Flexible Pavement Performance Models in MEPDG

Flexible Pavement Performance Flexible Pavement Performance Models in MEPDGModels in MEPDG

Lev KhazanovichLev Khazanovich University of MinnesotaUniversity of Minnesota

Seminar on Pavement Design Systems and Pavement Performance Models

March 22 –

23, 2007 Reykjavik, Iceland

AcknowledgementsAcknowledgementsGuide for Design of New and Rehabilitated Pavements Structures (NCHRP 1-37A and 1-40D).• Arizona State University (Prof. Matt Witczak, Mohamed El-Basyouny, & many others) • University of Maryland (Prof. C. Schwartz)• University of Illinois (Prof. W. Butler)• Several consultants around the world

Many slides in this presentation were developed under the above projects

OutlineOutline

• Overview of the MEPDG• Load Related Cracking• Rutting Models• Thermal Cracking• Roughness models • Conclusions

Design ProcessDesign ProcessFoundation

AnalysisClimate Materials

PropertiesTraffic Analysis

Trial Design

Pavement Response Model

Calibrated Damage-Distress/IRI Models

MeetPerformance

Criteria?

ModifyDesign

Inputs

AnalysisNo

Yes

Damage AccumulationOver time

OutputsIRIRutAlligator Ck

Long CkTemp Ck

Damage Accumulation Damage Accumulation -- Incremental Incremental Damage ConceptDamage Concept

• Design life is divided into time increments of:– 1 month for rigid pavements– 15 days for flexible pavements

Design life

Incremental Changes Over Pavement Life Incremental Changes Over Pavement Life

Time, years

CTB Modulus

Each load application

Granular Base Modulus

2 8640

Subgrade Modulus

Traffic

AC Modulus

SubSub--Layering for Structural AnalysisLayering for Structural Analysis

Asphalt

Asphalt

Unbound

Unbound Compacted Natural

Bedrock

• Cracking: εt at surface + bottom of all bound layers

• Rutting: εc at midthickness of all layers+ top of subgrade

Critical Response ValuesCritical Response Values

εtεc

εtεc

Critical Response LocationsCritical Response Locations

x

y

8 in 8 in 8 inSx

Sy

CL

A1 A2 A3 A4 A5 A6 A7 A8 A9 A10

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10

4 in4 in

70 Evaluated Points in X-Y Plane

Top of the AC layer

Mid-dept

ProblemProblem

•

In 2002 DG, layered elastic analysis is required for each month of the pavement design life. (260 design increments for 20 years design

life) •

Up to 20 layers in each model

•

70 evaluated points in each layer, up to 1400 points for each time increment

Single design iteration takes between Single design iteration takes between 30 to 60 min on a typical PC30 to 60 min on a typical PC

MNLAYER vs. JULEA and BISARMNLAYER vs. JULEA and BISAR

0

5

10

15

20

25

30

80 160 240 320 400

No. of Evaluated Points

Tim

e (S

econ

d)

MNLAYERJULEABISAR

Flexible Pavement PerformanceFlexible Pavement Performance

Fatigue Cracking

Thermal Cracking

Longitudinal Cracking

IRI

Rut Depth

HMA Fatigue Modeling HMA Fatigue Modeling

•Bottom – Up Crack Propagation:

•Top – Down Crack Propagation

(Classical Fatigue Mechanism)

Temperature &Speed of Loading

E* Varies w/HMA Layers

High Shear Stress Contact Pressure

Aging @ Surface High E @ Surface

Fatigue Damage Accumulates Over TimeFatigue Damage Accumulates Over Time

TIME

FATIGUECRACKING

DesignPeriod

Criteria

( )∑∑= = ⎥

⎥⎦

⎤

⎢⎢⎣

⎡=Δ

m

k

j

i ki

i

tN

nDI1 1 ε

SeasonLoadTop DownBottom Up

Allowable Number of Load ApplicationsAllowable Number of Load Applications

( )( ) ( ) ( ) 332211

ffff kHMA

ktfHff ECCkN ββεβ=

Nf = Allowable number of axle load applications εt = Tensile strain at critical locations EHMA= Dynamic modulus of the HMA, psi kf1, kf2, kf3= Global field calibration parameters βf1, βf2, βf3= Local calibration constants; =1.0 by default

Allowable Number of Load Applications (cont.)Allowable Number of Load Applications (cont.)

