ( 1 ) P(2) P(3) P(4) I(3)RL(1)P(1) P(2) P(3) P(4) R L ( 2) P ( 1 ) P(2) P(3) P(4) observer: 1 34 35 45 61 50 49 60 81 41 41 58 83 68 77 88 119 61 70 105 133 97 114 129 174 2 40 45 60 80 63 75 85 106 56 58 89 114 86 107 122 177 76 92 146 176 125 160 173 252 3 42 47 64 81 70 80 98 131 65 61 99 126 99 126 142 201 84 109 146 183 134 170 181 288 4 39 43 62 81 66 79 93 117 49 65 89 126 96 118 133 176 73 90 120 166 134 166 163 211 5 41 50 63 96 78 75 100 137 51 67 91 128 95 125 151 182 79 86 124 151 116 158 148 219 6 39 47 66 97 75 87 101 130 57 70 96 139 96 144 139 193 88 105 150 163 129 179 174 240 7 44 54 72 95 69 85 104 146 65 77 104 137 115 136 161 230 96 115 157 200 152 204 196 300 8 45 52 71 90 59 79 85 120 53 72 104 125 93 122 131 201 93 114 184 214 164 194 202 319 9 39 48 70 94 76 84 89 147 60 71 107 138 115 142 143 216 92 111 166 217 168 204 206 307 average 40 47 64 86 67 77 91 124 55 65 93 124 96 122 135 188 82 99 144 178 136 172 175 257 108 Annex B B.4Factor analysis of the experimental visibility distances TableB.3showsacomparisonoftheaverageexperimentalvisibilitydistancesDlto distances obtained by factor analysis D2. Table B.3 -Comparison of average experimental visibility distances Dl, distances obtained by factor analysis D2 and by the basic equation D3. test conditions (see table B.l) I(1)RL(1)P(1) P(2) P(3) P(4) RL( 2) P( 1) P(2) P(3) P(4) I(2)RL(1)P(1) P(2) P(3) P(4) RL( 2) P( 1) P(2) P(3) P(4) I(3)RL(1)P(1) P(2) P(3) P(4) RL( 2) P( 1) P(2) P(3) P(4) average st. deviation DlD2AD2 4041-1 4748-1 64613 86833 67634 77752 9195-4 124129-5 5560-5 6571-6 93903 1241222 96924 12211012 135139-4 1881880 8286-4 99102-3 14412915 1781753 1361324 17215814 175199-24 257270-13 1131130 8 (6%) D1 is the average observation distance given in table B.2 D2 is an approximation to D1 by the following factor analysis where:P(l)P(2)P(3) FP=1,0001,1941,503 RL( DRL( 2) FR=1,0001,549 KD1(2)1(3) H=1,0001,4622,099 D3 is obtained by the basic equation A.l for VL = 7,2 D3 40 48 59 85 64 79 97 130 57 71 88 124 95 113 137 194 82 101 127 176 136 162 192 255 113 AD3 0 -1 5 1 3 -2 -6 -6 -2 -6 5 0 1 9 -2 -6 0 -2 17 2 0 10 -17 2 0 7 (5%) D2 = 40,7xFPxFRxFI m; P(4) 2,042 Thefactoranalysisreproducestheexperimentaldataquitewell,withasmalldifference AD2 = D1-D2. The standard deviation is 8 m, and 6% in terms of percentage difference. 109 COST 331 Varianceanalysis(notshown)confirmsthatthevariationofthevisibilitydistanceis accountedforby the parameters, with only a small remaining variation. Eachparameterhas a strongly significantinfluence. It may be notedthat the variationof the visibilitydistanceis accountedforby theparameters inanindependentway.Forexample,achangeofroadmarkingpatternhasacertaineffect regardless of the retroreflectionlevel. It may also be notedthat the amount of roadsurfacecovered by road marking, as reflectedby the patterns P(l), P(2), P(3) and P(4), is as importantas the retroreflectionlevel. As an example, the patternP(3) correspondsto 4 times more roadmarkingsurfacethanP(l), whiletheretroreflectionlevelRL(2)correspondstoanaverageof4,3timeshigher retroreflectionthanRL(1)- The factorvalues are respectively1,503and1,549. These two featuresindicate that the conditions of the experimentare in the domainof Ricco's law, where size is as importantas luminance (see section A.5). B.5Comparison to the basic equation forvisibility ThebasicequationA.l,seesectionA.2,canbeappliedwhenvaluesfortargetsize,target