FRP Improves Bridge Deck Life

Fiber-reinforced polymer shortened construction time and lengthened deck life in West Virginia.

by David Holman Deitz

The use of fiber-reinforced polymers has become a more viable method for improving the life-cycle costs of structural components. This is particularly true for bridge decks in colder climates, such as West Virginia, where deicing salts are used for snow and ice removal in winter months.

The deicing salts increase the rate of reinforcement corrosion for more typical reinforced concrete bridge decks, eventually resulting in bridge deck replacement.

Since FRP bridge decks are immune to this type of corrosion, they can potentially increase the service life of the bridge deck.

Other advantages of FRP bridge decks include their light weight and a reduction in bridge deck construction time.

Disadvantages of FRP bridge decks include low modulus of elasticity translating into increased deflections, greater initial expense, and unfamiliarity to many engineers and constructors.

The site

Howell’s Mill Bridge carries County Route 1 over the Mud River in Cabell County, West Virginia, near Huntington. The welded plate girder bridge consists of two 120-foot continuous spans and supports the AASHTO LRFD HL-93 live loading. The bridge deck is 32.5-feet wide and uses reinforced concrete barriers. A 0.75-inch asphalt overlay was selected as a riding surface. Howell’s Mill is a jointless bridge, continuous over the pier. Abutments one and two are integral and semi-integral, respectively.

The West Virginia Department of Transportation’s goal in selecting Howell’s Mill Bridge for an FRP deck was to move towards making FRP a more common material for use in bridge applications. To accomplish this objective, WVDOT selected Howell’s Mill because it was a normal bridge design application, rather that an ideal bridge for an FRP bridge deck. In addition, they selected an engineering firm with experience in the design of FRP structural components and one that was also involved in more typical bridge designs in West Virginia.

Design considerations

Initially, the project team wanted a common design that would provide several different FRP bridge deck panel manufacturers the opportunity to supply the deck panels, keeping with the goal of having the project proceed as a normal bridge project. However, the characteristics of the available deck panels, such as longitudinal stiffness, method of connecting the deck to the girders, deck thickness, and so on, vary too widely to allow for a common design. It was decided that a specific deck panel would have to be selected prior to initiation of the design. The panel selection was based largely on the amount of published test data available for the panels that was directly applicable to the project.

The transverse girder spacing of 6.75 feet was selected for several reasons. A review of previous projects showed that the FRP deck fabricator’s panels could span 8 feet. However, there were few applications of the deck with girders spaced greater than this. Rather than using a spacing of greater than 8 feet, the project team decided to go with the 6.75-foot girder spacing.

Second, a hydraulic analysis limited the height of the girders to 3.5 feet. The 6.75-foot girder spacing allowed the girders to meet this height requirement and still be designed with more common plate sizes.

Once the girder spacing was selected and agreed upon by the design engineer and the deck fabricator, it was up to the deck panel fabricator to ensure that the deck panels could support the required vehicle loading. Though the bridge design is based on the AASHTO LRFD HL-93 loading, the design team felt that the deck should be capable of supporting an additional load case equal to 1.25 times the HL-93 design truck. This is similar to HS-25 truck loading currently used by many transportation departments. It is possible a single-wheel load could control the design of the deck, and the project team agreed that providing this overload check was prudent.

The AASHTO LRFD code does not provide specific guidelines for the distribution of wheel/axle loads to girders in the transverse direction for bridges constructed with FRP deck panels. For cases not specifically listed, AASHTO LRFD specifies that the lever rule be used to distribute vertical deck loads to the girders. The lever rule assumes that the deck is simply supported between girders, conservatively neglecting the continuous beam behavior of the deck. The design team believed this type of distribution was appropriate for the FRP deck panels.

Composite action between the girders and the FRP deck panels was conservatively neglected in the design of the girders for the applicable AASHTO LRFD strength requirements even though the FRP deck panels are connected to the girders by shear studs. Neglecting the composite action would ensure that the girders could support the required design loads without relying on the strength of the FRP deck. Composite action was considered for checks of the girders against the service, fatigue, and deflection requirements provided in the AASHTO LRFD code. An effective stiffness for the deck panels was taken from laboratory testing performed by the deck panel manufacturer.

