PST 522E – Synthesis and Characterization of Macromolecules

CHAPTER 4
DYNAMIC MECHANICAL ANALYSIS OF EPOXY-CARBON FIBER COMPOSITES

1. THEORETICAL PART Composite materials have been used widely for daily life applications for many years. Polymer composite systems have a large scale for industry or research area due to their light weight, design flexibility, and processability [1,2]. Using polymer composite is widespread to use for special engineering materials such as aerospace industry, automotive and civil engineering structures because of their outstanding mechanical properties. Polymer composite theories were established on properties of composite constituents, volume fraction, shape, matrix-inclusion interface [3,4,1,2]. Mechanical properties of polymer composites are significant when they use for a building or an instrument which has critical points about temperature and strength properties [5-7]. Composite materials properties have been developed such as chemical, physical and mechanical, so their application area comes much bigger. Because of improvements on composites materials, their industry are growing on automotive, aerospace, electronics and biotechnology [8,9]. The important point of using nanocomposites is large scale, low cost, easy applicable [10]. In composites, two materials are mixed to make improvement in the mechanical properties. Matrix material is dominant and including material. The second material is particle, filler or fiber. Composites are manmade materials. Materials have properties such as strength (compressive, tensile, flexural…), toughness (impact resistance), stiffness (modulus of elasticity), wearing resistance, thermal insulation, fatigue life. These properties of materials can be improved by composites. 1.1. Carbon Fiber Carbon fiber (graphite fiber) is one of the popular materials that used preparing composites. The synthetic carbon industry starts with the foundation of National Carbon Company in Ohio, and continues Union Carbide Corp. High performance carbon fibers were developed at the Parma Technical Center by Dr. Roger Bacon in 1958. The atomic structure of carbon fiber is similar to graphite. Carbon fibers are classified by the tensile modulus of the fiber; low modulus, standard modulus, intermediate modulus, high modulus, ultrahigh modulus. Polyacrylonitrile is one of the raw materials that used to manufacture carbon fiber. Fibers are used for plastic industry because of their considerable properties such as hardness, elastic modulus, mechanical strength, impact strength and dimensional control [11]. Fiber reinforced polymer matrix materials in which nano and micro-scale particles have been studied because of their potential and capacity of performance for twenty years [12]. Due to their high specific and stiffness properties, carbon fiber reinforced composites are

- 25 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

widely used structural materials for aerospace, marine, armor, automobile, railways, civil engineering structures, sport goods industry [13]. Carbon fibers are used for many different industries also carbon fiber composites. For carbon fiber composites, matrix has an important role. Carbon fiber is used for many polymer matrix composites. Carbon fiber and resin matrix interface is very important for determining properties of composites such as toughness or environmental endurance [14]. Nature of interface, bonding properties and surface morphology can change the composite fracture toughness [15]. Characterization of matrix resin is related with fiber, compression and thickness properties for composite polymer matrix [16]. 1.2. Epoxy-Carbon Fiber Composites The first epoxy products were synthesized in 1891. In the 1930’s, Pierre Castan (from Switzerland) and Sylvan Greenlee (in USA) patented their works at the same time, independently. The first commercial products marketed in the 1940’s, and it was result of product was bisphenol A and epichlorhydrine reaction. One of the epoxy components is hardener, epoxy resins depend on the reactivity of hardeners [17]. Epoxy resins are cured with different curing agents like amines, amides, acids, acid anhydrides and amine adducts. Cycloaliphatic curing agents better properties for applications of epoxy area such as weather ability, low blush and water spotting, and chemical resistance. By using suitable epoxy resin and hardener we can obtain materials which have significant mechanical, physical and chemical properties in order to use in a widely usage area. Because of resin chemistry and available applications, epoxy resins have been available for many major applications such as surface coating, tooling, civil engineering, molding compounds. Epoxy resins has widely used in industry for sixty years such as aircraft applications, the car industry, reinforcement to various concrete structures also thin film coating, electronic circuits [18]. Epoxy resin is mostly used as a matrix in material engineering due to their high stiffness, resistance of creep and chemical, also good adhesion [19]. Epoxy resins are mostly used with fiber to obtain reinforced composites. An epoxy resin with fiber reinforced has advantages such as good stiffness, specific strength, stability, chemical resistance, and show good adhesion to the fiber [20]. Due to their strength, high modulus and light weight, epoxy-carbon fiber composites can be used for aircraft industries.