luminance and backgroundluminance are available. In this connection, the target is the roadmarking while the background is the roadsurface. The luminance provided by a single headlamp at a given location is determinedas the product of the illuminanceat the locationand the relevantRL value(due to the definitionofRL).The illuminanceisto be determinedona plane perpendicularto thedirectionofillumination(in practice a vertical plane will do). The RL value must correspondto the geometricalsituationand,atleastin principle,thetotal luminanceprovidedbymorethanoneheadlampsimultaneouslymustbedeterminedasthe sum of individualcontributions. However,whentheheadlampsaremountedatthesameheight,thesameRLvalueapplies, andthetotalluminancemaybefoundasthetotalilluminancetimestheRLvalue. Thetotal illuminance,ontheother hand,isthe totalluminousintensityI oftheheadlampsdividedby the distance squared D2. This results in the followingexpression: L=IXRL/ D2(Equation B. 1) whereLis the luminance of the road marking or the road surfaceat a location. Iis the total luminous intensity of headlamps towards the location. RLisanRLvalue reflectingthe roadmarkingorthe roadsurfaceandthe geometricalsituation, andDis the distance fromthe headlamps to the location. The RL values of table B.lwere measuredin the standard geometry specifiedin the European standard EN1436, correspondingto a headlamp mounting height of 0,65m,an observereye height of1,2 m and a distance of 30 m. 110 Annex B Thetestvehicleaimsatthesameheightsofheadlampandobservation,whiletherelevant distancescover a range fromabout 40 m and upwards. However,accordingto the report No. 6oftheNordicResearchCooperationfornighttraffic,theRLvalueisroughlyconstantat distances from30 m and upwards. Accordingly,measuredRL valuesgivenintableB.lareuseddirectlyinequationB.lto provide luminance values. Remainingquestionsare addressedinviewofthe factthatconditionsduringtheexperiment are inthedomainofRicco'slaw, wherethestimulusto the eye istheilluminanceat theeye fromthe target. TheilluminanceattheeyedEfromashortsectionofaroadmarkingoflengthdData distance D is given by Ix RLxWxHoXdDxD~5, where W is the widthofthe line andH0 is the eye height. For a continuousline, whenassuminga constantintensity I ofthe headlamps, the total illuminance becomes: E= 0,25xlxRLXWxHoXdDxD/xUjO-iDih)4)(Equation B.2) whereDlandD2aredistancestothefrontandthebackendsofthemarking respectively Thevalue ofthe lastterminthis expression(l,0-(Di/D2)4)rangesfrom0to unity fora very shortandaverylongroadmarkingrespectively.Itisinterestingthatrelativelyshortroad markingsprovidealmostthefullvalueoftheterm.Forinstance,whenthelengthD2-D1 is 50% of the distance to the frontDi, the value is as high as 0,82. Thisshows that those parts of the roadmarking closest to the observer contribute muchmore to visibility than parts furtheraway. Theroadmarkingsintheexperiment,ofa lengthof75m,may beconsideredasverylong, except perhaps in cases corresponding to the longest visibility distances. Furthermore, the intensity value I of the headlamps towards the frontend of the roadmarking is more important than values towards more distant locations. When consulting table B.l,this shows that an assumptionof a constant intensity is relativelysafe. Inany case,theilluminanceattheeyecanbedeterminedfora continuouslinebymeansof equationB.2. A circulartargetofa uniformluminanceL anda size given by a solidangle co provides the same stimulus when Lxco = E, where co is measured in steradians. AsensiblechoiceofL istheluminanceofthe roadmarkingatthe front,asthisisprobably wheretheobserverislookingatthemomentofdetection.Thesolidangleisdeterminedby co =E/Ltobeconvertedtothediameteraofacirculartarget.Theluminanceoftheroad surfaceat the same location serves as the background luminance Lb. ExpressionsliketheoneofequationB.2couldbederivedforthebrokenlinesofpatterns P(l)and P(2). However,a detailed analysis of a pattern at a certain instant of time is probably not justifiedinviewofthe visualprocess,whichcertainlytakessometime. Inany case, the resultswouldbeclosetothoseobtainedbyequationB.2foracontinuouslineofthesame surfacearea as the broken line. Il l COST 331 Accordingly, the broken lines of patterns P(l) and P(2) are handled as continuous lines with reduced widths of respectively 2,5 and 5 cm. On this basis, visibility levels VL are computed for the visibility distances Dlgiven in table B.3. The VL's turn out to be on average 7,2 with some variation from case to case (values are not shown). Adjusting the distances to provide a VL of 7,2 results in the computed distances D3 also given in table B.3 and shown in figure B.5 as well. Thefitbetweencomputedandexperimentalvisibilitydistancesisgood,withsmall differencesAD3 = D1-D3. The standarddeviationis7 m,and5% intermsofpercentage difference. The fit issurprisingly good in view of the adjustmentof only one parameter value(VL of 7,2), translation from complex to simple conditions, uncertainty regarding true conditions and approximations regarding constant headlamp intensity I and length of road markings, etc. This result indicates that the translation is sensible, that conditionsare well controlledand that approximations are permissible. It may be noted that the only real uncertainty of the translationis the choice of background luminance, which is not very important in the conditions of the experiment. Itmayalsobenotedthatitdoesrequirequitelargechangesinconditionstoproducea significantchangeinthevisibilitydistance.Thiscanbeverifiedbyexamplesbasedon equation B.2, showing that the stimulus changes in proportion to the distance to the power of -4. Finally, a VL of 7,2 has been used because it provides the best fit to the experimental data. Otherwise,thevaluecannotbejustifiedandmaycontainsomesortofcorrectionfor conditionsorapproximations. However,it isnotunreasonablethattheexperimentshould resultina lowervalueofVL thanthevaluerecommendedforpracticaltrafficsituations (VL = 10, see section A.5 of Annex A). 112 Annex B mc a l c u l a t i o n(D3) 300 200 100 100200 exper i ment(D3) 300m Figure B.5 -Comparison between experimental and calculated visibility distances 113 AnnexC Annex CExample of calculation of visibility distance The example developed in this annex is for night driving on high beam with thefollowing assumptions and data: 1)Youngdriver-implyingthatsubstitutionsforagedescribedinannexAarenot required. 2)There is no glare or other veiling luminance - implying that the substitution for glare described in annex A is not required. 3)The headlamp intensity I is constant at 10.000 cd in all relevant directions. 4)The vehicle is a passenger car with two headlamps and an observer eye height H, of 1,2 m. 5)Theroadmarkingiscontinuous, hasa widthWof0,1mandstartsatadistance D = 96 m in front of the driver. 6)The roadsurfacehas an RL value of15 mcdm^lx"1 which implies a relatively, but not very dark, road surface. 