Available online at www.sciencedirect.

com

Procedia Engineering 10 (2011) 918

ICM11

Characterization and prediction of some engineering properties of polymer - Clay/Silica hybrid nanocomposites through ANN and regression models
Lingaraju.Da,*, Ramji.Kb, N B R Mohan Raob, Rajya lakshmi Ua
a

Associate Professor, Aditya Institute of Technology and Management, Tekkali, A.P, India b Professor, Andhra University, Visakhapatnam, India

Abstract The search from time immemorial for stronger, lighter and more durable materials used in construction and tools continues to this day. This urge to discover, invent and synthesize new materials has resulted in innumerable kinds of alloys, plastics and composites owing to the very exacting demands from different industries like aerospace, automobile, chemical, marine and so on. Glass fibers reinforced in polymers have received considerable attention during the last century. Studies conducted during the last decade reveal that adding small amounts of foreign particles (like clay or silica) of nanosize significantly improves the engineering properties of the polymers. This study presents the effect of reinforcing epoxy polymer with Halloysite nanoclay on mechanical properties. In the present work the clay in terms 1, 2 and 3 percent by weight was surface treated with a suitable silane agent. The hybrid nanocomposite was prepared by the hand lay-up technique. Characterization of the nanoclay was done by X-ray diffraction and Scanning electron microscopy. ASTM standards have been employed for investigating hardness, dynamic mechanical properties, wear, damping and post underwater shock. All the tests were done for both silica and clay as reinforcement materials. Analysis and Results have been presented for impact of nanoclay reinforcement on all the parameters. Radial basis function network, a tool of ANN, is employed to predict the above outcomes. Mathematical regression models have been employed for theoretical prediction. Comparative study reveals that ANN Tool predictions have better agreement with measured values than the other model. Thus, it can be confidently concluded that ANN Tool can be used to predict the properties of hybrid nanocomposites before actual manufacture. This will result in considerable savings in terms of project time, effort and cost.

2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of ICM11
Keywords: Artificial neural networks, regression models, polymer hybrid nanocomposites, Halloysite, post underwater shock, dynamic mechanical properties.

1877-7058 2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of ICM11 doi:10.1016/j.proeng.2011.04.004

