nanomaterials Article

Monitoring Damage Propagation in Glass Fiber Composites Using Carbon Nanofibers Ahmed Al-Sabagh 1 , Eman Taha 1 , Usama Kandil 1 , Gamal-Abdelnaser Nasr 2 and Mahmoud Reda Taha 3, * 1 2 3

*

Egyptian Petroleum Research Institute, Nasr City, Cairo 11727, Egypt; [email protected] (A.A.-S.); [email protected] (E.T.); [email protected] (U.K.) Department of Physics, Faculty of Science, Cairo University, Giza 12613, Egypt; [email protected] Department of Civil Engineering, University of New Mexico, Albuquerque, NM 87131, USA Correspondence: [email protected]; Tel.: +1-505-277-1258; Fax: +1-505-277-1988

Academic Editor: Juan Hinestroza Received: 14 June 2016; Accepted: 1 September 2016; Published: 10 September 2016

Abstract: In this work, we report the potential use of novel carbon nanofibers (CNFs), dispersed during fabrication of glass fiber composites to monitor damage propagation under static loading. The use of CNFs enables a transformation of the typically non-conductive glass fiber composites into new fiber composites with appreciable electrical conductivity. The percolation limit of CNFs/epoxy nanocomposites was first quantified. The electromechanical responses of glass fiber composites fabricated using CNFs/epoxy nanocomposite were examined under static tension loads. The experimental observations showed a nonlinear change of electrical conductivity of glass fiber composites incorporating CNFs versus the stress level under static load. Microstructural investigations proved the ability of CNFs to alter the polymer matrix and to produce a new polymer nanocomposite with a connected nanofiber network with improved electrical properties and different mechanical properties compared with the neat epoxy. It is concluded that incorporating CNFs during fabrication of glass fiber composites can provide an innovative means of self-sensing that will allow damage propagation to be monitored in glass fiber composites. Keywords: carbon nanofibers; glass fiber composites; self-sensing; damage monitoring

1. Introduction In recent years, textile fabric composites have attracted widespread attention owing to their superior properties, making them attractive materials for many applications like smart fabrics and the aerospace, marine, and automobile industries [1,2]. Glass fiber textile composites, in particular, are being widely researched currently due to their relatively high strength-to-weight ratio at low cost. The properties of glass fiber textile-reinforced polymer composites greatly depend on the fabric structure as well as the matrix properties. Many attempts have been made to improve their properties for high performance textile fabric composites. Several studies reported considerable improvement in fracture toughness of glass fiber reinforced polymer (GFRP) when fumed silica, carbon black, carbon nanotubes (CNTs) [3] and carbon nanofibers (CNFs) [4] were incorporated in polymer nanocomposites. The tensile strength of GFRP was improved using carbon nanofibers [5] and CNTs [6]. Significant enhancements in thermal and electrical conductivities of GFRP with carbon nanomaterials were reported [7–9]. In this sense, nanomaterials play an important and significant role in controlling the intrinsic properties of glass fiber composites. Carbonaceous nanomaterials such as CNTs and CNFs are the most common and the most promising additives used to fabricate multifunctional glass textile nanocomposites with enhanced capabilities. For instance, CNFs are attractive candidates in conductivity-related applications because of their relatively high electrical conductivity [10]. A percolated concentration Nanomaterials 2016, 6, 169; doi:10.3390/nano6090169

www.mdpi.com/journal/nanomaterials

Nanomaterials 2016, 6, 169

2 of 12

must be incorporated to transfer the composite from an insulate state into conductive state, which is known as the percolation threshold. At this concentration, the conductivity is improved by several orders of magnitude and percolation network is achieved between conductive particles [11]. Yang et al. [12] found that the electrical conductivity of CNFs-filled polystyrene (PS) composite improved by ten orders of magnitude over that of the neat PS with percolation threshold 4 wt %, providing high electromagnetic interference shielding (EMI) at low filler loading. Yang at el. [13] confirmed the percolation threshold of CNFs to be 5 wt % with a decrease in electrical resistivity by 11 orders of magnitude. Lozano at el. [14] examined the electrical behavior of CNFs-filled polypropylene (PP) for electrostatic dissipation (ESD) applications, and the percolation was found to be 9–18 wt %. It was also found that the percolation threshold of CNFs/polypropylene (PP) composites was about 3–5 vol % depending on the degree of graphitization (graphite perfection) of CNFs [15,16]. Despite the fact that CNFs significantly improve the electrical conductivity of polymers, there has been very limited research using CNFs for self-sensing applications. Park et al. [17] found that CNFs/epoxy composites have reliable self-sensing capability under both static and cyclic loading conditions. The aspect ratio of CNFs was reported to have significant effects in the formation of electrically percolated networks. In addition, prior works [18,19] have confirmed that damage monitoring using conductive CNFs composites is strongly dependent on forming a three-dimensional conductive network inside the polymer matrix. To date, the ability of CNFs in monitoring damage propagation in glass textile-reinforced polymer composites has not yet been reported. In this study, the ability of CNFs to improve electrical conductivity and alter the mechanical properties of epoxy nanocomposites is investigated. The conductive CNFs/epoxy nanocomposite is then used to monitor damage propagation in glass textile-reinforced polymer composites under static loading. 2. Materials and Methods 2.1. Materials and Fabrication CNFs were supplied by Nanostructured & Amorphous Materials Inc. They had diameter of 80–200 nm and a length of 0.5-20 µm and thus an aspect ratio ranging between 6.3 and 100. The epoxy used in fabrication was EPOTUF® 37-127 epoxy system supplied by U.S. Composites, Inc. (West Palm Beach, FL, USA) The epoxy resin is low viscosity, 100% reactive diluted liquid based on Bisphenol-A containing glycidyl ether. The hardener was Aliphatic Amine EPOTUF® 37-614. The resin to hardener mixing ratio was 2:1. The bidirectional S-Glass fiber fabric was supplied by ACP Composites, Inc. (Livermore, CA, USA). The performance of the CNFs-polymer nanocomposite is affected by the homogeneity of nanofibers dispersion into the polymer matrix. Several dispersion methods were recommended in the literature to obtain homogeneously dispersed CNFs in the polymer matrix and to avoid agglomeration. CNFs with different contents (0, 0.3, 0.5, 1.0, 1.5, 2.0 and 2.5 wt %) were first hand-stirred into the epoxy resin, and then sonicated in a path sonicator for 1 h at 40 ◦ C and frequency of 40 kHz. The resin-CNFs mixture was further dispersed using a high shear mixer at speed 11,000 rpm at a temperature of 90 ◦ C for 1 h. The resin-CNFs mixture was then mechanically stirred at temperature of 90 ◦ C for 2 h and a speed of 800 rpm. The resin-CNFs mixture was degassed to remove the bubbles for 30 min at 50 ◦ C and then left to cool for 1 h at room temperature. After cooling, the epoxy hardener was hand-stirred into the resin-CNFs mixture for 5 min and left overnight. CNFs/epoxy nanocomposite was then cured for 2.5 days at 110 ◦ C to ensure full curing. To prepare glass fiber reinforced (GFRP) composites, after adding the epoxy hardener to the resin-CNFs mixture, that mixture was then used to fabricate the GFRP using the hand layup technique. Six layers of bidirectional plain weave glass fiber textile fabrics were laid in 0◦ direction, and then vacuum pressure was applied for 24 h. The glass fiber composite plates were then cured for 2.5 days at 110 ◦ C to insure complete curing. 2 wt % CNFs were used to fabricate glass fiber composites.

