Materials and Design 65 (2015) 780–788

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Suitability of carbon fiber-reinforced polymers as power cable cores:Galvanic corrosion and thermal stability evaluation

http://dx.doi.org/10.1016/j.matdes.2014.10.0050261-3069/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +55 19 3705 6589.E-mail addresses: gcunha@cpqd.com.br, gibguara@gmail.com (G.C. Vasconce-

los).1 Present address: Department of Mechanical Engineering, Center for Technology

and Geosciences, Universidade Federal de Pernambuco (UFPE), Recife, PE, Brazil.

T.F.A. Santos a,1, G.C. Vasconcelos a,c,⇑, W.A. de Souza b, M.L. Costa c, E.C. Botelho c

a CPqD, Campinas, SP, Brazilb Companhia Energética de Minas Gerais (CEMIG), Belo Horizonte, MG, Brazilc Department of Materials and Technology, Universidade Estadual Paulista (UNESP), Guaratinguetá, SP, Brazil

a r t i c l e i n f o

Article history:Received 3 July 2014Accepted 1 October 2014Available online 23 October 2014

Keywords:Aluminum conductorCarbon fiber-reinforced polymerGalvanic corrosionTransmission cables

a b s t r a c t

The increasing demand for electrical energy and the difficulties involved in installing new transmissionlines presents a global challenge. Transmission line cables need to conduct more current, which createsthe problem of excessive cable sag and limits the distance between towers. Therefore, it is necessary todevelop new cables that have low thermal expansion coefficients, low densities, and high resistance tomechanical stress and corrosion. Continuous fiber-reinforced polymers are now widely used in manyindustries, including electrical utilities, and provide properties that are superior to those of traditionalACSR (aluminum conductor steel reinforced) cables. Although composite core cables show good perfor-mance in terms of corrosion, the contact of carbon fibers with aluminum promotes galvanic corrosion,which compromises mechanical performance. In this work, three different fiber coatings were tested(phenol formaldehyde resin, epoxy-based resin, and epoxy resin with polyester braiding), with measure-ments of the galvanic current. The use of epoxy resin combined with polyester braiding provided the bestinhibition of galvanic corrosion. Investigation of thermal stability revealed that use of phenol formalde-hyde resin resulted in a higher glass transition temperature. On the other hand, a post-cure processapplied to epoxy-based resin enabled it to achieve glass transition temperatures of up to 200 �C.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Electricity use in Brazil has been increasing steadily over thepast decade. According to the Brazilian Ministry of Mines andEnergy (MME), consumption has risen from 28.51 � 103 toes(tonne of oil equivalent) in 2000 to 42.86 � 103 toes in 2012 [1].Energy supply companies are therefore developing new transmissioncables in order to decrease sag and to improve creep resistanceunder high current loads. This will enable greater distancesbetween towers and help to reduce costs. The use of compositeswith thermosetting matrices reinforced with continuous carbonfibers enables substantial weight reduction combined with highstrength- and stiffness-to-weight ratios, low thermal expansioncoefficients, excellent fatigue properties, and good corrosionresistance [2–6]. These benefits, allied with the ability of thepultrusion process to produce materials with high strength-to-

weight ratios and dimensional stability, make these cables suitablereplacements for traditional steel core transmission cables, reduc-ing the sag between towers and improving electricity transmission[4,7,8]. In other applications, the high corrosion resistance ofthese materials makes them an attractive alternative for retrofit-ting steel structures, improving the stiffness of steel bridgegirders [9,10], strengthening concrete [11,12], and replacing steelcables in bridges and other civil structures [13]. However, the mostimportant issue concerning their mechanical performance is theoccurrence of galvanic corrosion between carbon and aluminum,which can compromise cable performance for the purposes ofthe application proposed in this work [14–16]. According Davis[17], under most environmental conditions normally encounteredin service, aluminum and its alloys are the anodes in galvanic cellswith most others metals, protecting them by corroding sacrifi-cially. This corrosion is more severe in solutions of halide salts aschloride solutions. It is known that graphite/aluminum compositesexhibit accelerated corrosion in marine environments whengraphite fibers and aluminum are exposed simultaneously.According to Oldfield [18], galvanic corrosion is caused by thedifference in potential between two conducting materials. Thenobler material acts as a cathode, while the more reactive material

