Beta Testing and Validation of HMA PRSonlinepubs.trb.org/onlinepubs/webinars/HMASlides.pdf · Validation of HMA PRS Performance Related ... §Final Report and presentation of results

NCHRP 9-22 Beta Testing and Validation of HMA PRS 1

Beta Testing and Validation of HMA PRS

April 27, 2009

NCHRP 9-22 WEBINAR


Introduction to Project

§ Project Team§ Specification Developments

• Materials and Method Specifications• Performance Based Specifications• Performance Related Specifications

§ NCHRP 9-22• Project history• Objectives


Project Team

§ Fugro Consultants, Inc., Austin, TX• Prime contractor and project management• Manual of Practice of the HMA PRS

§ Arizona State University• Technical program development• Spreadsheet solution for the three MEPDG

distresses

§ Transmetric America Inc.• C++ Software Development lead• Users Manual


Specification Development

§ Materials and Method Specifications• Direct the contractor to use specified

materials in definite proportions and specific types of equipment and methods to complete the work

• Each step is usually directed by a representative of the agency

NCHRP 9NCHRP 9 --22 Beta Testing and 22 Beta Testing and Validation of HMA PRSValidation of HMA PRS


§ Quality Assurance Specifications• Materials and construction (M&C) variables

used in process control and acceptance• HMA volumetric properties (air voids, asphalt

content, aggregate gradation, etc. ) are assumed to relate to performance

• These properties referred to as Acceptance Quality Characteristics (AQCs)

• Primarily tied to performance through intuition, engineering judgment, or both

5



• The AQCs of the mixture and the pavement are defined in terms of performance using measured fundamental engineering properties and prediction models

• Specifications describe how the finished product should perform over time

• Has had limited acceptance in the industry because of timely acceptance testing issues and with no consensus as to the performance time period

§Performance Based Specifications (PBS)

6



• Similar to Performance Based Specifications• AQCs are correlated to fundamental

engineering characteristics and in turn to performance through prediction models

• Connection to performance through valid empirical or mechanistic prediction models

• Relationships between material properties such as HMA layer thickness, strength, and pavement distresses to fundamental engineering properties

§Performance Related Specification (PRS)

7



• Performance of the pavement predicted based on the as-designed and the as-constructed properties.

• Difference in predicted performance between the as-designed and as-constructed pavement is the basis calculating incentive/disincentive for contractors.

§Performance Related Specification (PRS) con’t

8


Performance Related Specifications

§ In 1994 the FHWA funded the design, construction, and application of loads on a test track project, Westrack, that provided the basis for the development of a prototype PRS § Quantified the effects of variation in

materials and construction (M&C) properties on overall pavement performance

9


Performance Related Specifications

§ The project team developed empirical relationships and formed the basis for the performance prediction models used in the supporting software, HMA Spec§ Limitations were identified in the HMA

Spec software and project NCHRP 9-22 was initiated

10


NCHRP 9-22 Project History

§ Awarded to Fugro Consultants LP (Fugro), October 2000 § Project Scope

• Evaluate and refine the HMA PRS and supporting HMA Spec software

• Calibrate and validate the Level I and II performance models

• Develop a training course to assist the implementation of the HMA Spec software

11


NCHRP 9-22 Project History

§§ Scope modified in April 2001 to include Scope modified in April 2001 to include the MEPDG Prediction Models the MEPDG Prediction Models §§ Use the MechanisticUse the Mechanistic--Empirical (MEmpirical (M--E) E)

Design Guide Software produced in Design Guide Software produced in NCHRP Project 1NCHRP Project 1--37A as the 37A as the ““engineengine”” for for performance prediction models in the performance prediction models in the HMA PRS. HMA PRS.

12


NCHRP 9NCHRP 9--22 Project History22 Project History

§ Project “on hold” awaiting development of the MEPDG models§ MEPDG software “run times” for

specification simulation were unacceptable§ Project “on hold” to resolve run time

problem

13


NCHRP 9NCHRP 9--22 Project History22 Project History

§ Scope revised in 2005 to develop spreadsheet solutions § Excel® based distress prediction model

simulation runs were instantaneous§ Since Excel® software is “version

dependent”, decision made in 2006 to convert to C++ software

14


NCHRP 9NCHRP 9--22 Project Objectives22 Project Objectives

§ Development of Spreadsheet Solutions (rutting, fatigue, thermal cracking)

§ Preliminary integration of spreadsheet solutions into HMA Spec

§ Rename software to Quality Related Specification (QRSS)

