Fiber Reinforced Asphalt Pavements

54th ANNUAL IDAHO ASPHALT CONFERENCE

October 23, 2014Moscow, Idaho

Kamil E. Kaloush , Ph.D., P.E.

AgendaASU’s experience in testing and evaluation of FRAC

• Fiber Reinforced Asphalt Concrete (FRAC)

• Field projects constructed

• Process and production

• Laboratory tests

• On Going Research

3

Why do we use modifiers in HMA?• Mitigate both traffic and climateinduced pavement distresses– Rutting Resistance

• Stiffer asphalt binders

– Reduce Cracking• Eliminate or delay cracking

• Inhibit crack propagation

– Reduce Surface Wear: Raveling

4

Types of Modifiers• Polymers

Plastomers

Elastomers

• FillersGilsonite Mineral fillers

• Fibers1960’s Asbestos

polyester & polypropylene, glass, carbon, celluloseAramid (Kevlar)

over 31+ recycled waste fibers

Crumb / tire rubber

5

Why Fibers in HMA?• Additional tensile strength and fracture energy

• Reinforcing / load transfer element & minimize crackpropagation and severity

• Favors slight increase in optimum bitumen content

• Reduce drain down of bitumen in HMA

• Favorable cost

• Literature:– Increased dynamic modulus

– Better moisture susceptibility & freeze thaw resistance

– Rutting resistance

– Reduce, delay or eliminate of cracking

FHWAALFTests

Fibers in HMAMixtures

Tire Fibers

Coconut Fibers

Cellulose Fibers

(Chowdhary et al.)

(Passos, 2005).(FORTA®)

HuaxinChen et al(2009)

Tire Fibers Early Research in Arizona

The effect of polypropylene fibers on asphalt performance – Tapkin,ScienceDirect 2007

https://carpetrecovery.org/wp-content/uploads/2014/04/NyconG.pdf

Rheological properties of fiber reinforcedasphalt binders – Ye and Wu, IndianJournal of Engineering & MaterialsSciences, 2008

Scanning Electron Microscope(SEM):

Cellulose: flocculent materialwith porous and ribbon typesurface (strong absorption ofbinder)

Polyester and mineral fibers:round and smooth surface,much lower absorption.

FORTA® AR® FiberBlend of collated Polypropylene and Aramidfibers

FORTA® Fibers fibrillated

Physical Characteristics

Materials Polypropylene Aramid

Specific Gravity 0.91 1.45

Tensile Strength (MPa) 483 3000

Length (mm) 19 19

Acid/Alkali Resistance inert inert

DecompositionTemperature (°C) 157 >450

ASU Early Projects

• Boeing Mixture –Pilot Study

• City of Tempe –Evergreen Drive

Binder Mix Design Data Mix Type

Binder Type Design AC (%) Target Va (%) Gmm FORTA Boeing PHX D-1/2 PG 70-10 5.10 7 2.4605

Binder Mix Design Data Mix Type Binder Type Design AC (%) Target Va (%) Gmm

PHX C-3/4 Control PG 70-10 5.00 7 2.428 PHX C-3/4 1 lb/Ton PG 70-10 5.00 7 2.458 PHX C-3/4 2 lb/Ton PG 70-10 5.00 7 2.471

Evergreen Drive

Boeing

Laboratory Tests• Binder Tests• Triaxial Shear Strength• Dynamic Modulus• Permanent Deformation

– Repeated Load– Static Creep

• Beam and Axial Fatigue• Indirect Tensile Strength and Creep• Fracture and Crack Propagation

Polypropylene Modified Binder

Mix time and temperature:Time: 30 minTemperature: 329 and 365 F(165-185 C)

Conventional Tests

Superpave / SHRP Tests

Penetration AASHTO T49-93Softening Point AASHTO T53-92Rotational Viscosity AASHTO TP48

Dynamic Shear Rheometer (DSR): AASHTO PP1Bending Beam Rheometer(BBR): AASHTO TP1-98

Binder Tests

0.0

0.2

0.4

0.6

0.8

1.0

2.70 2.75 2.80 2.85 2.90 2.95

Log (Temp, oRankine)

Log

(Log

vis

cosi

ty, c

P)

(41) (103) (171) (248) (335) (432)(deg F)

