1 Copyright 2011 by ASME

INTER-LAMINAR MECHANICAL PROPERTIES IMPROVEMENTS IN CARBON NANOTUBES REINFORCED LAMINATED NANOCOMPOSITES

Davood Askari1 and Mehrdad N. Ghasemi-Nejhad

2

1Department of Engineering The University of Texas at Brownsville

Brownsville, TX 78520 davood.askari@utb.edu

2Department of Mechanical Engineering University of Hawaii at Manoa

Honolulu, HI 96822 USA nejhad@hawaii.edu

ABSTRACT

Owing to their superior mechanical properties, carbon nanotubes (CNTs) can be used as an additional reinforcement to improve the mechanical properties of laminated composite materials. To incorporate the excellent properties of CNTs into the existing traditional composite technology, vertically aligned high density arrays of CNTs were grown perpendicular to the surface of 2-D woven cloths and tows of various fibrous materials. The nano-forest like structures of fabrics is used to fabricate 3-D reinforced nanocomposites. Due to the presence of aligned CNTs in through-the-thickness direction, it is expected that the inter-laminar and through-the-thickness properties of the composite laminate will be improved considerably. To demonstrate the effectiveness of our approach, various composite single lap-joint specimens were fabricated for inter-laminar shear strength testing. Carbon woven cloths with and without CNTs nano-forests were inserted in between the single lap-joints using epoxy adhesive to measure the inter-laminar shear strength improvements due to the presence of through-the-thickness aligned CNTs nano-forests. It is observed that single lap-joints with carbon cloth insertion layers having CNTs nano-forest can carry up to 12% higher shear stress and 16% higher strain-to-failure. The failures of samples with nano-forests were completely cohesive while the sample with carbon woven cloth insertion failed adhesively. This concludes that the adhesion of adjacent carbon fabric layers can be considerably improved due to the growth of vertically aligned CNTs nano-forest in through-the-thickness direction. KEYWORDS: Carbon Nanotubes, Laminated Composites and Nanocomposites, Material Properties, Inter-laminar Shear Strength and Strain.

INTRODUCTION

The main advantage of composites, in addition to their high specific strength and stiffness, is their ability to be tailored towards a specific loading condition, i.e., placing the load carrying fibers where the loadings and stresses are. Nearly one-dimensional fiber materials with their anisotropic properties have been used as the reinforcements, along with a bonding material called the matrix, to manufacture structural composites in which mechanical loads are to be transferred through the embedded fibers. Fibers usually are very strong in the longitudinal direction but weak in lateral direction. Therefore, when they are used to make structural composites, the final product will have weak through-the-thickness mechanical properties. In addition, when 1-D unidirectional composites and/or 2-D woven laminated composites are manufactured the interlaminar properties are controlled by the matrix properties, since the adjacent layers are bonded by the matrix only, yielding poor interlaminar and through-the-thickness properties. This weakness often leads to interlaminar failures (such as delamination) in composites [1]. To overcome this problem, 3-D composites such as 3-D stitching and 3-D braiding have been proposed [2, 3]. The 3-D braided fibers, as raw materials, do not solve general purpose applications since the part thickness should be known in advance. In addition, in 3-D braided materials, the fiber directions are not orthogonal. As a result, the use of 3-D braided fiber architecture is limited to some specific applications and geometries. As far as the stitching is concerned, the thickness should be determined and then stitching performed. In this case, the fibers can be orthogonal; however, the post operation of stitching is performed only after the structure is designed to determine the thickness to be stitched to provide through-the-thickness fibers. In addition, while stitching can improve some through-the-thickness properties, it reduces the in-plane properties [4, 5].

Proceedings of the ASME 2011 International Mechanical Engineering Congress & Exposition IMECE2011

November 11-17, 2011, Denver, Colorado, USA

IMECE2011-64283

2 Copyright 2011 by ASME

Above all, the traditional composites lack any rooms for multifunctionality.

