ICAO Seminar on Aerodrome Physical Characteristics and

ICAO Seminar on Aerodrome Physical Characteristics and

Pavements By Dr.M.W.Witczak Invited Speaker Held at ICAO South American Regional Office Lima, Peru 6 – 9 August 2013

An Overview of the New AASHTO MEPDG Pavement

Design Guide

FEATURES OF THE AASHTO M-E PAVEMENT DESIGN GUIDE

Developed under the US NAS (National Academy of Sciences)–NCHRP (National Cooperative Highway Research program)

$10,000,000 – 7 Year Effort (Largest Single US Transportation Research Project in the History of the US)

Project Team Leaders AC/Flexible Pavements: Dr. M.W.Witczak Rigid Pavements: Dr.M.Darter

4

Introduction Road and Highways are a very significant cost for agencies to

construct, maintain and rehabilitate (US Infrastructure worth $1,000,000,000,000)

Pavement design is a very complex process that involves many variables as well as the variation of each variable. It is one of the most complex Civil Engineering structures to design because we demand a FS=1.0

Mechanistic concepts provide a more rational and realistic methodology for pavement design; however, pavement response models are mathematically very complex and do not have single closed form equation solution.

The M-E PDG provides a consistent and practical method to design a pavement for a desired level of reliability.

The MEPDG considers a wide range of AC Flexible pavement structural sections for : New pavement systems Overlay pavement systems

5

Conventional Flexible Pavements

Deep Strength HMA Pavements

Full-Depth HMA Pavements

"Semi-Rigid" Pavements

6

HMA Overlay over Existing HMA: New Existing AC Conventional AC AC Deep strength HMA pavements AC Full depth asphalt AC Semi-rigid pavements

HMA over JPCP

HMA over CRCP

7

HMA over Fractured JPCP Crack and Seat Rubbilization

HMA over Fractured CRCP Rubbilization

8

The primary distresses considered in the MEPDG for flexible pavements are: Permanent Deformation (rutting)

AC Layers Unbound Base/Subbase/Subgrade Layers Total Rut Depth

Fatigue Cracking Top Down-Longitudinal Cracking Bottom Up- Alligator Cracking

Thermal Cracking

In addition, pavement smoothness (IRI) is predicted based on these primary distresses and other factors.

9

Asphalt 10

Major Asphalt Pavement Distresses

Major pavement distresses Permanent deformation Fatigue cracking Transverse (Thermal) cracking

•How can we simulate these problems in the lab?

11

Hierarchical Input Process

Level 1 (High Reliability) Analysis of special problems Usually will incorporate Testing High Visibility/Risk/Cost Projects

Level 2 (Medium Reliability) Standard Design - Most Cases (Rigorous but practical)

Level 3 (Lower Reliability) Lower impact/risk projects

12

HIERARCHIAL APPROACH (AC MODULUS)

x

x,σII2

f(ψ)

f(ψ)

f(ψ)

x

x

x,σI2

LEVEL MIX BINDER RELIABILITY 1 E* Lab Test G*,δ Lab Test

2 E*Predictive equation

G*,δ Lab Test

3 E*Predictive equation

AC Grade to properties

x,σIII2

13

Hierarchical Approach in NCHRP 1-37A

Major Reasons for Presence in M-E PDG

Allows for a Quantifiable Decision to be Made, Based on Benefit / Costs Regarding the Utility of Using Detailed Engineering Tests and Data Collection / Analysis Techniques Relative to Simple, Empirical Correlations or Engineering Guesses

14

Hierarchical Approach in MEPDG

Major Reasons for Presence in M-E PDG

Provide Quantifiable Methodology for Agency to Prove Certain High Profile, High Importance and High Cost Projects Justified

“Most Advanced State of the Art Technology is

Mandated to Save Significant Cost Benefits”

15

Hierarchical Approach in MEPDG

Major Reasons for Presence in M-E PDG

Collary is also True

“Many Projects do not Require Sophisticated , Advanced Engineering Approaches”

