NCATReport16-02ENHANCEDCOMPACTIONTOIMPROVEDURABILITYANDEXTENDPAVEMENTSERVICELIFE:ALITERATUREREVIEWByNamTran,Ph.D.,P.E.,LEEDGAPamelaTurnerJamesShambleyApril2016

1.ReportNo.NCATReportNo.16-02

2.GovernmentAccessionNo

3.Recipient’sCatalogNo

4.TitleandSubtitleEnhancedCompactiontoImproveDurabilityandExtendPavementServiceLife:ALiteratureReview

5.ReportDateApril2016

6.PerformingOrganizationCode

7.AuthorsNamTran,Ph.D.,P.E.,LEEDGA;PamelaTurner;JamesShambley

8.PerformingOrganizationReportNo.NCATReportNo.16-02

9.PerformingOrganizationNameandAddressNationalCenterforAsphaltTechnologyatAuburnUniversity277TechnologyParkwayAuburn,AL36830

10.WorkUnitNo.

11.ContractorGrantNo.DTFH61-11-H-00032

12.SponsoringAgencyNameandAddressFederalHighwayAdministrationOfficeofAssetManagement,Pavements,andConstruction1200NewJerseyAvenue,SEWashingtonDC20590

13.TypeofReportandPeriodCoveredFinal

14.SponsoringAgencyCodeFHWA

15.SupplementaryNotesFHWATechnicalContact:TimothyAschenbrener,P.E.

16.AbstractThis literature review was conducted to provide information to support the FHWA Asphalt Pavement Technology

Program strategic directiononextendingpavement service life throughenhanced field compaction. The results from thepaststudiesclearlyindicatetheeffectoflowin-placeairvoidsonthefatigueandruttingperformanceofasphaltpavements.A1%decreaseinairvoidswasestimatedtoimprovethefatigueperformanceofasphaltpavementsbetween8.2and43.8%and the rutting resistanceby7.3 to66.3%. Inaddition,a1%reduction in in-placeair voidscanextend theservice lifebyconservatively10%.Basedontheseresults,alifecyclecostanalysis(LCCA)wasconductedontwoalternativesinwhichtheexactsameasphaltoverlaywouldbeconstructedto93%and92%densitiestoillustratetheeffectofin-placeairvoidsonthelifecyclecostofasphaltpavements.TheLCCAresultsshowthattheuseragencywouldseeanet-present-valuecostsavingsof$88,000ona$1,000,000pavingproject(or8.8%)byincreasingtheminimumrequireddensityby1%.

Duetoitssignificanteffect,thecostofprovidingincreasedin-placedensitycanbesignificantlylessthantheoperation,maintenance,and roadusercost savings realizeddue toextendedservice lifeof thepavements. InanAASHTOsurveyofstate agencies’ targets for field compaction conducted in 2007, the majority of states responding to the survey had acompaction target of 92 percent, but over one-third of the responding agencies had compaction targets less than 92percent.Mostofthesein-placedensityrequirementscurrentlyadoptedbystatesweredeterminedbasedonwhatlevelsofin-placedensitycouldbeachievedinthepastusingpriorconstructiontechnologies.Sincein-placedensityhasasignificantimpact on the performance of asphalt pavements, agencies may consider implementing a higher in-place densityrequirement that can be achievable by following best practices and adopting new asphalt pavement technologies andknowledgegainedfromrecentresearch.Someofthesetechnologiesandknowledge,includingwarmmixasphalt,intelligentcompaction, improved construction joints, and improved agency specifications to incentivize achieving higher in-placedensities,arebrieflydiscussedinthisreport. 17.KeyWordsIn-placedensity,airvoids,fieldcompaction,durability,servicelife

18.DistributionStatementNorestriction.ThisdocumentisavailabletothepublicthroughtheNationalTechnicalInformationService5285PortRoyalRoadSpringfield,VA22161

19.SecurityClassif.(ofthisreport)Unclassified

20.SecurityClassif.(ofthispage)Unclassified

21.No.ofPages25

22.Price

FormDOTF1700.7(8-72) Reproductionofcompletedpageauthorized

ENHANCEDCOMPACTIONTOIMPROVEDURABILITYANDEXTENDPAVEMENTSERVICELIFE:ALITERATUREREVIEW

FinalReportBy

NamTran,Ph.D.,P.E.,LEEDGAPamelaTurnerJamesShambley

NationalCenterforAsphaltTechnologyAuburnUniversity,Auburn,Alabama

April2016

iv

ACKNOWLEDGEMENTS

ThisprojectwasmadepossiblebytheFederalHighwayAdministration.Theauthorswouldliketothankthemanypersonnelwhocontributedtothecoordinationandaccomplishmentoftheworkpresentedherein.SpecialthanksareextendedtoTimothyAschenbrener,PEforservingasthetechnicalrepresentative.

DISCLAIMERThecontentsofthisreportreflecttheviewsoftheauthorswhoareresponsibleforthefactsandaccuracyofthedatapresentedherein.Thecontentsdonotnecessarilyreflecttheofficialviewsorpoliciesofthesponsoringagency,theNationalCenterforAsphaltTechnologyorAuburnUniversity.Thisreportdoesnotconstituteastandard,specification,orregulation.Commentscontainedinthispaperrelatedtospecifictestingequipmentandmaterialsshouldnotbeconsideredanendorsementofanycommercialproductorservice;nosuchendorsementisintendedorimplied.

v

TABLEOFCONTENTS

1.INTRODUCTION..........................................................................................................................1

2.EFFECTOFAIRVOIDSONPAVEMENTPERFORMANCEANDLIFECYCLECOST...........................22.1.EffectofAirVoidsonFatiguePerformance........................................................................22.2.EffectofAirVoidsonRuttingPerformance.........................................................................42.3.EffectofIn-PlaceAirVoidsonServiceLife..........................................................................62.4.EffectofIn-PlaceAirVoidsonLifeCycleCost.....................................................................62.4.Summary..............................................................................................................................7

3.NEWTECHNOLOGIESFORACHIEVINGHIGHERIN-PLACEDENSITY............................................83.1.WarmMixAsphalt...............................................................................................................83.2.FieldCompaction...............................................................................................................103.3.ImprovedConstructionJoints............................................................................................123.4.AgencySpecifications........................................................................................................133.4.1.ProjectSelection.........................................................................................................133.4.2.LiftThickness..............................................................................................................133.4.3.MixDesign..................................................................................................................143.4.4.Criteria........................................................................................................................143.4.5.PerformanceIncentives..............................................................................................15

4.SUMMARY................................................................................................................................15

REFERENCES.................................................................................................................................17

Tran,Turner,Shambley

1

1.INTRODUCTIONThe Federal Highway Administration (FHWA) Pavement Policy in Title 23, Part 626, Code ofFederalRegulations (23CFR626.3)states that“pavementshallbedesignedtoaccommodatecurrentandpredictedtrafficneedsinasafe,durable,andcosteffectivemanner.”Whilesafetyisaparamountaspectofpavementthatmustalwaysbemaintained,improvingthedurabilityofpavement while maintaining cost effectiveness is becoming increasingly important due tofinancialrealitiescurrentlyfacingstatehighwayagencies(SHAs).

Significant advancements in technology and techniques related to asphalt pavementdesignandconstructionyieldthepotentialforincreasingbothdurabilityandcosteffectiveness.Someof these advancements can be employed immediately to enhance field compaction toreducein-placeairvoidsandimprovemixturedurabilityandpavementservicelife.

Currently, most SHAs use a percentage of the theoretical maximum specific gravity(Gmm)toquantify in-placeairvoids(or in-placedensity)foracceptanceofconstruction.Otheragenciesmayuseacontrolstripormethodspecificationstocontroldensity.Figure1showstheresults of a 2007 American Association of State Highway Transportation Officials (AASHTO)Survey of SHAs’ targets for field compaction normalized to Gmm. The majority of statesresponding to the survey have a compaction target of 92%, but over one-third of therespondingSHAshavecompactiontargetslessthan92%.