7)The roadmarking hasanRL valueof100 mcdm^lx"1 justmeetingtheminimum requirement of class R2 in the European standard EN 1436. 8)The vehicle geometry is such that RL values measured for the standard geometry (i.e. accordingto EN1436) can be applieddirectly forboth headlamps, assuming some simplification. ThedefinitionofthecoefficientofretroreflectedluminanceisRL =UE,whereL isthe luminanceoftheroadmarkingorroadsurfacecreatedbyoneheadlamp,andEisthe illuminance(lx)createdby thatheadlampatthelocationoftheroadmarkingona plane perpendicular to the direction of illumination. Accordingly,theluminancecreatedbyoneheadlampisL=RLxE.Thedistancelawof illuminationsays that E = VD2,so that E = (10.000 cd)/(96 m)2 = 1,09 be. The luminance of the road surface created by one headlamp is L = (15 mcdm^lx"1) x (1,09 lx) = 16,4 med-m"2 =(16,4mcdm"2)/(1.000mcdm"2/cdm"2)=0,0164cdm"2.Thetwoheadlampstogether producetwicethisluminance,i.e.approximately0,033cdm"2.Inthesameway,the luminance of the road marking is 0,218 cdm"2. Having determined the luminances of the road surfaceand the road marking, the equivalent targetsizeoftheroadmarkingmustbedeterminedaswell.Thedefinitionsaysthatthe equivalent target size is the size of a circular target of the same solid angle (sr) as obtained by luminance weighted integration over the surface of the road marking. Thisintegrationisgenerallytedious;thecomputerprogrammestoredontheCD-ROM attachedtothisreportperformsitnumerically.Inthiscase,however,theintegralcanbe performed analytically; the result of ra = 0,25xHoXW/D2 is provided in chapter 5. This results in = 0,25x(l,2 m)x(0,l m)/(96 m)2 = 0,0000032 sr. 115 COST 331 It should be taken into account that the driver sees the road markingas a triangle witha base equaltothelinewidthWandaheightequaltohiseyeheightHo,havinganareaof 0,5XHQXWand a solid angle of 0,5xHoXW/D2. This result is true if the luminance of the road marking is constant. In headlamp illumination, however, this is not the case making the factor value decrease from0,5 to 0,25. Theequivalenttargetsizeaismeasuredastheangulardiameteroftheequivalentcircular target in the unit of minutes of arc ('). According to chapter 5: a= 3879Vu) which leads, for this example, to a value for a= 6,94'.___ Finally, the visibility level VL is foundby use ofthe basicequationVL =ALxa2/(A+Bxa)2. Thevalue forthe equivalenttargetsizeaof6,94'canbeinserteddirectly.Thevalueofthe luminancedifferenceAL is determinedas the differencebetweentheluminancesofthe road markingandthe roadsurface,leadingto0,218- 0,033 cdm2=0,185cdxm"2.AandBare functionsof the background luminance Lb,for which the road surfaceluminance serves. The expressions for A and B are: fA=log(10,086xLb0>2509)+0,27154xLb0'5867 f > 0,6:< IlB =o.ogsssxLb0'466 I I logA =0,2355+0,173xlogLb U{< 0,00418:{ II logB =-0,6835+0,5275xlogLb+0,0227(logLb)2 I If logA =0,1355+0,3372xlogLb+0,0866(logLb)2 I in between:{ [ logB =-1,0485+0,3190xlogLb The third case 'in between' is relevant for the value of 0,033 cdm"2. Results are A = 0,670 and B = 0,030. The basic equation gives: VL = (0,185 cdm"2)x(6,94')2/(0,670+0,030x6,94')=10 Thisisthevisibilitylevelrequiredforrealtrafficsituations,referredtoinchapter5. Accordingly,thevisibilitydistanceisactuallythe96musedintheexample.Itshouldbe notedthat repeating the calculations forother distancesshowsthat VL isstronglydependent on the distance. 