INFRARED SCANNING OF FRP COMPOSITE MEMBERS

U. B. Halabe1, G. Bangalore1, H. V. S. GangaRao1, and P. Klinkhachorn2

Department of Civil and Environmental Engineering, Constructed Facilities Center,West Virginia University, Morgantown, WV 26506-6103, USA

2Lane Department of Computer Science and Electrical Engineering, Constructed FacilitiesCenter, West Virginia University, Morgantown, WV 26506-6104, USA

ABSTRACT. Fiber Reinforced Polymer (FRP) composite is rapidly emerging as an alternativematerial for the infrastructure industry, and as a supplement to the conventional material such as steel,concrete, and timber. However, the long-term behavior of these materials has not been fullyunderstood. In order to study the durability issues, it is important to develop a nondestructiveevaluation (NDE) system for continuous monitoring of structural members built with FRP materials.This paper presents the results of an experimental study on delamination detection in FRP compositemembers using infrared thermography. Simulated delaminations of various sizes were inserted intoseveral FRP box sections and deck sections during the pultrusion process to create subsurface defects.The defective specimens were then tested in the laboratory using infrared thermography to predict thelocation and planar extent of these subsurface delaminations. The infrared tests yielded good results,which indicate that the technique can be developed for long-term in-service monitoring of FRPstructural members in the field environment.

INTRODUCTION

Structural members made of Fiber Reinforced Polymer (FRP) composite materialsare fast gaining popularity in civil engineering applications. They are being pultruded indifferent shapes and sizes for various applications. Apart from the strength tests(destructive tests) such as tension, compression, torsion and fatigue tests carried out on thespecimens made of these materials, tests are also necessary to detect the presence ofdelaminations, cracks, voids, etc. by nondestructive methods. These defects can developduring the manufacturing of these members (pultrusion process), during the construction,or during the service life of the structure when it is subjected to the loads. The presence ofdefects in a composite member may adversely affect the structure both locally andglobally. Since the durability aspects of the FRP composite materials are not fullyunderstood, it is very important to monitor them continuously in the field in order to ensurethe integrity of the structure. Infrared thermography has been identified as an effectivenondestructive testing and evaluation method for testing of these members, speciallybecause of its portability and easy-to-operate instrumentation.

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/$20.00

1003

EXPERIMENTAL SETUP

The experiments consisted of two steps. The first step was the creation of thedelaminations of known shapes and sizes in the specimens and the second was to locatethem using infrared thermography. Figure 1 shows the schematic view of the experimentalset up of infrared thermography used in the laboratory. The thermal images were recordedusing FSI Prism SP (Single Point) infrared camera with a resolution of 0.2°C. The camerawas capable of detecting the infrared radiation in the medium wavelength (3 to 5 jim)spectral region and displaying a real-time image showing relative intensity of the radiation.

A 9-inch TV monitor with built-in VCR was used to view and record the real timethermal images captured by the camera. The real time thermal images (in both gray scaleand color) were first recorded on a VHS tape and then transferred to a personal computerin the form of still images using special equipment (Snappy Video Snapshot).

The main heating source used in the laboratory was a quartz tower heater with twosettings, 750W and 1500W. To impart high energy in a short time the 1500W setting wasused during the experiments. It was found that heating the specimens by placing the heaterat a distance of about 4 to 6 inches from the specimen surface helped in achieving moreuniform heating. It was also observed that studying the thermal images by heating thespecimen for a small duration of time and then observing the images during the coolingcycle was much more effective than heating the specimen for a long duration and studyingthe images as the specimen cools down.

Two types of glass fiber reinforced polymer (GFRP) composite sections weretested in the laboratory. They were: (a) 2"x 5" hollow box sections with 3/16" thickness,and (b) bridge deck specimens (dimensions provided in a later section). The delaminationswith different dimensions were inserted in these box sections during the pultrusion process- the process by which the FRP sections are manufactured in the plant.