Figure 1: Different types of fiber composites

- 26 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

1.3. What is DMA? Dynamic Mechanical Analysis, otherwise known as DMA, is a technique where a small deformation is applied to a sample in a cyclic manner. This allows the materials response to stress, temperature, frequency and other values to be studied. The term is also used to refer to the analyzer that performs the test. DMA is also called DMTA for Dynamic Mechanical Thermal Analysis. Dynamic Mechanical Analysis has become more popular because of their significant properties and to provide information about materials in particular polymers. The first experiments to measure elasticity of materials could be done by Poynting in 1909 [21]. In the 1950’s, the Weissenberg Rheogoniometer and the Rheovibron instruments were invented for usage of commercially [22]. J. Starita, C. Macosko and Bohlin developed a commercial dynamic mechanical analyzer in the 1970’s. DMA gives the information about rheological and thermal properties of polymers. Rheology is very sensitive to small changes of the material’s polymer structure – thus ideal for characterization of polymers. The rheology structure relationship is the key to the development of new materials. Dynamic Mechanical Analysis (DMA) measures the mechanical properties of materials as function of temperature, frequency and time and also it is a thermal analytical method by which the mechanical response of a sample subjected to a specific temperature program is investigated under periodic stress. Dynamic mechanical analyzer is a thermal analytical instrument used to test the mechanical properties of many different materials. The Dynamic Mechanical Analysis is a high precision technique for measuring the viscoelastic properties of materials. Viscoelasticity is about elastic behaviors of material. Most real-world materials exhibit mechanical responses that are a mixture of viscous and elastic behavior. Dynamic Mechanical Analyzer (DMA) deforms a sample mechanically and after that it measures the sample response. When a force is applied on a material it suffers a change in shape, that is, it deforms. The deformation can be applied sinusoidally, in a constant (or step) fashion, or under a fixed rate. The response to the deformation can be monitored as a function of temperature or time. A force to resist the deformation is also set up simultaneously within the material and it increases as the deformation continues. If the material is unable to put up full resistance to external action, the process of deformation continues until failure takes place. The deformation of a body under external action and resistance to deform are referred to by strain and stress respectively. Polymers are viscoelastic fluids, which behave viscous or elastic, depending on how fast they flow or are deformed in the process. For instance, glass transition temperature and damping behavior can be used to determinate material’s using conditions such as temperature, stiffness. In addition, DMA measurements explain how a material behaves at the moment and future. Dynamic Mechanical Analyzer is useful for these tests: mechanical properties, morphology of polymers, loss factor (Tan delta), loss angle (delta), impact resistance, dynamic viscosity, curing kinetics, correlation with materials formulation, ageing, damping, glass transition temperature (Tg), industrial products stiffness, rheological properties, secondary transitions,

- 27 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

specimen stiffness, young Modulus, thermal stability, creep behavior, tension test, stressstrain, vs. There are seven types of clamps in DMA. Each of clamps is used with suitable materials. Single cantilever bending clamp, dual cantilever bending clamp, 3-point bending clamp, tension film clamp, tension fiber clamp, compression and shear clamp are used for DMA measurements. Each of clamps can be used for suitable materials.

Figure 2: Clamp of Film Tension for DMA 1.4. How does DMA differ from Thermomechanical Analysis? Thermomechanical Analysis, or TMA, applies a constant static force to a material and watches the material change as temperature or time varies. It reports dimensional changes. On the other hand, DMA applies an oscillatory force at a set frequency to the sample and reports changes in stiffness and damping. DMA data is used to obtain modulus information while TMA gives coefficient of thermal expansion, or CTE. Both detect transitions, but DMA is much more sensitive. Some TMAs can do limited DMA and the PerkinElmer DMA 8000 is the only DMA that can do TMA. 1.5. How does a DMA work? DMA works by applying a sinusoidal deformation to a sample of known geometry. The sample can be subjected by a controlled stress or a controlled strain. For a known stress, the sample will then deform a certain amount. In DMA this is done sinusoidally. How much it deforms is related to its stiffness. A force motor is used to generate the sinusoidal wave and this is transmitted to the sample via a drive shaft. One concern has always been the compliance of this drive shaft and the affect of any stabilizing bearing to hold it in position. A schematic of the analytic train of the DMA 8000, Figure 3, shows its innovative design that requires neither springs nor air-bearings to support the drive shaft.

