Yi BianErwin KohlerJohn Harvey

Laboratory Evaluation of Fiber-Reinforced Polymer Dowel Bars for Jointed Concrete

Pavements

TRB Annual Meeting, Washington D.C.January 24th, 2007

Introduction• LTE controls ride quality and structural life • LTE by aggregate interlock or dowel bars• Aggregate interlock reduced by

– abrasion & shrinkage– slabs thermal contraction

• Steel dowel bars are susceptible to corrosion• Low LTE cracking & faulting

UCPRC Project on Dowel Bar Retrofit

• General objective: Evaluate DBR and best options for implementation

• Work components1. APT testing2. Lab testing

– steel dowel corrosion– FRP

3. Field Live Traffic Testing 4. …Modeling (FEM, LCCA)

APT Testing

HVS testing, Ukiah

Results: – No damage to any of the DBR

joints, or loss of LTE– Greater increase in deflections at

non-DBR joints

Epoxy-coated steel dowels

HVS testing, Palmdale

Results: – No damage to any of the DBR

joints, or loss of LTE– Fatigue cracking of the slab

1. Epoxy-coated steel2. Hollow stainless steel dowels3. Fiber-reinforced polymer (FRP)

500 1 2 3 5 6 7 8 9 10 11 1213 14 15 16 17 184 19 20 21 22 23 24 4647 4849454443424140393837363534

333231

302926 282725

554 LT554 HVS553 HVS 555 LT553 LT555 HVS

7 8 9 10

15 16 17

Ukiah DBR

Epoxy

Epoxy

32 33 34

Epoxy

Hollow

Epoxy

Epoxy

35 36

Epoxy (3)

Epoxy (3)38 39

Epoxy

Epoxy

Epoxy

Epoxy

559 HVS

32 33 34 35 36 37 38 39 40 41 42 43 44 45

557 HVS 558 HVS556 HVS

Epoxy

EpoxyEpoxy

Hollow Epoxy

Epoxy

Epoxy (3)Epoxy (3)

Epoxy

Epoxy

Epoxy

Epoxy Epoxy FRP

FRP

FRP

Palmdale DBR

41 42 43

FRP

FRP

FRP

FRP

Epoxy

Epoxy

Epoxy FRP

Dowel types in Palmdale DBR1. Epoxy-coated

steel

2. Hollow stainless steel dowel

3. Fiber-reinforced polymer(FRP)

Lab: Corrosion of steel dowels

Lab testing: corrosion1. bare carbon steel2. stainless steel clad3. grout-filled hollow stainless steel4. microcomposite steel5. carbon steel coated with flexible epoxy (green)6. carbon steel coated with non-flexible epoxies (purple)7. carbon steel coated with non-flexible epoxies (gray)

Results: 1. Recommend that uncoated carbon steel

dowels not be used2. Epoxy dowels present risk of corrosion at

scrapes and the ends3. Recommend use of stainless steel clad,

hollow stainless steel, or micro-composite for locations with risk of high chloride exposure

Regular steel dowels

Epoxy-coated steel

• Defects are inevitable – pinholes, voids and mechanical scrapes & scratches

(macroscopic and microscopic)

• Localized corrosion initiated at the defects accumulated oxide will further lift the coat

Problems at joints w/corroded bars

• Corrosion products expand and lock the joints – Expansive products Fe(OH)2), (Fe3O4) and (Fe2O3)

Volume 6 times greater

• Decreased LTE due to volume reduction after the corrosion products are washed away

• Cracking concreteCorrosion spalling & transverse cracking

Lab: FRP Evaluation

FRP

• High strength-to-weight ratio • Excellent resistance to electrochemical

corrosion• Used extensively to repair and strengthen

reinforced concrete beams and columns

Caltrans’ Questions

• Are the mechanical properties of FRP dowels adequate to perform acceptably (compared to steel dowels)?

• Are the FRP mechanical characteristics negatively affected by environmental factors?

Experimental Set Up • The study consisted of evaluating the flexural and shear

properties of glass FRP dowel bars• 1.5-in diameter and 18-in long, from 2 manufacturers

Properties of the FRP dowel bars

Type A Type B

Glass fiber content 70% Min. 65%, typical 72-73%

Glass type E-type glass E-type glass

Matrix type Polyester resin Epoxy Vinyl Ester resin

Conditioning

1. Alkaline solution • glass FRP could be highly sensitive to alkaline attack • specimens submerged in alkali solution for 3 months • pH level for the alkali solution was 13.5

2. Water• Simulate high moisture content• submerged in water for 3 months in plastic tanks

3. Ultraviolet radiation• bond dissociation between fiber and matrix• bars exposed to direct sunshine for 2 months (July and August)

Results outline

• Flexural Tests–Flexural Stiffness–Flexural Strength–Flexural Fatigue

• Shear Tests–Shear Strength–Shear Fatigue

• Direct shear strength

Flexural Stiffness

1. Type B is 20% stiffer than Type A bars2. Type A bars unaffected by conditioning

processes3. Stiffness in conditioned Type B bars decreased:

- 4% for water conditioned - 6% for UV conditioned specimens

- 20°C - Avg loading freq. of 2, 6, and 10 Hz - Two replicates- COV <2%

0

5

10

15

20

25

30

35

40

45

50

Type A Type B

Stiff

ness

(GPa

)

Control Alkali conditioned Water conditioned UV conditioned

5.3x106 psi 6.4x106 psi

Effect of loading frequency, temperature, and conditioning

1. Type B is stiffer than Type A bars

2. Frequency (2 to 10 Hz) had no effect on dowel stiffness

3. Temperature effect

25

30

3540

45

50

2 Hz 6 Hz 10 HzLoading Frequency

Stif

fnes

s (G

Pa) Test at 5°C

Type B

Type A

253035404550

2 Hz 6 Hz 10 Hz

Loading Frequency

Stif

fnes

s (G

Pa) Test at 20°C

Type B

Type A

253035404550

2 Hz 6 Hz 10 Hz

Stif

fnes

s (G

Pa) Control

AlkaliWaterUV

Test at 40°C

Type B

Type A

5°C 40°C

20°C

Stiffness versus temperature

25

30

35

40

45

50

0 5 10 15 20 25 30 35 40 45Temperature (C)

Stiff

ness

(GPa

)

Control Alkali

Water UV

Type B

Type A

At 40°C bars are softer than at 20°C.

9 % for Type A 15%for Type B

Should not significantly impair the pavement’s performance:

FEM max concrete stress is mostly unaffected when dowel stiffness changes 10 – 20%

Flexural Strength

0

100

200

300

400

500

600

700

800

900

1000

Type A Type B

Flex

ural

stre

gth

(MPa

)Control Alkali conditioned Water conditioned UV conditioned

1. Type B strength is 80% greater than Type A2. 1% < COV < 4%, except for one (two replicates).3. Type B alkali conditioned bars:

• COV =19% (four replicates)• More on that

72x103 psi 130x103 psi

Damage on Type B bars after exposure to alkaline solution

• Visible cracks on one or both ends of the dowel• Cracks could be observed before any load application • Only on alkali conditioned specimens

Flexural Fatigue

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

100,000,000

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Stress ratio, S

Flex

ural

fatig

ue L

ife, n

(cyc

les)

Type A - control Type B - controlType A - alkali Type B - alkaliType A - water Type B - waterType A - UV Type B - UVType A regression Type B regressionCombined Types A & B

n = 10(12.75 – 12.74×S), R2 = 0.9475

Stiffness during flexural fatigue

0

5

10

15

20

25

30

35

40

45

0 1 2 3 4 5 6Load cycle (millions)

Elas

tic M

odlu

lus (

GPa

)

Type A - control, S=0.54 Type B - control, S=0.45Type A - control, S=0.58 Type B - control, S=0.50Type A - alkali, S=0.49 Type B - alkali, S=0.49Type A - water, S=0.57 Type B - water, S=0.48Type A - UV, S=0.56 Type B - UV, S=0.49

All tested specimens tend to fail when the stiffness drop to 15 – 20 GPa

Shear strength

0

100

200

300

400

500

600

700

800

900

1000

Type A Type B

Shea

r Stre

ngth

(MPa

) .

Control Alkali conditioned Water conditioned UV conditioned

Shear fatigue life

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

100,000,000

1,000,000,000

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Stress ratio, S

Shea

r fat

igue

life

, n (c

ycle

s) .

Type A - control Type B - controlType A - alkali Type B - alkaliType A - water Type B - waterType A - UV Type B - UVType A regression Type B regression

Type A > Type B bars in shear fatigue life

Direct shear

0

50

100

150

200

250

300

350

400

450

500

Type A Type B

Dire

ct s

hear

stre

ngth

(MP

a) .

ControlAlkali conditionedWater conditionedUV conditioned

Direct shear strength

Summary and conclusions

1. Stiffness for FRP bars:• It’s not influenced by loading frequency

(in the range of 2 to 10 Hz)

• It is affected by testing temperature (<20% at 40°C)2. Type B bars are ~20% stiffer, 80% stronger in

bending, and 100% stronger in shear than Type A bars

In bending, Type B are two times stronger than typical steel barsType A bars are ~30% stronger than steel

Summary and conclusions

3. Strength of Type B bars might be reduced by the high pH environment within the concrete slabs

– Type A bars were not affected by any of the three conditioning types

– Water and UV conditioning had no effect in either Type A or B bars

Summary and conclusions

4. Fatigue– both types of bars offer similar flexural

fatigue performance – greater number of shear cycles can be

expected from Type A bars – At low stress ratio (0.3 to 0.4), both types of

FRP bars will likely survive more than 100 millions wheel load repetitions

Thanks

Erwin KohlerUniversity of California Pavement Research Center

Project Scientist, PhDCivil and Environmental Engineering, UC-Davis

530-754-8699

ekohler@ucdavis.edu