Beruflich Dokumente
Kultur Dokumente
EDWIN R. SCHMECKPEPER
Assistant Professor
Department of Civil Engineering
University of Idaho
Moscow, Idaho 83844-1044
CHARLES H. GOODSPEED
Associate Professor
Department of Civil Engineering
University of New Hampshire
Durham, New Hampshire 03824
(Received July 15, 1993)
(Revised October 11, 1993)
ABSTRACT: The objective of this paper was to evaluate the suitability of Fiber- Reinforced
Plastic (FRP) grids for use as a structural reinforcement in slab-type concrete structures. The
behavior of concrete beams reinforced with FRP grids was experimentally investigated. Two
different types of FRP grids were tested; the first type used carbon fibers, the second type used
a mixture of carbon and E-Glass fibers. The mechanical properties of these two FRP grids were
determined. For each of the two types of reinforcement, five concrete beams were tested to
failure. The flexural behavior, as characterized by the load-deflection response, was monitored
throughout the tests. The results from the flexural tests on FRP reinforced concrete beams
showed that the failure mode, measured deflections and ultimate loads were consistent with
predictions.
INTRODUCTION
as slabs or walls. The use of FRP grids solves the problems of poor bond per-
formance by developing bond through direct concrete bearing on members which are
placed transverse to the longitudinal axis of the main reinforcing members [10]. In
addition to solving the problems with poor bond, the use of grids, rather than
individual bars, may result in fabrication and installation efficiencies. When made
into a two dimensional structure, these grids are particularly suited for slab-type
structures, such as bridge decks or wall panels, which do not require shear
reinforcement. A typical two-dimensional FRP grid is shown in Figure 1.
EXPERIMENTAL STUDY
The research presented in this paper has been directed towards determining the
performance characteristics of concrete beams reinforced with FRP grids. This work
is one of the steps required to determine the suitability of the FRP material for
structural applications.
FRP Grids
Fiber Reinforced Plastic (FRP) composites are composed of relatively high-
strength parallel fibers enclosed in a resin matrix, which binds the fibers into a single
structure. The resin matrix provides the means to transfer applied stresses to the
fibers and protects the fibers from moisture and deleterious interactions with the
environment, such as oxidation or corrosion.
The longitudinal and transverse members of the FRP Grids used in this study were
fabricated using a process in which bundles of carbon or glass fiber filaments were
impregnated with a vinyl ester resin, and then woven in two- or threedimensional
patterns to form the bars of the reinforcement grid. The finished grids were then
pressed between heated steel plates which flatten the upper and lower surfaces of the
grids. The result of this “built-up” fabrication process was that while the fiber content
of the grids was accurately controlled, the bars which make up the grids do not have
a uniform cross section. As shown in Figure 2, the bars in the grid are smooth on the
top and bottom, but have irregular sides.
Ultimate Young’s
Modulus Maximum
Strength Elongation
MPa Texture
(E) €
Density ult
MPa %
Fiber Strand (ksi) (e) (T)
(ksi)
Type Type g/cm3 g/1000 m
H-S tow 4820 1.82 800 230,300 1.4
Carbon (700) (33,400)
E-Glass roving 1380 2.54 2220 72,400 2.0
(200) (10,500)
H19 FRP bars. The bars of both of these grids were designed to have the same
ultimate tension capacity as a 19 mm Grade 60 steel bar. The FRP grids used in this
test program were produced with equal sized longitudinal and transverse bars,
equally spaced in both directions. The centerline to centerline bar spacing was 100
mm (4 "). The fiber areas and relative proportions of carbon and glass fibers are
shown in Table 2.
Material Composition
The mechanical properties of the individual bars in the FRP grids are dependent
upon the area of fibers and the ratio of fibers to resin. Material composition, or
“Burn-Off”, tests were conducted to determine their volume and weight fractions of
the two different types of fiber-reinforced plastic composites used in this test
program [12]. As a control, tests were simultaneously conducted on samples of the
vinyl-ester resin used in the manufacture of the FRP grids. The results of the material
composition tests, shown in Table 3, were used to calculate the theoretical section
properties of the FRP grids.
Section Properties
The gross cross sectional properties were determined by averaging the results
from repeated volumetric measurements on samples from the different types of
Table 2. Fiber area of bars in FRP grids.
