Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Research Home   |   Pavements Home
Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-06-106
Date: September 2009

Design and Evaluation of Jointed Plain Concrete Pavement With Fiber Reinforced Polymer Dowels

Previous | Table of Contents | Next

CHAPTER 3. MATERIALS, EQUIPMENT, AND LABORATORY TESTING PROCEDURES

INTRODUCTION

Mechanical properties of dowel bars affect the behavior and performance of JPCP provided with FRP and steel dowels. This chapter discusses mechanical properties of FRP and steel dowels and test setup in the Major Units Laboratory of WVU. Seven full-scale JPCPs (two with a dimension of 30.48 by 30.48 by 304.8 cm (12 by 12 by 120 inches) and five with a dimension of 30.48 by 27.94 by 304.8 cm (12 by 11 by 120 inches) were cast with simulated contraction or sawcut joints. The specimens were subjected to both static load and fatigue load with a frequency of about 4.0 Hz.

MATERIAL PROPERTIES

Class K concrete (25.856 and 31.026 MPa (3,750 and 4,500 psi) and two different diameters (3.81 and 2.54 cm (1.5 and 1.0 inch)) of GFRP dowel and epoxy-coated steel dowels were used for casting the concrete pavements in the structural laboratory. Relevant material properties are provided in this section.

The two slabs that were used for preliminary testing consisted of fc ′ of 25.856 MPa (3,750 psi); the remainder of the five slabs consisted of fc ′ of 31.026 MPa (4,500 psi), as found through cylinder tests.

GFRP Dowels

GFRP dowels provided by independent manufacturers are pultruded with continuous E glass filaments and polyester resin. Typically, filaments are drawn through a resin bath, sized by an appropriate die, to form the dowel bar. An ultraviolet inhibitor is added to the resin to resist effects of sunlight. Dowels with 3.81- and 2.54-cm (1.5- and 1.0-inch) diameters were used in this project. Other researchers have commented on using 4.45-cm (1.75-inch)-diameter FRP dowel bars instead of 3.81-cm (1.5-inch)-diameter dowel bars. However, researchers have noted that increased bar diameter result in larger RD. Hence, it was decided to use lower diameter bars in this research.(9)

Figure 3 shows two types of GFRP dowels used in this research. Table 1 and table 2 show properties of FRP bars listed by the manufacturer.

Figure 3. Photo. FRP dowels (2.54 and 3.81 cm (1.0 and 1.5 inches) in diameter). 
The photo shows two circular glass fiber reinforced polymer (FRP) dowel bars used in this research, with diameters of 2.54 and 3.81 cm (1.0 and 1.5 inches) and a length of 45.72 cm (18 inches). They were manufactured through a pultrusion process using continuous E-glass filaments and polyester resins.
Figure 3. Photo. FRP dowels (2.54 and 3.81 cm (1.0 and 1.5 inches) in diameter).

Table 1. Modulus of elasticity (MOE) test results of FRP rod specimens—ASTM D3916

Type Nominal Diameter (inches) Actual Diameter (inches) Area Square (inches) Ultimate Load (lbs) Tensile Stress (psi) Elongation (percent) MOE (Msi)
A 1.0000 0.9713 0.7394 66,539 89,986 0.094 6.0
1.0000 0.9700 0.7390 65,204 88,235 0.091 6.0
Average 1.0000 0.9707 0.7392 65,872 89,111 0.093 6.0
B 1.5000 1.4932 1.7512 134,440 76,772 0.219 5.2
1.5000 1.4933 1.7514 137,000 78,226 0.179 5.0
Average 1.5000 1.4933 1.7513 135,720 77,499 0.199 5.1

1 inch = 2.54 cm
1 inch2 = 2.54 cm2
1 psi = 0.006895 MPa
1 million psi (Msi) = 6.895 GPA

Table 2. Shear test results of FRP rod specimens—single-shear fixture.

Type Nominal Diameter (inches) Actual Diameter (inches) Area Square (inches) Ultimate Load (lbs) Shear Stress (psi)
A 1.0000 0.9703 0.7390 28,600 38,700
1.0000 0.9700 0.7380 30,300 41,060
1.0000 0.9702 0.7390 27,200 36,810
Average 1.0000 0.9702 0.7387 28,700 38,857
B 1.5000 1.4923 1.7490 33,982 19,429
1.5000 1.4935 1.7519 31,782 18,142
Average 1.5000 1.4929 1.7505 32,882 18,786

1 inch = 2.54 cm
1 inch2 = 2.54 cm2
1 lb = 0.454 kg
1 psi = 0.006895 MPa

Steel Dowels

The steel dowels used in this research, shown in figure 4, were grade 40 plain uncoated steel or epoxy-coated steel. Dowels with 3.81- and 2.54-cm (1.5- and 1.0-inch) diameters were used for laboratory experiments and field installation.

Figure 4. Photo. Steel dowels 3.81 and 2.54 cm (1.5 and 1.0 inches) in diameter. The photo shows two circular steel dowel bars of 3.81 and 2.54 cm (1.5 and 1.0 inches) diameter with a length of 45.72 cm (18 inches). These bars are grade 40 plain uncoated steel or epoxy coated steel and have been used for laboratory experiments and field installations.
Figure 4. Photo. Steel dowels 3.81 and 2.54 cm (1.5 and 1.0 inches) in diameter.

