3/12/2009

1

Permeable Friction Course (PFC) Mixtures are Different!

36th Rocky Mountain Asphalt Conference and y pEquipment Show

1st Annual Flexible Pavement Research Symposium

Amy Epps Martin, Allex E. Alvarezy pp ,

February 18, 2009

TxDOT

OUTLINE

1. Introduction 2 Current Research2. Current Research 3. Experimental Design and Results4. Summary and Recommendations - Future

Work

TxDOT

3/12/2009

2

1. Introduction1. IntroductionDenseDense--graded mixtures graded mixtures VsVs Porous friction course mixtures Porous friction course mixtures (PFC or OGFC) as (PFC or OGFC) as surface coursessurface courses

Kringos et al., 2007

TxDOT

PFC AdvantagesPFC Advantages•• Reduce splash and sprayReduce splash and spray•• Improve skid resistance in wet conditionsImprove skid resistance in wet conditions•• Decrease noiseDecrease noise•• Produce cleaner runoffProduce cleaner runoff

TxDOT

3/12/2009

3

2. Current Research2. Current Research

Improve TxDOT Improve TxDOT PFC mix design procedure and PFC mix design procedure and recommend construction practices recommend construction practices based on: based on:

VolumetricsDurabilityDrainabilityDensification Effects

TxDOT

3. Experimental Design and Test Results3. Experimental Design and Test ResultsSelected MixturesSelected Mixtures

Mixture Asphalt Type OAC (%) Aggregate Other

MaterialsI-35

PG 76-22

6.1 Sandstone, Limestone

Lime (1%), Fibers (0.3%)

US-59Y 5.8 Limestone

IH-30 6.6 Sandstone

US-83 6.4 Limestone

IH-20 6.5 Limestone

US 59 5 9 Granite LimestoneUS-59 5.9 Granite, Limestone

US-281Asphalt Rubber

8.1 Sandstone, Limestone

NoneUS-290 8.3 Sandstone

US-288 8.0 Granite, Limestone

TxDOT

3/12/2009

4

3.1 3.1 VolumetricsVolumetrics

(%)100*1 ⎟⎟⎠

⎞⎜⎜⎝

⎛−=

mm

mb

GGContentAVTotal

a. Total AV Content

Gmm: theoretical max. specific gravity of the mixtureGmb: bulk specific gravity of the compacted PFC mixture

⎠⎝ mm

Current practice:Current practice:pp• Total AV content (or corresponding density)• Vacuum method or dimensional analysis for Gmb

• Measured Gmm

Theoretical Max. Specific Gravity, Theoretical Max. Specific Gravity, GGmmmm

mixWGMeasured =

Method 1-measured Gmm

mixture at the design asphalt range (6 to 10%)

mixwmixpycwpycmm WWW

GMeasured+− + ,,

Method 2 calculated G

ormixture at low binder content (3.5 to 4.5%)

Method 2-calculated Gmm

b

b

mm

bse

GP

G

PG

−

−=

100100

b

b

se

bmm

GP

GP

GCalculated+

−=

100100

3/12/2009

5

tVW

DG

Method 2-dimensional

WVacuumG b =

Bulk Specific Gravity, Bulk Specific Gravity, GGmbmb

Method 1-vacuum

w

tmb

VDimGρ

=CFWWWW

VacuumGb

wbsb

mb

−−+ ,

Vt=π*r2*h

GGmmmm Comparison and Variability, AR Mixtures Comparison and Variability, AR Mixtures

2.372.412.452.49

cal M

axim

umfic

Gra

vity

Shift due to asphalt loss

2 372.412.452.49

l Max

imum

c

Gra

vity

0.02

0.03

Dev

iatio

n

Results and DiscussionResults and Discussion

Calculated Gmm: less variability and less asphalt-loss error

2.252.292.332.37

3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

Asphalt Content (%)

Theo

retic

Spec

if

Average Meas. Gmm Calculated GmmGmm-Ignition Sample Gmm Compacted Sample

2.252.292.332.37

3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

Asphalt Content (%)

