National Trends In Pavement Design - Pavement Preservation

National Trends inPavement DesignSoutheastern StatesPavement Management AssociationPavement Management and DesignMay 11, 2009New Orleans, LouisianaGary CrawfordFederal Highway Administrationgary.crawford@dot.gov

First Engineered RoadConcrete(Stone & othermaterial withLime)Squared StonesFine Dry Soil (Well-compacted)2

Roman Roads Roman Empire built over 3000 miles ofroads in Britain alone by 200 A.D. Used ditches to aid in drainage Varied thickness over weaker soils Indicated that Romans had someunderstanding of basic soil mechanics3

First Modern Roads 1764 France (Tresaguet) Labor costs too high; smaller stones, thinner sections4

First Modern RoadsUse of Tar and Asphalt 1830’s USA - England (McAdam) Impervious surface; asphalt/tar mixed hot; sandadded to fill voids5

First Concrete Pavement 1891 George Bartholomew built firstconcrete pavement in Bellefontaine,Ohio 8-ft.-wide strip of Main Street6

First Concrete PavementRensselaer County, New York, 1908 -farm-to-market thoroughfares.Early dump truck hauls materials to paving site.7

Concrete Design History 1916 – Typical Pavements 5 – 9”1906/1918 – Patent for skewed jts1920 – First CRCP in MD1970 – 15 states had built CRCP sections 1922 - No jointed pavements, thickenedcenter section to prevent cracking 1925-45 – Exp jts 50 – 120 ft, contraction jts15-60 ft8

Concrete Desing History 1940 to 1950 – Use of random jts.1987 – 15 states using random jts.1975 – 18 states using skewed jts.1987 – 25 states using skewed jts.9

First Asphalt Pavement Paris – 1854 Used natural rock asphalt,i.e., limestone rock impregnated withasphalt. Pavement provided a quiet, easilycleaned surface, but the skidresistance was very low in wet weather10

First Asphalt Pavementin US 1870, Design by Edmund DeSmedt Sand Mix placed in front of City Hall inNewark, New Jersey11

First Asphalt Pavement First asphalt concrete specificationsappeared in the US in the 1890’s. Concurrently, coal tar/aggregatemixtures were being used in Europe. First hot-mix asphalt plants weredeveloped in the late 1920’s12

Early Construction First modern asphalt paver introducedmid-1950’s Prior to 1950’s, asphalt was placed byform-riding finishers similar to PCC13

Flexible PavementsConventionalDeep AsphaltAsphaltAsphaltUnboundBaseAsphalt ralSubgradeNaturalSubgrade14

AASHO Road Test(late 1950’s)AASHO, 1961)15

AASHO Road TestAchievements Serviceability concept Traffic damage factors Structural numberconcept Empirical Process Simplified PavementDesign(AASHO, 1961)16

1950s Vehicle Loads.(AASHO, 1961)17

History of the Current AASHTOPavement Design Guide Empirical design methodology based on AASHORoad Test in the late 1950’s Several versions: 1961 (Interim Guide), 1972 1986 version refined material characterization 1993 revised version More on rehabilitation More consistency between flexible, rigid designs Current version for flexible design procedures 1998 Supplemental Guide for rigid pavement design18

Why is pavement designSO HARD ? Predicting thefuture Indecision ondesign Political influence Funds Materials Construction quality19

Design Methodologies c20

Pavement ThicknessAASHO LoadingLimitationsCurrent design trafficis far beyond roadtest limitsDataLimits(AASHORoadTest) 2 MillionCurrentDesigns 100 MillionAxle Load Repetitions21

The Concept of MechanisticDesign Fundamental engineering theories andmaterial properties are used to calculatecritical strains in the pavement due totraffic loadδεt22

Major Benefits of MEPDG overAASHTO More defensible design procedure More realistic pavement thickness athigh ESAL’s Cost Savings Ability to integrate with PRS, LCCA,warranty projects Not much different than AASHTO atlower ESAL ranges ( 500,000)23

Key Advantageof M-E Design“Comprehensive” designprocedure:Not Just Thickness!M-E models directly consider trueeffects and interactions of inputs onstructural distress and ride quality.Design optimization possible whereall distress types are minimized!24

