Effect of different carbon nano-fillers on rheological properties and lap shear strength of epoxy adhesive joints

Composites: Part A 82 (2016) 53–64

Contents lists available at ScienceDirect

Composites: Part A journal homepage: www.elsevier.com/locate/compositesa

Effect of different carbon nano-fillers on rheological properties and lap shear strength of epoxy adhesive joints Panta Jojibabu a, M. Jagannatham a, Prathap Haridoss a, G.D. Janaki Ram a, Abhijit P. Deshpande b,⇑, Srinivasa Rao Bakshi a,⇑ a b

Department of Metallurgical and Materials Engineering, Indian Institute of Technology Madras, Chennai 600036, India Department of Chemical Engineering, Indian Institute of Technology Madras, Chennai 600036, India

a r t i c l e

i n f o

Article history: Received 1 July 2015 Received in revised form 6 December 2015 Accepted 11 December 2015 Available online 17 December 2015 Keywords: A. Particle-reinforcement B. Adhesion D. Mechanical testing E. Joints/joining

a b s t r a c t In this work, the rheological properties, thermal stability and the lap shear strength of epoxy adhesive joints reinforced with different carbon nano-fillers such as multi-walled carbon nanotubes (CNT), graphene nanoplatelets (GNP) and single-walled carbon nanohorns (CNH) have been studied. The nanofillers were dispersed homogeneously using BrabenderÒ Plasti-CorderÒ. The epoxy pre-polymer with and without the nano-fillers exhibited shear thinning behavior. The nano-filler epoxy mixtures exhibited a viscoplastic behavior which was analyzed using Casson’s model. Thermo-gravimetric analysis indicated an increase in the thermal stability of the epoxy with the addition of carbon nano-fillers. Carbon nanofillers resulted in increased lap shear strength having high Weibull modulus. The joint strength increased by 53%, 49% and 46% with the addition of 1 wt.% CNT, 0.5 wt.% GNP and 0.5 wt.% CNH, respectively. The strength of the joints having high filler content (>1 wt.%) was limited by mixed mode type of failure. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Adhesive bonding is an easy and economically alternative method applicable for joining a variety of materials. Adhesive bonding offers unique advantages such as uniform stress distribution throughout the joint and good strength to weight ratio [1,2]. Epoxy-based adhesives find vital applications in automotive, aeronautics, electronics and packaging industries due to their excellent mechanical properties [1,3]. They have outstanding adhesion to various substrates such as steel, aluminum and carbon fiber reinforced plastics [4,5]. However, the strength of the epoxy adhesive joints is low compared to bolted, riveted and welded joints and is limited by strength of the epoxy. Efforts have been made to improve the strength of epoxy joints by adding nano-fillers such as carbon nanotubes (CNT) [2,6–10], nano-SiO2 [11,12], nanoAl2O3 [13–15], carbon black [16] and nano-CaCO3 [13] with varying degrees of success. Since 1990s, carbon based nanomaterials have attracted lot of attention due to their extraordinary mechanical, electrical and thermal properties. The large ratio of the strength and elastic modulus of the carbon nanomaterials to that of ⇑ Correspondingc authors. Tel.: +91 4422574169 (A.P. Deshpande), +91 4422574781 (S.R. Bakshi). E-mail addresses: [email protected] (A.P. Deshpande), [email protected] (S.R. Bakshi). http://dx.doi.org/10.1016/j.compositesa.2015.12.003 1359-835X/Ó 2015 Elsevier Ltd. All rights reserved.

polymers makes them an excellent reinforcement [17]. Carbon nanotubes have been shown to significantly improve the mechanical, electrical, and thermal properties of polymer composites [18–20]. Recently, two-dimensional (2D) nanostructures such as graphene nano-platelets (GNP), graphene, and graphene oxide have emerged as favorable fillers for polymer matrices [21–23]. GNP is promising reinforcement for polymer composites because of its high aspect ratio (length to thickness ratio), unique graphitized planar structure, and low manufacturing cost. Furthermore, the high surface area of GNP results in large contact area with polymer matrix, resulting in good load transfer and improvement in mechanical properties. Single-walled carbon nanohorns (CNH) are recently reported novel materials having a typical diameter of 2 nm and a length of 30–50 nm which have not yet been commercialized. The individual CNH have a tendency to couple together and form stable dahlia flower like particles with a narrow diameter distribution of 80–100 nm [24]. The strengthening derived from nanofillers depends on their intrinsic strength as well as their shape (aspect ratio). The mechanical behavior of polymer matrix composites also depends strongly on the interface between the filler and the matrix which influences the transfer of the mechanical load and hence is dependent on the shape of the particles. It is therefore expected that CNT, GNP and CNH may not give the same properties in composites.