Figure1.NormalizedCompactionTargetsbyState(Source:AASHTO2007SOMSurvey)

MostofthecompactiontargetsshowninFigure1weredeterminedbasedonhistorical

data and experience with prior construction technologies. They were based more on whatlevelsofin-placedensitycouldbeachievedinthepastthanonoptimalin-placedensities.With

10

2

9

43

22

32

1 1

0

5

10

15

20

25

89.0%89.5%90.0%90.5%91.0%91.5%92.0%92.5%93.0%93.5%94.0%

Num

bero

fStates

NormalizedCompac[onTarget(PercentofGmm)

Tran,Turner,Shambley

2

modern design and constructionmethods and equipment, optimal in-place densities can beachievablewithappropriateguidance,specifications,andincentives.

Past studies have shown that a small increase in in-place density through roadwaydesign, mix design, and/or construction can lead to a significant increase in service life ofasphalt pavements. According to Epps and Monismith (1971), Finn and Epps (1980), andPuangchi, et al. (1982), fatigue life (the time fromoriginal construction to significant fatiguecracking)ofasphaltpavementsisreducedapproximately10to30%forevery1%increaseinin-place air voids. Thus, for the states currently requiring a compaction target below 92%, anincreaseof10to30%inthepavementservice lifecanbeachievedbyraisingthecompactiontargetby1%.

Duetoitssignificanteffect,thecostofprovidingthisincreasedin-placedensitycanbesignificantly less thantheoperation,maintenance,androadusercostsavingsrealizedduetoextended service life of the pavements. As Noel stated at the 1977 Association of AsphaltPavingTechnologists(AAPT)meeting,“thesinglemostimportantconstructioncontrolthatwillprovideforlong-termserviceabilityiscompaction”(Hughesetal.1989).

Recognizing the importance of in-place density in building cost effective asphaltpavements, this literature review was tasked to provide information to support the FHWAAsphaltPavementTechnologyProgramstrategicdirectiononextendingpavementservice lifethrough enhanced field compaction. In the following sections, the effect of air voids onlaboratoryandfieldperformanceofasphaltmixturesisfirstdiscussed,followedbyasummaryofbestpracticesforachievinghigherin-placedensity.2.EFFECTOFAIRVOIDSONPAVEMENTPERFORMANCEANDLIFECYCLECOSTSeveral studies have been conducted to determine the effect of air voids on performancecharacteristicsofasphaltmixtures,includingfatiguecrackingandrutting.Keyfindingsfrompastlaboratoryandfieldstudiesontheeffectofairvoidsoneachoftheperformancecharacteristicsarefirstsummarizedinthissection,followedbyalifecyclecostanalysistoquantifytheeffectofin-placeairvoidsonthelifecyclecost.

2.1.EffectofAirVoidsonFatiguePerformanceEppsandMonismith(1969)conductedoneofthefirstmajorlaboratorystudiesontheeffectofairvoidsonlaboratoryfatigueperformanceofasphaltmixturesattheUniversityofCaliforniaatBerkeley (UCB). In this study, constant stress fatigue testingwas conductedon threeasphaltmixtures. The first mix was referred to as British Standard 594 Grading with 7.9% bindercontent.ThesecondmixwasreferredtoasCaliforniaFineGradingwith6%bindercontent,andthethirdmixwasreferredtoasCaliforniaCoarseGradingwith6%bindercontent.Twenty-sixtestswerereportedforthefirstmixwithspecimenairvoidsvariedbetween4%and14%,22tests reported for the second mix with air voids varied between 5% and 8%, and 20 testsreportedforthethirdmixwithairvoidsvariedbetween2.5%and7%.Basedonthetestresultsreported,theeffectof1%increaseinairvoidsonfatigueperformancereductionwasestimatedtobe20.6%,43.8%, and33.8% for the first, second, and thirdmix, respectively (Seedset al.2002).

Tran,Turner,Shambley

3

HarveyandTsai(1996)conductedanotherlaboratoryfatigueexperimentatUCBusingthe flexural bending beam test developed under the Strategic Highway Research Program(SHRP).Inthisexperiment,adense-gradedasphaltmixturewithanAR-4000binderwastestedbasedonafullfactorialdesignwiththreelevelsofairvoidsandfivelevelsofbindercontents.Thethreeairvoidlevels,including1%to3%,4%to6%,and7%to9%,wereselectedbasedontherangetypicallyobtainedinCaliforniawiththespecificationatthetimeandtoprovidedatafor the evaluation of higher compaction standards. A relationship between cycles to fatiguefailure,airvoids,bindercontent,andappliedstrainwasdeterminedbasedon97fatiguetestresults.Basedonthisrelationship,a1%increaseinairvoidswasestimatedtoresultina15.1%reductioninfatigueperformance(Seedsetal.2002). Amajorresearcheffortthatincludedbothlaboratoryandfieldstudieswasconductedafewyears lateraspartof theWesTrackproject (Eppsetal.2002) to investigate theeffectofchangesinairvoidcontent,bindercontent,andaggregategradationonmixtureandpavementperformancecharacteristics.Threeasphaltmixtureswithfine,fine-plus,andcoarseaggregategradations were evaluated. During construction, the binder content and in-place air voidcontentwerevariedtoyieldsevenuniquetreatmentcombinationsforeachmix.Threelevelsofin-placeair voidswere selected, including4%,8%,and12%.The intermediate level (i.e.,8%)was selected to representa typical in-placeair void content inpavements.The lowandhighlevels (i.e., 4% and 12%)were selected to represent expected extremes in in-place air voidsexperiencedinasphaltpavementconstruction.Theseparationof4%betweenthethreeairvoidlevelswasconsideredsufficienttoensurethatstatisticaldifferenceswouldexistbetweenthelow-andmedium-levelsectionsandbetweenthemedium-andhigh-levelsections.Theeffectof changes in mixture properties on the fatigue performance was investigated both in thelaboratoryandinthefield. Aspartofthelaboratoryexperiment,fatiguetestingwasconductedonbeamsextractedfromtheWesTrack(Eppsetal.2002).Sevenuniquemixcombinationsplustworeplicatesweretestedforthefinemix,sevenuniquemixcombinationsplusonereplicatetestedforthefine-plusmix, and seven uniquemix combinations plus two replicates tested for the coarsemix.Basedonthelaboratorytestresults,arelationshipbetweencyclestofailure,airvoidcontent,binder content, pavement temperature at 6-inchdepth, andmaximum strainwasdevelopedfor each of the three mixes. Based on these relationships, a 1% increase in air voids wasestimated to reduce the fatigue performance of fine, fine-plus, and coarsemixes by 13.5%,13.3%,and9%,respectively(Seedsetal.2002).

Forthefieldexperiment,theperformanceofWesTracktestsectionsvariedsignificantly,and some sections never showed fatigue cracking at the end of the experiment. Therelationships between load applications or number of equivalent single axle loads (ESALs) to10%fatiguecrackingandothermixtureproperties, includingairvoidcontent,bindercontent,and fines content,weredevelopedbasedonaprobabilistic regressionapproach to take intoaccount theperformanceof sections thatneverexhibited fatiguecracking.Twomodelsweredeveloped:onefor17testsectionswithfineandfine-plusmixes(14uniquemixcombinationsplusthreereplicates);andtheotherforninetestsectionsusingcoarsemixes(sevenuniquemixcombinationsplus tworeplicates) (Eppsetal.2002).Basedontherelationshipsdeveloped inthefieldexperiment,theeffectof1%increaseinin-placeairvoidsonfieldfatigueperformance

Tran,Turner,Shambley

4

reduction was estimated to be 21.3% and 8.2% for the fine/five-plus and coarse mixes,respectively(Seedsetal.2002).

