Note: Descriptions are shown in the official language in which they were submitted.
CA 02561453 2006-09-28
- 1 -
FIBER REINFORCED THERMOPLASTIC
COMPOSITE PANEL
TECHNICAL FIELD
The present invention relates to a fiber
reinforced thermoplastic composite panel for use in the
construction of bridge decks, as an example, and wherein the
components of the panel are fabricated from commingled glass
fiber reinforced polypropylene material.
The composite panel of the present invention was
developed to replace current timber decks which are used in
the construction of short span bridges and which could be
more durable with an estimated service life of at least 75
years and which has less weight per surface area than these
timber slabs making it much easier to install and replace if
necessary while providing an excellent fatigue resistance.
BACKGROUND ART
Timbe:r slabs conventionally used in these bridges
are made of standard E-P-S logs 9197x203x4268 mm WxHxL).
The logs are typically arranged in a staggered formation
every 406 mm (clear spacing of 209 mm) A wearing surface
made of the same wood is furnished atop of the cross log
strips (203x96~:1800 mm WxHxL) are used. This makes the
total depth of the deck slab equal to 299 mm. The strips of
the wearing surface are connected to the transverse logs
using 010 bolts every 812 mm on the longitudinal direction.
A review of the available literature into FRP
composite bridge deck slabs reveal that some developments
have taken place in the last decade. Several demonstration
projects have been completed and running in several
countries, but with no mass production. The research in
North America is still in its early stage, and no field
application or demonstration has been made to date.
CA 02561453 2006-09-28
- 2 -
Luke, S., et al, "The design, installation, and
monitoring of an FRP bridge at West Mill Oxford", in
Lightweight Bridge Decks, European Bridge Engineering
Conference, Rotterdam, The Netherlands, March 27,-28, 2003,
tested a modular bridge deck made of E-glass and polyester,
the deck was made of several cells 500 mm wide each
connected together using epoxy resin. The cross section of
the cell was 225 mm deep and had triangular openings with
7.75 mm thick webs. The authors acknowledged that the
triangular configuration used was not the optimum profile,
which was the arched web profile, the manufacturing
capabilities and cost-efficiency prohibited the fabrication
of the arched profile. The deck was tested approved for
only 40 t load as specified by the UK's BS-5400 code.
Hayes, M.D., Lesko, J., Haramis, J., cousins,
T.E., Gomez, J., and Massarelli, P., "Laboratory and Field
Testing of Composite Bridge Superstructure", Journal of
Composites for Construction, ASCE, Vol. 4, No. 3, May 2000,
pp. 120,-128, and Hayes, M.D., Ohanehi, D., Lesko, J.,
Cousins, T., and Witcher, D., "Performance of Tube and Plate
Fiberglass Composite Bridge Deck", Journal of Composites for
Construction,. ASCE, Vol. 4, No. 2, May 2000, pp.48-55,
tested 1224 mm wide bridge deck made of glass fiber and
polyester. The deck was made up of 12 pultruded square
tubes 102x102x6.36 mm thick sandwiched between two pultruded
9.53 mm thick plates. The section was thus 121 mm deep.
The elements were all connected using epoxy adhesive, all
mated surfaces were abraded before applying the epoxy and
pressure was applied to the assembled deck until curing was
complete. Additional lateral restraint was provided by
through fiber bolts 25.4 mm in diameter at 300 mm spacing on
the longitudinal direction of the deck.
Shekar, V., Petro, S.H., and GangaRao, H.V.S.,
"Fiber Reinforced Polymer Composite Bridges In West
CA 02561453 2006-09-28
- 3 -
Virginia", 8th International Conference on Low-Volume Roads,
Reno, Nevada, June 22-25, 2003, reported using E-glass and
vinylester resin in construction of four highway bridge
decks in the U.S. The cross-section of the decks used was
203 mm (8 in) deep using double trapezoidal and hexagonal
connectors. The units were assembled at the production
plant using polyurethane adhesive and mechanical pressure.
Assembled sections were 2.43 m in width (in direction of the
traffic) and their length were equal to the width of the
bridge in each case except for one bridge where the width
was covered with c.wo sections connected together
longitudinally over the central beam. The decks had
different spans in the four bridges (distance between beams)
of 762, 889, 1829, and 2591 mm. The construction of the
largest deck (1'7x54 m) was completed in 6 days, about 10% of
the time needed for conventional concrete deck, and using
only five workers.