54

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Although many studies on CNT and graphene/GNP reinforced epoxy composites are already reported [19,25–36], very limited literature is available on adhesive bonding using CNT–epoxy [2,6–10] and GNP–epoxy [37,38] composites and there is no work on CNH–epoxy composites/adhesive joints. Yu et al. have studied the strength and durability of up to 5 wt.% CNT reinforced epoxy adhesive joints using the Boeing Wedge test under water at 60 °C [2]. They observed that both the bond strength and durability increased with the addition of CNT. The 1 wt.% CNT reinforced joint showed very high fracture toughness (7.4
106 J m2) compared to pure epoxy joint (1983 J m2) [2]. The durability of CNT/ epoxy adhesive was also found to be better compared to pure epoxy joints. Gerson et al. [7] have studied the effect of CNT on the curing kinetics of epoxy adhesives and found that CNT increased the cross-linking density and glass transition temperature of epoxy (Tg). Srivastava has used 3 wt.% CNT reinforced epoxy to join carbon/carbon (C/C) and carbon/carbon–silicon carbide (C/C–SiC) composites [6]. The C/C–C/C and C/C–SiC–C/C–SiC joints exhibited a 25% and 41% increase in joint strength respectively with CNT reinforcement, compared to pure epoxy joint. To prevent the agglomeration of CNT, Wolf et al. [9] attached a thermally stable protein SP1 to CNT and observed that the CNT/SP1 reinforced adhesive joints showed 50% higher peel strength and 25% higher shear strength over neat epoxy joints. Sydlik et al. have modified CNT using multiple covalent functionalization methods with different zwitterionic surfactants and studied its effect on the shear strength of the adhesive joints [8]. It was observed that the lap shear strength of 1 wt.% functionalized CNT–epoxy joints was 36% and 27% higher than that of neat epoxy joint and non-functionalized CNT–epoxy joint, respectively. Zhang et al. [10] have used silane treated CNT to reinforce epoxy joints of carbon/carbon composites and observed that the silanized CNT were dispersed uniformly and had a good interface strength. The C/C composite joints with 0.2 wt.% silanized CNT had an average shear strength of 10.4 MPa, which was 31% higher than that of the neat epoxy joint. These results indicate that functionalization of nano-fillers can further improve the strength of the joints. But it is to be noted that functionalized CNT are expensive. There have been a few reports recently on use of GNP as reinforcement. Soltannia and Taheri have investigated the effect of different carbon fillers such as CNT, GNP and carbon fibers (CNF) in epoxy adhesive [37]. It was observed that the GNP/epoxy adhesive joints showed increased strength among the three carbon fillers. It was also shown that higher strain rate testing resulted in higher adhesive strength [37]. Guadagno et al. have synthesized tensile butt joints and measured the joint strength of up to 4 wt.% GNP reinforced epoxy adhesive. They found that the 1 wt.% GNP reinforced epoxy had nearly two times strength compared to pure epoxy joint. It was observed that higher addition of GNP lead to poor dispersion leading to decrease in the joint strength [38]. It is seen from literature that the maximum strengthening is observed at an intermediate level of nano-filler addition, which is dependent on the type of nano-filler. It is to be noted that in an adhesive joint, both the cohesive strength of the adhesive and the strength of adhesion with the substrate determine the overall strength. Dispersion of carbon nano-fillers is known to be a challenge due to their large specific surface areas and high viscosity of the epoxy resin. Poor dispersion of nano-fillers is known to result in ineffective strengthening and may also affect adhesion with the substrate and cause premature failure. Several dispersion techniques such as ultrasonication (bath and probe type), shear mixing, calendering and combination of these techniques have been used to disperse nano-fillers uniformly in polymer matrices with different degrees of success [39]. So, there is a need for novel processes which can result in improved dispersion, especially for high viscosity materials such as epoxy resin.

The objective of present work was to carry out a systematic study of the effect of different carbon nano-fillers, namely CNT, GNP and CNH, on the strength of epoxy adhesive joints. One of the prime novelty of this study is use of BrabenderÒ Plasti-CorderÒ, which is high energy shear mixer generally employed for blending of thermoplastic materials under controlled temperature conditions, to mix the epoxy and nano-fillers. This is the first study employing it for dispersing carbon nano-fillers in epoxy as per the author’s knowledge. It is also a first study on CNH as reinforcement. The effect of the type and nano-filler content on the rheological properties, thermal stability and the strength of the epoxy joints is presented and discussed. 2. Experimental 2.1. Materials A di-glycidyl ether of bisphenol-A (DGEBA) based two part epoxy adhesive (EP 415, Rotex polymers, Chennai, India) was used in this study. CNT of diameter 10–20 nm and length 10–30 lm were purchased from Cheaptubes Inc. (Texas, USA). GNP having thickness between 3 and 10 nm were obtained from Redex Nano (Ghaziabad, India). The CNH were synthesized in-house by using a DC arc discharge technique. In this method, pure graphite rod of 11 mm diameter and 260 mm length was used as anode. The cathode was a rotating graphite disc of 300 mm diameter. The discharge was carried out in a water cooled stainless steel chamber in helium gas static atmosphere to reduce contamination. The experimental conditions used for synthesis of CNH were 150 A current at 32 V under the pressure of 500 Torr. With an electrode separation of 1 mm, the arc was produced which consumed the anode gradually. The CNH were collected from the inner and upper wall of the reaction chamber. The epoxy mixture and joints containing CNT, GNP and CNH are named as EP–CNT, EP–GNP and EP–CNH respectively. 2.2. Preparation of carbon nano-filler epoxy mixtures The epoxy resin and carbon nano-fillers were taken in required quantities to prepare 0.2, 0.5, 1, and 2 wt.% nano-filler/epoxy mixture. The constituents were charged into the BrabenderÒ PlastiCorderÒ and mixed at 100 rpm for 15 min. The hardener was then added in the requisite weight ratio (100:40 by weight) and mixed manually for 10 min using a Teflon rod. This mixture was used for preparing lap shear joints of AA6061 sheets as described later. 2.3. Rheological studies Rheological measurements were carried out for carbon nano-filler epoxy mixtures obtained from BrabenderÒ Plasti-CorderÒ having different filler contents. A stress-controlled rheometer (ANTON PAAR, Physica 301, Germany) having a 25 mm parallel plate geometry and operating in steady mode was used to measure the viscosity. The measurements were carried out with a gap size of 1 mm at 25 °C over a shear rate of 0.1–100 s1. 2.4. Thermal stability studies The thermal stability of the cured adhesives was studied using a thermo-gravimetric analyzer (TGA, SDT Q600). The samples were heated to 800 °C at a heating rate of 10 °C/min under nitrogen atmosphere. The weight change as a function of temperature was analyzed and the onset temperatures for degradation were compared.