Area
Area E-
% % Carbon sq. Glass
Carbon E-Glass mm (sq. in) sq. mm
Bar Size Tows Rovings (sq. in)
H19 34 136 14.9 118.9
(0.023) (0.184)
C19 135 0 59.3 0
(0.092) (0)
reinforcing grids. There were measurable differences among grids of the same type
and size. For example, the depth of the H19 reinforcing grids varied from 17 mm to
25 mm, averaging 20 mm. Figure 2 shows the presence of voids and the uneven
distribution of fibers caused by the manufacturing process. In addition, the carbon
fibers were not uniformly distributed throughout the bars of the H-Type grids. The
fabrication process created bundles of carbon fibers surrounded by bundles of glass
fibers. The two different fiber types were not intermixed.
The theoretical cross sectional area of the solids was then determined using the
calculated cross sectional fiber area and the relative volume fraction for each
component. The cross sectional area of each fiber component is calculated as
follows: where
Note that the fiber manufacturers refer to a bundle of parallel glass fibers as a
“Roving”, and refer to a bundle of continuous carbon filaments as a “Tow”.
Using the specific gravities of each component, the theoretical cross sectional area
of the bars of each grid type was calculated. The measured and theoretical cross
sectional properties are shown in Table 4. A comparison of the measured area to the
theoretical area indicates that the H19 type grids contain voids adding as much as an
additional 20% to the apparent cross-sectional area. For this reason, the theoretical
cross sectional area was used for all stress calculations [13].
Mechanical Properties
The mechanical properties of the bars which make up the FRP reinforcing grids
were determined using axial tensile tests. These tests were conducted using
the full cross section of the FRP bars as per ASTM D3039 [14]. The strains in the
reinforcing material were recorded using an extensometer and electric resistance
strain gages placed upon opposing faces of the test coupons, the measured strains
were then averaged to determine the modulus of elasticity for the sample. As shown
in Figure 3, the bars did not fail along a “failure plane”, but rather at random
locations along the test specimen, exhibiting the characteristic behavior of continuous
fiber-reinforced composites [15]. The stress-strain relationships for the FRP and steel
reinforcing bars are shown in Figure 4.
Notes: 1) Beam Depth “d” measured from top of beam to center of reinforcement. 2) For ail beams:
Width = 300 mm (12"), and Height = 200 mm (8")-
h
0.025 m- 0.2 m
w
0.3 m
Figure 5. Test beam cross sections.
The beams reinforced with the C19 grids were designed to have the same rein-
forcement rigidity as the steel reinforced beams. Note that in order to obtain the same
axial stiffness as the steel reinforcement, the C-Type FRP area was 2.4 times as great
as the steel area. However, the weight of the C-Type FRP reinforcement was less
than one half of the equivalent steel weight.
The axial rigidity of the H-Type reinforcements was only 63% of that of the steel
reinforcement. The area of reinforcement in the beams reinforced with the H19 grids,
which was limited due to concrete placing considerations, was three times that of the
steel reinforced beam. While the weight of the H-Type FRP reinforcement was only
67 % of the steel weight. In order for the axial rigidity of the H-Type reinforcements
to be equal to that of the steel reinforcements, the area of the H-Type FRP would be
required to have been 4.8 times greater than that of the steel area. In this case, the
weight of the H-Type FRP reinforcements would be 1.07 times that of the equivalent
steel reinforcement.
Test Setup
The test beams were loaded monotonically to failure in four point symmetrical
bending, using a 500 kN (120 Kip) hydraulic jack and 445 kN (100 Kip) load cell as
shown in the Figure 6. The test beams were simply supported on a convex
bearing plate at one end of the beam and a concave plate with roller at the other end
of the beam. Deflection gauges were placed at midspan, at load points, and midway
between the supports and the load points on each specimen. Deflections were
measured and recorded after each load increment.
Test Results
The average test results are summarized in Table 6. The measured maximum
loads compared favorably to the calculated theoretical maximum loads.
Figure 7 shows the crack pattern in a steel reinforced beam subjected to a load of
17.8 kN (4 Kips) [18]. Figure 8, shows the crack pattern in a FRP grid reinforced
beam at the same load level. As predicted the crack pattern in the FRP reinforced
beams occurred at the location of the transverse bars of the FRP grid. However, as
shown in these two photographs, at any given load level the number
of cracks and their apparent sizes were not significantly greater in the FRP reinforced
beams than in the steel reinforced control beams.
The Load vs. Midspan Deflection results for the test beams are shown in Figure 9.
In the steel reinforced beam the load gradually increased up to an average value of
approximately 22 kN. Additional deflections resulted in little or no increase in load.