Theo

retic

aSp

ecifi

c

Calculated Gmm Measured GmmStandard deviation

0.00

0.01

Stan

dard

GGmbmb Dim:Dim:Simpler, faster, less expensive, cleaner, required equipment is readily available, and data can be directly used to analyze X-ray CT images

3/12/2009

6

b. Connected AV ContentWaterWater--Accessible AV ContentAccessible AV Content

Method 1-vacuum method

(%)100*)(_

,

,

⎟⎠⎞

⎜⎝⎛−−+

−−⎟⎠⎞

⎜⎝⎛−+

=

CFWWWW

WWCFWWWW

AVWAb

wbsb

SVb

wbsb

Method 2-dimensional analysis

(%)100

)(

dim ×

−−

=td

w

std

ensional V

WWVWAAV ρ

Method 2 dimensional analysis

Interconnected AV Content Interconnected AV Content -- XX--ray Computed ray Computed Tomography and Image AnalysisTomography and Image Analysis

Object

SourceDetector

3D d

Grayscale image B&W image

3D render

TxDOT

3/12/2009

7

Comparison of WaterComparison of Water--Accessible AV and Accessible AV and Total AV Content Total AV Content

90

110cc

es. A

V (%

)Vacuum, PG Vacuum, ARDimensional, PG Dimensional, AR

Most AV in PFC are water accessible

50

70

12 16 20 24 28Total Air Voids (%)

Rat

io W

ater

-Ac

/Tot

al A

V

20

25

30

conn

ecte

d A

V,

Surfa

ce A

V (%

) Equality LinePMLC-PGPMLC-ARCores-PGCores-AR

10

15

10 15 20 25 30Water-Accessible AV (%)

Inte

rcW

ith S

Dimensional analysis with vacuum

Good agreement for interconnected AV and

water-accessible AV

Summary and RecommendationsSummary and Recommendations--VolumetricsVolumetrics

• Use dimensional analysis for determining both Gmb and water-accessible AV content• Use calculated Gmmmm

• The methods used for determining Gmm and Gmb affect: OAC, mixture aggregate gradation, and fibers content

Include mixture-durability test for PFC mix design

Future work:Future work:• Explore the use of connected AV content for mix designand evaluation

3/12/2009

8

3.2 Durability3.2 Durability

Hamburg WheelHamburg Wheel Tracking Test (Hamburg)Tracking Test (Hamburg)

Current practice:Current practice:• No durability test applied

2 4’’4

8

12

16

Rut

(mm

)

Hamburg WheelHamburg Wheel--Tracking Test (Hamburg)Tracking Test (Hamburg)

2.4’’0

0 5000 10000 15000Number of Cycles

Wet ConditioningWet Conditioning

TTI Overlay Test (Overlay)TTI Overlay Test (Overlay)

Zhou et al., 2003Zhou et al., 2003

0

100

200

300

400

500

Load

(lb) Cracking life

No Conditioning (dry)No Conditioning (dry)

-300

-200

-100

0 50 100 150 200

Number of Cycles

3/12/2009

9

CantabroCantabro Loss Test (Loss Test (CantabroCantabro))

300 rev. Before (W0) After (Wf)

100*W-W(%)LC t b fo

4.5’’

100*(%)LossCantabro fo

oW=

No Conditioning (dry)No Conditioning (dry)

Test Specimen Preparation

Air Voids Variability

(COV)

Availability of Equipment (in

Texas)

Testing Time

(hours)

Results Variability

(COV)

Comparison of Comparison of Mixture Evaluation TestsMixture Evaluation TestsResults and DiscussionResults and Discussion

Hamburg Saw trimming 0.030 Medium 5 0.02 to 0.57

Overlay

Saw cutting, drying, final AV checking, and

gluing

0.030 Low 2 0.22 to 1.17

Cantabro Not required 0.016 High 0.3 0.07 to 0.36q g

Additional Additional CantabroCantabro Testing: Testing: wet (24 hrs @ 60°C + drying), cold (3°C), & aged (3 & 6 months @ 60°C)

3/12/2009

10

CantabroCantabro Results Results -- Effect of ConditioningEffect of Conditioning

40(%)

Dry WetLow Temp. 3 Months Aged6 Months Aged

0

20

5.5 5.7 5.9 6.1 6.3 6.5 6.7

Can

tabr

o Lo

ss (

PG 76-22 mixtures

Asphalt Content (%)

