8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
1/21
CHARACTERIZATIONTESTSFORMINERALFILLERSRELATEDTOPERFORMANCEOFASPHALTPAVINGMIXTURES
By
PrithviS.Kandal
CynthiaY.LynnFrazierParker
PaperpublishedinTransportationResearchBoard,TransportationResearchRecord1638,1998
277TechnologyParkway Auburn,AL36830
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
2/21
CHARACTERIZATIONTESTSFORMINERALFILLERSRELATEDTO PERFORMANCEOFASPHALTPAVINGMIXTURES
By
PrithviS.Kandhal
AssociateDirectorNationalCenterforAsphaltTechnology
AuburnUniversity,Alabama
CynthiaY.Lynn
GraduateStudentNationalCenterforAsphaltTechnology
AuburnUniversity,Alabama
FrazierParkerDirector
HighwayResearchCenterAuburnUniversity,Alabama
PaperpublishedinTransportationResearchBoard,TransportationResearchRecord1638,1998
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
3/21
DISCLAIMER
Thecontentsofthisreportreflecttheviewsoftheauthorswhoaresolelyresponsibleforthefactsandtheaccuracyofthedatapresentedherein.ThecontentsdonotnecessarilyreflecttheofficialviewsandpoliciesoftheNationalCenterforAsphaltTechnologyofAuburnUniversity.Thisreportdoesnotconstituteastandard,specification,orregulation.
i
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
4/21
ABSTRACT
Variousstudieshaveshownthatthepropertiesofmineralfillerespeciallythematerial
passing0.075mm(No.200)sieve(generallycalledP200material)haveasignificanteffectontheperformanceofasphaltpavingmixturesintermsofpermanentdeformation,fatiguecracking,andmoisturesusceptibility.However,researchershaveemployeddifferentcharacterizationtestsforevaluatingtheP200materials.
ThisstudywasundertakentodeterminewhichP200characterizationtestsaremostrelatedtothe
performanceofasphaltpavingmixtures.SixP200materialsrepresentingawiderangeofmineralogicalcompositionandparticlesizeswereused.TheseP200materialswerecharacterizedbysixtestsincludingRigdenvoids,particlesizeanalysis,andmethylenebluetest.Mixeswerepreparedwithtwofines/asphaltratios(0.8and1.5)byweight.MixvalidationtestsincludedtheSuperpavesheartestforevaluatingpermanentdeformationandfatiguecracking,andtheHamburgwheeltrackingtestandAASHTOT283forevaluatingmoisturesusceptibilityofthe12mixturescontainingdifferentP200materialsandfines/asphaltratios.
Theparticlesizesinmicronscorrespondingto60and10percentpassingandthemethylenebluetestweredeterminedtoberelatedtotheperformanceofasphaltpavingmixtures.
ii
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
5/21
Kandhal,Lynn,&Parker
CHARACTERIZATIONTESTSFORMINERALFILLERSRELATEDTO
PERFORMANCEOFASPHALTPAVINGMIXTURESPrithviS.Kandhal,CynthiaY.Lynn,andFrazierParkerJr.
INTRODUCTION
Accordingtovariousstudiesthepropertiesofmineralfillerespeciallythematerialpassing0.075mm(No.200)sieve(generallycalledP200material)haveasignificanteffectontheperformanceofasphaltpavingmixturesintermsofpermanentdeformation,fatiguecracking,andmoisturesusceptibility.However,researchershaveemployeddifferentcharacterizationtestsforevaluatingtheP200materials.ThisstudywasundertakentodeterminewhichP200characterizationtestsaremostrelatedtotheperformanceofasphaltpavingmixtures.
REVIEWOFLITERATURE
Numerousstudieshaveshownthatthepropertiesofmineralfiller(especiallythematerialpassingNo.200sieve)haveasignificanteffectonthepropertiesoftheHMAmixtures.Theintroductionofenvironmentalregulationsandthesubsequentadoptionofdustcollectionsystem(baghouse)hasencouragedthereturnofmostofthefinestotheHMAmixture.Amaximumfiller/asphaltratioof1@2to1@5,basedonweight,isusedbymanyagenciestolimittheamountoftheminus200material.However,thefinesvaryingradation,particleshape,surfacearea,voidcontent,mineralcomposition,andphysico-chemicalpropertiesand,therefore,theirinfluenceonthepropertiesofHMAmixturesalsovaries(1).Therefore,themaximumallowableamountshouldbedifferentfordifferentfines.
FinescaninfluencetheperformanceofHMAmixturesasfollows.
1.Dependingontheparticlesize,finescanactasafillerorasanextenderofasphaltcementbinder(2,3,4).Inthelattercaseanover-richHMAmixcanresultleadingto
flushingand/orrutting.Inmanycases,theamountofasphaltcementusedmustbereducedtopreventalossofstabilityorableedingpavement(5).2.Somefineshaveaconsiderableeffectontheasphaltcementmakingitactasamuch
stiffergradeofasphaltcementcomparedtotheneatasphaltcementgrade(1,3,6,7),andtherebyaffecttheHMApavementperformanceincludingitsfracturebehavior(8,9).
3.SomefinesmaketheHMAmixturessusceptibletomoisture-induceddamage(1).Water-sensitivityofonesourceofslagbaghousefineshasbeenreportedintheUnitedStates(5),andthewater-sensitivityofotherstonedustshasbeenreportedinGermany(10).StrippingofHMAmixturesasrelatedtothepropertiesoffiller/asphaltcombinations(fillerswereobtainedfromoperatingHMAplants)hasbeenreportedinJapan(11).
ItisveryimportanttocharacterizethefinessothattheperformanceparametersofHMApavements(resistancetopermanentdeformation,stripping,andfatiguecracking)arenotcompromised.