T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788 781

acts as an anode, which corrodes. According to the galvanic series(ASTM: G82), aluminum is a reactive metal that should have lowcorrosion resistance. Nevertheless, a thin, compact, and adherentaluminum oxide film is formed on the surface, which providesaluminum alloys with high corrosion resistance. Additionally, thegalvanic series shows that the galvanic effect will be greater forsubstances that are farther apart in the series. Aluminum andcarbon (graphite) are distant from each other, so it is expected thatgalvanic corrosion will be an important issue in applicationsinvolving these two materials, affecting the lifetime and reliabilityof carbon fiber-reinforced polymer (CFRP) cores of power transmis-sion cables. The occurrence of galvanic corrosion between CFRPand aluminum is well known in the aerospace industry[16,19,20], and similar effects have been observed for CFRP usedwith concrete and steel in civil engineering [9,11,12,14,21]. Onthe other hand, several galvanic pairs can show behavior differentto that expected from the galvanic series, due to environmentaleffects and the cathodic/anodic surface ratio [22,23]. To minimizecorrosion of aluminum in contact with more cathodic material,the ratio of the exposed area of aluminum to that of the morecathodic material should be kept as high as possible. Paints andother coating can also be applied to both the aluminum and thecathodic material or to the cathodic material alone [17]. Irelandet al. [24] investigated the occurrence of galvanic corrosionbetween aluminum 7075 alloy and glass fiber-reinforced polymermade from epoxy resin modified with multi-walled carbonnanotubes (MWCNTs), and attributed galvanic corrosion to thepresence of the MWCNTs. In other work, aluminum-based matrixmetal composites were found to be more susceptible to corrosionattack, with Al/SiC composites exhibiting strong corrosion in NaClenvironments due to galvanic coupling [25].

This work investigates galvanic corrosion involving aluminum1350 and carbon fiber-reinforced polymers, and assessesimprovements in corrosion resistance achieved by applyingprotective layers of phenol formaldehyde resin or epoxy-basedresin. A suitable protective surface layer can act as a barrier,altering the corrosion rate [26] and extending the service lifetime.The thermal stability of CFRP composites is also evaluated, becausecomposite core transmission cables can reach temperatures of upto 200 �C during energy consumption peaks.

2. Materials and experimental procedure

Pultruded rods composed of three different polymeric matricesreinforced with continuous carbon fibers were manufactured usingphenol formaldehyde resin (D26-GPB, Georgia-Pacific), epoxy-based resin (CY179CH, Huntsman), and epoxy-based resin withpolyester braiding.

Samples of aluminum 1350 that complied with the ASTM: B231standard were provided by Alubar. This material is often used as anelectrical conductor in traditional transmission cables. During the

Fig. 1. Experimental setup for zero resistance amm

manufacturing process, the cast ingot is reduced by rolling todecrease the cross-section and is then submitted to forging toobtain a diameter of 0.5 mm. Final thermal treatments areconducted before tempering, according to the application.

Phenol formaldehyde resin, epoxy-based resin, and epoxy resinwith polyester braiding were applied to the aluminum surface inorder to study their effectiveness in preventing galvanic corrosion.

2.1. Electrochemical corrosion tests

The electrochemical corrosion test apparatus is illustrated inFig. 1. A PAR/EG&G Model 283 potentiostat/galvanostat was usedto perform zero resistance ammeter (ZRA) tests in order toevaluate the effectiveness of the different coatings for preventionof galvanic corrosion between the carbon fiber-reinforced polymerand the aluminum. These tests were conducted in accordance withthe ASTM: G71 standard procedure. In this, the working electrodeand the counter-electrode were replaced by specimens of alumi-num and carbon fiber, and the electrical current density betweenthem was monitored. The ZRA tests were carried out in neutral3.5 wt.% NaCl solution (artificial seawater). Five tests wereperformed; two of the results (the highest and lowest values) werediscarded and the remaining three were used to calculate theaverage values. During the test, the electrolyte was not moved oraerated. According to Möller et al. [27], seawater is not easilysimulated in the laboratory for corrosion-testing purposes, due tothe complexity of the variables involved, such as dissolved oxygenlevels, salinity, and concentrations of minor ions, biological activ-ity, and presence of pollutants. A 3.5 wt.% NaCl solution is normallyused to represent seawater, and is known to be more aggressive. Innatural seawater, the presence of calcium and magnesium acts toprovide a physical barrier against oxygen diffusion, hence decreas-ing the corrosion rate.

