Mixed mode properties of CNT reinforced composites using Arcan test rig

Abstract

Composite materials reinforced with carbon nanotubes were mechanical tested using Arcan test rig under Mode-I, Mode-II and mixed mode loading conditions to obtain their fracture properties. The butterfly composite specimens were fabricated with 0.02, 0.05 and 0.1 wt % CNTs. The polyester/CNT composite was fabricated using VRTM (Vacuum Resin Transfer Molding) where the CNTs were first functionalised to reach an optimum properties. Arcan test rig was designed and fabricated to work with the Shimadzu testing machine. The results show that the functionalised CNTs have improved the fracture behavior by acting as bridge between the cracked face. In addition, the fracture properties were not improved for the higher weight fraction of 0.1 wt% CNTs.

Keywords

CNT / composites / Arcan test rig / stress intensity factor

Introduction

Composite materials have been existence for many years and researchers are currently focusing on improving their properties using nano-fillers than the conventional macro fibers. Carbon based nanomaterials such as carbon nanotubes (CNTs) and carbon nanofibers (CNFs) have been used as preferable reinforcements for different types of composites. Since CNTs exhibit excellent mechanical and physical properties, they are more preferable reinforcements than their counterparts. Recent research work has proved that the small addition of CNTs can have a greater impact on improving the properties of composites [ 1, 2]. Moreover, due to the higher aspect ratio (length to diameter), CNTs seize the propagation of nano-cracks by acting like a bridge between the cracked plan and demand higher energy for crack propagation. Lastly if the CNTs are uniformly distributed in to the matrix, they prevent the failure modes with significant strength due to their excellent inherent properties. There are few research works focused on understanding the effect of CNTs in improving the fracture properties of composites. The load bearing ability and the fracture energy of MWCNTs reinforced composites were studied by Karapappas et al. [ 3] and found that for the addition 1%wt CNTs, the Mode-I and Mode-II fracture energy have improved to 60% and 40% respectively.

In many structural applications, complex stress states predominate and a potential defect can have an adverse effect on the functionality of any material. Hence the failure propensity cannot be determined using uniaxial stress fields. Many experimental tests were used to measure the fracture behaviors of both metallic materials and composites such as double cantilever test [ 4] and compact test specimen [ 5] for Mode-I and end-notched flexure test [ 6] for Mode-II respectively. However the structural crack growth is generally not as a result of pure Mode-I and Mode-II loading alone. Hence it is important to understand the mixed-mode loading fracture behavior. Several tests, such as the edge delamination tension [ 7], the crack-lap shear and mixed mode flexure [ 8] have been used to obtain the mixed-mode fracture properties. Nevertheless all these tests have one or more problems which limit their usage for these kinds of testing. The Arcan test method developed by Arcan et al. [ 9], was proven to be a suitable tool for investigating Mode-I, Mode-II and almost any combination of Mode-I and Mode-II (mixed modes) loading conditions with the same test specimen configuration [ 10].

Hence the purpose of the current investigation is to design and fabricate a modified Arcan test rig which will be further be used to characterize the fracture behavior of CNTs reinforced polyester composites. In addition, the effect of CNTs functionalization and the weight fractions of CNTs on improving the Mode-I and Mode-II and mixed mode fracture behavior will also be analyzed using the unique setup.

Experimental

Materials and specimen preparation

The multi-wall carbon nanotubes (MWCNTs) (refer Fig. 1) supplied by SABI NANO of South Africa were used as the nano-filler for producing CNT reinforced composites. The MWCNTs have the diameter in the range of 40− 45 nm and the average length of 100 µm with greater than 90% purity. The range of relative density is 1.7−2.1 mg/cm³. The matrix GP LAM 410 polyester resin and MEKP curing agent were supplied by AMT composites South Africa. A MR8 release agent was applied on the steel mold for the easy removal of the finished components. All reagents used were of analytical grade.

Functionalization

The performance of CNTs as reinforcements mainly depends on the dispersion behavior of CNTs in to the matrix which can be improved by attaching the functional groups on CNTs surface and termed as functionalization. Our recent research works [ 1, 2] have proved that the functionalized CNTs could be dispersed uniformly into the matrix and thus led to improved fracture properties. Hence the MWCNTs were functionalized before blending with the polymer matrix. As received multi-wall carbon nanotubes were functionalized with the acid mixture of 1:3 volume ratio of 55% concentrated nitric acid (HNO₃) and 50% concentrated sulfuric acid (H₂SO₄). First a paraffin filled beaker was heated up to a controlled temperature of 100°C, then the acid-CNTs solution was added and mechanically stirred in a round bottom flask with an attached Liebig condenser and shown in Fig. 2.