Figure 3: Schematic of the DMA 8000 analytic train

- 28 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

1.6. What does DMA measure? DMA measures stiffness and damping, these are reported as modulus and tan delta. Because we are applying a sinusoidal force, we can express the modulus as an in-phase component, the storage modulus, and an out of phase component, the loss modulus, see Figure 4. The storage modulus, either E’ or G’, is the measure of the sample’s elastic behavior. The ratio of the loss to the storage is the tan delta and is often called damping. It is a measure of the energy dissipation of a material.

Figure 4: The relationship of the applied sinusoidal stress to strain is shown, with the resultant phase lag and deformation. 1.7. How does the storage modulus in a DMA run compare to Young’s modulus? While Young’s modulus, which is calculated from the slope of the initial part of a stress-strain curve, is similar conceptually to the storage modulus, they are not the same. Just as shear, bulk and compressive moduli for a material will differ, Young’s modulus will not have the same value as the storage modulus. 1.8. What is damping? Damping is the dissipation of energy in a material under cyclic load. It is a measure of how well a material can get rid of energy and is reported as the tangent of the phase angle. It tells us how good a material will be at absorbing energy. It varies with the state of the material, its temperature, and with the frequency. 1.9. Why would I want to scan modulus as a function of temperature? Modulus values change with temperature and transitions in materials can be seen as changes in the E’ or tan delta curves. This includes not only the glass transition and the melt, but also other transitions that occur in the glassy or rubbery plateau, shown in Figure 3.

- 29 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

These transitions indicate subtler changes in the material. The DMA 8000’s unique low starting temperature of -190 °C in the standard furnace and of -196 °C in the Fluid Bath let you easily look for these small molecular motions, Figure 5.

Figure 5: Modulus values change with temperature and transitions in materials can be seen as changes in the E’ or tan delta values. 1.10. How do I get good data? Good data requires several things: a properly calibrated instrument, a properly prepared specimen with a reasonable aspect ratio, using the right geometry, and applying both reasonable strains and heating rates. A properly calibrated instrument requires calibration for both temperature and force. A well prepared specimen should be of even thickness with parallel sides and right angle. Assuming the correct choice of geometry for the sample, a deformation of 50 microns and heating rates of 2-3 °C/minute normally work fine. 1.11. How do I know what geometry to use? The choice of the geometry you run your sample in is dictated by the sample’s physical state at the beginning of the experiment, its difficulty in loading, and the experiment you want to run. For example, a stiff bar of polymer can be run in all of the flexure fixtures, but single cantilever is often used because it is simple to load and allows thermal expansion of the specimen, shown in Table 1. Uncured thermosets are often run in shear. The DMA 8000 not only has the a full range of fixtures covering the normal 3-point bending, single cantilever, dual cantilever, tension, compression and shear fixtures, but also offers the novel Material Pocket for holding powders and soft samples that can not support their own weight. In addition, the design and flexible software make it possible to develop custom fixtures for your application.

- 30 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

Table 1: Preferred geometry for indicated sample size

1.12. How can I detect a Tg? The glass transition (Tg) is seen as a large drop (a decade or more) in the storage modulus when viewed on a log scale against a linear temperature scale, shown in Figure 6. A concurrent peak in the tan delta is also seen. The value reported as the Tg varies with industry with the onset of the E’ drop, the peak of the tan delta, and the peak of the E’ curve being the most commonly used.

Figure 6: The glass transition (Tg) in the storage modulus and tan delta

- 31 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

1.13. How do I know it’s really a Tg? Running a multi-frequency scan and calculating the activation energy of the transition allows you to decide if the transition is really a Tg. The activation energy for a Tg is roughly 300-400 kJmol-1. In comparison a Tb has an activation energy of about 30-50 kJmol-1 and at the melt or Tm, the frequency dependency collapses. 1.14. Why does my Tg value sometimes not agree with my DSC value? That’s actually not surprising. The glass transition is really a range of behavior where scientists have agreed to accept a single temperature as the indicator per certain standards. Different industries have used different points from the same data set that can vary as much as 15 °C. DSC, TMA, and DMA measure different processes and therefore, the numbers vary a bit. You can see as much as a 25 degree difference in data from a DSC to DMA data reported as peak of tan delta. See Figure 6 for an example. 1.15. Can I do TMA in my DMA? It depends on what you are looking for. Most tests like flexure, penetration, creep or a simple stress-strain run can be done. In the past, most dynamic mechanical analyzers have not been able to generate coefficient of thermal expansion (CTE) data, but the DMA 8000 can run TMA type experiments and obtain excellent CTE values for a wide range of samples run in extension. CTE tells you how your material will expand as a function of temperature. This information is vital for products where dissimilar materials will be heated together (for example motors and circuit boards) as well as curing systems where contraction on curing occurs. 1.16. Can I use DMA to study curing? DMA is commonly used to study curing of materials as this process involves a dramatic increase in the modulus values. It is commonly used to get both the point of gelation and the point of vitrification for thermosetting materials. Cures can be studied with temperature ramps and isothermally at a fixed temperature. The DMA 8000 can be configured with optional quartz windows and special fixtures to allow the study of photo-curing systems.

