纳米复合材料基GFRP板材耐久性研究(Durability of GFRP composite made of epoxy-organoclay nanocomposite

© 2009 Taylor & Francis Group, London, ISBN 978-0-415-46850-3

Durability of GFRP composite made of epoxy/organoclay nanocomposite

Hong-Gang Zhu & Christopher K.Y. Leung

Department of Civil Engineering, Hong Kong University of Science and Technology, Hong Kong, China

Jang-Kyo Kim

Department of Mechanical Engineering, Hong Kong University of Science and Technology, Hong Kong, China

ABSTRACT: With a suitable amount of organoclay introduced into epoxy resin, intercalated/exfoliated epoxy/organoclay nanocomposite showing improved barrier property and thermal stability can be formed. GFRP composite with epoxy/organoclay nanocomposite matrix has been fabricated in this study. The main purpose of our work is to investigate the effect of epoxy/organoclay nanocomposite matrix on durability improvement of GFRP composites. GFRP composite laminates with neat epoxy matrix or 3wt% epoxy/organoclay nano-composite matrix are conditioned in the following environments: (i) standard laboratory condition; (ii) heating for 1h at 40°C, 50°C, and 60°C; (iii) immersion in alkaline solution at 60°C. After predetermined periods of conditioning, aged specimens are tested in uniaxial tension to obtain the tensile strength, tensile modulus and failure strain of GFRP composites. Results show that the tensile strength of aged GFRP composites is reduced to different degrees for different environmental exposure condition. However, the degradation rate is reduced when epoxy/organoclay nanocomposite is used as GFRP matrix. This can be attributed to the improved barrier property and thermal stability of the epoxy/organoclay nanocomposite. When GFRP composites are employed in the rehabilitation of concrete structures, the use of epoxy/organoclay nanocomposite matrix is likely to improve the durability of the rehabilitated structure as well.

1 INTRODUCTION

service life requirement of 80–100 years, has been hindered by the poor durability of GFRP composites.

In recent years, glass fiber-reinforced polymers The degradations of GFRP composites caused by (GFRPs) have been widely used in the construc-environmental exposure mainly relate to four basic

tion industry. Examples include GFRP components environmental factors: gas, humidity (with or with-as replacement of concrete bridge decks and GFRP out chemicals), thermal and UV radiation. A possible sheets for beam and column strengthening. GFRP way to enhance the durability of GFRP composites composites used in the civil engineering applica-is to use polymer (composite) matrix with improved

tions may be exposed to various environments, such barrier property, thermal stability and UV radiation as atmospheric humidity, alkaline environment, ele-resistance.

vated temperatures as well as temperature cycling Polymer/organoclay nanocomposite is a new class and freeze-thaw cycling. Under these conditions, all of material developed by incorporating organoclay constituents of GFRP composites including matrix, (organically modified layered-silicate) into a poly-fiber and fiber/matrix interface are subject to physi-mer matrix. Once the organoclay particles compris-cal and chemical changes, resulting in degradations ing of silicate platelets with very small thickness in mechanical properties. A number of investigations (about 1 nm) and very high aspect ratio are properly have been conducted to study the durability of GFRP dispersed, a few weight percent is sufficient to cre-composite exposed to various environments and sig-ate very large surface area for strong interactions

nificant reductions in mechanical properties caused by with the polymer. As a result, the polymer/organoclay various environmental exposures have been reported nanocomposites exhibit markedly improved mechani-(Kajorncheappunngam et al. 2002, Karbhari et al. cal and thermo-mechanical property, thermal stabil-2002, Hale et al. 1998). The widespread utilization ity, water/chemical ion/gas barrier property and UV of GFRP composites in the civil infrastructure, with radiation resistance (Mai & Yu 2006). There is hence

1195

Neat epoxy resin and 3wt% epoxy/I.30P nanocom-posite were prepared for the fabrication of GFRP composites.