MATERIALS,TESTS,ANDTESTDATA
P200MaterialsandTests
Sixaggregatesources(Table1)werechosentorepresentawiderangeofmineralogicalcompositionandparticlesizes.Thesematerialswereobtainedbydrysievingfineaggregateparentrockovera75:m(No.200)sieve.
1
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
6/21
Kandhal,Lynn,&Parker
Table1.P200AggregateTestsResults
P-200TypeTest
P200-1 P200-2 P200-3 P200-4 P200-5 P200-6Natural Lime- Dolomite Granite Blast LimerockSand stone Furnace
Slag
SpecificGravity 2.558 2.760 2.955 2.872 3.043 2.798
RigdenVoid(British 39.3 35.4 32.3 41.4 40.6 34.3Standard),%
RigdenVoid 53.8 38.0 38.9 45.5 49.8 38.5(PennStateModified),%
FinenessModulus 3.67 2.46 4.99 4.50 4.47 2.81
D10(micron) 1.54 1.26 4.18 3.30 2.41 1.38D30(micron) 6.45 3.27 22.58 15.64 12.67 4.23
D60(micron) 26.92 9.98 51.95 40.89 43.41 14.60
SpecificSurfaceArea 12900 17968 6207 7206 8752 15603(cm2/ml)
MethyleneBlue 18.7 1.3 0.3 2.1 2.0 9.5
PlasticityIndex 29 NP NP NP NP NP
GermanFillerTest 35 70 80 60 55 75
ThefollowingtestswereusedtocharacterizetheP200materials.RigdenVoids(BritishStandard)-BS812RigdenVoids(PennStateModified)-Reference13ParticleSizeAnalysisMethyleneBlueTest-OhioDOTProcedurePlasticityIndex-AASHTOT90GermanFillerTest-KochMaterialsCompanyProcedure
Voidcontentinfines(generallycalledRigdenvoids)compactedtomaximumdensityhasbeenusedbyresearchersforcharacterizingthefines.Voidcontentisregulatedbyfourbasicpropertiesoffinesparticleshape,particlesize,particle-sizedistribution,andparticlesurfacestructure(1).Asampleofvacuum-ovendrysampleoffinesiseithervibratedinagraduatecylinder(1)orcompactedinasmallmoldbyacompactionhammer(12,13)tomaximum
packing.Mass(g)ofthecompactedfinesisdividedbythecompactedvolume(cm3)tocalculatebulkspecificgravity(GfB)ofcompactedfines.Apparentspecificgravity(GfS)ofthefinesolidsisdeterminedbyAASHTOT133usingkerosene.Voidcontent(V)inthefinescompactedtomaximumdensityisthencalculatedasfollows:BothBritishStandardBS812(12)andPennStatemodifiedequipment(13)wereusedtodetermineRigdenVoids.Theyarebasedonthesameconceptbutuseadifferentcompactive
2
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
7/21
Kandhal,Lynn,&Parker
effort.
ParticlesizeanalysiswasconductedwithaCoulterLF200Particlesizeanalyzer.Fromthis
analysis,severalparametersweredetermined.Therelativefinenessofanaggregatecanbe
determinedbythecalculatedfinenessmodulus(FM).FinenessmodulusoftheP200materialwascalculatedbydividingby100thesumofthepercentagesofP200materialcoarserthan75,50,30,20,10,5,3,and1microns.Thefinertheaggregate,thesmallerthefinenessmodulus.ParametersD10,D30,andD60werealsodeterminedfromparticlesizeanalysis.Theseparametersaretheparticlesizesthatcorrespond,respectively,to10,30,and60%ofthematerialpassing.Thespecificsurfacearea(cm2/ml)orSAwasthefinalparameterobtainedfromthisanalysis.
Themethyleneblue(MB)testisusedbytheInternationalSlurrySealAssociation(ISSA)to
quantifytheamountofharmfulclaysofthesmectite(montmorillinite)group,organicmatterandironhydroxidespresentinfineaggregate(14).Theprincipleofthetestistoaddquantitiesofastandardaqueoussolutionofthedye(methyleneblue)toasampleuntiladsorptionofthedye
ceases.TheGermanfillertestisameasureoftheamountofmineralfillerrequiredtoabsorb15grams
ofhydraulicoil.Thehydraulicoilisputinasmallbowl,then45gramsofmineralfillerisaddedandmixed.Anattemptismadetoformaballwiththemixture.Ifaballisformedandholdstogether,moremineralfiller,in5-gincrements,isadded.Thisprocessiscontinueduntilthemixturelosescohesion.Atthispoint,allofthehydraulicoilisfixedinthevoidsoftheP200materialandthereisnoexcesstoholdtheparticlestogether.ThetotalamountofP200addedtothehydraulicoilisreportedasthetestvalue.
Table1containstheresultsoftheP200characterizationtests.Eachvalueistheaverageofthree
replicates.
Table2containsthecorrelationmatrixbetweenaggregateproperties.Thecorrelationcoefficients(Rvalues)arethefirstnumberineachcell.Thesecondnumberineachcellisthestatisticalsignificancelevel(P)correspondingtothecorrelationcoefficient.
Rigdenvoids,BritishmethodandRigdenvoids,PennStatemodifiedhaveagoodcorrelation
(R=0.78,P=0.06)witheachotherbecausebothmeasurethevoidsinthecompactedP200materialalthoughwithdifferentcompactiveefforts.Rigdenvoids,PennStatemodified,hasanexcellentcorrelationwiththeGermanfillertest,whereasRigdenvoids,Britishmethod,hasonlyafaircorrelationwiththeGermanfillertest.