Dynamic Modulus Test Protocol

Follow Latest AASHTO Protocols Test Factorial

5 Temperatures (14, 40, 70, 100, and 130 deg F) 6 Frequencies (25, 10, 5, 1, 0.5, 0.1 Hz)

Recommend 3 Replicates per Mix Recommend 3 LVDT’s per Specimen Critical Attention to Specimen Flatness/

Perpendicularity (Use Capping if Problem)

16

17

Dynamic Modulus Test

ARIZONA STATE UNIVERSITY 18

Compressive Dynamic Modulus (|E*|) and Phase Angle (φ)

Time, t

φ/ωσosinωt

εosin(ωt-φ)

σ, εσ0 ε0

0

0|*|εσ

=E itωφ =

Stress – Strain Relationship

)sin(0 tt ωσσ =

)sin(0 φωεε −= tt

0

0*εσ

=E

)sin()sin(*

0

0

φωεωσ−

=t

tE

Dynamic Modulus Test (Level 1)

20

Construction of E* Master Curve

AASHTO TP62-03

5 Temperatures: 14, 40, 70, 100 and 130 oF

6 Frequencies: 25, 10, 5, 1, 0.5 and 0.1 Hz

ARIZONA STATE UNIVERSITY 21

Manual Shifting

0.01

0.1

1

10

-8 -4 0 4 8Log Reduced Time, s

E* 1

06 psi

14 °F

40 °F

70 °F

100 °F

130 °F

SC-64-22

22

Construction of E* Master Curve

0.01

0.1

1

10

-10 -8 -6 -4 -2 0 2 4 6 8Log Reduced Time, s

Dyn

amic

Mod

ulu

s, 1

06 psi

14 oF

40 oF70 oF

100 oF

130 oF

14 oF

40 oF70 oF

100 oF

130 oF

Log Time, sTime, sTime, s

Dyn

amic

Mod

ulus

, 106

psi

Time-Temp. Superposition Use any arbitrary temperature

value as a reference Normally this value is set to be

at 70°F Shift E* test results at other

temp. to reference temp. by time-temp superposition

E* results are not changed Can calculate E* values at any

temp. and freq. from master curve

0.01

0.1

1

10

-10 -8 -6 -4 -2 0 2 4 6 8Log Reduced Time, s

Dyn

amic

Mod

ulus

, 106 p

si

ARIZONA STATE UNIVERSITY

0.01

0.1

1

10

-10 -8 -6 -4 -2 0 2 4 6 8Log Reduced Time, s

Dyn

amic

Mod

ulus

, 10

6 psi

14 oF

40 oF70 oF

100 oF

130 oF

14 oF

40 oF70 oF

100 oF

130 oF

Log Time, sTime, sTime, s

Dyn

amic

Mod

ulus

, 106

psi

0.01

0.1

1

10

-10 -8 -6 -4 -2 0 2 4 6 8Log Reduced Time, s

Dyna

mic M

odulu

s, 10

6 psi

8

6

4

2

0

-2

-4

-6

-8 0 20 40 60 80 100 120 140

Temperature, ºF

Log

Shi

ft F

acto

r, lo

g a(

T) Log a(T) = a T2 + b T + c

log( *) (log )Ee tr

= ++ +δ

αβ γ1

E* Master Curves Shifting Concept

ARIZONA STATE UNIVERSITY 24

Master Curve

0.01

0.1

1

10

-8 -4 0 4 8Log Reduced Time, s

E* 1

06 psi

14 °F40 °F70 °F100 °F130 °FPredicted

SC-64-22

Witczak Predictive Equation (WPE)

)log393532.0log313351.0603313.0(34

238384

4

220020010

10055.0)(000017.000396.00021.0872.3

82208.0058097.000284.0

)(00177.002923.024994.1E*log

η

ρρρρ

ρ

ρρ

−−−++−+−

+

+−−−

−+−=

f

abeff

beffa

e

VVV

V

Where E* = dynamic modulus (105 psi) η = binder viscosity (106 poise), log log η = Ai + VTSi logT T = pavement temperature (Kalvin), Ai = Intercept of Viscosity-Temperature Regression Equation VTSi = Slope of Viscosity-Temperature Regression Equation Va = Air voids (%) Vbeff = Effective Binder Content by Volume (%) ρ34, ρ38, ρ4 = Cumulative Retained on 3/4“, 3/8“, and #4 Sieves, respectively (%) ρ200 = Passing on #200 Sieve (%)