0.0

0.2

0.4

0.6

0.8

1.0

2.70 2.75 2.80 2.85 2.90 2.95

Log (Temp, oRankine)

Log

(Log

vis

cosi

ty, c

P)

(41) (103) (171) (248) (335) (432)(deg F)

Pen59, 77oF

Soft. Point 139oF

Brookfield Viscosity 200-350oF

Viscosity – Temperature

BINDER TEST RESULTS

Laboratory MixingProcedure

Field Laboratory Mixes

Field Laboratory Mixes

Visual Observation

DYNAMIC COMPLEX MODULUS (E*)E* = Dynamic Complex Modulus = o / o

o = peak dynamic stress amplitude (kPa / psi)

o = peak recoverable strain (mm/mm or in/in)

= phase lag or angle (degrees) = VISCOELASTIC PROPERTY

Time

Stre

ss-S

train

o o

o sin( t – )

o sin( t)

E* TESTRESULTS

747,179

479,964

1,097,269

242,350133,365148,180

0

200,000

400,000

600,000

800,000

1,000,000

1,200,000

Salt River 3/4" PG70-10 Salt River 3/4" PG64-22 Fiber-Reinforced

E*,

psi

100 F 130 F

Evergreen

Boeing

4,027,411

5,132,366

1,500,000818,000

466,000294,000

0

1,000,000

2,000,000

3,000,000

4,000,000

5,000,000

6,000,000

7,000,000

Control 1 lb/Ton

E*, p

si

40 F 100 F 130 F

Repeated Load PermanentDeformation Tests

Secondary

Primary

Tertiary

FN Defines N/Time WhenShear Deformation Begins

Repeated Load Permanent Deformation Test(Flow Number) Fiber Reinforced Asphalt

0.0

0.4

0.8

1.2

1.6

2.0

2.4

0 20000 40000 60000 80000 100000Cycles, N

Acc

umul

ated

Str

ain

(%) WesTrack Section 19

Fiber-Reinforced Asphalt

(a)

Boeing

Flow Number Evergreen Dr.

Fatigue Test Results

Nf = 6.16E-15. -5.15

R2 = 0.89

Nf = 1.54E-16. -5.42

R2 = 0.95Nf = 2.13E-13 -4.67

R2 = 0.87

10

100

1000

1000 10000 100000 1000000 10000000

Number of Cycles (N)

Stra

in L

evel

()

Conventional 3/4" MixPG 70-10

Fiber-Reinforced Mix

Conventional 3/4" Mix PG 64-22

Fatigue Test Results

Flexural Strength and Post Peak Energy

Force vs Deflection-FEC20

0

50

100

150

200

250

300

0.000 0.050 0.100 0.150 0.200 0.250 0.300

Deflection, in

Forc

e, L

bs

Comparative Plot – Cyclic Load

0

50

100

150

200

250

0.000 0.050 0.100 0.150 0.200 0.250 0.300

Deflection (in)

Forc

e (lb

)

Control1 lb/Ton2 lb/Ton

Peak =176 lb Peak = 170 lb

Peak = 111 lb

Indirect Tension Tests

• Disk shape specimen (150 x 38 mm) withvertical and horizontal LVDTs on both sides

• The tensile creep– Three temp: 0, 10, and 20oC– Static load along the diametral axis of a

specimen– Deformations used to calculate tensile creep

compliance as a function of time

• The tensile strength– Determined immediately after the tensile

creep test– Constant rate of vertical deformation to

failureIndirect Tensile Tests Loading Frame and Specimen with LVDTs

Tensile Strength Evergreen

371

408410

610

571

468

598

579

467

300

350

400

450

500

550

600

650

700

0 10 20 30 40 50 60

TEMPERATURE (oF)

TE

NSI

LE

ST

RE

NG

TH

(ps

FEControl FE1lb-Ton FE2lb-Ton

Total Fracture EnergyTOTAL FRACTURE ENERGY OF FORTA EVERGREEN MIXES

279.3

163.2113.0

198.0

284.5430.4

206.7294.3

550.6

0

100

200

300

400

500

600

700

0 10 20 30 40 50 60

TEMPERATURE (oF)

TO

TA

L F

RA

CT

UR

E E

NE

RG

Y (l

bs-p

s

FEControl FE1lb-Ton FE2lb-Ton

VERTICAL DEFORMATION

LOA

D (l

bs)