The vacuum-assisted hand lay-up method was adopted for fabricating GFRP laminate at room tem-2 EXPERIMENTAL PROGRAM

perature (23°C). In this technique, one ply of dry glass fabric was wetted by a layer of epoxy resin (nanocom-2.1 Materials

posite) on a rigid PP (polypropylene) plate covered

Glass strand woven fabric, epoxy resin and organoclay with stencil paper. The resin was incorporated onto particles were used to fabricate GFRP composites the dry glass fabric using a brush. Then another ply of in this study. A commercial available thermosetting dry glass fabric was staked on the wetted glass fabric epoxy resin used for FRP matrix was employed. and wetted by another layer of epoxy resin (nanocom-This room temperature cured epoxy resin is made posite). The procedure was repeated until the desired up of MRL-A115 resin which is a diglycidyl ether of number of GFRP plies was laid. One more layer of bisphenol A (DGEBA) resin and MRL-B1 hardener epoxy resin (nanocomposite) was applied onto both (both supplied by Reno, Taiwan) in the mass ratio of two sides of the GFRP laminate. After the lamination 100:35. The organoclay used for the fabrication of procedure, the uncured GFRP laminate was sand-epoxy/organoclay nanocomposite is Nanomer I.30P wiched between two rigid PP plates covered with (octadecylamine modified montmorillonite) obtained stencil paper. To control the thickness of GFRP lami-from Nanocor Inc. Glass woven fabric (supplied by nate, two steel rods of specified diameter were placed Taiwan Electrical Insulator Co. Ltd) with an areal between the PP plates. The whole laminate (includ-weight of 200 g/m2 was used as fiber reinforcement.ing PP plates) was then degassed for 1h in a vacuum

oven at room temperature to remove the air bubbles trapped in the GFRP composite. The GFRP laminate

2.2 Nanocomposite fabrication

was then cured at room temperature under the pres-The in-situ polymerization method was employed sure from a specific dead load. After 7 days of room to fabricate the epoxy/organoclay nanocomposite. temperature curing, a postcure for 24h at 70°C was Before the fabrication process, required amount of performed in an oven.

In this study, both neat epoxy GFRP compos-I.30P organoclay was dried in an oven for 12h at

75°C to remove moisture while the MRL-A115 resin ite and 3wt% epoxy/I.30P nanocomposite GFRP was stirred using a high speed shear mixer (supplied composite were fabricated. All the GFRP laminate by Ross, Inc.) at a shear rate of 1000 rpm for about products fabricated through this process exhibit few 15 min at room temperature to lower its viscosity. The defects (such as voids, air bubbles, delaminations and dry organoclay was then added into the resin under contaminations), small thickness variation and good continuous action of the shear mixer. The organoclay architecture.

The constituent content of fabricated GFRP com-suspension was stirred continuously for 1h under room

temperature at a shear rate of 3000 rpm to enhance posites was determined using method II described in

ASTM D3171, based on the size and weight of GFRP the dispersion of organoclay within the resin. The sus-pension was subsequently sonicated in an ultrasonic laminate, the area weight and density of glass fab-bath at 70 W and 42 kHz for 3 hr at 60°C to further ric, as well as the density of epoxy (nanocomposite)

a good potential to improve the durability of GFRP

by using polymer/organoclay nanocomposite as the matrix.

In our research program, epoxy/organoclay nano-composite resin and GFRP composite with epoxy/organoclay nanocomposite matrix were fabricated using the in-situ polymerization method and vacuum-assisted hand lay-up method respectively. The study on the durability of GFRP composite with neat epoxy matrix and epoxy/organoclay nanocompos-ite matrix is currently in progress. A wide variety of environmental factors are considered, including high humidity and/or high temperature, alkaline environ-ment, wet/dry cycling and freeze/thaw cycling, etc. The present paper mainly focuses on the elevated temperature condition and the alkaline environment. The effect of using epoxy/organoclay nanocomposite matrix on the durability improvement of GFRP com-posite under such conditions will be evaluated.

exfoliate/intercalate the organoclay particles within the resin. This was followed by a degassing proce-dure in a vacuum oven for 2 hrs at 60°C to remove the air bubbles generated during shear mixing. After the resin/organoclay mixture was cooled to room temper-ature, a stoichiometric amount of MRL-B1 hardener was added into the resin/organoclay mixture under manual mixing. The obtained homogeneous mixture was used to fabricate GFRP composites.