Even though the bridge deck was considered to be non-composite for strength design, it is still necessary to check the shear studs for strength requirements. The shear studs offer resistance to the applied loading and a stud failure could result if an adequate number of studs are not provided. Of course, the shear studs were checked for the fatigue requirements since the bridge was designed as a composite structure for this condition. This check would still be necessary even if the bridge was designed as non-composite for fatigue. The studs will be present to connect the bridge deck to the girders and, regardless of the design assumptions, will act to transfer stresses between the two components.

Detailing

Howell’s Mill Bridge lies in a tangent section of County Route 1. Though the project’s typical roadway cross section provided for a crowned pavement surface in tangent sections, engineers decided to maintain a constant 2% slope through the bridge. Fortunately, a right hand curve was located immediately off the bridge allowing the 2% cross slope to be implemented without resulting in driver discomfort.

If the deck cross section were crowned, the deck panels would have been placed horizontally with 0% cross slope, and the crown would be built by varying the thickness of the asphalt overlay. For Howell’s Mill, the overlay would have varied by a thickness of almost 4 inches. This would have increased the dead load of the bridge significantly, resulting in a less efficient design.

Another benefit of the constant 2% cross slope is that water permeating the overlay can drain more readily since the panels are sloped rather than horizontal. Other FRP deck panel manufacturers can provide a roadway crown by other means, such as building the crown into the deck itself or by breaking the deck up into two pieces and splicing the pieces in the field.

Typical stay-in-place form angles of the girders are used to form the haunch and support the FRP deck panels until the haunches are grouted. The project specifications did not allow welding of stay-in-place form angles to the top flange of the girder in any locations. The stay-in-place angles are attached to the girders using a common form support angle, or strap, spanning above the top flange. Once the deck panels are positioned on top of the stay-in-place forms, shear studs are attached to the top flange of the girder through stud pockets located along the top of each girder at a 2-foot spacing. The stud pocket size is adequate to attach the shear studs using a standard stud gun. Finally, non-shrink grout is poured through the stud pockets to form the haunch.

The connection of the deck to the abutment was made by passing bent #4 bars through the end void of the deck panel. The end void was filled with concrete during construction of the backwall. It is important that the FRP deck panel cut outs were large enough to allow the #4 rebar to pass through the end void and to allow concrete to readily fill the end void during construction.

The deck panels arrived on site with a skid-resistant top surface. This skid-resistant surface both increases the bond between the deck and the overlay and provides a non-slip surface during construction. A 0.75-inch polymer asphalt overlay was selected for the bridge. The polymer asphalt has a higher flexibility than more conventional asphalt concrete. The increased flexibility is expected to increase the overlay’s resistance to cracking in the negative moment region over the pier and to cracking caused by thermal stresses.

Cost comparison

A cost comparison was made between the bridge with the FRP deck panels and two conventional bridge types that would have been suitable for the site. The first alternate is an adjacent prestressed concrete box beam bridge made up of eight 42-inch-deep by 48-inch-wide box beams with a 5-inch-thick composite deck. The second is a welded steel plate girder bridge with the same girder web depth and spacing as the proposed bridge, but supporting an 8-inch-thick reinforced concrete deck.

Costs considered are for the superstructures only. Substructures for the three bridges would be approximately the same. Costs studied for the proposed bridge are actual project bid costs from the contractor who was awarded the job. The costs for the two more conventional bridges were estimated based on West Virginia average unit bid prices and information provided by the contractor who bid the proposed FRP bridge. Results of the study show that the initial cost of the proposed bridge with the FRP bridge deck is approximately 2.4 times greater than the adjacent prestressed concrete box beam bridge.

The bridge deck was constructed in a shorter amount of time than the two other bridges. The fabricator estimated that the deck panels could be placed in as little as two to three days, compared to the time required to form, place concrete, and allow concrete to cure for the other two bridge types. Though the time savings to the contractor are included in the costs, the inconvenience to the public associated with the longer construction time frame is not.

In addition, the cost comparison does not account for the longer service life expected from the FRP deck panels compared to the conventional reinforced concrete decks.

Reprinted from Better Roads Magazine
November 2002

Click Here to return to article index