TheKenyamaintenancestudyonunpavedroads: researchondeterioration by T. E.Jones TRANSPORTandROAD RESEARCHLABORATORY DepartmentoftheEnvironment DepartmentofTransport TRRLLABORATORYREPORT1111 THEKENYAMAINTENANCESTUDYONUNPAVEDROADS: RESEARCHONDETERIORATION by TEJones TheworkdescribedinthisReportformspartoftheprogramme carriedoutfortheOverseasDevelopmentAdministration,but theviewsexpressedarenotnecessadythoseoftheAdministration. OverseasUnit TransportandRoadResearchLaboratory Crowthorne,Berkshire 1984 ISSN0305–1293 CONTENTS Abstract 1.Introduction 2.Designofthestudy 2.1Background 2.2Experimentaltestsections 3.Measurementofexperimentalvariables 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11Gravelloss Surfacelooseness Surfaceroughness Rutdepth Journeytimes Trafficvolumes Trafficloading hboratorytestingofmaterials In-situtestingofmateriak Ctimate Roadgeometry 4.Analysisanddiscussionofresults 4.1 Gravelloss 4.1.1 4.1.2 4.1.3 4.1.4 4.1.5 4.1.6 4.1.7 4,1.8Gravellossfromunpavedroads Soilerodibitity Groundslope Rainfa~factor Parametricpredictionofgravelloss Comparisonofresultswiththosefromaneartierstudy Resultsfromothercountries RegraveMngrequirementsPage 1 1 2 2 2 4 4 4 4 4 5 5 5 6 6 6 6 6 6 6 6 7 7 7 8 9 9 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 Surfacelooseness Surfaceroughness 4.3.1Surfaceroughnesschangesforeachgraveltypeinvestigated 4.3.2ComparisonwithpreviousTRRLstudy Rutdepth 4.4.1Developmentofrutswithcumulativeloading Journeytimes 4.5.1Speedmeasurementsoverthetestsectionsatthebeginningand endofthestudy 4.5.2Comparisonswithpreviousstudiesofvehiclespeedsonunpavedroads Trafficloading 4.6.1Afleloadsurveymeasurements 4.6.2Heavilyloadedsitevehiclesusedinroadconstruction ~mate 4.7.1Changesinroughnessrelatedtocumulativerainfallforuntraffickedroads 4.7.2Changesinroughnessrelatedtocumulativerainfdfortraffickedroads Geometry 4.8.1Geometricconstra~tsondeterioration RoadCamber 4.9.1Changesinroadcamberduringthestudy Shouldercompaction 5. Summaryandconclusions 6.Acknowledgements 7.References 8.Appendix1 9.Appendix2 10.Appendix3Page 9 10 10 11 13 13 13 13 14 14 14 15 15 15 16 17 17 17 17 18 18 18 19 41 43 46 @CROWNCOPYRIGHT1984 Extractsfromthetextmaybereproduced,exceptfor commercialpurposes,providedthesourceisacknowledged THEKENYAMAINTENANCESTUDYONUNPAVEDROADS: RESEARCHONDETERIORATION ABSTRACT Thisreportdescribesastudyofthedeteriorationofunpavedroadswhichwas undertakeninKenya.Theworkformedpartofalargerstudyoftheeffectiveness ofvariousmaintenancestrategiesontheperformanceofunpavedroads. Theprimaryobjectiveoftheresearchdescribedinthisreportwasto providenewandimproveddeteriorationrelationshipsforunpavedroadsforuse inacomputermodelforestimatingtheconstructioncosts,maintenancecosts andvehicleoperatingcostsofroadsindevelopingcountries. Experimentaltestsectionswerelocatedonthepublicroadnetworkin Kenyaandtheirratesofdeteriorationweremeasuredandquantifiedintermsof gravellossofthewearingcourse,surfaceroughness,deptiofloosesurface materialandrutdepths.Thetestsectionsweremonitoredforaperiodofover 2years.Thedeteriorationofthetestsectionswasrelatedtocumulativetraffic, originaldesignstandardsandconstruction,maintenancestrategyandclimate. 1.INTRODUCTION Indevelopingcountries,unpavedroadscanaccountforover90percentofthetotalroadnetwork.Inmanyof thesecountries,notonlyarethesecondaryandfeederroadsunpaved,butdsoasignificantproportionofthe primaryroads. Unpavedroadscanconvenientlybecategorisedaseitherengineeredgravelroadsorearthroads. Engineeredgravelroads,astheirnameimpties,aredesignedandbudttoanengineeringspecification.They arenorrndlyconsideredtobe‘aflweather’roads.Earthroadsarebuflttoawidespectrumofconstruction standards.Thesecanrangefrombeingsimflartotheengineeredgravelroadspecification,butwithoutanall- weathersurfacing,downtothewideningofafootpathorafarmtrack.Thisreportisconcernedonlywith engineeredgravelroads. Themaintenance~equirementsofanyunpavedroadwtildependlargelyonitsoriginalconstructionstandard, thetrafficthatitcarriesandtheprevailingclimate.Itfoflowsthatroadswithdifferentsurfacingmaterialsw1ll havedifferentratesofdeteriorationandthatthestandardofthemaintenancecarriedoutwfllaffecttherateof deterioration. In1979,theOverseasUnitoftheTransportandRoadResearchhboratory(TRRL),inconjunctionwith theKenyanMinistryofTransportandCommunications(MOTC),initiatedastudyontheeffectofvarious maintenanceinputsonthedeteriorationratesofunpavedroadsinKenya. Thestudywasdesignedtoinvestigatethetechnicalaspectsofthemaintenanceofunpavedroadsandwas notdirectlyconcernedwithorganisationalandmanagementissues. ThisReportdescribesthatcomponentofthestudyconcernedwiththedevelopmentofnewgravelroad deteriorationrelationshipsforincorporationintotheTRRLroadinvestmentmodelfordevelopingcountriesl. Thisincludedaninvestigationofthettie-relatedweatheringofgravelroadswhichwereuntrafficked.Other aspectsofthestudyconcernedwiththedevelopmentofoptimumgradingstrategiesforarangeofparameters andequipmentarethesubjectofaseparatereport2. Theresultsofthestudy,althoughofdirectbenefittoKenya,wdlalsobeapplicabletomanyother developtigcountries. 2.DESIGNOFTHESTUDY 2.1Back&ound Between1971and1974,ajointWorldBank/TRRL/MOTCprojectstudiedtheinter-relationshipsbetween construction,maintenanceandvehicleoperatingcostsforbothpavedandunpavedroads3‘4inKenya.This resultedinthedevelopmentofacomputermodeldesignedtoaidinvestmentdecisionswithintheroadsectorand whichisknownastheTRRLRoadInvestmentModell.Amodelhasalsobeendevelopedonthebasisofthiswork bytheWorldBankandthisisknownastheHighwayDesignandMaintenanceStandardsModels. Thecurrentstudyrequiredthedeterminationandquantificationofrelationshipsbetween:– Gravelloss Surfacelooseness Surfaceroughness Rutdepth Journeytimes and:– Trafficloading Climate Geometry Camber Shouldercompaction FieldworkstartedbycarryingoutanappraisaloftheconditionoftheexistinggravelroadnetworkinKenya. Modesofdeteriorationoftheroadswereidentifiedandmethodsofmeasuringtheseweredeveloped.Testsections ofroadwereselectedandtheexperfientalworkandmonitoringofroadperformancecommenced.Resultswere co~ectedoveraperiodoftwoyears.Thesewerethenanalysedandconclusionsdrawn. 2.2Experimentaltestsections ThestudywasconcentratedonthreegeographicalareasofKenyawheretheMOTCwerecarryingout rehabditationworksonlargenetworksofunpavedroads.Thisrehabilitationprogrammewaspartofjointprojects betweentheMOTCandtheCanadian(CIDA)andAmerican(USAID)donoragencies.Theuseoftestsectionswas necessaryinordertorelatethedeteriorationratesofunpavedroadstothestandard,frequencyandtypeof maintenance,thetrafficspectrum,thenaturalenvironmentandthephysicalpropertiesoftheparticulargravel. Inaddition,thetestsectionswereneededtoincludevariationsingeometricdesignstandardsintermsofvertical gradientandhorizontalcurvature.Therangeofgeometricstandardsstudiedwasbasedontheresultsoftheinitial 2 fieldappraisalandcoveredtypicalgeometricstandardsfoundineachexpefientalarea.Thecombinationof horizontalandverticalalignmentsstudiedisshowninTable1. TABLE1 Geometricframework Horizontalcurvature LowMediumHigh <300/km30°–900/km>900/km Flat<1% Vertical Intermediate1–3% gradient Steep>3% Theannualrainfallonthetestsectionsvariedbetween500and2000mmperyearand,ineachexperimental area,testsectionswereduphcatedinorderthatdifferentlevelsofmaintenancecouldbecarriedout;dltestsections were300metresinlength.Thelevelsofmaintenancechosenwereasfoflows:— (a)High:gradedeverysixmonths, (b)Normal:gradedeveryninemonths, (c)Nil:notgradedduringthestudy. Inadditiontotheselevels,itwasfoundpossibletoincludeasmaflnumberoftestsectionswithmaintenance frequenciesofthreeandtwelvemonths.Inallcases,normaldrainageandroadsidemaintenancewerecarriedout onaroutinebasis.Oneachgroupoftestsections,variousmaintenanceinputswereappfiedatdifferentfrequencies. AtypicallayoutofanexpertientalsiteisgiveninFigure1. Itwasexpectedthatthepropertiesofthegravelwearingcoursewouldmakeasignificantcontributiontothe rateofdeterioration.Testsectionswerethereforechosentoutilisefourbasicgraveltypeswhichwereasfoflows:— (a)Latenticgravels.Theseareformedbytheaccretionofnodulesofoxidesofironandaluminium.Thetest sectionsintheBungomaareawerealllatcritic. (b)Qwrtziticgravels.TheseareroundedgravelsderivedfromthebasementrockoftheareaeastofMount KenyaandwerethewearingcoursematerialsusedonsomeofthetestsectionsintheMeruarea. (c)Volcanicgravels.TheseareangulargravelsderivedfromthevolcanicrocksoftheMountKenyaareaand wereemployedaswearingcoursematerialsonothertestsectionsintheMeruarea. (d)Sandstonegavels.Thesearefinegrainedgravelsusuallycontainingquantitiesofquartzandfeldspar foundneartheKenyacoast.ThewearingcoursematerialsontheKalolenisiteswereallsandstoneinorigin. ThelocationoftheexperimentaltestsectionsisillustratedkFigure 2. 