The microstructures of the aluminum and carbon fiber sampleswith different coatings were evaluated using optical microscopy(Axiophot, Zeiss) and scanning electron microscopy (Model5800LV, JEOL).

2.2. Thermal analyses

Thermogravimetric analyses (TGA) were performed using an SIIExtra instrument (Seiko Nanotechnology) operated with a constantflow of nitrogen (50 mL/min). Samples (�20 mg) of the twopultruded systems (epoxy and phenol formaldehyde resinsreinforced with continuous carbon fibers) were placed in astandard platinum pan and a dynamic analysis was performed byheating from 50 to 800 �C at a rate of 10 �C min�1.

Dynamic mechanical analysis (DMA) was performed with aDMTA-3E Test Station (Patel Scientific). Samples of the same twopultruded systems were measured in dual cantilever bending

eter (ZRA) tests to evaluate galvanic coupling.

782 T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788

mode, using a frequency of 1 Hz and heating from 30 to 300 �C at aconstant rate of 5 �C min�1.

3. Results and discussion

The objective of this work was to study galvanic corrosionprevention using a protective layer, as well as the thermal stabilityof the layer, envisaging the use of a pultruded aluminum conductorcomposite core (ACCC) as a replacement for traditional steel ACSRpower cables. The two systems to be used as cores of ACCC powercables are illustrated in Fig. 2 and Fig. 3. Two different geometrieswere employed, corresponding to the Raven and Linnet core con-figurations of traditional aluminum and steel-reinforced cables.

Fig. 2. Single composite core manufactured by the pultrusion process.

Fig. 3. Twisted composite core manufactured by the pultrusion process.

Fig. 4. Scanning electron micrographs of the carbon fiber-reinforced polymer (CF

The mechanical performance of the single core shown in Fig. 2was described in an earlier paper [28].

3.1. Corrosion performance

Scanning electron micrographs of the carbon fiber-reinforcedpolymer submitted to a pultrusion process with epoxy resin areshown in Fig. 4, where it is possible to observe raw fibers, theepoxy coating, and continuous fibers embedded in the epoxy coat-ing. The composition of the samples was approximately 80 wt.% ofcarbon fiber and 20 wt.% of resin.

Fig. 5 displays the current density measurements for 1350aluminum alloy in 3.5 wt.% NaCl solution. During the manufactur-ing process, an aluminum ingot was firstly obtained, followed bythe rolling and drawing procedures. Evaluation was made ofgalvanic corrosion for the CFRP in contact with aluminum samplesunder three conditions (ingot, rolled, and drawn samples), accord-ing to the ASTM: G71 standard procedure.

According to Wolstenholme [29], in corrosion evaluation agalvanic current generated by dissimilar metals immersed in anelectrolyte can only be measured correctly with a zero electricalresistance (using a zero-resistance ammeter for measuring largegalvanic currents) [30]. The results of the galvanic currentmeasurements (Fig. 5) revealed minor differences in the galvaniccurrents obtained for the different aluminum specimens and theCFRP. A higher galvanic current was developed during the initialcoupling period, after which a plateau was reached. The galvaniccurrent density is shown in Fig. 5 (with a 95% confidence interval).It can be seen that the current ranged between 20 and 25 lA cm�2

for all the aluminum samples. Wang et al. [31] studied corrosion incarbon fiber metal laminates and observed a galvanic current of20 lA between 2024-T3 aluminum alloy and CFRP. When a com-posite coating was applied, there was a decrease in the galvaniccurrent to 30 nA. On the other hand, in a study of aluminum-basedmetal matrix composites reinforced with carbon fibers, Colemanet al. [32] observed high galvanic currents of 465 and 320 lA for357-aluminum alloy/carbon fiber and 2124-aluminum alloy/carbon, respectively.

In this work, the galvanic current between drawn aluminumwires and CFRP in a 3.5 wt.% NaCl solution (artificial sea water)was measured for up to 60 h. Three samples were used for eachcondition. The behavior of the galvanic current was similar to thatdescribed previously. The average value was around 25 lA (Fig. 6)for both immersion times (16 and 60 h), confirming that a plateauwas reached.

RP), showing raw fibers, epoxy-based polymer, and fiber with epoxy coating.

Fig. 5. Current density measurements using (a) cast aluminum, (b) rolled ingot, and (c) rolled and drawn ingot. The samples were exposed in a 3.5 wt.% NaCl solution for 16 h.The average current densities are shown in (d).