Challenges of MEPDGMethod Many inputs required Availability of models representing localconditions Availability of required materialproperties Predicting future traffic loads andclimate What can not be addressed?25

Why move to a ME baseddesign procedure ? EconomicsDeficiencies in current proceduresPolitical climateMethod based on sound engineeringprinciples26

Everyone is important in theMEPDG analysis process27

AASHTO MEPDG InterimGuide Balloted successfully in 2007 AASHTO currently devAASHTOWare version28

What’s New in MEPDG Topics to be covered CapabilitiesReliabilityCompare AASHTO Guide to MEPDGInputs ClimateTraffic ACPPCCP Unbound materials Calibration Testing29

Capabilities Wide range of pavement structures New Rehabilitated Explicit treatment of major factors Traffic – Over-weight trucksClimate – Site specific and over timeMaterials – New and differentSupport – Foundation and existingpavement30

Capabilities Models to predict change in distressover time User establishes acceptance criteria Distresses and smoothness31

What’s New in DesignReliability? Different than AASHTO 1986/93 Based on predicted distress and IRI User selects reliability levels andperformance criteria for distress and IRI32

M-E Design amageResponseTimeDamageAccumulationDistress33

What’s New About theDesign Guide?1993 sMEPDGInputs for EICMThermal PropertiesWind SpeedAir TemperatureWater Table DepthSun RadiationPrecipitation34

What’s New About theDesign Guide?1993 GuideESALsMEPDGAxle Load SpectraTraffic Truck Equivalency Truck SpeedFactorsGear/AxleConfigurationAxle/Tire SpacingTire PressureTraffic WanderMonthly, DailyDistribution Factors35

What’s New About theDesign Guide?1993 Guide MEPDGFoundationResilientModulus“k” valuesUniversal non-linearResilient modulusModel36

What’s New in FlexiblePavement Design?1993 GuideMEPDGLayer Coefficient Dynamic Modulus HMASPTMaterials(Level 1 &Master curves)ResilientModulus (68 F)Poisson’s ratio37

What’s New in RigidPavement Design ?1993 GuidePCCMaterialsModulus, FlexStrength, TensileStrength(28-day)MEPDGModulus of Elasticity(7, 14, 28 & 90 day)Flexural, TensileStrengthPoisson’s ratioPCC Thermal PropsDrying ShrinkageCoefficient of ThermalExpansion38

What’s New and Different1993 GuideOutputsStructural NumberRigid PavementThicknessMEPDGTime Series Distressand SmoothnessPrediction39

MEPDG Outputs –FlexibleThermalCrackingFatigueCrackingRut DepthLongitudinalCrackingIRI40

MEPDG Outputs - RigidTransverseCrackingJoint FaultingPunchoutIRI

MEPDG is an AnalysisProgramTrial design

What’s New and Different1993 GuideInputLevelsMEPDGLevel ThreeSingle ValueLevel TwoLevel One43

Design Inputs - HierarchicalLevelsInput levels can be mixed andmatchedDamage calculations are exactlythe same regardless of designinput level44

Climatic Data45

Climatic InputsInputLevel123 Identify weather station Pick from one of 800 site Create virtual by averaging surrounding orsimilar sites Create EICM file Depth to water table46

Climate Model (EICM)Asphalt DesignAdjustments: Unbound Resilient modulus Moisture content AC Hourly temperature profile Thermal cracking Rutting47

Hourly Temperature Profilefor AC 6/14/976/14/986/14/99TIMEDepth 0 in.Depth 3 in.Depth 6 in.48

EnvironmentConcrete Design EICM used to predict Hourly temperature profile Monthly moisture gradient49

Concrete Slab Temperatureand Moisture GradientsCurlingWarpingSlab wetter on topSlab dryer on top50

TRAFFIC INPUTS51

Traffic Hierarchical InputLevels Level 3 – AADT & % trucks with TTCClassification GroupLevel 2 – AADTT withRegional/Statewide AVC & WIM dataLevels 1 – AADTT with site specificAVC & WIM data52

Traffic Module Inputs - OverviewInput ParametersInput Level1 2 3Inputs Required to Compute AADTTAADTT for Base YearAADT and Percent Trucks for Base YearDirectional Distribution FactorLane Distribution Factor Truck Traffic Volume Adjustment FactorsTruck Distribution Factors - Base YearTruck Traffic Classification (TTC) FactorAxle Load Distribution FactorsMonthly Distribution Factors 53

