DSR-PAV TF Update ETG Meeting · Wes Cooper (Asphalt Institute) David Anderson (Consultant) 1. Stage 1 Completed DSR equilibrium time ... Anderson, D.A. and T.W. Kennedy, “Development

DSR-PAV TF UpdateETG Meeting

May 2018

Pavel Kriz (Imperial Oil/ExxonMobil)Gerry Reinke (Mathy)

Mike Anderson (Asphalt Institute)Wes Cooper (Asphalt Institute)David Anderson (Consultant)

1. Stage 1 Completed DSR equilibrium time • Time to equilibrium is not controlled among different DSRs the same,

however its impact on data variability is not dominant

2. Stage 2 Completed effect of strain magnitude & plate size on variability • Modified test setup did not result in desired improvement in test

variability – precision improved, however accuracy worsen

• Current DSR-PAV test is not discriminatory to varying binder qualities, yet it is a limiting specification

• Phase angle showed as very reliable parameter any lab can measure, and which can discriminate asphalts and temperature changes

• Data fully support current efforts on finding an alternate property (NCHRP 09-59 and 09-60)

Reconnect from September 2017

NC-DNC-B

3.65

3.60

3.55

3.50

3.45

NC-DNC-B

3.65

3.60

3.55

3.50

3.45

25mm, 0.1%

Sample

Log

|G*|

sind

, T h

igh,

Pa

25mm, 1.0%

8mm, 0.1% 8mm, 1.0%

Individual Value Plot of Log |G*|sind, T high, Pa

Panel variables: Geometry, Strain

|G*| is a very unreliable measurement

Phase Angle is a very reliable measurement

NC-DNC-B

5652484440

NC-DNC-B

5652484440

0.1 %, 25mm

Sample

delta

T H

igh

0.1 %, 8mm

1 .0%, 25mm 1 .0%, 8mm

NC-BNC-D

Sample

Individual Value Plot of delta T High

Panel variables: Strain, Geometry

Complex, Storage & Loss Moduli

5

0

Stra

in, 𝜸𝜸

𝝅𝝅 𝟐𝟐𝝅𝝅𝟑𝟑𝟐𝟐𝝅𝝅

𝟏𝟏𝟐𝟐𝝅𝝅

Stre

ss, 𝝈𝝈

𝜎𝜎 = 𝜎𝜎0 sin 𝜔𝜔𝑡𝑡 + 𝛿𝛿

𝛿𝛿

𝜎𝜎0𝛾𝛾0

𝑇𝑇 = period in s

Time, 𝒕𝒕

𝜔𝜔 = 2𝜋𝜋𝑇𝑇

= 2𝜋𝜋𝜋𝜋 (analogous to �̇�𝛾) ℜ

ℑ

𝛿𝛿𝐺𝐺𝐺𝐺

𝐺𝐺𝐺

symbol modulus energy response𝐺𝐺𝐺 storage stored elastic𝐺𝐺𝐺𝐺 loss dissipated viscous