Fig. 6. Measurements of the galvanic current between drawn aluminum and CFRPimmersed in 3.5 wt.% NaCl solution for 60 h. The inset shows the average currentvalues (with 95% confidence intervals) for the 16 h and 60 h tests.

T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788 783

Optical microscopy images of the drawn aluminum samplesafter galvanic coupling are shown in Fig. 7 for immersion timesof 16 and 60 h. Pitting corrosion can be seen, resulting from thedissolution reactions that occurred as the current density wasincreased. The results obtained here indicated that there was nodegradation of the carbon fibers.

The results of the ZRA tests using the drawn aluminum wiresand the CFRP are shown in Fig. 8. Three configurations were tested,using CFRP produced with phenol formaldehyde resin, epoxy resin,and epoxy resin plus polyester braiding. The carbon fiber compos-ites with the phenol formaldehyde resin matrix provided lesseffective protection against galvanic corrosion (Fig. 8a), because

the galvanic current was higher than that measured betweenaluminum and raw carbon fibers. This can be explained by thepresence of exposed carbon fibers, enabling a galvanic current toflow between the materials. Fig. 9 shows an SEM image of rawcarbon fibers in the phenol formaldehyde resin matrix after ZRAtesting for 16 h. There was probably a synergistic effect involvingexposure of the raw carbon fibers and degradation of the phenolformaldehyde resin, resulting in higher galvanic coupling currentflow. The higher current density indicated a faster corrosion rate,reflecting easier corrosion in seawater.

Greater galvanic corrosion resistance was achieved for thecarbon fiber composite with an epoxy-based matrix. The currentdensity flow between the aluminum and the CFRP decreased to6.1 ± 0.6 lA cm�2 (Fig. 8b), indicating that there was less stimula-tion of galvanic corrosion, compared to aluminum and the rawcarbon fibers. Nonetheless, although the epoxy coating provideda degree of protection, there was probably some contact betweenthe raw fibers and the aluminum. The integrity of the epoxysurface coating is shown in the scanning electron micrograph ofFig. 10, where the arrows indicate the presence of defects. It isimportant to emphasize that the measured current is acceptablein terms of corrosion, because it is of similar magnitude as thatof the material in a passive state. According to [33] normally, agood passivity should correspond to a relatively drastic decreasein corrosion rate of the immersed metal and could correspond toa corrosion rate of a few microamperes per square centimeterinstead of mA/cm2. Liu et al. [34] studied the effect of naturalinhibitors on corrosion of AA7075 alloy exposed in a 3.5% NaClsolution and reported that the corrosion current decreased from52 to 3 lA cm�2 when 1000 ppm of an organic inhibitor was used.There was therefore an improvement in corrosion resistance in the

Fig. 7. Optical microscopy images of aluminum samples after ZRA tests between CFRP and aluminum: (a) 16 h; (b) 60 h. The arrows indicate corrosion pits.

Fig. 9. Scanning electron microscopy images of the CFRP, showing the phenolic resin matrix and raw fibers.

Fig. 8. Measurements of current density between drawn aluminum wire and the CFRP with (a) phenolic resin, (b) epoxy resin, and (c) epoxy resin/polyester braiding. Theaverage galvanic currents are presented in (d).

784 T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788

Fig. 10. Scanning electron microscopy images of the CFRP with epoxy resin matrix.

Fig. 12. Dark field optical microscopy image of the CFRP sample with phenolic resincoating.

T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788 785

presence of the organic inhibitor. Tavakkolizadeh and Saadatm-anesh [14], for example, estimated the corrosion rate of the steelwhen coupled with the carbon in seawater solution using the Evandiagram and galvanic corrosion measurements. The authorsreported a corrosion rate of 227 lA cm�2 and 85 lA cm�2 by Evandiagram and galvanic corrosion measurements, respectively. Whenthey applied a thin epoxy coating, it was reported a decrease ofcorrosion rate to 18 e 8.5 lA cm�2 for epoxy thickness of 0.25and 0.10 mm using galvanic corrosion measurements, respectively.The authors highlighted an improvement of corrosion resistance ofthe coupled system with epoxy coating.

Among the systems analyzed, the CFRP with epoxy-based resinand polyester braiding showed the best corrosion resistance. Therewas no current between the two electrodes (Fig. 8c); the magni-tude of the current was of the same order as the instrumentalnoise. Fig. 8d shows the large differences in the average galvaniccurrents obtained for the three coatings. Fig. 11 shows the integ-rity of the coating, no defect is highlighted.

