NDE OF FRP BRIDGE BEAMS AND DECKS
John C. Duke, Jr., Scott Case, and John J. Lesko
Engineering Science and Mechanics, Virginia Tech, Blacksburg, VA 24061
ABSTRACT. Over 100 fiber reinforced polymer (FRP) innovative bridge projects have been undertaken
throughout the US in the past 5 years. However, little if any effort has been devoted to developing the
capability to nondestructively evaluate these components. Assumptions that experience with NDE of
FRP for aircraft applications can be adapted to these applications are naive.
Results of efforts to develop NDE for these types of materials and components are presented. The use of
infrared thermal imaging, ultrasonic examination, AE monitoring, Acousto-ultrasonic evaluation are
discussed. Research with distributed embedded strain sensors for overcoming the inadequacies of
conventional NDE is described.
INTRODUCTION
Increasing numbers of deficient bridge structures place greater and greater stress on
limited repair, rehabilitation, and replacement budgets. Advancements that might lead to
more durable future structures, or provide for less costly repair or rehabilitation are being
sought. FRP materials are being considered because of their potential in both arenas.
Resistance to deterioration from de-icing chemicals and high specific strength and
stiffness are reasons for considering FRP materials for new bridge structures as bridge
beams or as replacement bridge decks.
As with any structural elements used for bridge structures it is necessary to be able to
inspect FRP structural elements. Although the principle means of inspecting bridge
structures is visual inspection, such inspection has proven to be cost-effective and capable
of detecting deterioration before it progresses to a critical stage. In some instances the
visual inspection is supplemented by various nondestructive evaluation (NDE) methods.
FRP materials, because of their micro structure, tend to deteriorate in ways that are not
easily detected by visual examination.
FRP DETERIORATION MECHANISMS
Fiber reinforced materials have been developed to synergistically combine two, or
more, materials to achieve improved performance. In addition, because of the ability to
orient the reinforcement fiber and adjust the concentration these materials can be tailored to
achieve different properties. Despite the variation of fiber architecture, fiber material, or
matrix material these FRP materials exhibit essentially the same basic deterioration
mechanisms: matrix cracking, fiber breakage, and delamination or planar separation
between layers of reinforcement.
CP657, Review of Quantitative Nondestructive Evaluation Vol. 22, ed. by D. O. Thompson and D. E. Chimenti
© 2003 American Institute of Physics 0-7354-0117-9/03/S20.00
1178
FIGURE 1. Examples of FRP bridge beam materials depicting fiber layer waviness, resin rich areas, and
voids. The image on the right is an enlarged view of the upper surface of the bridge beam on the left that has
been expose when the material above separated.
The deterioration development basically occurs throughout the material, with deterioration
occurring in regions where the reinforcement fibers are oriented at an angle to the load
directions. As this deterioration develops, more and more load is redistributed to the fibers
that are oriented parallel to the load direction. As the deterioration occurs it tends to be
globally distributed until it eventually becomes localized and the structure experiences final
failure. Typically this deterioration will not be readily visible and NDE methods must be
used as supplements.
The NDE may be made more difficult by the presence of manufacturing imperfections.
Unlike FRP material components manufactured for aerospace applications the FRP
materials for infrastructure applications tend to contain many more manufacturing
imperfections, including voids, fiber layer waviness, resin rich and starved areas, Fig. 1.
NDE OF FRP BRIDGE BEAMS AND DECKS
Almost no research directed at development of NDE methods for FRP bridge beams or
decks has been sponsored FHWA, state DoTs, or material manufacturers. The latter of
course are seeking to penetrate the very lucrative transportation infrastructure market so
identifying manufacturing imperfections has been a low priority. The former organizations
have typically considered FRP projects novelties so the need for inspection capabilities
again has been given a low priority. Nevertheless, an effort has been made to develop
procedures for examining FRP beams and decks. Fig. 2 shows the cross section of 8 inch
and 36 inch deep glass and carbon (hybrid) fiber reinforced/ vinyl ester pultruded Double
Web I Beams and the cross-section of a overall 7 inch deep glass fiber reinforced/ vinyl
ester pultruded deck constructed of tubes and plates bonded together with epoxy and coated
with a polymer concrete wear surface; these were all manufactured by Strongwell, Inc.
Infrared thermal imaging was used to examine the exterior surfaces of 20-foot long, 8inch deep beams that were used to rehabilitate the Toms Creek Bridge in Blacksburg
Virginia1. The beams were translated past a quartz-heating lamp and then past the thermal
imaging system.
Two significant types of imperfections were detected, however, the sensitivity to small
size flaws was not established. Fig. 3 shows thermal images of the two types of
imperfections a delamination, probably caused when the beam was cut, and a resin starved
region. A portion of the resin-starved region is visible to the unaided eye; however, the
region detected by thermal imaging was of much greater extent that the visible region.
1179
a)
b)
FIGURE 2. a) 8" and 36" deep glass and carbon (hybrid)/vinyl ester pultruded Double Web I Beams
(DWIB), b) 7 inch thick deck section of glass/vinyl ester pultruded tubes and plates.
a)
b)
FIGURE 3. Thermal image of a) a delamination (gray) that begins on the left at the cut edge of the beam,
b) resin starved region (black region surrounded by gray).
Efforts have also been devoted to developing a procedure for field examination of an
FRP deck section. In the field-testing configuration utilized a deck section was installed
over a small pit in an interstate truck weigh in motion station. This installation restricted
access to the lower surface of the deck, so that only the top surface was available for
observation.
To create a thermal gradient hot air was forced through the deck tubes, see Fig. 2b, and
the thermal image of the top surface observed. Initial attempts to examine the deck during
operation were frustrated because of heat transferred to the deck heat from the truck tires.