15


NCHRP 9-22 Project Objectives, con’t

§ Prepare QRSS alpha version

§ Prepare and test QRSS beta version

§ Final Report and presentation of results to the NCHRP Panel

16

NCHRP 9-22 Beta Testing and Validation of HMA PRSNCHRP 9NCHRP 9 --22 Beta Testing and 22 Beta Testing and Validation of HMA PRSValidation of HMA PRS

17


Introduction and OverviewDr. M. W. Witczak

Professor, Civil and Environmental Engineering

Program DemonstrationDr.M.El-Basyouny

Assistant Research Professor

TRB Webinar SeriesTRB Webinar Series

A Rational Methodology to Assess Performance A Rational Methodology to Assess Performance Related Pay Factors for Asphalt PavementsRelated Pay Factors for Asphalt Pavements


STUDY PURPOSESTUDY PURPOSE

Ultimate Goal in Pavement Technology isUltimate Goal in Pavement Technology is……

ME PDGPavement

DesignNCHRP 1-37A/NCHRP 1-40

AsphaltMix

DesignNCHRP 9-19

(SPT)

SimultaneousAsphalt Mix and

Structural PavementDesign

HMA Probabilistic PRS Methodology: NCHRP 9-22


MEME--PDG Version 1.0 SolutionPDG Version 1.0 Solution--Necessity for Necessity for Rapid SolutionRapid Solution

No of Variables

Comp Hrs Comp Days Comp Yrs

6 32 1.33 -

10 512 21 .06

20 524,288 21,845 59.9

50 5.63xE^14 2.35xE^13 6.43xE^10

64 9.2xE^18 3.84xE^17 1.05xE^15


General ApproachGeneral Approach

qqDevelop Closed Form Solutions for Major MDevelop Closed Form Solutions for Major M--E E PDG HMAC Distress Types PDG HMAC Distress Types ØØHMAC RuttingHMAC RuttingØØHMAC Fatigue (Alligator)HMAC Fatigue (Alligator)ØØHMAC Thermal FractureHMAC Thermal Fracture

qqSolutions Need to Be:Solutions Need to Be:ØØRapid (Seconds or Minutes)Rapid (Seconds or Minutes)ØØHighly Accurate When Compared to MHighly Accurate When Compared to M--E PDG E PDG

SolutionsSolutionsØØSimple to ImplementSimple to Implement

qqCode MethodologyCode MethodologyØØSpreadsheet to C++ Spreadsheet to C++


General ApproachGeneral Approach

qqRun simulation of MRun simulation of M--E PDG E PDG ØØMatrix of runs for all key variablesMatrix of runs for all key variables

qqDevelop accurate statistical closed form Develop accurate statistical closed form predicted modelpredicted modelØØDamage MDamage M--E PDG E PDG ≈≈ Damage PMDamage PM

qqDevelop spreadsheet (Excel) system solutionDevelop spreadsheet (Excel) system solutionØØDeterministic Job Mix assessment vs. Plant Mix assessment.Deterministic Job Mix assessment vs. Plant Mix assessment.

qq Incorporate Probabilistic solutionIncorporate Probabilistic solutionØØProbabilistic Job Mix assessment vs. Plant Mix assessment Probabilistic Job Mix assessment vs. Plant Mix assessment

(Rosenblueth, Monte Carlo simulation, Taylor series).(Rosenblueth, Monte Carlo simulation, Taylor series).

qqDevelop C++ Develop C++ user friendly programuser friendly programØØ Insure that spreadsheet deterministic yields accurate solution tInsure that spreadsheet deterministic yields accurate solution t o Mo M--

E PDG. E PDG.


General Approach (ContGeneral Approach (Cont’’d)d)

qq Traffic is represented by ESALs. Traffic is represented by ESALs. qq No Seasonal changes for unsaturated E No Seasonal changes for unsaturated E

(base/subbase/subgrade).(base/subbase/subgrade).qq PLD: Service Life Difference between Job Mix PLD: Service Life Difference between Job Mix

design (Lab) AC mix and Field (Indesign (Lab) AC mix and Field (In--situ) AC mixsitu) AC mixqq Predicted Life Difference is the basis of the Predicted Life Difference is the basis of the

Pay FactorsPay Factorsqq Deterministic and Probabilistic Analysis.Deterministic and Probabilistic Analysis.qq Final Product :User Friendly C++ Code (Mimics Final Product :User Friendly C++ Code (Mimics