3 3.MEASUREMENTOFEXPERIMENTALVARIABLES 3.1Gravelloss Therateofgravellosswasrecordedastheverticallossinmfilimetresofmaterialfromtheroadsurface. Measurementsweremadeatthreemonttiyintervalsandwerecarriedoutusingopticalsurveytechniques. Ineachexperimentalsection,representativegridswereselectedforthesurveys.At5metreintervalsalonga 60metrelength,profilesweretakenat25cmincrementsacrosstheroadandintothedrainagechannel.Atthe sideofeachtestsection,concretemonumentswereinsta~edpar~eltothecentreofthegridandapproximately 15metresfromthecentrelineoftheroad(seeFigure3).Anadditiondbenchmark,consistingofa300mm squaremetalplate10mmthick,wasplacedintheroadbeyondthesurveygridsatsubgradeleveltomonitorany differentialmovementbetweentheroadstructureandtheconcretebenchmarks. 3.2Surfacelooseness Loosematerialonaroadwdlleadtolossoftractionwhichwfllaffectvehiclespeeds,journeytimesandfuel consumption.Consequently,measurementsofsurfaceloosenessweremadeimmediatelybeforeandafterany maintenanceinputonthesections.Thesurfaceloosenessonthenflmaintenancesectionswasrecordedatthree montMyintervals.AUmeasurementswerecarriedoutatthreepoints:atthebeginningandendoftheindividud gridareas,anddsoattheendofthesectionitself. Themeasurementswerecarriedoutbysettinga1metrex0.25metremetalframeoneachwheeltrackwith the1metrelengthplacedtransverselyacrosstheroad.Nloosematerialwithintheframewasthensweptintoa calibratedcylinderandrecordedastheaveragedepthofloosematerialwithintheframe.Additionalmeasurements weremadetodeterminethemoisturecontentoftheloosematerialandtheCBRvalueofthetop15mmofthe surface. 3,3Surfaceroughness Measurementsofsurfaceroughnessweremadeusinga5thwheeltowedbump~tegrator6’7supplemented byavehiclemountedintegratorunit.SeePlates1and2.Thetowedbumpintegratorwaspufledbyaspecially modifiedFordTransitvanwhilstthevehiclemountedunitwasinstafledinaFordCortinaMkIVEstatecar.The vehiclemountedunitwascalibratedperiodicallyagainstthetowedbumpintegratorunitandatypicalcrdibration curveisillustratedinFigure4.Toensurethatthesuspensionofthetowedunitwasnotchangingwithtime,this instrumentwasalsocalibratedperiodicallyoverroadsingoodstructuralconditionbutexhibitingawidespectrum ofroughnesslevels. Inthestudy,threemeasurementsofsurfaceroughnessweretakenineachwheelpathofeachtestsection whenusingthetowedunitandineachdirectionwhenusingthevehiclemountedunit.Additionalmeasurements werealsotakenwiththetowedunitinthecentreoftheroad.Thefrequencyofmeasurementwasrelatedtothe variousmaintenanceinputs,buttheintervalswereneverlargerthanthreemonths. 3.4Rutdepth Therutdepthwasrecordedforeachwheeltrackusingatwometrestraightedgeandawedgecalibratedin millimetres.Theserecordingsweremadeattheendofeachtestsectionandatthebeginningandendofeach gravellosssurveygridwithineachtestsection,Inaddition,rutdepthswerealsomeasuredfromtheprofilesofthe 4., roadsurfacewhichwereplottedaspartofthegravellossassessment.Measurementsofrutdepthweretakenat thesamefrequencyasthesurfaceroughness. 3.5Journeytimes Journeytimeisausefulindexofthesurfaceconditionofaroad.Additionally,valueoftimeisanelement oftheroadusercostandwillvarywithvehiculartype,surfaceconditionandalignmentoftheroad3.Measurements weremadeofthespeedsoftheeverydaytrafficthatpassedoverthetestsections,Thiswascarriedoutbyplacing twoobserverswithsynchronisedstop-watchesatthebeginningandendofeachtestsection.Thesemeasurements weretakenovera12-hourperiodfrom6.00am–6.00pm,usuaHybeforeandafteranymaintenanceactivity. 3.6Trafficvolumes Classifiedtrafficcountsforvehiclestrave~ingineachdirectionweretakenineachgroupofexpertiental sectionsatthreemonthlyintervalsduringthestudy.TheclassificationofthetrafficisillustratedinTable2.Site trafficbelongingtotheconstructionunitswereclassifiedseparatelyfromthenormaltraf~cflows.Thesurvey periodswerefor5days(6.00am–6.00pm)and1night(6.00pm–6.00am).Preliminarysurveyshadshown thattrafficontheseruralroadsafter6,00pmwasminimal,usuallylessthan5percentofthe24hourperiod. TABLE2 Vehicleclassification Typeofvehicle Motorcars Lightgoods Mediumgoods HeavygoodsDescription Thisclassincludespassengervehiclesseatingnotmorethanninepersons (includingthedriver).Estatecars,taxis,andhirecarsaregenerally includedbutnot‘bnd-Rover’typevehiclesormini-buses. Goodsvehiclesoflessthan1500kgunladenweightorvehicleswitha payloadcapacityoflessthan760kg.Thisclassspecifica~yincludes ‘bnd-Rover’typevehiclesandmini-buses. Thisclassincludesau2-axledgoodsvehiclesofmorethan1500kg unladenweightorvehicleswithapayloadcapacitygreaterthan760kg. Ingeneralmediumgoodsvehiclesdifferfromlightgoodsvehiclesin thattheyhavemorethanonetyreateachendoftherearadeietwin- tyres. Thisclassconsists”ofallgoodsvehicleswithmorethantwoaxles. BusesThisclassconsistsofallregularpassengerservicevehiclesandcoaches. 3.7Trafficloading Thetrafficloadingoneachgroupoftestsectionswasmonitoredat4monthlyintervalsusingtheTRRL 9Thesurveyswerecarriedoutinconjunctionwith portableweighbridge8togetherwithadigitalread-outunit. theclassifiedtrafficcensusperiodsof5daysand1night.Theywerecarriedoutusingproceduresandtechniques developedbytheTRRLanddocumentedinRoadNote4010. 5 3.8Laboratoqtestingofmaterials Attiebeginningofthestudyandimmediatelyaftertheunpavedroadshadbeenreconstructed,samplesof thegravelbothfromtheroadworksandfromthequarrysourceswerecoflectedandtested.TesBinthelaboratory consistedofcompaction,plasticityandliquidlimit,particlesizedistributionandlinearshrinkagecarriedoutin accordancewithproceduresdescribedinBritishStandard1377/7511. 3.9In-situtestingofmaterials Themostfrequentlyusedin-situtestinthestudywasthedensitytestcarriedoutatthebeginningofthe studyandbothbeforeandaftereachmaintenanceoperationinvolvingcompaction.Duringinitialsamplingofthe roads,thistestalsodeterminedthethicknessesofthewearingcoursesafterconstruction.Inaddition,in-situCBR testswerecarriedoutonwearingcourseandsubgradematerials. 3.10Climate Theprevailingclimatewasmonitoredbysettingupweatherstationsadjacenttoeachgroupofexpertiental sections.Instrumentscapableofmeasuringboththevolumeandintensityofrainfallwereinsttiedateachweather station.Themeasurementofbothintensityandvolumewasnecessarytoidentifyanyrainstormwhichmight causedisproportionatedamagetotheroadsurface. 3.11Roadgeometry Onunpavedroads,changesinveticularbehaviourduetothetignrnentcanhaveaneffectontherateof deteriorationoftheroad.Itwasthereforenecessaryforeachgroupofsectionstohavesimflargeometry.Test sectionswerethereforechosenwithcontinuousverticalgradientsinonedirectionandthehorizontalgeometryof sectionsinanysin~egrouphadsimilarratesofchangeofcurvatureaswellastotalradiiofcurvature.The geometricandothercharacteristicsoftheindividudtestsectionsaregiveninTable1ofAppendix2. 4.ANALYSISANDDISCUSSIONOFRESULTS 4.1Gravelloss 4.1.1Gravellossfromunpavedroads.Thelossofgravelfromthewearingcoursewi~eventuallyleadto permanentdamageoftheroadstructureunlessremedialtreatmentiscarriedoutintime.Inthepast,muchofthe researchonmateriallosshasbeencarriedoutinthefieldofagriculture.WorkintheUnitedStatesofAmerica12 onsoilconservationhasfoundthatthemechanismofmateriallossfromagriculturallandisdefinedbythe foflowingequation:– Soilloss=SoilerodibilityxGroundslopefactorxRainfallfactor 4.1.2Soilerodibility.Thisisthecharacteristicofthesoilwhichexplainswhyerosiontakesplaceatdifferent rates,irrespectiveofusage,whenotherfactorsareconstant.Theerodibifityofasodistoalargeextentcontrolled byitsphysicalproperties.Themostimportantpropertiesaretheparticlesizedistributions,stabflityofthe grading,plasticityvrdues,organiccontent,permeabilityofthestructure,mineralcontentandshape,andthe initialcompactionofthematerial. Theabilityofaparticularsoiltoresisterosionissometimesdefinedintermsofitserodibilityfactorwhich 13Therateoferosiondoesnotremainconstantbut expressestheestimatedsoillossinmetrictonnesperhectare. 6 varieswiththechangeofphysicalstateofthesoflmass. 4.1.3Groundslope.Inhighwayterms,theeffectofgroundslopeonthesod’sabititytoresisterosionislargely governedbytheroaddrainagesystemand,inparticular,thelongitudinalandtransversealignment.Erosionis frequentlymanifestintheformoflongitudinalgulhesalongthesurfaceofsteeproadswithgradientshigherthan aboutfivepercentandthisisespeciallythecaseinhighrainfallareas.ThisproblemisfllustratedinHate3which showstheformationofsuchgufliesonanunpavedroadinKenya.Onemethodfordedingwiththisproblem whichisbeingstudiedbyTRRListheconstructionofhorizontaltransversedrainagesystemstoremovethe rainfalltothesidedrainsasshowninHate4. Thesedrainagesystemswereinstalledin1981onroadsinamountainousareaofKenyawithanaverage annualrainfallofover2000mm.Theroadsusedinthispdotstudypossessedverticalgradientsrangingfrom6to 9percent.Theperformanceofthisdrainagesystemwfllbemonitoredoverafurtherperiodof1to2years. 