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

2.5. Preparation and testing of single lap shear adhesive joints AA6061 substrates were grit blasted using a grit blasting machine (Sandstorm, SB-9090, Bangalore, India) with F400 aluminum oxide grit under a pressure of 4–5 bar to obtain a rough surface. After grit blasting, the substrates were ultrasonically cleaned for 15 min in acetone followed by drying. The lap shear joints were prepared as per the ASTM D1002 standard with 25.4 mm
10 mm bond area as shown in Fig. 1. A bond line thickness of approximately 0.2 mm was maintained. After preparation of the joints, samples were cured at 60 °C for 2 h. Lap shear tests were conducted using tensile testing machine (Instron 3367, 30 kN). The specimens were tested at 1.3 mm/min crosshead displacement as per the ASTM D1002 standard. The distribution of shear strength of the adhesive joints was studied qualitatively by using Weibull distribution analysis. For measuring the bonding strength, at least 5 samples were tested for each combination. 2.6. Microstructural characterization Morphology of the carbon nano-fillers was observed using transmission electron microscope (Philips CM12, Holland) operated at a voltage of 120 kV and high resolution scanning electron microscope (HITACHI S4800, Japan). Fracture surface of the lap shear joints were analyzed for studying the dispersion of the carbon nano-fillers in epoxy matrix using SEM. The fractured adhesive samples were sputter coated with gold before the SEM observations. Inferences were drawn by observing more than 10 images of the fracture surface corresponding to de-bonded areas and cohesive failure areas. 3. Results and discussions 3.1. Morphology of the carbon nano-fillers Fig. 2 shows the TEM and SEM (inset) images of the CNT, GNP and CNH used in the study. Entangled CNT and stacks of GNP were observed as seen in Fig. 2a and b, respectively. The SEM image in Fig. 2c shows spherical shapes of CNH aggregates. CNH have dahlia flower structure as seen in Fig. 2d. These images of CNH are similar to that reported by Iijima et al. [40] 3.2. Rheological properties of epoxy nano-filler mixture The rheological behavior of the nano-filler/epoxy mixture is of practical significance and must be understood thoroughly. Fig. 3 shows the variation of the viscosity of carbon nano-filler/epoxy mixture before the addition of hardener at different weight fractions. We see from Fig. 3 that the behavior is clearly non-linear in the log scale which indicates that the Power Law model g ¼ mðc_ Þn1 does not apply. It is evident that all the samples including the neat epoxy exhibit a shear thinning behavior with the viscosity decreasing with increase in strain rate. Kim et al. have

Fig. 1. Schematic showing the ASTM D1002 standard for lap shear joint test. The thickness of the sample is 1 mm and the adhesive layer is about 0.2 mm thick.

55

also carried out similar studies with epoxy–CNT mixtures [41]. They observed shear thinning behavior with significant increase in the viscosity of epoxy with the increase in CNT content. The present results show similar behavior for both CNT and CNH while only a small increase in viscosity is observed with GNP addition. As it is apparent from Fig. 3b, the addition of GNP to epoxy does not alter the viscosity much for all weight fractions but it exhibits the shear thinning behavior. The small effect of GNP on viscosity could be due to the following two reasons. The density of the CNT and CNH are lower than GNP, and hence their volume fraction is greater than GNP for the same weight fraction. So the changes with GNP addition may not be that dramatic. It is also reported that the GNP might align in the flow field easily on the application of shear due to their plate like structure, and hence the viscosity is not affected much [41]. Fig. 3d represents the shear stress of 1 wt.% nano-filler/epoxy mixture as a function of shear rate. From the inset, the nano-filler added epoxy show a viscoplastic behavior with some initial yield stress. Below this yield stress, the material does not flow and behaves like solid. The nanoparticles form a network structure which is firm at low strain rates leading to infinite viscosity (no flow). When the stress exceeds the yield stress, the network gets disrupted and shearing is observed. Casson has proposed an empirical model to represent this viscoplastic behavior given by following equation [42]. 1

r2 ¼ r20 þ ðg1 c_ Þ2 1

1

ð1Þ

where r is the shear stress, r0 is the yield stress of the material, g1 final viscosity and c_ is the strain rate. Several studies have used Casson’s model to determine the yield stress of viscoplastic polymers [43,44]. The values of r0 and g1 were calculated using a linear fit 1 1 of r2 and c_ 2 for different materials as shown in Fig. 4. The variation of the yield stress as a function of the amount of nano-filler is shown in Fig. 4d. For EP–CNT, the yield stress was found to be 137 Pa for 2 wt.% of CNT, which might be due to the formation of interconnected CNT with the molecular chains in epoxy. The highest yield stress was observed for EP–CNH equal to 190 Pa. This might due to the larger volume fraction and the interconnected nature of CNH in the epoxy matrix along with its small size, resulting in larger interaction with the molecular chains in the epoxy. For EP–GNP, the increase was much lower at 63 Pa for 2 wt.% GNP addition. This might be due to the fact that the GNP donot form network structure easily due to their flat morphology as well as lower volume fraction compared to other carbon nanofillers. From these results it could be said that high shear rates will be helpful in actual applications due to the viscoplastic nature and the shear thinning behavior. 3.3. Thermal stability The weight loss vs. temperature curves for cured samples is presented in Fig. 5. It is observed that the first stage of degradation was observed at temperatures below 200 °C, for both neat and nano-filler/epoxy composite. This is due to the decomposition of low molecular weight volatile compounds such as 1-propanol in the epoxy adhesive [45]. The highest weight loss occurred in second stage of degradation which is related to the higher molecular weight compounds in the epoxy adhesive [2] produced after curing. The onset of 2nd stage of decomposition temperatures are given in Table 1. The maximum increment in the decomposition temperature was 11.5 °C for 1 wt.% CNT, 9.5 °C for 0.5 wt.% GNP, and 7.7 °C for 0.5 wt.% CNH. At higher fractions of carbon nano-fillers, thermal stability of reinforced epoxies was observed to reduce, although still being better than the neat epoxy. The adverse

56

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Fig. 2. TEM and SEM (inset) images of (a) CNT, (b) GNP, (c) SEM image of CNH and (d) TEM image of CNH.

Fig. 3. Steady state viscosity measurements of (a) EP–CNT, (b) EP–GNP, (c) EP–CNH joints and (d) shows plots of shear stress vs. shear rate of 1 wt.% carbon nano-filler epoxy mixture showing the viscoplastic behavior.

effect of the poor dispersion at higher filler fractions is probably responsible for this effect. As mentioned earlier, there are several reports on the thermal stability of CNT/epoxy [2] and graphene oxide (GO)/epoxy composites [46] but limited literature is available on CNH/epoxy composites. The present observations of improved thermal stability of nano-filler/epoxy composites are consistent with the earlier results.