It is reported that the addition of certain nanostrutured

materials as secondary reinforcement may lead to improvement of the composite materials if a properly processed and optimum amount of nanomaterials is used [6-10]. Owing to their superior material properties CNTs are one of the best candidates to be used as an effective reinforcing material [11-13]. Cao et al. [14] and Veedu et al. [15] developed a 3-D multifunctional hierarchical nanocomposite, where a new technique was introduced to grow carbon nanotubes in the perpendicular (through-the-thickness) direction on silicon carbide (SiC) fibers and woven cloths similar to a nano-brush and nano-forest. Using the nano-forest layers, they fabricated a truly 3-D laminated nanocomposite with superior through-the-thickness properties. Moreover, they have introduced multifunctional capabilities in their novel nanocomposite, such as increased mechanical properties as well as manipulation and control of coefficient of thermal expansion, electrical conductivity, thermal conductivity, and structural damping [15, 16]. They have successfully shown that the material properties of their 3-D laminated nanocomposite such as fracture toughness GIC, GIIC, flexural modulus, flexural strength, flexural toughness, damping, coefficient of thermal expansion, through-the-thickness thermal conductivity, and through-the-thickness electrical conductivity have improved, substantially. These results show the effectiveness of the proposed solution for improvement of the through-the-thickness materials properties and multifunctinality of the laminated composites by means of additional radially aligned CNTs reinforcements over the fiber cloths. However, in their work, they only grew CNTs on SiC fibers/cloths and it was not possible to directly grow CNTs on comonly used non-SiC (such as Glass, Kevlar, and Carbon) fibers/cloths.

Here in this work, we developed a new techniques to grow radially alligned CNTs on non-SiC fibers and fiber cloths [16]. Once CNTs are grown on fibers and fiber cloths, the same procedures for matrix impregnations, lay-up laminations, and curing, as used in a traditional wet lay-up technique for composites manufacturing [1], can be used to develop 3-D hierarchical nanocomposites with superior through-the-thickness properties and multifunctionality, as demonstrated by Ghasemi-Nejhad and co-workers [14-16]. To demonstrate the effectiveness of our approach, various composite single lap-joint specimens were fabricated for inter-laminar shear strength testing. Carbon woven cloths with and without CNTs nano-forests were inserted in between the single lap-joints using epoxy adhesive to measure the inter-laminar shear strength improvements due to the presence of through-the-thickness aligned CNTs nano-forests. It is observed that single lap-joints with carbon cloth insertion layers having CNTs nano-forest can carry up to 12% higher shear stress and 16% higher strain-to-failure. The failures of samples with nano-forests were completely cohesive while the sample with carbon woven cloth insertion failed adhesively. This concludes that the adhesion of adjacent carbon fabric layers can be considerably improved due to the growth of vertically aligned CNTs nano-forest in through-the-thickness direction.

GROWTH OF CARBON NANOTUBES OVER 2-D FABRICS OF VARIOUS MICROFIBERS

Chemical Vapor Deposition (CVD) technique is one of the very simple and economical ways of producing CNTs and was introduced in 1993 by Endo et al. [17]. In general, a mixture of hydrocarbon (e.g., exylene) and metal catalyst particles (e.g., ferrocene) is introduced (i.e., in vapor phase), through an inert gas flow (e.g., Ar or mixture of Ar and H2), into a quartz tube furnace at the center of which a substrate is placed. One of the main advantages of CVD growth technique is that aligned CNTs can be grown with desired architectures and at previously patterned substrates for desired applications. For our research, the CVD growth technique, similar to that introduced by Andrews et al. [18], was chosen due to its simplicity and ability for substantial control over the important growth parameters such as: CNTs length, alignment, and pattern of growth. It also allows us to directly grow CNTs on various types of substrates. Depending on their composition, not all substrates are suitable for CNTs growth. Silicon and silicon dioxide based solid substrates are most widely used as substrates in CVD growth of CNTs. However, other types of substrates may also be used if coasted or doped with catalysis, prior to growth process. Several methods (e.g., SiC nanoparticles coating, SiC based pre-ceramic polymer coating, silicon dioxide puttering or chemical treatment) have been used to directly grow CNT arrays over fibrous materials [16]. In this study, chemical treatment with diluted HF acid was used to directly grow CNTs on different types of fabrics.