Based on X-Ray diffraction (XRD) measurements and TEM observations, the epoxy/organoclay com-posite samples fabricated through this process were characterized as intercalated/exfoliated nanocompos-ite. Details can be found in Woo et al. (2007).2.3 GFRP composite fabrication

1196

2.5.1 Specimen preparation and aging conditionGFRP laminates (275 mm long, 160 mm wide and 0.9 mm in thickness) containing 4 plies of glass woven fabric were fabricated and immersed in alka-line solution at 60°C in an environmental chamber.

2.4 Short term thermal aging at high temperaturesThe alkaline solution was made by dissolving 10. g

CaCO3 and 5.95 g Ca(OH)2 in 1 liter of deioned water.

2.4.1 Specimen preparation

The pH value of the alkaline solution is about 11.5 at

GFRP laminates (260 mm long, 85 mm wide and

60°C. Elevated temperature of 60°C instead of room

1.4 mm in thickness) containing 6 plies of glass

temperature was employed to accelerate the diffusion

woven fabric were fabricated. To determine the ten-of alkaline solution into GFRP composites and hence

sile behavior, 3 coupon specimens (250 mm long,

accelerate the degradation of GFRP composites. The

24 mm wide and 1.4 mm in thickness) were cut from

predetermined aging periods are 0 days, 30 days,

each GFRP laminate for the performance of tension

45 days and 60 days. For each case, one neat epoxy

tests. The cut edges of tensile specimens were grinded

GFRP laminate and one 3wt% epoxy/I.30P nanocom-using #600 grinding paper to eliminate the cracks

posite GFRP laminate were prepared.

fabricated during cutting. Both ends of the tensile

After each aging period, two GFRP laminates (one

specimens were attached with beveled tabs (8°–10°,

with epoxy matrix and the other with nanocomposite

60 mm long and 2 mm in thickness). The actual gauge

matrix) were taken out, cleaned using flushing deioned

length of tensile specimen is 130 mm long.

water and dried using tissue Weight measurement is then performed. The GFRP laminates were subse-2.4.2 Aging condition and tension test

Material testing system, MTS 810, attached with a quently kept at standard laboratory condition (23°C, 681 environmental chamber, was employed for the 70% RH) until the moisture content reached its origi-tension test after short term thermal aging at elevated nal level. They were then cut according to the pattern

mm temperature. The tensile specimens placed on the shown in Figure 1. 5 coupon specimens (250

long and 24 mm wide, marked as A, B, C, D and E) adiabatic support were heated at a specified tempera-ture for 1h first in the environmental chamber. The were obtained from each GFRP laminate to deter-aged specimens were subsequently tested in tension mine the tensile behavior of aged GFRP composite.

at same temperature in the environmental chamber. The cut edges of coupon specimen were grinded using The temperatures considered include 23°C, 40°C, #600 grinding paper and both ends of the tensile speci-50°C and 60°C. They are all below the Tg of epoxy mens were reinforced by 60 mm long wrapping tabs.

As expected, the tensile properties obtained from resin (nanocomposite) which is about 69°C.

For each temperature considered, both neat epoxy coupons A and E are inferior to those from coupons

B, C and D, due to the presence of an “edge effect”. GFRP tensile specimens and 3wt% epoxy/I.30P nano-composite GFRP tensile specimens were prepared. As fibers and fiber/matrix interfaces may be directly exposed to the environment on the edges of the lami-And for each case, 3 specimens were tested.