( )( ) ( ) ( ) 332211

ffff kHMA

ktfHff ECCkN ββεβ=

MC 10= ⎟⎟⎠

⎞⎜⎜⎝

⎛−

+= 69.084.4

bea

be

VVV

M

( )HMAH

H

e

C

49.302.111003602.0000398.0

1

−++

=

( )HMAH

H

e

C

8186.2676.15100.1201.0

1

−++

=

Bottom-up cracking

Top-down cracking

Vb e= Effective asphalt content by volume, percent Va = Percent air voids in the HMA mixture CH = Thickness correction term

BottomBottom--Up Cracking Up Cracking

⎟⎠⎞

⎜⎝⎛

⎟⎠⎞

⎜⎝⎛+

= + 601*

e16000

100))*log10(D**C'C*C'(C 2211bottomFCwhere:

FCbottom

= bottom-up fatigue cracking, percent lane area

D

= bottom-up fatigue damageC1

= 1.0'2

'1 2CC −= 12 =C

856.22 )1(*748.3940874.2' −+−−= hacC

TopTop--Down CrackingDown Cracking

where:FCtop

= top-down fatigue cracking, ft/mileD

= top-down fatigue damage

( )( ) ⎟⎠⎞

⎜⎝⎛+

= − TopDILogCCTop eC

FC211

56.10 4

Factors Affecting Fatigue Cracking in Factors Affecting Fatigue Cracking in Flexible Pavements Flexible Pavements

• HMA layer thickness.• HMA layer dynamic modulus.• Binder grade in the HMA

mixture.• Air voids in the asphalt layers.• Effective binder content in the

asphalt layers.

Factors Affecting Fatigue Cracking in Factors Affecting Fatigue Cracking in Flexible Pavements Flexible Pavements

• Base thickness.• Subgrade modulus.• Traffic load configuration.• Traffic load, contact area and

tire pressure.• Traffic load repetitions.• Temperature and environmental

conditions.

BottomBottom--Up Fatigue (Alligator) Up Fatigue (Alligator) Cracking CalibrationCracking Calibration

0

10

20

30

40

50

60

70

80

90

100

-4 -3 -2 -1 0 1 2 3

Log Damage (%)

Alli

gato

r Cra

ckin

g (%

of T

otal

Lan

e A

rea)

Se = 5.01%Se/Sy = 0.815N = 405R2 = 0.275

Log Damage (%)Alli

gato

r Cra

ckin

g (%

of

Tota

l Lan

e A

rea)

TopTop--Down Fatigue (Longitudinal) Down Fatigue (Longitudinal) Cracking CalibrationCracking Calibration

0

1000

2000

3000

4000

5000

6000

7000

0 1000 2000 3000 4000 5000 6000 7000

Measured Cracking (ft / mile)

Pre

dict

ed C

rack

ing

(ft /

mile

)

R2 = 0.544Se = 582.8 ft /mileSe/Sy = 0.688N = 312

Measured Cracking (ft/mile)

Pred

icte

d C

rack

ing

(ft/m

ile)

Effect of AC Thickness on CrackingEffect of AC Thickness on CrackingBottom Up Cracking - Alligator

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0 36 72 108 144 180 216

Pavement Age (month)

Alli

gato

r Cra

ckin

g (%

)

50 mm

75 mm

100 mm

150 mm

Permanent Deformation Accumulates Over TimePermanent Deformation Accumulates Over Time

TIME

RUTDEPTH

DesignPeriod

Criteria

( )( )[ ]∑∑∑= = =

=Δm

k

j

i

l

dikddP hRD

1 1 1,

εLoad Month Depth

Accumulation of RuttingAccumulation of Rutting

∑=

×ε=N sub-layers

1i

iip hPD

Load, P

AC Layer

Base Layer

Subgrade

See Fig. A.

Fig. A

εp from pred. Eq.

Sub-layer

Similar for unbound layers

Permanent Deformation in AC LayerPermanent Deformation in AC Layer

iβ

where:εp =Accumulated plastic strain at N repetitions of load (in/in)εr = Resilient strain of the asphalt material as a function of mix

properties, temperature and time rate of loading (in/in)N = Number of load repetitionsT = Temperature (deg F)ai

= Non-linear regression coefficients= field calibration factors

rrTNkh HMArzrHMAHMAp

HMAp 32 *5606.1*4791.035412.3)(1

)(

)( 10 ββεβε

−==Δ

Permanent Deformation in Unbound Permanent Deformation in Unbound Layer (Layer (Tseng and Tseng and LyttonLytton

Model)Model)

Δp(Soil) = Permanent or plastic deformation for the layer/sublayer N = Number of axle load applications εo, β, and ρ = material properties obtained for the resilient strain εr εv = Average vertical resilient or elastic strain in the layer/sublayer hSoil = Thickness of the unbound layer/sublayer, inches ks1 = Global calibration coefficients; =1.673 for granular materials =1.35 for fine-grained materials βs1 = Local calibration constant