10

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

1. Introduction Researchers all over the world are focusing on polymer hybrid nanocomposites because of the unique combination of improvement in corrosion resistance and in mechanical, tribological and thermal properties. Halloysite nanotubes (HNTs) were a kind of naturally deposited aluminosilicate with a high aspect ratio as stated by Joussein E [1] and Raw Halloysite purification method was studied by Shchukin DG [2]. From the structural point of view, two types of polymer/phyllosilicate nanocomposites are possible, Intercalated and Exfoliated stated by Giannelis EP [4]or a mix of these types as reported by Ray.S.S [3]. Gang Zhou et al. [5] reported that high-performance thermo set polymer composites were synthesized by using long fibers and nanoclays. Mechanical and thermal properties of synthesized composites were compared with long-fiber-reinforced composites and polymer layered silicate composites. Mingxian Liu et al. [6] reported that naturally occurring nanotubes, halloysite nanotubes, were modified by silane and incorporated into epoxy resin to form nanocomposites. Varada Rajulu L et al [7] reported on the structure, properties and surface modification of clay minerals, followed by the processing and characterization techniques of polymer nanocomposites. The mechanical properties of nylon 6-clay nanocomposite synthesized by in-situ polymerization were first demonstrated by researchers at the Toyota Central Research Laboratories (Kojima Y) [8]. Such nanocomposites exhibited significant improvement in strength and modulus. Chandradass J. et al. [9] studied that the damping efficiency increased with the addition of nanoclay material for limited weight percentage. Velmurugan R. et al [10] studied the conventional metals replaced by the composites in most of the high performance areas. Fillers like nano metal powders or inorganic clays with some treatment like chemical modification to its surface provided better performance than the natural structure. Mohan.T.P.et al [11] reported a well dispersed exfoliated layered silicate clay nanocomposites based on epoxy resins and synthesized by using organo modified alkyl quarternary ammonium modified or unmodified Semitic clay. Thermal stability of series of each nanocomposite system with different organo clay [OC] and unmodified clay [UC] concentrations were determined by using thermo gravimetric analysis (TGA). The performance of filled fiber reinforced polymer composites is generally determined on the basis of the interface attraction of filler, fiber, and polymer. In a recent investigation [12], the effects of SiC particles as fillers in glassepoxy (GE) composite systems on the mechanical and tribological properties have been discussed. Velmurugan at el studied the effect of nanoclay incorporation with various types of loading in the epoxy/glass fiber laminates [10]. Nanoparticles make a better inter phase property than the fibers in which the inter phase properties affects the damping character of a composite [13]. In high performance applications like aerospace applications vibration is an important parameter to be considered. In addition to surface ships and submarines, marine applications of FRP primarily involve offshore platform components, docks and piers. In offshore applications, FRP materials have been used for tension leg platforms, tether lines, risers, cables, tubing and drill pipes [14, 15]. Other environmental factors to be considered in a marine environment are the combined effects of moisture, salt, temperature and ultraviolet (UV) radiation [16]. Kie et al. [17] studied the dynamic mechanical properties and inferred that the damping capacity of the nanocomposites increased due to the addition of the nanoclay particles. The clay addition increased the storage modulus (E'), and the damping loss factor (tan = E"/E'). The increased E" caused such damping improvement in the nanocomposites. Batra R.C. et al. [18] reported three-dimensional finite element code, with rate dependent damage evolution equations for anisotropic bodies, and used it to numerically ascertain the damage developed in a fiber-reinforced composite due to shock loads produced by an underwater explosion. Mouritz A.P. [19] used the four-point bend test to measure the residual flexural strength of a glass reinforced polymer (GRP) laminate after it had been impulsively loaded by an underwater shock wave produced by an explosion. Will et al. [20] pointed out that for high velocity impacts, the structure responded in a local mode, a little energy was used to deform fibers and the structure, and a significant amount of energy was dissipated in mechanisms such as delaminating, debonding and fiber pull-out. Mouritz A.P. et al [21] examined and reported the development of micro structural damage in a glass reinforced polymer (GRP) laminate subjected to explosive shock loading in water. Velmurugan R. et al [22] studied the effect of clay addition on improving the mechanical properties of epoxy polymer and glass fiber-reinforced epoxy-clay hybrid composites.

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

11

Asifullah khan et al [23] studied and developed the artificial neural networks (ANN) and regression models to correlate the dynamical mechanical properties of polymer clay nanocomposites (PCN) with temperature and clay composition. Venkatesan et al [24] reported the study on the application of supervised feed forward Radial Basis Function (RBF) neural network with one hidden layer. It used relatively smaller number of locally tuned units adaptive in nature. RBFs were suitable for pattern recognition and classification. Performance of the RBF neural network was also compared with the most commonly used multilayer perceptron network model and the classical logistic regression. Diabetes database was used for empirical comparisons and the results showed that RBF network performed better than other models Reported by Lawrence and Peterson [25]. Very few works have been reported for modeling and prediction of wear properties of MMCs by using ANN technique. K.Velten [26], Jones [27], K.Genel [28] have used different ANN models in tribological applications. Their modeling results confirmed the feasibility of ANN and its good co-relation with the experimental results. In this study, Radial Basis Function (RBF) neural network which is a widely used algorithm is employed in training. The training of the RBF network is radically different from the classical training of standard FFNNs. In this case, there is no changing of weights with the use of the gradient method aimed at function minimization. In RBF networks with the chosen type of function, training resolves into selecting the centers and dimensions of the functions and calculating the weights of the output neuron. The centre, distance scale and precise shape of the radial function are parameters of the model, all fixed if it is linear. Selection of the centers can be understood as defining the optimal number of basis functions and choosing the elements of the training set used in the solution. This was done according to the method of forward selection [29]. In the present investigation the experimental results hardness, damping ratio, wear rate, post shock tensile strength and glass transmission temperature are predicted by Regression and ANN models and reported in the paper. 2. Characterisation of clay/silica XRD characterisation technique is used to find out the phases present in the composite and also to find the structural properties of the phases identified. X-ray diffractometer is used for the residual strains of the silica phase presence in the composite and characterisation of the prepared samples (20x20 mm). Only a few intensity lines coincide with the diffraction peaks. Profiles/peaks that match with the Intensity lines for FWHM curve have been selected as shown in Figure-1(a).

Figure-1(a): XRD pattern of nanoclay, (b) Planes representation of XRD pattern of silica nanocomposite.