Dynamic Modulus Master Curve AC Surface with PG76-22

10

100

1,000

10,000

-8.0 -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0

E* (k

si)

Log Reduced Time (sec)

Dynamic Modulus Master Curves with Witczak Predictive Equation (PG76-22, AC Surface)

PG76, Class Y (AC Surface)

ARIZONA STATE UNIVERSITY 27

MC Sigmoidal Predictive Equation

tr = Time of loading at reference temperature δ = Minimum value of E* δ+α = Maximum value of E* β, γ = Parameters describing the shape of the sigmoidal function

logE* )(log1 rte γβαδ ++

+=

ARIZONA STATE UNIVERSITY 28

Time-Temperature Superposition: Shifting

tr = Time of loading at reference temperature t = Time of loading a(T) = Shift factor as a function of temperature T = Temperature

tr = )(Ta

t rt

)( tTa =

ARIZONA STATE UNIVERSITY

E* Master Curve Mathematical Formulation

log( *) (log )Ee tr

= ++ +δ

αβ γ1

[ ]log( ) log( ) log ( )t t a Tr = −And

Where:

E* = Dynamic Modulus (psi)

δ, α, β, and γ = Sigmoidal Parameters

tr = Reduced Time

t = Time (sec)

a(T) = shift factor dependent on temperature, T (in oF)

SOURCE OF VARIATION OF

METHODS

Final Master Curve Equation

[ ])log()log( 2

1*)log(

cbTaTteE

++−+++=

γβ

αδ

Optimized simultaneously to get the 7 parameters (δ, α , β, γ, a, b, and c)

Master Curve Equations

( )rlog tLog E*1 eβ+γ

α= δ +

+

tr = 1/fr

2log a(T) aTemp bTemp c= + +

( )( ) ( ) ( )rlog a T log t log t= −

( )effeq

17.6f2 a h

ν=

+

Definition of Time (Period)

T: Called “Period” but it is actually the time required for the response to begin repeating itself

The fundamentally accepted definition (exclusive of rheologists) is that:

T = tload = 1 / f

T

i.e., f = 10Hz implies 10cycles/sec

or tload = 0.1sec

(tload)

Typical Calculated Frequency Values as Function of Speed

Type Road Facility

Design Speed (mph)

Location Frequency (Hz)

Representative AC Layer (4”-12”)

Thin AC Layers Wearing

Surface (1”-3”)

Thick AC Layers

Binder/Base (3”-12”)

Interstate 60 Mid 15 - 40 45 - 95 12 - 25

Bottom 5 - 20 28 - 55 5 - 15

State Primary

45 Mid 10 - 30 35 - 70 15 - 20

Bottom 5 - 15 21 - 42 5 - 10

Urban Street

15 Mid 5 - 10 10 - 25 5 - 10

Bottom 1 - 4 7 - 14 1.5 - 5

Intersection 0.5 Mid 0.1 – 0.5 0.5 – 1.0 0.1 – 0.25

Bottom 0.05 – 0.25 0.25 – 0.5 0.05 – 0.15

Introduction to the Ai-VTSi Analysis

Relationships Used in the Ai-VTSi Analysis

Loglog ή(cp) = Ai + VTSi* Log Tr ή(cp) – in units of centipoise Tr – Rankine Temperature (Tr=Tf+459.7)

36

ASTM Ai-VTSi Viscosity Model log log η = A + VTS log TR

Viscosity in Witczak E* Model (Part of Level 2)

y = -2.5807x + 8.163R² = 0.9993

0.0

0.2

0.4

0.6

0.8

1.0

1.2

2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05

Log

Log

(Visc

) (c

P)

Log (Temp) (R)

Temperature - Viscosity Relationship

Relationships Used in the Ai-VTSi Analysis

Conversion of Pen (5 sec; 100 gm) to ή

ή (in Poise) = 10.5012-2.2601 * log Pen + 0.00389* (log Pen)^2

Relationships Used in the Ai-VTSi Analysis ή at Trb (Softening Point) = 13,000 Poise (Shell Oil) ή (cp) = ή (cs) * (1 / Gb)