Fig. 4 displays an infrared thermal image obtained at night, during a period when the weigh
in motion station was closed. Separation of the top plate from the tubes, or of the wear
surface from the plate could be detected if the size was relatively large. Smaller
separations or delaminations associated with the initiation of this form of damage are
obscured by inherent material variations. This is especially evident in an ultrasonic C-scan
of a portion of pultruded material, Fig. 5.
Furthermore, it is clear from AE monitoring during proof testing of the FRP bridge
beams that small-scale damage occurs at loads well below the service load. Fig. 6 shows
the AE results obtained, with sensors placed on the flange experiencing tension and on the
flange experience compression during a proof-load to just above the service load level.
The load was increased, then decreased and reloaded, which explains the two sets of AE
curves; the AE from the tension side clearly exhibits the Kaiser effect, while that from the
compression side reoccurs prior to reaching the previous high load, but upon reaching it the
AE from both sides increases abruptly.
1180
FIGURE 4. Thermal image of a FRP bridge deck section, observed during in-situ field-testing. Hot air was
forced through different tubes. Imperfections are detected by comparing the thermal image to the baseline
image obtained in the laboratory.
FIGURE 5. Ultrasonic C-scan of a portion of a pultruded FRP component depicting inherent material
variations present that might obscure detection of delaminations or separation of bonded components.
40000 -r
§ 35000 - 8
30000 -
> 25000
20000
15000
10000 -5000 -LU
0
<
0
-tension, 1st loading
-tension, 2nd loading
-compression, 1st loading
-compression, 2nd loading
500
1000
1500
2000
2500
3000
3500
4000
Strain (microstrain)
FIGURE 6. AE from the proof load of an FRP bridge beam with AE sensors on the flanges while the beam
is loaded in flexure. The load was increased in stages and upon reloading significant AE begins to reoccur at
the new load levels. Extensive AE is observed from both flanges throughout the loading that was stopped at
just above the anticipated service load.
1181
:>
Hollow-co re
alignment sleeve
/
Resonant cavity
Attachment
(*a,ft,?a)
in ode fiber
Reflector fiber
t±
M
l°l
- I
FIGURE 7. Schematic EFPI-AE sensor depicting the details of the extrinsic Fabry-Perot interferometer,
manufactured by Luna Innovations, Inc.
AN ALTERNATIVE APPROACH: HEALTH MONITORING
While or until the quality of the FRP materials for infrastructure improves to where the
level of imperfection allows for detection of damage initiation and development an
alternative approach is being considered. The approach involves the use of embedded or
attached sensors to monitor the deformation and condition of the structural component.
Three optical fiber sensor types have been given special consideration: an Extrinsic FabryPerot Interferometric Acoustic Emission (EFPI AE) sensor, Fig. 7, a Long-Period Grating,
(LPG) Fig. 8, and, a Distributed Sensor System, (DSS) Fig. 9.
Fundamental
guided mode
laddingmode
FIGURE 8. Schematic of a long-period grating, with a specially formulated coating that changes the
transmission properties of the grating when the coating absorbs moisture, manufactured by Luna Innovations,
Inc.
1182
FIGURE 9. Schematic diagram of the distributed strain sensing system where fiber Bragg intrinsic inference
gratings are used to sense either strain or temperature. FRM- Faraday rotating mirrors, D-Detector, R- partial
reflectors. Luna Innovations, Inc. manufactures the DSS system under license from NASA/LaRC.
The EFPI-AE is particularly sensitive to in-plane displacements that would be
associated with fiber breakage, which if aligned with the critical load carrying directions
could signal the onset of catastrophic failure. Detection of moisture that can degrade the
properties of the composite material directly or indirectly through freeze-thaw damage is
possible using the LPG sensor. While the DSS sensing system can essentially blanket the
critical member allowing the strain throughout the component to be monitored making
detection and assessment of the effects of defects unnecessary since the strain provides a
direct measure of assessment locally by the sensor and globally by the collective
distribution.
SUMMARY
FRP materials for transportation infrastructure applications typical contain
manufacturing imperfections that are distributed throughout the component and tend to
obscure deterioration that initiate and gradually propagate while in-service in typical
environments. NDE development has been frustrated except when imperfections are quite
large. Health monitoring offers an alternative to NDE when it is desired to use these
materials. However, frequent consideration of the health monitoring data is necessary to
determine if deterioration has occurred to the extent that the structural performance is
compromised.
ACKNOWLEDGEMENT
The support of the Virginia Transportation Research Council, the Federal Highway
Administration, the Virginia Department of Transportation, Luna Innovations, Inc., and
Strongwell Inc. is acknowledged.
1183
REFERENCES
Hayes, Michael D., Ohanehi, D., Lesko, J. I, Cousins, T. E., and Witcher, Dan,
"Performance of Tube and Plate Fiberglass Composite Bridge Deck", Journal of
Composites for Construction, May 2000, pp 48-55.
Miceli, Marybeth, "Assessment of Infrared Thermography as an NDE Method for
Investigation of FRP Bridge Decks," Masters Thesis, Virginia Polytechnic Institute
and State University, Blacksburg, 2000.
Duke, Jr., J. C., "Improved Detection of Fiber Breakage in Fiber-Reinforced
Materials Using an In-Plane Displacement AE Sensor," Proceedings of the 10th USJapan Conference on Composite Materials, ed. F-K Chang, Stanford, California,
2002, pp. 170-176.
Childers, B. and Duke, Jr., J. C. Recent Developments in Fiber Optic Sensors, 81st
Annual Transportation Research Board Meeting, Washington, DC, January 2002.
1184