Each delamination was made by joining two polypropylene sheets (the materialused for making microwave safe containers) with an enclosed air pocket in between them.This material can withstand high temperatures of the pultrusion process (= 190 °C). Figure2 shows the raw materials used for making the delamination. The small strips were gluedalong the boundary as spacers in order to enclose an air pocket in between the two sheets.Additional strips were placed in the central region as stiffeners in case of largerdelaminations. Finally, this delamination was sealed in between two latex sheets (obtainedby cutting disposable gloves) so that the air is fully trapped inside. The delaminations werethen inserted at selected locations inside various FRP specimens.

FIGURE 1. Schematic view of the infrared thermography set up in the laboratory.

1004

FIGURE 2. Preparation of a delamination by enclosing an air pocket between two polypropylene sheets.

LABORATORY TESTING AND RESULTS

The specimens were heated by placing the quartz heater at a distance of 4-6 inchesfrom the surface. The surface temperature profiles were observed using the infrared cameraand were also recorded on a VHS tape for later computer acquisition.

Box Specimens

Four box specimens (2"x 5" hollow box sections with 3/16" thickness) were tested in thelaboratory, and the results from these specimens are described next.

Specimen - Box 1

A 3"x3" delamination made by trapping air between two latex sheets (obtained bycutting disposable gloves) was inserted in the specimen (Figure 3.a). Since the latex sheetswere very flexible, it was not possible to maintain a 3"x3" shape during the pultrusionprocess. Figure 3.b shows the gray-scale infrared image of the specimen Boxl. As seen inthe image the bright spot on the specimen indicates the delamination. Since thedelamination for this specimen was created using latex sheets, the originally intendeddimension of 3"x3" could not be obtained (as the latex is very flexible). The image clearlyshows the delamination size being different (approximately 1.5"x3").

FIGURE 3.a. Specimen - Box 1. FIGURE 3.b. Gray scale infrared image.

1005

FIGURE 4.a. Specimen - Box 2. FIGURE 4.b. Gray scale infrared image.

Specimen - Box 2

A 2"x2" delamination with 1/16" thickness was made by trapping air between twopolypropylene sheets as explained earlier, and was inserted in the specimen during thepultrusion process (Figure 4.a). Since the delamination is located very close to the surface,it is visible to the naked eyes (the difference in the colors of the box section material andthe delamination material could also be a reason) in Figure 4.a. The infrared image inFigure 4.b shows the delamination very clearly.

Specimen - Box 3

This specimen was made by inserting a 3"x3" delamination of 1/16" thicknessmade from polypropylene sheets (Figure 5.a). Even in this specimen, the delamination isvisible to the naked eyes. The infrared image (Figure 5.b) clearly shows the delaminationof size 3"x3" in the specimen.

FIGURE 5.a. Specimen - Box 3. FIGURE 5.b. Gray scale infrared image.

1006

FIGURE 6.a. Specimen - Box 4.

Specimen - Box 4

FIGURE 6.b. Gray scale infrared image.

This specimen was made by inserting delaminations of two sizes - one of 3"x3"size and the other of 2"x2" size (both with 1/16" thickness). Further, to conceal thedelaminations a layer of yellow paint was applied on the surface of the specimen (Figure6.a). Hence, unlike specimens Box 2 and Box 3, the locations of the delaminations inspecimen Box 4 are not visible to the naked eye. However, they could be located veryeasily using infrared thermography. Figures 6.b shows the infrared thermal image of thespecimen where both the 2"x2" and 3"x3" delaminations are clearly visible.

Bridge Deck Specimens

Figure 7 shows the cross-section of a FRP bridge deck module. Each module of thebridge deck is 24" wide, 8" deep and 12 feet long (pultruded at Bedford ReinforcedPlastics Inc., PA.). Several such modules are attached together to form a bridge deck inactual field construction. The bridge deck is composed of E-glass fibers and polyester resincombination. The glass fibers are continuous strand roving and triaxial fabrics. The fibervolume fraction is about 45%.

There are many regions within a FRP deck where there is a possibility ofoccurrence of delaminations under service loads. Some of the possible regions wheredelaminations could occur are: within the flange of any given module, between the flangejoints of two modules, and between the wearing surface and the deck. Taking thesepossibilities into consideration, different specimens with various subsurface delaminationconfigurations were made and tested using infrared thermography.