Following ASTM D3039, the uniaxial tension nate, degradation occurs to a higher extent in coupons tests were conducted with wedge grips. Loading was A and E. In the next section, only the test results from applied under the displacement control mode at a pace coupons B, C and D (which are far from the edges of rate (crosshead speed) of 0.25 mm/min. The loading direction was parallel to the longitudinal direction of tensile specimen, which is the warp direction (or fill direction) of the fiber yarn in glass fabric. The lon-gitudinal strain was measured with an extensometer (25 mm in gauge length) attached to the middle part of tensile specimen while the corresponding load-ing and displacement were obtained from the load cell mounted on the loading fixture. All signals were recorded concurrently.

For all specimens, the tensile strength of GFRP composites was evaluated from the peak values of measured tensile stress while the elastic modulus was determined as a chord modulus, which is the slope

of stress-strain curve between strain values of 0.001 Figure 1. Cutting pattern of aged GFRP composite and 0.003.laminate.

matrix. The weight fraction of glass fabric in neat epoxy GFRP laminate and epoxy nanocomposite GFRP laminate are about 55% and 52% respectively, while the volume fraction of glass fabric in both two kinds of GFRP laminate is about 35%.

2.5 Long term aging in alkaline solution at 60˚C

1197

the laminate) are used for comparison with the results from control specimens kept under standard labora-tory condition (i.e., without aging).

2.5.2 Tension test

The procedures of the tension test were identical to those described in 2.4.2, except that the coupon speci-mens were tested under room temperature.3 RESULTS AND DISCUSSION

3.1 Short term thermal aging at high temperatures

The GFRP composites containing woven fabric exhib-ited a nonlinear tensile stress-strain curve with knees between linear stages, regardless of the type of matrix and the aging temperature. Transverse or angled frac-ture surrounded by short length splitting failure was observed in the specimens. Failure occurred mainly in three locations: at the end of bevel tab, within the gauge length but near the end of the tab and within the middle part of the gauge length, (see Figure 2). The measured tensile properties, however, are not sensitive to the fail-ure location. With increasing aging temperature, espe-cially when the temperature is close to the Tg of epoxy matrix, delamination failure gradually appeared.

The plots of tensile strength, elastic modulus and failure strain of both materials versus aging tempera-ture are shown in Figure 3. At room temperature, the tensile strength and failure strain of the composites containing organoclay were marginally lower than the control samples without organoclay (Woo et al. 2008). However, it was unexpected that the composite with a stiffer epoxy nanocomposite matrix also exhibited a lower elastic modulus than the control. The reason for the small differences in virgin (un-aged) tensile properties between two different composites is that the tensile properties of GFRP composite are deter-mined mainly by the glass fiber.

With increasing aging temperature, the tensile strength and elastic modulus of both composites decreased significantly due to the thermal soften-Figure 3. (a) Tensile strength (b) elastic modulus and

ing of matrix material. 16% and 11% reductions in (c) failure strain of both neat epoxy GFRP composite and

3wt% epoxy/I.30P nanocomposite GFRP composite versus temperature.

tensile strength, and 18% and 12% reductions in elas-tic modulus were noted for the composites with neat epoxy and epoxy nanocomposite, respectively. At elevated temperature, the mobility of epoxy molecule increases and the epoxy matrix becomes soft, result-ing in the reduction in elastic modulus of composite. The thermal softening of epoxy matrix would also

Figure 2. Typical failure modes of GFRP tensile reduce the effectiveness of stress transfer from matrix

to fiber. The tensile properties are hence degraded.specimens.