Nanomaterials 2016, 6, 169

3 of 12

Nanomaterials 2016,used 6, 169 for producing the glass fiber composites was based on the electrical percolation 3 of 13 The CNFs content observations of epoxy-CNFs nanocomposites discussed below. Figure 1 presents schematically the Nanomaterials 2016, 6, 169 3 of 13 electrical percolation observations of epoxy-CNFs nanocomposites discussed below. Figure 1 glass fiber composite fabrication method. Fiber volume fraction of glass fiber composites incorporating presents schematically the glass fiber composite fabrication method. Fiber volume fraction of glass electrical observations of epoxy-CNFs nanocomposites discussed below.ToFigure 1 the 2.0 wt % CNFspercolation was determined using ASTM D3171 [20] and was found to be 55.6%. examine fiber composites incorporating 2.0 wt % CNFs was determined using ASTM D3171 [20] and was presents schematically the glass fiber composite fabrication method. Fiber volume fraction of glasswere dispersion of CNFs in the epoxy matrix, fractured surfaces of the epoxy-CNFs nanocomposites found to be 55.6%. To examine the dispersion of CNFs in the epoxy matrix, fractured surfaces of the fiberwith composites incorporating 2.0 wtthen % CNFs was determined using ASTM D3171Scanning [20] and was covered a layer of gold and were the Field Emission epoxy-CNFs nanocomposites were covered investigated with a layer ofusing gold and were then investigated using Electron the found to be 55.6%. To examine the dispersion of CNFs in the epoxy matrix, fractured surfaces of the Microscope (FESEM) using Quanta FEI Company’s 250. Fourier infrared Field Emission Scanning Electron250, Microscope (FESEM)Quanta using Quanta 250, FEItransform Company’s Quanta(FTIR) epoxy-CNFs nanocomposites were covered with a layer of gold and were then investigated using the 250.was Fourier transformusing infrared (FTIR) was also recorded using Nicolet IS-10 FTIRwithin spectra also recorded Nicolet IS-10spectra FTIR spectrophotometer-Thermo Fisher Scientific Field Emission Scanning Electron Microscope (FESEM) using Quanta 250, FEI Company’s Quanta −1 − 1 spectrophotometer-Thermo 400–4000 cm wave number. Fisher Scientific within 400–4000 cm wave number. 250. Fourier transform infrared (FTIR) spectra was also recorded using Nicolet IS-10 FTIR

spectrophotometer-Thermo Fisher Scientific within 400–4000 cm−1 wave number.

Figure 1. Schematic representation of glass fiber composite fabrication incorporating CNFs.

Figure 1. Schematic representation of glass fiber composite fabrication incorporating CNFs. Figure Schematic representation fiber composite fabrication incorporating CNFs. 2.2. Electrical and1.Mechanical Measurementsofofglass the Composites

2.2. Electrical and Mechanical Measurements of the Composites

The electrical conductivity of CNFs/epoxy nanocomposites was determined according 2.2. Electrical and Mechanical Measurements of thepolymer Composites The electrical of CNFs/epoxy polymer determined to ASTM D257 conductivity [21]. Measurements were performed usingnanocomposites a Keithley 2636b was source meter and according strip The electrical conductivity of CNFs/epoxy polymer nanocomposites was determined according electrodes a standard two-probewere configuration as shown 2. 2636b source meter and strip to ASTM D257via [21]. Measurements performed usingin a Figure Keithley to ASTM D257 [21]. Measurements were performed using a Keithley 2636b source meter and strip electrodes via a standard two-probe configuration as shown in Figure 2. electrodes via a standard two-probe configuration as shown in Figure 2.

Figure 2. The strip electrode used to determine the electrical resistance of CNFs/epoxy nanocomposites. Figure paint 2. The strip electrode used to determine the electrical resistance of CNFs/epoxy Silver used to ensure good contact between the specimens and nanocomposites. thenanocomposites. electrode. The Figure 2. The stripwas electrode used to determine the electrical resistance of CNFs/epoxy electrical conductivity (σ) was calculated using Equation (1). Silver paint was used to ensure good contact between the specimens and the electrode. The electrical conductivity (σ) was calculated usingcontact Equationbetween (1). Silver paint was used to ensure good the specimens and the electrode. (1)

= The electrical conductivity (σ) was calculated using Equation (1). (1) A is the cross sectional area, L is the length, and R=is the measured electrical resistance. L properties of CNFs/epoxy nanocomposites, To identify the significance of CNF on the elastic σ= A is the cross sectional area, L is the length, and R is the measured electrical resistance. three specimens of 20 mm × 10 mm × 3 mm were AR tested. Dynamic mechanic analysis (DMA) was

To identify the significance of CNF on the elastic properties of CNFs/epoxy nanocomposites, performed on a Triton Instruments, operating in the tension mode at an oscillation frequency of 1 Hz.

(1)

A is the cross sectional L is themm length, andwere R is tested. the measured three specimens of area, 20 mm × 10 × 3 mm Dynamicelectrical mechanicresistance. analysis (DMA) was Data were collected from room temperature to 150 °C at a scanning rate of 10 °C/min. For static performed onthe a Triton Instruments, operating in elastic the tension mode at of an oscillation frequency of 1 Hz. To identify significance of CNF on the properties CNFs/epoxy nanocomposites, tension properties of CNFs/epoxy nanocomposites, three specimens of 20 mm × 10 mm × 2 mm were Data were collected from room temperature to 150 °C at a scanning rate of 10 °C/min. For static was threetested specimens of 20 Instruments, mm × 10 mm × 3 mm were tested. Dynamic mechanic analysis (DMA) using Triton operating in the static tension mode with preload force 0.01 N, load tension properties of CNFs/epoxy nanocomposites, three specimens of 20 mm × 10 mm × 2 mm were performed on a Triton Instruments, operating in the tension mode at an oscillation frequency of 1 Hz. rate 0.2 N/min and to maximum force 2 N. using Triton Instruments, operating in the static mode with preload force 0.01 N, load ◦ Ctension ◦ C/min. Data tested were collected from room temperature to 150 at a scanning rate of 10 For Three GFRP composite coupons of 19 mm × 150 mm were tested under off-axis (i.e., load was static rate 0.2 N/min and to maximum force 2 N. applied at 45° with respect to thenanocomposites, fiber direction) static monotonically increasing The were tension properties of CNFs/epoxy three specimens of 20 mm ×tension 10 mmstress. × 2 mm Three GFRP composite coupons of 19 mm × 150 mm were tested under off-axis (i.e., load was reason for choosing off-axis loading was toinsimulate realistic loading conditions where stresses are tested using Triton Instruments, operating the static tension mode with preload force 0.01 N, load applied at 45° with respect to the fiber direction) static monotonically increasing tension stress. The generated in any direction to the glass fiber composite. It is also well established that behavior of the rate 0.2 N/min and to maximum force was 2 N.to simulate realistic loading conditions where stresses are reason for choosing off-axis loading fiber composites in the off-axis direction is governed by the polymer matrix rather than the fibers [6]. Three GFRP composite coupons 19 composite. mm × 150 tested under off-axis of (i.e., generated in any direction to the glassof fiber It ismm also were well established that behavior the load Therefore, off-axis tension would best show the significance of CNFs. The static tension tests were was applied at 45◦ with the fiber direction)bystatic monotonically increasing fiber composites in therespect off-axisto direction is governed the polymer matrix rather than thetension fibers [6].stress. Therefore, off-axis tension would bestwas show significance of CNFs. The static tension testsstresses were are The reason for choosing off-axis loading tothe simulate realistic loading conditions where