Fibers often are coated with a very thin layer of compatible

materials (i.e. sizing) for protection, ease of handling, and improved adhesion to the matrix. Thus, CNTs cannot directly grow on most of the commercially available fibers. To overcome this problem, the fiber coatings were removed and functionalized with diluted HF acid, and hence the main materials of the fibers were exposed to which the iron catalyst particles had tendency to attach during the CVD processing. Consequently, upon CVD process arrays of vertically aligned CNTs were successfully grown on chemically treated Sic, carbon, glass and Kevlar fibers.

Figure 1 shows a photograph of HF acid treated/etched carbon woven cloths before and after CVD processing. The CVD processing time for the layers from bottom to top was varied with increments of 20 minutes from 0 minutes to 60 minutes, respectively. It can clearly be seen that the CNTs yield for the top layers was much higher than the bottom layers, as one would expect. The size of the carbon woven cloths shown in Fig. 1 was roughly 25 mm by 25 mm. It should be noted that since the optimum growth zone for the growth of CNTs in our current CVD system is only 3.5 cm by 7 cm, the fabrication of larger samples may not be easily possible and require a larger furnace.

3 Copyright 2011 by ASME

Figure 1: HF treated/etched carbon fiber woven cloths subjected to various CVD processing times, with 20 minutes

increments, for the growth of CNTs nano-forests.

Figures 2 and 3 show the scanning electron micrographs (SEM) of uniform growth of the radially aligned CNTs on chemically treated carbon fiber cloth subjected to approximately 1 hour of the CVD process, with various magnifications. The average length of the grown CNTs was measured to be ~ 100 m. Similar results have been obtained for HF treated SiC, Kevlar, and glass fibers.

Figure 2: Low (left) and high (right) magnification SEM images of the radially aligned uniform growth of CNTs over

HF acid treated/etched carbon fiber-cloth, after approximately 1 hour of CVD process.

Figure 3: Low (left) and high (right) magnification SEM images of the radially aligned uniform growth of CNTs (Inset: helical CNT) over the HF acid treated/etched carbon fiber-cloth edge, after approximately 1 hour of CVD process.

SEM images with various magnifications provided in this section, demonstrates our success in CVD growth of radially aligned multi-walled carbon nanotubes (MWCNTs) on different micro-fibers. The average diameter of the examined micro-fibers were 6 m to 20 m and the diameter and length

of the radially grown MWCNTs were approximately 35 nm to 50 nm and 300 m to 400 m, respectively. It is observed that the rate of the CNTs growth on SiC fibers, SiC coated non-SiC fibers, and HF acid treated/etched glass fibers is almost similar (i.e., 300-400 m/hour) and considerably higher than the growth rate of CNTs over the HF acid treated/etched Kevlar and carbon fibers (i.e., ~100 m/hour). These unique structures provide very large chemically-physically available active surfaces which may have potential applications for nanocleaning, painting micro-surfaces, selective chemical absorption and filtration, high efficiency heat dinks and thermal-electrical conductors. In addition, they can be used to substantially enhance mechanical and physical properties of the composite materials. [14, 15]. Since the CNTs are radially grown on the fibers, it is expected that they will considerably contribute to the improvement of not only mechanical properties (e.g., strength, strain-to-failure, fracture toughness, coefficient of thermal expansion, and damping), but also physical properties (e.g., thermal and electrical conductivities) [15] of the traditional laminated composites. As a result, the use of these CNTs grown fibers/cloths facilitates the fabrication of high performance 3-D laminated nanocomposites with multifunctional capabilities. For further demonstration, CVD processed HF acid treated carbon fiber cloths, were used fabricate single joint samples to mechanically measure the shear strength improvements. The presence of CNTs in between the fabric layers and through-the-thickness direction is expected to considerably improve the interface properties that will translate into enhancement of the fiber-matrix adhesion. SAMPLE PREPARATION AND TESTING