4.1.4Rainfallfactor.Therainfallfactorisavalueusedtodescribethecapacityoflocalisedrainfa~toerode soflfromanunprotectedface.ThesefactorshavebeendevelopedmainlyfromstudiesconductedbytheUS AgricultureServiceforperiodsofover40years.Theresultsofthesestudiesindicatethatwhenotherfactorsare constant,stormsoillossesaredirectlyproportionaltotheproductoftworainstormcharacteristics. (a)totalkineticenergyofthestorm. (b)itsmtimum30minuteintensity. Theparametersofsoilerodibihty,groundslopeandrainfallintensityclearlyaffecttherateofsoilloss, irrespectiveofwhetheritisonagriculturalland,oronunpavedroads. 4.1.5Parametn.cpredictionofpavelloss.Forroadsthesoillossequationmustdsoincludetheeffectoftraffic onthewearingcoursematerialastheinteractionbetweentrafficandrainfallcontributessignificantlytotheloss ofmaterialfromagravelsurfacedroad.Thestudydsokcludedthemaintenanceinputasaparameteraffecting therateofgravelloss.Thevariablesinvestigatedinthestudywerethereforeasfollows:– (a)trafficvolumesandloading (b)rainfallvolumeandintensity (c)roadti~ent (d)graveltype (e)maintenanceinput Thegravellossdatafromthestudywasfirstplottedasafunctionofthecumulativetrafficvolumeforeach testsection.Thedatawhichwastakenfrommeasurementsatapproximatelythreemontiyintervalsshowedno significantdifferencesduringwetordryseasons.Aselectionofresultsfromeachofthefourgraveltypes investigatedisshowninFigures5and6. 7 4.1.6&mparisonofresultswiththosefromanearlierstudy.Previousstudiesofgravellossonunpavedroads bytheTRRL4hadproducedanequationforpredictingtheannualgravellossforlateritic,quartzitic,volcmic andcordgravelsasfollows:– () TA2 ‘LA=f TA2+~~(4.2to.092TA t3.50R~tl.88VC).....eqn4.1 whereGLAistheannualgravellossinmiflimetres TAistheannualtrafficinbothdirectionsmeasuredinthousandsofvehicles RListheannualrainfallmeasuredinmetres Vc isthepercentagegradient f isaconstantwhichfor; lateriticgravels=0.94 quartziticgravels=1.1 volcanicgravels=0.7 coralgravels=1.5 Usingtherelevantvalues”ofannualtraffic,annualrainfall,verticalgradientandgravelconstantsinthis equation,thepredictedannuallossofgravelwasevaluatedforeachtestsectioninthecurrentstudy.Thesevalues wereplottedagainsttheactualvaluesobtainedinthestudyandillustratedinFigure7.Itcanbeseenfromthis Figurethattheannuallossofgravelmeasuredinthisstudyisgreaterthanthepredictedannualgravellossby about37.5percent.Tobringthepredictedgravellossintoagreementwiththeactuallossitwouldbenecessary toincreasethematerialconstantsfintheequationasfollows:– lateriticgravels 0.94to1.29 quartziticgravelsl.loto1.51 volcanicgravels 0.70to0.96 coralgravels1.50(notinvestigatedinthecurrentstudy) sandstonegravels1.38 Thedifferencesbetweenthegravellosspredictedbytheoriginalequationsandthegravellossmeasuredin thisstudycanbeattributedtothefollowing:– (a)Intheoriginalanalysis,theactionofrainfallaloneinproducinggravellosswascalculatedtheoretically usingtheresultsofWischmeieretal14.Anestimatedvalueofsotierodibilitywasusedwhichgavea predictedmaximumannualgravellossof2mmanditwasthenassumedthatwhenthetrafficvolumewas zero,thegravellosswouldbene#i@ble. Inthecurrentstudy,twoexperimentallengthsofunpavedroadwereconstructedwhichwerenot subjectedtotraffic.Thesepilotexperimentsweredesignedtostudytheeffectofclimateonthe deteriorationofunpavedroadsinisolationfromotherparameters.Oneofthevariablesmeasuredwasthat ofgravellossoveragridofroadarea60metresby7metres.meresultsofthesemeasurementsindicated (b) (c) (d)valuesofannualgravellossof4.3and7.5mmrespectively.Thesefiguresimplythattheerodingeffectof rainf~doneishigherthanthatcalculatedbyeitherWischmeierorbyusingtheoriginalTRRLanalysis.On exposedroads,theeffectofwinderosion,particularlyonsurfacesthatareslightlydisturbedbytrafficor rainfallcoulddsobesignificant.Theuntraffickedtestsectionsweresubjectedtoapprotiately0.8metre ofrainftiinthetwelvemonthperiod.Thisbyitselfcouldaccountforhdfthedifferencebetweengravel losspredictedfromtheoriginalequationsandthegravellossmeasuredinthisstudy. Intheoriginrdstudy,thethicknessofthewearingcoursesofthegraveltestsectionsvariedfrom28mmto 223mm,whflst,fortherecentstudy,thevariationswerecontainedwithinabandof121–165mm. Althoughgravellosscanbeseenasalossoftheuppersurfacinglayer,anyinherentweaknessesorstrengths duetothevaryingdepthofwearingcoursemustinfluencetheresistanceoftheroadtodeformation. Whilstnodirectcomparisonshavebeenmadeofspeedsoveraperiodoftimeforthesameroad,alignments ofunpavedroadsinKenyahave,ingeneral,greatlyimprovedinthelast7–1Oyears.Inaddition,the conditionofvehiclesonunpavedroadsindeveloptigcountrieshasalsoimprovedinrecentyears.Thus,not onlyaretheroadscapableofcarryingtrafficathigherspeedsbecauseofimprovedalignments,butvehicles nowalsohavetheabflitytotravelfaster.Theseincreasedspeedswfllhaveincreasedtheratesofgravelloss. Acontributoryfactortotheincreasedlossofgravelisthebehaviourdchangesinroadusersassociatedwith theintroductionofnewvehicles.OnruralroadsinKenya,thelastfiveyearshaveseenrapidincreasesinthe numbersof‘matatus’.Thesearelightgoodsvehicleswithanominalloadofl–11Atonnes,convertedto carryupto40passengers.Onsomeruralroads,thistypeofvehiclecanaccountfor30to50percentofthe totaltraffic.Thesevehicles,invariablyoverloaded,areengagedinahigtiycompetitivebusinesswhich producesveryaggressiveroaduserbehaviour.Theproductofoverloading,speedingandirregularmanoeuvres resultsinashearingmovementwhichdamagestheroadsurface.Inthestudy,itwasnotpossibletoisolate thesevehiclesfromothertraffic,sothereforequantificationofthedamagecausedwasnotpossible. 4.1.7Resultsfromothercountries. Anindicationoftherangeofgravellossesthatcanoccurinpracticeis illustratedinFigure8wheretheresultsofstudiesinanumberofdifferentcountriesareshown15,16,17,18me datafromtheKenyanresearch,bothin1971–74and1979–81,refertotestsectionswith3percentvertical gradientand1250mmofannualrainfall.RoadswithsteepergradientsinwetterareasWUhavehigherlosses, wtisttheconversewfllbetrueofroadswithflatter/lowergradientsindrierareas. Figure8furtherdlustratesthepointthatannualgravellossonunpavedroadsWNvarybetween10mmand 30mmper100vehiclesperdayandwfllbedependentonclimateandroadalignment.Thismeansthat,annu~y, 70to210cubicmetresofgravelwdlbelostfromeachkilometreofroadper100vehiclesperday. 4.1.8Regravellingrequirements.Theseratesofgravellossprobablyonlyholdforthefirstphaseofthe deteriorationcyclelastingpowiblyfortwoorthreeyears.Theyshouldnotbeconsideredtoholdoveralong periodoftime.Asthewearingcourseisreducedinthickness,otherdevelopmentssuchastheformationofruts wfllaffectthelossofgravelmaterial.Howevertheratesoflossgivenabovecanbeusedasanaidtotheplanning forregravellinginthefuture. 4.2Surfacelooseness Previousresearchhasshownthatloosematerialongravelroadsincreasesfuelconsumptionforawide spectrumofvehicles3. 9 Therearetwoprincipalreasonsforthepresenceofloosematerialongravelroads.Firstly,itoccursasa directresultofattritionoftheroadsurfacebytheactionoftrafficandrainfti.Ifthereissufficientmoisturein thematerialitWMbecompactedbytraffic,butodyinthewheelpaths.Ifthereisinsufficientmoisture,thenthe loosedrymaterialwdlbedispersedacrosstheroadbytrafficandwind.Itcanalsobeduetothemaintenance techniqueofgradingmateriallosttotheditchesandshouldersbackontotheroad. Mostofthetestsectionsinthisstudywerecompactedaspartofthemaintenanceinput.Thedatarelative totheloosenessmeasurementsdiscussedinthischapterarethereforederivedfromthenilmaintenancesections andthetestsectionswherecompactionwasnotcarriedout. Theresultsoftheloosenessmeasurementsforeachgraveltypehavebeenplottedasafunctionofcumulative trafficandareMustratedinFigures9and10. Althoughtherewasamaximumof10mmofloosematerialimmediatelyaftergrading,thedepthreduced rapidlywithtraffictoaconstantlevel.Inthecaseofthequartziticgravel,thislevelwasreachedafterapproxtiately 5000trafficpasseswhilsttheothergravelstookapproxtiately6000–10000trafficpassestoreachasimilar asymptote.Routinemeasurementsonthecompactedsectionsgavevahresoflessthan1mminaflcases. Therelationshipsbetweendepthofloosematerialagainstcumulativetrafficwerefoundtobethefoflowing: Lateriticgravel DLM=6.Olle–‘.334Tt0.5 .....4.2 Quartziticgravel DLM=5.895e–‘.358T+1.0 Volcanicgravel DLM=6.748e–‘.183Tt1.0 Sandstonegravel DLM=6.925e–‘.187Tt10 whereDLM=depthofloosematerialinmm T=cumulativetrafficpassesinbothdirectionsinthousandsof vehiclessincegrading. 4.3Surfaceroughness.....4.3 .....4.4 .....4.5 4.3.1Surfaceroughnesschangesforeachgravel~peinvestigated.Theprincipalindexofunpavedroad deteriorationwhichaffectsvehicleoperatingcostsisroughness(surfaceirregularity).Inturn,trafficisthepara- meterthathasthemostsignificanteffectontherateofchangeofroughness.Intheanalysis,theroughnessof thesectionswasplottedasafunctionoftraffic.Toobtainthebestfitofcurveforthemeasurementsthedata wasanalysedusingpolynomialregressiontechniques.Insomecases,ascanbeseenfromFigures11and12,the fitwasonlymarginallybetterthanthatobtainedusinglinearregression. 