Outline

ƒ Pavement structures
ƒ Comparison to current AASHTO 1993 Guide
New Pavement Design Guide
(Mechanistic-Empirical Pavement Design Guide) ƒ Calibration
ƒ What’s not included in MEPDG
ƒ I l
Implementation
t ti needs
d
ƒ AASHTO DARWin Task Force timeline
Linda Pierce, PE ƒ WSDOT implementation
i l t ti efforts
ff t
State Pavement Engineer
Washington State Department of Transportation

Pavement Structures Capabilities – Performance Indicators

ƒ Hot Mix Asphalt ƒ Concrete ƒ Hot Mix Asphalt ƒ Concrete

– Conventional (HMA over – Jointed-plain – Rut depth – JPCP
crushed stone base) ƒ New ƒ Total ƒ Mean joint faulting
– Deep strength (HMA over ƒ Overlay
O l ƒ HMA ƒ T
Transverse cracking
ki
stabilized base) – Continuously reinforced ƒ Unbound layers ƒ Joint spalling
– Full-depth (HMA over ƒ New – Transverse cracking ƒ Smoothness
subgrade) ƒ Overlay – Alligator cracking (fatigue – CRCP
– Semi-rigid (HMA over – Restoration cracking) ƒ Crack spacing and width
cementitious stabilized ƒ Diamond grinding – Top-down cracking ƒ Load transfer efficiency
materials) ƒ Punchouts
– Reflective cracking
– Full-depth reclamation ƒ Smoothness
– Smoothness
– Overlays

3 4
 Climate Comparison to 1993 Guide - General
Traffic
Materials
ƒ 1993 Guide ƒ MEPDG
– Outputs – Outputs
ƒ Structural number or ƒ Distress over time
Structure ƒ Concrete
C t pavementt ƒ Smoothness
S th prediction
di ti
thickness – Input Levels
Iterations – Input levels ƒ Level 1
ƒ Single value ƒ Level 2
Damage
ƒ Level 3

Time
Response
p Damage Distress
Accumulation
5 6

Hierarchical Input Levels Comparison to 1993 Guide - Climate

– Level 1 ƒ 1993 Guide ƒ MEPDG

ƒ Based on site specific data and testing – Seasonal adjustments – Inputs for EICM
ƒ Used for critical projects, forensics, research – Drainage coefficients ƒ Thermal properties
ƒ Highest testing cost ƒ Wi d speed
Wind d
– Level 2 ƒ Air temperature
ƒ Regression equations ƒ Depth to water table
ƒ Routine
R ti pavementt d designs
i ƒ S radiation
Sun di ti
– Level 3 ƒ Precipitation
ƒ Best guess or default values
– Input values can be mixed and matched
– Damage calculations are exactly the same regardless of
the design level

7 8
MEPDG Climatic Inputs Climate Inputs

ƒ HMA structure ƒ Concrete structure

– Identify weather station – Unbound materials – Hourly temperature profile
ƒ Resilient modulus – Monthly moisture gradient
– Choose from 800 sites
ƒ Moisture
M i t center
t – Used for curling and warping
– Create virtual by – HMA hourly temperature predictions
averaging profile
surrounding sites ƒ Thermal
Th l cracking
ki
– Insert depth to water table ƒ Rutting

9 10

Comparison to 1993 Guide - Traffic Traffic Input Levels

ƒ 1993 Guide ƒ MEPDG ƒ Level 1

– ESALs – Axle load spectra – Average annual daily truck traffic (AADTT)
– Truck equivalency factors – Truck speed – Site specific AVC
– Gear/axle configuration – WIM data
– Axle/tire spacing
– Tire pressure
ƒ Level 2
– Traffic wander – AADTT
– Monthly, daily distribution – Regional/statewide AVC and WIM data
factors
acto s ƒ Level 3
– Average annual daily traffic
– % trucks by truck class