Figure 7: DMA 8000 with special fixture to allow the study of photo-curing systems

- 32 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

1.17. Why should I be concerned about frequency scans and multiple frequency runs? Most materials can see many frequencies in their final product. An example is the rubber used in a windshield wiper which sees a range of operating frequencies and temperatures in use. Modulus-frequency plots can tell you how your material will change as frequency changes. For viscous materials, this can give useful information about its flow. It is often advisable to not just look at modulus-frequency at one temperature, but to scan many frequencies as you heat a material. This allows you to see how transitions shift under the influence of frequency. For example, in some polymers a shift from 1 to 100 hertz will move a Tg by 14 degrees, which could cause a material to fail if the high frequency is not considered in its design. 1.18. What does Time-Temperature Superposition (TTS) tell me? The Williams-Landel-Ferry model, or WLF, says that under certain conditions, time and temperature can be mathematically interchanged. A TTS, shown in Figure 8, lets you use data collected as frequency scans at a range of temperatures to predict behavior at frequencies that are not directly measurable. The DMA 8000 has advanced software that makes this a fairly simple process. The data is often converted to time to predict lifetime performance. One should note that TTS calculations rest on some assumptions and is often invalid if these assumptions aren’t met. One basic assumption is a single relaxation time and is tested by using a wicket or Cole-Cole plot.

Figure 8: Time Temperature Superposition 1.19. How can I tell if a TTS is valid? You can tell if a TTS is valid by using the DMA 8000’s software to generate a Cole-Cole or wicket plot. Plotting E’ against either E’ or tan delta should give a nice half circle plot if the assumptions of the William Landel Ferry model are met. If they aren’t, then the material is not rheologically simple and a WLF superposition will fail.

- 33 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

1.20. Why would I want to run my samples immersed in a fluid? Certain solvents can cause a material to soften while they are exposed to them. Others can react with and harden the material. Both of these effects can cause failure in biomedical devices, coating, paints and gaskets to name only a few. In addition, the effect of a stress and a solvent often is different than just soaking a material in a solvent and then testing it in air. The DMA 8000’s Fluid Bath allows collecting data of sample immersed in solutions under a variety of conditions.

Figure 9: DMA 8000 with fluid bath 1.21. Can I use DMA to see if humidity affects my sample? Humidity is known to have tremendous effects on the properties of materials from polymers to papers to natural products. The DMA 8000 has the option of being configured with an integrated humidity generator that allows precise control of the humidity in the furnace. This permits accurate and precise studies on how humidity affects the properties of your materials. 1.22. Is UV curing an important application for my DMA if I have a photo-calorimeter? Yes, a photo-calorimeter only looks at the energy of photo-curing. Photo-curing in the DMA lets you see how the physical properties change and when the modulus of the curing material has reached acceptable limits in terms of strength and stiffness. This information is important for cost effective design of your cure cycle. Using a DMA with UV also allows you to investigate the degradation of materials and so evaluate additive packages, formulations, etc. 1.23. DMA Modulus Parameters 1.23.1 The Modulus Modulus is an intrinsic material property (does not change with material size or shape), defined as the ratio of stress/strain in a body under a particular mode of deformation (such as