The first FRP bridge deck in New York State was
placed in late 1999 to replace a deteriorated concrete deck
and allow for higher live load on the bridge,
Chiewanichakorn, M., Aref, A.J., and Alampalli, S., "Failure
Analysis of Fiber-Reinforced Polymer Bridge Deck System",
Journal of Composites, Technology & Research, Vol. 25, No.
2, April 2003, pp. 121-129. The deck panels were made of E-
glass stitched fabric wrapped around foam blocks. The deck
was designed using fini.te element analysis and stresses in
the composite materials were limited to only 20% of their
ultimate strength, deflection was also limited to Span/800.
This extremely conservative design was due to the lack of
data and experience on composite bridge decks. Composite
action between the FRP deck and the steel girders was
deliberately eliminated during design. Field tests after
installation showed as-designed stresses in the composites
at about 2.9 MPa ( fufrp = 221 MPa) .
CA 02561453 2006-09-28
- 4 -
An important numerical study was reported by Gan,
L.H., Ye, L., and Mai, Y_, "Design and evaluation of various
section profiles for pultruded deck panels", Composite
Structures, Elsevier, Vol. 47, 1999, pp. 719-725, on the
assessment of different cross-sectional configurations for
pultruded FRP bridge deck panels. The authors used
commercial finite element package ABAQUS to compare the
behavior of seven shapes most used in research and
application, hexagonal (honeycomb), triangular, rectangular,
square, thick-top square, enhanced triangular, and enhanced
channel were analyzed. The analysis was based on equal
cross-sectional area for all seven shapes to reach the
optimum cost-performance one since the area of the cross
section is the main parameter governing the cost of the
bridge deck. The overall depth of the shapes was also kept
constant. The authors assumed orthotropic properties for
the material and used 3-D block and shell elements for the
static and buckling analyses, respectively.
SUMMARY OF INVENTION
It is a feature of the present invention to
provide a fiber reinforced thermoplastic composite panel
constructed of commingled glass fiber reinforced
polypropylene (FRP) components and which provides several
advantages over traditional wood timber slabs and over the
known thick bridge panels have been constructed of synthetic
materials such as E-glass, polyester and fiberglass fabrics.
Another feature of the present invention is to
provide a fiber reinforced thermoplastic composite panel
which is impermeable to moisture/water, which is non-
corroding, which is environmentally friendly, which has a
high fatigue and durability resistance and which provides a
thinner wearing surface as opposed to wood timber decks.
CA 02561453 2006-09-28
- 5 -
Another feature of the present invention is to
provide a fiber reinforced thermoplastic composite panel
which is lightweight, easy to install, which has a service
life which is five times that of timber decks and which may
have several other uses such as in the construction of
shelters, guard rails and a multitude of other applications.
According to the above features, from a broad
aspect, the present invention provides a fiber reinforced
thermoplastic composite panel which comprises two corrugated
sheets formed of commingled glass fiber reinforced
polypropylene (FRP). Each of the corrugated sheets define a
plurality of opposed and offset, spaced-apart connecting
ridge sections and integrally formed intermediate trough
sections therebetween. The two corrugated sheets are
ls interconnected in superimposition by connection means
interconnecting at least some of the ridge sections disposed
in abutting relationship and forming hollow core spaces by
the trough sections formed therebetween. A flat plate
constructed of FRP is connected to a respective one of
opposed sides of the interconnected corrugated sheets and
secured to at least some of the ridge sections disposed on
the opposed sides. A rigid filler material is disposed in
the hollow core spaces defined by the trough sections
between the two corrugated sheets and the two corrugated
sheets and the flat plates.
According to a still further broad aspect of the
present invention there is provided a fiber reinforced
thermoplastic composite panel comprising a pair of flat
plates constructed of commingled glass fiber reinforced
polypropylene (FRP) interconnected in spaced parallel
relationship by a plurality of elongated FRP channel members
disposed transversely between the flat plates and spaced-
apart in parallel relationship. Each channel member has a
flat transverse wall and opposed connecting flanges
CA 02561453 2006-09-28
- 6 -
extending at right angles along opposed longitudinal edges
thereof. The flat plates are connected to the opposed
connecting flanges by connection means. Diagonal braces are
secured between the flat plates and each of the plurality of
elongated FRP channels and span between adjacent ones of the
channels. Hollow spaces are defined between the channel
members, the flat plates and the diagonal braces. A filler
material is disposed in the hollow spaces.