Measurements of sample cross-sections were performed usingoptical microscopy. The fibers produced using phenolic resin,epoxy resin, and epoxy resin with braiding showed thicknessesof 50 ± 6, 21 ± 1, and 53 ± 4 lm, respectively. A dark field opticalmicroscopy image of the CFRP sample with phenolic resin coatingis shown in Fig. 12. Although it was anticipated that a thickercoating would provide a more effective physical barrier, hencedecreasing the possibility of galvanic corrosion, the presence of

Fig. 11. Scanning electron microscopy images of the

raw fibers was observed for the samples produced with phenolicresin. Consequently, direct contact between the aluminum andthe raw carbon fibers favored the occurrence of galvanic coupling.Despite the thinner coating, the combination of epoxy coating andbraiding provided effective avoidance of direct contact between

CFRP with epoxy resin and polyester braiding.

Fig. 13. TGA results showing weight loss as a function of temperature for the CFRP with (a) epoxy-based resin and (b) phenol formaldehyde resin.

786 T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788

the aluminum and the carbon fibers, hence preventing galvaniccoupling.

3.2. Thermal performance of the pultruded samples

The results of the TGA analyses of the CFRP with epoxy-basedand phenol formaldehyde resins are shown in Fig. 13. The thermaldecomposition of the epoxy resin system began at 300 �C (onset

Fig. 14. DMA spectra showing E0 , E00 , and tan delta curves for the phenolic resinsystem. The glass transition temperature was around 150 �C.

Fig. 15. DMA spectra showing E0 , E00 , and tan delta curves for the epoxy-based resin systransition temperature of the aged sample increased to around 140 �C.

temperature), with subsequent rapid decomposition in a singlestep (Fig. 13a). In the final phase of the process, about 83% of themass remained (approximately the carbon fiber content of thecomposite). The decomposition of the phenol formaldehyde resinsystem occurred at a slower rate, compared to the epoxy system,indicating that the system was thermally more stable. The decom-position process started at 320 �C and occurred in at least twosteps (Fig. 13b). The remaining residue was equivalent to about87% of the initial mass and corresponded mainly to the carbon fibercontent.

The results of the DMA analyses of the CFRP with phenol form-aldehyde and epoxy-based resins are shown in Fig. 14 and Fig. 15,respectively. The tan d plots indicated that the glass transitiontemperatures (Tg) of the composites were 154 and 124 �C for thephenol formaldehyde and epoxy matrices, respectively. The Tg isthe temperature at which the polymer shows a significant increasein mobility, and can be directly related to the working temperatureof the final composite, because above the Tg the mechanical prop-erties of the composite become compromised and creep increases.Therefore, initial comparison of the composites suggested that thephenol formaldehyde resin system should provide better perfor-mance at higher temperatures. However, the epoxy-based resinwas developed for post-cure processes. The DMA test was there-fore repeated nine months after fabrication, under room tempera-ture conditions, in order to determine whether the post-cureprocess occurred at room temperature, and what effect it mighthave on Tg. The tan d plot temperature was 141 �C, which was

tem. The glass transition temperature was around 120 �C. After 9 months, the glass

Fig. 16. Glass transition temperature measured using DSC for the epoxy-basedresin system aged for 12 h at 180 �C.

T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788 787

similar to the temperature obtained for the phenol formaldehyderesin composite, showing that the working temperatures of thesematerials were not significantly different. (See Fig. 15)

Another consideration is that an increase in Tg can also occurunder field conditions, where ACCC cables are constantly heatedby the Joule effect, which can itself increase the working tempera-ture and extend the cable lifetime. A DSC measurement was there-fore made of the epoxy-based resin system after post-curing for12 h at 180 �C. The Tg value obtained was 200.7 �C (Fig. 16), indicat-ing that it should be possible to achieve higher Tg values using opti-mized curing procedures.