10 Figures11and12refertotheexperimentalsectionsonwhichnomaintenancewascarriedoutfortheperiod ofthestudy.TheywereconstruttedinthethreeprincipalresearchareasofMefi,BungomaandKaloleniand consistedofwearingcoursesmadewithlateritic,quartzitic,volcanicandsandstonegravels.Ntheexperimental sectionswerebudtusingthesametechiqueandstandardofconstruction.Eachsectionhadadesignatedwearing courseof150mmcorrespondingtotheMOTCspecification.Inpractice,thetticknesswas’foundtobeinthe rangeof121–165mm. Inthemajorityofcases,thecorrelationbetweenroughnessandtrafficforeachgroupoftestsectionsisgood withregressioncoefficientshigherthan0.93.However,whentheresultsfromthetidividudsections,builtwith thesamegraveltype,areplottedtogetherthereareapparentdifferencesinthedeteriorationcycles,particularlyin thecasesofthequartziticandsandstonegravels.Themaindifferenceinperformanceofthequartziticgravels occursatatrafficlevelof15,000vehicleswheretherearetwosignificantincreasesinsurfaceroughnessonthe lowertraffickedroads.Thesetwohighreadingstookplaceaftertheheaviestandmostintensiver~nstormrecorded inthestudyhadfaflen.TMsparticularstormhadanintensityof40m~ouroveraperiodof1%hoursandwas probablyresponsibleforthesuddendeteriorationofthisroad. Thetwosandstonegravelscamefromborrowpits30kilometresapart,andalthoughofthesamegeological origin,showedmarkeddifferencesintheirmaterialproperties.Oneofthegravelscontainedlargequantitiesof micaandfeldspar,whilsttheotherpossessedamuchcoarserparticlesizedistribution.Thelatterformedthe wearingcourseontheodyroadmonitoredintiestudythatwasbuiltbyprivatecontractor.Initiafly,the constructionofthisroadappearedcorrectintermsofthegravelwearingcoursettickness,particlesizedistribution andcompaction.However,thesubsequentdeteriorationoftheroadshowedtheexistenceofpotholeswhichhad resultedfromthepresenceofoversizematerialinthewearingcourse.Thisoversizematerialcomprisedsandstone fragmentswhichhadbeensptitbyweatheringprocessesintothinflatslabs.Wheresuchmaterialoccurred,the compactioncouldnothavebeeneffectiveandtheareawouldhavebecomea‘softspot’withtheoversizematerial beingexposedandeventu~yremovedbytheactionoftraffic. Additionalanalysisofthephysicalpropertiesofthesetwogravelswascarriedouttoseeiftherewasany correlationbetweensurfaceroughnessandparticlesizedistribution,butnonewasfound. Thelateriticgravelsdeterioratedataslowerratethanalltheothergravelsinvestigatedinthestudy.After 97,000trafficpassesaroughnessof8000mm/kmwasreachedwhichwmmuchlowerthantheroughnessIevek obtainedwithothergravelsatlowertrafficvolumes.Thegravelusedforthewearingcoursescamefromtwo largequarriesandthemateriakwereverysimilarintermsoftheirphysicalproperties.Thedifferencesbetween thevariouslateriticgravekintheirratesofdeteriorationwerethoughtiniti~ytobeduetothediffetiggeometry butnocorrelationwasfoundbetweenroughnessandeitherverticalgradientorhorizontalcurvature. Volcanicgravekdeterioratedatafasterratethananyoftheothergravels,reachingaroughnesslevelof 8000mm/kmafteronly30,000trafficpasses.Thethreeroadsbufltwithvolcanicgravelwearingcourseswere constructedtothreedistinctlevelsofgeometry,butagainnocorrelationwasfoundbetweentherou&essof thesectionsandtheirparticulargeometric.standards. 4.3.2ComparisonwithpreviomTMLstidy.InthepreviousstudyconductedbyTRRLin1971–1974, lateritic,volcanicandquartziticgravelsweregroupedtogetherandequatedtoageneraldeteriorationcurveof roughnessasafunctionoftraffic(seeFigure13).Thiswasduetoanabsenceofnflmaintenancesectionson volcanicgravelroadsandnoapparentdifferenceinratesofdeteriorationbetweenlateriticandquartziticgravels. 11 Inthecurrentstudy,nilmaintainedsectionswereestablishedonvolcanicgravelroadsanddistinctrelationships werefoundforaflfourgravelsinvestigated.Athoughthequartziticgravelswerefoundtoperformbetterthanthe volcanicorsandstonegravels,theysti~deterioratedmorequicklythanthelateriticgravelledroads. Inthedeteriorationcycles,thereissomeevidenceintheequations,particularlyinthevolcanicandlateritic sections,thattherateofchangeofroughnesstendstodecreaseatthehighlevels.Itwasobservedduringthestudy thattherewasachangeofbehaviourbyroadusersatthehigherlevelsofroughness,particularlyonthevolcanic grave~edroadswhichhadonaveragethehighestlevelofroughnessrecorded.Generallyvehiclestendedtofollow aregularwelldefinedlineoftravelalongaroad,thepositionsofwhichvariedaccordingtothenumbersof vehiclesusingtheroadandthegeometryoftheroad.Forexampleonroadscarryinglessthan100veticlesper day,mostofthetrafficstraddledthecentreoftheroadformingtwoconcentratedwheelpaths.Onroadswith morethan100vehiclesperday,vehiclestendedtousetheircorrectlanemoreoften,andusuallyfourseparate wheelpathsdeveloped.Thepatternofbehaviourappearedalsotobeconstrainedbythehorizontalandvertical alignmentoftheroad,especirdlyonroadswithradiiofcurvaturegreaterthanninetydegreesperkm.Onmore heavilytraffickedroads,thedefinedwheelpathsbecamemoreestablishedanddeterioratedtoapointwheremuch ofthetraffic,particularlycarsandlightgoodsvehiclesattemptedtotravelonsmootherandrelativelyuntrafficked partsoftheroad. Thusifthemeasurementsofroughnessareonlytakeninthewelldefinedwheelpaths,thelevelswould apparentlyremainthesamebecausemostofthetraffichaddevelopednewwheelpathselsewhereontheroad. Theexistingwheelpathsthenodydeterioratefromtheeffectofrainfallandoccasionaltrafficking.Duringthe study,measurementsofroughnesswerealsotakeninthemiddleoftheroadand,indlcasesexceptonsome oftheroadsonsteepalignments,theroughnesslevelswerethelowestherethananywhereelseontheroad.The towedbumpintegratorcanbeusedtomeasuretheroughnessofindividualwheeltracks,butitisnotpracticalto carryoutsuchamonitoringprogrammewiththisinstrumentoveralargenetworkofroads. InthestudybytheTRRL3’4in1971–1974,therelationshipbetweenroughnessandtrafficwasageneral onecoveringlateritic,volcanicandquartziticgravels,plusaseparaterelationshipforcoralgravelsbasedonlimited data.Coralgravelswerenotincludedinthepresentstudy. ThegeneralrelationfiipderivedfromtheoriginalTRRLstudyindicatedalargechangeinroughnessfor thesegravelsafterapproxtiately80,000trafficpasses(seeFigure13).Itcanbeseenthatbetween80,000and 100,000trafficpasses,theroughness,infact,increasesbyapproxtiately50percent.Thisisbecauseofthe influenceofthehighestrecordedroughnessinthatstudy.Thissingemeasurementof14,500mm/kmonone ofthelateriticgravelsectionshasheavflybiasedtheshapeofthecurve,astheprevioushighestroughnesslevel recordedwasonly9,000mm/km.Morerecentinvestigationofthelateriticsectioninquestionhasfoundthat theoriginalwearingcoursewasonly65mmthickascomparedtotheMOTCspecifiedminimumthicknessof 150mm.Outofthe42sectionsinthatstudy,37sectionshadwearingcoursessubstantiallythickerthan65mm. Ifthesin~emeasurementofroughnessof14,500mm/kmhadbeenomittedfromtheanalysis,thechangein roughnessbetween80,000and100,000trafficpasseswouldhavebeenreducedtoanincrementofapproximately 15percentratherthan50percent.Afteratraffickingof100,000vehicles,theroughnesswouldthenbe approximately9,000mm/kminsteadof12,000mm/km.Thiscompareswitharoughnesslevelofapproximately 7,500mm/kmforthelateriticgravelintherecentstudyafter100,000trafficpasses. TheregressionequationsrelatingroughnesstotrafficvolumeforeachofthegravelsillustratedinFigures 11and12are:– 12 hteriticgravels R=3008+22.300T–0.531T2–0.0033T3 QuartziticgravelsR=3384t52.287TtO.143T2t0.0157T3 VolcanicgravelsR=3442–2.519T+8.111T2–0.0934T3 IR=3505+221T–6.360T2+0.0736T3 Sandstonegravels R=3545–39.lT+5.060T2t0.149T3 whereR=meanroughnessinthewheeltracksmeasuredinmm/km byatowedbumpintegratortowedat32km/h T~Trafficpassingthesectionsinbothdirectionssincegrading andmeasuredinthousandsofvehicles. 4.4Rutdepth 4.4.1Developmentofrutswithcumulativetraffic.Intheanalysisofrutdepth,themeasurementswereplotted asafunctionoftrafficusingpolynomialtechniquesandareillustratedinFigures14and15. OntheIateriticgravelsections,thedevelopmentofrutswaslesspronouncedthanonsectionsusingother graveltypesandtherewasevidenceofareductioninslopeor‘flattening’ofthecurvesimflartothatobtainedwith theroughnessmeasurements.Thevolcanicgravelsectionsdidnothavehighvaluesofrutdepthconsideringtheir highlevelsofroughness.Thisisattributedtothefactthat,ontheseroads,itwasobservedthatextrawheelpaths werebeingdevelopedasdiscussedearlier. Theregressionequationswhichhavebeenfittedtothedataobtainedfromtherutdepthmeasurementsare .....eqn4.6 ..... eqn4.7 .....eqn4.8 ..... eqn4.9 ..... eqn4.10 asfo~ows:— hteriticgravelsRD=7.18–0.081Tt0.0069T2–0.000036T3 QuartziticgravelsRD=7.49t0.17lT+0.014T2–0.00009T3 VolcanicgravelsRD=10.11+0.314Tt0.00031T2t0.00002T3 SandstonegravelsRD=7.09+0.573T–0.0128T2+0.00024T3 whereRD=Rutdepthinmmundera2metrestraightedge T=Trafficpassingthesectionsinbothdirectionssince gradingandmeasuredinthousandsofvehicles. 