This change in slope occurred when the steel reinforcement began yielding. The
beams would not collapse until either the steel or the concrete could no longer
accommodate additional strain. The deflected shape of the steel reinforced beam at
ultimate load is shown in Figure 10. This type of elastic-plastic behavior, which is
referred to as “Tension Reinforcement Yielding”, is generally preferred to a purely
brittle failure due to the energy being dissipated and since the increasing deflections
provide some warning prior to collapse.
In contrast, while the FRP reinforced beams exhibited significantly higher ultimate
loads than the steel reinforced beams, they did not exhibit the same type of behavior.
As shown in Figure 11, which shows an FRP reinforced beam at failure, the concrete
strength limited the ultimate load in the FRP reinforced beams. The behavior of the
FRP reinforced beams was characteristic of brittle materials. However, as indicated
in Table 6, unlike the reinforcement in the steel reinforced beams, the FRP grids
were loaded significantly below their ultimate capacity. In addition, Figure 9 also
shows that energy dissipation in the FRP reinforced beams, as measured by the area
under the load vs. deflection curve, was comparable to that of the steel reinforced
beams.
Figure 12. Measured and predicted load vs. deflection for beam C-1.
Deflections were calculated using the ACI 318 equations derived by Branson and
the CEB bilinear method [19]. As shown in Figure 12, the deflections predicted by
the CEB bilinear method and the ACI method compare favorably with the measured
deflections.
Based upon statistical data, the ACI code for concrete design limits the allowable
live load on a steel reinforced concrete beam by using load magnification factors to
reflect the variability of the load, and strength-reduction factors to account for
construction tolerances and uncertainties in materials. These load magnification
factors and strength-reduction factors result in a factor of safety which typically
ranges between 1.55 and 2.4. The strength-reduction factors for FRP reinforced
concrete have not been developed.
However, in addition to strength requirements, the ACI code also includes
serviceability requirements which limit live load deflections to a maximum of the
l/180th of the beam length. Therefore, the maximum allowable deflection for the test
beams used in this study would be (3.6 m/180) = 20 mm. At this deflection the steel-
reinforced beams resisted an average load of 19.1 kN (4.3 Kips), the C-Type
reinforced beams resisted an average load of 19.0 kN (4.3 Kips), and the H-iype
reinforced beams resisted an average load of 11.2 kN (2.6 Kips).
Using the load at the maximum deflection as the allowable live load, 19.0 kN, the
C-Type reinforced beams C2, C3, and C5 would have a factor of safety of 4.1 with
respect to the maximum test load of 79.2 kN. Similarly, the H-Type reinforced beams
H2, H3, and H5 would have an allowable live load of 11.4 kN, and a factor of safety
of 4.8 with respect to the maximum test load of 54.8 kN. The factors of safety for
both the C-Type and the H-Type FRP reinforced concrete beams are greater than
those normally provided in steel-reinforced concrete beams.
DISCUSSION OF RESULTS
The ultimate capacity of the steel-reinforced beam was limited by yielding in the
CONCLUSION
FRP grids are a potentially viable replacement for steel-in-concrete reinforcement.
Static tests have indicated that the flexural behavior of FRP reinforced structures can
be predicted such that serviceability requirements are satisfied. Satisfactory static
performance is only the first step in the long process leading towards widespread
acceptance of FRP as a reinforcement for concrete. Ongoing research, being
conducted at the University of New Hampshire and the University of Idaho, is
investigating long term effects on full scale FRP reinforced bridge deck panels [20]
and precast cladding panels. In addition, research on FRP reinforced concrete beams
subjected to adverse environments is being conducted as part of the Graduate
Research Fellow program at the FHWA’s Turner- Fairbank Highway Research
Center.
The acceptance of FRP as a reinforcement for concrete will be influenced by its
lack of “Tension Reinforcement Yielding” and its low modulus of elasticity. The low
modulus of elasticity will result in larger deflections than are found in a concrete
member reinforced with an equal strength steel reinforcement. The brittle behavior
and low modulus of elasticity of FRP reinforcements requires greater consideration
of concrete shear, compression, and deflections than in steel-reinforced concrete.
However, since the FRP materials behave in a predictable manner and are not
susceptible to electrochemical corrosion, the use of FRP as a concrete reinforcement
will continue to increase.
ACKNOWLEDGEMENTS
Portions of the work described formed part of the doctoral dissertation of the first
author, at the University of New Hampshire, under the direction of the second author.
In addition, portions of the work described were supported by the
Federal Highway Administration Graduate Research Fellowship Program, under the
direction of Eric Munley of the FHWA Structures Division.