1 Pa-s = 10 Poise

Summary of Ai-VTSi Values for Example (With Mix / Compaction Temperatures)

Impact of Aging Upon E* Master Curves

Change of E* Due to Field Aging Time for 2 Differing Environmental Locations

43

Advantages: E* allows hierarchical characterization takes care of aging takes care of vehicle speed can be linked to PG Binder E* approximates FWD back-calculated modulus provides rational mechanistic material property for

distress prediction FHWA – AASHTO test protocols available Distress predictive models available

Dynamic Modulus (E*)

44

Indirect Tension Creep Test

45

Beam Fatigue Test

46

Rotational Viscometer

spindle

torque

sample

samplechamber

47

Dynamic Shear Rheometer

height (h)

radius (r)

torque (T)deflection angle (Θ)

49

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08

Total Number of ESALs

AC R

uttin

g (in

)

ESALs

Load Spectra

Actual Traffic load spectra yields higher levels of rutting and cracking compared to the classical E18KSAL’s.

Traffic repetitions is a significant parameter influencing pavement distress.

50

51

-

0.05

0.10

0.15

0.20

0.25

PG 82-22 PG 70-22 PG 64-22Binder Grade

AC R

uttin

g (in

)

Binder stiffness has a significant influence upon AC rutting.

As the binder stiffness increases, AC rutting decreases.

In fact, as the entire HMA mix stiffness increases, AC rutting decreases.

52

53

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 10 20 30 40 50 60 70

Traffic Speed (mph)

AC

Rut

ting

(in)

Traffic Speed Influences The AC Rutting.

Creep Speed (Parking Lot, Intersection Analysis) Causes Much More Damage To The Pavement Compared To Faster Highway Speeds.

54

55

-

0.05

0.10

0.15

0.20

0.25

0.30

Phoenix Dallas Atlanta Minneapolis

Environmental Location

AC

Rut

ting

(in)

MAAT (47.2oF)

MAAT (62.1oF)

MAAT (66.5oF)

MAAT (75.1oF)

For all variables being the same, the higher the temperature of an environmental location, the higher the AC rutting becomes.

56

57

0

5

10

15

20

25

30

35

40

2 4 6 8 10 12 14

AC Thickness (in)

Alli

gato

r Cra

ckin

g (%

)

Criteria

Design

Use of M-E PDG Analysis as a Design Tool

AC thickness has a significant influence upon Alligator fatigue cracking. As the AC thickness increases, the amount of alligator (bottom-up) fatigue cracking decreases.

58

59

-0.40

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00-50 -40 -30 -20 -10 0 10 20 30 40 50

Distance (in)

AC

Rut

ting

(in)

Wander = 0Wander = 10 inWander = 24 in

60

-0.40

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00-50 -40 -30 -20 -10 0 10 20 30 40 50

Distance (in)

Bas

e R

uttin

g (in

)

Wander = 0Wander = 10 inWander = 24 in

61

-0.40

-0.35

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00-50 -40 -30 -20 -10 0 10 20 30 40 50

Distance (in)

Subg

rade

Rut

ting

(in)

Wander = 0Wander = 10 inWander = 24 in

The more channelized that the vehicular traffic becomes, the more severe the pavement rutting becomes.

The severity of the rutting is magnified for layers near the surface.

62

63

1 ft

3 ft 5 ft

10 ft

AC

Base

Natural SG Compacted SG

4,000

5,000

6,000

7,000

8,000

9,000

10,000

11,000

12,000

1 3 5 7 9 11

GWT Depth (ft)

Mod

ulus

of S

ubgr

ade

(psi

)

Compacted Subgrade

Natural Subgrade

Presence of GWT near / within unbound material layers can significantly alter the material moduli and hence increase pavement damage.

64

65

0

500

1000

1500

2000

2500

PG70-22 PG64-28 PG58-34Binder Grade

Ther

mal

Cra

ckin

g A

mou

nt

(ft/m

ile)

Binder stiffness has the greatest influence upon Thermal Fracture within a cold environment.

As the binder stiffness (or surface layer stiffness) increases, the AC Thermal Fracture increases.