3.4. Lap shear strength of adhesive bonded joints The lap shear test results of neat as well as nano-filler reinforced epoxies are shown in Fig. 6 and show that there is a significant difference in the strength for neat and nano-filler/epoxy adhesive joints. The strength reported here is an average of five specimens with the error bars corresponding to standard deviation. As the amount of carbon nano-filler in the adhesive increases,

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Fig. 4. Plots showing linear behavior of with the filler content.

57

r1=2 vs. c_ 1=2 as per Casson’s model for (a) EP–CNT, (b) EP–GNP, (c) EP–CNH mixtures and (d) shows the variation of the yield stress

Fig. 5. TGA curves of epoxy adhesives reinforced with (a) CNT, (b) GNP and (c) CNH.

58

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Table 1 Onset temperatures for 2nd stage decomposition of neat and nano-filler reinforced epoxy. Sample

EP–CNT EP–GNP EP–CNH

Onset temperature for 2nd stage decomposition (°C) 0 (wt.%)

0.2 (wt.%)

0.5 (wt.%)

1 (wt.%)

2 (wt.%)

325.1 325.1 325.1

326.9 328.4 328.9

331.5 334.6 332.8

336.6 330.7 331.2

331.2 329.1 328.7

MansourianTabaei et al. have dispersed CNT up to 3 wt.% in epoxy matrix using ultrasonication followed by mechanical stirring [45] and prepared lap-shear joints as per the ASTM D1002 with strain rate of 1.27 mm/min. The strength was found to increase by 50% for 3 wt.% CNT reinforced epoxy adhesive joints. Wolf et al. have used both shear mixer and 3-roll mill to disperse CNT in epoxy matrix up to 0.7 wt.% [9]. An increase in the lap shear strength of 16.2% was observed for 0.35 wt.% CNT dispersed using shear mixer while 25% increment was observed for 0.7% CNT dispersed using 3-roll mill. In this study, 49% of shear strength increment was observed with 1 wt.% of CNT addition by using BrabenderÒ Plasti-CorderÒ which is a significant achievement for such low reinforcement percentage. This is due to the homogeneous dispersion of nano-filler in the epoxy matrix using BrabenderÒ Plasti-CorderÒ. The strength achieved in the present study is a function of the intrinsic strengthening effect of different carbon nano-fillers as well as their effect on the adhesion between the epoxy and substrate which is discussed in Section 3.8. Fig. 6. Lap shear strengths of nano-carbon reinforced epoxy adhesive joints.

the joint strength increased and attained the highest value of 21.4 MPa for 1 wt.% EP–CNT, 20.7 MPa for 0.5 wt.% EP–GNP and 20.3 MPa for 0.5 wt.% EP–CNH joint, which corresponds to a 53%, 49% and 46% increase in the strength compared to the neat epoxy joint respectively. The high tensile strength and modulus of these carbon nano-fillers contributed to the enhanced shear strength of the adhesive joints. At lower fractions, carbon nano-fillers were dispersed uniformly and there was a strong interfacial bonding between the matrix and the fillers which enabled effective load transfer between them. At higher fractions of the carbon nano-fillers, the strengthening effect started to reduce due to the formation of agglomerates which are expected to have a weak interface with epoxy. From Fig. 6 the lap strength of the 2 wt.% EP–CNH joints was observed to be lower than that of the neat epoxy adhesive. This can be attributed to the severe agglomeration of CNH and the network structure of CNH which may not get infiltrated properly with epoxy. It was also noted that the mode of failure progressed from cohesive (within the adhesive) to a mixed mode of cohesive and adhesive (interface between adhesive and substrate) failure. Among the three nano-fillers, CNT were found to be the most effective at higher loading (1 wt.%) while GNP and CNH were better at relatively lower loadings of 0.5 wt.%. It is also observed that the joint strength for EP–CNH and EP–GNP reduce significantly when their content is increased from 0.5 to 1 wt.% as compared to EP–CNT. Sydlik et al. [8] have reported that the lap shear strength of 0.5 wt.% and 1 wt.% of non-functionalized CNT was increased 14% and 7% higher than the pure epoxy adhesive joints with mechanical mixing and 1 min of sonication. They also report that using 1 wt.% of functionalized CNTs increased the strength by 36%. Soltannia et al. showed the 25.6% increment by the addition of 0.5 wt.% GNP with calendaring dispersion technique. It is observed that the percentage increase in joint strength in this study is very high compared to reported literature indicating that BrabenderÒ Plasti-CorderÒ could disperse the nano-fillers effectively.

3.5. Weibull distribution analysis The distribution of the lap shear strength of adhesive joints was analyzed using a two parameter Weibull distribution given below [47]:

  m  ri Pðri Þ ¼ 1  exp 

r0

ð2Þ

Here the subscript i denotes the specimen number (as per increasing order of lap shear strength), P(ri) is the failure probability of the ith specimen, ri is the lap shear strength of ith specimen, r0 is a characteristic strength (determined from fitting), and m is the Weibull modulus. The Weibull modulus, which is also called the shape parameter, indicates the deviation in the measured strength. A small value of Weibull modulus indicates a large scatter in the joint strength values and vice versa. The failure probability P(ri) for stress of ri was calculated using the following equation:

Pðri Þ ¼

i  0:5 N

ð3Þ

Fig. 7 shows the Weibull distribution plots and the Weibull modulus for different nano-filler reinforced epoxy adhesive joints. It was observed that Weibull modulus values were different for different carbon nano-fillers at different weight fractions. Li et al. [48] have studied the Weibull distribution model of epoxy-carbon fiber adhesive bonds and obtained a Weibull modulus of 21.56. In their studies it was explained that the Weibull modulus of greater than 20 was larger enough to represent much reliable bonding strength of the specimens. In the present study, most of the adhesive joints are found to have a Weibull modulus more than 20 and it is believed that these joints have reliable bonding strength. The Weibull modulus seems to be higher for low weight percent filler concentration while with increasing concentration, the scatter was found to increase. This is in accordance with the switching of failure mode from cohesive to a mixed mode. The Weibull modulus of 0.2 wt.% CNT was found to be small which could be due to the fact that the weight percentage of CNT was below the percolation

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

59

Fig. 7. Weibull plots and Weibull modulus for the nano-carbon reinforced adhesive bonded joints.