Nanomaterials 2016, 6, 169

4 of 12

generated in any direction to the glass fiber composite. It is also well established that behavior of the fiber composites in the off-axis direction is governed by the polymer matrix rather than the fibers [6]. Therefore, off-axis tension would best show the significance of CNFs. The static tension tests were performed using MTS® Bionex servo hydraulic machine. A displacement control protocol was used in the static tension tests according to the ASTM standards methods D3039/D3039M [22] with a loading Nanomaterials 2016, 6, 169 4 of 13 rate of 1.0 mm/min. The electrical resistance of the glass fiber composite specimens was measured ® Bionex during the tension test using a Keithley 2636B source meter. Conductivecontrol electrodes were performed using MTS servo hydraulic machine. A displacement protocol wasapplied used at in the static tension tests according to the ASTM standards methods D3039/D3039M [22] with a the glass fiber composite coupon using silver paint at two points spaced by 50 mm to allow electric loading rate of 1.0 mm/min. The electrical resistance of the glass fiber composite specimens was resistance measurements. Schematic representation of the electrical resistance measurement during measured during theincreasing tension testtension using a test Keithley 2636Bin source meter. the static monotonically is shown Figure 3. Conductive electrodes were applied at the glass fiber composite coupon using silver paint at two points spaced by 50 mm to allow Damage in glass fiber composite coupons was estimated in terms of the change of the electrical electric resistance measurements. Schematic representation of the electrical resistance measurement conductivity during loading. The electrical conductivity was measured and a metric of damage based during the static monotonically increasing tension test is shown in Figure 3. on electrical conductivity change denoted DE (t) was wasestimated calculated using Equation (2). Damage in glass fiber composite coupons in terms of the change of the electrical conductivity during loading. The electrical conductivity was measured and a metric of damage based σ (t) on electrical conductivity change denoted (t) was DE (tD) E= 1 − calculated % using Equation (2). (2)

σ ( t0 )

= 1−

σ

%

(2)

where DE (t) is the electrical damage measured at time σt, σ(t0 ) is the initial electrical conductivity of the composite application at time t0 , andatσ(t) conductivity the composite whereprior DE(t) to is load the electrical damage measured timeist,the σ(telectrical 0) is the initial electrical of conductivity of at time t.the Moreover, a metric ofload damage based on modulus composite prior to application at change time t0,ofand σ(t) is of theelasticity, electricalrepresenting conductivity mechanical of the composite at timeDt. Moreover, a metric of damage based on change of modulus of elasticity, damage and denoted (t) was calculated using Equation (3): M representing mechanical damage and denoted DM(t) was calculated using Equation (3):

D M (t) = 1 −

E (t)

%

= 1 − E (t0 )%

(3)

(3)

0) isinitial the initial tangent modulus of elasticity glassfiber fibercomposite composite coupon coupon at t00 and wherewhere E(t0 ) E(t is the tangent modulus of elasticity of of thethe glass and E(t) is themodulus tangent modulus of elasticity of thefiber glasscomposite fiber composite coupon at time t. A minimum is the E(t) tangent of elasticity of the glass coupon at time t. A minimum tangent tangent modulus of zero was assumed to account for the descending stress-strain. modulus of zero was assumed to account for the descending stress-strain.

Figure 3. Schematic of electrical resistance measurement of glass fiber composite coupons during

Figure 3. Schematic of electrical resistance measurement of glass fiber composite coupons during tension tests. tension tests.

3. Results and Discussion

3. Results and Discussion 3.1. Mechanical Properties of CNFs/Epoxy Nanocomposites

3.1. Mechanical Properties of CNFs/Epoxy Nanocomposites

The mechanical properties of CNFs/epoxy nanocomposites can be observed from a static tension

The properties of CNFs/epoxy nanocomposites can observed from a static tension test, mechanical shown in Figure 4a,b. The stress-strain relationship observed forbe CNFs/epoxy nanocomposites at roomintemperature and thestress-strain change in elastic modulus (E) versus the concentrations are shown at test, shown Figure 4a,b. The relationship observed forCNFs CNFs/epoxy nanocomposites in Figure 4a,b respectively. The results show strong influence of the presence of the CNFs room temperature and the change in elastic modulus (E) versus the CNFs concentrations on arethe shown mechanical properties of CNFs/epoxy nanocomposites, a gradual increase of material stiffness in Figure 4a,b respectively. The results show strong influence of the presence of the CNFs by on the increasing the amount of CNFs within the epoxy matrix up to 1.0 wt % of CNFs and then a sharp mechanical properties of CNFs/epoxy nanocomposites, a gradual increase of material stiffness by decrease in the material stiffness for composite filled 1.5 wt %, 2.0 wt % and 2.5 wt % CNFs. Moreover, increasing the amount of CNFs within the epoxy matrix up to 1.0 wt % of CNFs and then a sharp the elastic modulus increases from 200 MPa to 700 MPa with increasing the CNFs concentrations from 0 to 1.0 wt % and drops to 200 MPa at 1.5 wt % CNFs, to 90 MPa at 2.0 wt % CNFs and to 100 MPa at 2.5 wt %. The reduction of elastic modulus at 2.0 wt % and 2.5 wt % CNFs/epoxy composite

Nanomaterials 2016, 6, 169

5 of 12

decrease in the material stiffness for composite filled 1.5 wt %, 2.0 wt % and 2.5 wt % CNFs. Moreover, the elastic modulus increases from 200 MPa to 700 MPa with increasing the CNFs concentrations from 0 to 1.0 wt % and drops to 200 MPa at 1.5 wt % CNFs, to 90 MPa at 2.0 wt % CNFs and to 100 MPa at 2.5 wt %. The reduction of elastic modulus at 2.0 wt % and 2.5 wt % CNFs/epoxy composite could be Nanomaterials 2016, 6, 169 5 of 13 attributed to the effect of CNFs on the epoxy network. We suggest that high content of CNFs reduced could bedegree attributed theepoxy effect ofmatrix. CNFs onFurther the epoxyin-depth network. We suggest that high content of CNFs below to the crosslinking of to the investigations were conducted reduced the crosslinking degree of the epoxy matrix. Further in-depth investigations were conducted examine this hypothesis. below to examine this hypothesis.