The process was carried out for three different functionalization durations of 6, 24 and 48 h and left to cool to room temperature. They were then washed with deionized water until neutral pH and vacuum filtered using a 3 µm porosity paper. The filtered functionalized MWCNTs were dried for an hour at 100°C. Three different functionalization times were selected to obtain the optimum functionalization time. TEM images were used to study the possible morphological changes due to the proposed chemical functionalization.

Figure 3 shows the TEM images of the functionalized MWCNTs. Before functionalization the carbon nanotubes have the traces of the catalyst and the amorphous carbon materials as shown in Fig. 1. After functionalization of 6 and 24 h, the MWCNT walls were clean and the catalyst along with the amorphous carbon were removed (Figs. 3(a) and (b)) from the nanotube structures. However, as the functionalization was increased to 48 h, the nanotubes appeared to be exfoliated and the tube structure were damaged due to the long duration of functionalization. It was also clear that the aggressive acid functionalization for 48 h, has destroyed the tube structure and shorten the length, which could affect the CNTs properties as a strengthening reinforcements. Based on the results, for the current analysis, the MWCNTs were functionalized for 24 h only.

Fabrication of fCNTs reinforced composite specimens

The CNTs reinforced composites were fabricated using a Vacuum Resin Transfer Molding (VRTM). As a first step, functionalized CNTs (fCNTS) were mechanically mixed in acetone for 30 min. The polyester resin was then added in to the acetone/CNTs solution and sonicated for 2 min. Finally the resin mixture was stirred by hand and kept ready for composite fabrication. The MWCNTs reinforced composites were fabricated with 0.02, 0.05 and 0.1 wt% of fCNTs respectively. Before starting the fabrication process, the steel mold was cleaned using acetone and then dried using a dry cloth. As part of the priming procedure, a thin film of MR8 release agent was applied to the mold surface. Figure 4 shows a schematic of the three piece VRTM mold used for the fabrication of composite specimens. After setting up the VRTM, the resin solution was mixed with the curing agent and kept ready at the inlet port. The resin feed valve was then slowly opened. As the resin filled the mold, the vacuum pressure was maintained at a constant value of 18 kPa. The feed line was kept open until the resin exited the mold. Finally, the resin was allowed to cure at room temperature. A similar procedure was followed for all three different weight fractions of fCNTs.

Mixed mode characterization

Test specimen

Arcan types tests are used to study the material behavior in Mode-I and II (tensile and shear respectively) and mixed mode behavior of composites. The butterfly specimens (Fig. 5) were cut from the fabricated composites panels. The specimens were pre-cracked with three different crack lengths of 3, 6 and 9 mm respectively using a thin band saw of 0.8 − 1 mm thickness. Each three specimens were tested for the individual weight fractions of 0.02, 0.05 and 0.1 wt% of fCNTs respectively.

ARCAN test rig and testing procedure

For the proposed research work, an Arcan test rig (Fig. 6) was designed and fabricated to conduct experiments using a Shimadzu tensile testing machine. The two circular test rig pieces and other accessories were machined from a 300 mm diameter by 20 mm thick EN8 steel plate. The machined parts were then case hardened. The fabricated test rig can be loaded in five different angular positions Mode I (0°) mixed mode (22.5°, 45°, 67.5°) and Mode II (90°) respectively.

Before starting with actual experiments, the Arcan test rig was mounted on the Shimadzu tensile testing machine with dummy specimens to check whether the designed Arcan test rig was working as per the testing requirements. The specimens were loaded into the test rig before mounting with the tensile tester. Figure 7 shows the four different loading positions of the designed Arcan test rig. The rate of loading for all the test was kept at 1 mm/min. During experiments, special care was taken to mount the butterfly specimens at the center of the Arcan test rig so that the center of the specimen will be aligned with the testing machine loading axis.