TheGermanfillertestisbasedindirectlyontheRigdenvoidsconcept.IftheRigdenvoidsare
high,theamountofP200materialneededtoreachtheendpointofthetestisrelativelylowbecausemorehydraulicoilisfixedbythehighvoids.TheGermanfillertestdoesnotrequireany
specialequipmentandisverysimpletoperformandcanpotentiallybesubstitutedforRigdenvoids,PennStatemethod.
3
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
8/21
Kandhal,Lynn,&Parker
Table2.CorrelationMatrixofP200AggregatePropertiesa
Rigden Rigden FM D10 D30 D60 SA MB GermanVoid,BS Void,PS Filler
Rigden 1.0 0.784 0.235 -0.073 -0.051 0.186 -0.250 0.166 -0.749Void,BS 0.065 0.655 0.890 0.923 0.724 0.633 0.753 0.086Rigden 1.0 0.326 -0.092 -0.013 0.277 -0.268 0.557 -0.949Void,PS 0.529 0.863 0.980 0.595 0.607 0.251 0.004FM 1.0 0.899 0.936 0.996 -0.990 -0.302 -0.058
0.015 0.006 0.0001 0.0001 0.561 0.913D10 1.0 0.992 0.907 -0.919 -0.538 0.321
0.0001 0.013 0.010 0.270 0.535D30 1.0 0.947 -0.942 -0.503 0.261
0.004 0.005 0.309 0.617D60 1.0 -0.983 -0.360 -0.001
0.0004 0.483 0.998SA 1.0 0.347 0.000
0.500 0.999MB 1.0 -0.680
0.137German 1.0Filler
aTopvaluesarecorrelationcoefficientsRandbottomvaluesaresignificancelevelsPineachcell.
Thetestparametersfinenessmodulus(FM),specificsurfacearea(SA),D10,D30,andD60arestronglyrelatedwitheachotherasshowninTable2.Allcorrelationsaresignificantatthe5%level.
Rigdenvoids,PennStatemethod,andRigdenvoids,Britishstandard,donothaveany
correlationwiththeparticlesizeparameters.Asmentionedearlier,Rigdenvoidsareregulatedbyparticleshapeandparticlesurfacetexturebesidesparticlesize.
MixtureValidationTests
TwelveHMAmixeswereevaluatedinthisstudy.ThesixP200aggregatesinboth0.8and1.5F/Aratioswerecombinedwithlimestonecoarseandfineaggregatetoproducevalidationmixes.Alllimestonewaswashedovera75:m(No.200)sievepriortobatchingtoremovetheP200material.Limestonewaschosenasthecoarseandfineaggregatesothatmoisturesusceptibilitywouldnotbecausedbythebaseaggregate.Moisturesusceptibilitydifferences,ifany,canthenbeattributedtotheeffectoftheP200material.Figure1showstheHMAmixgradationusedforthe0.8F/Aratio(5%passing75:msieve).TheHMAmixgradationforthe1.5F/Aratiowassameasthatfor0.8F/Aratioexceptithad8%passing75 :msieve.
ASuperpavePG64-22gradeasphaltcementwasusedinallHMAmixtesting.Optimum
asphaltcontentwasdeterminedbySuperpavevolumetricmixdesignforamixcontainingalllimestoneaggregate(includingtheP200fraction)usingthe0.8F/Aratio.Anasphaltcontentof5.3%gave4%airvoidsatNdesign(119gyrations,forintermediatedesigntrafficlevelof10
7ESALs).Thisasphaltcontentandsamecompactiveeffortwasusedforallvalidationmixes.However,relativelylowairvoids(averageof2.8percent)weregenerallyobtainedinvalidationsamples.ItisquitepossiblethatinsomecasesitcanbeattributedtoP200material'spotentialactionasanextender.
4
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
9/21
Kandhal,Lynn,&Parker
Figure1.HMAMixGradation
PermanentDeformationandFatigueCracking
SpecimensweremadetobetestedbytheSuperpavesheartester(frequencysweepatconstant
heightandsimpleshearatconstantheight)andtheindirecttensiletesterforevaluatingtheHMAresistancetopermanentdeformationandfatiguecracking(15,16).ThetestingofthecompactedspecimenswasperformedbytheAsphaltInstitute.
ThefollowingthreeindividualtestparameterswhichareusedintheSuperpaveintermediatemix
analysiswereusedtodeterminethepropensityoftheHMAmixturestopermanentdeformation(rutting)andfatiguecracking.
1.G*/sin*at0.1hertz.
G*/sin*oftheHMAmixissimilartoG*/sin*(ruttingparameter)ofPGgraded0
asphaltbinder.ItisameasureofHMAstiffnessathighpavementtemperature(40C)ataslowrateofloading(0.1cycle/second).HighervaluesofG*/sin*indicateincreasedstiffnessofHMAmixturesand,therefore,increasedresistancetorutting.
5
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
10/21
Kandhal,Lynn,&Parker
G*isthecomplexmodulusand*isthephaseanglewhenHMAistestedunderdynamicloading.
2.Slope(m)ofthefrequencyvsG*plot.Themvaluewasobtainedfromthefrequencysweepatconstantheightconductedby
theSuperpavesheartester(SST)athigheffectivetemperature(40C)forpermanentdeformationorTeff(PD)withfrequenciesrangingfrom0.01hertzto10hertz.Inotherwords,G*(stiffness)ofthecompactedHMAspecimenismeasuredatdifferentfrequencies.Theslope(m)ofthebestfitlineonthefrequencyvsG*plotiscalculated.ThissloperepresentstherateofdevelopmentofruttingforthetestedmixandisusedintheSuperpavemodelassuch.Thelowerthemslope,thebetteristhemix'sresistancetorutting.