(𝝈𝝈 in phase with 𝜸𝜸)elastic

(𝝈𝝈 out of phase 𝜸𝜸)viscous

𝑮𝑮∗ = 𝐺𝐺′ + 𝑖𝑖𝐺𝐺′′

𝑮𝑮∗ = 𝐺𝐺𝐺2 + 𝐺𝐺𝐺𝐺2 =𝜎𝜎0𝛾𝛾0

tan 𝛿𝛿 = ⁄𝐺𝐺𝐺𝐺 𝐺𝐺𝐺

𝜎𝜎 = 𝜎𝜎0 cos 𝛿𝛿 sin𝜔𝜔𝑡𝑡 + 𝜎𝜎0 sin 𝛿𝛿 cos𝜔𝜔𝑡𝑡

𝝈𝝈 in phase with 𝜸𝜸 𝝈𝝈 out of phase with 𝜸𝜸

𝜎𝜎 = 𝛾𝛾0𝜎𝜎0𝛾𝛾0

cos𝛿𝛿 sin𝜔𝜔𝑡𝑡 +𝜎𝜎0𝛾𝛾0

sin 𝛿𝛿 co𝑠𝑠 𝜔𝜔𝑡𝑡

𝐺𝐺𝐺 𝐺𝐺𝐺𝐺

Representation in Complex Plane

Science Behind DSR-PAV

𝛿𝛿𝐺𝐺𝐺𝐺

𝐺𝐺𝐺

𝑮𝑮∗ = 𝐺𝐺′ + 𝑖𝑖𝐺𝐺′′

𝑮𝑮∗ = 𝐺𝐺𝐺2 + 𝐺𝐺𝐺𝐺2 =𝜎𝜎0𝛾𝛾0

tan 𝛿𝛿 = ⁄𝐺𝐺𝐺𝐺 𝐺𝐺𝐺

𝑮𝑮∗ � sin𝜹𝜹 = 𝑮𝑮∗𝐺𝐺′′

𝑮𝑮∗ = 𝐺𝐺𝐺𝐺

1 2

low phase angle = brittlehigh phase angle = ductile

𝐺𝐺1′′ = 𝐺𝐺2′′𝐺𝐺′′ 𝐺𝐺′′

𝐺𝐺𝐺 𝐺𝐺𝐺

5000 kPa

3 4

low phase angle = brittlehigh phase angle = ductile

𝐺𝐺1∗ = 𝐺𝐺2∗𝐺𝐺′′ 𝐺𝐺′′

𝐺𝐺𝐺 𝐺𝐺𝐺

5000 kPa

DSR-PAV can not capture fundamental differences

0

1

2

3

4

5

6

7

8

9

-40 -20 0 20 40 60 80 100 120 140 160

log(

|G*|

.sin

δ),

Pa

Temperature, deg C

0

10

20

30

40

50

60

70

80

90

100

-40 -20 0 20 40 60 80 100 120 140 160

Phas

e An

gle,

deg

Temperature, deg C

Coating grades

BURA IIIAir-rectified paving

Straight run PG & RAF Coating grades

BURA IIIAir-rectified paving

Straight run PG & RAF5000 kPa

• Two asphalts (PG 64 & PG 46) were oxidized to variety of products ranging from 1 PG stiffer paving grade to roofing coating grades

• Phase angle offers clear differentiation between these binders

Data from: P. Kriz, et al, Rheological Properties of Simple Bitumen, E&E Congress, Istanbul, Turkey, 2012.

Colloidal Stability & Aging Index

• Set of samples of varying phase stability prepared in the lab (by adding oil, asphaltenes, oxidation)

• Are ΔTc, phase angle & aging susceptibility related to colloidal stability?

0-5-10-15-20-25-30

60

50

40

30

20

Delta Tc

Phas

e An

gle

at D

SRPT

16-1919-2222-2525+below 16

=5000 kPaT at |G*|sind

12108642

60

50

40

30

20

RTFO Vis Ratio

Phas

e An

gle

at D

SRPT

16-1919-2222-2525+below 16

=5000 kPaT at |G*|sind

The Proposal

• Short term: Allow good asphalts in the specification• For asphalts tested above G”>5000 kPa, review phase angle• If phase angle is sufficiently high (e.g. above XX degrees)

allow them to pass• TF needs help to collect data – field performance & phase

angle at intermediate temp. volunteers?

• Long term: develop a new fatigue parameter to limit fast aging materials and poor phase stable materials (e.g. NCHRP, D. Christensen)

• There were reasons for aging index & ductility in the specification – Superpave already measures these properties

Appendix

Science Behind DSR-PAV, cont’d

Temp or time

log |G*|δ (phase angle)Max limit on

“creep stiffness”

Low T PG

S(t) promotes lower stiffness, m-val more ductile binder to prevent cracking

High T PG

|G*|/sin delta - Promotes higher stiffness &More elastic binders to prevent rutting

Intermediate PG|G*| sin delta promotes less viscous, more elastic (i.e. more brittle) binders? Does this makes sense?

Better binder

Poorer binder

Test Setup - Conclusion

• Very large variability for modulus measurement• Data from 3 labs had to be excluded from analysis• Variability increases with increasing stiffness• 25mm PP at 0.1% showed lower variability for

individual labs however larger dispersion among labs

25 mm 8mm

COV for Phase Angle about 10x lower

COV Delta, %COV |G*|sin delta, %NC-DNC-BNC-DNC-B

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0.00

COV,

%

NC-BNC-D

Sample_1

Chart of COV |G*|sin delta, %, COV Delta, %

Phase Angle Measurement is Less Variable

Over a narrow range of temperature, G’ & G’’ change relatively proportionally, thus a change in phase angle would be much less significant than a change in |G*|

G’

G’’

δ

𝐺𝐺∗ � sin 𝛿𝛿 = 𝐺𝐺𝐺𝐺

• 5000 kPa limit suggested on very limited data developed from tests on asphalts used in the Zaca-Wigmore Test Road1

• Deacon et al.2 showed that a general trend between G” & fatigue can only be observed when material properties are vastly different (2 – 18 MPa)

Limit Increase

5MPa6MPa

5MPa6MPa

1Anderson, D.A. and T.W. Kennedy, “Development of SHRP Binder Specification”, J AAPT, Vol. 62, 1993, pp. 481-507.2A.A. Tayebali et al., “Fatigue Response of Asphalt-Aggregate Mixes”, SHRP-A-404, National Research Council, Washington, DC, 1994.

Change to 6 MPa is NEGLIGIBLE with respect to mix data