The phase identified in nanoclay powder is the Halloysite-10A, Al2Si2O5 (OH) 4 and crystal structure is triclinic. From Figure-1(a), the intensity line 2=12.402 deg, coincides with the diffraction peak maximum. The FWHM value has to be considered for this particular Braggs angle. By using 2 and FWHM values, the crystal size is measured as 52nm. The XRD pattern from Figure-1(b) shows some crystalline nature (as indicated by a few sharp peaks) and these peaks closely match with silica (SiO2) peaks. From the XRD pattern, it is observed two sharp peaks occur at 14.13 and 17.2. The silica phase is identified and the structural properties are found from XRD pattern by Jade Evaluation Software.

12

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

3.0. Experimentation The synthesis of glass fiber reinforced epoxy with particle reinforced hybrids consists of mixing epoxy resin with nanoparticle by sonication process and then placing the mixture in the vacuum oven to remove the air bubbles at room temperature. This procedure results in a well dispersed state of nanofillers in the epoxy resin. The 50wt% of E-glass fiber woven rove mat, together with 50wt% of epoxy resin (pure, 1wt%, 2wt% and 3wt% of nanofillers) are used for preparation of hybrid composite by the hand-layup technique. The laminates are cured at room temperature and left in the mould for 24hrs for complete curing. Experiments are conducted using the Pin-on-disc type Friction and Wear monitor (DUCOM; TR-20-M26) with data acquisition system to evaluate the wear behaviour of the composite against hardened ground steel disc (En-32) having hardness 65 HRC and surface roughness (Ra) 0.5 m and Hardness is measured by Barcoll Hardness tester. Dynamic mechanical analysis (DMA) is done with the EPLEXOR (GABO) - 1500N machine.

3.1. CASE STUDY: Experimental Investigation of underwater shock test Exploding concentrated underwater charges to damage underwater structures such as ship hulls is a part of naval warfare strategies. Careful design of structures to withstand the shock loads still remains a challenge to structural designers. Hence, there is a continuing need to understand the structural behaviour of hull panels due to non-contact underwater explosion. The structural analysis is quite complex involving large deformation, high strain rates, material non-linearity and fluid-structure interaction [32]. A major threat to ship structures and marine vessels is their exposure to severe shock loads which could be due to the underwater explosion of a mine or a torpedo or the structure striking a partially submerged object in water, and/or the slamming pressure that occurs at high sea states when the forefront of the vessel rises above the water surface and then rapidly re-enters the water. These shock waves generally generate impulses of very high pressures of short durations, resulting in extremely high strain rates, which may result in severe structural damages [30,31]. The underwater shock tests were performed to simulate the conditions experienced by the GRP hull of a naval vessel when subjected to an underwater blast from an exploding mine, bomb or torpedo. The tests were performed in a water-filled pit as shown schematically in Figure-2. This small-scale shock testing facility consists of a steel cylinder confined within a concrete slab.

Figure-2: Schematic representation of the facilities for the underwater explosive shock testing of FRP laminates

The inside steel lining of the pit was covered with a thin plastic sheet containing small air bubbles, which was used to minimize the internal reflection of shock waves from the pit wall following the explosion. The pit has a diameter of 5m and a depth of 5m. Hybrid nanocomposites laminates were cut into square specimens of 850mm side for testing. The laminates were placed at the top of the fixture and were clamped on all sides at their ends which

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

13

allowed the FRP to bend under shock loading. The laminates were backed with water or air. The experimental setup is shown in Figure-3.

Figure-3: Assembly before dipping into the water pit

4.0. Results and Discussion 4.1. Hardness

46 45

44 42 40

Hardness

Hardness

44 43 42 41 40 0 0.25 0.5 0.75 1 1.25 1.5 1.75 w t% of nanoclay 2 2.25 2.5 2.75 3 Experimental ANN Regression

38 36 34 32 30 28 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3 w t% of nanosilica Experimental ANN Regression

Figure-4(a): Predicted and experimental values for hardness of clay hybrid nanocomposites, (b) Predicted and experimental values for hardness of silica hybrid nanocomposites

The Barcoll hardness increased from 41 to 45 by increasing the HNT nanoclay reinforcement. For nanosilica hardness increased from 41.1 to 42.8 for 1wt% and 2wt% and then decreased to 31.4 for 3wt% nanosilica. The Hardness results of regression model and neural network model are compared with the experimental results with different nanofiller reinforcements as shown in Figures-4(a&b). It is observed that maximum error is 3% for regression model and 0.036% for neural network model. 4.2. Damping ratio The increased natural frequency is due to good dispersion of nanoclay in the matrix which results in improved stiffness. It may be noted that the addition of clay has considerable effect in the damping behaviour of the hybrid nanocomposites.

14

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

The damping ratio for pure epoxy resin is 0.0139. For 3wt% nanoclay addition, the damping ratio is 0.0167. The addition of 3wt% nanoclay improves the damping ratio by 20.14%. For higher velocities the damping increases due to low contact time and dissipated low vibration energy in the laminate. Hence it is clear that addition of nanofiller results in good damping is shown in Figure-5.
DAMPING RATIO ( %) AT FIRST MODE
25

0.016
Damping Ratio

20
% Improvement

15

0.014 Damping ratio % Improvement

10

0.012 0 1 w t% nanoclay 2 3

Figure-5: Damping ratio and % Improvement Vs nanoclay reinforcement.

It is observed from the impact response that, clay addition controls the amplitude of oscillations and hence decreases impact damage. The increased natural frequency is due to good dispersion of nanofiller (nanosilica and nanoclay) in the matrix which results in improved stiffness. It may be noted that the addition of clay/silica has considerable effect in the damping behaviour of the hybrid nanocomposites. The damping ratio for 0wt% (pure) is 0.014. For 3wt% nanoclay addition the damping ratio is 0.017. The addition, of 3wt% nanoclay improves the damping ratio by 20.14%. For 3wt% nanosilica addition, the damping ratio is 0.0198. The addition of 3wt% nanosilica improves the damping ratio by 42.44%. For higher velocities the damping increases due to low contact time and low vibration energy dissipated in the laminate. Hence it is clear that, addition of nanofiller gives good damping. Damping ratio results from regression model and neural network model are compared with the experimental results for different nanofiller reinforcements. These are shown in Figures-6(a&b). While Neural Network model error was 0.098% and regression model indicated 2.47% error maximum.
1.7 1.65 1.6 1.55 1.5 1.45 1.4 1.35 1.3 0 0.25 0.5 0.75 1 1.25 1.5 1.75 wt% of nanoclay 2 2.25 2.5 2.75 3
Experimental ANN Regression
Damping ratio
2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 0 0.25 0.5 0.75 1 1.25 1.5 1.75 w t% nanosilica 2 2.25 2.5 2.75 3 Experimental ANN Regression

Figure-6: (a) Predicted and experimental values for damping ratio of clay hybrid nanocomposites, (b) Predicted and experimental values for damping ratio of silica hybrid nanocomposites

4.3. Wear rate Wear rate results of hybrid nanocomposites from regression model and neural network model are compared with the experimental results for different nanofiller reinforcements as shown in Figures-7(a&b). It is observed that maximum errors in hybrid nanocomposites are 0.192% and 0.193% for neural network model and regression model

Damping ratio

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

15

respectively. The wear rate of epoxy decreased from 0.24 to 0.04 micrometers/sec at load 25N and a travel of 1.81km as compared to 2% reinforced nanoclay hybrid nanocomposites and it increased to 0.16 micrometers/sec by reinforcing 3% nanoclay. The highest wear rate is seen with pure epoxy composite and lowest wear rate is at 2% nanoclay reinforcement.

Figure-7: (a) Predicted and experimental values for wear rate of clay hybrid nanocomposites, (b) Predicted and experimental values for wear rate of silica hybrid nanocomposites

The wear rate of epoxy decreased from 0.24 to 0.04 micrometers/sec at load 25N and a travel of 1.81km as compared to 2% reinforced nanosilica hybrid nanocomposites and it increased to 0.08 micrometers/sec by reinforcing 3% nanosilica. The highest wear rate is seen with pure epoxy composite and lowest wear rate is at 2% nanosilica reinforcement. From this experiment it can be concluded that, wear rate varies inversely with hardness of the hybrid nanocomposites. 4.4. Underwater shock The shock factor is a measure of the severity of the attack, and relates the charge weight to the distance between the point of ignition and the target plate; the shock wave pressure varies with the charge weight, the standoff distance, and the relative attack orientation. A higher value of the shock factor implies that a larger portion of energy of the underwater explosion is imparted to the ship. For submarines, this factor is called the Hull Shock Factor (HSF) Under water shock tests of 15gms of explosive charge is taken for the study. An explosive PEK-1 at 0.46 m was taken as the stand-off distance (D) from the laminate. The rapid rise in pressure at the start of the profile represents the front of the shock wave striking the FRP. In this case the maximum pressure of the shock wave was measured to be 31 MPa. Calculated Shock factor is 0.122, where standoff distance (D) is 0.46meters and Charge weight (W) is 0.015kgf. The post shock hybrid laminates are shown in Figure-8(a-c).

Figure-8: (a) Deformed laminate after underwater shock test, (b) 1wt% nanoclay reinforced hybrid nanocomposites after underwater shock test, (c) Pure FRP composite laminate after underwater shock test

16

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

For the experiment results it is concludes that nanofiller reinforced hybrid nanocomposites gives better impact shock to underwater explosives. Pure FRP laminate shows less post shock tensile strengths when compared with nanoclay and nanosilica reinforced hybrid nanocomposites. It can be concluded from the experiment that nano reinforced clay and silica up to 2wt% will give better results in underwater explosive conditions. Post underwater shock Tensile strength for hybrid nanocomposites are shown in Figure-9.

post underwater shock tensile test results

185
Tensile strength(MPa)

180 nanoclay 175 170 165 160 155 0 1 w t% nanoparticle 2 3 nanosilica

Figure-9: Post underwater shock Tensile strength for hybrid nanocomposites

Post underwater shock tensile strength increased by 3.9%, 10% and 4% over pure FRP for 1wt%, 2wt% and 3wt% of nanoclay reinforcement respectively. 4.5. Dynamic Mechanical Analysis Figure-10 shows that the glass transmission temperature (Tg) of epoxy-fiber composites occurs at 99.30C. On addition of nanoclay, Tg increases to 109.3 0C for 1wt% nanoclay, Tg value for 2 and 3wt% nanoclay is 129.20C and 108.80C, respectively. The increase in Tg up to 2wt% nanoclay is due to increase in crosslink density between organo ions in nanoclay and epoxy matrix or epoxy-fiber interfacial strength. For higher clay content (3wt% nanoclay), Tg decreases but is still above the Tg of epoxy-fiber composites. Increasing clay platelets further restricts the cross-linking of epoxy matrix. Glass transition temperature value corresponding to wt% reinforcement is measured and the improvement in Tg with 2wt% reinforced hybrid nanocomposites is observed to be 30.1%.
0.7 0.6 0.5
Loss factor
9.00E+04

pure
Storage modulus MPa

8.00E+04 7.00E+04 6.00E+04 5.00E+04 4.00E+04 3.00E+04 2.00E+04 1.00E+04 pure 1nc 2nc 3nc

1nc 2nc 3nc

0.4 0.3 0.2 0.1 0 -0.1 0

50

100 Tem perature C

0

150

200

0.00E+00 0 20 40 60 80 100 120 140 160 180 200 Tem peture 0C

Figure-10: (a) Loss factor Vs Temperature for clay hybrid nanocomposites, (b) Storage modulus Vs Temperature for clay hybrid nanocomposites

Figure-10(b) shows the DMA results of epoxy-fiber composites with nanoclay reinforcement. Epoxy fiber composites have the storage modulus of 61.9 GPa. Addition of 2wt% nanoclay has increased the storage modulus to 82.7 GPa. For higher clay content (3wt% nanoclay) storage modulus has decreased to 80.3 GPa. The decrease in modulus at higher clay content is due to segregation of particles, unexfoliated clay, etc. Glass transition temperature

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

17

showed an improving trend of 9%, 30%, and 10% over pure FRP for 1wt%, 2wt% and 3wt% respectively of nanoclay reinforcements. Glass transition temperature (Tg) is analysed for clay and silica reinforced hybrid nanocomposites as shown in Figure-11.
140

Glass transition temperature (Tg)

130 120 110 100 90 80 70 60 0 1 wt% nanoparticle 2 3

nanoclay nanosilica

Figure-11: Glass transition temperature vs. wt% nanoparticles

For nanoclay, Tg increases by 30% from 99.30C (for pure) to 129.20C (for 2wt %). For nanosilica, Tg increases by 19% from 99.30C (for pure) to 119.10C (for 1wt %) and falls gradually to 89.40C (for 3wt %). Comparison of predicted properties with experimental results in all the tested cases indicate that the error is less than 5.02% in the prediction using regression model and 1% for neural network model, validating both the models. Higher error in regression model predictions are obtained since some of the factors that contribute to wear rate like condition of the applied load, track distance, travel, etc. were not taken into account. However, since neural network was trained with the experimental data and the influence of all the factors is inherently present in the data, the model is devoid of above limitations. No generalization of the data is done in neural networks, as in case of regression model for finding best-fit curve. Neural network maps the inputs to the outputs in multi-dimensions and takes care of non-linearity present in the case of study leading to more accurate predictions. 5. Conclusion: The present work is an attempt to synthesizing new materials, especially inorganic compounds at the nanosize which have better engineering properties and hence are applicable in industries like aeronautical, marine, automobile etc,. Characterization by XRD and SEM of the nanoparticles used, has established the sizes to be 36 nm for HNT halloysite nanoclay and 66 nm for nanosilica. The following findings are reported. Damping ratio at first mode also improved monotonically with 7%, 19% and 20% over pure FRP at 1wt%, 2wt% and 3wt% respectively of nanoclay reinforcement. Hardness exhibited improvements of 4%, 11% and 10% over pure FRP at 1wt%, 2wt% and 3wt% of nanoclay respectively. Post underwater shock tensile strength also increased by 4%, 10% and 4% over pure FRP for 1wt%, 2wt% and 3wt% nanoclay respectively. As expected from the enhanced mechanical properties like strength, hardness etc., wear rate decreased by 14%, 82% and 33% over pure FRP at 1wt%, 2wt% and 3wt% nanoclay respectively Glass transmission temperature increased by 10%, 30% and 10% over pure FRP for 1wt%, 2wt% and 3wt% of nanoclay . Storage modulus also showed a similar trend with improvements of 17%, 34% and 20% over pure FRP for 1wt%, 2wt% and 3wt% nanoclay respectively. Reinforcing with nanosilica instead of nanoclay also produced results with similar trends excepting that most of the properties peaked at 1wt% nanosilica instead of 2wt% as in the case of nanoclay ANN and Regression models have predicted the results with maximum error of 1% and 5.02% respectively. Hence ANN is a highly suitable tool for predicting mechanical and tribological properties and can be used with reasonable confidence for predicting the behaviour of synthetic materials even before they are actually manufactured, thus saving considerable time, effort and money.

18

Lingaraju.D et al. / Procedia Engineering 10 (2011) 918

Acknowledgements The authors are grateful to the support of the Naval Science Technological Laboratories, India, for providing financial, experimental setup and guidelines. References
1. Joussein E, Petit S, Churchman J, Theng B, Righi D, Delvaux B, Halloysite clay minerals a review, Clay Miner, 40, 383426, 2005. 2. Shchukin DG, Sukhorukov GB, Price RR, Lvov YM, Halloysite nanotubes as biomimetic nanoreactors, 1, 510513, 2005 3. Ray SS, Okamoto M, Polymer/layered silicate nanocomposites: a review, Prog Polym Sci, 28,1539, 2003 4. Giannelis EP. Review of polymer layered silicate nanocomposites, Adv. Mater, 8, 29-35, 1996 5. Gang Zhou, Siva Movva, James Lee, Nanoclay and Long-Fiber-Reinforced Composites Based on Epoxy and Phenolic Resins, Journal of Applied Polymer Science, Vol. 108, 37203726, 2008 6. Mingxian Liu, Baochun Guo, Mingliang Du,Yanda Lei, Demin Jia, Natural inorganic nanotubes reinforced epoxy resin nanocomposites, J Polym Res 15:205212, 2008 7. Varada Rajulu L, Ganga Devi L, Rama Devi R, Effect of Orientation of Fibers and Coupling Agent on Tensile Properties of Phenolic/Silica Composites, Journal of Reinforced Plastics and Composites, Vol. 27, No. 6/2008 8. Kojima Y, Usuki A, Kawasumi M, Okada A, Y.Fukushima, Kurauchi T, Kamigaito O, Mechanical properties of nylon 6clay hybrid, J. Mater. Res. 8, 1185, 1993 9. Chandradass J, Ramesh kumar M, Velmurugan R, Effect of nanoclay addition on vibration properties of glass fiber reinforced vinyl ester composites, Materials letter, 61, 4385-4388, 2007 10. Velmurugan R, Jeyaprakash P, Balaganesan G, Damping study of hybrid nano composites by low velocity impact, Proceedings of International Conference on Aerospace Science and Technology, Bangalore, India, 26-28 June 2008 11. Mohan T.P, Ramesh Kumar M, Velmurugan R, Thermal stability and damping characters of high performance epoxy layered silicate nanocomposites, International Symposium of Research Students on Materials Science and Engineering, Chennai India, ISRS-200204 12. Suresha B, Chandramohan G, Sadananda Rao P. R, Sampathkumaran P, Seetharamu S., Influence of SiC Filler on Mechanical and Tribological Behavior of Glass Fabric Reinforced Epoxy Composite Systems, Journal of Reinforced Plastics and Composites; 26, 565, 2007 13. Jihua Gou, Scott OBriant, Hichang Gu, Gangbing Song, Damping augmentation of nano composites using carbon nano fiber paper, Journal of nano materials, 1-7, 2006 14. Williams J.G, Opportunities for composites in the Offshore Oil Industry National conference on the use of composite materials in Load-bearing marine structures. Vol.II conference Preceedings, p.91, National Academy Press, 1990 15. ASM Engineered materials handbook, vol. 1, composites, pp.45-48 (ASM International), 1993 16. Diveyev B, Smolskyy A, Sukhorolskyy M, Dynamic rigidity and loss factor prediction for composite layered panel, International Scientific Journal, Archives of Materials Science and Engineering, Volume 31, Issue 1, Pages 45-48, 2008 17. Kie, Xu, Polymer, 44 1009, 2004 18. Batra R.C, Hassan N.M, Response of fiber reinforced composites to underwater explosive loads Composites: Part B, 38, 448468, 2007 19. Mouritz A. The effect of underwater explosion shock loading on the flexural properties of GRP laminates Int J Impact Eng,18(2),129 39, 1996 20. Will MA, Franz T, Nurick GN, The effect of laminate stacking sequence of CFRP filament wound tubes subjected to projectile impact Compos Struct;58(2), 25970, 2002 21. Mouritz A.P, Saunders D.S, Buckley S, "The damage and failure of GRP laminates by underwater explosion shock loading, Composites, volume 25, number 6, 431-37, 1994 22. Velmurugan R, Mohan T.P Epoxy-Clay Nanocomposites and Hybrids: Synthesis and Characterization, Journal of Reinforced Plastics and Composites, 28; 17-37, 2009 23. Asifullah Khan, Mohtashim H. Shamsi, Tae-Sun Choi, Correlating dynamical mechanical properties with temperature and clay composition of polymer-clay nanocomposites. Computational Materials Science, 45, (2009) 257265 24. Venkatesan P, Anitha S, Application of a radial basis function neural network for diagnosis of diabetes mellitus, current science, vol. 91, no. 9, 10th november 2006
th

25. Lawrence M, Petterson A, BrainMaker Users Guide and Reference Manual, 7 Edition. California Scientific Software, Nevada City, CA 95959, 1998 26. Velten K, Reinicke R, Friedrich K, Wear volume prediction with artificial neural networks, Tribology International, Vol. 33 , pp 731 736, 2000. 27. Jones, Jansen, Fusaro, Preliminary investigation of neural network techniques to predict tribological properties, Tribology Transactions, Vol. 40(2), p 312, 1997 28. Genel K, Kurnaz S.C, Durmanb M, Modeling of tribological properties of alumina fiber reinforced zincaluminum composites using artificial neural network. Materials Science and Engineering, Vol. A363, pp 203210, 2003 29. Kishor S. Trivedi, "Probability and statistics with reliability, queuing and computer science applications", 2nd Edition, John Wiley and Sons Ltd., Chichester, UK, 2001 30. Mouritz A. The effect of underwater explosion shock loading on the fatigue behaviour of GRP laminates. Composites;26(1):39, 1995 31. Mouritz A. The damage to stitched GRP laminates by underwater explosion shock loading. Compos Sci Technol;55(4): 36574,1995 32. Keil AH, The response of ships to underwater explosion David Taylor Model Basin Report No. 1576, 1961