LEFM

Linear Elastic Fracture Mechanics (LEFM) has been found to be a useful tool for investigating cracks in steel and plastic materials. The purpose of the fracture toughness testing is to determine the ability of the material resistance to fracture which is an important parameter for designing structural members. ASTM standards E 399 [ 11] and D 5045 [ 12] specify some guidelines for plane strain Mode-I fracture toughness (K_IC) for metals and plastics. The stress intensity factor (K_c) at the tip of the pre-crack in a compact tension specimen is given by

(1) K_{\mathrm{c}}=Y{\displaystyle \frac{P_{\mathrm{c}}}{wt}}\sqrt{\mathrm{\pi }a}.

The stress intensity factor at the crack tip for a modified version of an Arcan specimen were calculated by using [ 12]

(2) K_{\mathrm{I}\mathrm{c}}=Y_I{\displaystyle \frac{P_{\mathrm{c}}}{wt}}\cos a\sqrt{\mathrm{\pi }a},

(3) K_{\mathrm{I}\mathrm{I}\mathrm{c}}=Y_{\mathrm{II}}{\displaystyle \frac{P_{\mathrm{c}}}{wt}}sina\sqrt{\mathrm{\pi }a},

where P_c is the fracture load, a is the crack length, w is the specimen width, t is the specimen thickness, α is the loading angle and Y is a geometrical factor. The K_Ic and K_IIc are obtained using the geometrical factor values and given below

(4) Y_I=-\mathrm{2.0841}+\mathrm{27.296}(\mathit{a}/\mathit{w})-\mathrm{66.859}{(\mathit{a}/\mathit{w})}^{\mathrm{2}}+\mathrm{62.614}{(\mathit{a}/\mathit{w})}^{\mathrm{3}},

(5) Y_{\mathrm{II}}=\mathrm{1.0459}+\mathrm{0.2528}(\mathit{a}/\mathit{w})-\mathrm{1.5878}{(\mathit{a}/\mathit{w})}^{\mathrm{2}}+\mathrm{2.4909}{(\mathit{a}/\mathit{w})}^{\mathrm{3}}.

Results and discussion

Mode-I and Mode-II fracture behavior

For the current investigations, the test specimens were loaded with tensile mode (α = 0°), shearing mode (α = 90°) and mixed mode with α = 45° loading conditions. Fracture tests were carried out using the designed Arcan test rig with Shimadzu tensile testing machine. For all the tests, the fracture loads and the displacements were recorded to generate the load-displacement curves which were then used to obtain the fracture load for further calculations. Since the CNTs reinforced composites were brittle in nature, the crack initiation stage was very difficult to identify. Hence the specimens were allowed to fail completely and the maximum force was recorded. The average values of fracture loads were used to determine the critical stress intensity factors. Figure 8 shows the Mode-I and Mode-II stress intensity factors (SIF) for 0.1wt% fCNT reinforced composites with three different crack lengths of 3, 6 and 9 mm respectively. As expected, the increase in crack length led to increase in both Mode-I and Mode-II stress intensity factors. Figure 9 shows a typical failed specimens with α = 0° (axial) and α = 90° (shearing) loading conditions.

Similar trends were also observed for other two (0.02 and 0.05 wt%) fCNTs weight fractions (Fig. 10). It could be concluded from the results that the fCNTs have improved the fracture properties of the composites. The Mode-I stress intensity factor for a 0.02 wt% fCNTs reinforced composite specimen with 3 mm crack length was 60 MPa·mm^1/2. As the CNTs weight fraction was increased to 0.05 wt%, the SIF value for the same crack length was increased to 80 MPa·mm^1/2, which was 33.3% increase in value. The increase in the fracture properties could be attributed to the uniform distribution of the functionalized CNTs within the polymer matrix. Here it should be pointed out that the functionalization has improved the distribution of the CNTs and thus improved the fracture properties of the composites. Moreover the uniformly distributed functionalized CNTs could have acted like bridges between the cracked face (bridging mechanism) which eventually increased the required energy to propagate the crack further. This bridging effect could be a main factor for increasing the stress intensity factor values as the fCNTs weight fraction was increased.

However for an increase of 0.1 wt% of CNTs, the increase in stress intensity factor was only 15%, which should be the result of agglomeration (refer Fig. 11) of fCNTs within the matrix. It could be assumed that at the higher weight fraction of fCNTs, agglomeration could not be prevented and resulted in ununiformed distribution of fCNTs into the matrix and thus led to reduction in fracture properties. Moreover it should be pointed out that the Mode-I fracture toughness is larger than the Mode-II fracture toughness [ 13]. This reflects that the cracked specimens were tougher in tensile loading than the shear loading conditions.