- 34 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

shear, bending, torsion, etc.). Thus modulus is a measure of materials overall resistance to deformation. The modulus is the ratio of a component of stress to a component of strain, in rheology. G = Stress/Strain 1.23.2. The Elastic (Storage) Modulus The storage modulus is measure of elasticity of material. It is also called “the ability of the material to store energy”. It is equivalent to the ability of a sample to store energy, i.e. its elasticity. Energy storage occurs as molecules are distorted from their equilibrium position by application of a stress. Removal of the stress results in a return to equilibrium position of the molecular segments. G' = (stress/strain)cosδ 1.23.3. The Viscous (loss) Modulus Loss modulus represents the capability of a material to dissipate energy (mechanical, acoustic) as heat, owing to viscous motions inside the material itself. It is limited to the molecular motion within the sample that dissipates energy as heat. In rheology, loss modulus is the imaginary part of the complex modulus. G" = (stress/strain)sinδ 1.23.4. Tan Delta It is measure of material damping - such as vibration or sound damping. Damping refers to damping the loss of mechanical energy as the amplitude of motion gradually decreases. It means also the ability of a material to dissipate mechanical energy by converting it into heat. Tan Delta is a useful index of material viscoelasticity since it is a ratio of viscous and elastic moduli. Tanδ is an important indication of viscoelasticity of materials, it is independent from the shape and dimension of samples and it is dimensional. Tan δ= Loss Modulus/Storage Modulus=G"/G 1.24. Stress- Strain Curves Stress expresses an applied force or system of forces that tends to strain or deform a body, strain means that a deformation produced by stress. Strain is defined the deformation from a specified reference state, measured as the ratio of the deformation to the total value of the dimension in which the change occurs. In another way, it defines the measurement of deformation, relative to a reference configuration of length, area, or volume. Strain is also called relative deformation. The stress-strain curve characterizes the behavior of the material tested. It is most often plotted using engineering stress and strain measures, because the reference length and cross-sectional area are easily measured. Stress-strain curves generated from tensile test results help gain insight into the constitutive relationship between stress and strain for a particular material. The maximum stress level on the stress-strain curve corresponds to the strength of material while the maximum strain is defined as ultimate strain. If the slope s steep, the sample has a

- 35 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

high tensile modulus and it resists deformation. If the slope is gentle, the sample has a low tensile modulus and it is easily deformed. Various regions and points in a stress-strain curve are shown in Figure 10.

Figure 10: Various regions and points of stress strain curve.

2. EXPERIMENTAL Mechanical properties were investigated for epoxy and epoxy-carbon fiber composites. The values of modulus and stress-strain curves were compared for epoxy and epoxy-carbon fiber composites. 2.1. Materials and Apparatus Epoxy resin and hardener were used to prepare composites. Epoxy resin and hardener was taken from Tekno Construction Chemicals Co.(Istanbul, Turkey). Its commercial names are Teknobond 300 A and Teknobond 300 B. Epoxy resin is a product of bisphenol A(epichlorhydrin)%Concentration:76.00-88.00 and 1,6-hexanedioldiglycidyl ether%Concentration:14.00-22.00. It is obtained that epoxy resin’s number average molecular weight has lower from 700. Hardener is a product of aliphatic and cycloaliphatic polyamines. Carbon fibers were used to prepare composite materials from epoxy. Carbon fibers were polyacrylonitrile based carbon fiber. Carbon fibers (TORAYCA T700 12K, Warp = 12.000, Warp = 7 micron, tex (g/1000m) = 880)

- 36 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

2.2. Preparation Two of samples were prepared. First one is “neat” epoxy and the second one is carbon fiberepoxy composites. Firstly, epoxy A and B components (hardener) were mixed for 10 minutes in a 50 ml flask. Then, epoxy resin casted on the glass substrate and were cured in 100 oC heated owen for one hour. Secondly, carbon fibers were cut into small species with a simple scissor and added to homogenous mixtures as % 0.5 as weight. Then the continuous mixing was employed 20 minutes. After that, carbon fiber containing resins casted on the glass substrate and the prepared composite resins were cured in 100 oC heated owen for one hour. DMA was used to measure modulus values of the epoxy and epoxy-carbon fiber composites. 2.3. Procedure 2.3.1. Procedure # 1 The material was heated from +30 0C to +130 0C with a heating rate 5 0C/min for epoxy and epoxy-carbon fiber composites. The applied frequency was 1 Hz. Materials were tested using tensile film clamp. The relationship with temperature and modulus also stress strain curves at different temperatures were also studied, after five days. 2.3.2. Procedure # 2 The ambient temperature at 30 oC and 18 N maximum loads was applied with a ramp force rate of 1 N/min, after 5 days. Their stress-strain curve at 30 oC and their mechanical properties such as maximum stress, maximum strain, and young modulus were investigated. (Stress-strain curve)