(a)

(b) Figure 4. (a) The stress-strain relationship for CNFs/epoxy nanocomposites; and (b) the change in

Figure 4. (a) The stress-strain for CNFs/epoxy nanocomposites; and (b) the change in elastic modulus versus therelationship CNFs concentrations for CNFs/epoxy nanocomposites. elastic modulus versus the CNFs concentrations for CNFs/epoxy nanocomposites. To better understand the significance of fillers/fibers on the mechanical properties of polymer composites, a number of models exist in the literature [23–25]. We further examine here the model by better understand the significance of fillers/fibers on the mechanical properties of Christensen [24] to predict the elastic modulus of CNFs/epoxy nanocomposites:

To polymer composites, a number of models exist in the literature [23–25]. We further examine here the model by 1 (4) = 1 Christensen [24] to predict the elastic modulus 1of − CNFs/epoxy 6 4 6 nanocomposites: " ! and # Vf is the filler/fiber where Ec, Em and Ef are the moduli of composite, matrix and filler respectively Vf Vf Vf 2 1 volume fraction. Ec = E + 1+ + Em (4) 6 f 1 − Vf Young’s4modulus6of CNFs/epoxy nanocomposites Figure 5 compares the experimentally measured with the values predicted using Christensen’s model [25]. The predicted values given by model showed a good agreement with the experimental observation of Ec except at where Ec , EChristensen’s m and Ef are the moduli of composite, matrix and filler respectively and Vf is the filler/fiber CNFs contents above 1.0 wt % CNFs. At low CNFs contents, the significance of the CNFs on the volume fraction. epoxy matrix is negligible, and thus CNFs work as fibers that reinforce the epoxy matrix and improve Figure 5 compares theas experimentally measured Young’s modulus CNFs/epoxy its mechanical properties a composite. This behavior follows classical composite theory of and thus nanocomposites with the values predicted using Christensen’s model [25]. The predicted values given by Christensen’s model showed a good agreement with the experimental observation of Ec except at CNFs contents above 1.0 wt % CNFs. At low CNFs contents, the significance of the CNFs on the epoxy matrix is negligible, and thus CNFs work as fibers that reinforce the epoxy matrix

Nanomaterials 2016, 6, 169

6 of 13

Nanomaterials 2016, 6, 169

6 of 12

good agreement with the model can be observed. Conversely, at 2.0 wt % and 2.5 wt % CNFs (i.e., higher CNFs 2016, concentrations), CNFs start to induce a chemical effect on the epoxy matrix, hindering Nanomaterials 6, 169 6 of 13 and improvebetween its mechanical properties a composite. This behavior follows classical the reaction the resin and the as hardener and affecting epoxy crosslinking. Thiscomposite adversely theory thus good with the model Conversely, can be observed. Conversely, at wt and goodand agreement with agreement the model can observed. at 2.0 wt % effect and 2.5ofwt %2.0 CNFs (i.e., affects the mechanical properties of be CNFs/epoxy nanocomposite. The CNFs on % epoxy 2.5 wt % CNFs higher CNFs CNFs start induce athe chemical effectmodels on the epoxy higher CNFs concentrations), CNFs start to induce a chemical effect epoxy matrix, hindering crosslinking is (i.e., not considered inconcentrations), Christensen’s model [25] to and all on other composite in the the reaction between the resin hardener andthe affecting epoxy crosslinking. Thiscrosslinking. adversely matrix, hindering the reaction between the resin and hardener andThis affecting literature, which only account forand thethe reinforcing effect of the fillers. modelepoxy therefore failed to affects mechanical properties ofproperties CNFs/epoxy nanocomposite. The effect high of The CNFs oncontent epoxy This adversely affectsofthe mechanical of CNFs/epoxy nanocomposite. effect of CNFs predict thethe modulus elasticity of CNFs/epoxy nanocomposite incorporating CNFs as crosslinking is not considered in Christensen’s model [25] and all other composite models in the on epoxy is not considered in Christensen’s model and(at all the other composite models in shown in crosslinking Figure 5. Interestingly, composite filled with 1.5 wt %[25] CNFs transition between low literature, which only account for the reinforcing effect of the fillers. This model therefore failed to the literature, which only account for the reinforcing effect of the fillers. This model therefore failed to and high CNFs contents) has no positive or negative effect on the elastic modulus. The 1.5 wt % CNFs predict modulus elasticityofof CNFs/epoxy nanocomposite incorporating content as as predict thethe modulus ofof elasticity CNFs/epoxy nanocomposite incorporating highCNFs CNFs content is not high enough to chemically affect epoxy crosslinking. In addition, this high concentration failed to shown in Figure 5. Interestingly, composite filled with 1.5 wt % CNFs (at the transition between low shown in Figure 5.filler, Interestingly, filled withmodulus 1.5 wt % of CNFs (at thenanocomposite. transition between low act as reinforcing having nocomposite effect on the elastic the epoxy high CNFs contents) hasno nopositive positiveor ornegative negative effect effect on on the %% CNFs andand high CNFs contents) has the elastic elasticmodulus. modulus.The The1.5 1.5wtwt CNFs is not high enough to chemically affect epoxy crosslinking. In addition, this concentration failed to is not high enough to chemically affect epoxy crosslinking. In addition, this concentration failed to act act as reinforcing filler, having no effect on the elastic modulus of the epoxy nanocomposite. as reinforcing filler, having no effect on the elastic modulus of the epoxy nanocomposite.

Figure 5. The experimental and predicted values of Young’s modulus of CNFs/epoxy nanocomposites. The experimental andpredicted predicted valuesof ofinteraction of nanocomposites. Figure 5. 5. The experimental and values of Young’s modulus modulus ofCNFs/epoxy CNFs/epoxy nanocomposites. ToFigure further confirm the above hypothesis between the CNFs and epoxy and the

chemical significance of high CNFs on epoxy, FTIR analysis was performed. FTIR is utilized to To further confirm the above hypothesis of interaction between the CNFs and epoxy and the observe chemical changes the epoxy matrixofafter incorporating The FTIR spectra of the the To further confirm theof above hypothesis interaction betweenCNFs. the CNFs and epoxy and chemical significance of high CNFs on epoxy, FTIR analysis was performed. FTIR is utilized to epoxy matrix with and without CNFs is presented in Figure 6. Figure 6 shows the absorption bands chemical significance of high CNFs on epoxy, FTIR analysis was performed. FTIR is utilized to observe observe chemical changes of the epoxy matrix after incorporating CNFs. The FTIR spectra of the corresponding to C–H cm−1after ), epoxide ring (~830 cm−1), N–H of primary amines (1590–1640 chemical changes of the(2800–2970 epoxy matrix incorporating CNFs. The FTIR spectra of the epoxy matrix epoxy matrix with and without CNFs is presented in Figure 6. Figure 6 shows the absorption bands −1) and ether (~1250 cm−1). cm−1),and O–H groupsCNFs (3200–3600 cm−1),in C–N (1040–1120 cm with without is presented Figure 6. Figure 6 shows the absorption bands corresponding corresponding to C–H (2800–2970 cm−1), epoxide ring (~830 cm−1), N–H of primary amines (1590–1640 −1 ), epoxide ring (~830 cm−1 ), N–H of primary amines (1590–1640 cm−1 ), O–H to C–H cm(3200–3600 cm−1),(2800–2970 O–H groups cm−1), C–N (1040–1120 cm−1) and ether (~1250 cm−1). groups (3200–3600 cm−1 ), C–N (1040–1120 cm−1 ) and ether (~1250 cm−1 ).

Figure6. 6. FTIR FTIR spectra spectra of neat epoxy epoxy and CNFs/epoxy CNFs/epoxy nanocomposites. Figure nanocomposites. Figure 6. FTIR spectraofofneat neat epoxy and and CNFs/epoxy nanocomposites.

