New Concepts in Controlling Asphalt Strain for Perpetual ...onlinepubs.trb.org/Onlinepubs/webinars/151005.pdfThis schematic shows the basic ideas behind perpetual pavements.\爀 䐀攀猀椀最渀

New Concepts in Controlling Asphalt Strain for Perpetual Pavement Design

Dr. Mary M. Robbins Dr. David H. Timm, P.E.

Goal of Perpetual Pavement Design • Design so there are NO deep structural distresses

– Bottom up fatigue cracking – Structural rutting

• All distresses can be quickly remedied from surface • Result in a structure with ‘Perpetual’ or ‘Long Life’

Perpetual Pavement in the U.S. • Asphalt Pavement Alliance Awards Program • 35+ years of service • Minimal structural improvements • No deep structural distresses

– Problems at surface easily and quickly remedied

http://www.asphaltmagazine.com/dotAsset/4af55c62-42dd-4bbe-a6cc-8f647455eb22.jpg

U.S. Perpetual Pavement Award Winners

Designing Perpetual Pavements

Newcomb, 2001

M-E Perpetual Pavement Design

No Damage Accumulation

Log N

Log ε

Threshold Strain

E1 E2 E3

D1 D2 D3

P

A

εt

εv

What is the Endurance Limit for HMA? • 1972 – Monismith estimates about 70 µε • 2001 – I-710 designed at 70 µε • 2002 – 70 µε used by APA • 2007 – NCHRP 9-38 Lab Study

• 100 µε for unmod binders • 250 µε for mod binders • More severe than field

• 2007 – MEPDG uses 100 to 250 µε • 2008 – Field measurements at NCAT Test Track show much

higher strains

http

://w

ww

.pav

emen

tinte

ract

ive.

org/

wp-

cont

ent/

uplo

ads/

2011

/07/

Devi

ce1.

jpg

Endurance Limit in Laboratory

Normal Fatigue Testing Results VersusEndurance Limit Testing

0

200

400

600

800

1000

1200

1000 100000 10000000 1.1E+08

Number of Loads to Failure

Stra

in, (

10E

-06)

Endurance Limit

Normal Range forFatigue Testing

0

200

400

600

800

1000

1200

1000 100000 10000000 1.1E+08

Number of Loads to Failure

Stra

in, (

10E

-06)

Endurance Limit

0

200

400

600

800

1000

1200

1000 100000 10000000 1.1E+08

Number of Loads to Failure

Stra

in, (

10E

-06)

Endurance Limit

Normal Range forFatigue Testing

http

://w

ww

.pav

emen

tinte

ract

ive.

org/

wp-

cont

ent/

uplo

ads/

2011

/07/

Devi

ce1.

jpg

Endurance Limit in Field

Endurance limit

Strains at 90th percentile are about

2.18 times the lab endurance limit

Need for Distribution-Based Design • Pavements experience wide range of loading and

environmental conditions – Results in wide range of strain responses

• Traditional M-E design uses transfer functions and Miner’s Hypothesis to sum damage over time – Fatigue transfer functions difficult to develop and may not

provide sufficient accuracy – Transfer functions not needed with perpetual pavement

design • Designing with a strain distribution will limit fatigue

cracking and avoid transfer functions – Also arrive at reasonable perpetual (maximum) pavement

thicknesses

DEVELOPMENT OF DISTRIBUTION-BASED DESIGN CRITERIA

Perpetual Pavement Design Framework

Traffic

Climate Material Properties

Pavement Structure

Mechanical Analysis

σ, ε

Response Criteria: Fatigue Cracking

Rutting

NCAT Pavement Test Track: Opelika, AL • 1.7 mile closed-loop test facility • 46 asphalt test sections (200 ft/section) • 10 Million ESALs in two years