Fig. 8. SEM images if fracture surface of (a) 0.2 EP–CNT, (b) 0.5 EP–CNT, (c) 1 EP–CNT and (d) 2 EP–CNT samples.

threshold. MansourianTabaei et al. [45] found two percolation thresholds of 1.125 wt.% and 3 wt.% in their study while measuring the lap shear strength of the CNT reinforced epoxy adhesive joints. In this present study, 0.2 wt.% CNT are in below the percolation threshold which could lead to different concentration at different places leading to a larger scatter and a lower Weibull modulus. More detailed experiments may be needed to support this.

3.6. Dispersion of carbon nano-fillers in the epoxy matrix In order to study the homogeneity of the carbon nano-fillers dispersion in the epoxy joint, the fractured surfaces were investigated by SEM. Figs. 8–10 show the representative fracture surfaces of EP–CNT, EP–GNP and EP–CNH joints, respectively. At lower weight fractions, nano-fillers were distributed uniformly in the

60

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Fig. 9. SEM images of fracture surface of (a) 0.2 EP–GNP, (b) 0.5 EP–GNP, (c) 1 EP–GNP and (d) 2 EP–GNP samples.

Fig. 10. SEM images of fracture surface of (a) 0.2 EP–CNH, (b) 0.5 EP–CNH, (c) 1 EP–CNH and (d) 2 EP–CNH samples.

epoxy matrix while at higher fractions (above 1 wt.%) they formed agglomerates. As shown in Fig. 8, CNT were dispersed homogeneously up to 1 wt.% without apparent agglomerates while agglomerates could be seen in the 2 wt.% samples. At higher magnifications, these agglomerates were observed to be partially infiltrated with the epoxy. For GNP and CNH, up to 0.5 wt.%, uniform dispersion was observed and agglomerates of GNP and CNH were observed with further addition as shown in Figs. 9 and 10.

It has been demonstrated in the present study that the lap shear strength of the adhesive joints increased with the incorporation of carbon nano-fillers. This increased shear strength is due to the increase in the strength of the epoxy, while maintaining the interfacial adhesion with the substrate unaffected. Jiang et al. have used a term strengthening efficiency (R) for quantifying the effect of carbon nano-fillers on the strength of the composites which is given in Eq. (4) [50]:

R¼ 3.7. Strengthening of carbon nano-fillers Several micromechanics models have been proposed for strengthening in fiber and particulate reinforced composites [49].

rc  rm V f rm

ð4Þ

Here Vf is the volume fraction of carbon nano-filler, rc is the strength of the composite (taken as lap shear strength of nano-filler reinforced epoxy joints in this case), rm is the strength of the matrix

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

(taken as lap shear strength of the neat epoxy adhesive joint). This expression gives the change in strengthening efficiency with increasing CNT which can be caused by change in dispersion. For the lap joints failing in cohesive mode, the strength of the joint is due to the intrinsic strength of the adhesive and hence Eq. (4) can be used to model the lap shear strength of the adhesive joints. It was found that only the 0.2 and 0.5 wt.% nano-filler reinforced epoxy joints had 100% cohesive failure. The volume fractions were calculated from weight fractions using density of epoxy as 1.25 g cm3 (as per manufacturer), density of CNT as 1.8 g cm3, [39] density of GNP as 2.2 g cm3, [39] and density of CNH as 1.2 g cm3 [51]. Fig. 11 shows the fractional increase in strength with filler content for different reinforcements. The value of R was obtained using a liner fit for up to 0.5 wt.% filler addition. From Fig. 11, the highest strengthening efficiency was observed for GNP (167) which is 1.5 times and 2.2 times higher than the CNT (114) and CNH (77) respectively. The low strengthening efficiency observed for CNH could be due to their clustered nature. The TEM images in Fig. 1c shows that the dahlia like structures are interconnected with a lot of porosity in between. Clusters of CNH were also observed at higher fractions in the epoxy matrix. For the joints with filler concentrations above 1 wt.%, the strengthening is seen to be extremely low compared to predicted values (given by the linear fit). This drop is attributed to presence of clusters as well as a change from cohesive failure to mixed mode failure as described later. It is known that interfacial load transfer from the matrix to the nano-filler is the mechanism for the strengthening of the composite. For very strong particle–matrix interfacial bonding, Pukanszky et al. have provided an empirical relationship that relates the effect of composition and interfacial interaction on tensile yield stress of polymers reinforced with particles given by Eq. (5) [52,53].

rc ¼



 1  Vf rm expðBV f Þ 1 þ 2:5V f

ð5Þ

where the pre-exponential term is strength of the matrix considering the fillers to be equivalent to porosity (no load transfer condition) and B is an interaction parameter that quantifies the extent of load transfer. The parameter B is in turn related to the macroscopic characteristics of the filler–matrix interface which depends on the specific surface area of fillers, density of fillers, filler surface roughness and interfacial bonding energy [54]. rc, rm and Vf are same as described in Eq. (4). A value of B equal to zero indicates lack of interfacial bonding, where the particles do not carry any load and

Fig. 11. Strengthening efficiency of carbon nano-fillers in epoxy adhesive joints. The solid line represents the linear fit. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