AASHTO MEPDG)AASHTO MEPDG)


Comparison of Rut Depth between MEPDG 1.0 Comparison of Rut Depth between MEPDG 1.0 and ASU Rutting Modeland ASU Rutting Model

y = 0.997xR² = 0.996

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Rut

Dep

th fr

om N

ew A

SU M

odel

(in)

MEPDG 1.0 Rut Depth (in)


Comparison of Fatigue Damage between MEPDG Comparison of Fatigue Damage between MEPDG 1.0 and ASU Fatigue Model1.0 and ASU Fatigue Model

y = 0.998xR² = 0.998

-8

-6

-4

-2

0

2

4

6

-8 -6 -4 -2 0 2 4 6

Log

Fati

gue

Dam

age

(%) f

rom

'New

' ASU

Mod

el

MEPDG Ver. 1.0 Log Fatigue Damage (%)


Comparison of Thermal Cracking between MEPDG Comparison of Thermal Cracking between MEPDG 1.0 and ASU Thermal Model1.0 and ASU Thermal Model

y = 1x + 0.053R² = 1

0

500

1000

1500

2000

2500

0 500 1000 1500 2000 2500

TC

am

ount

from

M-E

PD

G (

ft/m

ile)

TC amount from ASU program (ft/mile)

17 sites, 5 AC thick, 5 PG and 8 mixes(Filtered)


Effective Temperature for Fatigue CrackingEffective Temperature for Fatigue Cracking(Cont(Cont’’d)d)

qq TTeffeff = effective temperature for bottom= effective temperature for bottom--up fatigue cracking, up fatigue cracking, ooFF

qq ffeffeff = effective frequency= effective frequencyqq TT′′ff = climatic factor= climatic factorqq MAAT = mean annual air temperature, MAAT = mean annual air temperature, ooFFqq σσMMATMMAT = standard deviation of the mean monthly air = standard deviation of the mean monthly air

temperature within a given year, temperature within a given year, ooFFqq Sunshine = mean annual sunshine, %Sunshine = mean annual sunshine, %qq Wind = mean annual wind speed, mphWind = mean annual wind speed, mphqq Rainfall = mean cumulative rainfall depth, inchesRainfall = mean cumulative rainfall depth, inches

feffeff T691387.17f016814.1T ′+−−=

Rainfall)(071.0)Sunshine(337.0)Wind(132.1)(254.1)MAAT(110.1T MMATf ++−σ+=′

Note that this climatic factor is the same for both AC rutting and Alligator Fatigue Cracking.

Based on 7776 M-E PDG Runs


Effective AC Modulus (E*Effective AC Modulus (E*effeff))

E*E*effeff = effective Asphalt Mix Dynamic Modulus, in 10= effective Asphalt Mix Dynamic Modulus, in 1055 psi.psi.ηηTeff Teff = bitumen viscosity in 10= bitumen viscosity in 1066 poise (at effective temperature for fatigue).poise (at effective temperature for fatigue).ffeffeff = effective loading frequency in Hz.= effective loading frequency in Hz.Va =Va = air voids in the mix, by volumeair voids in the mix, by volumeVbeff =Vbeff = % effective bitumen content, by volume% effective bitumen content, by volumep34 = % retained on the p34 = % retained on the ¾¾ inch sieve, by total aggregate weight (cumulative)inch sieve, by total aggregate weight (cumulative)p38 = % retained on the 3/8p38 = % retained on the 3/8--inch sieve, by total aggregate weight inch sieve, by total aggregate weight

(cumulative)(cumulative)p4 =p4 = % retained on the No. 4 sieve, by total aggregate weight (cumula% retained on the No. 4 sieve, by total aggregate weight (cumulative)tive)p200 =p200 = passing the No. 200 sieve, by total aggregate weightpassing the No. 200 sieve, by total aggregate weight

( )

( )( )

( ) ( )( )effTeff log393532.0flog31335.0603313.034

238384

eff

eff

42

200200eff*

e1

p00547.0p000017.0p003958.0p0021.087197.3

VaVbVb

8022.0Va058097.0

p002841.0p001767.0p029232.0249937.1ELog

η⋅−⋅−−+

⋅+⋅−⋅+⋅−+

+⋅−⋅−

⋅−⋅−⋅+−=


Input Variables for Each DistressInput Variables for Each Distress

qThree INPUT parts 1. Initial User Input

• General and Basic information for the project(Project Info, Traffic, Climatic, Structure…)