In this section of, sample preparations and mechanical single lap-joint shear strength testing based on the ASTM D5868-01 is presented for three different sets, denoted by A, B, and C. Lap joints are widely used in adhesive joints, as they are simpler to make and assemble, and the stress developed in the adhesive is almost always shear [19]. For each set of samples, a layer of carbon woven cloth with or without a CNTs nano-forest on one or both sides, is placed in between two rectangular bars made of carbon fiber laminated composite (see Fig. 4). Next tensile loads were applied to break the lap joints and measure the corresponding shear strength values. The measured shear strengths were in fact the adhesion strengths of middle carbon fabric layers to the adjacent laminae (i.e., the interface between the plain fabrics and the fabrics with CNTs nano-forests).

Figure 4: Schematic of the single lap-joint shear strength test based on ASTM D5868-01.

4 Copyright 2011 by ASME

The carbon woven cloth used for set A samples is plain in both sides with no CNTs. The set B samples have a layer of carbon woven cloth with vertically aligned CNTs nano-forest grown on both sides. Specimens in the set C samples have a layer of carbon woven cloth with CNTs nano-forest grown only on one side. These samples are directly used to demonstrate the adhesion enhancement (i.e., higher inter-laminar shear strength) of the carbon fabric layers with vertically aligned through-the-thickness CNTs nano-forests as compared to the plain carbon cloths with no CNTs. The average length of the CNTs grown on carbon fabrics was roughly estimated to be ~ 40 micron, after a 30 minute 30 min CVD process. To manufacture the rectangular carbon fiber/epoxy laminated composite adherends, 8 layers of stain weave (5-harness) carbon prepregs [20] were hand laid-up on a solid aluminum plate followed by a vacuum bagging and an autoclaving technique. Note that a symmetric quasi-isotropic stacking sequence was used and then the manufactured laminated composite was cut to obtain samples with the dimensions suggested by the ASTM standard D5868-01, (see Figs. 4). To assemble the adherends (i.e., composite laminated bars and carbon plain weaves with or without CNTs nano-forest), a very thin layer of SC-15 epoxy resin & hardener [21] is used as an adhesive in between the adherends, and then the single lap-joint samples were uniformly compressed in the overlapped adhesion area using two solid disks and a C-clamp and then is placed inside a mechanical convection oven followed by a cure cycle according to the manufacturers suggestion [21]. Initially, the temperature of the convection oven was increased from the room temperature of 23 C to 150 C at a rate of 1.5 C/min and then it was kept at 150 C for nearly 1 hour followed by natural cooling. Five samples were prepared for each set (i.e., sets, A, B, and C) as suggested by ASTM D 5868-01. Finally, prepared single lap-joint shear test samples were tested using an Instron testing machine (see Fig. 5), and then the average values of shear strengths and strains-to-failure for each set of samples were obtained. The next section of this report presents and compares the results obtained from single lap-joint shear tests and discusses the effects of the additional through-the-thickness CNTs nano-forest reinforcements on inter-laminar shear strength of the laminated composites.

Figure 5: A single lap-joint composite sample under shear stress due to the application of axial tensile load.