Nanomaterials 2016, 6, 169

7 of 12

Comparing the FTIR spectrum of 2.0 wt % CNFs/epoxy with other spectra of CNFs/epoxy polymer nanocomposites and the neat epoxy, it can be observed that incorporating 2.0 wt % CNFs in the epoxy matrix caused an increase in the epoxide ring and primary N–H band intensity. In addition, the band of the hydroxyl group is significantly shifted to a lower intensity and to a lower wave number value (3410–3320 cm−1 ) at this wt % of CNFs compared with other wt %. It is obvious from the FTIR spectra in Figure 6 that for the neat and low concentration CNFs samples (0.3, 0.5 and 1.0 wt %) the O–H peak appeared almost around 3410 cm−1 . This is attributed to the fact that at such low CNFs concentrations, there is a minor effect from CNFs on epoxy curing. Incorporating CNFs concentrations of 1.5, 2.0 and 2.5 wt % into the epoxy matrix appeared to shift the O–H peak to 3370 cm−1 , 3320 cm−1 and 3310 cm−1 , respectively. This may be attributed to the effect of CNFs on the curing behavior of the epoxy matrix [26]. It is well known that the broad complex band of the hydroxyl stretching vibration region at about 3200–3600 cm−1 is attributed to the combined effect of the differently associated hydroxyl groups, i.e., hydrogen bonding between hydroxyl and hydroxyl/carbonyl groups of different strength and hydrogen bonding of water molecules. In addition, a matrix having O–H groups could undergo two modes of hydrogen bonding; inter- and intramolecular hydrogen bonds between O–H groups [27]. Consequently, it could be concluded that incorporating 2.0 wt % CNFs in the epoxy matrix reduced its network formation process via lowering the crosslinking bonds and consequently changed the ratios of hydrogen bonding modes which lead to different geometry with different force constants and consequently shifting the wave number absorption value. Overall, the O–H band intensity decreases and is shifted to a lower wave number. This observed progressive shift of the (υ O–H) band from about 3410 cm−1 toward lower wave numbers (3320 cm−1 ) can be attributed to redistribution in the arrangement of hydroxyl group association due to the different geometry caused by lowered cross-linked matrix. To further confirm the above hypothesis, DMA testing was conducted on neat epoxy and CNFs/epoxy nanocomposites. According to the rubber elasticity theory modified by Nielsen [28], the crosslinking density was determined from the storage modulus of composites in rubbery plateau using the following equations [29,30]: E υ= (5) 3RT where υ is the crosslinking density, E is the storage modulus at (Tg + 50), T is the absolute temperature of (Tg + 50) K and R is the gas constant. The molecular weight between crosslinks was given using Equation (6). ρ (6) Mc = υ where Mc is molecular weight between crosslinks and ρ is the density of the polymer composite. The degree of crosslinking Xlink of the polymer composites is suggested as an inverse for the molecular weight between crosslinks in a unit volume as in Equation (7): X Link =

1 MC

(7)

The calculated degree of crosslinking Xlink is presented in Figure 7. The results show that incorporating 0.3, 0.5, 1.0 and 1.5 wt % CNFs in the epoxy resin slightly reduced the crosslinking by 14.8%, 19.7%, 13.8% and 20.5% respectively as can be observed in Figure 4a,b. Such a relatively low decrease in epoxy crosslinking does not have a significant negative effect on mechanical behavior of CNFs/epoxy nanocomposites due to the compensation for that decrease with the high increase in mechanical properties (e.g., modulus of elasticity) by CNFs as fibers in the matrix. On the contrary, incorporating 2.0 and 2.5 wt % CNFs in the epoxy resin reduced the crosslinking by 44% and 38% respectively owing to strong interaction between CNFs and the epoxy matrix, leading to degradation of epoxy crosslinking resulting in sharp drop in elastic modulus. The above analysis, consistent with

Nanomaterials 2016, 6, 169

8 of 12

the FTIR observations, proved that the degree of crosslinking plays a significant role in the mechanical behavior of CNFs/epoxy nanocomposites.

Figure 7. The degree of crosslinking for CNFs/epoxy nanocomposites.

Figure 7. The degree of crosslinking for CNFs/epoxy nanocomposites. Figure 7. The degree Nanocomposites of crosslinking for CNFs/epoxy nanocomposites. of CNFs/Epoxy 3.2. Electrical Properties

3.2. Electrical Properties of CNFs/Epoxy Nanocomposites

The epoxy matrix has inherently poor electrical conductivity. The most effective method to

3.2. Electrical Properties of CNFs/Epoxy Nanocomposites

The epoxy matrix hasisinherently poor electrical conductivity. mostconductivities effective method to overcome this limitation to add conducting fillers. Figure 8 shows theThe electrical (σ) The matrixishas inherently poor filler electrical conductivity. The most effective method to (σ) for CNFs/epoxy nanocomposites at different loadings. The unfilled neat epoxy (atconductivities 0 wt %) is an overcome thisepoxy limitation to add conducting fillers. Figure 8 shows the electrical −7 S/m. When overcomethat thisnanocomposites limitation is electrical to add fillers. Figure 8The shows the electrical conductivities (σ) is an insulator exhibits low conductivity ofloadings. approximately 2.1 × 10 CNFs were for CNFs/epoxy at conducting different filler unfilled neat epoxy (at 0 wt %) for CNFs/epoxy nanocomposites at different filler loadings. The The unfilled neat−epoxy (at 0 wt %) iswas an 7 added to epoxy resin, the electrical conductivity increased. percolation phenomenon insulator that exhibits low electrical conductivity of approximately 2.1 × 10 S/m. When CNFs were −7 S/m. When CNFs were insulator that exhibits low electrical conductivity of approximately 2.1 × 10nanocomposites observed when, at low CNFs content, the conductivity of the CNFs/epoxy was very added to epoxy resin, the electrical conductivity increased. The percolation phenomenon was observed addedto to resin, the electrical conductivity The CNFs percolation phenomenon close theepoxy conductivity of the epoxy matrix, becauseincreased. no conductive network was formed.was As when,observed at low CNFs content, the conductivity of the CNFs/epoxy nanocomposites was very close when, low CNFs content, the conductivity of the nanocomposites was veryto the more CNFs wereatadded, the fibers moved close together, andCNFs/epoxy at a certain concentration (percolation conductivity of conductivity the epoxy matrix, because no conductive CNFs network was formed. As more close to the of the epoxy matrix, because conductive CNFs network wasconductivity formed. AsCNFs threshold), a continuous conductive network of carbonno nanofibers was formed and the wereincreased added, the fibers moved together, and at a certain (percolation threshold), more CNFs added, theofclose fibers moved close together, and atconcentration a certain concentration (percolation bywere several orders magnitude. a continuous conductive network of carbon nanofibers formed and the conductivity increased by threshold), a continuous conductive network of carbonwas nanofibers was formed and the conductivity increased orders of magnitude. several ordersby of several magnitude.