Fracture EnergyArea under curve

Comparison of Indirect TensileStrength at 5 Temperatures

00.5

11.5

22.5

33.5

44.5

-10 0 10 21.2 37.8

Indi

rect

Ten

sile

Stre

ngth

, N/m

m2

Temperature, °C

IDT Comparison

Control 1 lb/ton 2 lb/ton

FRACTURE AND CRACK GROWTHMODELC* LINE INTEGRAL

• C* Line Integral–analog of the Jintegral where strains anddisplacement replaced by their rates(time dependent materials)

• Defined as the difference of 2identically loaded bodies havingincrementally differing crack lengths

dU*=Change in energy rate for a load P and a crack extension dCB=thickness

dCdU

BC *1*

C* MULTIPLE SPECIMENS METHOD

• Specimenssubjected to different constant displacementrates

• Load and crack length measured as a function of time

T (time)

P (L

oad)

a (c

rack

leng

th)

Step 1

a

P

(displacement rate)

P (L

oad)

Step 2

a1

a2

a3

a1>a2>a3

Load vs displacement rate for fixed crack lengthArea under the curve= rate of work done U* per unit of crack plane thickness

a (Crack Length)

U* (

Ene

rgy

Rat

e)

Step 3

1

2

3

1< 2< 3

Crack growth rate are plotted versus crack length

(displacement rate)

C* L

ine

Inte

gral

Step 4

Area under the curve ( Step 3) is plotted against crack lengthSlope of curve is C*

Crack growth rate is plotted as a function of C*The higher the slope the more resistant the material

C* L

ine

Inte

gral

Step 5

(Crack Growth rate)

C* LINE INTEGRAL Literature Results

Abdulshafi and Kaloush (1988)

C* Fracture Test

Develop test geometry, protocol, temperature andloading rate dependency, FE analysis and predictivemodels

(Jeff Stempihar, PhD 2013)

C* LINE INTEGRAL

ACTUAL TEST PERFORMED20 minutes

C* Fracture Test

Post Test Failure

10/26/2014

What is the field telling us?Boeing – Mesa, AZ

infield placed in 20082010, broken pipe caused a sink 2.5’ W x 8’ L x 2.5” DOnly (1) 10” crack found

10/26/2014

Field MaintenanceAsphalt saw-cut & removed in 2 pieces 3’W x 8’L x 2”Thick

One center anchor on 560lbs slabRemoved in one piece!

Evergreen DriveDecember 2013

Fibers Extraction

xylene, toluene and trichloroethylene

Centrifuge Procedure

Fibers Recovery

Microscopic Observations ofRecovered Fibers

Summary of Laboratory Tests (Fibers Benefits)– The viscosity temperature susceptibility relationshipshowed positive and desirable modification process.

– Higher Dynamic Modulus E* values compared to theconventional mixtures.

– Gradual accumulation in permanent strain and highertertiary flow values => desirable properties to resist rutting.

– Higher fatigue life and fracture energy– Higher crack propagation resistance as represented by theC* Fracture Test.

– Fiber extraction Process• Good estimate of actual fiber content• Quality Control / Quality Assurance

USE of Data in the MEPDG (PavementME)

Extrapolated Layer Coefficient of Fiber ReinforcedAsphalt Concrete Mixture

600,000

0.44

630,000

Fiber Reinforced Mixture

Typical Dense Graded Mixture

0.53

Laboratory Evaluation

Mix Property JACMixture

SHRMixture

(Control)

SieveSize (US)

SieveSize (SI)

JACMixture

% Passing

SHRMixture

% Passing

FAA P-402 Control Points

Gradation Open Open 1" 25.4 - - -

Binder PG 64-34 PG 64-34 3/4" 19 100 100 100

Asphalt Content 5.70% 5.60% 1/2" 12.7 82 85 70-90

Laboratory Target Air Voids

13,15%* 15,17%* 3/8" 9.5 57 52 40-65

Gmm 2.416 2.540 No. 4 4.76 22 19 15-25

Hydrated Lime (%) 0.75% - No. 8 2.38 12 13 8-15

Fiber Reinforcement 1 lb/ton (0.5 kg/MT) None No. 30 0.6 6 7 5-9

Mixing/Compaction Temp, F ( C)

325/300 (163/149)