116 Annex D Annex DVisibilityof longitudinalroadmarkingsinheadlamp illumination D.lIntroductionanddiscussion Thisannexcontainsananalysisofvisibilitydistancesoflongitudinalroadmarkingsin headlamp illumination. The longitudinal road markings have the geometries accounted for in fifteencountries in the state of the art report (refer to section D.2). The visibility distances have been calculated using the programme described in Appendix 4. Refer to sections D.3 and D.4 regarding conditions for calculations, results and discussion. Someofthecalculationsarecarriedoutforidealconditions,assumingayoungdriver, powerful headlamps and absence of glare and other veiling luminance. For such conditions, the visibility distance is mostly close to the distance at which the cutoff of the low beam of the headlamps meets the road. In the circumstancesof the calculations, this distance is about 60 m for road markings to the leftof the driver, and about100 m for road markings to theright5. Assuming that drivers need a minimum of 2 second preview time (see chapter 6), the shorter of the above-mentioned distances of 60 m allows driving speeds up to about110 km/h. The longer distance of 100 m allows high driving speeds. The almost constant visibility distance is obtained in spite of wide ranges of effectivewidths and reflectivity(coefficientof retroreflectedluminance, RL). Visibility distances in excess of the cut-offare obtained only for combinations of large effectivewidths and high reflectivity. Other calculationsare carried out for less good conditions exemplifiedby greater age of the driver, lessefficientheadlampsandoccurence ofglare. For theseconditions, the visibility distance decreases down to about half of the distance of the cut-off of the low beam. Such conditionsare realistic, more commonthan ideal conditions, and will occur formost roads, at least for some drivers in some periods. Even worse conditions occur during rain or wetness, but have not been included in the calculations. The distance is longer for road markings to the right of the vehicle, due to the elevated part of the beam being to the right for right-hand traffic. For left-hand traffic, the longer distance is for road markings to the left. 117 COST 331 D.2Geometryoflongitudinalroadmarkings Thegeometriesforfifteencountriesdescribedinthestate oftheart reportare included.The countries are indicatedin table D.l,while geometries are accountedforin tables D.3, D.4 and D.5formotorways,interurbandualcarriagewayroadsandinterurbansinglecarriageway roads respectively. TablesD.3,D.4andD.5showalsotheeffectivewidthoftheroadmarkings,whichisthe averagewidthalongtheroadtakinggapsofbrokenlinesintoaccount.Therangesofthe effectivewidth are given in Table D.2. Theabove-mentionedtablesshowthatthefifteencountriesuseseveraldifferentroad marking geometries and that the ranges of effectivewidths are quite large. Table D.l-Fifteencountries. Symbol: B CH D DK E F FIN G ICE IRL NL P S SLO UK Country: Belgium Switzerland Germany Denmark Spain France Finland Greece Iceland Ireland the Netherlands Portugal Sweden Slovenia UnitedKingdom Table D.2-Rangesof effectivewidths. Road type: Motorways Dual carriageway Single carriageway Line type: leftline laneline rightline leftline laneline rightline centre Une edgeline Effectivewidth: 7,5 to 30 cm 2,2 to5 cm 7,5 to 30 cm 3,3 to 30 cm 1,7to10 cm 3,3 to 30 cm 2,5 to10 cm 3,3 to 30 cm 118 Annex D Table D.3 Geometry of longitudinal road markings on motorways. Effective width Left line: 