Concrete

Class K ready-mixed concrete was used for laboratory slab casting. The compressive strength of the concrete was 25.856 MPa (3,750 psi) for 30.48-cm (12-inch)-thick slabs and 31.026 MPa (4,500 psi) for 27.94-cm (11-inch)-thick slabs. Concrete was poured in formwork, and forms were removed after 24 hours. The concrete beams were cured 28 days by wet burlap and plastic sheet covering.

Base

To simulate a stiff subgrade used in the field, a base layer of limestone aggregates was prepared and compacted to a depth of 40.64 cm (16 inches) in a wood-framed bin shown in figure 5. The modulus of subgrade reaction k was determined from plate loading tests in the laboratory. Based on the measured values, an average of 11.072 kg/cm3 (400 lbs/pci) was used for all tests employing aggregate base. The value was also applied to the theoretical calculations.

FORMWORK

Figure 5 and figure 6 show wood formworks made for casting concrete slabs in the laboratory.

TEST SETUP

Specimen Fabrication

Seven different slabs were cast with FRP and steel dowels using different spacing and diameters.

This photo shows wood formwork used for casting jointed concrete slabs with dowel bars in the laboratory. The photo shows dowel and supporting dowel bar, dowel basket, and positioning of steel plate in the middle of the formwork.
Figure 5. Photo. Wood formwork.

A three-dimensional drawing of the wood formwork is shown in this photo with a half length of 152.4 cm (60 inches), width of 30.48 cm (12 inches), and height of 27.94 cm (11 inches). A 0.635-cm (0.25 inch)-thick steel plate with a height of 9.525 cm (3.75 inches) is shown at the joint location between two adjacent slabs.

1 inch = 2.54 cm
Figure 6. Diagram. Dimensions of formwork.

Material Preparations

FRP and steel dowels were prepared with slots at locations where strain gauges were going to be placed.

Uniaxial strain gauges were bonded to dowels at the slot position shown in figure 7. Strain gauges were protected using M-Coat-J (polysulfide liquid polymer).

This drawing shows the side view and the top view of the circular dowel bars with slot positions on both sides of the bar at mid-length where the uniaxial strain gauges will be mounted on the dowel bars.

1 inch = 2.54 cm
Figure 7. Diagram. Trimmed dowel bar.

Pavement Slab Casting

The inside walls of the wood formworks were oiled so that the concrete pavement slabs could be easily demolded. Strain gauge instrumented dowels were then placed in dowel baskets to properly center them in the joint (figure 8). A steel plate with 0.635-cm (0.25-inch) thickness was placed in the middle to simulate a contraction or sawcut joint in the concrete pavement (figure 5, figure 6, and figure 8).

Figure 9 through figure 12 show the slab casting. Class K concrete conforming to the West Virginia Department of Transportation (WVDOT), Department of Highway (DOH) specification was used for casting (figure 11). Concrete cylinders were cast simultaneously to obtain concrete compressive strength. Twenty-four hours after casting the beams and cylinders, curing was carried out using wet burlaps.

This photo shows inside walls of wood formworks that have been oiled for easy demolding of the concrete pavement slabs. It also shows dowel bars instrumented with strain gauges, which are supported on dowel baskets and centered equally on both sides of the steel plate. A steel plate with the dimensions of the slab joint is placed in the middle of the slab to be lifted off after concrete curing to simulate a contraction or sawcut joint in the concrete pavement.
Figure 8. Photo. Instrumented dowels and steel plate positioned in the wood formwork.

 

This photo shows ready-mixed concrete being unloaded from the mixing truck through chutes and poured into the wooden formwork. Three people are spreading and vibrating the concrete mix in the formwork.
Figure 9. Photo. Placing concrete into the formwork.

 

This photo shows a dowel bar with strain gauge and wire attachment placed in the wooden formwork as it is being covered with the concrete pouring down from the chute of a concrete mixer.
Figure 10. Photo. Dowel being covered by concrete.

 

This photo shows graduate students casting concrete cylinders and beams using plastic cylinders and wood formwork in a laboratory.
Figure 11. Photo. Casting concrete cylinders.

 

This photo shows surface-finished jointed concrete pavement slabs just after they are cast in wood formwork. Hooks are provided on both sides across the joints, and strain gauge wires coming out of the slab from the dowel bars are visible.
Figure 12. Photo. Surface finished specimens.

Test Specimens

Two specimens with dimensions of 30.48 by 30.48 by 304.8 cm (12 by 12 by 120 inches) (figure 13) were cast for preliminary tests. Only FRP dowels were used as load transfer devices in these two specimens.

A three-dimensional diagram of the concrete slab shows 152.4 cm (60 inches) of its length on one side of the joint and a partial length of the slab on the other side. The slab is 30.48 cm (12 inches) in height and width. A 0.635-cm (0.25-inch)-thick steel plate that is 30.48 cm (12 inches) wide and 9.525 cm (3.75 inches) high is shown at the joint location between two adjacent slabs. One fiber reinforced polymer dowel is placed at a height of 15.24 cm (6.0 inches) from the top surface with equal projection on both sides of the joint.