325/300 (163/149) No. 200 0.074 2 2.5 1-5

*Air voids for cylinder and beam samples, respectively (Corelok Method)

Jackson Hole Airport •Temperature changes from: -40oF to 41o F (winter) & up to 104oF in the summer•Elevation requires higher approach speeds•Short runway length•Accommodates planes such as the 757 and A320•Snow plowing caused raveling in existing pavement

Raveling Test

Location Replicate Wi Wf % Loss Average CV

Jackson HoleAirport

1 986.8 959.6 2.8%

2.6% 16.9%2 970.1 940.2 3.1%

3 967.6 945.1 2.3%

4 1031.5 1009.7 2.1%

Control

1 987.8 953.4 3.5%

3.7% 33.3%2 998.9 971.9 2.7%

3 963.5 910.9 5.5%

4 999.3 968.4 3.1%

•Cantabro Test •LA abrasion machine without steel balls•Mashall specimen size•Test temperature = 25 C

•Recommend a lower test temperature forsoft binders

CO2 EQUIVALENT EMMISSION COMPARISON

Service life (Y)

Total kg Annual CO2 Eq. /km runway (lb/mi runway) % Change

10 128279 (455,703) 100.0% 15 85519 (303,801) 33.3%

20* 64139 (227,850) 0.0% 25 51312 (182,283) -20.0% 30 42760 (151,902) -33.3%

The transport distance was assumed to be 25 km (15.5 miles), the density ofasphalt concrete was taken as 2275 kg/m3 (142 lb/ft3) and a runway width of 45.7 m(150 feet). The use of FRAC as the FAA P-401 surface course can result in a 33%decrease in total kg of annual CO2 equivalent per km of runway. This is based on theassumption that the dynamic modulus increases by 50% to 300,000 psi (1,723) forFRAC and is also limited by the current FAA design procedures.

* Standard FAA design life

ItemJackson Hole Airport Sheridan County Airport

Unit $/unit Total Cost Unit $/unitTotal Cost

P-402 Porous Friction Course, tons 8530 48.5 $413,705 5800 58.6 $339,880

PG 64-34 Modified Binder, tons 640 1000 $640,000 520 890 $462,800

Total Cost of HMA Mixture$1,053,70

5 $802,680Cost per Ton of HMA Mixture $124 $138

ItemJackson Hole Airport Sheridan County Airport

Unit $/unit Total Cost Unit $/unitTotal Cost

P-402 Porous Friction Course, tons 8530 50 $426,500 5800 60.4 $350,320

PG 64-34 Modified Binder, tons 640 1000 $640,000 520 890 $462,800Fiber Reinforcement Additive,

lbs 8530 7 $59,710 5800 8.35 $48,430

Total Cost of FRAC Mixture$1,126,21

0 $861,550Cost Per Ton of FRAC Mixture $132 $149

Cost Comparison

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

4 6 8 10 12 14 16

EquivalentUniform

Ann

ualCost

Service Life (Years)

Jackson Hole

Jackson Hole with Fibers

Sheridan Co

Sheridan Co with Fibers

Cost Comparison

Project in ColumbiaAutopistas del Café

Downhill Curves (Location: Pereira)

• Downhill curve with high trafficvolume

• Raveling problems occur within 2month and yearly repairs whererequired (2009, 2010 & 2011);

• In January 2013, FORTA FI wasused with a dense HMA (“MDC2”) and a conventional 60 70binder;– Work done on 2 curves: milled

5cm and repaved with 5cm;– 16 month later, both curves are

in excellent condition;– No cracks, no raveling

Direccióndeltráfico

Bus stop(Medellín)

• Bus stop on a downhillalongside Parque VillaHermosa;

• Severe rutting within 2years; (May 2013)

• Work done: milling of15cm and paving a7.5cm base MSC 1 and a7.5 cm surface course ofMDC 2 with FORTA FI;

• 1 Year later road has norutting

Devimed PeajeDeviMed highway (Medellin to Bogota): already paved more than 50km x 9m wide by 6 cm average.

Toll booth, Peaje Guarne is one of the most congested roads out of Medellin. Problems: reflective cracking

Paved with FORTA-FI, 7 cm thick by PAVIMENTAR

Other on going research

Thank You!

kaloush@asu.edu

kaloush@asu.edu