2. As-Design Job Mix Formula Input• Mean and Standard Deviation of necessary

variables for the Stochastic Analysis(Volumetric Properties, Gradation…)

3. As-Constructed Data (each lot)• Raw data collected from the field• From the raw data, the mean and standard deviation

will be computed


Development of Probabilistic Development of Probabilistic Based PRS Model RequirementsBased PRS Model Requirements

qqAs Built si^2 (Variance) of all Key Specific As Built si^2 (Variance) of all Key Specific Mix and Pavement Cross Section Variables Mix and Pavement Cross Section Variables for a Unique Project in a Given Environmentfor a Unique Project in a Given EnvironmentØØMix Volumetrics (Va%, Vb%)Mix Volumetrics (Va%, Vb%)ØØMix Gradation (P200, P4 etc..)Mix Gradation (P200, P4 etc..)ØØPG Grade (Gb*, Viscosity, Pen)PG Grade (Gb*, Viscosity, Pen)ØØHMA Mass DensityHMA Mass DensityØØPavement Cross Section (hacPavement Cross Section (hac…….).)

qqStart Thinking of these Parameters for Two Start Thinking of these Parameters for Two ConditionsConditionsØØLab or Job Mix Design (Demand Function)Lab or Job Mix Design (Demand Function)ØØActual Product Produced by Contractor inActual Product Produced by Contractor in--Situ Situ

(Capacity Function)(Capacity Function)


Job Mix Vs. Plant Produced VariabilityJob Mix Vs. Plant Produced Variability

Job Mix (Design) Formula – Demand Function

µi = Xi

Xi (use Job Mix Formula)

σi

(use historical literature review for “typical” plant variability expected)

Plant Produced – Capacity Functionµi = Xi

Xi (use Job Mix Formula)

σi

(use as produced variation of parameters from plant production)


Input Variables for Each DistressInput Variables for Each Distress

qCritical Inputs for Probabilistic Analysis of JMF (As-Design)

Critical Inputs UnitRutting Fatigue .C. Thermal C.

Mean S.D. Mean S.D. Mean S.D.Target In-Situ

Air Voids % T H.V. T H.V. T H.V.

Effective Binder Content by Vol. % C C C C C C

Cum. Retained 3/4 % T H.V. T H.V. - -

Cum. Retained 3/8 % T H.V. T H.V. - -

Cum. Retained #4 % T H.V. T H.V. - -

Passing #200 % T H.V. T H.V. - -

Gmm N/A D H.V. D H.V. D H.V.

Gmb N/A D H.V. D H.V. D H.V.

Gsb N/A D H.V. D H.V. D H.V.T = Transported from Initial InputC = Computed by Other Variables

D = Direct User Input or Default ValueH.V . = Historical Value or User can input


Calculation of the Variance Associated Calculation of the Variance Associated with the Service Lifewith the Service Life

q Up to 1000 Monte Carlo simulations are performed on the Witczak Dynamic Modulus Predictive equation.

q Each of the variables in the equation is treated as a random number following a Normal Probability distribution with a mean and standard deviation calculated from the statistical analysis of the field measured values for the as-constructed mix.

q For the JMF (JMD), historical (National or State) standard deviations of these variables are used.

q For the “As Produced” product, actual Lot-Project standard deviations of same key variables are used.

q The AC distress type mean and variance is then predicted based on these E* values from Monte Carlo simulation runs.


Cumulative Frequency Distribution of Cumulative Frequency Distribution of the Service Lifethe Service Life

50

100

0Design life (Years) (Target)

Years (increasing)Cum

ulat

ive

Freq

uenc

y D

istri

butio

n,

CFD

Good Mix

Bad Mix

Design Mix

- Years + Years


Calculation of Predicted Life Calculation of Predicted Life Difference (PLD)Difference (PLD)

qq The program allows the user to select The program allows the user to select the PLD to be calculated using either the PLD to be calculated using either one of the following methods:one of the following methods:a)a) PLDPLDTT = (In Situ Mix Life = (In Situ Mix Life -- Target Life)Target Life)

In this case, the predicted life difference In this case, the predicted life difference ““PLDPLDTT”” is calculated as the average difference is calculated as the average difference between the cumulative frequency distribution between the cumulative frequency distribution curve of the In Situ Mix Life and the constant curve of the In Situ Mix Life and the constant Target Design Life. Target Design Life.

b)b)PLDPLDJJ = (In Situ Mix Life = (In Situ Mix Life -- JMF Life)JMF Life)In this case, the predicted life difference In this case, the predicted life difference ““PLDPLDJJ”” is calculated as the average difference is calculated as the average difference between the cumulative frequency distribution between the cumulative frequency distribution curves for the In Situ Mix Life and the Job Mix curves for the In Situ Mix Life and the Job Mix Life. Life.