1 inch = 2.54 cm
Figure 13. Diagram. Concrete slabs for preliminary tests.

Five different concrete slabs were cast with FRP and steel dowels with different spacings and diameters (table 3). Two embeddable strain gauges were positioned vertically on both sides of a dowel across the joint to measure concrete strain at loaded side and unloaded side. Details of concrete specimens and dowels are provided in figure 14, figure 15, and table 3.

Table 3. Dowel details in specimens.

Specimen Number Slab Depth (inches) Dowel Material Dowel Diameter (inches) Spacing (inches) c/c Number of Dowels in Each Specimen

Preliminary group

PG-1 12 FRP 1.5 12 1
PG-2 FRP 1.5

Group 1

1 11 Steel 1.0 6 2
2 FRP 1.0
3 FRP 1.5

Group 2

4 11 FRP 1.5 12 1
5 Steel 1.5

c/c = Center to center
1 inch = 2.54 cm

 

Figure 4. Photo. Steel dowels 3.81 and 2.54 cm (1.5 and 1.0 inches) in diameter.

specimen is symmetrical about the joint plan (unit: inch)
1 inch = 2.54 cm
Figure 14. Diagram. Concrete slabs containing two dowels.

 

 

A three-dimensional diagram of this concrete slab is shown with a length of 152.4 cm
specimen is symmetrical about the joint plan (unit: inch)
1 inch = 2.54 cm
Figure 15. Diagram. Concrete slabs containing only one dowel.

Test Setup and Instrumentation

Jointed concrete slabs were placed on an aggregate base inside a wooden box to simulate field conditions. The base was 15.24 cm (6 inches) high, 45.72 cm (18 inches) wide, and 304.8 cm (120 inches) in length. The modulus of subgrade reaction k was obtained from tests on this base through load application on a standard steel plate.(10)

Pavement load was applied on one side of the joint (figure 16 and figure 17) using a 244.65-kN (55-kip) hydraulic actuator system through a controller. A 0.635-cm (0.25-inch)-thick steel plate 25.4 by 40.64 cm (10 by 16 inches) was connected to the actuator to simulate the load from a dual tire wheel load. An additional neoprene pad was used between the steel plate and the concrete surface to prevent any local damage during the test. Two calibrated linear variable differential transformers (LVDTs) were used to measure joint deflections at the loaded and unloaded sides of the joint. Strain gauges, LVDTs, and the load cell from the actuator were connected to a data acquisition system to automatically record data during the tests.

This photo shows the experimental setup with the slabs being positioned on top of an aggregate-filled rectangular wooden bin representing the subgrade. Load is being applied to one of the slabs near the joint with a 244.652 kN (55 kip) hydraulic actuator system suspended from a loading frame.
Figure 16. Photo. Experimental setup.

 

This photo shows a close-up view of a slab at the joint location where two linear variable differential transformers (LVDTs) are used to measure slab deflections at loaded and unloaded sides of the joint as the load is applied from the hydraulic actuator system through a steel plate.
Figure 17. Photo. LVDTs positioned on both sides of the joint.

 

Static Testing

Load was applied at different increments to simulate an HS25 wheel load and/or higher load. Deflections and strains were recorded automatically by the data acquisition system. Details of the load applied on different concrete slabs are shown in table 4.

Table 4. Details of static testing.

Specimen Number Load Range
Slab number 1 0–12.5 kips
Slab number 2 0–12.5 kips
Slab number 3 0–12.5 kips
Slab number 4 0–11 kips
Slab number 5 0–11 kips

1 kip = 4.448 kN

Fatigue Testing

The load from the hydraulic actuator system was set at the required range of fatigue cycles (table 5) and was applied using a sine wave. After every 1 million cycles, a static test was conducted on the pavement system to measure strain and deflections. Details of fatigue tests are provided in table 5.

Table 5. Details of fatigue testing.

Specimen Number Load and Range (kips)
Number of cycles (million)
0–1 1–1.25 1.25–2 2–3 3–4 4–5
Slab number 1 2–12.5 2–18.75 2–18.75 2–18.75 2–18.75 2–18.75
Slab number 2 2–12.5 2–18.75 2–18.75
Slab number 3 2–12.5 2–12.5
Slab number 4 2–11 2–11 2–11 2–11 2–11 2–11
Slab number 5 2–11 2–11 2–11 2–11 2–11 2–11

— No loading at corresponding cycles.
For specimens 1, 2, and 3, an overload factor of 1.5 (i.e., 1.5 by HS25) was used to apply higher load during fatigue tests after 1 million cycles.
1 kip = 4.448 kN

It should be noted that slabs number 1 through 3 had 55.60 kN (12.5 kips), and slabs number 4 and 5 had 48.93 kN (11 kips) of loading corresponding to HS25 loading as described in chapter 6 of this report.

Previous | Table of Contents | Next