Traffic Module Inputs - OverviewInput ParametersHourly Distribution FactorsTruck Traffic Growth Function/FactorInput Level1 2 3 Axle Load Distribution FactorsAxle Load Distribution Factors General Traffic InformationNo. of Axle Types per Truck Class Axle SpacingAxle Load GroupsTire Spacing/Axle ConfigurationTire Pressure 54

Lateral Truck Traffic WanderInputLevel1 2 3 Mean wheel location Traffic wander standard deviation Design lane width55

Unbound Materials(Aggregates and Subgrade) Resilient Modulus Level 3 Defaults Level 2 Correlations Level 1 Materials specific testing Variability None Seasonal Values EICM56

Unbound Material GeneralPropertiesInputLevel1 2 3 Unbound Material Type - select from list of: AASHTO Classification (AASHTO M 145) Unified Soil Classification System(ASTM D 2487) Other (e.g. crushed stone, cold recycled AC) Layer Thickness: thickness of the layer ininches57

For PCCPSubgrade resilient modulus isconverted to a k-value thatproduces equivalent surfacedeflections for each month in year58

ASPHALTMATERIALPROPERTY ANDDESIGN INPUTS59

Mix Dynamic Modulus Level 3 – Predictive equation and binder class 2 – Predictive equation and binder tests 1 – Laboratory mix tests Predictive equation Gradation Air Voids Asphalt content Binder information60

PCC MATERIAL PROPERTYAND DESIGN INPUTS61

CRCP Design Features - InputsInput Reinforcement Level1 Bar diameter Spacing Percent steel Base properties Base type Erodibility Base/slab friction coefficient Crack spacing (optional)2 3 62

JPCP Design Features - Inputs Joint DetailsInputLevel Joint spacing Sealant type Dowel diameter and spacing1 2 3 Edge Support Shoulder type and LTE Widened slab Base properties Base type Interface type, i.e. bonded or unbonded Erodibility63

Three Step Process Verification – assuring generalreasonableness Calibration – minimize differencebetween predicted and observeddistress Validation – confirm accuracy ofcalibrated model64

Performance Verification Procedure evaluates the trial design todetermine if it meets the desiredperformance criteria at individually setreliability levelsTrial design65

MEPDG Guide Calibration Done with national LTPP data Default values also from LTPP Confirm/change national defaults66

Implementation –Calibration Requires extensive experimentalstudies, including: Field testing programs Laboratory testing Data analysis67

Field Testing Programs Select test sites in each agency (LTPPand others) that includes range of: Climate types and areas in the agency Pavement types AC (all types), PCC (all types) Types of overlays and rehab. Base and subgrade types Joint types in PCCTraffic characteristics Typical preservation techniques 68

Field Testing Programs, Cont. Obtain pavement performance data Distress surveysFWD and core testingPavement profileMaterial related distresses Determine in-place material properties69

Data AnalysisLocal calibration will involverecalibrating the distressmodels using data collectedfrom the selected local sections70

Actual Field PerformanceRegional/LocalCalibration Processβsβs Agency Calibration FactorCalibrated National Predicted Performance71

MEPDG Survey Conducted in2007 52 responses, 50 states plus DC & PR 65 questions on: Current Design ProceduresMEPDG KnowledgeImplementation ActivitiesPartnering ActivitiesTraining Needs72

Asphalt Design ProcedureAASHTO 197263%13%8%4%AASHTO 1993State DesignProcedureAASHTO/StateDesign ProcedureOther12%73

Concrete Design ProcedureAASHTO 197212%AASHTO 198119%10%AASHTO 1993AASHTO 199817%36%4%2%State DesignProcedureAASHTO/StateDesign ProcedureOther74

How does actual performancecompare to design life?12%Less than design life33%Similar to design lifeMore than designlifeDon't Know45%10%75

Does SHA Use or Plan to UseMEPDG?N0 -12YES - 40AlaskaHawaii76

EvolutionThe MEPDG is not perfect .BUT;The MEPDG provides a reasonable andstructured platform for continuousimprovement.77