4.5Journeytimes.....eqn4.11 .....eqn4.12 .....eqn4.13 .....eqn4.14 4.5.1Speedmeasurementsoverthetestsectionsatthebeginningandendofthestudy.Measurementsof journeytimesweretakenat6monthsintervalsforthenormaltraffictravelingoverthetestsection.Theperiod ofmeasurementwasforoneday,from8.00amto6.00pmongroupsoftestsections.Ontheroadswhichcarried lessthan100vehiclesperday,themeasurementswererepeatedthefollowingday.Typicalresultsofmeasurements atthebeginningandendofthestudyareillustratedinTable1ofAppendix3. 13 ItcanbeseenfromthisTablethatchangesinmeanvehiclespeedovertheperiodofthestudyweresmd. Inmostcases,therewasaslightreductioninspeedinresponsetothedeteriorationoftheroad.Ononeroadin Meruwhichhadthehighestlevelofroughnessinthestudy,thereductionsinspeedwerelarge.Elsewhere,the differenceswererarelymorethan2kilometresperhourand,insomecases,therewasasma~increaseinspeed. Samplesizesofcars,heavygoodsvehiclesandbusesweresmallbecauseofthelowtotalnumberofvehiclesusing theroadbut,inthecaseofthelightgoodsvehiclesandmediumgoodsvehicles,thesamplesizewasnormallyin therangeof50to100vehicles.Thesmdnumberofcars,busesandheavyvehiclesWUleadtoahighdegreeof variabilityinanymeasurementsofspeed,althoughthepreviousTRRLstudy3foundthattheroadconditionhad onlyasmalleffectonbusspeeds.Onanyparticularroad,vehiclesineachclasstravelledatasimilarspeedwith usuallyasmallnumberofvehiclesinthatclasstravelingeitheratveryfastspeedsoratexcessivelyslowspeeds. 4.5.2Comparisonswithpreviousstudiesofvehicle speedsonunpavedroads.The previouswork byTRRL relatedchangesinvehiclespeedtochangesinroadconditionintermsofroughnessandrutdepth.Itwasfound thatforroughnesschangesof5000mm/km,thespeedsofcars,lightgoodsvehicles,mediumandheavygoods vehicles,andbuseswerereducedby4.4,4.8,3.0and1.8km/hrespectively.Ruttingoftheroadsurfaceproduced simflarresults,forexamplerutdepthsof20mmreducedthespeedsofcarsandheavygoodsvehiclesby3.7and 5.3km/hrespectively.Changesinroughnessoftheorderof5000mm/kmandrutdepthsof20mmwereonly foundonthenilmaintenancesectionsinthisrecentstudy.WiththeexceptionoftheroadinMeruwiththe highestroughnesslevel,nolargechangesinvehiclespeedwereobservedonthetestsections,includingthenil maintenancesectionsastheroadsdeteriorated.Thenflmaintenancesectionswerelocatedneartestsections whichhadreceivedmaintenanceanditisthoughtthatthismayhaveinfluencedroaduserbehaviour.Theatie loadsurveysshowedthatupto80percentofthetrafficontheseroadswereregularusers.Thissignifiedthat manyofthedriverswerefarnfliarwiththedifferentfeaturesoftheroadandthismayhavehadconsiderable influenceonthevehiclespeedrelationshipirrespectiveoftheroadcondition. 4.6Trafficloading 4.6.1Axleloadsurveymeasurements. Axleloadsurveyswerecarriedoutatapproximately4monthsintervals oneachgroupofroadscontainedwithinthestudy.Thesurveyslastedfor5daysfrom6.00amuntd6.00pmand foronenightfrom6.00pmuntfl6.00am.Theprecisetimingofthesurveysduringtheyearwererelatedasfaras possibletoexpectedchangesinthetypeofgoodsbeingtransported,suchaswouldresultfromtheharvestingof crops.ThemaingoodstransportedintheareasofMeruandBungomawerevegetables,fruit,tea,coffee,sugarand sunflowerseedswhflst,intheKdoleniarea,itwassisal,fruitandmfr.Ahighproportionofthetrafficflowwas composedofpassengercarrying‘matatus’asmentionedinsection4.1.IntheMeruandBungomaareas,matatu trafficconstituted30to50percentofthetotalflowontheexperimentalroads.”Thebulkoftheremaining trafficwasmediumgoodsvehiclescarryinglocalproduce.OntheroadsintheKaloleniareawhichwaslessdensely populated,thenumberofmatatuswassubstanti~yless,usuallyoftheorderof15to20percentofthetotal flow.Onalltheexperimentalroadstherewereordysmallnumbersofcars,heavygoodsvehiclesandbuses. TheresultsofthesurveysareillustratedinFigures16and17whichshowtypicalaxleloaddistributions forthedifferentexperimentalroads.Itcanbeseenthatthenumberofaxlesexceedingthelegrdlimit,whichin Kenyais8tonnesmaximumonasingetie,isminimalandinmostcaseslessthan2percent. Initially,itwasplannedthattheanalysisofroadconditionintermsofroughnessandrutdepthwouldbe relatedtocumulativeatieloadingasweflastrafficvolumes.Theanalysisshowedhoweverthattherewasno significantdifferenceiftheroadconditionwasplottedasafunctionoftrafficvolumeorasafunctionoftie loading.Subsequently,rdldatarelatingtoroadconditionhasbeenpresentedinthisreportasafunctionof trafficvolume. 14 4.6.2Heavilyloadedsitevehiclesusedinroadconstruction.Acommonfeatureofnewunpavedroadsisthat thesiteconstructiontrafficisinvariablyheavierthanthetraffictheroadwflleventuallycarry.Ontheroadsbuflt undertheaegisoftheMOTC/USAID/CIDArehabilitationprogramme,thevehiclestransportinggravelwereUSUNY theInternationalTypeC–1954lorriescapableofcarrying8to10tonnesofmaterial.Theloadingofthesethree- afledvehicleswassttictlycontrolledandreartiesrarelyexceeded7.2tonnes.Thecentrolbythevarious organisationswasaimedatprotectingthesuspensionofthevehicleunderdifficultfieldconditionsratherthan complyingwiththelegalrequirements.VehiclesofthistypeWWexertsomeadditiondcompactiontotheroads whentheyarenewbutthemajoradvantageisthattheylocatesoftspotsandgeneralweakareasinthenew construction.Thismeansthatremedidtreatmentcanbeappliedimmediatelyatatimewhenitismosteffective andwfllprevent,oratleastpostpone,futureproblems. Thedifferencesinatieloadingbetweentrafficwhichincludessiteconstructionvehiclesandthenorrnd trafficareNustratedinFigure17.Here,resultsfromtwoade-loadsurveysontheBungomatoBokoliroadare shown.Thefirstsurveywascarriedoutsoonaftercompletionoftherebuildingofthisroadwhentheloadedsite vehicleswerestalltraveningovertheroadtoreachtheadjacentconstructionsite.Thesecondsurveywascarried outtwomonthslaterwhentherewasnositeconstructiontraffic.Themostnotabledifferenceintieloadingwas intherangefrom6to10tonneswhichcoveredtherangenorrndyfoundforthesecondandthirdaxleofthe heavylorrieshautinggravel.Thefirstsurveyshowedthattheroadwascarrying2300tonneseachdaywhdstthe secondsurveywiththenormaltrafficflowshowedthatthedaflyloadingwasonly900tonnes. 4.7~imate 4.7.1~angesinroughnessrelatedtocumulativerainfallforuntraffickedroads.Apartfromtraffic,other parameterswillaffectthedeteriorationrateofthegravelwearingcourses.Theprincipaloneoftheseisrainfall. Pdotexperimentsweresetupintwoareastotrytoestablishtheeffectofrainfdontheroadsurfacein isolationfromotherparameters.Thiswasachievedbyconstructingshortlengthsofroadadjacenttothemain testsectionsinvestigatedinthestudy.Theselengthsofroadwerebuflttothesamestandardofconstruction asthemainroadandwereleftuntraffickedexceptwhenmeasurementswerebeingtakenwiththetowedbump integrator.Toensurethatnotrafficwouldusetheroadaccidentdly,thewholeareawasfencedoffandaccess preventedbytheconstructionofditches100metresfromeachendoftheactualtestsectionoftheroad.men roughnessmeasurementswerebeingtaken,culvertpipeswereplacedintheseditchesandovertoppedwithgravel overawidthof4metres.Afterthemeasurementshadbeentaken,thegravelwasremoved,stockpiledandthe culvertsstoredinanearbyschool. Thetwosectionswerebufltwithquartziticandsandstonegravelwearingcourses.Ithadbeenintendedto duplicatethesesectionsinotherareasusinglateriticandvolcanicgravelwearingcoursesbutthiswasnotpossible duetoproblemsoflandacquisition. Selectionofthepreciselocationofsitesforthesetwoexperimentalroadswasgovernedbytheneedtolook atdifferentlevelsofrainfall.AnnurdrainfallfiguresderivedfromtablessupphedbytheKenyaMeteorological Department19giveaveragevahresof1250mmand630mmovera5yearperiodfortheareasinwhichthesetwo roadswereconstructed.Theactualrainfallrecordedatthesitesduringatwelvemonthperiodwas711and847mm respectively,whichwassomewhatdifferentfromtheaveragefigures. OnthequartziticgravelledroadnearMeru,therewasanannualincreaseofroughnessof787mm/kmfora cumulativerainfdof711mm.Thisindicatesaratioofannualroughnessincreasetorainfallof1108mm/kmper 15 metreofrainfrdl.IntheKdoleniarea,theroughnessincreasedinthesameperiodby821mm/kmforatigher rainfallof847mm.Thisgivesaratioof970mm/kmpermetreofrainfall.Theincreaseinroughnessatthetwo testsitesplottedasafunctionofcumulativerainfallisillustratedinFigure18. Theseresultsmeanthat,“onuntraffickedunpavedroadsinareaswheretheannurdrainfallisapproxtiately 1metre,theroughnessoftheroadwti-irrcreaseannuallybyapproximately970–1100mm/km,regardlessof traffic.RainfaHrecordsmaintainedonthetwositesshowedthattherainftiintensitieswerelow,withthehighest stormsrecordedduringthemonitoringperiodof10mm/h.Higherrainfallintensitieswouldhaveinfluencedthe changeinroughnessconsiderably.Therefore,inareassubjectedtoflashstorms,ienearmountains,coastaland