61

Fig. 12. Variation of the normalized lap shear strength of carbon nano-filler reinforced epoxy adhesive joints with the volume fraction of nano-fillers. Solid line represents fitting. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

essentially act as porosity [55]. Eq. (5) has been applied to analyze the interfacial load transfer to different nano-fillers as shown in Fig. 12. In Fig. 12, only data up to 0.5 wt.% were used for fitting since at higher content, extrinsic effects such as poor dispersion came into picture. The GNP were observed to have highest B value of 152, which was 1.5 times larger than the CNT (103) and 2 times larger than CNH (73). The higher surface area of GNP contributes the increased interfacial interaction in the case of GNP compared to the CNT and CNH. Lower B value was observed in CNH are again to the enclosed porosity which does not help in load transfer. 3.8. Mode of failure of adhesive joints The mode of failure of the 0, 0.5 and 2 wt.% carbon nano-filler reinforced epoxy adhesive joints is shown in Fig. 13. It is observed that the mode of failure for neat epoxy adhesive joints is 100% cohesive (within the adhesive). For the reinforced epoxy adhesive joints at lower weight fractions, the mode of failure was cohesive as well. As the weight fraction of nano-filler increased, the mode of failure progressed from cohesive (within the adhesive) to adhesive (at the interface between the adhesive and substrate) type. This is clearly observed in the fracture sample pictures of 0.5 wt. % and 2 wt.% nano-filler reinforced joints shown in Fig. 13. For the 0.5 wt.% reinforced joints, cohesive failure was observed with the fracture surface on both substrates containing epoxy layer adhering on to them. For the 2 wt.% nano-filler reinforced joints, a mixture of adhesive and cohesive failure was observed. The regions corresponding to cohesive failure was rough while the adhesively failed surfaces were smooth as shown in Fig. 13. In the smooth surface, the epoxy surface was flat and shiny and the corresponding region on the other substrate did not show any epoxy coating. Cohesively failed regions were observed in both substrates at the same location. Fig. 6 showed that the strength of the joints reduces after 1 wt.% filler addition. But it is found from several studies in literature that the intrinsic strength of the composite must increase monotonically up to 2 wt.%. For example, Allaoui et al. showed the strength of epoxy increases with CNT addition up to 4 wt.% [25]. Zaman et al. reported that the mechanical properties were increased with the incorporation of GNP up to 2.5 wt.% [22]. In order to understand the reasons for adhesive failure, SEM was carried out on the fracture surface of 2 wt.% nano-filler

62

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

Fig. 13. Mode of failure of neat epoxy and carbon nano-filler reinforced epoxy adhesive joints. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 14. SEM images of fracture surfaces of (a) & (b) 2 EP–CNT, (c) & (d) 2 EP–GNP and (e) & (f) 2 EP–CNH in epoxy matrix at 2 wt.%. Arrow marks represent the agglomerates of carbon nano-fillers.

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

reinforced joints. The mixed mode of failure in these joints was characterized by the presence of smooth regions (corresponding to adhesive failure) and rough regions (corresponding to cohesive failure). The distribution of carbon nano-fillers on the smooth and rough surface of the fractured adhesive joints was studied. Fig. 14 shows high magnification SEM images of the fracture surface. The smooth surfaces featured large nano-filler clusters. These large clusters are expected to have poor adhesion with the substrate and consequently act as nucleation sites for the growth of interface cracks leading to delamination. The rough surfaces resulting from cohesive failure exhibited the fibrous surface of the fractured epoxy and did not contain any nano-filler clusters. For the 2EP–GNP sample, the smooth surface of GNP which have settled parallel to the substrate lead to formation of delamination cracks at lower loads resulting in adhesive failure. Even in the 2EP–CNT and 2EP–CNH fracture samples, a lot of agglomerates were observed on the smooth surface as shown in Fig. 14a and e. These agglomerates are the reason for reducing the strength of the joints, even below that of the neat epoxy joint. The present study shows that even though the nano-fillers may increase the intrinsic properties of adhesive, the joint strength can be limited at higher contents of nano-filers due to mixed-mode failure caused by settling of particle clusters. 4. Conclusions BrabenderÒ Plasti-CorderÒ shear mixer can be used for synthesizing well dispersed mixtures up to 1 wt.% of carbon nano-fillers. Shear thinning behavior is observed for both the neat epoxy and the carbon nano-filler/epoxy mixtures. Viscoplastic behavior was observed for nano-filler/epoxy mixtures in accordance with Casson’s model. The yield stress of the epoxy adhesive increased with the addition of nano-filler. The increase was high for CNH and CNT addition while it was small for GNP. The flat morphology and relatively lower volume fraction of GNP is attributed to the lack of network formation leading to increased viscosity. Thermogravimetric analysis showed that the thermal stability increased marginally for the carbon nano-filler/epoxy composites. The 2nd stage decomposition temperatures were increased by 11.5 °C for 1 wt.% CNT, 9.5 °C for 0.5 wt.% GNP, and 7.7 °C for 0.5 wt.% CNH. At higher fractions of carbon nano-fillers, thermal stability started to show a reducing trend. The lap shear test results showed that the joint strength increased with increase of CNT, GNP and CNH content up to 1, 0.5, and 0.5 wt.% respectively. It was observed that for up to 0.5 wt.% filler addition, EP–CNH showed better properties than EP–CNT and was similar to EP–GNP joints. However, increasing the filler content from 0.5 to 1 wt.% resulted in significant reduction in strength of EP–CNH and EP–GNP joints compared to EP–CNT. Addition of 1 wt.% CNT, 0.5 wt.% GNP and 0.5 wt.% CNH resulted in an increase of lap shear strength by 53%, 49% and 46% respectively compared to the neat epoxy adhesive joints. Higher strengthening efficiency was observed for GNP among the three carbon nano-fillers. SEM results showed that, at relatively low weight fractions nano-fillers were dispersed uniformly in the epoxy matrix without any agglomerates and cohesive failures were observed. At higher fractions of carbon nano-fillers a mixed mode (cohesive and adhesive) failure was observed reducing the lap shear strength even below that of the epoxy joint.

Acknowledgement We would like to acknowledge Technology Development Board (TDB) of Department of Science and technology (DST) of India for funding under project ‘‘Multi-Join”.