7,5 cm 20cm 22,5 cm 25cm 30cm Lane line: 2,2 cm 2,5 cm 2,9 cm 3,3 cm 3,5 cm 3,8 cm 4cm 4,3 cm 5cm do Right line: 7,5 cm 14,3 cm 16,4 cm 20cm 25cm 30cm Geometry Length 20 m continuous continuous continuous continuous 2m 3m 5m 4m 2m 3m 3m 2,5 m 4m 5m 6m 20m 20m 38m continuous continuous continuous Gap 20m continuous continuous continuous continuous 7m 9m 12m 8m 7m 10m 9m 10m 10m 10m 12m 20m 4m 14m continuous continuous continuous Width 15cm 20cm 22,5 cm 25cm 30cm 10cm 10cm 10cm 10cm 15cm 15cm 15cm 20cm 15cm 15cm 15cm 15cm 20cm 22,5 cm 20cm 25cm 30cm Country IRL CH+E+FIN+ NL+P+SLO+ UK F CH+G B+D+DK+S UK NL+FIN E IRL UK F G+S B P CH+DK D+SLO IRL E F CH+FIN+NL +P+SLO+UK CH+G B+D+DK+S 119 COST 331 Table DA-Geometryoflongitudinalroad markings oninterurban dualcarriageway roads. Effective width Left line: 3,3 cm 7,5 cm 10cm 12cm 15cm 18cm 20cm 25cm 30cm Lane line: 1,7cm 2,2cm 2,5 cm 2,8 cm 2,9 cm 3,3 cm 3,3 cm 3,8 cm 4cm 4cm 4,3 cm 5cm 5cm 6,7 cm 10cm Right line: 3,3 cm 7,5 cm 10cm 10,6 cm 12cm 15cm 16,7 cm 20cm 25cm 30cm 1} used only on 'it is not quite Geometry Length 1m 20m continuous continuous continuous continuous continuous continuous continuous 1m 2m 3m 3m 5m 5m 4m 3m 4m 2,5 m 4m 5m 6m 6m continuous 1m 20m continuous 20m continuous continuous 20m continuous continuous continuous roads with speed lin certain that this line Gap 2m 20m continuous continuous continuous continuous continuous continuous continuous 5m 7m 9m 10m 12m 10m 8m 9m 8m 10m 10m 10m 12m 12m continuous 2m 20m continuous 14m continuous continuous 4m continuous continuous continuous oit of 60 km/h - is continuous Width 10 15 10 12 15 18 20 25 30 10 10 10 12 10 10 10 15 12 20 15 15 15 20 10 10 15 10 18 12 15 20 20 25 30 cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm Country S IRL NL+ICE+DK D NL+SLO+UK F CH+E+FTN+ G+P+UK CH B+DK UK1* UK FIN+ICE+S F E DK G+IRL S D B P SLO CH CH NL2) S IRL ICE+NL+DK F D NL+SLO+UK E CH+FIN+G+ P+UK CH+D B+DK 120 Annex D Table D.5 Geometryoflongitudinal road markings oninterurbansinglecarriageway roads. Effective width Centre line: 2,5 cm 2,5 cm 2,8 cm 2,8 cm 3cm 3,3 cm 3,3 cm 3,3 cm 3,4 cm 3,6 cm 3,8 cm 4cm 5cm 5cm 5cm 10cm 10cm 20cm Edge line: 3,3 cm 5cm 7,5 cm 8,3 cm 10cm 12cm 15cm 20cm 25cm 30cm l)it is not quite cei 2) double continuo! Geometry Length 2m 3m 3,5 m 3m 2,5 m 2m 4m 5m 4m 4m 3m 4m 2m 3m 5m continuous continuous continuous 1m 3m 20m 3m continuous continuous continuous continuous continuous continuous tain that this line i JS lines have been Gap 6m 9m 9m 10m 10m 4m 8m 10m 10m 7m 9m 8m 4m 6m 10m continuous continuous continuous 2m 3m 20m 3,5 m continuous continuous continuous continuous continuous continuous s continuous added Width 10cm 10cm 10cm 12cm 15cm 10cm 10cm 10cm 12cm 10cm 15cm 12cm 15cm 15cm 15cm 10cm 2*5cm 2*10cm 10cm 10cm 15cm 18cm 10cm 12cm 15cm 20cm 25cm 30cm Country CH+ICE FIN+S E F B UK G DK P G ERL+S D UK CH+UK SLO NL 2) 2) S ICE IRL F FIN+ICE+ NL+UK+D K D CH+E+NL + P+SLO+U K B+CH+G B DK 121 COST 331 D.3Calculationsandresults Calculations are performedby means of the computer programme (described in appendix 4) with use of the following input: 1)The right-hand trafficoption of the computer programme. Results apply forleft-hand traffic, when considering the left line to be the right Une and vice versa. 2)Ideal,mediumandadverseconditionsasdescribedintableD.6.Actualuseof conditions is mentioned in table captions. 3)The 'passenger car' option of the computer programme. 