lakeareas,theannualincreasesinsurfaceroughnessofunpavedroadsduetorainfallalonecouldbemuchgreater. Theequationsderivedfromtheregressionanalysisofthedataareasfollows:– Quartziticgravel R1=3303t1.117RR RegressioncoefficientR*=0.83337 Sandstonegravel RI=3550+1.021RR RegressioncoefficientR*=0.90966.....eqn4.16 .....eqn4.17 whereRI=routinessinmm/kmmeasuredbytowedbumpintegrator RR=rainfallinmm. Forthepurposeofpredictingroughnessintermsofrainfall,itisprobablyadequatetocombinethese equationstogive:– RI=3429t1.063RR .....eqn4.18 RegressioncoefficientR*=0.78199 4.7.2~angesinroughnessrelatedtocumulativerainfallfortraffickedroads.Anattemptwasmadetoseparate theeffectoftrafficfromrainfallontheexpertientalsectionsbyplottingroughnessasafunctionofcumulative rainfall.ThisisillustratedinFigure19wherethedataisplottedforconstanttrafficvolumesof8000vehiclessince grading.Ascanbeseen,thedataisscatteredand,usingthisformofanalysis,theredoesnotappeartobeadirect relationshipbetweenroughnessandtraffic.Otherlevelsoftrafficvolumeswereusedbutnosignificantrelation- shipswereapparentinthesecaseseither.Oneexplanationisthatonecannotisolatetrafficinthiswayonunpaved roadsassmallnumbersofoverloadedvehiclescancauseseveredamagetotheroadiftheytravelduringperiods whentheroadiswet. Theresultsofthemeasurementsontheuntraffickedsectionshaveyieldedmuchneededinformationonthe deteriorationofunpavedroadsduetorainfallalone.Inparticular,therelationshipsfoundfortheuntrafficked sectionsareprobablyvalidforlowvolumeroadscarryinglessthanabout10vehiclesperday. 16 Inordertoprovidethedatatocalibrateabetterrelationshipbetweentrafficandrainfallitwouldbe necessarytomonitortraffic,rainf~androadcontitiononadadybasis.Thedeterioratingeffectoftrafficcould thenbeweightedaccordingtotherainfdatthetimeoftrafficking.Thiswasnotpossiblewiththeresources avadableforthisstudy. 4.8Geometry 4.8.1Geometricconstraintsondeterioration.Itwasnotfoundpossibleintheanrdysistoisolatetheeffectof horizontalcurvatureandverticalgradientsonthedeteriorationofeithermaintainedornonmaintainedroads. Horizontalcurvature,withintherangemeasured,appearedtohavenoeffectonratesofdeterioration. Inititiy,ithadbeenincludedintheanalysisofgravelloss,butwaseventuallyrejectedasitmadenosignificant differencetotheresults.Themostlikelyinfluenceofthisparameterintermsofthedeteriorationofunpaved roads,wouldbethroughitseffectonvehiclespeeds.Reductionsinspeeddependmoreonthefrequencyandrate ofchangeofcurvatureratherthanthetotalamountofhorizontalcurvature. Theeffectoflongitudinalgradientondeteriorationwasmorenoticeableinthelatterstagesofthestudy withthedevelopmentofsmtigulliesalongthecentreofsomeofthetestsections.Inparticular,thevolcanic sectionsonsteeprdignrnentnearMeruwerebadlyaffected.Thismodeofdeteriorationwasnotrelatedtoanyof themaintenanceinputs,butoccurredastheresultofinadequaciesinthedesignoftheroad.Amajorproblemof unpavedroadsbufltonsteepalignmentsistheefficientremovalofsurfacewatertothesidedrains.Asthegradients increase,theproblembecomesmoreacuteirrespectiveofanyincreaseinthecrossfdloftheroad.Althesteep volcticsectionswereconstructedwithcrossfdlsof4percentwhichwouldhavedelayedtheformationofguflies but,afteraperiodofeighteenmonths,thesevolcanicsectionsrecordedthehighestlevelofgravellossandrutting ofanygroupofsectionsmonitoredinthestudy,resultinginasignificantreductionoftheinitialcrossfdl.Thenil maintenancesectiononthissteeprdigrrmenthadagravellossof38.7mmaftereighteenmonthscomparedtoa gravellossof22.2mmforanflmaintenancevolcanicsectiononflataligment.Therutdepthmeasuredinthe centreoftheroadonthesametwosectionshadvaluesof53mmforthesteepalignmentand21mmfortheflat alignment.Inthelattercase,therutsinthecentreoftheroadweretheresultoftheinteractionbetweenthe innerwheelsoftrafficandthegravelwearingcourseandnotasadirectresultoferosionbywater.Roadsonsteep alignmentswithcrossfdlslessthan4percentarecommoninmanydevelopingcountriesandoftendisplaythe deteriorationMustratedinPlate3. Theproblemofgullyerosionalongthecentreofunpavedroadswfllbeexacerbatedasverticalgradients increaseabovethevalueofthecrossf~.Therecanodybetwosolutionstothisproblem.Eitherthevertical geometrymustbephysicdyreducedbyrealigningthesectionofroad,ordrainagemeasures,suchasthose discussedinChapter4.1,mustbeintroduced. 4.9Roadcamber 4.9.1~angesinroadcamberduringthestudy.Attheendofthestudymostofthetestsectionshadretained aneffectivecamberofover3percent.Theonlyexceptionstothiswerethenflmaintenancesectionswherethe developmentofrutsandpronouncedwheelpathscausedthetransverseprofdestotakeupanirregularshape. Thetestsectionswhichhadinitialcambersof6percentappearedtobeinslightlybetterconditionthan their4percentcounterpartsectionswiththesamemaintenanceinput,butthedifferencewasnotsignificant.It ispossiblethat,ifthereweredifferences,theywouldassertthemselvesoveralongerperiodofmeasurementor 17 inamoreintensiverainfallarea.Thehighercamberedsectionswerenotfoundtohavecorrespondti~yhigher ratesofgravel10ss. 4.10Shouldercompaction Afrequentproblemonbothpavedandunpavedroadsisthedeformationoftheshoulderwhichoften precipitatesthestructuralfaihrreofthepavement.Inmanycases,thisistheresultofvehicles,particularlyheavy lorries,standingofftheroadduetobreakdownorovernightstopandsometimesasaresultofpassingvehicles strayingofftheedgeoftheroad.Itcanalsooccurasaresultofwaterleavingtheroadsurface,butstayingonthe shoulderbecauseofinsufficientcrossfdl.Anobjectiveofthestudywastoexaminewhetherimprovedperformance oftheroadcouldbeobtainedbyapplyingadditionalcompactiontoshouldersduringmaintenanceoperations. Duringthestudy,nosignificantdeteriorationwasfoundontheshouldersofanytestsection,whether additiondcompactionhadbeenapphedornot,implyingthattheadditiondcompactionwasumecessary. However,thenumbersofhea~vehiclesusingtheroadwerelowand,whentrafficdidstopbecauseofmechanicrd breakdown,itwasabletostopontheroaditselfbecauseofthelowtrafficintensity. 5.SUMMARYANDCONCLUSIONS TheunitrateofgravellossfoundinthestudywashigherthanthatpredictedbytheTRRLroadinvestmentmodel. Thetotrdlossofgravelfromunpavedroadsindevelopingcountriesisalsoincreasingannurdlybecauseofadditions totheroadnetwork.ThisproblemWMbecomeexacerbatedasroadnetworksexpandandthesourcesofgoodroad makinggravelcontinuetodwindle.Already,haulagedistancesof80kilometresforgravelexistinKenya,andin Africagenerallyhaulagedistancesformaterialarelengthening. Thestudyquantifiedseparatedeteriorationrelationshipsforlateritic,quartzitic,volcanicandsandstone gravels.Thiswasanimportantimprovementonpreviousstudiesofgravelroads.Therelationshipsaresummarised inAppendix1.Thesedeteriorationrelationshipswdlnowbecombinedwiththeresultsofresearchinother countriestoproducemoregeneralrelationshipsforinclusionintheTRRLroadinvestmentmodel. Theeffectofrainfa~inisolationfromotherparameterswasfoundtoincreasethesurfaceroughness annuallybybetween970and1100mm/kmpermetreofrainfallforsandstoneandquartziticgravels. Onwe~maintainedroadstheamountofloosematerialthatstayedonthesurfacewasnotsignificant,and isudikelytoinfluencevehicleoperatingcosts. Onlyasmallnumberofvehiclestravelingoverthetestsectionswerefoundtobeoverloaded.Overloaded vehiclesdidnotthereforesignificantlyinfluencetherateofdeterioration. 6.ACKNOWLEDGEMENTS TheworkdescribedinthisreportformspartoftheresearchprogrammeoftheOverseasUnit(UnitHead: MrJSYerrell)oftheTransportandRoadResearchbboratory.Theresearchwascarriedoutinco-operationwith theMinistryofTransportandCommunicationsoftheRepublicofKenya.Thesupportandencouragementshown byMrWPWmbura(ChiefEngineer,RoadsandAerodromes),andMrNBOnduto(ChiefMaterialsEngineer) andtheirstaffisgratefullyacknowledged. 18 7.REFEMNCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.PARSLEY,UNDALand~CHARDROBINSON.TheTRRLroadinvestmentmodelfordeveloping countries(RTM2):DepartmentoftheEnvironmentDepartmentofTransport,TRRLLaboratoryReport 1057.Crowthorne,1982(TransportandRoadResearchbboratory). JONES,TE.‘TheKenyamaintenancestudyonunpavedroads:optimummaintenancestra~es. &parmentoftheEnvironmentDepartmentofTransport,TMLLaboratoryReport1112. Crowthorne,1984@ransportandRoadResearchbboratory). DEH,SWABAYNAYAKA,IASAYERandRWYATT.TheKenyaRoadTransportCostStudy: researchonvehicleoperatingcosts.DepartmentoftheEnvironmentTMLReportLR672.Crowthorne, 1975(TransportandRoadResearchhboratory). HODGES,JW,JROLTandTEJONES.TheKenyaRoadTransportCostStudy:researchonroad deterioration.DepartmentoftheEnvironmentTRRLLaboratoryReport673.Crowthorne,1975 (TransportandRoadResearchhboratory). HARRA~CGetal.TheHighwayDesignandMaintenanceStandardsModel(HDM):modelstructure, empiricalfoundationsandapphcations.PTRCSummerAnnualMeeting,UniversityofWarwick,9–12 July1979. DEPARTMENTOFSCIENTIFICANDINDUSTWALRESEARCH.RoadResearchhborato~, Laborato~ResearchNoteRN/3328/WJOS.TheBumpIntegrator.