RESULTS AND DISCUSSIONS

As mentioned earlier, five specimens were tested for each set of samples from which average values of shear strength and strain to failure were obtained. To calculate the shear strength of each sample, the maximum tensile load was simply divided by the overlapping bonded area and to calculate the strain-to-failure value, the axial extension was simply divided by the gauge length of the specimen. The average values of shear strength and strain-to-failure for the set A specimens (i.e., single lap-joint samples with bare carbon plain weave cloth) were 11.62511 MPa and 0.01471, respectively. It was observed that the fracture of the set A specimens had occurred through the adhesive layer and the separation of adherends had occurred within the adhesive bound, suggesting an adhesive failure mode. The bare carbon weave layers were entirely left on one side of the fractured surface without any tear or fiber pullout/distortions (see Fig. 6). It can be seen that the fracture surface is very clean and shiny, meaning that the failure had occurred right at the interface of the carbon fabric layer and surface of the composite laminate, at the adhesive.

Figure 6: A typical fracture surface observed for set A specimens, after single lap-joint shear test.

In general, there are two dominant mechanisms of failure for adhesively bonded joints, namely; adhesive failure and cohesive failure. Adhesive failure in fact is the interfacial failure between the adhesive and one of the adherends, which is indicative of a weak-boundary layer adhesion. On the other hand, cohesive failure is when the fracture results in a layer of adhesive remaining on both adherend surfaces and, more rarely, when the adherend fails before the adhesive, with fracture almost contained in the adherend [19]. This later failure is known as cohesive failure of the substrate. It should be mentioned that the ideal type of failure is when cohesive failure occurs through the adhesive or one of the adherends. It can be concluded that with this type of failure mode, the joining system is a strong one.

Next, the specimens from set C samples (i.e., single lap-joint samples with CNTs nano-forest grown only on one side of the carbon fabric layer) were tested, for which the average values were almost similar to set A samples, with no considerable differences. While assembling the set C samples and prior to the shear strength testing, the sides with CNTs nano-forests were marked to examine whether the specimens will break at

5 Copyright 2011 by ASME

the bare side or the CNTs nan-forest side. As was expected, fractures occurred on the bare side of the carbon weave layers where no CNTs were grown. The rupture of lap-joints for set C samples where within the adhesive layer resulting in adhesive/interface failure. Note that the fracture surfaces where similar to those of set A samples, as shown in Fig. 6. Based on the results from the testing of set A and set C samples alone, it is evident that the inter-laminar shear strength of the adhesive bonds between the carbon fabric layers with CNTs nano-forests and composite laminates is much higher than that of bare carbon fabrics with no through-the-thickness CNTs reinforcements. It can be stated that the presence of vertically aligned CNTs perpendicular to the surface of the 2-D carbon weave cloths (i.e., through-the-thickness direction) considerably contributes to a more efficient shear stress load transfer and enhances the interface properties of the laminated composite structures. In other words, additional through-the-thickness vertically aligned CNTs reinforcements improves the adhesion of the adjacent layers, resulting in laminated nanocomposite with much higher inter-laminar shear strength, as well as those properties improvements demonstrated by Veedu et al. [15].

To quantify the inter-laminar shear strength enhancements due to the presence of vertically aligned CNTs nano-forests, set B samples were tested similar to sets A and C samples. The average values of shear strength and strain-to-failure for the set B specimens (i.e., single lap-joint samples with vertically aligned CNTs nano-forests grown on both sides of the carbon fabric layer) were 13.00745 MPa and 0.01709, respectively. These values are considerably higher than those obtained for sets A and C samples and show nearly 12% and 16% improvements in shear strength and strain-to-failure, respectively. It should be mentioned that the results from testing different samples were very consistent with standard deviations of less than 3%. Close examination of the fracture surfaces on set B samples reveals that the fracture occurred within and through the inserted carbon fabric layer and not at the interface regions (see Fig. 7). It is clearly evident that the carbon fabric is completely torn apart and has remained on both sides of the composite lap joint. The rupture region is within the inserted carbon fabric layer and not the interfaces, suggesting that the inter-laminar shear strength of vertically aligned CNTs nano-forest carbon fabrics is even much higher than the values obtained for set B samples.

Figure 7: Typical fracture surfaces observed for set B specimens, after single lap-joint shear test for both sides of the

sheared portion of the specimen.

Therefore, this is a cohesive failure where the adherend has failed before the adhesive, and the fracture occurred through the adherend, which is an ideal type of failure. These observations simply demonstrate that through-the-thickness material properties (e.g., inter-laminar shear strength and strain-to-failure) of laminated composite structures, using vertically aligned 3-D CNTs nano-forest woven fabrics have been substantially improved. However, a more accurate estimate of the inter-laminar shear strength properties enhancements could be obtained, using this test, if the properties of the inserted carbon fabric layers were much stronger in shear. To accurately measure GIC and GIIC properties, DCB and ENF tests [15] have to be carried out.

For further verification, SEM images of the fractured surfaces also are obtained for sets A and B samples, some of which are shown in Figs. 8 and 9, respectively.

Figure 8: A typical SEM image of the fracture surfaces observed for set A specimens, after single lap-joint shear test.

SEM image in Fig. 8 shows the direct failure of the adhesive layer between the inter-layer and the laminated specimen. On the contrary, one can clearly observe the presence of fractured bare carbon fibers in SEM image of Fig. 9 that is a verification of a cohesive failure. The surface of the inter-layer carbon fabric in Fig. 8 does not show any fiber distortion or breakage where the surface is covered with a thin layer of adhesive and where the failure occurs in a clean shear mode. This is an evidence for adhesive failure mode in set A samples. However, the fiber breakage and pull-out can clearly be seen in Figs. 7 and 9.

Figure 9: A typical SEM image of the fracture surfaces observed for set B specimens, after single lap-joint shear test.

Inset: close-up view of the fibers fracture.

6 Copyright 2011 by ASME

CONCLUSIONS

Composite materials have a wide range of applications, depending on their compositions, shape, size, and properties. Due to their light weight, multifunctional characteristics, and high strength, composite materials have been widely used in space and aerospace structures, as well as in sport and automotive industries. The development of nanocomposite materials with improved properties and characteristics can significantly benefit the applications where high performance materials are greatly desired. Chemical vapor deposition technique have been employed to grow vertically aligned arrays of CNTs perpendicular to the surface of chemically treated 2-D woven cloths and tows of various fibrous materials. The nano-forests like fabrics can then be used to fabricate 3-D reinforced laminated nanocomposites. Due to the presence of aligned CNTs in through-the-thickness direction and in between the adjacent fabric layers of laminated composite, it was expected that the inter-laminar and through-the-thickness properties of composite laminates would be improved, considerably, and the fabricated composite structure would possess multifunctional capabilities. To demonstrate the effectiveness of our approach, composite single lap-joint specimens were fabricated for inter-laminar shear strength testing. Carbon woven cloths with and without CNTs nano-forests were inserted in between the single lap-joints using epoxy adhesive to measure their inter-laminar shear strength improvements. It is observed that single lap-joints with carbon cloth insertion layers having CNTs nano-forest can carry up to 12% higher shear stress and 16% higher strain-to-failure. Fractured surfaces were examined under SEM. The failures of samples with nano-forests insertions were completely cohesive, while the samples with plain carbon woven cloth insertions failed adhesively. This concludes that the adhesion of adjacent carbon fabric layers can be considerably improved due to the presence of vertically aligned CNTs nano-forests in through-the-thickness direction.

ACKNOWLEDGMENTS

The authors acknowledge the Office of Naval Research (ONR) for the financial support under the government grant number of N00014-05-1-0586 for the ADPICAS project. The first author also acknowledges the Department of Engineering and the College of Science, Mathematics, and Technology of the University of Texas at Brownsville for their support.

REFERENCES

1. Mazumdar, S. K., Composites Manufacturing: Materials, Product, and Process Engineering (CRC Press, Boca Raton, USA, 2002).

2. Tong, L., Mouritz, A. P. & Bannister, M. 3D Fiber reinforced polymer composites, (Elsevier Science Ltd, Oxford, UK, 2002).

3. Dow, M. B., & Dexter, H. B. Development of stitched, braided and woven composite structures in the ACT program and at Langley Research Center (http://techreports.larc.nasa.gov/ltrs/PDF/1997/tp/NASA-97-tp206234.pdf). NASA Technical Report, NASA/TP-97-206234 (1997).

4. Mouritz A. P. Fracture and tensile fatigue properties of stitched fiberglass composites. Journal of Materials: Design and Applications 218, 87-93 (2004).

5. Dexter, H. B., & Funk, J. G. Impact resistance and interlaminar fracture toughness of through-the-thickness reinforced graphite/epoxy. 27th AIAA Structural Dynamics and Materials Conference, AIAA Paper 86-1020, 700709 (1986).

6. Tai, N. H., Yeh. M. K., & Liu. J. H. Enhancement of the Mechanical Properties of Carbon Nanotube/phenolic Composites using a Carbon Nanotube Network as the Reinforcement. Carbon 42, 774777 (2004).

7. Calvert, P. Nanotube composites: a recipe for strength. Nature 399, 210-211 (1999).

8. Thostenson, E. T., Ren, Z. F. & Chou, T. W. Advances in the science and technology of carbon nanotubes and their composites: a review. Composites Science and Technology 61, 1899-1912, (2001).

9. Thostenson, E. T., Li, C. & Chou, T. W., Nanocomposites in contextes. Composites Science and Technology 65, 491516 (2005).

10. Ghasemi Nejhad, M. N., Veedu, P. V., Askari, D., Yuen, A. Polymer Matrix Composites with Nano-scale Reinforcement. June 2007, US Publication Number: 20070142548, May 2008, International Publication number: WO/2008/060294.

11. Ajayan P. M., Schadler L. S., and Braun P. V. Nanocomposite Science and Technology (Wiley-VCH, Weinheim, 2003).

12. Ajayan, P. M. Nanotubes from Carbon. Chemical Reviews 99, 1787-1799 (1999).

13. Cao A., Dickrell P. L., Sawyer W. G., Ghasemi-Nejhad M. N., Ajayan P. M. Super-Compressible Foamlike Carbon Nanotube Films. Science 310, 307-310 (2005).

14. Cao A., Veedu V. P., Li X., Yao Z., Ghasemi-Nejhad M. N., and Ajayan P. M. Multifunctional Brushes Made from Carbon Nanotubes. Nature Materials 4, 540545 (2005).

15. Veedu V. P., Cao A., Li X., Ma K., Soldano C., Kar S., Ajayan P. M., and Ghasemi-Nejhad M. N. Multifunctional Composites Using Reinforced Laminae with Carbon Nanotube Forests. Nature Materials 5, 457-462 (2006).

16. Ghasemi Nejhad, M. N., Veedu P. V., Cao, A., Ajayan, P., Askari, D. Three-Dimensionally Reinforced Multifunctional Nanocomposites. June 2007, US Publication Number: 20070128960, August 2008, International Publication Number: WO/2008/054409.

17. Endo, M., Takeuchi K., Shiraishi M. and Kroto H.W. The Production and Structure of Pyrolytic Carbon Nanotubes (PCNTs). Journal of Physics and Chemistry of Solids 54, 1841-1848 (1993).

18. Andrews, R., Jacques, D., Rao, A. M., Derbyshire, F., Qian, D., Fan, X., Dickey, E. C. & Chen, J. Continuous Production of Aligned Carbon Nanotubes: a Step Closer to Commercial Realization. Chemical Physics Letters 303, 467-474 (1999).

19. Messler, R. W., Joining of Advanced Materials, (Butterworth-Heinemann, Troy, NW, USA, 1993).

20. Product Specification Sheet, Hexcel Corporation, Dublin, CA, (2007).

21. Products Data Sheet, Applied Poleramic Inc., Benicia, CA, (2008).