L»-5.0in.-U—6.0 in.

FIGURE?. Cross-section of a FRP bridge deck module.

1007

Specimen - BD1

This specimen was made by connecting two 12" long pieces (cut from a 12' longbridge-deck module). The pieces were connected to each other using a structural adhesive("Pliogrip" manufactured by Ashland Specialty Chemical Company) in the same way as itis done at the construction site. Two delaminations of sizes 3"x3" and 2"x2" (both with1/16" thickness), made using polypropylene sheets as explained earlier, were insertedwithin the joint between the two modules. The depth of the delaminations from the decksurface was about 5/16". Figure 8 shows the photographic view of the specimen andFigures 9 and 10 show the location of the delaminations.

FIGURE 8. Bridge deck specimen - BD1.

AA

i(2"x2")

!(3"x3")

FIGURE 9. The delamination location in the bridge deck specimen - BD1.

' /

/////

/

\DELAMINATION

FIGURE 10. Section at A-A (in FIGURE 9) and the enlarged view of the delamination.

1008

FIGURE ll.a. Bridge deck specimen- BD1. FIGURE ll.b. Gray scale infrared image of BD1.

Figure ll.a shows the front view of the bridge deck specimen BD1 tested in thelaboratory. Figure ll.b shows gray scale thermal image of the specimen. Though theimage shows some differential temperature zones, only a part of the larger delaminationcan be identified and the smaller delamination is not even visible. Therefore, it can beconcluded that 1/16" thick delaminations at a depth of 5/16" from the surface in a GFRPdeck are not distinctly visible using the current infrared thermography technique. Use of amore refined technique (background image subtraction) is currently being investigated.

Specimen - BD2:In this specimen (which is a 12" long piece cut from a 12' long bridge deck

module), the delaminations were inserted between the bridge deck and the 3/8" thickwearing surface. The wearing surface is a non-skid, flexible, hybridized, and copolymeroverlay system, with a commercial name Mark-163 FLEXOGRID (manufactured by Poly-Carb, Inc., OH). It is made of a specially selected blend of aggregates (Glacial gravel -Basalt, Quartzite, Granite) used along with two-part liquid polymer system (mixed at the .job site). Two delaminations of sizes 2"x2" and 3"x3" (both with 1/16" thickness) wereplaced on the surface of the FRP deck specimen. This was followed by the application of a3/8" thick wearing surface as per the directions given in the product technical data sheetsprovided by Poly-Carb, Inc. Figures 12 and 13 show the locations of the delaminations.

It is impossible to locate the presence of the delaminations by visual observation ofthe specimen (Figure 14.a). Infrared thermography was very successful in predicting thelocations as well as the shapes and relative sizes of the two subsurface delaminations ascan be seen from the gray scale infrared images of this specimen (Figure 14.b).

1009

A.B

(3"x3") i(2"x2")

4"

FIGURE 12. Location of delaminations in specimen BD2. FIGURE 13. Section at B-B (in FIGURE 12)and the enlarged view of the delamination.

FIGURE 14.a. Bridge deck specimen - BD2.

CONCLUSIONS

FIGURE 14.b. Gray scale infrared image of BD2.

After conducting infrared thermography tests in the laboratory on four FRP boxspecimens and two FRP bridge deck specimens, the following conclusions were drawn.• In the box-section specimens, the subsurface delaminations were clearly visible in the

thermal images. This is probably due to the presence of delamination very close to thesurface. The presence of a thick coating of paint on specimen Box 4 made thedelaminations invisible to the naked eye, though it could be easily detected usinginfrared thermography.

• The delamination of 1/16" thickness inserted at the flange-to-flange junction inbetween two bridge deck modules (specimen BDl), at a depth of about 5/16" from thesurface, was not detectable using the current infrared thermography technique.

• The delaminations of 1/16" thickness between the 3/8" thick wearing surface and thebridge deck surface were easily detectable using infrared thermography. This isbecause of higher thermal conductivity of the wearing surface material (thermalconductivity of gravels ~ 1 W/m/C) compared to the composite deck (thermalconductivity « 0.3 to 0.38 W/m/C).

1010