BRIEF DESCRIPTION OF DRAWINGS
The invention will now be described with reference
to the accompan-~~ing drawings in which:
FIG. 1 is an end section view of a fiber-
reinforced thermoplastic composite panel constructed in
accordance with one embodiment of the present invention;
FIG. 2A is a fragmented top view of Figure 1;
FIG. 2B is an enlarged section view showing the
construction of the corrugated sheets and their connections
together and to the opposed FRP flat plates;
FIG. 3 is a schematic view showing a test
platform to illustrate how a fiber reinforced thermoplastic
composite panel constructed in accordance with the
embodiments of the present invention was tested to determine
its performance characteristics;
FIG. 4 is a graph showing the results of the
tests;
FIG. :D is a further graph showing the results of
the tests;
FIG. 6 is an end section view showing the
construction of a fiber-reinforced thermoplastic composite
panel in accordance with a second embodiment of the present
invention; and
FIG. 7 is a graph illustrating the performance
characteristics of both composite panels when tested on the
CA 02561453 2006-09-28
- 7 -
test platform as illustrated in Figure 3 under increasing
loads.
DESCRIPTION OF PREFERRED EMBODIMENTS
Referring now to the drawings and more
particularly to Figures 1 and 2, there is shown generally at
a fiber reinforced thermoplastic composite panel
constructed in accordance with one embodiment of the present
invention. The panel comprises two corrugated sheets 11 and
10 12 formed of commingled glass fiber reinforced polypropylene
(FRP) by a continuous roll forming process. Each of the
corrugated sheets 11 and 12 define a plurality of opposed
and offset spaced-apart connecting ridge sections 11', 11"
and 12' and 12" and integrally formed intermediate trough
ls sections 13 and 14 formed between the connecting ridge
sections. The two corrugated sheets 11 and 12 are
interconnected in superimposition, as hereinshown, by
connection means which is hereinshown by high tensile bolts
interconnecting abutting ridge sections together. A web
of FRP material having a thickness of about 1.6 mm is shown
at 29 and interconnected by the bolts 15 between the
connecting ridges between opposed drop sections 11' and 12',
as shown in Fiqure 2B, whereby to provide stability in the
assembly of the two corrugated sheets 11 and 12.
Additionally, an epoxy glue 16 can be interposed between the
abutting connecting ridge sections. These ridge sections
also have a flat top wall 17 to provide good contact and
flat surfaces for the securement of the tensile bolts 15 as
is better seen in Figure 2B.
Hollow core spaces 18 are defined by the trough
sections 11" and 12". Flat plates 19 and 19' constructed of
FRP material are connected to a respective one of opposed
sides of the interconnected corrugated sheets 11 and 12 and
secured to at least some of the ridge sections 11' and il",
CA 02561453 2006-09-28
- 8 -
herein all ridge sections by tensile bolts 15. A filler
material 20 is disposed in the hollow core spaces 18 defined
by the trough sections 11" and 12" between the two
corrugated sheets and the two corrugated sheets and the flat
plates 19 and 19'. The filler material can be an expanding
soft foam material or light concrete or sand.
In the embodiment of the panel as shown by
reference numeral 10, the corrugated sheets 11 and 12 have a
thickness of 4.0 mm and the flat plates 19 and 19' have a
thickness of 5.0 mm. The fastener bolts 15 are spaced-apart
a distance of 150 mm to interconnect all of the ridges 11'
and 12' in abutting relationship. The panel as herein
formed is a bridge deck panel and was tested on the test
platform 25, as illustrated in Figure 3, and the panel was
made of flat plates 19 and 19' fabricated in one piece
having 1,200x3,300 mm each and the web sections or the core
between these flat plates was fabricated in half sizes
(600x1650 mm), and joined together, as above-described, to
form full webs in the longitudinal direction (600x3300 mm).
Tests have shown that the panel 10 was capable of resisting
a load of 515 kN between two points, herein designated by
reference numerals 26 and 27, spaced-apart a distance of
1800 mm which is representative of a pair of truck axles.
After tests it was observed that the panel 10 recovered all
deformation upon the removal of the load exerted by the
pistons 28 and 281, each having been loaded to 110 kN. It
is to be noted that this panel is considered to be a very
lightweight structural panel as opposed to timber bridge
decking structures. The panel also, as previously
mentioned, has several applications such as providing
supports in mine shafts or for the construction of shelters
for military application. Such a panel has been found to
resist direct impact by 55 mm caliber bullets.
CA 02561453 2006-09-28
- 9 -
With reference to Figure 3, finite element
analysis (FE) was performed using known commercial software
package SAP2000xM to predict deflections and stresses in the
deck profile. The analyses were made on 1/ model to make use
of the symmetry of the bridge deck and loading. The model
of the deck was 725 mm in the longitudinal direction (1/2
span of 1450 mm between bridge girder), and 625 mm wide (1/2
of 1250 mm which is equivalent to the width of three timber
log spacing carrying the wheel load), and 204 mm deep (equal
to the depth of timber deck currently used) The maximum
deflection, including any possible local buckling of the top
flange or webs was the governing parameter in the analysis.
As specified by CHBDC, the maximum allowable deflection in
the deck is L/400 which gives: 1450/400 = 3.63 mm.
Stresses and strains in the materials at the service load
level were much less than the material ultimate capacity.
The load transfer posts 28 and 28' were secured to
a hydraulic actuator mounted on a loading steel frame to
apply the loading to the deck panel. The deck panel 10 was
placed on three parallel I-beams 1.5 m wide, with 83 mm wide
flanges. The two loading areas were 250x600 mm and were
separated by 1.8 m center in center, which represents the
width of truck axle. Plywood boards of 20 mm were placed
between the supporting I-beams and the slab, and between the
loading steel plates and the slab to insure no compressive
force will be exerted on the heads of deck bolts. In field
applications, bolts can be protected by using a soft
inexpensive wearing layer such as asphalt. The deck was
simply deposited on the supports and was retained laterally
by steel clamps on each side. The deck was instrumented
with 34 electrical strain gauges on the FRP on the top and
bottom surfaces, and on the webs. In addition, nine LVDT
were used to measure the deflection at different spots.
CA 02561453 2006-09-28
- 10 -
The panel or deck 10 was loaded four times with
increasing load in each test at 220, 380, 450 and 515 M.
It was observed that the panel displayed a linear behavior
in all the four tests. The four elastic lines were almost
identical in slope, indicating no loss of stiffness due to
the repeated loading and unloading. During the test of 380
kN, it was observed that the loading steel plates (25 mm
thick) started to bend upward. The panel withstood the four
tests and showed typical flexure behavior with no buckling
or failure in bolts.
Figure 4 shows the load-deflection curves for the
test. It should be noted that the three lines to the left
represent the deformation at the supports, which is due to
the compression. of the plywood strips above the I-beams.
Taking this into consideration, the net maximum deflection
should be around 28-6=22 mm, as indicated by the dashed
lines. At the service load level, the net maximum
deflection was approximately 9 mm (14-5 as indicated by the
dashed lines) or L/l60 (the target L/400 is 3.6 mm) . The
difference can be easily justified due to the use of very
soft foam instead of stiff one used. In addition, the
absence of rubber pad under the steel loading plate caused
the deformation to be almost uniform under the plate, which
in return extended the deflection line much farther outside
the plate area.
With respect to stress levels during the tests,
the stresses recorded in the composite parts were generally
low. For the test of 515 kN, stress measurements are
presented in Figure 5 with a maximum of 80 MPa tensile
stresses on the bottom surface under the loading area. This
level of stress accounts for only 27% of the capacity of the
material in tension (300 MPa) . The maximum compressive
stress was 35 MPa at the side of the loading area, or 25% of
the capacity of the material in compression (140 MPa). Some
CA 02561453 2006-09-28
- 11 -
cracking noises were heard near the end of the test (around
400 kN) and few sharp snaps occurred, probably due to
plywood breaking, but without any visible damage to the
deck. In conclusion, the deck panel 10 tested was able to
withstand more that double the service load. The recorded
deflections were higher than the allowable limits, but that
was acceptable when considering the very soft foam used, and
the absence of rubber pads under the rigid loading steel
plates. The stress levels in the deck were relatively low,
which provides enough room for fatigue and creep effects
that will take place under service condition, without
jeopardizing the integrity of the deck.