It is important to point out that environmental exposure canalso affect Tg, depending on the polymeric system and the bonding.Bound water can act as a plasticizer and lower the Tg, with disrup-tion of the initial inter-chain van der Waals’ forces and hydrogenbonds resulting in increased chain segment mobility. In contrast,bound water can lessen the extent of Tg depression in water-satu-rated materials, due to secondary cross-linking. Zhou and Lucas[35] evaluated the influence of hygrothermal effects on the Tg ofepoxy systems and found that changes in Tg did not solely dependon the water content alone, but were also influenced by the hygro-thermal history of the material. Longer times and higher exposuretemperatures resulted in higher Tg values. It is therefore importantto consider the influence of environmental exposure on the Tg

value of the system, in order to be able to predict the maximumworking temperature of the conductor cables.

4. Conclusions

Two configurations (Raven and Linnet) are proposed for ACCCcables. In conventional ACSR cables, the steel core is in contactwith drawn aluminum. The replacement of the steel core by a com-posite core means that galvanic corrosion between the carbon fibercomposite and the aluminum becomes a concern. Three differentcoatings were therefore evaluated in terms of their ability to pre-vent galvanic corrosion between the composite and the aluminumwires. Coatings using epoxy resin and epoxy resin with polyesterbraiding were shown to be effective in preventing galvanic cur-rents, with the latter eliminating the galvanic current. In contrast,a phenol formaldehyde resin coating was unable to prevent thegalvanic current, due to direct contact of the raw carbon fibers withthe aluminum.

It is important to consider the thermal stability of ACCC cables.The polymers used in composite cores should be resistant up to

200 �C, which is the working temperature during periods of highdemand. The phenol formaldehyde resin system showed a higherglass transition temperature and greater thermal stability, com-pared to the epoxy-based system. On the other hand, the epoxyresin system could achieve higher glass transition temperaturesof up to 200 �C following post-curing. This, together with a greaterlevel of protection against corrosion, makes the epoxy system suit-able for use in ACCC cables.

Acknowledgments

The authors would like to express their deepest gratitude toCemig and the ANEEL R&D program for the provision of financialsupport, to Alubar for kindly providing the aluminum samples,and to Stratus Compostos Estruturais Ltda for manufacturing thecore cables. MLC and ECB thank FAPESP. TFAS is grateful to Profes-sor C.S. Fugivara (UNESP) for support throughout the corrosiontests, and to F. Montoro (LNNano/CNPEM) for invaluable assistanceduring the SEM tests.

References

[1] Report: consolidated data – final energy consumption by source: electricity.<http://www.mme.gov.br/mme/menu/todas_publicacoes.html>. [accessed25.04.12] [in Portuguese].

[2] Batchelor J. Use of fibre reinforced composites in modern railway vehicles.Mater Eng 1981;2:172–82.

[3] Botelho EC, Silva RA, Pardini LC, Rezende MC. A review on the developmentand properties of continuous fiber/epoxy/aluminum hybrid composites foraircraft structures. Mater Res 2010;9:247–56.

[4] Baker AA. Carbon fibre reinforced metals – a review of the current technology.Mater Sci Eng A 1975;17:177–208.

[5] Ramakrishna S. Microstructural design of composite materials for crashworthystructural applications. Mater Des 1997;18:167–73.

[6] Bora MO, Çoban O, Sinmazcelik T. Effect of fiber orientation on scratchresistance in unidirectional carbon-fiber-reinforced polymer matrixcomposites. J Reinf Plast Compos 2010;29:1476–89.

[7] Ma C-CM, Yn MS, Chen C-H, Chang CJ. Processing and properties of pultrudedthermoplastic composites I. Compos Manuf 1990;1:191–6.

[8] Chen C-H, Ma C-CM. Pultruded fibre-reinforced polyurethane composites III.Static mechanical, thermal, and dynamic mechanical properties. Compos SciTechnol 1994;52:427–32.

[9] Rizkalla S, Dawood M, Schnerch D. Development of a carbon fiber reinforcedpolymer system for strengthening steel structures. Composites: Part A2008;39:388–97.

[10] Arronche L, Gordon K, Ryu D, La Saponara V, Cheng L. Investigation of galvaniccorrosion between AISI 1018 carbon steel and CFRPs modified with multi-walled carbon nanotubes. J Mater Sci 2013;48:1315–23.

[11] Torres-Acosta AA, Sem R. Electrochemical behavior of coupled carbon fiber-reinforced polymer (CFRP) rods in concrete. J Appl Electrochem2005;35:529–37.

[12] El Maaddawy T, Soudki K. Carbon-fiber-reinforced polymer repair to extendservice life of corroded reinforced concrete beams. J Compos Constr2005;9:187–94.