(RoadResearchbboratory1958). JORDAN,PGandJCYO~G.DevelopmentsinthecalibrationanduseoftheBumpIntegrator. JournaloftheInstitutionofHighwayEngineers,April1981. POTOCKI,FP.Aportableweighingunitanddatarecorder.DepartmentoftheEnvironment,RRLReport LR391.Crowthorne,1971(RoadResearchbboratory). SNMTH,MS.Adigitalportableweighbridge.TheHighwayEngineer,Vol25,No7,July1978,pp9–12. TRANSPORTANDROADRESEARCHLABORATORY.Aguidetothemeasurementofatieloadsin developingcountriesusingaportableweighbridge.DepartmentoftheEnvironmentDepartmentof Transport,RoadNote40.1978(HMSO). BWTISHSTANDARDSINSTITUTION.MethodsofTestforsoilsforengineeringpurposes.British StandardsBS1377/75.hndon(BritishStandardsInstitution). WISC~IER,WHandJVMANNENNG.Relationofsoilpropertiestoitserodibfity.SoilScience SocietyofAmericaWashingtonDC,(1968). AGWCULTUWLRESEARCHSERVICE.Auniversalequationforpredictingrainfall-erosionlosses. ARSSpectilReportARS22–66.WashingtonDC,1961(UnitedStatesDepartmentofAgriculture). 19 14.WISCHMEIER,WH,CBJOHNSONandBVCROSS.Asoilerodibifitynomographforfarrrdandand constructionsites.Journalofsoilsandwaterconservation,1971,26,189–93. 15.ODIERL,RSMILLARD,PI~NTELdosSANTOSandRSMEHRA.bwcostroads.Design, constructionandmaintenance.London1971(Butterworthsfor~ESCO). 16.ROBINSON,Rhvestigationofweatheredbasaltgravel:performanceoftheGhion-Jirnmaroad experiment.EthiopianRoadAuthon.tyandTransportandRoadResearchLaboratory.JRRPReport21. AddisAbaba,1980(EthiopianRoadAuthority). 17.LOMBARD,P,JBML~N,MR~DERSONandMND~ELRoadmaintenance.Conferenceon HighwayMaintenanceinAfrica,AddisAbaba,1974.(~TEDNATIONSECONO~CCOM~SSION FORAFWCA). 18.ROBERTS,PWDH.TheperformanceofunsealedroadsinGhana.DepartmentoftheEnvironment DepartmentofTransportTRRLLaboratoryReport1093Crowthorne,1983(TransportandRoad Researchhboratory). 19.KENYAMETEORO~GICALDEPARTMENT.AnnualrainfaflinKenya1975–1979(Kenya MeteorologicalDepartment). 20 1 I 2 34567 8 Gradedatsix monthsintervals bymotorgrader, compactedby vibratingrollerGradedatsix monthsintervals bytowedgrader, compactedby vibratingrollerAssection1 butroadbuilt with6percent camberNilmaintenance Gradedatsix monthsintervals bvmotorgrader, notcompacted afterwardsGradedatsixGradedatnine monthsintervalsmonthsintervals bvtowedgrader,bymotorgrader, notcompactedcompactedbv afterwardsvibratingrollerGradedatsix monthsintervals bvmotorgrader, compactedbv vibratingroller afteradding water Allsectionswere300metresinlengthseparatedbvtransitionsofminimumlength25metres. Insomecasesthetransitionsweremuchgreaterbecauseofthegeometrvoftheroad. Fig.1Bungoma –MungatsiRoad,SiteB ,. ,, “..”., .. “.. .“ ,, ., .“,. ,,.,, # 1 TANZANIA Namanga ... .. ~/”....J<.. Ar;sha Moshi Fig.2RainfallmapofKenyashowinglocationofexperimentalroads Concrete Centralrodusedasbenchmark standing5mmproudofmonument / / k b 12mmsteelrods o . 3m — n Fig.3Diagrammaticlayoutofbenchmarkmonument 6000 5000 4000 3000 2000 1000 A RTBI=795.526+1.479RVBI , A A o1000 200030004000 Fig.4 Roughness(mm/km)measuredby vehiclemountedintegratorunit–RVBl Relationshipbetweentowedbumpintegratorunitandvehicle mountedintegratorunit 30 20 10 0 40 30 20 10 0 40 30 20 10 0 40 30 20 10 0 Volcanicgravel GL=O.12+0.27T ~2=0,93 l 10 20 3040 5060 Trafficpassesx103,T Volcanicgravel :/“‘2=098 GL=049+068T o 1020 30405060 Trafficpassesx 103, T Sandstonegravel -/:L=:’y45T o 102030 40 50 60 Trafficpasses x103,T Sandstonegravel :<,,, GL=–I.06+1.17T r2=0.g8 o10 20 3040 5060 Trafficpassesx103,T Fig.5Gravellossagainsttrafficpassesforsomevolcanicandsandstonegravelroads 1~ Lateriticgravel 30 l l l 20 GL=1.46t0,35T 10r2=O.go o 0. 2040 60 80100120 Trafficpassesx103,T 40 J Lateriticgravel u 30 z ~ z20 0 — % > 10 l GL=–0.93+0.22T $ 011111 0 2040 6080100120 40 30 20 10 0 40 30 20 10 0Trafficpassesx103,T Quartziticgravel l l GL=–1.42+0.67T r2Go,g4 1II11 010 203040 50 60 Trafficpassesx103,T Quartziticgravel :~:~lT o10 20 30 405060 Trafficpassesx103,T Fig.6GravellossagainsttrafficpassesforsomeIateriticandquartziticroads 35 30 25 20 15 10 5 0 . . . . 0Lateritic A Quartzitic AVolcanic 0Sandstone o I o 510 1520 2530 35 Actualgravelloss(mm) Fig.7Predictedannualgravellossagainstactualgravellossforroadsmonitoredinthisstudy 60 50 40 z >m 30 G 20 10 0q Kenva,1971(Coral) XKenva,1971(Quartzitic) OKenva,1971(Lateritic) AKenva,1971(Volcanic) l Kenva,1979(Lateritic) AKenva,1979(Volcanic) 25mm(1in)per100ADT15 / / / / /Ethiopia,Robinson16 /./ ,fl /,~< / Cameroon,Lombard17 /“~ Niger,Lombard” IvorvCoast,Lombard” Ghana,Roberts18 III o100200 300400 500600 Averagedailvtraffic,ADT Fig.8Gravellossratesfordifferentcountries 10 8 6 4 2 ,0 Quartziticgravels –0.358 T Y= 5.895e +1.0 o5 10 15 20 25 30 Cumulativetrafficpassesx10-3 Volcanicgravels –O.183T y=6,748 e +1.0 o 510 15 202530 Cumulativetrafficpassesx10–3 Fig.9Relationshipbemeendepthofloosematerialandtrafficpassesforquartzitic andvol=nicgravels 10 8 6 4 2 0 8 6 4 2 0 Fig.10 +1.0 Sandstones . –0.18664 y=6.925e . 0 0 . 0 0 u ~ III11 0 510 15202530 Cumulativetrafficpassesx10–3 . Laterites () () i o –0.334 y=6.Olle+0.5 00 0 nOo uw uo +II1PI o 5 10 15 20 2530 Cumulativetrafficpassesx10–3 Relationshipbetweendepthofloosematerialandtrafficpassesforsandstone andIateriticgravels 8000 6000 4000 2000 12000 10000 8000 600 400 200 Lateriticgravel AD2581all1 0D258lb/4 AD279/5 q D279/18 oA AA Ao A 0 R=3008+22.300T+0.531T2–0.0033T3 II 1 1 I I I1 0 2040 6080100120140 Trafficx10–3 . ) Volcanicgravel A D4841a17 O D484tb/5 AD482/7 7 0 A A 00A / A A * R= 3442–2.519T+8.111T2–0.0934T3 A o10 203040506070 Trafficx10–3 Fig.11RelationshipbetweenroughnessandtrafficforIateriticandvolcanicgravels 10000 800 600 400 [ 200 0 12000 10000 8000 600 4000 2000 ‘m Sandstone A AC107N/5 O C107SI8 R=3545–39,1T+5.06T2~ +0,149T3 R=3505+221T–6.36 +0.0736T3T2 10 20 30 40 Trafficx10–3 50 60 70 . . . . b E Quartzitic AE774/8 oD 4841C17 o / o R=3384+52.287T+0.143T2 +0.0157 T3 A A o 10 203040 5060 70 Trafficx10–3 Fig.12Relationshipbetweenroughnessandtrafficforsandstoneandquartziticgravels 1600[ 1400C 1200( 10000 z < ~ z w 8000 2 y $ 6000 4000 2000 0 0 Fig.13- . . 1 Im . G8 l G1OO G19 n G29 q G32 A l o 0 A A o q A ) o 20 4060 80100120 Trafficx10–3 RelationshipbetweenroughnessandtrafficforIateritic,quartziticandvol~nic gravelsderivedfromTRRLstudy1971–1974 m q A OD258 b14 AD279JS q D279/18 A Ao o 0 RD=7.18–0.081T+0.0069T2 –0.000036T3 1 1 I 1I1 0 2040 6080100120 140 Cumulativetrafficx10–3 A / A < RD=10.11+0.314T+0.00031 AA +0.00002T3 o A oT2 o 2040 6080100120140 Cumulativetrafficx10–3 Fig.14RelationshipbetweenrutdepthandtrafficforIateriticandvolcanicgravels ,U Quartzitic AE774/8 OD4821c17 o RD=7.49+0.171T+0.014T2 o –0.00009T3 . A o 0 0 0 10 20 3040506070 Cumulativetrafficx10–3 mSandstone AC 107S/8 OC107N/5 . . RD=7.09+0.573T–0.0128T2 A +0.00024T3 . A o . A o10 20304050 6070 Cumulativetrafficx10–3 Fig.15Relationshipbetweenrutdepthandtrafficforquartziticandsandstonegravels 40 30 20 10 0 0 2 . . Mariakani-KinangoRoad 98commercialaxles I I 4681012 Axleloads(tonnes) o 24 Kaloleni-LusitangiRoad 102commercialaxles 6 Axleloads(tonnes) Fig.16DistributionofaxleloadsonKinangoRoadand81012 LusitangiRoadinKaloleniarea 40 30 20 10 0 . . . .d o 2 4 6 . . . 0 Fig.172 4Withsitetraffic 420commercialaxles 1 8 10 12 Axleloads(tonnes) Withoutsitetraffic 300commercialaxles I I I I 6 810 12 Axleloads(tonnes) DistributionofaxleloadsonBungoma/BokoliRoadinBungomaarea withandwithoutsitetraffic 6000 Quartziticgravel A A RO=3303tl.117RF r2=0.83 01 II 1 1 0 300600901200 150 Cumulativerainfall(mm),RF 6000 Sandstonegravel o u Q ;o 400 0 2 (~ ~ & % a : ~ 200R.=3550+1.020RF Kr2=O.g, n1II103006009012001500 6000 o Cumulativerainfall(mm),RF Fig.18Roughnessagainstcumulativerainfallfornon-traffickedroads 8 m O Sandstonegravels :A: q Volcanicgravels 58%0~: AAAQuartziticgravels o~’;AAALateriticgravels I 1II --- ---.--- .F“ o300600 YuuILuuI300 Cumulativerainfall(mm),RF Fig.19Relationshipbetweenroughnessandcumulativerainfallforconstanttrafficvolumes Neg.no.E471a3 Plate1Towedbumpintegratortrailer Neg.no.E491a3 Plate2VehiclemountedbumpintegratorinCortinaEstateCar Plate3Gullyerosiononsteepgradient Neg.no.E46/83 Plate4Transversedrainageinstallations 8.APPENDIX1 SUMMARYOFT=UNPAVEDROADDETEWORATIONRELATIONS~PS 1.Gravelloss GLA= f= GLA= TA= RL= Vc=f (T::,,), (42t0.092TAt3.50RL2t1.88VC) 1.29forlateriticgravels 1.51forquartziticgravels 0.96forvolcanicgravels 1.38forsandstonegravels annualgravellossinmm annualtrafficvolumeinbothdirectionsmeasuredinthousandsofvehicles annualrainfallinmetres riseandfall(gradient)givenasapercentage. 