Investigation of Thermal Cracking at a Test Site near Rochester, Minnesota

Mike Farrar & Changping SuiWestern Research Institute

July 17, 20092009 PAVEMENT PERFORMANCE PREDICTION SYMPOSIUMLaramie, Wyoming

Principal Researchers: Fred Turner, Mike Harnsberger, Shin-Che Huang, Troy Pauli, Eric Kalberer, Will Grimes, Ron Glaser, Ryan Boysen, Will Schuster, Fran Miknis, Bill Tuminello, technical staff.

www.westernresearch.org

Acknowledgements

Minnesota DOT

Abatech TSARTM and RHEATM software

This work is being performed in supportof FHWA Contract DTFH61-07D-00005

Evaluation of apparent thermal cracking

1. How does the observed transverse cracking compare to BBR results using laboratory aged binders?

2. How does the observed transverse cracking compare to DSR (4 mm diameter parallel plates) results using laboratory aged binders?

3. How well do the BBR and DSR results compare?

Four asphalts from different sources were used to construct 5 comparative test sections at the Rochester, MN site. The crude oil source used to produce the asphalt binders is the only significant variable from section to section, except one section includes 20%RAP. One of the test sections is showing significantly more thermal cracking than theother 4 test sections.

Rochester, MN Comparative test sites

Constructed August 2006

Asphalt binder source

Asphalt Code Description

MN1-2 Canadian blend, Elvaloy modified

MN1-3 Canadian blend (a different blend compared to MN1-2)

MN1-4 Blend of Arab heavy, Arab medium, and Kirkuk

MN1-5 Venezuelan blend

31P NMR of MN1-2 Evaloy Modified binder

FTIR

Apparent Thermal Cracking

BBR vs. transverse cracking

Modified Tcr

MN 1-3 RTFO

0

2

4

6

8

10

12

14

16

18

-46 -40 -34 -28 -22 -16 -10

Temp (°C)

Ther

mal

Str

ess

(Mpa

)

Tsar+ Two-Point Six-Point

Two-Point Tcr = -33.9°CSix-Point Tcr = -32.3°C

Modified Tcr based on: Shenoy, A., “Single –Event Cracking Temperature of Asphalt Pavements Directly from the Bending Beam Rheometer Data”, Journal of Transportation Engineering, Vol. 128, No. 5, September 2002.

Modified Tcr vs. transverse cracking

Modified Tcr based on: Shenoy, A., “Single –Event Cracking Temperature of Asphalt Pavements Directly from the Bending Beam Rheometer Data”, Journal of Transportation Engineering, Vol. 128, No. 5, September 2002.

T BBR vs. Tcr

Dynamic Shear Rheometer (DSR) with 4 mm diam. parallel plates

---- Problem: Error due to machine compliance at low temperature---- Solution: Correct compliance error using method developed by Schröter et al.1

1 K. Schröter, S. A. Hutcheson, X. Shi, A. Mandanici, and G. B. McKenna, J. Chem. Phys. 125, 2006, 214507.

Plate radius = 2 mmSample gap = 1.5 ~ 1.75 mmSample volume = 18.84 ~ 21.98 mm3

TA Instruments has adjusted the ARES rheometer software to correct for machine compliance

WRI’s ARES rheometer

K*mes is the measured complex torsional stiffness K*s is the actual complex torsional stiffness of the sample Kt is the machine torsional stiffness. G is the shear modulush is the gapR is the plate radius

Compliance correction cont.

Compliance Correction for Dynamic Data

tsmes KKK1

*1

*1

+=

hRGK

4

2π

= tsmes KRGh

RGh 1

*2

*2

44 +=ππ

From Schröter et al 2006

Assume the machineis a spring in series with the viscoelastic sample

BBR vs.. DSR (4-mm diam. parallel plates)

BBR DSR (PP4)

Materials required for one specimen

~ 10 g ~ 25 mg

Temperature for preparing specimen

Above 135 oC 50 – 70 oC

Pre-mold for specimen Yes No

Temperature conditioning Alcohol/glycol Nitrogen gas

Time needed to run one isotherm

~ 5 hrs ~ 1/2 hr (5.5 hrs for 10 isotherms)

Material, fabrication, time requirements

Comparison: BBR, and 4,8, and 25 mm diam. plates G* master curves

Time at T+10C & G(t) = (G(t) at T, 7200s)

Interconversion - dynamic to creep stiffness

1.E+07

1.E+08

1.E+09

1.E+10

1.E-07 1.E-05 1.E-03 1.E-01 1.E+01 1.E+03 1.E+05 1.E+07

Red. Time (s)

Cree

p St

iffne

ss S

(t) M

Pa

MN1-5 RTFO S(t) from dynamicdata (RHEA)MN1-5 RTFO BBR S(t)

MN1-5 RTFO S(t) from dynamicdata (RHEA Leaderman)

Reference temperature = -20˚C

BBR and DSR low temperature Performance Grade

Slope = relaxation rate mr

7200

Log reduced time, s

log

G(t),

Pa

60Log loading time, s

Log

S(t)

, Pa

Slope = mc-value

BBRCreep Stiffness S(t) and m-value

DSRRelaxation modulus G(t) and the

slope at 2 hours (4-mm diam. plates)

Similarity of two plots: (1) Both S(t) and G(t) represent the stiffness of material(2) Slopes of both plots are related to relaxation rate of

material

Temperature = Low PG + 10 ˚C Ref. temperature = Low PG

BBR S(t) vs. DSR G(t)

S (t) = 300 MPa at 60 S

G (t) = 156 MPa at 2 hr

Tem

pera

ture

= L

ow P

G +

10˚C

Temperature = Low PG

6 WRI/FHWA validation site asphalts

2.0x107 4.0x107 6.0x107 8.0x107 1.0x108 1.2x1084.0x107

8.0x107

1.2x108

1.6x108

2.0x108

2.4x108

Linear Regression: S(t) = A + B * G(t) A 1.74237E7 B 1.80759R2 0.9396

S( t

) at 6

0 s,

Pa

G( t ) at 2 hours, Pa

BBR m-value (mc) vs. the relaxationmodulus slope (mr)

mc = 0.3 at 60 S

mr = 0.258 at 2 hr

Temperature = Low PG

Tem

pera

ture

= L

ow P

G +

10˚C 6 WRI/FHWA

validation site asphalts

0.275 0.300 0.325 0.350 0.375

0.32

0.34

0.36

0.38

Linear Regression: mc = A + B * mrA 0.12529B 0.67632R2 0.8925

mc v

alue

S(t)

on

BBR

at 6

0 s

mr value from G(t) on DSR at 2 hours

T BBR vs. T 4mm DSR

T (4 mm DSR) vs. transverse cracking

MN1-4 Loss Tangent

Transverse cracking vs. loss tangent

Summary and Questions

• BBR and 4 mm diam. parallel plate DSR did not seem to explain the higher level of cracking in the MN1-4 section compared to the other sections.

• Will the current trend in transverse cracking continue?• BBR and 4 mm DSR appear to correlate well.• Consider fracture: DTT or ABC Device?• Consider ductility using G’/(η’/G’) at low temperatures (< 0 ˚C).• Consider converting the relaxation modulus from the 4 mm

DSR to thermal stress using Boltzmann’s hereditary integral.• Would it be useful to extract the asphalt from two or three year

old cores and evaluate using 4 mm DSR?• Do we need to consider physical hardening?