63

References [1] Banea M, da Silva LF. Adhesively bonded joints in composite materials: an overview. Proc Inst Mech Eng, Part L: J Mater Des Appl 2009;223(1):1–18. [2] Yu S, Tong MN, Critchlow G. Use of carbon nanotubes reinforced epoxy as adhesives to join aluminum plates. Mater Des 2010;31(Suppl. 1(0)):S126–9. [3] Higgins A. Adhesive bonding of aircraft structures. Int J Adhes Adhes 2000;20 (5):367–76. [4] Burkholder GL, Kwon YW, Pollak RD. Effect of carbon nanotube reinforcement on fracture strength of composite adhesive joints. J Mater Sci 2011;46 (10):3370–7. [5] Kweon J-H, Jung J-W, Kim T-H, Choi J-H, Kim D-H. Failure of carbon compositeto-aluminum joints with combined mechanical fastening and adhesive bonding. Compos Struct 2006;75(1–4):192–8. [6] Srivastava VK. Effect of carbon nanotubes on the strength of adhesive lap joints of C/C and C/C–SiC ceramic fibre composites. Int J Adhes Adhes 2011;31 (6):486–9. [7] Gerson AL, Bruck HA, Hopkins AR, Segal KN. Curing effects of single-wall carbon nanotube reinforcement on mechanical properties of filled epoxy adhesives. Compos Part A: Appl Sci Manuf 2010;41(6):729–36. [8] Sydlik SA, Lee J-H, Walish JJ, Thomas EL, Swager TM. Epoxy functionalized multi-walled carbon nanotubes for improved adhesives. Carbon 2013;59:109–20. [9] Wolf A, Buchman A, Eitan A, Fine T, Nevo Y, Heyman A, et al. Improved adhesives containing CNT/SP1 nano fillers. J Adhes 2012;88(4–6):435–51. [10] Zhang J, Luo R, Yang C. A multi-wall carbon nanotube-reinforced hightemperature resistant adhesive for bonding carbon/carbon composites. Carbon 2012;50(13):4922–5. [11] Kinloch A, Lee J, Taylor A, Sprenger S, Eger C, Egan D. Toughening structural adhesives via nano-and micro-phase inclusions. J Adhes 2003;79(8– 9):867–73. [12] Bhowmik S, Benedictus R, Poulis J, Bonin H, Bui V. High-performance nanoadhesive bonding of titanium for aerospace and space applications. Int J Adhes Adhes 2009;29(3):259–67. [13] Zhai L, Ling G, Li J, Wang Y. The effect of nanoparticles on the adhesion of epoxy adhesive. Mater Lett 2006;60(25):3031–3. [14] Zhai L, Ling G, Wang Y. Effect of nano-Al2O3 on adhesion strength of epoxy adhesive and steel. Int J Adhes Adhes 2008;28(1):23–8. [15] May M, Wang H, Akid R. Effects of the addition of inorganic nanoparticles on the adhesive strength of a hybrid sol–gel epoxy system. Int J Adhes Adhes 2010;30(6):505–12. [16] Park SW. Adhesion strength of glass/epoxy composite embedded with heattreated carbon black on the surface. Compos Part A: Part Appl Sci Manuf 2010;41(11):1597–604. [17] Schadler L, Giannaris S, Ajayan P. Load transfer in carbon nanotube epoxy composites. Appl Phys Lett 1998;73(26):3842–4. [18] Im H, Kim J. Thermal conductivity of a graphene oxide–carbon nanotube hybrid/epoxy composite. Carbon 2012;50(15):5429–40. [19] Yang K, Gu M, Guo Y, Pan X, Mu G. Effects of carbon nanotube functionalization on the mechanical and thermal properties of epoxy composites. Carbon 2009;47(7):1723–37. [20] Ghaleb Z, Mariatti M, Ariff Z. Properties of graphene nanopowder and multiwalled carbon nanotube-filled epoxy thin-film nanocomposites for electronic applications: the effect of sonication time and filler loading. Compos Part A: Appl Sci Manuf 2014;58:77–83. [21] Teng C-C, Ma C-CM, Lu C-H, Yang S-Y, Lee S-H, Hsiao M-C, et al. Thermal conductivity and structure of non-covalent functionalized graphene/epoxy composites. Carbon 2011;49(15):5107–16. [22] Zaman I, Phan TT, Kuan H-C, Meng Q, La LTB, Luong L, et al. Epoxy/graphene platelets nanocomposites with two levels of interface strength. Polymer 2011;52(7):1603–11. [23] Kamar NT, Hossain MM, Khomenko A, Haq M, Drzal LT, Loos A. Interlaminar reinforcement of glass fiber/epoxy composites with graphene nanoplatelets. Compos Part A: Appl Sci Manuf 2015;70:82–92. [24] Fraczek-Szczypta A, Blazewicz S. Manufacturing and physico-mechanical characterization of carbon nanohorns/polyacrylonitrile nanocomposites. J Mater Sci 2011;46(17):5680–9. [25] Allaoui A, Bai S, Cheng H-M, Bai J. Mechanical and electrical properties of a MWNT/epoxy composite. Compos Sci Technol 2002;62(15):1993–8. [26] Zhou Y, Pervin F, Lewis L, Jeelani S. Fabrication and characterization of carbon/ epoxy composites mixed with multi-walled carbon nanotubes. Mater Sci Eng A 2008;475(1):157–65. [27] Xu X, Thwe MM, Shearwood C, Liao K. Mechanical properties and interfacial characteristics of carbon-nanotube-reinforced epoxy thin films. Appl Phys Lett 2002;81(15):2833–5. [28] Tang L-C, Wan Y-J, Yan D, Pei Y-B, Zhao L, Li Y-B, et al. The effect of graphene dispersion on the mechanical properties of graphene/epoxy composites. Carbon 2013;60:16–27. [29] Chatterjee S, Nafezarefi F, Tai N, Schlagenhauf L, Nüesch F, Chu B. Size and synergy effects of nanofiller hybrids including graphene nanoplatelets and carbon nanotubes in mechanical properties of epoxy composites. Carbon 2012;50(15):5380–6. [30] Wang F, Drzal LT, Qin Y, Huang Z. Mechanical properties and thermal conductivity of graphene nanoplatelet/epoxy composites. J Mater Sci 2015;50 (3):1082–93.