Morerecently, theAsphalt Institute (AI)conducteda laboratorystudy toevaluate theeffectofairvoidsontheperformanceofatypicalKentuckyasphaltmixture(BlankenshipandAnderson2010).The flexuralbendingbeamtestwasconductedat fivestrain levelsonbeamspecimenspreparedat variousair void levelsbetween4%and11.5%.The laboratory testingresults showed that theeffect of air voidson the fatigue life of the asphaltmixturebecamemorepronouncedatlowerstrainlevels(i.e.,350and400microstrain).At350microstrain,thefatigueperformancereducedby42%astheairvoidcontentincreasedfrom7%to11.5%(Fisheretal.2010).Thiscorrespondstoa9.2%reductioninfatiguelifefora1%increaseinairvoids. In summary, results from the past studies clearly indicate the adverse effect ofincreased in-place air voids on the fatigue performance of asphalt pavements. Table 1summarizes these results. Depending on themix type and experiment, a 1% decrease in airvoidswas estimated to improve the fatigue performance of asphalt pavements between 8.2and43.8%.Table1.EffectofAirVoidsonFatiguePerformance

Study Lab/FieldExperiment

MixType AirVoidsEvaluated

IncreaseinFatigueLifefor1%DecreaseinAirVoids

UCB(Eppsand

Monismish1969)

Lab BritishStandard 4-14% 20.6%1CaliforniaFine 5-8% 43.8%1

CaliforniaCoarse 2.5–7% 33.8%1UCB

(HarveyandTsai1996)

Lab CaliforniaDense-Graded

1-3%4-6%7-9%

15.1%1

WesTrack(Eppsetal.2002)

Lab Fine 4,8,12% 13.5%1Fine-Plus 4,8,12% 13.3%1Coarse 4,8,12% 9.0%1

Field Fine/Fine-Plus 4,8,12% 21.3%1Coarse 4,8,12% 8.2%1

AI(Fisheretal.2010)

Lab 9.5mmDense-Graded

4–11.5% 9.2%

1(Seedsetal.2002)2.2.EffectofAirVoidsonRuttingPerformanceThe effect of changes inmixture properties, including air void content, binder content, andaggregategradation,onruttingperformancewasalsoinvestigatedinafieldstudyaspartoftheWesTrackproject(Eppsetal.2002).Unlikethefatiguecrackingevaluationpreviouslydescribed,this investigation was only conducted based on rutting data measured on WesTrack testsectionsthroughthefirsttwomillionequivalentsingleaxleloads(ESALs).Threeanalyseswereconductedfortheruttingdata.Level1Aanalysiswasbasedonadirectregressioncorrelatingrutdepthtotraffic,airtemperature,andmixtureproperties.Thisanalysiswasconductedon26original and eight replacement sections. Level 2 analysis included mechanistic-empiricalanalysesofallthedataassumingthepavementbehavesasamultilayerelasticsystem.Level1B

Tran,Turner,Shambley

5

analysis was a regression correlating rut depth to traffic, air temperature, and mixtureinformationthatwasgeneratedbasedontheLevel2analysisof23testsectionsthatexhibitednoorlittlefatiguecracking.ResultsofLevel1Aand1Banalysesareofparticularinteresttothisliteraturereviewandarediscussedinthefollowingparagraphs. FortheLevel1Aanalysis,threeseparateruttingmodelsweredeveloped.Thefirstonewasdevelopedforthefineandfine-plusmixesbasedonthedatacollectedin17testsections(14 uniquemix combinations plus three replicates). The second onewas generated for nineoriginalcoarsemixtures(sevenuniquemixcombinationsplustworeplicates).Thethirdmodelwas developed for eight coarse mixtures in the replacement sections (seven unique mixcombinations plus one replicate) (Epps et al. 2002). Based on the Level 1A models, a 1%increase in in-placeairvoidswasestimated to reduce therutting resistanceof fine/fine-plus,originalcoarse,andreplacementcoarsemixesby11.5%,9.6%,and66.3%,respectively(Seedsetal.2002). FromtheLevel1Banalysis,onemodelwasdevelopedwithdummyvariables(i.e.,0or1)to account for the effect of mix type on rutting. This model essentially consists of threeseparatemodels:oneforfineandfine-plusmixes,onefororiginalcoarsemixes,andtheotherforreplacementcoarsemixes.Basedonthismodel,a1%increaseinairvoidscorrespondstoa7.3%reductioninruttingresistanceforthefine,fine-plus,andcoarsemixes.Inaddition,a1%increase in air voids was estimated to cause a 10.9% decrease in rutting resistance for thereplacementcoarsemixes(Seedsetal.2002). IntheAIstudy(BlankenshipandAnderson2010),theeffectofairvoidsontheruttingperformance of an asphaltmixturewas evaluated using the flownumber test in accordancewith AASHTP TP 79. The flow number test results showed that the rutting resistance of themixturedecreasedby34%as theairvoids increased from7%to8.5%.Thiscorresponds toa22.7%reductioninruttingresistancefora1%increaseinairvoids.

Insummary,resultsfrompaststudiesclearlyindicatetheadverseeffectofincreasedin-place air voids on the rutting of asphalt pavements. The effect is summarized in Table 2.Dependingonthemixtypeandanalysis,a1%decreaseinairvoidswasestimatedtoimprovetheruttingresistanceofasphaltmixturesby7.3to66.3%.Table2.EffectofAirVoidsonRuttingPerformance

Study Lab/FieldExperiment

MixType AirVoidsEvaluated

FinalFieldRutDepth(mm)

DecreaseinRuttingfor1%Decreasein

AirVoidsWesTrack(Eppsetal.

2002)

FieldLevel1A

Fine/Fine-Plus 4,8,12% 9-35 11.5%1OriginalCoarse 4,8,12% 13-36 9.6%1

ReplacementCoarse 4,8,12% 12-26 66.3%1Field

Level1BFine/Fine-Plus/Coarse 4,8,12% 9-36 7.3%1ReplacementCoarse 4,8,12% 12-26 10.9%1

AI(Fisheretal.

2010)

Lab 9.5mmDense-Graded 4–11.5% N/A 22.7%

1(Seedsetal.2002);N/A=NotApplicable

Tran,Turner,Shambley

6

2.3.EffectofIn-PlaceAirVoidsonServiceLifeIna recent study todevelopperformance-relatedpayadjustment factors for theNew JerseyDepartment of Transportation (NJDOT), Wang et al. (2015) analyzed construction andperformancedataofpavementsinNewJersey.Amongseveralfactorsidentifiedbytheauthors,theaverage in-placeair voidsmeasured through field core testingwere found to impact theservicelifeofasphaltmixtures,asillustratedinFigure2.Theservicelifeisdefinedasthetimefrom initial construction until the next rehabilitation activity for each pavement section. Thedata were mined from the NJDOT’s pavement management system and Materials Bureauqualityassurancetestingdatabasefor55pavementsections.ThecorrelationshowninFigure2indicatesthattheservicelifedecreasesapproximatelyoneyearforevery1%increaseinthein-placeairvoids.Thiscorrespondstoanapproximate10%increaseinasphaltmixtureservicelifefora1%decreaseinin-placeairvoids.

Figure2.CorrelationsbetweenAverageAirVoidsandServiceLifeofAsphaltMixtures(Wang

etal.2015)2.4.EffectofIn-PlaceAirVoidsonLifeCycleCostTheresultsofthepaststudiesontheeffectofairvoidsonthefatigueandruttingperformanceand service life of asphalt mixtures, as discussed in the previous sections, suggested thatincreasingtheasphaltpavementdensityby1%wouldhavetheeffectofincreasingtheservicelifeoftheasphaltmixturebyconservatively10%.Forillustrationpurposes,thismeansthatanasphaltoverlayconstructedto93%densitymightbeexpectedtolast20years,whiletheexactsame asphalt overlay constructed to 92% density would only be expected to last 18 years.Figure3showsananalysiscomparingthelifecyclecostsofthetwooverlaysoveraperiodof20years. Also, it is assumed that the cost of providing a 1% increase in in-place density isnegligible;therefore,thesameinitialcostisusedforboththealternativesinthisanalysis.

y=-1.1012x+16.575R²=0.32261

02468

101214161820

0 1 2 3 4 5 6 7 8 9 10

TimeAb

erCon

struc[on

(years)

AverageIn-PlaceAirVoids(%)

Tran,Turner,Shambley

7

InAlternative1,theasphaltoverlayconstructedto93%densitywasexpectedtolast20years.Theagencycostfortheinitialoverlayconstruction(i.e.,thepresentvalue(PV)atyear0)wasassumed tobe$1,000,000.Atyear20, theoverlaywouldbe replaced; thus, the salvagevalue(SV)forthisoverlayatyear20wouldbe$0.Thenetpresentvalue(NPV)forAlternative1wouldbe$1,000,000. InAlternative2,theexactsameasphaltoverlaywouldbeconstructedto92%density.Theinitialconstructioncostforthisoverlay(i.e.,PVatyear0)wasassumedtobe$1,000,000.Thisoverlaywasexpectedtolast18years.Atyear18,itwouldbereplacedwithanewone.Thefuturecost(i.e.,futurevalue(FV)atyear18)fortheexactsamereplacementoverlaywouldbe$1,000,000. Itwasalsoexpected to last18years.Thus,atyear20, the remaining life for thereplacementoverlaywouldbe16years,anditssalvagevalueof$889,000wouldbeaproratedshare of the replacement overlay cost. Using a real discount rate of 4%, the replacementoverlay cost (i.e., FV at year 18) and its remaining value (i.e., RV at year 20) would bediscountedbacktoyear0.TheNPVforAlternative2wouldbe$1,088,000.

Figure3.LifeCycleCostAnalysis(per$1,000)

Based on the life cycle cost analysis (LCCA) for the two alternatives, the user agency

wouldseeanNPVcostsavingsof$88,000ona$1,000,000pavingproject(or8.8%)bysimplyincreasing the minimum required density by 1%. This was a conservative estimate withouttakingintoaccountpotentialhigheruserdelaycostsduetotheearlierreplacementoverlayatyear18.2.4.SummaryThe results from thepast studies clearly indicate the adverse effect of increased in-place airvoidsonthefatigueandruttingperformanceofasphaltpavements.A1%decreaseinairvoidswas estimated to improve the fatigue performance of asphalt pavements between 8.2 and43.8%, to improve the rutting resistance by 7.3 to 66.3%, and to extend the service life byconservatively10%. Toillustratetheeffectofin-placeairvoidsonthelifecyclecostofasphaltpavements,anLCCAwasconductedontwoalternativesinwhichtheexactsameasphaltoverlaywouldbeconstructedto93%and92%densities.Usingtheconservative10%increaseinservicelife,the

SV SV

Tran,Turner,Shambley

8

LCCA results showed that the user agency would see an NPV cost saving of $88,000 on a$1,000,000pavingproject(or8.8%)byincreasingtheminimumrequireddensityby1%.3.NEWTECHNOLOGIESFORACHIEVINGHIGHERIN-PLACEDENSITYMostofthein-placedensityrequirementscurrentlyadoptedbySHAsweredeterminedbasedon what levels of in-place density could be achieved in the past using prior constructiontechnologies. Since in-place density has a significant impact on the performance of asphaltpavements, a higher in-place density, which results in a longer pavement service life, mayprovide a significant cost saving to SHAs and the traveling public. Agencies may considerimplementingahigherin-placedensityrequirement,whichcanbeachievablebyfollowingbestpracticesandadoptingnewasphaltpavementtechnologiesandknowledgegainedfromrecentresearch.Thesetechnologiesandknowledgeincludewarmmixasphalt,intelligentcompaction,improved construction joints, and improved agency specifications to incentivize achievinghigher in-place densities. In the following sections, each of these technologies is brieflydiscussed.3.1.WarmMixAsphaltThe term warm mix asphalt (WMA) refers to asphalt mixtures that can be produced attemperatures that are typically 25oF to 90oF lower than standard hot mix asphalt (HMA)productiontemperatures.TheWMAtechnologiescanbeconsideredcompactionaids.Theycanbe used to improve the workability of the asphalt binder, to increase time for mixturecompactionduringnormalpavingoperations,andtoenhancecompactionduringcoldweatherpaving.

ThereareseveralcategoriesofWMAtechnologiesandprocessescommerciallyavailable,includingasphaltfoamingtechnologies,chemicaladditives,organicadditives,andcombinationsof thesetechnologies (Prowelletal.2012).Theasphalt foamingtechnologiesmayusewater-injectionsystems,dampaggregate,orhydrophilicmaterialssuchaszeolitetofoamasphalt.Thefoaming process temporarily expands the binder volume and fluid content, which helpsimprovecoatingandcompaction.Availablechemicaladditivesoftenincludesurfactantstoaidin coating and lubrication of the asphalt binder in the mixture. The organic additives aretypically special types of waxes that cause a decrease in binder viscosity above themeltingpointofthewax.Tousetheseadditives,waxpropertiesshouldbecarefullyselectedbasedontheplanned in-servicepavementtemperaturestoreducetheriskofpermanentdeformation.Approximately 20 WMA technologies are currently marketed in the United States. Moreinformation about each WMA technology commercially available can be found in QualityImprovementPublication125,Warm-MixAsphalt:BestPractices(Prowelletal.2012).

SeveralstudieshavebeenconductedtoevaluatetheabilityofWMAtoprovidesimilarpavementquality as standardHMAmixtures. TheWMAmixturesused in these studieswereproduced at much lower temperatures than the comparable HMA mixtures. Some recentstudiesarediscussedinthefollowingparagraphs.

In2009,Estakhrietal.evaluatedWMAforuse in thestateofTexas.A sectionofOldAustinHighway(Loop368)wasresurfacedbytheTexasDepartmentofTransportation(TxDOT)

Tran,Turner,Shambley

9

using a standard HMA mixture and an Evotherm WMA mixture. Both mixtures met therequirements for a TxDOT Item 341 Type C dense-graded mix, and both binders met therequirements for PG 76-22. The production temperature for the HMAwas 320°F, while theproduction temperature for theWMAwas reduced to 220°F. At the time of placement, theHMAtemperaturewasaround305°FwhiletheWMAtemperaturerangedfrom170°Fto210°F.Bothmixeswerecompactedtoa2-inchmatthicknessusingthesamerollerpatternoverthecourseof threenights.Nucleardensity testson the compactedmat showed that the controlHMAmixtureachievedanaveragein-placedensityof94.2%andtheWMAdensitiesaveragedapproximately93.6%.OverallitwasdeterminedthattheWMAcouldbecompactedtosimilartoslightlyhigherin-placedensitiesastheHMAmixtureatreducedcompactiontemperatures.

A study performed by the National Center for Asphalt Technology (NCAT) in 2010evaluatedthreeWMAtechnologiesusedinafieldtrialsponsoredbythestateofMissouri.TheprojectbeganasanoverlayofanexistingconcretepavementthathadpreviouslybeenoverlaidwithHMAmix.Duringtheoverlayprocess,thecontractornoticedthatthepavementhadpoorinitialsmoothnessresults,whichwerebelievedtobeduetotherubberizedcracksealantusedon the concrete pavement. The high temperatures used for the HMA caused the sealant toexpand,increasingtheroughnessofthenewoverlay.ThecontractorthenapproachedthestateabouttheuseofWMA,whichtheybelievedmightreducethesofteningofthecracksealantandprovide a smoother pavement. The mix placed was a 12.5-mm Superpave mix with 10%reclaimedasphaltpavement (RAP)andapolymermodifiedPG70-22binder.ThethreeWMAtechnologies evaluated were Aspha-min zeolite, Sasobit, and Evotherm ET. A control HMAsectionwasplacedforcomparison.Allsectionswereplacedduringa10-dayperiodinMayof2006.ThecontrolHMAwasproducedat320°F,andtheWMAmixeswereinitiallyproducedat275°F.Onceitwasdeterminedthatin-placedensitiesandconstructabilitywereacceptable,theEvothermandSasobitproductiontemperatureswere loweredto225and240°F,respectively.Field cores were used for evaluating in-place density. Overall, the researchers reported nodifficultyobtainingtherequiredin-placedensityvalueswiththeWMAmixes(Hurley2010).

AsimilarstudyinColoradoevaluatedAdvera,Sasobit,andEvothermDATcomparedtoacontrolmixplacedonI-70,about70mileswestofDenver.Theprojectwasapproximatelyninemiles longandconsistedofa2.5”mill followedbya2.5”overlay.EachWMAwasplacedinaone-miletestsectionadjacenttoanHMAsectionforcomparisonpurposes.TheHMAmixwasproduced at a target temperature of 305°F, while the WMA mixes were producedapproximately50°Flower.In-placedensitiesweremeasuredusinganucleardensitygauge.Theinitial construction data showed that the Advera and Sasobit sections had in-place densitiesthatweresimilarorslightlylowerthantheHMAcontrolsection.TheEvothermin-placedensitywas approximately 1% higher than that of the control section. All of the WMA and HMAsectionseasilymetthetargetinitialin-placedensityrequirements(i.e.,92%to96%ofGmm).Athree-yearevaluationof the test sectionsalso found that theWMAperformanceduring thattimewassimilartotheHMAcontrolsection(Aschenbrener2011).

AnevaluationofpilotprojectsusingWMAinthestateofConnecticutin2010and2011evaluated the effect of warm mix additives on pavements placed in several locationsthroughout the state.Warmmix technologies used in this study included: Sasobit (with andwithout SBS polymer), Evotherm, Advera, SonneWarmix, andmechanical foaming (with andwithoutSBSpolymer).Thepavementswereplacedoverthecourseoftwopavingseasons,with

Tran,Turner,Shambley

10

Connecticut Asphalt Paving Laboratory (CAPLab) personnelmonitoring the placement of thepavement. Reduction in compaction temperature variedwithWMA technology, but inmostcaseswasaround30to50°FlowerthanthetemperaturesatwhichtheHMAcontrolsectionswere placed. The only exceptionwas themix containing Sasobit plus SBS polymer. Thismixinitiallyexperienceddifficultyachievingthetargetin-placedensity(92%ofGmm).Torectifythisissue, theproduction temperaturewas increased.Construction records from theConnecticutprojects showed that, with few exceptions, the mixes containing WMA additives providedcomparablein-placedensitiestothosewithHMA(Zinke2014). In2014,theWashingtonDOTpublishedareportontherehabilitationofapproximately10milesofInterstate90.Forthisproject,threeinchesofexistingpavementweremilledoffandreplacedbyanequaldepthofeitherHMAorWMA(Sasobit).Theprojectuseda12.5-mmmixplacedovera5-dayperiodinJuneof2008.Thesameequipmentandmethodswereusedforbothmixes.HaultimesfortheHMAmixwerearound30–45minutes.HaultimesfortheWMAmixwerearound30minutes.TheonlyissuereportedduringconstructionwasthepresenceofclumpsintheWMAmix.ItwasdeterminedthattheclumpshadformedduringthehaulingoftheWMAandwereduetocoolingofthemix.ItwasrecommendedthattheWMAberemixedin a windrow device to reheat it prior to compaction in order to maintain a consistenttemperatureandtoeliminateanyclumpsthathadformed.Measurementsofdensityshowedthatonaverage,theHMAwascompactedtoadensityof93.5%(standarddeviation=1.58)andtheWMAwas compacted toanaverageof93.7%density (standarddeviation=1.36). Itwasalso noted that theWMA had fewer instances of failing density than the HMA.Overall, theresearchersfeltthattheseresultsindicatedthattheWMAmixcouldbecompactedtothesamelevelofdensityastheHMAatlowercompactiontemperatures(Anderson2014). BasedonareviewofstudiescomparingthecompactionofWMAtothecompactionofHMA, it appears that WMA can be compacted to similar in-place densities at much lowercompaction temperatures. The implications of this include improved in-place densities forprojectsrequiring longerhaultimes(increasedtemperature lossduringtransit)andimprovedin-placedensitiesduringcoldweatherconstruction.3.2.FieldCompactionAsphalt pavement density does not increase linearly with additional compaction; rather, itchanges randomly “due to continuous reorientation of aggregates and the randomness ofaggregate shapes and textures” (Beainy et al. 2014). In general, compaction uniformity andoverall compaction are increased through additional roller passes. There have been manyrecent advances in compaction equipment, and construction practices regarding compactionhave been analyzed much more closely. The use of vibratory rollers, oscillatory rollers, orvibratory pneumatic tire rollers can achieve optimized in-place density when properlyemployed. The rollingpattern, frequency, drum spacing, amplitude, and temperature controls ofvibratoryrollersarecriticaltoachievepropercompactionwithoutcausingaggregatedamagetothe asphalt pavement structure. Rolling patterns should be optimized based on the drum-width-to-panel-width relationship.Vibration frequencyanddrumdiameter shouldbeused todetermine the appropriate rolling speed for double-drum vibratory rollers. Advances in

Tran,Turner,Shambley

11

vibratoryrollermanufacturinghaveledtotheadventofhighfrequencyrollerstoenablefasterrollingspeedswhereavibratoryrollercancompletebreakdownrollingandkeepupwiththepaver,evenwhilecompletingaseven-passpattern.Vibratorydrumspacingshouldbebasedondrumdiametertoensurethesmoothnessofpavementsurfaces.Finally,monitoringthesurfaceHMA pavement temperature zones through the use of real-time infrared sensors can allowoperatorstomonitoridealcompactiontimes(Starry2006).Whetherasphaltmixturesarestiffor tender, breakdown or initial rollers should be used immediately following the paver toensure that the mixture is compacted while hot (Scherocman 2006). By optimizing andautomating these variables, the effectiveness of achieving higher in-place densities withvibratoryrollerscanbegreatlyimproved. The asphalt paving industry has also seen the introduction of new vibratory rollersequippedwith an integrated Intelligent Compaction (IC) system. This systemmay include anonboard computer, Global Positioning System (GPS) based mapping, and optional feedbackcontrols. Itallowsreal-timemonitoringofcompactionandadjustmentsasneededtoachieveoptimum density and uniform coverage. In addition, color-coded mapping provides acontinuous record showing the location of the roller, number of roller passes, andmaterialstiffnessmeasurements. During compaction, the location of the roller, its speed, number ofpasses,andcoveragecanbemonitoredusingtheGPS.Compactionmetersoraccelerometersmountedinthedrummonitortheappliedcompactioneffort,frequency,andmaterialresponse.Some rollers also have temperature instrumentation that allows monitoring of the surfacetemperature of asphalt pavingmaterials.Optional feedback controls can continuously adjusttheforceandfrequencyofthecompactordrum.TheICdisplayinformstheoperatorwhenthedesired level of compaction is reached, eliminating unnecessary passes and making thecompactionprocessmoreefficient.TheICdisplayalsonotifiestheoperatorwhenthedesiredcompaction levelhasnotbeen reached,allowing for furtheranalysisand remedialactions tocorrecttheissue,thuspotentiallyachievingbetterfinaldensityvalues. SeveralstudieshaveattemptedtoevaluateifacorrelationcouldbedevelopedbetweenIC stiffnessmeasurements and in-place densitymeasurements, thereby eliminating the needforotherqualitycontrol/qualityassurance(QC/QA)densitytests(Minchinetal.2001,Maupin2007, Chang et al. 2011, Chang et al. 2014). The results of these studies show that therelationship between IC measurements and in-place density is inconsistent. Several factorswere found to affect the ability of the IC to correlatewith traditionalmeasurements (eithernucleardensitygauge (NDG)or laboratory testingof cores)of in-placedensity.These factorsinclude the stiffness of the underlying layers, temperature of the pavement mat duringcompaction, reliability of the nuclear density gauge (NDG) readings, differences in materialversusmechanicalpropertiesbeingmeasured,andinstrumentation.Forthisreason,itappearsthatICmeasurementsarecurrentlynotagoodcandidateforreplacinglaboratorycoredensitytesting as an acceptance test for HMA paving. The use of IC does, however, show somepotentialasareal-timemeasureofcompactionandmaybeuseful forqualitycontrolandforidentifying locations on the pavingmat thatmay not have achieved the desired compactionlevelandthatmaynotbeidentifiedbyNDGtesting.

In summary, there have been several advances, such as the integrated IC system, incompaction equipment and practices in the past few years. These new technologiesmake iteasier to optimize and automate some compaction parameters, such as rolling pattern,

Tran,Turner,Shambley

12

frequency, drum spacing, amplitude, and temperature control, to achieve higher in-placedensities. In addition, the use of GPS based mapping provides real-time monitoring ofcompactionandacontinuousrecordthatshowsthelocationoftheroller,thenumberofrollerpasses,andmaterialstiffnessmeasurementstoachieveuniformcoverage.WhiletheICsystemhelps improve the compaction process, it is not used in place of traditional laboratory coredensitytestingasanacceptancetestforHMApavingatthistime.3.3.ImprovedConstructionJointsMany asphalt pavement failures can be attributed to insufficient compaction of longitudinaljoints. These failures are primarily affected by the density of the free edge of a lane; thecompactionof thematerial in the joint;andhowwell thehotsideof the joint iscompacted.The construction of longitudinal joints requires precise workmanship to achieve optimalcompaction.Onesequenceofmethodstoachieverequiredcompactionistocompactthefirstlanewith the rolleroverhanging theedgebysix inches, followedbyplacing thesecond layerwithaonetoone-and-a-half inchoverlapofthefirst layerdictatedbytheedgerplateonthepaverscreed.Finally,lanetwoshouldbecompactedfromthehotsidewiththeoutsidetireofarubber tire roller directly on the joint or by a steel drum rollerwith the drum extending sixinchesoverthetopofthejoint(BensonandScherocman2006).

Several technologies have also been implemented in recent years to improveconstruction of longitudinal joints. One of these technologies employed to enhance thecompaction of the free edge is the proper selection and application of tack coat orasphalt/rubberizedsealants (Brown2006).Tackcoatscanminimizemovementof theasphaltmix under the roller; however, the application of tack coats does not necessarily lead toenhancedcompaction.Another recent technologyemployed is theuseof infraredheaters toreheattheedgeofpavementwhenplacinganadjacentlane.Intheory,theheatshouldmaketheasphaltmoreworkableandhelpensureproperbindingbetweenlayers.However,infraredheatersoftenoverheatsomeof thematerialandcausedamagetotheasphaltbinder,or thepavement is under-heated, thereby yielding minimal benefit (Brown 2006). Yet anotherconstructionpracticetoimprovelongitudinal jointcompactionistheuseofataperedjointtofacilitate the transitionand compaction fromapreviouslyplacedasphalt lane toanadjacenthotlane(Brown2006).Thequalityofcompactionandalignmentcanbehardtoachievewithinthis wedge joint; however, this practice also has the added benefit of easing the transitionbetweenlanesforthemotoringpublicduringconstruction.Inthiscase,itmaybebeneficialtocutbacktheunder-compactedmix inthewedgeofthefirst laneas it isspecifiedforairfieldsandbysomestateagencies.

Materialdirectlyinthelongitudinaljointmaylackpropercompactionduetotoosmallofanoverlapbetweenadjacentlayersortoolittleasphaltbeingappliednearthejoint.Theseproblemscanbecausedbypoorscreedalignment,pooraugeroperationonthepaver,orluteoperationsthatremoveandstarvethejoint.Thecompactionofmaterialonthehotsideofthejoint canalsobe significantly impactedby thequantityofmaterialplacedat the joint.Whencompacted,looseasphaltmixoftendecreasesinvolumebyabout20%.Ifthisvolumechangeisnot taken into account, too little asphaltmixmay be placed adjacent to a previously placedlayercausinganinsufficientlevelofin-placedensitytobemet.Oncethenewlyplacedasphalt

Tran,Turner,Shambley

13

iscompactedtothesamethicknessasthepreviouslyplacedasphaltlane,verylittleadditionalcompactioncanbeachievedonthehotsideofthejoint(Brown2006).MoreinformationaboutbestpracticesforconstructionandspecifyingasphaltpavementlongitudinaljointsisavailableonAsphaltInstitute’swebsite(2016).

3.4.AgencySpecificationsAs early as 1989, Hughes et al. recommended a realistic target average value of 93% oftheoretical maximum density with a standard deviation of 1.5%. While some states haveadoptedhighertargetvaluesforin-placedensity,additionalimprovementsinroadwayservicelifecouldberealizedifspecificationsrequiredminorincreasesinthein-placedensities.Alackof universal in-place density guidance has made implementation of standards difficult asconstruction practices, test protocols, and materials have resulted in changes to pavementstructures(Seedsetal.2002).Withmodernmethodsandequipment,theseminorincreasesareveryfeasibleifappropriateguidanceandspecificationsareimplemented.

Itshouldbenotedthatseveralchallengesmayhindertheefficacyofincreasingin-placedensity of asphalt pavements to increase service life in a cost effective manner. First,appropriateprojectselectionmustbeconsidered.Thestructureofthepavementbasemustbeconsidered as a primary criterion for implementing increased in-place density requirements.The use of increased in-place density is most applicable to structural overlays rather thanfunctionaloverlays.Iffunctionaloverlaysareplacedonweakbases,itmaybeverydifficulttoobtain even minimal in-place density requirements. As a part of project selection, thepavement design must consider appropriate lift thicknesses based on nominal maximumaggregatesize(NMAS)andcoarsegradations.Also,themixdesignmustincludeproperasphaltcontentfordesiredin-placeairvoids.Finally,tofullyimplementarequirementforincreasedin-place density, test methods for measuring field compaction must be standardized andacceptancecriteriaandperformanceincentivesmustbeestablishedtoproperlymotivateandrewardconstructioncontractorperformance(Aschenbreneretal.2015).3.4.1.ProjectSelectionWhen choosing a project targeted for higher in-place density, SHAs should evaluate theunderlying base for new projects or asphalt layers during overlay operations. Optimalcompactionmaybedifficulttoachievewhenunderlyinglayersarenotsufficientlysupportive.Evenifoptimalcompactionisachievableinthenewasphaltlayers,theoverallgaininlong-termperformancemaybeminimizedbydefectsintheunderlyinglayers.3.4.2.LiftThicknessBased on studies by Moutier (1980) and further analysis by Zeinali (2014), compactioneffectiveness could be directly improved by increasing lift thickness. Asphalt pavementcompactionisprimarilyimpactedbytheabsolutethicknessofthelayerbeingcompacted,thethickness compared to the NMAS, and the uniformity of the thickness. As the absolute liftthicknessincreases,thetimeavailableforcompactionincreasesduetothethicker liftcooling

Tran,Turner,Shambley

14

moreslowly.Toprovidesufficientliftthicknessfortheaggregateparticlestore-orientandpacktogetherduringthecompactionprocess,theminimumliftthicknessshouldbethreeandfourtimes the nominal maximum aggregate size for fine and coarse dense-graded mixes,respectively.Rutsgreaterthanone-halfinchshouldbemilledbeforeoverlaysareplacedduetothe potential for roller bridging leading to uneven compaction.Whenusing vibratory rollers,“thedepthofpenetrationofthecompactionenergyimparted…dependsontheweightoftherolleraswellastheamplitudeandfrequencyofthevibrations.Foragivensettingofamplitudeandfrequency, thedensityachieveddependsonthethicknessofthematandtheunderlyingpavementlayers”(Beainyetal.2014).3.4.3.MixDesignOptimalfieldperformanceofasphaltmixturescanbeachievedbyusingqualitymaterialsandproperlycontrollingvolumetricproperties.Asphaltbindersshouldbeproperlyselectedbasedon their performance properties related to the conditions under which they are used.Aggregates need to be hard, sound, durable, angular, and properly graded for bestperformance.Finally,theseconstituentmaterialsneedtobeproperlycombinedinmixdesignsto meet specific volumetric requirements. State DOTs have specification requirements toensurethatsatisfactoryqualitymaterialsandmixturesareused.ThespecificationrequirementsaretypicallyadoptedfromtherecommendedSuperpavespecificationsorsometimesadoptedfromrequirementsbasedonexperiencesofthestateDOT.

Whenamixisproducedinthefield,itoftenhasdifferentpropertiesthanthemixdesignconductedinthelaboratory.Someadjustmentsmaybeneeded,butcareshouldbetakenwhenmakingthesefieldadjustmentsastheycanhaveasignificantimpactonthecompactabilityofthemix. During construction, it is essential that thegradation,binder content, andvolumetricproperties, suchasairvoidsandvoids inmineralaggregate,becloselycontrolledsothat thevariability is low.Most stateDOTshave construction tolerance requirementsandpay factorsrelatedtotheseproperties.Forexample,laboratoryairvoidsaregenerallycontrolledbetween3and5%fordense-gradedmixturesandthedesignistypicallyperformedat4%airvoids.Iftheairvoidsarealittlehigh,longtermdurabilityofthemixistypicallyreduced.Iftheairvoidsarea little low, bleeding (and possibly rutting) in the asphalt mixture may occur. Thus, thegradation,bindercontent,andvolumetricpropertiesmustbeuniformduringconstructionforbestfieldperformance.Thesepropertiescanalsoinfluencethecompactabilityofthemix.3.4.4.CriteriaTypical quality control (QC) and acceptance specifications rely on acceptance testing,comparisontesting,qualitylevelanalysis,andpayfactordeterminations.Basedonexperiencefrom the Port Authority of New York and New Jersey (PANYNJ), even with specifications ofspecifictypesoflongitudinaljoints,manyprojectshadlowdensityinthesejoints.PANYNJhasimplemented an end-result specification and a specific joint type mandate to incentivizeachievementofcompactioncriteriaregardlessofconstructionmethod(Bognacki2006).Agencyspecificationsmustuseappropriatemeasures forsettingrequirements for in-placepavement

Tran,Turner,Shambley

15

performance. In the past, state agencies used the density of laboratory samples for targetdensity, but this had the potential for greater variation in field compaction (Santucci 1998).Recently, most agencies have compared field compaction with the maximum theoreticaldensity.3.4.5.PerformanceIncentivesMany SHAs have developed performance incentives based on various asphalt acceptanceproperties.Thesepropertiesshouldincludein-placedensity,asphaltcontent,andliftthickness(Santucci 1998). Construction performance incentives should be established based on theeconomic impact to the highway agency. In general, inferior performance penalties andsuperiorperformancebonusesshouldbebasedonthecosttotheagencyduetomorefrequentor less frequent anticipated rehabilitation requirements (Santucci 1998). When the ArizonaDepartmentofTransportation(ADOT)implementedatrueincentivebasedspecificationin1990,average in-place air voids decreased from 8.5% to 7.5%. This specificationwas based on in-placeairvoidsinsteadofapercentageofin-placedensity.TheidealADOTspecificationwouldyield an in-place air void target of 7%. The 1% increase in in-place compactionwas a directresultofimplementationofthecompactionincentive(Nodes2006).Furtherimplementationofspecific construction performance incentives should encourage attainment of enhancedcompaction.4.SUMMARYThis literature review was conducted to provide information to support the FHWA AsphaltPavementTechnologyProgramstrategicdirectiononextendingpavementservicelifethroughenhancedfieldcompaction.Resultsfromthepaststudiesclearlyindicatetheadverseeffectofincreasedin-placeairvoidsonthefatigueandruttingperformanceofasphaltpavements.A1%decreaseinairvoidswasestimatedtoimprovethefatigueperformanceofasphaltpavementsbetween8.2and43.8%,to improvetheruttingresistanceby7.3to66.3%,andtoextendtheservice life by conservatively 10%. Based on these results, an LCCA was conducted on twoalternatives inwhich the exact same asphalt overlaywould be constructed to 93% and 92%densitiestoillustratetheeffectofin-placeairvoidsonthelifecyclecostofasphaltpavements.The LCCA results show that theuser agencywould see anNPV cost savingsof $88,000on a$1,000,000pavingproject(or8.8%)byincreasingtheminimumrequireddensityby1%.

Due to its significant effect, the cost of providing increased in-place density can besignificantly less thantheoperation,maintenance,androadusercostsavingsrealizedduetoextendedservicelifeofthepavements.Ina2007AASHTOsurveyofstateagencies’targetsforfieldcompaction,themajorityofrespondingstateshadacompactiontargetof92%ofGmm,butoverone-thirdoftherespondingagencieshadcompactiontargetslessthan92%.Mostofthecurrentin-placedensityrequirementsadoptedbystatesweredeterminedbasedonthelevelsof in-place density that could be achieved in the past using prior construction technologies.Since in-place density has a significant impact on the performance of asphalt pavements,agencies may consider implementing a higher in-place density requirement that can beachievableby followingbest practices andadoptingnewasphalt pavement technologies and

Tran,Turner,Shambley

16

knowledge gained from recent research. Some of these technologies and knowledge werebrieflydiscussed in this report, includingwarmmixasphalt, intelligentcompaction, improvedconstructionjoints,andimprovedagencyspecificationstoincentivizeachievinghigherin-placedensities.

Tran,Turner,Shambley

17

REFERENCESAnderson,Keith;Russell,Mark;Uhlmeyer,Jeff;Weston,Jim;Roseburg,Jerry;Mahoney,Jim;

andDeVol,Joe.(2014)“WarmMixAsphaltFinalReport.”ExperimentalFeaturesWA08-01,WashingtonDepartmentofTransportation.

Aschenbrener,Tim;Schielbel,Bill;West,Randy.(2011)“ThreeYearEvaluationofthe

ColoradoDepartmentofTransportation’sWarmMixAsphaltExperimentalFeatureonI-70atSilverthorne,Colorado.”ColoradoDepartmentofTransportation.

AsphaltInstitute.BestPracticesforConstructionandSpecifyingHMALongitudinalJoints.

http://www.asphaltinstitute.org/best-practices-for-construction-and-specifying-hma-longitudinal-joints/.AccessedMarch8,2016.

Beainy,Fares;Singh,Dharamveer;Cummuri,Sesh;andZaman,Musharraf.(2014)“Laboratory

andFieldStudyofCompactionQualityofanAsphaltPavement.”InternationalJournalofPavementResearchandTechnology.7(5),317-323.

Benson,JoshuaS.andScherocman,JamesA.(2006)“ConstructionofDurableLongitudinal

Joints.”TransportationResearchCircularE-C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,120-138.

Blankenship,P.andAnderson,R.M.(2010)“LaboratoryInvestigationofHMAModulus,Flow

NumberandFlexuralFatigueonSamplesofVaryingDensity.”JournaloftheAssociationofAsphaltPavingTechnologists,Vol.79,pp.497-518.

Bognacki,CasimirJ.(2006)“ConstructingLongitudinalJointswithLowerVoids.”Transportation

ResearchCircularE-C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,96-99.

Bonaquist,Raymond.(2011)“MixDesignPracticesforWarmMixAsphalt.”National

CooperativeHighwayResearchProgram(NCHRP)Report691:TransportationResearchBoard.

Brown,E.Ray.(2006)“BasicsofLongitudinalJointConstruction.”TransportationResearch

CircularE-C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,86-95.

Chang,George;Xu,Qinwu;Rutledge,Jennifer;Horan,Bob;Michael,Larry;White,David;and

Vennapusa,Pavana.(2011)“AcceleratedImplementationofIntelligentCompactionforEmbankmentSubgradeSoils,AggregateBase,andAsphaltPavementMaterials.”FHWA-IF-12-002,NationalTechnicalInformationService.

Tran,Turner,Shambley

18

Chang,George;Xu,Qinwu;Rutledge,Jennifer;andGerber,Sabrina.(2014)“AStudyonIntelligentCompactionandIn-PlaceDensity.”FHWA-HIF-14-017,NationalTechnicalInformationService.

Christensen,DonaldW.andBonaquist,RamonF.(2006)“VolumetricRequirementsof

SuperpaveMixDesign.”NationalCooperativeHighwayResearchProgram(NCHRP)Report567:TransportationResearchBoard.

DelPilarVivar,ElianaandHaddock,JohnE.(2006)“HMAPavementPerformanceand

Durability.”JointTransportationResearchProgram,FHWA/IN.JTRP-2005/14.Epps,JonA.;Hand,Adam;Seeds,Steve;Schulz,Todd;Alavi,Sirous;Ashmore,Colin;Monismith,

CarlL.;Deacon,JohnA.;Harvey,JohnT.;andLeahy,Rita.(2002)“RecommendedPerformance-RelatedSpecificationforHot-MixAsphaltConstruction:ResultsoftheWestrackProject.”NationalCooperativeHighwayResearchProgram(NCHRP)Report455:TransportationResearchBoard.

Epps,J.A.andMonismith,C.L.(1969)“InfluenceofMixtureVariablesontheFlexuralFatigue

PropertiesofAsphaltConcrete.”Proceedings,AssociationofAsphaltPavingTechnologists,Vol.38,pp.423-464.

Epps,J.A.andMonismith,C.L.(1971)“FatigueofAsphaltConcreteMixtures—Summaryof

ExistingInformation.”ASTMSpecialTechnicalPublication508,Philadelphia,PA:ASTM.Epps,J.A.;Seeds,S.B.;Monismith,C.L.;Harvey,J.T.;Alavi,S.H.;Scholz,T.V.;Leahy,R.B.;and

Ashmore,S.C.(2002)“RecommendedPerformance-RelatedSpecificationsforHot-MixAsphaltConstruction:ResultsoftheWestTrackProject.”NCHRPReport455,NationalCooperativeHighwayResearchProgram,TransportationResearchBoard,Washington,D.C.

Estakhri,Cindy;Cao,Rongji;Alvarez-Lugo,Allex;andButton,Joe.(2009)“Production,

Placement,andPerformanceEvaluationofWarmMixAsphaltinTexas.”GeotechnicalSpecialPublicationNo.193,AmericanSocietyofCivilEngineers.

Finn,F.N.,andEpps,J.A.(1980)“CompactionofHotMixAsphaltConcrete.”ResearchReport

214-21,CollegeStation,TX:TexasTransportationInstitute,TexasA&MUniversity.Fisher,Jonathan;Graves,Clark;Blankenship,Phillip,Hakimzadeh-Khoee,Salman;andAnderson,

R.Michael.(2010)“FactorsAffectingAsphaltPavementDensityandtheEffectonLongTermPavementPerformance.”KentuckyTransportationCenterResearchReportKTC-10-05/RSF14-05-1F,UniversityofKentucky,unpublishedpaper.

Harvey,J.T.andTsai,B.W.(1996)“EffectsofAsphaltContentandAirVoidContentonMix

FatigueandStiffness.”TRR1543,TransportationResearchBoard,Washington,D.C.

Tran,Turner,Shambley

19

Hughes,CharlesS.(1989)“CompactionofAsphaltPavement.”NCHRPSynthesisofHighway

Practice152,NationalCooperativeHighwayResearchProgram,TransportationResearchBoard,Washington,D.C.

Hurley,Graham;Prowell,Brian;andKvasnak,Andrea.(2010)“MissouriFieldTrialofWarmMix

AsphaltTechnologies:ConstructionSummary.”NCATReportNo.10-02,NationalCenterforAsphaltTechnology.

Linden,R.N.;Mahoney,J.P.;andJackson,N.C.(1988)“EffectofCom-pactiononAsphalt

ConcretePerformance.”TransportationResearchRecord1217,TransportationResearchBoard,NationalResearchCouncil,Washington,pp.20–28.

Maupin,G.W.(2007).“PreliminaryFieldInvestigationofIntelligentCompactionofHot-Mix

Asphalt.”VTRC-08-R7,VirginiaDepartmentofTransportation.Minchin,R.;Thomas,H.R.;andSwanson,David.(2001)“TheoryBehindVibratoryBased

OnboardAsphaltDensityMeasuringSystem.”TransportationResearchRecord1761,TransportationResearchBoard,NationalResearchCouncil,Washington,D.C.

Nodes,JulieE.(2006)“ImpactofIncentivesonIn-PlaceAirVoids.”TransportationResearch

CircularE-C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,163-168.

Nose,Yukinori.(2006)“VibratoryPneumaticTireRoller.”TransportationResearchCircularE-

C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,54-68.

PIARCTechnicalCommittee4.3RoadPavements.(2009)“LongLifePavementsandSuccess

Stories.”WorldRoadAssociation(PIARC).Prowell,B.;Hurley,G.;andFrank,B.(2012)“WarmMixAsphalt:BestPractices.”QIP125,3rd

Edition.NationalAsphaltPavementAssociation(NAPA).Puangchit,P.;Hicks,R.G.;Wilson,J.E.;andBell,C.A.(1982)“ImpactofVari-ationinMaterial

PropertiesonAsphaltPavementLife.”FHWA-OR-82-3,OregonDOT.Santucci,Larry.(1998)“TheRoleofCompactionintheFatigueResistanceofAsphalt

Pavements.”InstituteofTransportationStudies,TechnologyTransferProgram.Scherocman,JamesA.(2006)“CompactionofStiffandTenderAsphaltConcreteMixes.”

TransportationResearchCircularE-C105.FactorsAffectingCompactionofAsphaltPavements.TransportationResearchBoard,69-83.

Tran,Turner,Shambley

20

Seeds,StephenB.;Hicks,GaryR.;Elkins,GaryE.;Zhou,Haiping;andScholz,ToddV.(2002)“LTTPDataAnalysis:Significanceof‘As-Constructed’ACAirVoidstoPavementPerformance.”AppliedPavementTechnology,Inc.,NCHRPProject20-50(14),FinalReport.

Starry,DaleW.,Jr.(2006)“VibratoryRollers.”TransportationResearchCircularE-C105.Factors

AffectingCompactionofAsphaltPavements.TransportationResearchBoard,36-48.Verstraeten,J.(1995)“BituminousMaterialswithaHighResistancetoFlowRutting.”

PermanentInternationalAssociationofRoadCongresses(PIARC).Wang,H.;Wang,Z.;Bennert,T.;andWeed,R.(2015)“HMAPayAdjustment.”ReportFHWANJ-

2015-007.NewJerseyDepartmentofTransportation.Zeinali,Alireza;Blankenship,PhillipB.;andMahboub,KamyarC.(2014)“Evaluationofthe

EffectofDensityonAsphaltPavementDurabilitythroughPerformanceTesting.”TransportationResearchBoard93rdAnnualMeeting,TransportationResearchBoard.

Zinke,Scott;Mahones,James;andMorison,Kelly.(2014)“ConnecticutWarmMixAsphalt

(WMA)PilotProjects.”ResearchProjectSPR-2269,ConnecticutDepartmentofTransportation.