– 5 trucks operate 16 hours/day • Surface performance monitored on a weekly basis • Some sections feature embedded pavement

response instrumentation

Scope

PerRoad

• Stochastic M-E design framework

• Accounts for seasonal variation

• Variability in layer thickness and layer moduli

Scope Category Parameter Description

Traffic

• 5 triple trailer trucks • Range of axle loads

Climate

• Mid-depth Pavement Temperatures

Scope Category Parameter Description

Cross-section

• 2006 & 2009 Test Track Sections

• AC Layer: Average and COV of layer thickness

• Base Layer: Average and COV of layer thickness

Material Properties

• Base, and Subgrade: Backcalculate Mr

• Backcalculated AC Modulus E1, adjusted for Seasonal temperatures:

h Base

𝑬𝑬𝟏𝟏 = 𝜶𝜶𝟏𝟏(𝒆𝒆𝜶𝜶𝟐𝟐𝑻𝑻)

Scope Section Year Placed Cracking/Type

N3 2003 Yes/Top-down N4 2003 Yes/Top-down N8 2006 Yes/Fatigue N9 2006 Yes/Top-down

N10 2006 Yes/Fatigue S11 2006 Yes/Fatigue N10 2009 Yes/Fatigue N11 2009 Yes/Fatigue S8 2009 Yes/Fatigue S9 2009 Yes/Fatigue

S10 2009 Yes/Fatigue S11 2009 Yes/Fatigue

Development

Validation

Test Sections for Development of Criteria

Cumulative Distributions for Predicted Tensile Strain

FEL = 118 µε

Section FEL (µε) Percentile at FEL N3 (2006) 151 72nd N4 (2006) 146 71st N8 (2006) 203 70th N9 (2006) 203 99th

N10 (2006) 130 33rd S11 (2006) 118 19th

Cumulative Distributions for Predicted Tensile Strain

Fatigue Cracking

Use N3 and N4 to define

design distribution

Fatigue Ratio • Fatigue Ratio*:

– Rn = fatigue ratio at nth percentile – εn = tensile strain at bottom of AC, at nth percentile – εf = laboratory-determined fatigue threshold

𝑹𝑹𝒏𝒏 =𝜺𝜺𝒏𝒏𝜺𝜺𝒇𝒇

*Willis, J.R. and D.H. Timm. Field-Based Strain Thresholds for Flexible Perpetual Pavement Designs, Report No. 09-09, National Center for Asphalt Technology, Auburn University, 2009.

Results

Percentile

Maximum Design Fatigue

Ratio N8 (FC)

N9 (NC)

N10 (FC)

S11 (FC)

50% 0.68 0.66 0.29 1.37 1.93 55% 0.74 0.74 0.32 1.52 2.15 60% 0.81 0.82 0.35 1.70 2.41 65% 0.88 0.90 0.39 1.90 2.74 70% 0.96 1.00 0.43 2.10 3.02 75% 1.06 1.11 0.48 2.32 3.35 80% 1.18 1.26 0.53 2.61 3.75 85% 1.31 1.43 0.59 2.95 4.24 90% 1.49 1.62 0.66 3.25 4.69 95% 1.73 1.82 0.76 3.64 5.30 99% 2.19 2.14 0.91 4.40 6.31

Criteria

Percentile

Limiting Design Distribution for Predicted Strain

Maximum Design Fatigue Ratio for Predicted Strain

1% 29 5% 41

10% 48 15% 54 20% 60 25% 66 30% 71 35% 78 40% 84 45% 91 50% 100 0.68 55% 110 0.74 60% 120 0.81 65% 131 0.88 70% 143 0.96 75% 158 1.06 80% 175 1.18 85% 194 1.31 90% 221 1.49 95% 257 1.73 99% 326 2.19

VALIDATION RESULTS

Validation Dataset

Validation Dataset

Core on transverse crack

Core on longitudinal crack

Section S8

Validation Results

Section

Fatigue Endurance Limit (µε)

Percentile at Fatigue Endurance Limit based on

Predicted Strain Distribution

N10 (2009) 100 37th N11 (2009) 134 45th S8 (2009) 92 9th S9 (2009) 92 18th

S10 (2009) 99 16th S11 (2009) 94 6th

FEL = 134 µε

CRITERIA IN DESIGN

Perpetual Pavement Design Framework

Traffic

Climate Material Properties

Pavement Structure

Mechanical Analysis

σ, ε

Response Criteria: Fatigue Cracking

Rutting

Perpetual Pavement Design

Adjust cross-section to meet design limit

Perpetual Pavement Design

Perpetual Pavement Design

• Develop maximum AC thickness tables • Following SHRP R23 report*, developed tables

based on regional conditions – Minneapolis – Baltimore – Phoenix

*Jackson, N., J. Mahoney, and J. Puccinelli. Using Existing Pavement in Place and Achieving Long Life, Final Report, SHRP 2 R23, TRB, Washington, D.C., 2012.