3. REFERENCES [1] Park, J.H. and Jana, S.C., (2003). “The Relationship Between Nano- and MicroStructures and Mechanical Properties in PMMA-Epoxy-Nanoclay Composites”, Polymer, 44, 2091-2100. Wu, C.L., Zhang, M.Q., Rong, M.Z. and Friedrich, K., (2002). “Tensile Performance Improvement of Low Nanoparticles Filled-Polypropylene Composites”, Composites Science and Technology, 62, 1327-1340. Jordan, J., Sharaf, M., Jacob, K., Tannenbaum, R. and Jasiuk, I., “Experimental Trends in Polymer nanocomposites - A Review”, Georgia Institute of Technology, p. 1-29. Gersappe, D., (2002). “Molecular Mechanisms of Failure Nanocomposites.”, Physical Review Letters, 89(5), 058301-1-4. in Polymer

[2]

[3]

[4]

[5]

Kontou, E. and Kallimanis, A., (2006). “Thermo-visco-plastic behaviour of fibrereinforced polymer composites,” Composites Science and Technology, 66, 588–1596.

- 37 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

[6]

Al-Haik, M.S., Garmestani H and Savran A., (2004). “Explicit and implicit viscoplastic models for polymeric composites.”, Int J Plasticity, 20, 1875–907. Megnis, M. and Varna, J., (2003). “Micromechanics based modeling of nonlinear viscoplastic response of unidirectional composite.”, Compos Sci Technol, 63, 19–31. Evora, V.M.F. and Shukla, A., (2003). Mater. Sci. Eng. A361, 358–366. Iwahori, Y., Ishiwata, S., Sumizawa, T. and Ishikawa, T., (2005). Compos. A 36.

[7]

[8] [9]

[10] Zhang, M.Q., Rong, M.Z. and Friedrich, K., (2005). Application of non-layered nanoparticlesin polymer modification, in: K. Friedrich, S. Fakirov, Z. Zhang (Eds.), Polymer Composites: From Nano- to Macro-scale, Springer, New York, pp. 25–44. [11] Saliba, C.C., Orefice, R.L., Carneiro, J.R.G., Duarte, A.K., Schneider, W.T. and Fernandes, M.R.F., (2005). “Effect of the incorporation of a novel natural inorganic short fiber on the properties of polyurethane composites”, Polymer Testing, 24, 819– 824 [12] Chowdhury, F. H., Hosur, M. V. and Jeelani, S., (2007). “Investigations on the thermal and flexural properties of plain weave carbon/epoxy-nanoclay composites by hand-layup technique”, J Mater Sci, 42, 2690–2700. [13] Zhou, Y., Pervin, F., Rangari, K., Jeelani, S., (2007). “Influence of montmorillonite clay on the thermal and mechanical properties of conventional carbon fiber reinforced composites”, Journal of Materials Processing Technology , 191, 347–351. [14] Xu, Z., Huang, Y., Zhang, C. and Chen, G., (2007). “Influence of rare earth treatment on interfacial properties of carbon fiber/epoxy composites”, Materials Science and Engineering A , 444, 70–177. [15] Kim, JK. and Mai, Y-W., (1991). “High strength, high fracture toughness fibre, composites with interface control—a review.”, Compos Sci Technol , 41, 333–78. [16] Xu, Z., Pervin, F., Lewis, L. and Jeelani, S., (2007). Experimental study on the thermal and mechanical properties of multi-walled carbon nanotube-reinforced epoxy, Materials Science and Engineering A, 452–453, 657–664. [17] Edited by B. Ellis, (1993). Chemistry and technology of Epoxy Resins; Blackie Academic&Professional, p 1- 4. [18] Edited by B. Ellis, (1993). Chemistry and technology of Epoxy Resins;; Blackie Academic&Professional, p 203, 252-253. [19] Gonçalez, V., Barcia, FL. and Soares, BG., (2006). Composite Materials Based on Modified Epoxy Resin and Carbon Fiber;; J. Braz. Chem. Soc., Vol.17, No. 6, 11171123 [20] Donnet, J.B., (2003). Compos. Sci. Technol. 63, 1085–1088.

- 38 -

Chapter 4 – Dynamic Mechanical Analysis of Epoxy-Carbon Fiber Composites

[21] Poynting, J. H., (1990). Proceedings of the Royal Society, Series A, 82, 546. [22] Dealy, J., (1992). Rheometers for Molten Plastics, Van Nostrand Reinhold, New York, p136–137and 234–236.

- 39 -