4)The 'low beam' headlamp option of the computer programme. 5)Straight and flat roads. 6)Positions of the road marking 1,5 m to the leftof the vehicle for leftUnes, lane lines and centre lines, and 1,5 m to the right of the vehicle for right lines and edge Unes. 7)Effective widths of the road markings as accounted for in the previous section. 8)AnRL value forthe roadsurfaceof15 mcdm"2 lx"1 meantto representa relatively dark,butnotverydark,asphaltconcretesurface.AnadditionalRL valueof 30 mcdm"lx~used for motorways is meant to represent cement concrete surfaces. 9)RL values for the road marking of100, 150, 200 and 300 mcdm^lx"1 corresponding toclassesR2, R3, R4 and R5inEN1436 of re Actual use of conditions is shown in table captions toclassesR2, R3, R4 and R5inEN1436 ofrelevanceforwhiteroadmarkings6. Results of tables D.7 and D.8 are for motorways, while results of tables D.9 and D.10 are for dual carriageway interurban roads and single carriageway interurbanroads respectively. All tables include results for 'ideal', 'medium' and 'adverse' conditions as described in table D.6. DADiscussionofresults Figure D.lshows the reachofthe cut-offoflowbeam headlamps, whenmountedon the passenger car. The reachisabout 60 m and100 m forroadmarkingsto the leftandright respectively. The reach explains the visibility distance ofsome ofthe cases oftablesD.7, D.8, D.9 and D.10, where the road marking needs to have some part within the cut-off. Only combinations of'ideal' conditions, large effectivewidthand highreflectivityleadstovisibilitydistances beyond the cut-off. In some cases, on the other hand, the visibility distance is not close to the reach of the cut-off and may be as low as half of that reach. Thetablesalsoprovidethevisibilitydistancesthatareneededfor2,3and5seconds of preview at some relevant driving speeds. 6The above-mentioned RL values are in the scale of the standard measuring geometry on EN 1436 and do not necessarily apply at the actual conditions of the tables. 122 Annex D Forthe'ideal'conditionsoftableD.6,thedriverisyoung,theheadlampintensities correspond to powerful headlamps, and there are no oncoming vehicles causing glare. More relevantsituations may be reflectedby the'medium' and'adverse' conditions of table D.6, where drivers are not young, headlamp intensities are lower due to degradation and dirt and oncoming vehicles cause glare. Sometimes conditions are even worse, e.g. during rain or wetness, or during winter in some countries. d r i CZ car ( r i verposi t i on about 3 10m ght -handt r a f f i c ) abo ut60m z^-"^l e f tmar ki ng | 3m -~____^r i g h tmar ki ng about97in Figure D.l- Reach of the cut-off oflow beam headlamps fora passenger car driving in a 3 m wide lane. Table D.6 -Three sets of conditions labelled'ideal','medium ' and'adverse'. Conditions Ideal Medium Adverse Age of driver young 50 year 60 year Headlamp intensity*) 100% 50% 30% Glare Lv no glare 0,25 cd-rn-2 0,50 cdm"2 *) refersto percentage of10.000 cd in directions below the cut-offof the low beam 123 COST 331 Table DJ-Visibility distance of longitudinal road markings on motorways.The RL of the road surface is 15 mcdm'2'lx'1'. Motorways Effective width: Leftline: 7,5 cm 20cm 22,5 cm 25cm 30cm Lane line:2) 2.2 cm 2,5 cm 2,9 cm 3.3 cm 3,5 cm 3,8 cm 4cm 4,3 cm 5cm Right line: 7,5 cm 14.3 cm 16.4 cm 20cm 25cm 30cm RL of road marking(mcdm^lx1): 100150200300 visibility distance (m) for adverse-medium-idealconditions!) 42-56- 59 49-60- 66 49-60- 68 50-60- 70 52-60- 73 34-45- 54 35-46- 54 36-48- 55 36-49- 56 37-50- 56 37-51-57 38-51- 57 38-52- 57 39-54- 58 42-58- 79 47-66- 87 48-67- 89 49-70- 91 50-72- 94 52-75- 96 49-60- 61 57-60- 76 58-60-78 59-60- 80 60-60- 83 39-49- 57 40-51-57 41-52- 58 42-53- 58 42-54- 58 43-55- 58 43-55- 59 44-56-59 45-58- 59 49-67- 86 55-75- 95 56-77- 96 57-80- 98 59-83-101 61-84-102 53-60- 67 60-60- 84 60-60-86 60-60- 88 60-60- 93 42-53- 58 43-54- 58 45-55- 59 46-57- 59 46-57- 59 47-58- 59 48-59- 59 48-59- 60 50-60- 61 54-73- 91 61-82- 99 62-84-100 64-84-102 66-84-104 68-84-105 60-60- 76 60-60- 97 60-60-100 60-60-103 60-60-109 47-57- 59 49-58- 59 50-59- 61 51-60-63 52-60- 63 53-60- 65 53-60- 65 54-60- 67 55-60- 69 62-81- 97 69-84-104 71-84-105 73-85-106 76-89-107 78-92-109 1}Visibilitydistancesapply fornight-timedrivingina passengercaronlowbeam, onastraightandplaneroad.Adverse,mediumandidealconditionsrelatetothe age of the driver, to the intensity of the headlamps and to glare; refer to table D.6. 2)Lane lines are placed to the leftof the vehicle. required visibility distance: 110 km/h:120 km/h:130 km/h: 2 s preview time:61m67 m72 m 3 s preview time:92 m100 m108 m 5s preview time:153 m167 m181m 124 Annex D Table D.8 -Visibilitydistanceoflongitudinal road markingson motorways.TheRL of the road surface is 30 mcd-m'2 lx'1. Motorways Effective width: Left line: 7,5 cm 20cm 22,5 cm 25cm 30cm Lane line: 2) 2.2 cm 2,5 cm 2,9 cm 3.3 cm 3,5 cm 3,8 cm 4cm 4,3 cm 5cm Right line: 7,5 cm 14.3 cm 16.4 cm 20cm 25cm 30cm RL of road marking (mcdm^lx1): 100150200300 visibility distance (m) for adverse-medium-ideal conditionsl) 39-51-57 45-60- 60 45-60 -67 46-60 -62 47-60 -65 31-40-49 32-41- 50 33-42- 51 34-44-52 34-45- 53 34-45. 53 35-45- 53 35-46- 54 36-48- 55 39-53-72 43-60- 80 44-62- 81 45-64- 83 46-66- 87 47-68- 89 46-59- 59 54-60-71 55-60- 73 56-60-75 57-60- 78 37-47- 54 38-48- 55 39-49- 56 40-51-56 40-51-56 41-52- 57 41-53- 57 42-53- 57 43-53- 58 47-63- 81 52-72- 90 53-73- 91 55-76- 94 56-79- 97 58-82- 99 51-60-63 60-60- 80 60-60- 82 60-60- 84 60-60- 88 41-51-56 42-52- 57 43-53- 57 44-54- 58 45-55- 58 45-56- 58 46-56- 58 46-57- 59 48-59- 59 52-70- 87 58-79- 96 60-81- 97 62-84- 99 64-84-102 66-84-103 58-60- 73 60-60- 92 60-60- 95 60-60- 97 60-60-101 46-55- 58 47-56- 59 49-58- 59 50-59- 59 51-59-59 51-60-62 52-60- 63 53-60-64 54-60- 66 60-79- 95 68-84-102 69-84-103 72-84-104 74-87-106 77-90-107 1)Visibility distances apply for night-time driving in a passenger car on low beam, on a straight and plane road. Adverse, medium and ideal conditions relate to the age of the driver, to the intensity of the headlamps and to glare; refer to table D.6. 2)Lane Unes are placed to the left of the vehicle. required visibility distance: 110 km/h:120 km/h:130 km/h: 2 s previewtime:61 m67 m72 m 3 s previewtime:92 m100 m108 m 5 s previewtime:153 m167 m181m 125 COST 331 Table D.9 -Visibility distance of longitudinal road markings on interurban dual carriageway roads.The RL of the road surface is 15 (mcdm2ix'1). Dual carriageway roads Effective width Left line: 3,3 cm 7,5 cm 10cm 12cm 15cm 18cm 20cm 25cm 30cm Lane line:2) 1.7 cm 2.2 cm 2,5 cm 2.8 cm 2.9 cm 3.3 cm 3,8 cm 4cm 4,3 cm 5cm 6,7 cm 10cm Right line: 3,3 cm 7,5 cm 10cm 10.6 cm 12cm 15cm 16.7 cm 20cm 25cm 30cm RL of road marking (mcdm"2lx"'): 100150200300 visibility distance (m) for adverse-medium-ideal conditions" 36-48-56 42-56-59 44-59-