Example Comparison between Cumulative Example Comparison between Cumulative Distributions of Service Life Between Design Distributions of Service Life Between Design

and As Constructed Mix (Single Lot)and As Constructed Mix (Single Lot)

Lot No. 3, ADOT Signal Road Section Project

0.00.10.20.30.40.50.60.70.80.91.0

0 5 10 15 20 25 30 35

Service Life (years)

Cum

ulat

ive

Freq

uenc

y Di

strib

utio

n

Lot No. 3Lab Mix

Reliability = 96.28%

PLDT = (In Situ Mix Life – JMF Life)


Project Input Data (ContProject Input Data (Cont’’s)s)PAY ADJUSTMENT FACTOR (PF) DATA

7205-52.5-2.5-10

PLD for no Bonus, X3 PLD for no Penalty, X4PLD for Remove and Replace, X5

Max. Bonus (%), Y1Max. Penalty (%), Y2Max PLD (Years), X1Min PLD (Years), X2

707580859095

100105110115120

-10 -8 -6 -4 -2 0 2 4 6 8 10Predicted Life Difference (PLD), years

Pen

alty

/Bon

us F

acto

r (P/

B)

(X1, Y1)

(X3)

(X4)

(X2, Y2)

100 %

Zero +-

(X5)

Predicted Life Difference Versus Penalty/Bonus Factor for AC Rutting

70

75

80

85

90

95

100

105

110

115

120

-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12

Predicted Life Difference (PLD), years

AC

Rut

ting

Pena

lty/B

onus

Fac

tor (

P/B

), %

P/B = (RSL) +

P/B =

P/B = (RSL) +

2.80

P/B =

80.00

107.00

93.00

120.00

Remove and Replace


( ) ( ) ( )

( ) ( )sAC

sACt

tfkktfbAC

bACt

fmmf

bACbACt

rmmbrmAC

mACt

rllmrSAC

sACt

rkksrrPB

CN

BPnC

NBPn

CN

BPnPC

NBPn

PCN

BPnPC

_)_(

_)_(

_)_(

__)_(

__)_(

_1

1/

1/

1/

1/

1/

−+

−+

−+

−+

−=

∑∑

∑∑∑

ββ

β

∑ ∑= =

=L

i

Lp

jIRIijPB BPC

1

1.0/

12 )/(

21 PBPBPB CCC +=

Revised Penalty Bonus Cost Revised Penalty Bonus Cost EquationEquation

For Three Distresses(Three AC Layer System)

For IRI

P/B Cost for three distresses

P/B Cost for IRIFinal P/B Cost


Mode I OutputMode I OutputMODE I : Simultaneous Asphalt Mix and Pavement Structural Design (M-E PDG Shortcut)

SUMMARY OF PROGRAM INPUT

PROJECT GENERAL INPUT DATA PROJECT TRAFFIC AND CLIMATIC CONDITIONSProject ID H492001C Desired Speed (mph) 65

Project Location Signal Road Station Desired Traffic (ESALs) 5,900,000Date of Analysis 1/16/2007 Mean Annual Air T emperature (oF) 57.86Operator's Name Ajatshatru Patni Mean Monthly Air Temp St. Dev. (oF) 18.80

Mean Annual Wind Speed (mph) 8.24Mean Annual Sunshine (%) 89.41

Annual Cummulative Rainfall Depth (in) 8.23

SUMMARY OF PROGRAM OUTPUT

Layer Label ID SHRP 3/4"Layer Thickness (in) 5

Effective Frequency (Hz) ** 33.74Effective Temperature (oF) 103.16

SPT Recom. Frequency (Hz) 25 25 25SPT Recom. Temperature (oF) 100.95

Allowable Rut Depth (in) 0.30Allowable Layer E* (ksi) 557.178

Predicted Rut Depth (in) 0.34Predicted Layer E* (ksi) 461.119

Acceptable (Rut) ??? NOAcceptable (E*) ??? NO

Please Go Back and Revise your Design!!!!