1198

Figure 3 indicates that with the introduction of organoclay into epoxy matrix, the GFRP compos-ite always exhibited a smaller rate in both tensile strength and elastic modulus reduction, especially when the temperature was above 40°C. Although the GFRP composites with organoclay showed a marginally lower tensile strength and elastic modu-lus at room temperature, with temperature increas-ing, the differences in tensile properties between the two GFRP composites with and without organoclay became negligible. At 60°C, these tensile properties became higher for the GFRP composites with orga-noclay than for the neat epoxy GFRP composite. The reason is that the organoclay platelets properly dis-persed within the epoxy matrix can inhibit the mobil-ity of polymer molecules at elevated temperatures and hence can effectively reduce the thermal softening of epoxy. This ameliorating effect is more significant at a higher temperature, especially at temperatures closer to the Tg of polymer matrix.

A similar observation was made on the failure strains of GFRP composites with and without organo-clay. With increasing aging temperature, the number and total length of uniformly distributed transverse cracks within the epoxy matrix were reduced. As a result, the failure strain is decreased. With the rigid organoclay inclusion resisting premature failure of polymer matrix, GFRP with nanocomposite matrix shows superior failure strain after exposure to high temperature (60°C).

3.2 Long term aging in alkaline solution at 60°CIn this paper, only test results obtained for composite samples before and after aging for 30 days are pre-sented. Further aging on additional specimens are currently underway.

The control laminate samples (without aging) are generally transparent, and the GFRP composites with organoclay exhibit a slightly more fuscous appearance. However, the GFRP composite samples, with or with-out organoclay, became opaque after aging in alkaline solution. The opaqueness became even more obvious after the evaporation of absorbed moisture, as shown in Figure 4. The white and opaque representation arose from the ingress of alkaline solution contain-ing white CaCO3 and Ca(OH)2 chemicals, and stayed preferentially along the fiber-matrix interface region. This appearance transformation could be resulting from the residue of white alkaline chemicals in GFRP composite and could also be the result of chemical reaction between the alkaline solution with fiber and matrix (Ravindran & Cho 2006). Further studies need to be conducted to provide a proper explanation.

The weights of composite samples were measured before and after aging, and Table 1 lists the weight change after 30 days aging. Although all the GFRP

Figure 4. GFRP composites laminate: (A) Non-aged con-trol sample and (B) Alkaline solution aged sample.

Table 1. Weight changes of GFRP laminates after ageing. W1 W2 ΔW ΔWGFRP

g g g %

Neat epoxy 61.24 62.68 1.44 2.35 Nanocomposite 66.63 68.42 1.79 2.69Note: W1 and W2 are the weights of GFRP laminates before and after ageing, respectively. ΔW(g) = W2 − W1 is the mass increment after aging.

laminate samples showed nominally the same size and contained the same amount of fibers, the GFRP lami-nate containing 3wt% organoclay had a higher origi-nal weight than the neat epoxy composite because of the higher density of the epoxy nanocomposite than the neat epoxy. The mass increment after aging was higher for the nanocomposite GFRP laminate than the neat epoxy GFRP laminate. Higher microvoids content in the matrix of GFRP composite with orga-noclay, resulting from the higher viscosity and faster curing speed of epoxy/organoclay nanocomposite, may be responsible for this observation. It appears that degassing for 1h was not sufficient to remove all entrapped air in the epoxy nanocomposite matrix. This issue needs to be further investigated in the future.GFRP composites tested in this study also showed nonlinear tensile stress-strain curve and the three failure locations as shown in Figure 2. The charac-terization of failure mechanisms using microscopic techniques is currently underway.