βρ

εε

εβ⎟⎠⎞

⎜⎝⎛−

⎟⎟⎠

⎞⎜⎜⎝

⎛=Δ N

r

osoilvsssoilp ehk 11)(

Total Pavement Total Pavement -- RuttingRutting

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1 1.2

Average Measured Total Rutting (in)

Pre

dict

ed T

otal

Rut

ting

(in)

Predicted vs Measured Total Rutting Equality Line

R2 = 0.577N = 334Se = 0.107Se/Sy = 0.818

Average Measured Rutting

Ave

rage

Pre

dict

ed R

uttin

g

Effect of AC Thickness of RuttingEffect of AC Thickness of RuttingPermanent Deformation: Rutting

0

2

4

6

8

10

12

14

16

0 36 72 108 144 180 216

Pavement Age (month)

Rut

ting

Dep

th (m

m)

Hac=50 mmHac=75 mmHac=100 mmHac=150 mm

Thermal CrackingThermal Cracking

HMAHMA--Thermal FractureThermal Fracture

• Uses SHRP Thermal Fracture Model– Recalibrated Using Approximately 30 Sections in

NCHRP Project 9-19

• Thermal Fatigue (cyclic) – Propagation of Cracks Through the Asphalt Layer

• Thermal Stresses– Very Low Temperature– Mixture Properties– Friction

• Mixture Fracture Properties

Materials Characterization (IDT)Materials Characterization (IDT)

Schematic of Crack Depth Fracture Schematic of Crack Depth Fracture Model Model

Amount of Crack Propagation in aAmount of Crack Propagation in a Cooling CycleCooling Cycle

nKAC Δ=Δ

ΔC=

Change in the crack depth due to a cooling cycle.ΔK=

Change in the stress intensity factor

A, n = Fracture parameters for the asphalt mixture

Stress Intensity Factor ApproximationStress Intensity Factor Approximation

)C1.99 + (0.45 = K 0.56oσ

K

= stress intensity factorσ= far-field stress from pavement response

model at depth of crack tipCo

= current crack length

SchaperySchapery--MolenaarMolenaar--LyttonLytton

Model Model

⎟⎠⎞

⎜⎝⎛ +=

mn 118.0

( )n)**(E*2.52 - 4.389*m10 = A σβ log(

where:E=Mixture stiffness.σm =

Undamaged mixture tensile strength.

β=Calibration parameter.

Effect of AC Thickness on Thermal Effect of AC Thickness on Thermal CrackingCracking

Thermal Cracking: Total Length Vs Time

0

50

100

150

200

250

300

0 36 72 108 144 180 216

Pavement Age (month)

Tota

l Len

gth

(m/k

m) Hac=50 mm

Hac=75 mmHac=100 mmHac=150 mm

Pavement Smoothness Pavement Smoothness –– IRIIRI

IRI = IRIi + ΔIRID + Δ

IRISF

IRIi = Initial IRI at construction

ΔIRID = Change in IRI due to distress

ΔIRISF = Change in IRI due to site factors

(age, subgrade properties, non- load distress)

Generalized Smoothness ModelGeneralized Smoothness Model

Site FactorSite Factor

( ) ( ) ( )( )100064.01Pr008.0102.0 +++++= FIecipPIAgeSF

Age

= Pavement age, yearsPI

= Percent plasticity index of the soil

FI

= Average annual freezing index, degree F daysPrecip= Average annual precipitation or rainfall, inches

Generalized Smoothness ModelGeneralized Smoothness Model

( ) ( )( ) ( )RDTC

FCSFIRIIRI Totalo

0.400080.0400.00150.0

++++=

IRIo = Initial IRI after construction, in./mi. SF = Site factor FCTotal = Area of fatigue cracking ft2/mi TC = Length of transverse cracking ft./mi. RD = Average rut depth, inches

IRI Model CalibrationIRI Model Calibration

0

50

100

150

200

0 50 100 150 200Measured IRI, in/mi

Pred

cite

d IR

I, in

/mi

N = 1926R2 = 56 percentSEE = 18.9 in/mi

Flexible Flexible ──

Effect of Fatigue Cracking Effect of Fatigue Cracking (Wheelpath: Longitudinal & Alligator) (Wheelpath: Longitudinal & Alligator)

Initial IRI = 63 in/mi, cracking accumulated linearly over 25 years0

40

80

120

160

200

0 10 20 30 40 50 60 70

Fatigure Cracking, percent area

IRI,

in/m

i

ConclusionsConclusions• The MEPDG incorporated the following

performance prediction models– Load Related Cracking– Rutting Models– Thermal Cracking– Roughness

• The models are calibrated based on the performance data from the LTPP sections located throughout the US and Canada.

• Local calibration of the models is recommended

More InformationMore Information

www.trb.org/mepdg

• Guide Documentation • Software• Climatic database