Summary and RecommendationsSummary and Recommendations--DurabilityDurability

• Cantabro Loss test recommended• Cantabro test results suggest:

Mixture resistance to disintegration is affected more by aggregate than binder propertiesaggregate than binder properties

The test can be used as a screening tool for PFC mix design, but it may not provide enough sensitivity for selecting the OAC

Cantabro Loss values showed a direct relationship with water-accessible AV content

Future work:Future work:• Evaluate relationships between field and lab. responses• Use analytical performance models to improve PFC mix design

3/12/2009

11

3.3 Drainability3.3 Drainability

Current practice Current practice -- design (SGC specimens):design (SGC specimens):• Ensure total AV content (min. 18%)

• Measure lab permeability (min. 100 m/day)

Current practice Current practice -- fieldfield

• Measure field drainability: water flow value (max. 20 secs)

TxDOT

Field drainabilityField drainabilityLab drainabilityLab drainability

Laboratory and Field Measurement of DrainabilityLaboratory and Field Measurement of Drainability

Water flow value(outflow time)

TxDOT

Coefficient of permeability (k)

3/12/2009

12

Evaluation of Current PracticeEvaluation of Current Practice200

m/d

ay)

SGC - PGCores - PGNCAT Minimum

Results and DiscussionResults and Discussion

0

100

16 20 24 28 32 36Total Air Voids (%)

Perm

eabi

lty (m

Best fit lines

Watson et al.' relationship

TxDOT

Lack of correlation can be related to differences in: (i) Total AV content,(ii) Specimen thickness, and(iii) Internal structure of the mixture

Alternatives EvaluatedAlternatives Evaluated(i) Relationship of water-accessible AV content and

lab-measured permeability,(ii) Relationship of lab and field drainability, and(iii) A l ti l di ti f bilit (E t d l f(iii) Analytical prediction of permeability (Expected value of

permeability using modified Kozeny-Carman Eq.)

(ii) Relationship of lab (ii) Relationship of lab and field drainabilityand field drainability

R2 = 0.82, PG

R2 = 0.75, AR

100

1000

rmea

bilit

y /d

ay)

PG MixturesAR Mixtures

TxDOT

101 10 100 1000

Water Flow Value (s)

Cor

e Pe

r(m

/

NCAT Minimum

3/12/2009

13

200

d C

alcu

late

d

bilit

y (m

/day

) Expected Value of Perme.Calculated Perme.Equality line

200C

alcu

late

d

ty (

m/d

ay) Expected Value of Perme.

Calculated Perme.Equality Line

(iii) Expected Value of Permeability (iii) Expected Value of Permeability ((EE[[kk]) ]) and Calculatedand CalculatedPermeability Permeability (Modified (Modified KozenyKozeny--Carman Equation) Carman Equation)

Road cores

0

100

0 100 200

E[k]

and

Perm

eab

Measured Permeability (m/day)

0

100

0 100 200

E[k]

and

CPe

rmea

bili

Measured Permeability (m/day)

cores

Parameters for E[k]:

SGC specimens

TxDOT

• Average and variance of both aggregate-particle size (gradation) and total AV content (X-ray CT)

• Covariance of aggregate-particle size and total AV content• Empirical calibration coefficient• Aggregate, asphalt and fluid (water) parameters

Summary and RecommendationsSummary and Recommendations--DrainabilityDrainability

• Current practices led to poor drainability evaluation of field-compacted mixtures• Water-accessible AV content may be used as a surrogate of the total AV content to indirectly assesssurrogate of the total AV content to indirectly assess permeability• Use the Expected value (E[k]) as an estimator of permeability. Alternatively, the WFV can be used to asses field drainability

TxDOT

Future WorkFuture Work• Further assess permeability of field-compacted mixtures using laboratory-compacted mixtures

3/12/2009

14

Current CCurrent Construction Controlonstruction ControlAsphalt content, gradation Asphalt content, gradation Visual inspection: density, material variability, segregationVisual inspection: density, material variability, segregation