3.G*sin*at1.0hertzG*sin*oftheHMAmixissimilartoG*sin*(fatiguefactor)oftheasphaltbinder.ItisameasureofthestiffnessatintermediateeffectivepavementtemperaturesforfatiguecrackingorTeff(FC).G*sin*wasmeasuredat1.0hertztorepresentfastmovingtraffic.AT
eff(FC)of20Cwasused.HighvaluesofG*sin*at1.0hertz
indicatehighstiffnessatintermediatetemperaturesand,therefore,lowresistancetofatiguecrackingaccordingtoSuperpave.
MoistureSusceptibility(Stripping)
KochMaterialsCompanyperformedtheHamburgwheeltrackingtestintheirlaboratoryinTerre
Haute,Indiana.TheHamburgwheeltrackingdevice(HWTD)measuresthecombinedeffectsofruttingandmoisturedamagebyrollingasteelwheelbackandforthacrossthesurfaceofaHMAslabthatissubmergedinhotwatermaintainedat50C(122F).Thetestingdurationis20,000cyclesanddeformationisrecordedandplottedaftereachcycle.Onthecyclesversusdeformationplottwodistinctlinesaregenerallyobserved.Thefirstline(ruttingline)indicatesruttingintheHMAunaffectedbystripping.Thefollowingsecondline(strippingline)withasteeperslopeindicatesruttingduetostripping.Thepoint(numberofcycles)wheretheslopeof
theruttinglineandtheslopeofthestrippinglineintersectiscalledtheinflectionpoint.Thisisthepointwherestrippingisassumedtohavebeeninitiated.Inflectionpoint(expressedintermsofnumberofcycles)isthetestparameterofinterestforthisstudy.TheHMAslabsgenerallyhadanairvoidcontentof71percent.
AASHTOT-283wasalsousedtomeasurethemoisture-susceptibilityoftheHMAmixesin
termsoftensilestrengthratio(TSR).
MixtureValidationDataandStatisticalAnalysis
Table3containsallmixturevalidationtestresults.ThemainobjectiveofthestatisticalanalysisistocorrelatetheP200aggregatepropertieswithHMApropertiesdeterminedbythemixvalidationtests.Table4showsthecorrelationmatrixbetweentheP200aggregatepropertiesand
theHMApropertiesatthe0.8F/Agradation.Table5showsthecorrelationmatrixbetweentheP200aggregatepropertiesandHMApropertiesatthe1.5F/Agradation.
G*/sin*@0.1hzandm(theslopeofthebestfitlineonthefrequencyvsG*plot)werethetwoHMAparameterschosentoindicatetheruttingpotential,asmentionedpreviously.NeitheroftheseparameterscorrelatestoanyoftheP200aggregatetestsatasignificantlevel(P
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
11/21
Kandhal,Lynn,&Parker
Table3.P200MixtureValidationTestResults
RuttingParameters
HighTemperature(40C)
FatigueCrackingParameter
IntermediateTemperature(20C)
MoistureSusceptibilityParameters
MixDesignation
G*/sin*@m0.1hz,psi
G*sin*@1.0hzpsi
TSRPercent
InflectionPoint(Hamburg)
1A 8722 0.43373 98051 66.2 7000
2A 9830 0.40572 89829 64.6 8400
3A 11034 0.40931 97967 51.7
200004A 9682 0.43596 88811 55.0
86005A 11271 0.40282 85644 57.5
100006A 12934 0.39183 93707 64.1
100001B 21700 0.31559 109560 64.7 6000
2B 21752 0.30167 98782 59.9 14400
3B 10269 0.40396 93318 52.4 98004B
11872 0.37874 84319 54.7 64005B
9510 0.41171 86723 64.5 80006B
25900 0.29158 95502 68.1 5000
Table4.CorrelationMatrixBetweenP200AggregatePropertiesandHMAProperties(0.8
F/AGradation)a
Rutting Fatigue Stripping
G*/sin*@0.1 m G*sin*@1.0hzTSR Inflectionhz Point
RigdenVoids -0.468 0.599 -0.556 0.033 -0.665(British 0.35 0.21 0.25 0.95 0.15Standard)
RigdenVoids -0.526 0.599 -0.051 0.157 -0.463(PennState 0.28 0.21 0.92 0.77 0.36Method)
Fineness -0.094 0.365 0.039 -0.865 0.561Modulus 0.86 0.48 0.94 0.03 0.25
D10 0.007 0.248 0.110 -0.968 0.758
0.99 0.63 0.84 0.001 0.08D30 0.01 0.228 0.113 -0.961 0.766
0.99 0.66 0.83 0.002 0.08
D60 -0.048 0.294 0.016 -0.886 0.6060.93 0.57 0.98 0.02 0.20
SpecificSurface0.049 -0.367 0.009 0.896 -0.543Area 0.92 0.47 0.99 0.02 0.27
MethyleneBlue-0.255 0.306 0.534 0.693 -0.4760.63 0.56 0.28 0.13 0.34
GermanFiller 0.657 -0.644 -0.003 -0.409 0.6470.16 0.17 0.99 0.42 0.17
a
TopvaluesarecorrelationcoefficientsRandbottomvaluesaresignificancelevelsPineachcell.
7
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
12/21
Kandhal,Lynn,&Parker
Table5.CorrelationMatrixBetweenP200AggregatePropertiesandHMAProperties(1.5
F/AGradation)a
Rutting Fatigue Stripping
G*/sin*@0.1hz m G*sin*@1.0hz TSR InflectionPoint
RigdenVoids -0.293 0.251 -0.257 0.143 -0.355(British 0.57 0.63 0.62 0.79 0.49Standard)
RigdenVoids -0.201 0.221 0.206 0.286 -0.465(PennState 0.70 0.67 0.70 0.58 0.35Method)
Fineness -0.901 0.926 -0.488 -0.584 -0.249Modulus 0.01 0.008 0.33 0.22 0.63
D10 -0.838 0.836 -0.560 -0.824 -0.0320.04 0.04 0.25 0.04 0.95
D30 -0.871 0.881 -0.532 -0.771 -0.0490.02 0.02 0.28 0.07 0.92
D60 -0.920 0.948 -0.511 -0.590 -0.1920.009 0.004 0.30 0.22 0.72
SpecificSurface 0.895 -0.915 0.572 0.609 0.288Area 0.016 0.01 0.24 0.20 0.58
MethyleneBlue 0.615 -0.559 0.776 0.613 -0.5590.19 0.25 0.07 0.20 0.25
GermanFiller -0.05 0.060 -0.413 -0.361 0.359
0.92 0.91 0.42 0.48 0.49aTopvaluesarecorrelationcoefficientsRandbottomvaluesaresignificancelevelsPineachcell.
TherewerenosignificantcorrelationsbetweenP200propertiesandG*sin*(fatiguefactor)ineither0.8F/Aratioor1.5F/Aratiogradations.TSRcorrelatedwellwiththegradationparametersatthe0.8F/Aratio.However,withthisexception,nosignificantcorrelationswereseenbetweenstrippingparametersandP200aggregatetestsateitherthe0.8F/Aratioorthe1.5F/Aratio.
ThecorrelationsaregenerallybetterfortheF/Aratioof1.5(becausehigheramountsofP200
wereused)thanF/Aratioof0.8.Therefore,P200testscanbebetterrelatedtoHMAperformanceatF/Aratioof1.5whichwillbeprimarilyusedinthisstudytoselecttheP200testswhicharerelatedtoHMAperformance.ItappearsfromTable5thatthefinenessofP200materialexpressedbythetestparametersD60,D30,D10,finenessmodulus,andspecificsurfaceareaissignificantlyrelatedtopermanentdeformationofHMAathighconcentrationlevelsofP200inthemix.
Asmentionedearlier,nosignificantrelationshipsareobservedbetweenP200aggregate
propertiesandHMAruttingparameters(G*/sin*@0.1hzandm)orHMAfatigueparameter(G*sin*)ataF/Aratioof0.8.ThisindicatesthatatlowconcentrationlevelsofP200,theeffectonruttingandfatigueisnotstatisticallysignificant.
8
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
13/21
Kandhal,Lynn,&Parker
ItappearsfromTable5thatthefinenessoftheP200material(especiallyD10)hasasignificant
effectontheretainedtensilestrength(TSR).Sinceparentrocks(limestone)ofthecoarseaggregateandfineaggregatearethesameinallmixes(onlytheP200isdifferent),theeffectofbinderstiffening(causedbytheP200material)appearstobedominantinthesemixes.The
smallerthesizeofP200(especiallyD10),themorethebinderisbeingmodifiedand/orextendedandthusgivesincreasedresistancetostrippinginAASHTOT283test.
Surprisingly,noP200aggregatetestshasanysignificantrelationshipwithstrippingwhen
measuredbytheHamburgwheeltrackingdevice.Methylenebluehasthehighest(althoughinsignificantatthe0.05level)relationshipwiththeinflectionpoint.Obviously,theHamburgwheeltrackingtest,whichisconductedwithHMAslabssubmergedinhotwater(50C)andsubjectedtomechanicalaction,issignificantlydifferentthanthestrippingprocessinAASHTOT283,whichdoesnotinvolveanymechanicalaction.
Again,similartotheresultsobtainedinmixeswithF/Aratioof1.5,thefinenessoftheP200
materialhasasignificanteffectontheretainedtensilestrength(TSR)ataF/Aratioof0.8(Table
4).AsalsoobservedinmixeswithF/Aratioof1.5,noP200testhasanysignificantrelationship
withstrippingwhenmeasuredbytheHamburgwheeltrackingdevice.Methyleneblueistheonlyindependentvariablewhichhasthehighest(butnotsignificantatthe0.05level)correlationwithinflectionpoint(R=-0.48,P=0.34).
TheforwardselectionmultiplevariablesproceduregivenintheSASprogramwasusedtoselect
theP200testswhicharerelatedtoHMAperformanceparameters.Theforwardselectionprocedurebeginsbyfindingthevariablethatproducestheoptimumone-variablesubset,thatis,thevariablewiththelargestcoefficientofdeterminationorR2.Inthesecondstep,theprocedurefindsthatvariablewhich,whenaddedtothealreadychosenvariable,resultsinthelargestincreaseinR2andsoon.Theprocesscontinuesuntilnovariableconsideredforadditiontothe
modelprovidesanincreasesinR2
consideredstatisticallysignificantatthespecifiedlevel(P=0.05forthisstudy).
Table6containstheP200testsselectedintheforwardselectionprocedureandthe
correspondingregressionequationsrelatingtheP200aggregateteststotheHMAperformanceparameters.
TheselectionoftheP200aggregateteststhatbestrelatetotheHMAperformancepropertieswill
bebasedsolelyontheinformationtakenfromthe1.5F/Agradationtesting.Thisisbecause1.5F/AratioseemstocorrelatemuchbetterwiththeHMAperformancepropertiescomparedto0.8F/AratioduetoanincreasedamountofP200materialinthemix.Themodelsof0.8F/AratiogenerallyhavelowcoefficientofdeterminationorR2valuesandinsignificantPvalues.
PermanentDeformation
G*/sin*@0.1hertzatHighTemperatureHighG*/sin*valuesindicateincreasedresistancetopermanentdeformationorrutting.Thetwo-variablemodel(seeTable6andFigure2)givesD60astheprimaryindependentvariableandmethyleneblueasthesecondindependentvariable.ThecoefficientofdeterminationorR2valueofthismodelis0.94(P=0.015)whichisexcellent.Astheparticlesize(at60%passing)decreases,theG*/sin*(stiffnessorresistancetorutting)increases.ItappearsthatthefinertheP200material,themoreitmodifiestheasphaltbinderandstiffenstheHMAmix.Themodelalsoindicatesthatthehigherthemethylenebluevalue(anotherindicationofthepresenceofveryfineP200),thehigherisG*/sin*.
9
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
14/21
Kandhal,Lynn,&Parker
Table6.RegressionEquationsBetweenP200AggregateTestsandHMAValidationTests
0.8F/AGradation
Performance Step Dependent Independent Equation R2 P
Parameter
Permanent 1 G*/sin* German G*/sin 0.43 0.16Deformation @0.1hz Filler *=6845.29+59.737(German
Filler)
Permanent 2 G*/sin* Rigden G*/sin*=- 0.52 0.32Deformation @0.1hz Voids(Penn 8062.28+217.88(Rigden
State) Voids,PennState)+144.60(GermanFiler)
Permanent 1 m German m=0.457-0.000699(German 0.42 0.17Deformation Filler Filler)
Permanent 2 m D10 m=0.450+0.0076(D10)- 0.65 0.21Deformation 0.000876(GermanFiller)
Fatigue 1 G*sin* Rigden G*sin*=120620.9- 0.31 [email protected] Voids 760.41(RigdenVoids,British
(British Standard)Standard)
Fatigue 2 G*sin* Methylene G*sin*=123483.4-- 0.71 [email protected] Blue 906.9(RigdenVoids,British
Standard)+457.7(Methylene Blue)
1.5F/AGradation
Performance Step Dependent Independent Equation R2
PParameter
Permanent 1 G*/sin* D60 G*/sin*=28961.36- 0.85 0.09Deformation @0.1hz 387.56(D60)
Permanent 2 G*/sin* Methylene G*/sin*=25596.17- 0.94 0.01Deformation @0.1hz Blue 338.18(D60)+321.94(Methyl 5
eneBlue)
Permanent 1 m D60 m=0.256+0.003(D60) 0.90 0.00Deformation 4
Permanent 2 m Methylene m=0.275+0.0027(D60)- 0.95 0.01Deformation Blue 0.0019(MethyleneBlue)
Fatigue 1 G*sin* Methylene G*sin*=89153.7+981.26(Me 0.60 [email protected] Blue thyleneBlue)
Fatigue 2 G*sin* Rigden G*sin*=124586.23- 0.75 [email protected] Voids 965.21(RigdenVoids,British
(British Standard)+1064.7(MethyleneStandard) Blue)
10
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
15/21
Kandhal,Lynn,&Parker
Figure2.PredictedG*/sin*Valuesvs.ActualG*/sin*UsingaTwo-Variable(D60and
MethyleneBlue)Model
mValue(SlopeoftheFrequencyvs.G*Plot)atHighTemperature HighmvaluesindicateincreasingrateofrutdevelopmentinHMAmixes.Thetwo-variablemodel(R2=0.95,P=0.01)givesD60astheprimaryindependentvariableandmethyleneblueasthesecondaryindependentvariableaffectingm(seeTable6andFigure3).ThisissimilartoG*/sin*@0.1hertz.
ItisrecommendedtouseD60andmethyleneblueastheP200testswhicharerelatedtoHMA
performanceintermsofpermanentdeformation.
FatigueCracking
G*sin*@1.0HertzatIntermediateTemperatureHighvaluesofG*sin*indicatehighmixturestiffnessatintermediatetemperatureandthereforeincreasedfatiguecracking.Thetwo-variablemodelforG*sin*@1.0hertzindicatesmethyleneblueastheprimaryindependentvariableandRigdenvoids,Britishstandardasthesecondaryvariable.Thisissimilartothe
ruttingmodels2inthathighervaluesofmethyleneblueindicatestifferHMAmixes.However,the
modelhasaRvalueof0.75andalevelofsignificanceof0.12(greaterthanthedesired0.05).Therefore,itappearsthattheeffectofP200materialatthe1.5F/Aratioisstatisticallynotsignificantand,therefore,noP200testisrecommendedforfatiguecracking.
11
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
16/21
Kandhal,Lynn,&Parker
Figure3.PredictedRUTMValuesvs.ActualRUTMValuesUsingaTwo-Variable(D60andMethyleneBlue)Model
Stripping
Twomixvalidationtests:AASHTOT283(ModifiedLottman)andHamburgwheeltracking
device,wereusedtodetermineHMAperformanceintermsofresistancetostrippingormoisturesusceptibility.
AASHTOT283Higherretainedtensilestrength(TSR)obtainedbythistestindicatesincreased
resistancetostripping.Thetwo-variablemodelforTSRataF/Aratioof1.5(R2=0.82,P=0.08)consistsofD10(P200sizeat10%passing)astheprimaryindependentvariable(R2=0.68,
P=0.04)andspecificsurfacearea(SA)ofP200asthesecondaryindependentvariable(Table7).TSRincreasesastheP200becomesfinerat10%passing(D10decreases).ItappearsthatveryfinesizeP200at10%passinglevelisstiffeningtheF/Abinderandthusprovidingincreasedresistancetostripping.Theliteraturereviewhasindicatedthathighviscosityasphaltbindersofferincreasedresistancetostrippingcomparedtolowviscosityasphaltbinders,allotherthingsbeingequal.