Develop Maximum Thickness City

Overall Mean Temperature (°F)

Seasonal Information Month Duration Temperatures (°F)

Minneapolis

45°F Winter 17 weeks 21 Spring 13 weeks 45 Summer 13 weeks 70 Fall 9 weeks 56

Phoenix

70°F Winter 13 weeks 54 Spring 13 weeks 68 Summer 17 weeks 87 Fall 9 weeks 66

Baltimore

56°F Winter 13 weeks 35 Spring 13 weeks 54 Summer 17 weeks 74 Fall 9 weeks 53

Jackson, N., J. Mahoney, and J. Puccinelli. Using Existing Pavement in Place and Achieving Long Life, Final Report, SHRP 2 R23, TRB, Washington, D.C., 2012.

Develop Maximum thickness

• Developed tables based on varying subgrade and base moduli • Criteria:

– Fatigue Cracking = Design cumulative strain distribution – Rutting = 50th percentile < 200µε

Layer Modulus (psi) Poisson’s Ratio

Thickness

Input

Distribution (COV)

Input Average Distribution (COV)

AC Varied based on binder and

season

Log-normal (30%)

0.35 Varied Normal (5%)

Base 30, 50, 100, 250, and 500 ksi

Log-normal (40%)

0.4 6, 8 and 10 in. Normal (8%)

Subgrade 5, 10, and 20 ksi Log-normal (50%)

0.45 Semi-infinite Not applicable

Maximum Thickness Base = 6 inches

Subgrade Mr (ksi)

Base Mr (ksi)

Calculated AC Thickness (in.) Range of

Maximum Thicknesses (in.)

Minneapolis (PG 64-34)

Phoenix (PG 70-22)

Baltimore (PG 64-22) Average

5 30 12.5 15.5 14 14.0 12.5-15.5 5 50 12 15 14 13.7 12-15 5 100 12 14 13.5 13.2 12-14 5 250 8.5 12 10 10.2 8.5-12 5 500 8 11 9 9.3 8-11

10 30 10.5 14 12 12.2 10.5-14 10 50 10.5 13 12 11.8 10.5-13 10 100 10 12 11 11.0 10-12 10 250 8.5 11 9.5 9.7 8.5-11 10 500 7.5 10.5 8.5 8.8 7.5-10.5 20 30 9 12.5 10 10.5 9-12.5 20 50 8.5 12.5 9.5 10.2 8.5-12.5 20 100 8 12 9 9.7 8-12 20 250 7.5 10.5 8.5 8.8 7.5-10.5 20 500 7 10 8 8.3 7-10

Maximum AC Thickness

Conclusions

• No relationship between FEL, predicted strain levels and performance found for sections from 2006 and 2009 cycles

• Consider cumulative distribution of predicted strain rather than a single strain threshold

• All control points should be met on updated criteria – Design fatigue ratios should be less than all maximum

fatigue ratios at the 50th-99th percentiles

• Additional research to incorporate reliability – To guard against erroneous values due to high FEL

Recommendations

• Utilize strain distribution in place of conservative limiting strain of 70 µε – Re-design of PerRoad will include distribution

• Maximum AC thickness tables can be developed for state-specific climate, material, and subgrade conditions

• Maximum AC thickness tables can be used in conjunction with agency-specific design procedures to optimize pavement thickness

• For More information: NCAT Report 15-05

Acknowledgements

We would like to thank NAPA for supporting this research

Thank you

Dr. David H. Timm Brasfield & Gorrie Professor Department of Civil Engineering, Auburn University Email: [email protected] Dr. Mary M. Robbins Assistant Research Professor National Center for Asphalt Technology at Auburn University

Email: [email protected]