Figure 8. Change in electrical conductivity vs. weight percentage of CNFs in epoxy. The right sketch is a schematic describing the significance of altering CNFs content on the formation of the conductive Figure 8.inside Change in electrical conductivityvs. vs.weight weight percentage percentage of right sketch network Figure 8. Change inepoxy. electrical conductivity ofCNFs CNFsininepoxy. epoxy.The The right sketch is is a schematic describing the significance of altering CNFs content on the formation of the conductive a schematic describing the significance of altering CNFs content on the formation of the conductive network inside epoxy. The CNFs/epoxy percolated at 1.5 wt % which was determined by the conventional method, i.e., network inside epoxy.

the peak position of dlogσ/dC [31] and the electrical conductivity (σ) at 2.0 wt % is approximately percolated 1.5 wt8%iswhich was to determined bymechanism the conventional i.e., 1.81 ×The 103CNFs/epoxy S/m. The illustration inatFigure proposed explain the of themethod, percolation The CNFs/epoxy percolated ataspect 1.5 wtthe % electrical which was determined thewt conventional method, the peak position dlogσ/dC [31] and conductivity (σ) by at 2.0 %inisachieving approximately phenomenon. Theof relatively low ratio of CNFs played an important role good i.e., the peak dlogσ/dC and the conductivity at 2.0 wtof %theis percolation approximately 1.81 ×position 103 S/m. The illustration in Figure 8high iselectrical proposed to explain the (σ) mechanism dispersion andof resulted in a[31] relatively percolation limit compared with other nanofillers 3 phenomenon. Theillustration relatively low aspect ratio of CNFs played an important role in achieving good 1.81 × 10 S/m. The in Figure 8 is proposed to explain the mechanism of the percolation dispersion The and relatively resulted inlow a relatively highofpercolation limitan compared with other nanofillersgood phenomenon. aspect ratio CNFs played important role in achieving

Nanomaterials 2016, 6, 169

9 of 12

dispersion and resulted Nanomaterials 2016, 6, 169 in a relatively high percolation limit compared with other nanofillers9 reported of 13 in theNanomaterials literature2016, [19].6, 169 Figure 9 shows the FESEM images for 0.5 and 2.0 wt % CNFs in the epoxy matrix. 9 of 13 reported in the literature [19]. Figure 9 shows the FESEM images for 0.5 and 2.0 wt % CNFs in the The images show the absence of agglomeration of CNFs as an indication of the effective dispersion reported in theThe literature [19]. Figure 9 shows the FESEM images for 9c 0.5an and 2.0a wt %ofCNFs in the epoxy matrix. images show the absence of agglomeration of CNFs as indication the effective process using sonication, shear mixing and mechanical string. Figure shows close view of 2.0 wt % epoxy matrix. The images the absence agglomeration of CNFsstring. as an indication the effective dispersion process using show sonication, shearof mixing and mechanical Figure 9cofshows a close CNFs in epoxy with the network formation that exhibited percolation behavior. The blue arrows mark view of 2.0process wt % CNFs in sonication, epoxy with shear the network that exhibited behavior. The dispersion using mixingformation and mechanical string. percolation Figure 9c shows a close the conductive network which occurs as the CNFs are getting close to each other, creating electron view 2.0 wtmark % CNFs in epoxy with the network thatCNFs exhibited percolation The blue of arrows the conductive network which formation occurs as the are getting closebehavior. to each other, pathsblue through the matrix and thusthe increasing electrical conductivity. arrows mark the conductive network occurs as the CNFs are conductivity. getting close to each other, creating electron paths through matrixwhich and thus increasing electrical creating electron paths through the matrix and thus increasing electrical conductivity.

(a) (b) (c) (a) (b) (c) Figure 9. FESEM images of CNFs in the epoxy matrix; (a) 0.5 wt %; (b) 2.0 wt %; and (c) a close view Figure 9. FESEM images of CNFs in the epoxy matrix; (a) 0.5 wt %; (b) 2.0 wt %; and (c) a close view of of 2.0 wt % CNFSimages showsof theCNFs formation conductive Figure 9. FESEM in the of epoxy matrix;network (a) 0.5 wtinside %; (b)epoxy. 2.0 wt %; and (c) a close view 2.0 wt % CNFS shows the formation of conductive network inside epoxy. of 2.0 wt % CNFS shows the formation of conductive network inside epoxy.

3.3. Monitoring Damage Propagation of GFRP Composite under Static Tension 3.3. Monitoring Damage Propagation underStatic StaticTension Tension 3.3. Monitoring Damage PropagationofofGFRP GFRP Composite Composite under The stress-strain curves of GFRP composite coupons with and without 2.0 wt % CNFs under staticThe tension are presented in Figure 10. It can becoupons observed thatand the GFRP coupons show stress-strain curvesof ofGFRP GFRPcomposite composite with without 2.0 2.0 wt % underunder The stress-strain curves coupons with and without wtCNFs %nonlinear CNFs static tension are presented in Figure 10. It can be observed that the GFRP coupons show nonlinear behavior for both composites with and without CNFs. static tension are presented in Figure 10. It can be observed that the GFRP coupons show nonlinear behavior for both composites withand andwithout without CNFs. CNFs. behavior for both composites with

Figure 10. Typical stress-strain curves of neat glass fiber composite and GFRP incorporating 2.0% CNFs. Right-hand-side photoscurves show test and corresponding values on incorporating stress-strain. Figure 10. Typical stress-strain curves of instances neat glass fiber composite and GFRP 2.0% 2.0% Figure 10. Typical stress-strain of neat glass fiber composite and GFRP incorporating CNFs. Right-hand-side photos show test instances and corresponding values on stress-strain.

CNFs. Right-hand-side photos show test instances and corresponding values on stress-strain. The source of nonlinearity is in the off-axis direction; the polymer matrix dominates the tension The source nonlinearity in the off-axis matrix CNFs dominates theeffect tension behavior rather of than the fibers is[32]. It was alsodirection; observedthe thatpolymer incorporating had no on The source of nonlinearity is[32]. in the off-axis direction; theincorporating polymer matrix the tension behavior rather than the fibers It was also observed CNFs had no effect the initial elastic modulus. In the off-axis direction, the that composite behavior isdominates dominated by on the the initial modulus. In theItoff-axis direction, thethat composite behavior is dominated by The theon the matrix. Atelastic low stress (region 1),also the contribution of incorporating glass fiber shallCNFs not behad neglected. behavior rather thanapplied the fibers [32]. was observed no effect matrix. lowstiffness applied stress (region 1), the contribution of of glass fiber shall not be The relative high glass fiberdirection, counteracts the effect the CNFs. When theneglected. applied stress initial elasticAt modulus. In of the off-axis the composite behavior is dominated by the matrix. relative high stiffness of glass fiber counteracts the effect of the CNFs. When the applied stress increases (region 2), the effect of CNFs on the composite behavior starts to appear through At low applied stress (region 1), the contribution of glass fiber shall not be neglected. The relative high interlaminar debonding place as of aonthe result of reduced fiber-matrix bond due to through reduced increases (region 2), the taking effectthe of CNFs theCNFs. composite starts to appear stiffness of glass fiber counteracts effect Whenbehavior the applied stress increases (region 2), interlaminar debonding taking place as a result of reduced fiber-matrix bond due to reduced

the effect of CNFs on the composite behavior starts to appear through interlaminar debonding taking