Thetwovariablemodel(seeTable7)2forTSRobtainedat0.8F/Aratioismuchbetterthanthat
obtainedata1.5F/Aratio.IthasaRvalueof0.98(P=0.003)andithasD10andmethyleneblueastheprimaryandsecondaryindependentvariablesaffectingHMAstripping.This
12
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
17/21
Kandhal,Lynn,&Parker
Table7.RegressionEquationsBetweenP200AggregateTestsandStripping
0.8F/AGradation
Performance Step Dependent Independent Equation R2 P
Parameter
Stripping 1 TSR D10 TSR=71.21-4.844(D10) 0.94 0.002
Stripping 2 TSR Methylene TSR=68.54- 0.98 0.003Blue 4.19(D10)+0.201(Methylene
Blue)
1.5F/AGradation
Performance Step Dependent Independent Equation R2 PParameter
Stripping 1 TSR D10 TSR=70.78-4.29(D10) 0.68 0.04
Stripping 2 TSR Specific TSR=95.47-8.84(D10)- 0.82 0.08Surface 0.001(SpecificSurfaceArea)Area
0.8F/AGradation
Step Dependent Independent Equation R2 PPerformanceParameter
Rutting 1 Inflection D30 InflectionPoint= 0.59 0.076Point 5496.7+478.4(D30)
Rutting 2 Inflection Rigden InflectionPoint= 0.98 0.003
Point Voids 35183.4+458.4(D30)-792.2British (RigdenVoidsBritishStandard)Standard
1.5F/AGradation
Performanc Step Dependent Independent Equation R2 PeParameter
Rutting 1 Inflection Methylene InflectionPoint=9784.9-268.6 0.31 0.25Point Blue (MethyleneBlue)
Rutting 2 Inflection Surface InflectionPoint=5792.6-359.9 0.58 0.28Point Area (MethyleneBlue)+0.39
(SurfaceArea)
indicatesthatthefineness(D10)ofthematerialaswellasthenature(methyleneblue)oftheP200materialaffectsHMAresistancetostripping(Figure4).Itappearsthatthetwo-variablemodelforTSRobtainedata1.5F/AratiohadalowerR2valueandhigherPvaluebecauselargeamountsoffinesstiffenedtheasphaltbindertoomuchandmaskedtheeffectofthenatureofthefines.
13
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
18/21
Kandhal,Lynn,&Parker
Figure4.PredictedTSRValuesvs.ActualTSRValuesfortheTwo-Variable(D10and
MethyleneBlue)Model
BasedontheTSRobtainedbyAASHTOT283,D10andmethylenebluearetherecommendedP200aggregatetestswhicharerelatedtostrippingofHMAmixes.Asstatedearlier,D10indicatesthefinenessoftheP200materialandmethyleneblueindicatesbothfinenessandnatureoftheP200material.
HamburgWheelTrackingTheinflectionpointobtainedinthistestrepresentsthenumberof
passesatwhichstrippingstartstooccurintheHMAmix.Thelargertheinflectionpoint(numberofpasses),thehigheristhemix'sresistancetostripping.Thetwo-variablemodelforinflection
pointata1.5F/AratiohasalowR2(0.58)andhighPvalue(0.28)andtherefore,isnotconsideredstatisticallysignificant(Table7).Thismodelhasmethyleneblueandspecificsurfaceareaastheprimaryandsecondaryindependentvariablesaffectingstripping.Thehigherthemethylenebluevalue,thelowertheinflectionpointandhenceresistancetostripping.ThisisinagreementwiththeobservationmadeincaseofTSRobtainedbyAASHTOT283.
Thetwo-variablemodelforinflectionpointata0.8F/Aratioissignificantlybetterthanthatata
1.5F/Aratio(Table7).ThisisinagreementwiththetrendseeninthecaseofTSR.Thetwo-variablemodel(R2=0.98,P=0.003)hasD30astheprimaryindependentvariableandRigdenvoids,Britishstandardasthesecondaryindependentvariableaffectingresistancetostripping.It
14
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
19/21
Kandhal,Lynn,&Parker
shouldbenotedthatD30hasahighcorrelation(R=0.99,P=0.0001)withD10whichwas
selectedastheprimaryindependentvariableincaseofTSR.ItisnotunderstoodwhytheRigdenvoids,Britishstandardwasselectedasthesecondaryindependentvariablebythestatisticalanalysis.Normally,higherRigdenvoidscausestifferF/Asystemsand,therefore,shouldresult
inincreasedresistancetostripping(orhighervaluesofinflectionpoint).However,themodelshowsanoppositeeffect,becausetheslopeofregressionisnegative.
ItshouldberealizedthatAASHTOT283andtheHamburgwheeltrackingdeviceare
significantlydifferentstrippingtests,althoughtheVTMvaluesofthetestspecimensarecomparable(71percent).AASHTOT283involvescuringofthemixina60Covenfor16hoursfollowedbyvacuumsaturationofHMA,thenimmersioninahot(60C)waterbathfor24hours,transferto25Cwaterbathfortwohours,andthentestingfortensilestrength.Hamburgwheeltrackingdoesnotinvolveanyvacuumsaturation.ItdoesinvolveimmersionoftheHMAina50Cwaterbath.However,whileimmersed,thesampleissubjectedtotheloadedwheeltrackingdeviceandrutdepthmeasurementsaretaken.Arepeated,dynamicmechanicalloadisappliedtotheHMAinHamburgwheeltrackingdevicewhereasnosuchloadingisappliedin
AASHTOT283.TheconditioningandtestingofHMAissignificantlydifferentinthesetwotests.Therefore,itisnotsurprisingthatdifferentindependentvariableswereselectedinthesetwomixvalidationtests.