3.4 Densification Effects3.4 Densification Effects

Minimum smoothnessMinimum smoothness

•• No field density requirements for PFCNo field density requirements for PFC

Assess effects of densification on PFC based on:Assess effects of densification on PFC based on:

ObjectiveObjectiveAssess effects of densification on PFC based on:Assess effects of densification on PFC based on:

Internal structure (air voids [AV] characteristics)Internal structure (air voids [AV] characteristics)Macroscopic response (durability and functionality)Macroscopic response (durability and functionality)

FOR TWO COMPACTION LEVELSFOR TWO COMPACTION LEVELS

TxDOT

Comparison of Total AV ContentComparison of Total AV Content

26

30

34

Void

s (%

)

Road Cores PMLC Specimens

AV Design Range

Results and DiscussionResults and Discussion

Field AV content reproduced at 15 gyrations of the SGC

Field Vs Lab (SGC) air voids content

14

18

22

26

I-35-PG

US-59Y-PG

IH-30-PG

US-83-PG

US-281-AR

US-290-AR

US-288-AR

Tota

l Air

V

010 20 30 40

Air Voids Content (%)

m)

Ongoing Research! Ongoing Research!

voids content

TxDOT

Distribution of AV contentUS-59Y mixture

25

50

75

100

Posi

tion

(mm

Core, Total AV15G Total AV50G Total AV50G Interc. AV

3/12/2009

15

2

3

4

5

6

ge in

Hei

ght (

mm

)US-59Y-PG-50GUS-59Y-PG-15GUS-290-AR-50GUS-290-AR-15G

Compaction Curve and StoneCompaction Curve and Stone--onon--Stone ContactStone Contact

Stone-on-Stone Contact

0

1

0 10 20 30 40 50Number of Gyrations, N

Cha

ng

40

45

d -D

RC

(%) VCA-Mix 12G VCA-Mix 15G

VCA-Mix 50G

DRCmix VCAVCA <

Ongoing Research!Ongoing Research!

TxDOT

US-59Y-PG mixture30

35

0 5 10 15Replicate Specimen

VCA

-Mix

and

VCA-DRC

Ongoing Research! Ongoing Research!

Effect of Densification on DurabilityEffect of Densification on Durability

30

40

50

o Lo

ss (%

) 50 Gyrations15 Gyrations

50

75

100

Life

(Cyc

les)

Cantabro test Overlay test

0

10

20

Dry Wet LowTemper.

Aged 3Months

Aged 6Months

Can

tabr

o

0

25

0 10 20 30 40 50 60Number of Gyrations

Cra

ckin

g

Replicate ResultAverage

Mi t SGC Total AV Cycles to Failure Rut Depth @ 20000 Hamburg-Wheel Tracking test

TxDOT

Mixture Gyrations Contenty@ 12.5 mm

p @Cycles (mm)

12 22 - 11.4112 22.3 - 8.9650 17.6 - 4.8250 16.2 - 5.43

US-59Y-PG

3/12/2009

16

Effect of Densification on DrainabilityEffect of Densification on Drainability

200

300er

mea

bilit

y/d

ay)

US-290-AR-15GUS-290-AR-50GUS-59Y-PG-15GUS-59Y-PG-50G

Laboratory permeability

0

100

16 19 22 25 28Total Air Voids (%)

Wat

er P

e(m

/

10

20

30

r Flo

w V

alue

(s) Total AV=21.7%

Total AV=26.7%3 P

Total AV=23.4%

Field drainability (WFV)

0

10

0 1 2 3 4 5 6 7Number of Static Roller Passes

Wat

er

US-290-AR US-59Y-PG

3 Passes

TxDOT

• High levels of densification (after reaching stone-on-stone contact) are required for mixture durability

• These findings suggest the necessity of:Checking stone-on-stone contact during mix design

Summary and RecommendationsSummary and Recommendations--DensificationDensification

g g gIncluding a construction density control

Short-term action:Increase efforts to establish required roller patterns

Future WorkFuture Work• Develop techniques (e.g., nondestructive methods) to

TxDOT

evaluate the field density and enforce a density specification• Improve the current SGC compaction protocol• Evaluate long-term mixture performance to obtain final recommendations for field density control