Nanomaterials 2016, 6, 169

10 of 12

place as a result of reduced fiber-matrix bond due to reduced crosslinking. This is reflected in region 2016, 6, 169stiffness of CNFs/GFRP coupons compared with GFRP composite10 of 13 neat 2 andNanomaterials in the decreased with epoxy. At high applied stress levels (region 3), another effect of CNFs becomes apparent. The reduced crosslinking. This is reflected in region 2 and in the decreased stiffness of CNFs/GFRP coupons polymer crosslinking results in softer matrix compared with neat polymer matrix. Such softening compared with GFRP composite with neat epoxy. At high applied stress levels (region 3), another in the polymer matrix limits its ability to restrain lateral fiber movement under tension loads, thus effect of CNFs becomes apparent. The reduced polymer crosslinking results in softer matrix an apparent necking like behavior takes place. This in behavior is very pronounced withtoCNFs/GFRP compared with neat polymer matrix. Such softening the polymer matrix limits its ability restrain coupons compared with GFRP coupons with neat epoxy. Such necking results in reduced cross-section lateral fiber movement under tension loads, thus an apparent necking like behavior takes place. This and premature (region 4)with at relatively lower elongation in CNFs/GFRP compared behavior is failure very pronounced CNFs/GFRP coupons compared with GFRP coupons withwith neat neat GFRP.epoxy. The right side images in Figure 10 show the the off-axis tensile test Such hand necking results in reduced cross-section and deformed premature shape failure of (region 4) at relatively for both CNFs/GFRP neat/GFRP at allwith fourneat regions. neckhand like side feature in images lower elongation inand CNFs/GFRP compared GFRP. The The right images in Figure210and 3 showthe thesignificant deformed shape of the off-axis tensile test for bothinCNFs/GFRP andcoupons neat/GFRP at all four with confirms necking like behavior taking place CNFs/GFRP compared The neck feature in images 2 and 3 confirms the significant necking like behavior taking GFRPregions. composite withlike neat epoxy. place in11 CNFs/GFRP coupons compared with GFRPcomposite composite with neat epoxy. Figure shows damage propagation in GFRP coupons incorporating 2.0 wt % CNFs Figure 11 shows damage propagation in GFRP composite coupons incorporating 2.0observed wt % CNFsusing under static tension. The figure shows a comparison between the damage metric under static tension. The figure shows a comparison between the damage metric observed using electrical conductivity monitoring and that quantified from the mechanical test using Equation (3). electrical conductivity monitoring and that quantified from the mechanical test using Equation (3). It It is apparent that incorporating CNFs enables monitoring damage initiation and propagation with is apparent that incorporating CNFs enables monitoring damage initiation and propagation with reasonable accuracy. Furthermore, gradually and reached a relatively reasonable accuracy. Furthermore,both bothmetrics metrics increased increased gradually and reached a relatively flat flat plateau showing constant damage in GFRP at the peak stress. However, it can also be observed plateau showing constant damage in GFRP at the peak stress. However, it can also be observed that that the mechanical damage value was always theelectrical electrical damage value for same the same the mechanical damage value was alwayshigher higher than the damage value for the stress.stress. The mechanical damage increase rate (damage propagation) is much higher for mechanical the mechanical The mechanical damage increase rate (damage propagation) is much higher for the damage damage metric compared with the electrical damage metric. For instance, at 40 MPa (representing metric compared with the electrical damage metric. For instance, at 40 MPa (representing about 40% of about 40% ofof off-axis strength of GFRP), a significant level ofdamage mechanical damage metric (about 30%) in off-axis strength GFRP), a significant level of mechanical metric (about 30%) is observed is observed in the GFRP coupon compared with a very limited damage (about 5%) observed by the the GFRP coupon compared with a very limited damage (about 5%) observed by the electrical damage electrical damage metric. The difference between the mechanical and electrical damage metrics can metric. The difference between the mechanical and electrical damage metrics can be explained by the be explained by the difference in significance of microcracks and microcrack propagation on elastic difference in significance of microcracks and microcrack propagation on elastic modulus and electrical modulus and electrical conductivity. While the elastic modulus is known to be significantly affected conductivity. While the electrical elastic modulus is known to be significantly affected by by cracking by cracking [33], the conductivity might not be influenced at the same rate cracking[33], as the electrical might notconductive be influenced the bymatrix. cracking long that as alternative long conductivity as alternative electrically pathsat can besame foundrate in the Thisasmeans using electrically conductive pathsofcan be found in theinitiation matrix. and Thispropagation means thatinusing will provide CNFs will provide means monitoring damage GFRP CNFs but it might not be as sensitive to damage occurrence and propagation as mechanical damage. In essence, calibration means of monitoring damage initiation and propagation in GFRP but it might not be as sensitive to of the two metrics be necessary if electricaldamage. conductivity is to becalibration used to monitor damage occurrence andmight propagation as mechanical In essence, of the damage two metrics propagation in GFRP incorporating CNFs. Nevertheless, it is important to note that such damage might be necessary if electrical conductivity is to be used to monitor damage propagation in GFRP initiation and propagation monitoring is not possible in neat GFRP composites due to being nonincorporating CNFs. Nevertheless, it is important to note that such damage initiation and propagation conductive. monitoring is not possible in neat GFRP composites due to being non-conductive.

Figure 11. Stress-electrical damage (DE) and stress-mechanical damage (DM) for glass fiber composite

Figure 11. Stress-electrical damage (DE ) and stress-mechanical damage (DM ) for glass fiber composite incorporating CNFs under monotonically increasing static tension stress. incorporating CNFs under monotonically increasing static tension stress.

Nanomaterials 2016, 6, 169

11 of 12

4. Conclusions The mechanical and electrical properties of epoxy incorporating CNFs were examined. The results of the electrical measurements show that CNFs can significantly improve the electrical conductivity of epoxy. It was also observed that a significant improvement in the electrical conductivity of CNFs/epoxy nanocomposite is achieved with a percolation threshold equal to 1.5 wt %. The electromechanical behavior of GFRP incorporating 2.0 wt % CNFs was then examined under monotonically increased static loading. Electromechanical measurements of GFRP coupons under off-axis static tension tests showed that electrical damage based on the change in electrical conductivity of GFRP can be correlated well with growing damage in the GFRP coupons under static tension loads. The measurements also showed that damage propagation monitoring in CNFs/GFRP using electrical conductivity has a lower sensitivity than mechanical damage propagation. This might be attributed to the pronounced effect of microcracking on mechanical behavior compared with its influence on electrical conductivity. Nevertheless, it is evident that using CNFs during fabrication of GFRP composites allows damage initiation and damage propagation in GFRP to be monitored with acceptable repeatability and resolution. Acknowledgments: This research was funded by the National Science Foundation (NSF) Award # OISE-1103601 and by Air Force Office of Scientific Research (AFOSR) Award # FA9550-14-1-0021 to the University of New Mexico. The work is also funded by the Science and Technology Development Fund (STDF) for US-Egypt (STDF/NSF) Program; Award # 3713 and (STDF-CSE) Program (ID 5213) to Egyptian Petroleum Research Institute (EPRI), Cairo, Egypt. The authors greatly acknowledge this funding. Special thanks to Professor Sanjay Krishna and Brianna Klein at UNM Center for High Technology Materials (CHTM) for their help with the source meter unit. Author Contributions: Eman Taha fabricated and tested all specimens. Eman Taha and Usama Kandil conducted the microstructural investigations. Eman Taha, Usama Kandil and Mahmoud Reda Taha participated in data analysis and interpretation of the test results. Usama Kandil conducted analysis of microstructural data with FTIR. All five authors participated in writing the manuscript. Mahmoud Reda Taha led the team’s efforts. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3.

4. 5. 6. 7.

8. 9. 10. 11.