Figure 5 shows the tensile test results. As shown previously, the control neat epoxy samples before aging exhibited a slightly higher tensile strength and

1199

come from several mechanisms (1) plasticization and microcracking of polymer matrix; (2) debonding of fiber/matrix interface; and (3) corrosion and dissolution of fiber. According to Chu et al. (2004), most the deg-radation in tensile performance of GFRP was mainly related to fiber/matrix debonding and glass fiber disso-lution, rather than merely at the level of resin cracking.It is interesting to note that the reduction in tensile properties became smaller with the incorporation of 3wt% organoclay into the epoxy matrix. The absolute values of the aged properties were higher when organ-oclay nanocomposite matrix was used in place of pure epoxy matrix. Although 30 days aged nanocomposite GFRP composite exhibited a larger absorption rate of alkaline solution, its degradation rate in tensile prop-erties was much smaller than that of aged neat epoxy GFRP composite. In this case, the improved barrier property of epoxy nanocomposite matrix was respon-sible for the superior residual tensile properties of the composites with organoclay. It appears that the organo-clay served as the barrier to diffusion of alkaline solu-tion into the composite, thus reducing chemical attack on the glass fiber during the aging process. Therefore, only the matrix material was significantly degraded. For the control composite samples, due to the absence of effective diffusion barriers, the mechanical proper-ties of both the fiber and matrix are degraded.4 CONCLUSIONS

The durability of GFRP composites with and without organoclay in the matrix was studied. When GFRP composites were subject to short term thermal aging or long term alkaline solution aging at 60°C, their ten-sile properties were degraded but to different degrees. It is interesting to note that although the introduction of organoclay into epoxy matrix has little effect or even a slightly negative effect on the tensile properties of vir-gin GFRP composite, it reduces the degradation rate of tensile properties of aged GFRP composite, under ele-vated temperature or alkali environment. After aging, the absolute values of tensile properties are superior for the composite with organoclay nanocomposite matrix rather than pure epoxy matrix. As GFRPs are already commonly used in both new constructions and rehabil-itation of existing structures, the improved durability of GFRP composite with polymer/organoclay nanocom-posite matrix is a finding of practical significance that should be further explored in future investigations.

Figure 5. (a) Tensile strength (b) elastic modulus and (c) failure strain of both neat epoxy GFRP laminate and 3wt% epoxy/I.30P nanocomposite GFRP laminate versus aging period.

elastic modulus than the GFRP composites containing REFERENCESorganoclay, After aging, all these tensile properties

for both materials were reduced significantly due Chu, W., Wu, L.X. & Karhbari, V.M. 2004. Durability evalu-to the ingress of alkaline solution. The aging effects ation of moderate temperature cured E-glass/vinylester

system. Composite Structures 66:367–376.of alkaline solution on GFRP composites generally

1200

Hale, J.M. & Gibson, A.G. 1998. Coupon tests of fiber rein-forced plastic at elevated temperatures in offshore pro-cessing environments. Journal of Composite Materials 32:387–404 & 526–543.

Kajorncheappunngam, S., Gupta, Rakesh, K. & GangaRao, Hota, V.S. 2002. Effect of aging environment on degrada-tion of glass-reinforced epoxy. Journal of Composites for Construction 6(1):61–69.

Karbhari, W.M., Rivera, J. & Zhang, J. 2002. Low-temperature hydrothermal degradation of ambient cured E-glass/vinylester composites. Journal of Applied Poly-mer Science 86:2255–2260.

Mai, Y.W. & Yu, Z.Z. 2006. Polymer Nanocomposites. Cambridge: Woodhead Publishing Limited & Boca Raton: CRC Press LLC.

Ravindran, N. & Cho, E.H. 2006. Durability of glass-fiber-reinforced polymer nanocomposites in an alkaline environment. Journal of Vinyl & Additive Technology 12:25–32.

Woo, R.S.C., Chen, Y.H., Zhu, H.G., Leung, C.K.Y. & Kim, J.K. 2007. Environmental degradation of epoxy-organoclay nanocomposites due to UV exposure: Part I photodegradation. Composites Science and Technology 67:3448–3456.

Woo, R.S.C., Zhu, H.G., Leung, C.K.Y. & Kim, J.K. 2008. Environmental degradation of epoxy-organoclay nanocom-posites due to UV exposure: Part II residual mechanical properties. Composites Science and Technology In press.

1201