With reference now to Figure 6, there is shown a
much simpler design of the present invention and having
fewer assembly elements. The corrugated sheets 11 and 12 as
used in the core design of Figure 1 are substituted by a
core design having a plurality of elongated FRP channel
members 30 which are disposed transversely between the
opposed flat plates 31 and 32 formed of commingled glass
fiber reinforced polypropylene FRP. These channel members
30, also constructed of FRP, are spaced apart in parallel
relationship and each channel member 30 has a flat
transverse wall 33 and opposed connecting flanges 34 and 34'
extending at right angles along opposed longitudinal edges
of the flat transverse wall 33. The flat plates 31 and 32
are connected to the opposed connecting flanges 34 and 34'
by connection means constituted by high tensile bolts 35.
Diagonal braces 36 are secured between the flat
plates 31 and 32 and each of the plurality of elongated FRP
channels 30 and span between adjacent ones of the channels.
Hollow spaces 37 are defined between the channel members 30,
the flat plates 31 and 32 and the diagonal braces. A filler
material 38, such as an expanding soft foam material as
previously described, is injected in these hollow spaces 37
CA 02561453 2006-09-28
- 12 -
to add rigidity to the panel 50 as shown in Figure 6. As
previously described, this filler material can also be light
concrete or sand, depending on the application of the panel.
It is pointed out that the elongated FRP channel
members 30 are integrally formed in a continuous roll
forming machine developed by AS Composite Inc. These
channel members are formed as U-shaped channel members 30
with the opposed connecting flanges 34 and 34' projecting
from a common side of the flat transverse wall 33. The
diagonal braces 36 each extend from a free end edge of one
of the connecting flanges, such as connecting flange 341 to
an opposed end edge such as designated by reference numeral
39 of an adjacent channel member. Accordingly, there is
constructed a non-metallic panel, other than the high
tensile bolts utilized to interconnect the opposed flat
plates to the elongated FRP channel members. However, it is
preferred that these bolts be non-corrosive high tensile
type bolts.
Some of the advantages of the panel constructed in
accordance with the second embodiment is that it eliminates
longitudinal splices in the fabrication of the web. Also,
this design eliminates the use of epoxy adhesives and half
the amounts of tensile bolts the assembly is much faster.
The use of the U-shaped channel members 30 also eliminates
the need for pre-drilling bolt holes.
A prototype was constructed in accordance with the
panel 50 of the second embodiment. The prototype was 1.2 m
wide, 3.2 m long and 0.25 m thick. The base materials used
was TwinTexR fabric (commingled glass fiber reinforced
polypropylene) consolidated and formed to the desired shape.
The panel consisted of two 4.0 mm flat plates 31, 32
separated in t:he middle by multiple channel members 30
spaced 125 mm apart and diagonal braces 36. These
individual parts are connected together by high tensile
CA 02561453 2006-09-28
- 13 -
bolts 35. Expanding soft foam 38 was used to fill the inner
space 37 to prevent buckling of the channels.
The panel was tested under both monotonic and
cycling loading with one load on each span representing
truck wheel (two loads representing truck axle) . Static
(monotonic) loading test was performed first and a total
load of 510 kN was reached (limit of hydraulic actuator) and
the slab suffered no visible damage. The slab recovered all
the deformation upon load removal. The deflection recorded
was much less than that of the first panel 10 and was within
target. Fatigue test was performed until failure of the
slab at a frequency of 1 Hz and a load of 1-110 kN on each
span. A total of 2 million cycles were achieved without
significant damage, the slab failed soon thereafter when the
load was doubled in value.
Figure 7 is a graph illustrating the deflection
comparison between the panel 10 constructed in accordance
with the first embodiment and the panel 50 constructed in
accordance with the second embodiment. This graph clearly
shows the superior performance of the panel 50 constructed
in accordance with the second embodiment representative by
the curves 60 as opposed to the curves 61 of the first
embodiment.
It is within the ambit of the present invention to
cover any obvious modifications of the preferred embodiment
described herein, provided such modifications fall within
the scope of the appended claims.