Rivero, P.J.; Urrutia, A.; Goicoechea, J.; Arregui, F.J. Nanomaterials for functional textiles and fibers. Nanoscale Res Lett. 2015, 10, 501. [CrossRef] [PubMed] Govaert, F.; Vanneste, M. Preparation and application of conductive textile coatings filled with honeycomb structured carbon nanotubes. J. Nanomater. 2014, 2014, 1–6. [CrossRef] Wichmann, M.H.G.; Sumfleth, J.; Gojny, F.H.; Quaresimin, M.; Fiedler, B.; Schulte, K. Glass-fibre-reinforced composites with enhanced mechanical and electrical properties—Benefits and limitations of a nanoparticle modified matrix. Eng. Fract. Mech. 2006, 73, 2346–2359. [CrossRef] Bortz, D.R.; Merino, C.; Martin-Gullon, I. Carbon nanofibers enhance the fracture toughness and fatigue performance of a structural epoxy system. Compos. Sci. Technol. 2011, 71, 31–38. [CrossRef] Morales, G.; Barrena, M.I.; Gómez de Salazar, J.M.; Merino, C.; Rodríguez, D. Conductive CNF-reinforced hybrid composites by injection moulding. Compos. Struct. 2010, 92, 1416–1422. [CrossRef] Soliman, E.; Al-Haik, M.; Taha, M.R. On and off-axis tension behavior of fiber reinforced polymer composites incorporating multi-walled carbon nanotubes. J. Compos. Mater. 2011, 46, 1661–1675. [CrossRef] Hossain, M.K.; Hossain, M.E.; Dewan, M.W.; Hosur, M.; Jeelani, S. Effects of carbon nanofibers (CNFs) on thermal and interlaminar shear responses of E-glass/polyester composites. Compos. B Eng. 2013, 44, 313–320. [CrossRef] Li, J.; Wua, Z.; Huang, C.; Li, L. Multiscale carbon nanotube-woven glass fiber reinforced cyanate ester/epoxy composites for enhanced mechanical and thermal properties. Compos. Sci. Technol. 2014, 104, 81–88. [CrossRef] Wang, S.; Qiu, J. Enhancing thermal conductivity of glass fiber/polymer composites through carbon nanotubes incorporation. Compos. B Eng. 2009, 69, 1486–1498. [CrossRef] Al-Saleh, M.H.; Sundararaj, U. A review of vapor grown carbon nanofiber/polymer conductive composites. Carbon 2009, 47, 2–22. [CrossRef] Bauhofer, W.; Kovacs, J.Z. A review and analysis of electrical percolation in carbon nanotube polymer composites. Compos. Sci. Technol. 2009, 69, 1486–1498. [CrossRef]

Nanomaterials 2016, 6, 169

12.

13. 14. 15. 16. 17.

18. 19.

20. 21. 22. 23. 24. 25. 26.

27.

28. 29.

30. 31. 32. 33.

12 of 12

Yang, Y.; Gupta, M.C.; Dudley, K.L.; Lawrence, R.W. A comparative study of EMI shielding properties of carbon nanofiber and multi-walled carbon nanotube filled polymer composites. J. Nanosci. Nanotechnol. 2005, 5, 927–931. [CrossRef] [PubMed] Yang, S.; Lozano, K.; Lomeli, A.; Foltz, H.D.; Jones, R. Electromagnetic interference shielding effectiveness of carbon nanofiber/LCP composites. Compos. A Appl. Sci. 2005, 36, 691–697. [CrossRef] Lozano, K.; Bonilla-Rios, J.; Barrera, E.V. A study on nanofiber-reinforced thermoplastic composites (II): Investigation of the mixing rheology and conduction properties. J. Appl. Polym. Sci. 2001, 80, 1162–1172. [CrossRef] Finegan, I.C.; Tibbetts, G.G. Electrical conductivity of vapor-grown carbon fiber/thermoplastic composites. J. Mater. Res. 2001, 16, 1668–1674. [CrossRef] Gordeyev, S.A.; Macedo, F.J.; Ferreira, J.A.; Van-Hattum, F.W.J.; Bernardo, C.A. Transport properties of polymer-vapour grown carbon fibre composites. Physica B 2000, 279, 33–36. [CrossRef] Park, J.; Jang, J.; Wang, Z.; Kwona, D.; DeVries, K.L. Self-sensing of carbon fiber/carbon nanofiber-epoxy composites with two different nanofiber aspect ratios investigated by electrical resistance and wettability measurements. Compos. A Appl. Sci. 2010, 41, 1702–1711. [CrossRef] Wen, S.; Chung, D.D.L. Electrical-resistance-based damage self-sensing in carbon fiber reinforced cement. Carbon 2007, 45, 710–716. [CrossRef] Park, J.; Kim, D.; Kim, S.; Kim, P.; Kwona, D.; DeVries, K.L. Inherent sensing and interfacial evaluation of carbon nanofiber and nanotube/epoxy composites using electrical resistance measurement and micromechanical technique. Compos. B Eng. 2007, 38, 847–861. [CrossRef] Standard Test Methods for Constituent Content of Composite Materials; ASTM D 3171-99; ASTM International Standards: West Conshohocken, PA, USA, 1999. Standard Test Methods for DC Resistance or Conductance of Insulating Materials; ASTM D257-99; ASTM International Standards: West Conshohocken, PA, USA, 1999. Standard Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials; ASTM D3039/D3039 M-14; ASTM International Standards: West Conshohocken, PA, USA, 2014. Guth, E. Theory of filler reinforcement. J. Appl. Phys. 1945, 16, 20–25. [CrossRef] Manera, M. Elastic properties of randomly oriented short fiber-glass composites. J. Compos. Mater. 1977, 11, 235–247. [CrossRef] Christensen, R.M. Asymptotic modulus results for composites containing randomly oriented fibers. Int. J. Solids Struct. 1976, 12, 537–544. [CrossRef] Golru, S.; Attar, M.M.; Ramezanzadeh, B. Studying the influence of nano-Al2 O3 particles on the corrosion performance and hydrolytic degradation resistance of an epoxy/polyamide coating on AA-1050S. Prog. Org. Coat. 2014, 77, 1391–1399. [CrossRef] Mikhaylova, Y.; Adam, G.; Haussler, L.; Eichhorn, K.J.; Voit, B. Temperature-dependent FTIR spectroscopic and thermoanalytic studies of hydrogen bonding of hydroxyl (phenolic group) terminated hyperbranched aromatic polyesters. J. Mol. Struct. 2006, 788, 80–88. [CrossRef] Nielsen, L.E. Cross-linking: effect on physical properties of polymers. J. Macromol. Sci. C Polym. Rev. 1969, 3, 69–103. [CrossRef] Zeng, X.; Yu, S.; Lai, M.; Sun, R.; Wong, C. Tuning the mechanical properties of glass fiber-reinforced bismaleimide–triazine resin composites by constructing a flexible bridge at the interface. Sci. Technol. Adv. Mater. 2013, 14, 065001. [CrossRef] Liang, Y.L.; Pearson, R.A. Toughening mechanisms in epoxy–silica nanocomposites (ESNs). Polym. J. 2009, 50, 4895–4905. [CrossRef] Zhang, M.Q.; Xu, J.R.; Zeng, H.M.; Huo, Q.; Zhang, Z.Y.; Yun, F.C. Fractal approach to the critical filler volume fraction of an electrically conductive polymer composite. J. Mater. Sci. 1995, 30, 4226–4232. [CrossRef] Soliman, E.; Kandil, U.; Taha, M.R. Improved strength and toughness of carbon woven fabric composites with functionalized MWCNTs. Materials 2014, 7, 4640–4657. [CrossRef] Dowling, N. Mechanical Behavior of Materials: Engineering Methods for Deformation, Fracture and Fatigue, 4th ed.; Pearsons: New York, NY, USA, 2012. © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).