2.Surfaceroughness hteriticgravels R=3008 Quartziticgravels R=3384 Volcanicgravels R=3442 t22.300T–0.531T2 –0.0033T3 +52.287T+0.143T2t0.0157T3 — 2.519T+8.11lT2–0.0934T3 Sandstonegravels R=3505+221T–6.360T2t0.0736T3 whereR=roughnessmeasuredinmm/hbytowedbumpintegrator andT=cumulativetrafficinbothdirectionssincegradingandmeasuredinthousandsofvehicles. 3.Surfacerutting hteriticgravels RD=7.18–0.081Tt Quartziticgravels RD=7.49+0.171T+0.0069T2–0.000036T3 0.014T2–0.00009T3 41 Volcanicgravels m=10.11t0.314Tt0.00031T2t0.00002T3 Sandstonegravels ~=7.09t0.573T–0.0128T2t0.00024T3 where~=rutdepthinmmmeasuredundera2metrestraight-edge andT=cumulativetrafficinbothdirectionssincegradingandmeasuredinthousandsofvehicles. 4.Surfacelooseness bteriticgravels DLM=6.01le–0.334T+0.5 Quarttiticgravels DLM=5.895e–0.358Ttl.O Volcanicgravels DLM=6.748e–‘.183Tt1.0 Sandstonegravels DLM=6.925e–0.187T+l.O whereDLM=depthofloosematerialinmm andT=cumulativetrafficinbothdirectionssincegradingandmeasuredinthousandsofvehicles. 42 Sections areaand road Bungoma D279 Bungoma D258/a Bungoma D258~Section no 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 89.APPE~IX2 TAB~1 Characteristicsofexperimentalsections Mteofrise andfrdlh percent 0.8 0.9 0.7 1.0 1.1 2.8 2.7 2.3 2.9 3.0 2.4 2.6 2.7 2.8 2.2 2.5 2.7 2.8 1.9 2.1 1.8 1.7 1.7 1.4 1.5 2.8 2.5 2.5 2.2 0.3 0.3 0.2 0.2 0.2 0.2 0.2 0.3Horizontal curvature degreekm 8 10 10 14 11 19 27 30 16 11 12 15 14 12 10 11 15 12 38 33 28 25 25 21 29 35 40 45 34 5 6 8 0 0 5 5 6Annurd rainfall mm 940 940 940 940 940 1170 1170 1170 1170 1170 1360 1360 1360 1360 1690 1690 1690 1690 910 910 910 910 910 910 910 910 910 910 910 590 590 590 590 590 590 590 590Annual traffic 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 65700 81760 81760 81760 81760 81760 81760 81760 81760 81760 81760 81760 40150 40150 40150 40150 40150 40150 40150 40150 Gravel type 43 TAB~1(continued) Sections areaand road Wte ofrise andfa~~ percentHorizontal curvature degree~Annurd rahfafl mm Section noAnnual trafficGravel type Meru E7741 2 3 4 5 6 7 80.6 0.5 0.7 0.5 0.3 0.3 0.8 0.410 8 8 15 10 9 17 7 970 970 970 970 970 970 970 970 9125 9125 9125 9125 9125 9125 9125 9125 51100 51100 51100 51100 51100 51100 51100 Meru D4821.0 1.3 1.1 1.0 1.3 1.5 1.4 20 29 29 21 22 21 17 1250 1250 1250 1250 1250 1250 1250 1 2 3 4 5 6 7 Meru D484/a1 2 3 4 5 6 71.0 0.8 1.1 0.9 1.3 1.1 1.083 79 68 65 84 66 611250 1250 1250 1250 1250 1250 125025550 25550 25550 25550 25550 25550 25550 Meru D484P1 2 3 4 52.6 2.1 2.5 2.2 2.0108 95 89 109 931380 1380 1380 1380 138025550 25550 25550 25550 25550 Meru D484/c1 2 3 4 5 6 7 8 0.9 0.7 0.6 0.8 0.8 0.7 0.3 0.59 11 14 12 16 8 8 91380 1380 1380 1380 1380 1380 1380 138034675 34675 34675 34675 34675 34675 34675 34675 bloleni C107N1 2 3 4 51.6 1.1 1,5 1.8 1.528 30 26 23 29640 640 640 640 640 35040 35040 35040 35040 35040 44 TM~1(continued) Sections areaand road =loleni clo7/s filoleni D549Section no 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 fite ofrise andfafl~ percent 1.1 1.8 1.7 1.9 1.6 ‘1.4 1.5 1.2 0.5 0.7 0.6 0.8 0.8 0.8 0.4 0.3 0.3 0.5 0.3 0.7 0.5 0.5Horizontal curvature degree~ 58 69 60 84 57 39 55 48 8 11 11 7 0 0 3 8 0 6 6 5 0 4bud rainfall mm 410 410 410 410 410 410 410 410 320 320 320 320 320 320 320 320 320 320 320 320 320 320kud traffic 15330 15330 15330 15330 15330 15330 15330 15330 13870 13870 13870 13870 13870 13870 13870 13870 13870 13870 13870 13870 13870 13870Gravel type 45 Road no D279 Bungoma area D258(a) Bungoma area10.APPENDIX3 TABLE1 Summaryofvehiclespeedsoverexperimentalsections Section no 1–5 6–7 8–12 13–17 18 1 –7 8–10 11Veficle* class c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c E MG HG BBeginningofstudy speedkm/h 66.7 54.6 51.4 N/A 40.9 45.7 48.7 46.2 N/A 40.9 72.2 59.0 49.4 N/A 45.0 71.1 58.1 47.4 N/A 42.8 69.2 58.1 47.4 N/A 40.5 52.9 56.3 50.2 N/A 50.3 52.9 55.0 48.0 41.9 50.2 110.7 69.2 66.7 N/A 61.3Endofstudy speedkmfi 64.6 48.7 51.1 N/A 40.2 45.8 49.3 45.7 N/A 38.1 71.1 57.6 47.2 N/A 42.2 69.7 56.6 47.2 N/A 41.5 6?.6 55.5 45.3 N/A 41.5 51.8 54.1 47.6 40.9 50.6 51.9 53.9 47.6 40.2 51.7 112.9 66.3 64.4 N/A 59,9Difference inspeedkm/h –2.1 –5.9 –0.3 — –0.7 W.1 W.6 –0.5 — –2.8 –1.1 –1.4 –2.2 — –2.8 –1.4 –1.5 –0.2 — –1.3 –1.6 –2.6 –1.9 — +1.0 –1.1 –2.2 –2.6 — W.3 –1.0 –1.1 –0.4 –1.7 +1.5 +2.2 –2.9 –2.3 — –1.4 *C=CarsLG=LightgoodsMG=MediumgoodsHG=HeavygoodsB=Buses 46 TABLE1(continued) Road no D258(b) Bungoma area E774 Meru area D482 Meru areaSection no 1 –3 4 5–8 1–3 4–7 8 1–3 4–6 7 VeMcle class c LG MG HG B c E MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG B c LG MG HG BBeginningofstudy speedkmh 112.5 66.7 65.2 N/A 58.1 104.8 64.0 61.5 N/A 58.1 108.5 69.2 66.7 N/A 59.9 65.1 62.1 59.1 N/A 38.9 62.1 61.6 58.1 N/A 42.4 69.2 66.7 61.1 N/A 42.4 47.8 42.9 36.7 N/A 47.8 48.7 48.6 36.7 N/A 47.8 48.7 48.7 35.3 N/A 48.9Endofstudy speedkmh 112.3 65.8 63.9 N/A 59.8 110.6 61.4 60.9 N/A 58.8 112.3 65.3 63.1 N/A 59.1 63.6 60.1 58.0 N/A 39.1 63.8 60.2 58.2 N/A 42.1 64.1 60.1 59.3 N/A 41.9 46.3 46.7 33.5 N/A 49.1 46.8 46.8 35.1 N/A 49.9 47.2 44.9 34.0 N/A 48.8Difference inspeedkm/h 4.2 –0.9 –1.3 — tl.7 t5.8 –2.6 –0.6 — to.7 t3.8 –3.9 –3.6 — –0.8 –1.5 –2.0 –1.1 — to.2 tl.7 –1.4 to.1 — –0.3 –5.1 –6.6 –1.8 — –0.5 –1.5 t3.8 –3.2 — tl.3 –1.9 –1.8 –1.6 — +2.1 –1.5 –3.8 –1.3 — –0.1 47 TAB~1(continued) Road no D484(a) Meru area D484(b) Meru area 9 D484(c) Meru area cl07(s) bloleni areaSection no 1–6 7 1–2 3–4 5 1–4 5–8 14 5–7 Vehicle class c x MG HG B c LG MG HG B c a MG HG B c LG MG HG B c z MG HG B c u MG HG B c E MG HG B c LG MG HG B c LG MG HG BBeginningofstudy speedkm/h 66.7 55.3 43.8 N/A 42.2 64.3 54.5 39.1 N/A 39.1 48.7 34.6 40.9 N/A 42.9 47.6 34.0 41.9 N/A 41.9 48.7 36.0 43.9 N/A 41.9 58.1 64.3 47.4 N/A 39.1 59.1 58.1 47.4 N/A 42.9 N/A 58.1 48.7 29.0 51.4 63.0 57.8 44.3 26.9 45.0Endofstudy speedkmm 65.4 53.5 39.6 N/A 40.9 63.2 52.2 37.4 N/A 40.7 44.1 31.8 36.9 N/A 40.2 43.8 30.6 36.3 N/A 39.5 42.2 29.0 33.3 N/A 39.0 55.4 60.6 47.1 N/A 40.7 54.7 56.9 48.3 N/A 41.2 58.0 54.5 46.2 30.5 48.8 58.0 54.5 47.6 29.0 47.8 Difference inspeedkmb –1.3 –1.8 –4.2 — –1.3 –1.0 –2.3 –1.7 — +1.6 –4.6 –2.8 –4.0 — –2.7 –3.8 –3.4 –5.6 — –2.4 –6.5 –7.0 –10.6 — –2.9 –2.7 –3.7 –0.3 — tl.6 –4.4 –1.2 W.9 — –1.7 — –3.6 –2.5 tl.5 t2.6 –5.0 –3.3 t3.3 +2.1 t2.8 48 TAB~1(continued) Road no cl07(s) continued clo7~ =loleni areaSection no 8 1–4 5 Veticle class c a MG HG B c z MG HG B c w MG HG BBeginningofstudy speedkm/h 55.9 54.5 48.6 32.7 49.2 66.7 58.0 49.3 36.0 54.5 67.1 56.3 48.7 N/A 50.9Endofstudy speedkm/h 56.3 53.5 47.6 29.3 46.2 63.0 53.5 48.5 32.7 52.7 63.1 51.9 48.0 31.9 51.9 Efference inspeedkmm W.4 –1.0 –1.0 –3.4 –3.0 –3.7 –4.5 –0.8 –3.3 –1.8 –4.0 –4.4 –0.7 — +1.0 49 ABSTRACT THEKENYAMAINTENANCESTUDYONUNPAVEDROADS:RESEARCHONDETERIORATION: TEJones:DepartmentoftheEnvironmentDepartmentofTransport,TRRLbboratoryReport1111: Crowthorne,1984(TransportandRoadResearchbboratory).Thisreportdescribesastudyofthe deteriorationofunpavedroadswhichwasundertakeninKenya.Theworkformedpartofalargerstudyof theeffectivenessofvariousmaintenancestrategiesontheperformanceofunpavedroads. Theprimaryobjectiveoftheresearchdescribedinthisreportwastoprovidenewandimproved deteriorationrelationshipsforunpavedroadsforuseinacomputermodelforestimatingtheconstruction costs,maintenancecostsandvehicleoperatingcostsofroadsindevelopmentcountries. ExperimentaltestsectionswerelocatedonthepublicroadnetworkinKenyaandtheirratesof deteriorationweremeasuredandquantifiedintermsofgravellossofthewearingcourse,surfaceroughness, depthofloosesurfacematerirdandrutdepths.Thetestsectionsweremonitoredforaperiodofover2years. Thedeteriorationofthetestsectionswasrelatedtocumulativetraffic,originaldesignstandardsand construction,maintenancestrategyandclimate. ISSN0305–1293 ABSTRACT THEKENYAMflNTENANCESTUDYONUNPAVEDROADS:RESEARCHONDETERIORATION: TEJones:DepartmentoftheEnvironmentDepartmentofTransport,TRRLLaboratoryReport1111: Crowthorne,1984(TransportandRoadResearchLaboratory).Thisreportdescribesastudyofthe deteriorationofunpavedroadswhichwasundertakeninKenya.Theworkformedpartofalargerstudyof theeffectivenessofvariousmaintenancestrategiesontheperformanceofunpavedroads. Theprimaryobjectiveoftheresearchdescribedinthisreportwastoprovidenewandimproved deteriorationrelationshipsforunpavedroadsforuseinacomputermodelforestimatingtheconstruction costs,maintenancecostsandvehicleoperatingcostsofroadsindevelopmentcountries. ExperimentaltestsectionswerelocatedonthepubficroadnetworkinKenyaandtheirratesof deteriorationweremeasuredandquantifiedintermsofgravellossofthewearingcourse,surfaceroughness, depthofloosesurfacematerialandrutdepths.Thetestsectionsweremonitoredforaperiodofover2years. Thedeteriorationofthetestsectionswasrelatedtocumulativetraffic,originaldesignstandardsand constmction,maintenancestrategyandclimate. ISSN0305–1293