(Deterministic Solution Results)


Influence of the Traffic Repetitions upon Influence of the Traffic Repetitions upon the Critical E* for Signal Road Sectionthe Critical E* for Signal Road Section

Signal Road Section (H492001C),Relationship between Critical Layer E* & T raffic Repititions, αr1 =

1.000, xr1 = 0.623, Design Freq. = 25 Hz

y = 0.0055x0.868

R2 = 1

y = 0.0016x0.868

R2 = 1

y = 0.0007x0.868

R2 = 1

y = 0.0004x0.868

R2 = 1

1

10

100

1000

10000

100000

1000000

10,000 100,000 1,000,000 10,000,000 100,000,000

N, ESALs

Crit

ical

E*(k

si)

Target Rutdepth = 0.1(in.)Target Rutdepth = 0.2(in.)Target Rutdepth = 0.3(in.)Target Rutdepth = 0.4(in.)


Influence of Traffic Level upon Predicted Service Life Influence of Traffic Level upon Predicted Service Life (ADOT Signal Road Section Project (ADOT Signal Road Section Project –– ADOT Traffic ADOT Traffic

Known)Known)

0

5

10

15

20

25

30

35

40

100,000 1,000,000 10,000,000 100,000,000Traffic, ESALs

Pred

cite

d Se

rvice Li

fe, Y

ears

Rutting Criteria = 0.1 inRutting Criteria = 0.2 inRutting Criteria = 0.3 inRutting Criteria = 0.4 inRutting Criteria = 0.5 inActual Design Traffic

RD = 0.28


q The developed prediction models include Mix Characterization PropertyØDynamic Modulus

• For the analysis of rutting and fatigue crackingØCreep Compliance

• For the analysis of thermal fracture

q Two options available to obtain the propertiesØ1: Use of Lab-measured data (Simple Performance Test)

• Dynamic Modulus Test and Creep Compliance TestØ2: Use of Volumetric information

• WPE for E* and a set of regression equations for D

42

Simple Performance Test Procedure


-10 -8 -6 -4 -2 0 2 4 6 8

14.00 40.00 70.00 100.00

130.00 130.00 130.00 Predicted MC

- One (representative) master curve is developed based upon

the multiple E* results.- Obtain optimized seven parameters of Sigmoidal

Function

Development of E* Master Curve

Mix Design

Calculation of effective temp. and

freq.

- Two sets of a combination of effective temp. and freq.

(rutting and fatigue cracking)

)(071.0)(549.0

)(186.1)(876.0)(006.1)(332.2995.13 5.0

RainSunshineWindMMATMAATFreqTeff

+

+−σ++−−=

)(08.0)(333.0)(431.0)(718.1)(121.1)(940.10)(361.362.14

RainSunshineWindMMATMAATzFreqLnTeff

+

+−σ++−−=

RuttingRutting

FatigueFatigueeff

ac

17.6F2(a h )

υ=

+

effeff

17.6vF2(a z )

=+


Mix Design

- Based upon the master curve as well as eff. Temp. and freq. , E*eff s are calculated for rutting

and fatigue cracking

Calculation of Effective Dynamic

Modulus.

r

log(E*)1 exp( log t )

α= δ +

+ β+ γ

effr

1log(a(T)) log log(t )f

= −

2eff effa b c log(a(T))T T+ + =

-10 -8 -6 -4 -2 0 2 4 6 8

14.00 40.00 70.00 100.00

130.00 130.00 130.00 Predicted MC

E*eff


Summary of the 9Summary of the 9--2222Project Performance Related SpecificationsProject Performance Related Specifications

qq 99--22 represents the first true interaction of structural 22 represents the first true interaction of structural mix design and asphalt mix design methodologiesmix design and asphalt mix design methodologies

qq Development of this methodology has been an Development of this methodology has been an enormous, enormous challenge to interact,linkenormous, enormous challenge to interact,link

•• Structural Design: 1Structural Design: 1--37A, 137A, 1--40D40D•• Mix Design (Superpave / SPT): 9Mix Design (Superpave / SPT): 9--19, 919, 9--33 (A)33 (A)•• PRS (QRS): 9PRS (QRS): 9--2222

qq Recall that the process has been field calibrated to Recall that the process has been field calibrated to distress and pavement performancedistress and pavement performance

qq It is the It is the FUTUREFUTURE conceptual methodology that should conceptual methodology that should be used to reward and/or penalize contractors for their be used to reward and/or penalize contractors for their productproduct

Documents

Beta Testing and Validation of HMA PRSonlinepubs.trb.org/onlinepubs/webinars/HMASlides.pdf · Validation of HMA PRS Performance Related ... §Final Report and presentation of results