8
The deck prototype dimensions were 1.828 6.100 0.203 m (72 240 8 in.). The
fiber reinforcement was E-glass continuous roving and multi-axial stitched E-glass fabric
(90/45) (BTI, TH4000 / THX1501). The resin matrix was vinyl ester (Reichhold, Atlac 580-
05),
2.5 FRP Bridge Deck Fabricated by the Contact Molding Hand Lay-Up Process
Bridge #5 was fabricated based on the concept of sandwich construction using a low-density
honeycomb core sandwiched between two contact-molded hand lay-up face sheets (Figure 2).
The dimensions of the deck prototype were 1.828 6.100 0.203 m (72 240 8 in.).The resin
matrix was made of isophthalic / terephthalic polymer resin (AOC, Vibrin F457-BRP-25), which
was reinforced with bi-axial E-glass fabric (0/90) and mat (BTI, CM4810 & CCC A118).
3. EXPERIMENTAL SETUP
3.1 Self-Reacting Loading Frame
Each deck prototype was placed on three W36x182 steel girders, resulting in a continuous two-
span bridge structure Figure 3. A self-reacting steel test frame was designed for a maximum load
capacity of 270 kN (60,000 lb). The maximum deflection of the steel transverse beam was
limited to less than 0.25 mm (0.01 in.). Two actuators mounted on the two cross arms of this load
frame applied the load through two steel plates of 228 559 mm (9 22 in.) centered with
respect to the supports of the span, which simulate the AASHTO HS20-44 design truck wheel
load print. The long dimension of the plate was perpendicular to the girder direction. The inner
long edge of the plate was 178 mm (7 in.) away from the center of the deck. An elastomeric pad
was placed between each steel plate and the prototypes to provide uniform pressure that
simulates the wheel load action (Figure 4). The setup induced a positive bending moment under
the load and a negative bending moment on the central support.
Each test deck was instrumented with strain gages (EA5-500BL-350, MicroMeasurements),
thermocouples, and linear voltage differential transducers (LVDTs), which were supported by an
independent steel frame. The LVDTs were used to measure deflections on the top and bottom
surfaces of the decks. Seven thermocouples were used on four sides of each deck, and one
thermocouple was used for ambient temperature (Figure 4). The locations of the strain gages
bonded to the bottom deck surface were symmetrical with respect to the numerically matching
strain gages bonded to the top surface of decks. The complete instrumentation layout is shown in
Figure 5.
4. FATIGUE TEST PROCEDURE
The fatigue evaluation procedure consisted of applying four million simulated wheel load cycles
at 30C (22F) and another four million cycles at 50C (122F) (See Figure 6). The fatigue
performance of each FRP deck prototype was compared with the response of the conventional
reinforced-concrete deck. The fatigue load range was computed for an AASHTO HS20-44 truck
wheel with impact and dead load. A computed load of 115 kN (26,000 lb) was applied