[13] Meier U. Carbon fiber reinforced polymer cables: why? Why not? What if?Arab J Sci Eng 2012;37:399–411.

[14] Tavakkolizadeh M, Saadatmanesh H. Galvanic corrosion of carbon and steel inaggressive environments. J Compos Constr 2001;5:200–10.

[15] Sloan FE, Talbot JB. Corrosion of graphite-fiber-reinforced compositesI – galvanic coupling damage. Corrosion 1992;48:830–7.

[16] Wielage B, Dorner A. Corrosion studies on aluminum reinforced with uncoatedand coated carbon fibers. Compos Sci Technol 1999;59:1239–45.

[17] Davis JR, editor. Corrosion of aluminum and aluminum alloys. MaterialsPark: ASM International; 1999. p. 313.

[18] Oldfield JM. Electrochemical theory of galvanic corrosion. In: Hack HP, editor.Galvanic corrosion, ASTM STP 979. West Conshohocken, PA: ASTMInternational; 1988.

[19] Cai BP, Liu YH, Ren CK, Liu ZK, Tian XJ, Abulimiti ABB. Experimental study ofgalvanic corrosion behaviour of carbon fibre composite coupled to aluminiumin artificial seawater. Corros Eng Sci Technol 2012;47:289–96.

[20] Payan S, Le Petitcorps Y, Olive J-M, Saadaoui H. Experimental procedure toanalyse the corrosion mechanisms at the carbon/aluminium interface incomposite materials. Composites Part A 2001;32:585–9.

[21] Wootton I, Spainhour L, Yazdani N. Corrosion of steel reinforcement in carbonfiber-reinforced polymer wrapped concrete cylinders. J Compos Constr2003;7:339–47.

[22] Hack HP. Galvanic corrosion. Ann Arbor: ASTM; 1988. p. 358.[23] Neil W, Garrard C. The corrosion behavior of aluminum-silicon carbide

composites in aerated 3.5% sodium chloride. Corros Sci 1994;36:837–51.

788 T.F.A. Santos et al. / Materials and Design 65 (2015) 780–788

[24] Ireland R, Arronche L, La Saponara V. Electrochemical investigation of galvaniccorrosion between aluminum 7075 and glass fiber/epoxy composites modifiedwith carbon nanotubes. Composites Part B 2012;43:183–94.

[25] Turnbull A. Review of corrosion studies on aluminum metal matrixcomposites. Br Corros J 1992;27:27–35.

[26] Dorner A, Wielage B, Schurer C. Improvement of the corrosion resistance of C/Al-composites by diamond-like carbon coatings. Thin Solid Films 1999;355–356:214–8.

[27] Möller H, Boshoff ET, Froneman H. The corrosion behaviour of a low carbon steelin natural and synthetic seawaters. J South Afr Inst Min Metall; 106: 585–592.

[28] Vasconcelos GC, De Aragão BJG, Souza WA, Nascimento CAM. Evaluation ofpolymeric matrices for application in composite core of conductor. In: 1st BrazConf Compos Mater (BCCM1), Natal; 2012.

[29] Wolstenholme J. Inexpensive zero-resistance ammeter for galvanic studies. BrCorros J 1974;2:1–2.

[30] Nayak S. The analogue of potentiometer for current Zero Resistance Ammeter(ZRA). Resonance 2013;18:378–86.

[31] Wang W-X, Takao Y, Matsubara T. Galvanic corrosion-resistant carbon fibermetal laminates. In: 16th Int conf compos mater. Kyoto; 2007.

[32] Coleman SL, Scott VD, McEnaney B. Corrosion behavior of aluminum-basedmetal matrix composites. J Mater Sci 1994;29:2826–34.

[33] Ghali E. Active and passive behavior of aluminum and magnesium and theiralloys. In: Corrosion resistance of aluminum and magnesium alloys:understanding, performance, and testing. New Jersey: John Wiley & Sons,Inc.; 2010. p. 78–118.

[34] Liu W, Singh A, Lin Y, Ebenso EE, Tianhan G, Ren C. Corrosion inhibition ofAl-alloy in 3.5% NaCl solution by a natural inhibitor: an electrochemical andsurface study. Int J Electrochem Sci 2014;9:5560–73.

[35] Zhou J, Lucas JP. Hygrothermal effects of epoxy resin. Part II: Variations of glasstransition temperature. Polymer 1999;40:5513–22.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具