11 12
Traffic Inputs Traffic Inputs
Input Level Input Level
Input Parameters Input
p Parameters
1 2 3 1 2 3
Inputs Required to Compute AADTT Axle Load Distribution Factors
AADTT for Base Year √ √ Axle Load Distribution Factors √ √
AADT and Percent Trucks for Base Year √ Truck Traffic Classification (TTC) Factor √
Directional Distribution Factor √ √ √ General Traffic Information
L
Lane Di
Distribution
t ib ti F Factor
t √ √ √ No. of Axle Types per Truck Class √ √
Truck Traffic Volume Adjustment Factors Axle Spacing √ √
Truck Distribution Factors - Base Year √ √ Axle Load Groups √ √ √
Monthly Distribution Factors √ √ √ Tire Spacing/Axle Configuration √ √ √
Hourly Distribution Factors √ √ √ Tire Pressure √ √ √
Truck Traffic Growth Function/Factor √ √ √
13 14

Axle Load Spectra Comparison to 1993 Guide - Foundation

ƒ 1993 Guide ƒ MEPDG

Load Group
Year Month Hour Axle Type – Resilient modulus – Universal non-linear resilient
0-2 2-4 4-6 .. x-y – “k” values modulus model
i j k Single
Tandem
T id
Tridem
Quad

15 16
Unbound Materials Comparison to 1993 Guide - HMA

ƒ Resilient modulus (HMA) ƒ General properties ƒ 1993 Guide ƒ MEPDG

– Level 1 – Material type – Layer coefficient – Dynamic modulus
ƒ Materials specific testing ƒ AASHTO classification – Resilient modulus (68ºF) ƒ Level 1 - laboratory mix
– Level
L l2 ƒ Unified
U ifi d soilil classification
l ifi ti testing
ƒ Correlations to material ƒ Other (crushed stone, cold ƒ Level 2 – predictive
testing recycled, HMA, etc) equation and binder tests
– Thickness ƒ Level 3 – predictive
– Level
L l3
equation and binder class
ƒ Default
– Poisson’s ratio
ƒ Resilient modulus (PCC)
– Converted to k-value that – Predictive equation
produces equivalent surface
ƒ Gradation
deflections for each month
ƒ Air voids
ƒ Asphalt content
17 ƒ Binder information 18

Comparison to 1993 Guide - Concrete

ƒ 1993 Guide ƒ MEPDG

– Modulus – Modulus of elasticity (7, 14,
– Flexural strength 28 and 90 day)
– Tensile strength – Flexural tensile strength
– 28-day – Poisson’s ratio
– Thermal properties
ƒ Drying shrinkage
ƒ Coefficient of thermal
expansion

19 20
Concrete Design Features

ƒ CRCP ƒ JPCP
– Reinforcement – Joint details
ƒ Bar diameter ƒ Joint spacing
ƒ Spacing
S i ƒ Sealant
S l t type
t
ƒ Percent steel ƒ Dowel diameter and
– Base properties spacing
ƒ Base type – Edge support
ƒ Erodibility ƒ Shoulder type and load
ƒ Base/slab friction coefficient transfer efficiency
ƒ Widened slab
– Crack spacing (optional)
– Base properties
ƒ Base type
ƒ Interface type (bond or
unbonded)
ƒ Erodibility 21 22

Performance Evaluation Calibration

ƒ Procedure ƒ LTPP data used model development

evaluates trial
design to determine
ƒ Local calibration will require
if it meets
t the
th – Field testing program
desired ƒ Condition surveys
performance criteria ƒ FWD testing and coring
– Laboratory testing
ƒ Characterize material properties
– Data analysis

23 24
Step-by-Step
Step by Step Calibration Procedure Step-by-Step
Step by Step Calibration Procedure

1 Select hierarchical input level for each input

1. 8. Determine local calibration coefficients
8
parameter 9. Asses standard error for transfer function
2 Develop experimental design and matrix
2. 10 Improve precision of model
10.
3. Estimate sample size for each distress model 11. Interpretation of results (accuracy of calibration
4 Select roadway segments
4. coefficients)
5. Extract and evaluate roadway segments
6 Conduct field investigations of test sections
6.
7. Assess bias

25 26

Factors Currently not Included in MEPDG Factors Currently not Included in MEPDG

ƒ LTPP database was the primary source of data ƒ Friction,

Friction skid resistance and noise
used in the calibration/validation of the MEDPG ƒ Single and super single tires
ƒ If LTPP data did not include a particular factor
factor, – All axle loads are assumed to be dual tire
then it was not included in MEPDG – Can be simulated in software using special loading
feature
ƒ Future inclusion is dependent on material and ƒ Durability and mixture disintegration
performance data to support model development – Raveling, stripping, ASR, etc.
ƒ Volume change in problem soils
– Frost susceptible or
– Expansive-highly plastic clay soils

27 28
Factors Currently not Included in MEPDG Factors Currently not Included in MEPDG

ƒ Asphalt treated permeable base ƒ Staged construction

ƒ Geogrids and other reinforcing materials ƒ Ultra-thin concrete overlays
– Data not available for calibration – Minimum thickness of JPCP is 6 inches and CRCP is 7
inches
ƒ Semi-rigid pavements – Joint spacing is limited to 10 feet and above
– Can be evaluated in MEPDG ƒ Jointed reinforced concrete pavement
– Fatigue cracking and transfer functions have not been ƒ Early-age concrete opening to traffic
calibrated
– Assumes 28-day minimum
ƒ Pavement preservation programs ƒ Friction interface between HMA overlay and
existing
g concrete ppavement

29 30

AASHTO DARWin-ME
DARWin ME Timeline AASHTO DARWin Task Force Timeline

ƒ Fall 2007 ƒ Summer 2008

– Sign intellectual property agreement with NCHRP – Issue project solicitation package
– ARA system design documentation ƒ 10-12 states
ƒ Winter 2007/2008 ƒ Funding – could be $100k per state
– Issue contract for independent source code review ƒ Fall 2008
ƒ Is
I software
ft modular?
d l ?
– Develop and issue Request for Proposals
ƒ Ability to interface with third party software?
ƒ Spring
p g 2008 ƒ Winter 2008/2009
– Results of independent source code review – Select contractor
– Issue contract for request for information (identification – Commence DARWin-ME development (15-18 months)
of potential consultants)

31 32
AASHTO DARWin Task Force Timeline AASHTO DARWin-ME
DARWin ME User’s
User s Group

ƒ Current version of DARWin (v3

(v3.1)
1) ƒ Composed of all DARWin
DARWin-ME
ME licensee
licensee’s
s
– Sunset within 12 months following the release of the ƒ Annual meeting
DARWin-ME – Discuss software functionality
– Discuss needed enhancements
– Trouble shoot
– Networking

33 34

General Implementation Needs

ƒ Sensitivity analysis
– With several 100 inputs, what is really important?
– AASHTO JTCoP has submitted NCHRP research problem
WSDOT
S O Implementation
p e e tat o Efforts
o ts statement
ƒ Axle load spectra characterization
ƒ Materials characterization
ƒ S
Seasonal l climate
li effects
ff
ƒ Field evaluation
ƒ Calibration to local conditions
ƒ Laboratory protocols
ƒ Link to pavement management system

36
 WIM Site Evaluation (2005 study) Axle Load Spectra (2007 study)

ƒ 11 out of 52 sites possessed valid weight data ƒ 12 WIM sites with 12 months of data
ƒ No significant seasonal variations (ESALs are ƒ WIM sites divided into load patterns
constant throughout year) – Light (6 stations)
ƒ Consolidation of monthly gross vehicle weight – Medium (5 stations)
distributions into a yearly distribution is valid – Heavy y ((1 station))
ƒ No evident trends with respect to ƒ Analyzed single HMA pavement structure
– Geographic location
– Urban versus rural
– Interstate versus non-interstate

37 38

MEPDG Performance Prediction Axle Load Spectra (2007 study)

Longitudinal Cracking (ft/m i) Alligator Cracking (% of the total pavem ent area)

1000 50
ƒ Conclusion
800
Default
40
Default
– No difference in longitudinal cracking
600 Med 30 Med

400
High
20
High – No difference in alligator cracking
Low Low

200 10 – Small difference in IRI

0 0
0 5 10 15 20 25
Year since original construction
30 35 40 0 5 10 15 20 25
Year since original construction
30 35 40 – Less than 2.0 mm difference in rutting (40 year design)
Rutting (m m ) IRI (in/m i)
ƒ Recommendation
10 200 – WSDOT can use one axle load spectra for entire state
8
150
– If known haul, should use site specific axle load spectra
6

ƒ Report to be developed
Default 100 Default
4 Med
Med
High 50 High
2
Low Low

0 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
Year since original construction Year since original construction
39 40
Concrete Model Evaluation Concrete Model Evaluation

ƒ Study observations ƒ Transverse cracking

– Transverse cracking influenced by joint spacing – Results show very small amounts of transverse cracking
– Dowel bar use reduces faulting and smoothness which matches well with p
pavement management
g data
– Base type, traffic and climate have significant impacts ƒ Longitudinal cracking
on faulting and roughness – Primary distress noted in Washington State (and other
– Performance prediction and its relation to input values western states)
match pavement knowledge – MEPDG does not include longitudinal crack model

41 42

Concrete Model Evaluation Dynamic Modulus Study

ƒ Faulting ƒ Washington State University research study

– Calibration factors significantly different from default ƒ Evaluate HMA mixtures from 7 WSDOT projects
values – Class of mix
– Agree with pavement management data for undoweled ƒ ½ inch mix – 5 projects
pavements ƒ ¾ inch mix – 2 projects
– Prediction trends and values are reasonable – Binder type
ƒ Roughness ƒ PG58-22 – 2 projects
– Default roughness calibration underestimates actual ƒ PG64-22 – 1 project
WSDOT roughness
h ƒ PG64 28 – 2 projects
PG64-28
ƒ PG70-22 – 1 project
– For DBR, default calibration factors adequately predict
ƒ PG70-28 – 1 project
roughness
g
– WSDOT needs a studded tire wear prediction model ƒ M
Material
t i l ((aggregate
t and
d binder)
bi d ) sentt tto WSU ffor
43
dynamic modulus testing and evaluation 44
Dynamic Modulus Study Dynamic Modulus Study

ƒ Study objectives ƒ WSDOT intent

– Develop WSDOT dynamic modulus database – How does dynamic modulus affect HMA thickness
– Investigate sensitivity of dynamic modulus to HMA determination?
mixture properties (% passing #200) – Does dynamic modulus vary across the state? By
– Evaluate MEPDG in predicting performance based on aggregate source?
d
dynamici modulus
d l results lt – Can
C one ((or potentially
t ti ll several)
l) d
dynamic
i modulus
d l
ƒ Study in progress results (i.e. Level 1) be used statewide?
– How does aggregate gradation effect dynamic modulus
results
ƒ Dynamic modulus assumed during pavement design procedure,
is this similar to project mix design?

45 46

Calibration and Pavement Management Calibration and Pavement Management

ƒ Existing pavement structure ƒ Terrain ƒ Voluminous amount of pavement data

– Layer thickness (pavement, ƒ Posted speed
base and subbase) – Can it be successfully used to calibrate MEPDG?
ƒ Existing pavement condition
– HMA Binder type
yp data – Currently investigating…
– Aggregate class (½ or ¾ inch)
ƒ Traffic
– Pavement distress
ƒ WSDOT LTPP sites
– FWD results (currently in
– Currently ADT and number of separate database) – Since MEPDG is based on LTPP sites
trucks in 3 classes ƒ Investigate WSDOT LTPP sections for calibration
– 13 bin classification is readily
ƒ How does LTPP pavement condition assessment differ from
available
WSDOT pavement condition rating?
ƒ Referencing system
– Useful for automatically locating
– Currently investigating…
weather data

47 48
Development of Design Tools
Questions?
ƒ Catalog design
– Once MEPDG calibrated to WSDOT
– Base on as high input level as possible (more than likely Level 2)
– Applicable statewide
ƒ Potential for web-based tool
ƒ WSDOT design
g
– Link within pavement management system
ƒ Condition
ƒ Traffic
ƒ Climate
ƒ Material properties
ƒ Etc
Etc.

49 50