Mn/DOT Flexible Pavement Design Mechanistic-Empirical Method Pavement Design Systems and Pavement Performance Models March 22-23, 2007 - Reykjavik, Iceland Bruce Tanquist Assistant Pavement Design Engineer Minnesota Department of Transportation www.dot.state.mn.us

Acknowledgements • • • •

Icelandic Road Administration Haraldur Sigursteinsson Nordic Road Association NordFoU Project

Pavement Design at Mn/DOT • Current procedure – Subgrade soil R-value, traffic, rule-of-thumb materials properties – Relates to ride

• MnPAVE procedure – Modulus of all layers, base strength, repeated load damage in HMA and subgrade – Relates to structural distresses – cracking, rutting

Background WESLEA • Layered Elastic Analysis (5 Layers) • Developed at U.S. Army Engineer Waterways Experiment Station (Van Cauwelaert et al, 1986)

ROADENT

University of Minnesota (1996-1999) Dr. David Timm – Auburn University Dr. David Newcomb – NAPA Dr. Bjorn Birgisson – University of Florida

MnPAVE Pavement Design • Climate model • More material types and default properties • Lab and field test results • Updated performance models

Climate

Seasons MnROAD Cell 21

Modulus (ksi)

10,000

Asphalt Base Soil

1,000

100

10

1

0

50

100

150

200

Day of Year

250

300

350

Criteria for Determining the Beginning of MnPAVE Seasons Season

Criteria

Fall

3-day Average Temperature < 17 °C

Winter

Freezing Index > 90 °C-days

Spring Thaw

Thawing Index > 15 °C-days

Spring Recovery

2 Weeks After Start of Spring Thaw

Summer

3-day Average Temperature > 17 °C

Temperature Model

“Whiplash” Equation

n

Ti ∑ 2 i =1 Di T = n 1 ∑ 2 D i =1 i

5 Seasons vs. 52 Weeks • Number of times the asphalt modulus (stiffness) is calculated. • Affects the calculation time. • Wide range of pavements were simulated. • Design thickness differed by no more than 0.1 in. • Default MnPAVE procedure has 5 seasons.

Structure

Material Properties • New Asphalt • Existing Asphalt (overlay design) • Aggregate Base/Subbase • Soils

Witczak Equation log E = a0 + a1p200 + a2 (p200 )

2

Vbeff + a3 p4 + a4Va + a5 Vbeff + Va

a6 + a7 p4 + a8 p3 / 8 + a9 (p3 / 8 ) + a10 p3 / 4 + 1 + e (a11 +a12 log f +a13 2 logη ) 2

Where: E = 10-5 x Dynamic Modulus (psi)

f = Load frequency (Hz)

η = 10-8 x Dynamic Viscosity (cP)

Vbeff = Effective binder content (% by vol.)

Pb = binder content (% by wt. of mix)

p3/4 = Cumulative % retained on 3/4” sieve

Va = Air voids (% by volume)

p3/8 = Cumulative % retained on 3/8” sieve

Pba = Absorption (% by wt. of aggregate) p4 = Cumulative % retained on No. 4 sieve Gb = specific gravity of binder Gsb = Bulk specific gravity of aggregate

p200 = Percent passing No. 200 sieve

Thompson (ILLI-PAVE) Equation log10 E AC Where:

⎛ AREA ⎞ ⎛ AREA ⎞ ⎟⎟ + 0.26⎜⎜ ⎟⎟ = 1.48 − 1.76 log10 ⎜⎜ ⎝ D0 ⎠ ⎝ TAC ⎠ ⎛ 2 D1 2 D2 D3 ⎞ ⎟⎟ AREA = 6⎜⎜1 + + + D0 D0 D0 ⎠ ⎝

EAC = Modulus of the HMA layer (ksi) TAC = Temperature of the HMA layer (°F) D0 = Deflection at center of load (mils) D1 = Deflection at 12 in. (305 mm) from center of load D2 = Deflection at 24 in. (610 mm) from center of load D3 = Deflection at 36 in. (914 mm) from center of load

Cell 20 Modulus Comparison 1994-2003 Witczak and Thompson Equations 10,000

HMA Modulus (ksi)

7.8” 120/150 Pen. (PG 58-28)

1,000

100

1994 1997 2000 2003

10

20

30

1995 1998 2001 Witczak 40

50

1996 1999 2002

60 70 80 90 HMA Temperature (°F)

100

110

120

Modulus Reduction Factor – Modulus Ratio (R ) • R= E*(near crack)/E* (between cracks)

MnROAD Class 5 Moduli 1994-1996

250

40

Backcalculated E (MPa)

30 150

25 20

100 15 10

50

5

Cells 21, 28 0

0 0

30

60

90

120

150

180

Days from Spring Thaw

210

240

270

300

Backcalculated E (ksi)

35

200

Unsaturated Properties 100%

CL3 est. CL 5 est. CL 3 measured CL 5 measured

90% 80%

CL 4 est. CL 6 est. CL 4 measured CL 6 measured

Saturation

70% 60% 50% 40% 30% 20% 10% 0% 0.001

0.01

0.1

1

10

Suction (kPa)

100

1000

10000

Statewide HMA Deflections 1994 - 2005

Subgrade Modulus Superimposed on Soil Map

Subgrade Modulus Predictions Hogg (FWD) Predicted 85th % 15th % MEPDG High MEPDG Low MnPAVE 5.2

Esubgrade, MPa

200

150

Epredicted = eµ+σ

2

2

µ = -7.74(CLAY2SILT2) + 0.694(e-CLAYe-SILT) + 3.86 σbest fit = 0.34 (range: 0.11 to 0.47)

35 30 25

COV = 34% 20 15

100

10 50 5 0

0 Silty Clay

Silty Clay Loam

Clay

Clay Loam

Silt

Silt Loam Sandy Sandy Sandy Loamy Sand Loam Clay Clay Loam Sand Loam

Esubgrade, ksi

250

Traffic

Simplified Load Spectrum Input • • • • •

Analysis of WIM data Mn/DOT and FHWA vehicle types Axle distributions by truck type Assumptions about truck distributions Route types

MnROAD 2001 WIM Data

Repetitions

Steer Tandem Single Tridem

0

10,000

20,000 30,000 Axle Weight (lbs)

40,000

Vehicle Type 8

Repetitions

Steer Single Tandem Tridem

0

10,000

20,000 30,000 Axle Weight (lbs)

40,000

Burnsville I-35E (Feb. 1 - Mar. 1 1992) 10,924 AADT

Single Tandem

Repetitions

Tridem Steer

WIM MnPAVE

0

10,000

20,000

30,000

Axle Weight (lbs)

40,000

50,000

Output

Questions www.dot.state.mn.us