64

P. Jojibabu et al. / Composites: Part A 82 (2016) 53–64

[31] Rafiee MA, Rafiee J, Wang Z, Song H, Yu Z-Z, Koratkar N. Enhanced mechanical properties of nanocomposites at low graphene content. ACS Nano 2009;3 (12):3884–90. [32] Qin W, Vautard F, Drzal LT, Yu J. Mechanical and electrical properties of carbon fiber composites with incorporation of graphene nanoplatelets at the fiber– matrix interphase. Compos Part B: Eng 2015;69:335–41. [33] Wehnert F, Langer M, Kaspar J, Jansen I. Design of multifunctional adhesives by the use of carbon nanoparticles. J Adhes Sci Technol 2015:1–11 [ahead-ofprint]. [34] Fidelus J, Wiesel E, Gojny F, Schulte K, Wagner H. Thermo-mechanical properties of randomly oriented carbon/epoxy nanocomposites. Compos Part A: Appl Sci Manuf 2005;36(11):1555–61. [35] Szeluga U, Kumanek B, Trzebicka B. Synergy in hybrid polymer/nanocarbon composites. A review. Compos Part A: Appl Sci Manuf 2015;73:204–31. [36] Lachman N, Wagner HD. Correlation between interfacial molecular structure and mechanics in CNT/epoxy nano-composites. Compos Part A: Appl Sci Manuf 2010;41(9):1093–8. [37] Soltannia B, Taheri F. Influence of nano-reinforcement on the mechanical behavior of adhesively bonded single-lap joints subjected to static, quasistatic, and impact loading. J Adhes Sci Technol 2014:1–19 [ahead-of-print]. [38] Guadagno L, Sarno M, Vietri U, Raimondo M, Cirillo C, Ciambelli P. Graphenebased structural adhesive to enhance adhesion performance. RSC Adv 2015;5 (35):27874–86. [39] Ma P-C, Siddiqui NA, Marom G, Kim J-K. Dispersion and functionalization of carbon nanotubes for polymer-based nanocomposites: a review. Compos Part A: Appl Sci Manuf 2010;41(10):1345–67. [40] Iijima S, Yudasaka M, Yamada R, Bandow S, Suenaga K, Kokai F, et al. Nanoaggregates of single-walled graphitic carbon nano-horns. Chem Phys Lett 1999;309(3):165–70. [41] Kim K-S, Rhee K-Y, Lee K-H, Byun J-H, Park S-J. Rheological behaviors and mechanical properties of graphite nanoplate/carbon nanotube-filled epoxy nanocomposites. J Ind Eng Chem 2010;16(4):572–6. [42] Casson N. Rheology of disperse systems. In: Proceedings of a conference organized by the British Society of Rheology. New York: Pergamon Press; 1959. [43] Chapartegui M, Markaide N, Florez S, Elizetxea C, Fernandez M, Santamaría A. Specific rheological and electrical features of carbon nanotube dispersions in an epoxy matrix. Compos Sci Technol 2010;70(5):879–84.

[44] Morote-Martínez V, Pascual-Sánchez V, Martín-Martínez JM. Improvement in mechanical and structural integrity of natural stone by applying unsaturated polyester resin-nanosilica hybrid thin coating. Eur Polym J 2008;44 (10):3146–55. [45] Mansourian-Tabaei M, Jafari SH, Khonakdar HA. Lap shear strength and thermal stability of diglycidyl ether of bisphenol a/epoxy novolac adhesives with nanoreinforcing fillers. J Appl Polym Sci 2014;131(6). [46] Liu Q, Zhou X, Fan X, Zhu C, Yao X, Liu Z. Mechanical and thermal properties of epoxy resin nanocomposites reinforced with graphene oxide. Polym-Plast Technol Eng 2012;51(3):251–6. [47] Seo DW, Lim JK. Tensile, bending and shear strength distributions of adhesivebonded butt joint specimens. Compos Sci Technol 2005;65(9):1421–7. [48] Li J, Luo R, Bi Y, Xiang Q, Lin C, Zhang Y, et al. The preparation and performance of short carbon fiber reinforced adhesive for bonding carbon/carbon composites. Carbon 2008;46(14):1957–65. [49] Bakshi S, Lahiri D, Agarwal A. Carbon nanotube reinforced metal matrix composites – a review. Int Mater Rev 2010;55(1):41–64. [50] Cha SI, Kim KT, Arshad SN, Mo CB, Hong SH. Extraordinary strengthening effect of carbon nanotubes in metal-matrix nanocomposites processed by molecularlevel mixing. Adv Mater 2005;17(11):1377–81. [51] Murata K, Kaneko K, Kanoh H, Kasuya D, Takahashi K, Kokai F, et al. Adsorption mechanism of supercritical hydrogen in internal and interstitial nanospaces of single-wall carbon nanohorn assembly. J Phys Chem B 2002;106(43):11132–8. [52] Turcsanyi B, Pukanszky B, Tüdõs F. Composition dependence of tensile yield stress in filled polymers. J Mater Sci Lett 1988;7(2):160–2. [53] Pukánszky B, Tüdös F, Jancˇarˇ J, Kolarˇik J. The possible mechanisms of polymerfiller interaction in polypropylene-CaCO3 composites. J Mater Sci Lett 1989;8 (9):1040–2. [54] Metın D, Tihminliog˘lu F, Balköse D, Ülkü S. The effect of interfacial interactions on the mechanical properties of polypropylene/natural zeolite composites. Compos Part A: Appl Sci Manuf 2004;35(1):23–32. [55] Fu S-Y, Feng X-Q, Lauke B, Mai Y-W. Effects of particle size, particle/matrix interface adhesion and particle loading on mechanical properties of particulate–polymer composites. Compos Part B: Eng 2008;39(6):933–61.