66

67

0

500

1000

1500

2000

2500

Barrow Fargo Billings ChicagoEnvironmental Location

Ther

mal

Cra

ckin

g Am

ount

(ft/m

ile)

Min. Air Temp.(-47oF)

Min. Air Temp.(-34oF)

Min. Air Temp.(-29oF) Min. Air Temp.

(-15oF)

Thermal Cracking cumulatively increases over time. Combined property of binder content and air void has

an influence upon the Thermal Fracture. In general, AC Thermal Fracture decreases with an

increase of binder content and a decrease in air void.

68

ARIZONA STATE UNIVERSITY 69

Influence of AC Mix Stiffness on Alligator Cracking, (HAC = 1 in)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Low E* Med E* Hi E*

AC Mix Stiffness

Allig

ator

Bot

tom

Up

% C

rack

ing

Refe

renc

e Ba

sed

Upon

600

0 ft

2 / 500

ft

SG Mr= 3 ksi SG Mr= 8 ksi SG Mr= 15 ksi SG Mr= 30 ksi

ARIZONA STATE UNIVERSITY 70

Influence of AC Mix Stiffness on Alligator Cracking, (HAC = 10 in)

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

Low E* Med E* Hi E*

AC Mix Stiffness

Allig

ator

Bot

tom

Up

% C

rack

ing

Refe

renc

e Ba

sed

Upon

600

0 ft

2 / 500

ft

SG Mr= 3 ksi SG Mr= 8 ksi SG Mr= 15 ksi SG Mr= 30 ksi

ARIZONA STATE UNIVERSITY 71

Influence of AC Thickness upon Alligator Cracking

0%

10%

20%

30%

40%

50%

60%

70%

0 2 4 6 8 10 12 14

AC Thickness (inch)

Alli

gato

r Bo

ttom

Up

% C

rack

ing

Ref

eren

ce B

ased

Upo

n 60

00 ft

2 / 500

ft

Med E*, Mr= 3 ksi Med E*, Mr= 8 ksi Med E*, Mr= 15 ksi Med E*, Mr= 30 ksi

ARIZONA STATE UNIVERSITY 72

Influence of Subgrade Modulus upon Alligator Cracking

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5000 10000 15000 20000 25000 30000 35000

Subgrade Modulus (psi)

Alli

gato

r B

otto

m U

p %

Cra

ckin

g R

efer

ence

Bas

ed U

pon

6000

ft2/

500

ft

ARIZONA STATE UNIVERSITY 73

Influence of AC Mix Air Voids upon Alligator Cracking

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 2 4 6 8 10 12

AC Mix Air Voids (%)

Alli

gato

r B

otto

m U

p %

Cra

ckin

g R

efer

ence

Bas

ed U

pon

6000

ft2/

500

ft

ARIZONA STATE UNIVERSITY 74

Influence of Percent AC Binder by volume upon Alligator Cracking

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

6 7 8 9 10 11 12 13 14 15 16

AC Mix Effective Binder Content By Volume (%)

Alli

gato

r B

otto

m U

p %

Cra

ckin

g R

efer

ence

Bas

ed U

pon

6000

ft2/

500

ft

ARIZONA STATE UNIVERSITY 75

0

0.1

0.2

0.3

0.4

0.5

0.6

0 2 4 6 8 10 12 14 16

Thickness of Asphalt Layer (in)

Tens

ile S

trai

n

Eac = 100,000 Eac = 300,000 Eac = 500,000Eac = 1,000,000 Eac = 3,000,000

ARIZONA STATE UNIVERSITY 76

AC Rut Depth Prediction (M-E PDG)

Basic Model:

∑=2

1

pεz

zd dzR

3r2

1

β̂β̂r

r

p NTβ̂εε

=Lab Relationship

[ ] ( )[ ][ ]3r2r

1z

ββr

zo21yxz

zp NTβ β ZCC)σμ(σσ

E1ε ++−

=

( )3r2

1

β̂β̂rrp NTβ̂εε =

ARIZONA STATE UNIVERSITY 77

Unbound Base / Subbase / Subgrade Rut Depth Prediction (M-E PDG)

Basic Model:

∑=2

1

pεz

zd dzR

Not Linear Log-Log; Linear Semi log

β

Nρ

eεεβ̂

εε

r

o

r

p

−

=

=

−β

Nρ

eεεβ̂εε

r

orp

[ ]

+−

=

−β

Nρ

eεεβ̂)σμ(σσ

E1ε

r

oyxz

zp

ARIZONA STATE UNIVERSITY 78

Unbound Base / Subbase / Subgrade Rut Depth Prediction (M-E PDG)

(Cont’d)

β = f (wc) EICM / 1-37A Moduli – Moisture Interaction Critical here ∆w ∆Er

( )rr

o E β, ρ,fεε

=

Cο= f (Εr)

ρ = f (β)

[ ][ ] ( )[ ][ ]NEwf zcfc ,)σμ(σσE1ε yxz

zp β+−

=

Calibrated Model

Lytton Coefficients

ARIZONA STATE UNIVERSITY 79

Influence of MAAT upon Permanent Deformation

0.00

0.20

0.40

0.60

0.80

1.00

1.20

40 50 60 70 80

MAAT Temperature (oF)

AC

Rut

ting

(in)

Low AC E*

Medium AC E*

High AC E*

ARIZONA STATE UNIVERSITY 80

Effect of Traffic Speed upon Permanent Deformation

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 10 20 30 40 50 60 70Traffic Speed (mph)

AC

Rut

ting

(in)

AC Rutting Base Rutting Subgrade Rutting

ARIZONA STATE UNIVERSITY 81

Influence of AC Thickness upon AC Rutting as Function of Depth Within AC

Layer

0.00

0.02

0.04

0.06

0.08

0.10

0 1 2 3 4 5 6 7 8 9 10 11

Depth of Mid point of AC Sublayer (in)

AC

Rut

Dep

th a

t Sub

laye

r Mid

Poi

nt (i

n) 1 in AC Thickness

5 in AC Thickness

8 in AC Thickness

12 in AC Thickness

ARIZONA STATE UNIVERSITY 82

Effect of AC Thickness on Subgrade Rutting at Different Subgrade Modulus

(Medium AC Mix Stiffness)

0.000.050.100.150.200.250.300.350.400.450.50

0 2 4 6 8 10 12AC Thickness (in)

Subg

rade

Rut

ting

(in)

SG Modulus = 8000 psi

SG Modulus = 15000 psi

SG Modulus = 30000 psi

*3993.0*734.11552.31 10*1* NTk

r

p −= βεε

General εp/εr Relationship Used in the 2002 Design Guide

εp = plastic strain

εr = resilient strain

T = layer temperature (deg F)

N = no of load repetition

k1 = Confining Pressure, Depth Function.

βr1, βr2, βr3 = Calibration Factors

HMA Layer

βr1 βr2 βr3

Field Calibration Factors AC-Fatigue

3

24.1

5

11" f

f

1EKFβN

tff

ββ

σ ε−

=

Nf = number of repetitions to fatigue cracking

εt = tensile strain at the critical location

E = material stiffness

K1 = laboratory calibration parameter

βf1, βf2, βf3 = calibration factors

M-E PDG is the most powerful Pavement-Material Analysis-Design Tool ever developed.

M-E PDG will lead to a more fundamental analysis of the consequences associated with the material-structure - environmental interaction.

M-E PDG has the potential for increasing pavement performance and life while decreasing life cycle costs associated with new and rehab scenarios.

85

86

Implementation Considerations

Be careful of blind application of Modified asphalts in MEPDG.

E* value may be okay Distress performance prediction models (ac

rutting, fatigue cracking and thermal fracture) generally calibrated with conventional asphalt mixtures

Performance prediction of Modified AC Mixtures questionable

Suggest local calibration

87

Implementation Considerations

MEPDG is an excellent product and major enhancement to current technology; however the technology is still evolving: Do not expect perfect predictions

Need to locally calibrate to actual field performance Must be prepared to Conduct Trench Sections!!!!!!

Need to have a well defined nationally coordinated approach to develop planned model enhancements Reflective cracking Rutting and fatigue cracking model enhancements Chemically Stabilized Materials Calibration Performance of modified mixtures Refinement of level standard deviations for use in

reliability models