ItisrecommendedthatD10andmethylenebluebeusedforHMAperformanceintermsof
stripping,takingintoconsiderationbothAASHTOT283andHamburgwheeltrackingtests.
D10istheprimaryindependentvariableincaseofTSR.IthashighcorrelationwithD30
(R=0.992,P=0.0001)whichwasselectedastheprimaryindependentvariableinHamburgwheeltrackingtest.
MethyleneblueisthesecondaryindependentvariableincaseofTSR.Thistestindicatesthe
natureandfinenessoftheP200material.
CONCLUSIONSANDRECOMMENDATIONS
Thefollowingconclusionscanbedrawnfromthisstudy.
1.Permanentdeformation.ThepermanentdeformationdataobtainedbytheSuperpavesheartesterintermsofG*/sin*@0.1hertzandmvalue(slopeoffrequencyvs.G*plot)indicatesthatD60(theparticlesizeofP200materialat60%passing)istheprimaryindependentvariableandthemethylenebluevalueisthesecondaryindependentvariableaffectingpermanentdeformationofHMAmixtures.ItappearsthatthefinertheP200material,themoreitmodifiestheasphaltbinderandstiffenstheHMAmix.BothlowervaluesofD60andhighervaluesofmethyleneblueindicatefinerP200material.
2.Fatiguecracking.ThefatiguecrackingdataobtainedbytheSuperpavesheartesterin
termsofG*/sin*@1hertzdidnotindicateanystatisticallysignificantcorrelationwithanyoftheP200propertiesevaluatedinthisstudy.
3.Stripping.ThestrippingdataobtainedbyAASHTOT283indicatesthatD10(theparticlesizeofP200at10%passing)istheprimaryindependentvariableandmethyleneblueisthesecondaryindependentvariableaffectingthestrippingpotentialofHMAmixes.D10indicatesthefinenessoftheP200materialandmethyleneblueindicatesbothfinenessandnatureoftheP200material.
ThefollowingteststhatarerelatedtoHMAperformancearerecommendedforevaluating
aggregatesforhotmixasphaltmixtures.
15
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
20/21
Kandhal,Lynn,&Parker
PerformanceParameterPermanentDeformationFatigueCracking
Stripping
RecommendedP200TestD60andMethyleneBlueNone
D10andMethyleneBlue
16
8/9/2019 Charactrization Tests for Mineral Fillers Related to Performance of Asphalt Paving Mixtures
21/21
Kandhal,Lynn,&Parker
REFERENCES
1.
2.3.4.5.6.
7.
8.9.10.
11.
12.
13.
14.
15.16.
Kandhal,P.S.EvaluationofBaghouseFinesinBituminousPavingMixtures.JournaloftheAssociationofAsphaltPavingTechnologists,Vol.50,1981.
Puzinauskas,V.P.FillerinAsphaltMixtures.Proceedings,CanadianTechnicalAsphaltAssociation,Vol.13,1968.Anderson,D.A.,J.P.Tarris,andD.Brock.DustCollectorFinesandTheirInfluenceonMixtureDesign.JournaloftheAssociationofAsphaltPavingTechnologists,Vol.51,1982.Tunnicliff,D.G.BindingEffectsofMineralFiller.Proceedings,AssociationofAsphaltPavingTechnologists,Vol.36,1967.Schrimsher,T.BaghouseDustanditsEffectonAsphalticMixtures.CaliforniaDepartmentofTransportation,ResearchReportCA-DOT-TL-3140-1-76-50,October1976.Eick,J.M.,andJ.F.Shook.TheEffectsofBaghouseFinesonAsphaltMixtures.AsphaltInstitute,ResearchReport78-3 ,November1978.
Dukatz,E.L.,andD.A.Anderson.TheEffectofVariousFillersontheMechanicalBehaviorofAsphaltAndAsphalticConcrete.JournaloftheAssociationofAsphaltPavingTechnologists,Vol.49,1980.Anderson,D.A.,H.U.Bahia,andR.Dongre.RheologicalPropertiesofMineralFiller-AsphaltMasticsanditsImportancetoPavementPerformance.AmericanSocietyforTestingandMaterials,STP1147,1992.Anderson,D.A.,R.Dongre,D.W.Christensen,andE.L.Dukatz.EffectsofMinus200SizedAggregateonFractureBehaviorofDense-GradedHotMixAsphalt.AmericanSocietyforTestingMaterials ,STP1147,1992.Breuer,J.V.TestingtheWaterSensitivityofStoneDustFillers, Bitumen(Germany),Vol.29,No.3,May1967.Yamashita,H.,andK.Sakata.StrippingofAsphaltConcreteMixturesasRelatedtothePropertiesofFiller/AsphaltCombinations.Proceedings,12thJapanRoadCongress,
September1975.BS812:MethodsforSamplingandTestingofMineralAggregates,SandsandProperties.BritishStandardInstitution,London,1975.Anderson,D.A.GuidelinesonUseofBaghouseFines.NAPA,InformationServices101,1987.InternationalSlurrySealAssociation.TestMethodforDeterminationofMethyleneBlueAbsorptionValue(MBV)ofMineralAggregateFillersandFines.ISSA,Bulletin145,1989.TestMethodforDeterminingthePermanentDeformationandFatigueCrackingCharacteristicsofHotMixAsphalt(HMA)UsingtheSimpleShearTest(SST)Device.TP7-94,AASHTOProvisionalStandards,June1996.TestMethodforDeterminingCreepComplianceandStrengthofHotMixAsphaltUsingtheIndirectTensileTestDevice.TP9-94,AASHTOProvisionalStandards,June1996.
17