Mixed mode fracture behavior

The average values of the mixed mode stress intensity factor (Fig. 12) results are summarized in Table 1. The results depicted that the opening mode SIF is higher than the shearing mode. It was also clear that the specimens with 3 mm crack was capable resisting early crack initiation. Moreover the specimens with 0.05 wt% of fCNTs have improved the fracture behavior compared to 0.02 wt% fCNTs. However increase in fCNTs weight fractions to 0.1 wt% did not show appreciable improvement in fracture properties. The reason behind this phenomenon was explained in the previous paragraph. It could also be confirmed from the result was that for the mixed mode fracture, the tensile loading conditions were the main contributing factors than the shearing mode to initiate crack propagation.

Conclusion

CNT strengthened polyester composites were mechanically tested for obtaining the Mode-I, Mode-II and mixed mode fracture behavior. A modified Arcan test rig was designed and fabricated for conducting the experiments using a Shimadzu tensile testing machine. The butterfly testing specimens were fabricated with 0.02, 0.05 and 0.1 wt% fCNTs respectively. The stress intensity factor increased with increase in crack length. The fracture properties of the specimen improve as the fCNTs weight fraction was increased from 0.02 to 0.05 wt%. But similar trend was not seen for specimens fabricated with 0.1 wt% fCNTs. The experiments using Arcan test rig could further be expanded to fCNTs reinforced hybrid composites for realizing the potential of fCNTs as an additional reinforcements.

References

[1] Muthu J, Dendere C. Functionalized multiwall carbon nanotubes strengthen GRP hybrid composite: Improved properties with optimum fibre content. Composite Part B, 2014, 67: 84–94

[2] Jacob Muthu S D, Paskaramoorthy R. Double-Wall Carbon Nanotube-Reinforced Polyester Nanocomposites: Improved Dispersion and Mechanical Properties. Polymer Composites, 2012, 33(6): 866–872

[3] Karapappas P, Vavouliotis A, Tsotra P, Kostopoulos V, Paipetis A. Enhanced Fracture Properties of Carbon Reinforced Composites by the Addition of Multi-Wall Carbon Nanotubes. Journal of Composite Materials, 2009, 43(9): 977–985

[4] Wilkins D J, Eisenmann J R, Camin R A, Margolis W S, Benson R A. Characterizing Delamination Growth in Graphite-Epoxy. Damage in Composite Materials, ASTM STP 775, K. L. Reifsnider, ed. American Society for Testing and Materials, Philadelphia, 1982, 168–183

[5] American Society for Testing and Materials. Standard E399−06, Standard Test Method for Linear-Elastic Plane-Strain Fracture Toughness KIC of Metallic Materials. Annual Book of ASTM Standards. Philadelphia: ASTM, 2007

[6] Russell A J. On the Measurement of Mode II Interlaminar Fracture Energies, DREP Materials Report. 82−0, Defence Research Establishment Pacific, Victoria, December, 1982

[7] O'Brien T K. Mixed-Mode Strain-Energy-Release Rate Effects on Edge Delamination of Composites, Effects of Defects in Composite Materials, ASTM STP 836, D. J. Wilkins, ed. American Society for Testing and Materials, Philadelphia, 1984, 125–142

[8] Wolfenden A, Johnson W S. Stress analysis of the crack-lap-shear specimen: An ASTM round-robin. Journal of Testing and Evaluation, JTEVA, 1987, 15(6): 303–324

[9] Banks-Sills L, Arcan M, Gabay H. A mode II fracture specimen-finite element analysis. Engineering Fracture Mechanics, 1984, 19(4): 739–750

[10] El-Hajjar R, Haj-Ali R, 0. Rami Haj-Ali. In-plane shear testing of thick-section pultruded FRP composites using a modified Arcan fixture. Composite Part B, 2004, 35(5): 421–428

[11] ASTM E399. Standard Test Method for Plane Strain Fracture Toughness and Strain Energy Release Rate of Metalic Materials: Annual Book of ASTM Standards, 1983

[12] ASTM D5045. Standard Test Method for Plane Strain Fracture Toughness and Strain Energy Release Rate of Plastic Materials: Annual Book of ASTM Standards, 1995

[13] Ayatollahi M R, Shadlou S, Shokrieh M M. Mixed mode brittle fracture in epoxy/multi-walled carbon nanotube nanocomposites. Engineering Fracture Mechanics, 2011, 78(14): 2620–2632

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg