Effects of Rubber-Rich Domains and the Rubber-Plasticized Matrix on the Fracture Behavior of Liquid Rubber-Modified Araldite-F Epoxies.
The fracture behavior of a bisphenol A diglycidylether (DGEBA) epoxy, Araldite F, modified using carboxyl-terminated copolymer of butadiene and acrylonitrile (CTBN) rubber up to 30 wt%, is studied at various crosshead rates. Fracture toughness, [K.sub.IC], measured using compact tension (CT) specimens, is significantly improved by adding rubber to the pure epoxy. Dynamic mechanical analysis (DMA) was applied to analyze dissolution behavior of the epoxy resin and rubber, and their effects on the fracture toughness and toughening mechanisms of the modified epoxies were investigated. Scanning electron microscopy (SEM) observation and DMA results show that epoxy resides in rubber-rich domains and the structure of the rubber-rich domains changes with variation of the rubber content. Existence of an optimum rubber content for toughening the epoxy resin is ascribed to coherent contributions from the epoxy-residing dispersed rubber phase and the rubber-dissolved epoxy continuous phase. No rubber cavitation in the fra cture process is found, the absence of which is explained as a result of dissolution of the epoxy resin into the rubber phase domains, which has a negative effect on the improvement of fracture toughness of the materials. Plastic deformation banding at the front of precrack tip, formed as a result of stable crack propagation, is identified as the major toughening process.
Unmodified thermoset epoxy resins usually have poor impact resistance and fracture toughness, which limits their applications as structural materials. It is well known that addition of rubber to brittle epoxies can greatly improve the resistance of the materials to crack propagation . The toughening effect is generally accompanied by a relatively small loss of mechanical and thermomechanical properties.
Liquid rubber has been used over many years to improve the fracture toughness of brittle epoxies . Existence of rubber in rubber-epoxy blends is very complicated. Rubber can dissolve in an epoxy matrix as a plasticizer or flexibilizer, and/or separate from the matrix as a second phase during the curing process. The rubber, either dissolved in epoxy matrices or as a dispersed second phase, can significantly affect fracture properties of the blends. Many efforts have been made to understand the behavior of the dispersed rubber phase, which separates from the resin before gelation. Sultan and McGarry  found that toughening mechanisms are affected by the rubber particle sizes. A sub-micro rubber particle size favors shear deformation, whereas crazing is activated by large rubber particles (1.5-5 [micro] m). A simple quantitative model was proposed by Kunz-Douglass et al.  to describe rubber stretching and tearing mechanisms. This model can account for an increase of a factor of two in fracture toughness. Kinloch et al.  reported that rubber cavitation and plastic shear yielding in epoxy matrices are essential microdeformation mechanisms occurring at the crack tip, which dissipate energy and produce the toughening effect. Yee and Pearson [6-8] proposed a similar process, cavitation-induced shear deformation, in which the cavitation of rubber particles is a prerequisite for the localized shear yielding of the matrix. Rubber particles first enhance shear localization by acting as stress concentrators. Then, as bulk strain energy increases, cavitation takes over and dissipates the energy, making further shear localization possible. However, Huang and Kinloch [9-11] and Guild and Young  claimed that the role of cavitation is to initiate plastic dilatation within the matrix and that this process is independent of shear banding. Cavitation can occur either before or after shear banding . Some other minor toughening mechanisms, such as rubber particle bridging/crack deflection and rubber particle debonding, were also found . Concerning the conditions under which rubber particles cavitate, some researchers [14, 15] concluded that the rubber particle size is one of the main factors controlling rubber cavitation, while Sue and Garcia-Meitin  proposed that a triaxial stress state favors cavitation. It can be seen that a considerable debate still exists, focusing on the role of rubber cavitation and the formation conditions of rubber cavitation.Huang and Kinloch [9-11] and Guild and Young  claimed that the role of cavitation is to initiate plastic dilatation within the matrix and that this process is independent of shear banding. Cavitation can occur either before or after shear banding . Some other minor toughening mechanisms, such as rubber particle bridging/crack deflection and rubber particle debonding, were also found . Concerning the conditions under which rubber particles cavitate, some researchers [14, 15] concluded that the rubber particle size is one of the main factors controlling rubber cavitation, while Sue and Garcia-Meitin  proposed that a triaxial stress state favors cavitation. It can be seen that a considerable debate still exists, focusing on the role of rubber cavitation and the formation conditions of rubber cavitation.
Thermal behavior of a tetraglycidyldiaminodiphenylmethane (TGDDM)/CTBN blend, cured by diaminodiphenyl sulfone (DDS), was investigated by Kim et al. . It was observed that when bisphenol-A was added to the rubber/epoxy blend, [T.sub.g] of the epoxy-rich phase was reduced. Dissolution of the rubber in the epoxy-rich phase was assumed to be one of the reasons responsible for lower [T.sub.g]. However, [T.sub.g]'s did not change with the rubber content in the rubber/epoxy blends without bisphenol-A. In a study of rubber-modified epoxies of DGEBA/CTBN with BPA added to the resin formulation, Manzione et al. [18, 19] concluded that the dissolved rubber promotes plastic deformation and necking at low strain rates, providing a large increase in elongation at break. Dissolved rubber in an epoxy matrix can be expected to increase the matrix ductility and to reduce the level of stress at which shear bands initiate. Bussi and Ishida  studied fracture behavior of blends of DGEBA resin with hydroxyl-terminated, in ternally epoxidized polybutadiene rubber, and found that the improvement of fracture toughness of blends is due to the incorporation of rubber into the epoxy network, which causes plasticization, rather than to the presence of phase-separated rubber particles. It is usually difficult to assess the role of dissolved rubber owing to the contribution of cavitation of phase-separated rubber to the toughness. Although some work has been done in understanding the role dissolved rubber plays, it is still not very clear which part of rubber, dissolved or phase-separated, plays a major role in toughening the epoxies.
In this work, liquid CTBN rubber was used to toughen a standard DGEBA epoxy. The variation of fracture toughness with rubber content and cross-head rate was evaluated and the toughening mechanisms were investigated.
2. EXPERIMENTAL PROCEDURE
The epoxy resin used in this study is a diglycidyl ether of bisphenol A (DGEBA) resin, Araldite-F, produced by Ciba-Geigy, Australia. It is a high-viscosity standard base resin for general use. The curing agent is piperidine, mixed with the pure epoxy at a weight ratio of 5:100. CTBN 1800x13 liquid rubber is used as toughener, which contains 26% acrylonitrile and has a molecular weight of 3200. The epoxy-rubber mixture was cured at 120[degrees]C for 16 hours. The details for preparation of rubber-toughened epoxies can be found in a previously described procedure .
The tensile test specimens were machined from the cured resin plates with a thickness of 4.5 mm. Tensile properties were determined according to ASTM D638M-91a on an Instron 5567 testing machine. The strain was measured using a clip gauge with a gauge length of 50 mm, connected to a computer data acquisition system. At least five specimens were tested for each crosshead rate (0.5 mm/min to 500 mm/mm) or rubber content (up to 30 wt%). All specimens were polished using 1200 grade emery paper before testing.
CT specimens were used to evaluate fracture behavior of the modified epoxies. The width, W, is 50 mm, and thickness 12 mm. The precracks of specimens were introduced by first machining a 45[degrees] notch and then inserting a fresh razor blade by tapping . The sharp precrack induced by tapping a razor blade varies from 2 mm to 5 mm in length. At least three specimens were tested for each crosshead rate or rubber content. All tests were performed on the Instron 5567 testing machine. The data for the load and crack opening displacement were collected at a time interval of 2 ms for the crosshead rate of 500 mm/min. The fracture toughness, [K.sub.IC], was calculated according to ASTM D5045-93.
Fracture surfaces of compact tension specimens were examined using a C505 scanning electron microscope with an accelerating voltage of 20 kV. The surfaces were sputter-coated with gold to reduce any charge built up on the surfaces. Dynamic mechanical spectra were obtained at a frequency of 1 Hz on a DMA 2980 Dynamic Mechanical Analyzer (TA instruments). Dual cantilever clamps with a supporting span of 34.87 mm were used. Specimens were cooled to low temperatures by liquid nitrogen gas, and the spectra were scanned from -100[degrees]C to 180[degrees]C. The data were recorded at a sampling rate of 2 sec/point.
3. RESULTS AND DISCUSSION
3.1 Mechanical Properties
The typical stress-strain curves of the rubber-modified epoxies are shown in Fig. 1. All systems clearly illustrate a non-linear stress-strain response before final failure. As expected, the yield stress increases with increasing the crosshead rates. The failure elongation greatly increases after rubber was added to the pure epoxy. The failure elongation of the pure epoxy at a crosshead rate of 0.5 mm/min is about 7%, while it is increased to about 35% for the 30% rubber-modified epoxy at the same crosshead rate. The increase in resin ductility means an increase in capability of resin shear deformation, which results in improvement of fracture toughness if shear deformation plays a major role in the energy dissipation process during crack propagation. Visual inspection during and after the test on tested specimens shows all specimens fail through a necking process at the middle of the gauge length. A triaxial stress state, which is believed to be able to trigger rubber cavitation, exists in the necked region [16, 22], but SEM reveals that there is no rubber cavitation in the region. For a low rubber content, specimens show an intrinsic strain-softening effect, while for the 30% rubber-modified epoxy, a slight strain-hardening effect is observed at the crosshead rate of 0.5 mm/min. Yee and Person (6) saw an intrinsic strain-softening effect in Epon 828 (DGEBA)/CTBN 1300 X 13 systems with a rubber content up to 20%. (Epon is a product of the Shell Chemical Co.)
Young's modulus, E, and yield stress, [[sigma].sub.y], for the rubber-modified epoxies are presented in Fig. 2, respectively, as a function of crosshead rate. Assuming that the strain rate is proportional to the crosshead rate, the well-known effect of strain rate on the yield stress is observed in these neat and rubber-modified epoxies, well described by Eyring's law , ie, yield stress increases linearly with the logarithm of strain rate.
Figure 3 depicts Young's modulus and yield stress as a function of rubber content. Both modulus and yield stress decrease with rubber content. As predicted by Kerner's equation (24), a linear decrease of Young's modulus with volume fraction of dispersed rubber domains should be obtained, assuming a negligible modulus for the rubber. Yield stress was also found to linearly decrease with volume fraction of dispersed rubber domains . Therefore, it can be assumed that the deviation from a linear relationship in Fig. 3a indicates that volume fraction of dispersed rubber domains increases very slowly with added rubber at a low rubber content. The decrease in both the modulus and yield stress with rubber content was also reported by other investigators [6, 25, 26].
3.2 Fracture Toughness
The fracture toughness of the modified epoxies with different rubber contents is presented in Fig. 4. A significant improvement of fracture toughness, is obtained. [K.sub.IC] increases from 0.75 MPa.m1/2 to 1.48 [MPa.m.sup.1/2] when the rubber content is increased from 0% to 15%, and then it reduces to 1.07 [MPa.m.sup.1/2] with the further increase of rubber content to 30% at the crosshead rate of 0.5 mm/min. An optimum rubber content for the maximum [K.sub.IC] can be clearly seen. The crosshead rate has a discernible influence on the optimum rubber content, which exists around 15% at the crosshead rates of 0.5 and 5 mm/min, and shifts to about 20% at the higher crosshead rate. Yee and Person , Kinloch et al. , and Bascom et al  found that an optimum rubber content existed between 15% and 20% CTBN concentrations. Such behavior was seen as a result of the phase change of the elastomer from a dispersion phase to a blend at a high CTBN concentration. But the strain rate was found to have no clear influence on the o ptimum rubber content . The effect of crosshead rate on the fracture toughness, [K.sub.IC] is depicted in Fig. 5. The [K.sub.IC] generally decreases with increasing crosshead rate. However, the variation of [K.sub.IC] with crosshead rate is not linear. A slight reduction of [K.sub.IC] was observed for all rubber-toughened epoxy specimens when the crosshead rate was increased from 0.5 mm/min to 50 mm/min, and a remarkable drop is followed with further increasing the crosshead rate to 500 mm/min.
3.3 Morphology Observation
Visual inspection of fracture surfaces of CT specimens shows that there was no stress-whitened zone. The stress whitening is caused by scattering of visible light from a layer of scattering voids, i.e., cavitated rubber particles . The morphologies of rubber phase in the epoxy matrix were examined using scanning electron microscopy (SEM) and presented with the original rubber content in Fig. 6. For the 5% rubber-toughened epoxy, a diffuse morphology of rubber in epoxy matrix with tiny rubber-concentrated areas of an approximate size of 0.5 [micro]m (Fig. 6a) can be seen, without a clear phase boundary. The formation of microcrack-like regions on the surface cannot be explained at the moment, but this kind of microcrack-like region is not observed in other rubber-modified epoxies. Rubber domains in the modified epoxies with 10%, 15% and 20% rubber have sizes of 1 to 3 [micro]m in diameter, but the microstructure inside the rubber domains varies with the initial rubber content. Rubber domains in the modifie d epoxies with 15% rubber or more have a blending structure (Fig. 6c to 6d), consisting of the epoxy matrix phase. When the rubber content is increased to 25%, rubber domains prefer to form large clusters with a size of 8-12 [micro]m. The cluster contains a continuous epoxy phase connecting the scattered rubber particles. The size of rubber domains in the 30% rubber-modified epoxy dramatically increases and falls into a range of 200 to 300 [micro]m.
Plastic deformation bands due to stable crack propagation at the front of the precrack tip were observed in all rubber-modified epoxies. Figure 7 shows the effect of crosshead rate on morphologies of plastic deformation bands of the 5% rubber-modified epoxy. It is found that width of the plastic deformation band at the precrack tip is slightly dependent on the crosshead rate (25 [micro]m at 0.5 mm/min and 30 [micro]m at 500 mm/min, respectively). On the other hand, the epoxy in the plastic deformation band at the crosshead rate of 500 mm/min, compared with that at 0.5 mm/min, was less deformed, showing the more brittle fracture behavior. The effect of rubber content on the width of the plastic deformation band can be seen in more detail in Table 1. Epoxies modified with 5% and 30% rubber have an average plastic deformation band of 25 [micro]m and 125 [micro]m in width, respectively, at 0.5 mm/min. But the largest width of plastic deformation band is seen in the 15% rubber-modified epoxy (Fig. 8), being almos t 500 [micro]m. The variation of plastic deformation band width with rubber content, with a maximum for the 15% rubber-modified epoxy, is very similar to that of fracture toughness (Fig. 4), showing an optimum rubber content of around 15% at 0.5 mm/min. The formation of the plastic deformation band at the precrack tip is clearly correlated with the improvement of fracture toughness of rubber-modified epoxies. The plastic deformation of the epoxy in the band is believed to be the major energy dissipation process.
The morphologies of rubber domains of all rubber-modified epoxies either in the plastic deformation band at the precrack tip or ahead of the plastic deformation band are very similar (Fig. 6e). No rubber cavitation is evident, and no discernible trace of localized shear deformation around the rubber domains is identified. The fracture surface of the rubber domains in the plastic deformation band is relatively smooth and less plastically deformed, compared with that of the epoxy phase in the band. Without cavitation and discernible plastic deformation as well as significant promotion of localized shear deformation to the surrounding epoxy, the separated-rubber domains contribute little to the increase of fracture toughness of the modified epoxies. Increasing volume fraction of the dispersed rubber domains will reduce the area of the plastic deformation region of the epoxy in the plastic deformation band. As a result, the efficiency of rubber as a toughening agent to improve fracture toughness is reduced. The p henomenon that rubber domains do not cavitate was also found by Verchere et al.  in a system consisting of a DGEBA epoxy, cured by cycloaliphatic diamine, in the presence of an epoxy-terminated butadiene-acrylonitrile random copolymer. Dekkers et al.  observed similar behavior in the study of toughening mechanisms of blends with poly (butylene terephthalate) (PBT) and bisphenol-A polycarbonate (PC) toughened with core-shell impact modifier.
3.4 Dynamic Mechanical Spectra
Dynamic mechanical analysis (DMA) has been proved to be a useful tool to provide a qualitative understanding of the general behavior of multiphase polymer systems [24, 29-31]. It has been shown that a systematic analysis of dynamic data can provide a sound basis for examining the effects of crosslinking, plasticization, grafting procedure, and composition on the mechanical and fracture behavior of toughened polymers. Figure 9 shows the influence of rubber content on the storage modulus of the rubber-modified epoxies. The storage modulus decreases with increasing rubber content over the temperature range of -30 to 100[degrees]C, showing that the addition of the rubber clearly softens the epoxy matrix, which is consistent with the variation of elastic modulus, shown in Fig. 3a. Loss modulus and tan delta of the pure epoxy are presented in Fig. 10. Two damping peaks on each curve are observed. The peak at the low temperature is related to the [beta] relaxation of the epoxy, and the other at the high temperature is relevant to the [alpha] relaxation .
The influence of rubber content on the shape, position, and intensity of the low temperature relaxation loss modulus for the rubber-modified epoxies is given in Fig. 11. Besides a [beta] relaxation peak, a new peak appears on the high temperature side, which is related to the glass transition of the dispersed rubber phase in the epoxy matrix (named the rubber peak hereafter). It is noted that the rubber peak slightly shifts to higher temperatures with increasing initial rubber content. This is attributed to the epoxy dissolved in the separated rubber domains, which raises the onset of movement of rubber molecular segments to a higher temperature. The structural details of this rubber-rich domain are quite complicated and they change with the initial rubber content, shown in Fig. 6. Romanchik et al.  observed the presence of a core-shell morphology for the rubber-rich domain, with about 10% to 20% of the total volume being rubbery shell, while Person and Yee  found the presence of an epoxy phase inside the rubber-rich dispersed domains. The shifting of the rubber peak position observed here is in the direction opposite to that obtained by Manzione et al.  and Beck et al. , who explained the shifting in terms of the presence of interfacial tensile stresses induced by the difference in thermal expansion coefficients between the epoxy matrix and the rubber domains. It is also seen that the rubber peak broadens on the high temperature side, indicating the appearance of new and slower relaxation modes. This may be related to the movement of rubber-induced epoxy molecular segments. Finally, but most important, the intensity of the rubber peak changes with the initial rubber content. The intensities of the rubber peaks for the 5% and 10% rubber-modified epoxies are very weak, and the trace of rubber peak of the 5% rubber-modified epoxy, located around -44[degrees]C, can be identified only after the signals are magnified significantly. When the rubber content goes up to 15%, the intensity of the rubber pea k rapidly increases. Verchere et al.  reported an approximately linear increase of the magnitude of the rubber loss peak with the volume fraction of the dispersed rubber domains in a DGEBA/epoxy-terminated butadiene-acrylonitrile (ETBN) system. The nonlinear increase in intensity of the rubber peak with initial rubber content in Fig. 11 indicates that the increase of volume fraction of dispersed rubber domains is relatively slow at the low initial rubber content, and it rapidly speeds up after the initial rubber content is increased to 15%. The result is consistent with the measurement of Young's modulus (Fig. 3a), showing that the volume fraction of dispersed rubber domains at the low initial rubber content does not increase as much as at the high rubber content.
As discussed in the previous section, the improvement in fracture toughness of the rubber-modified epoxies largely originates from plastic deformation of the epoxy-rich phase in the plastic deformation band, as shown in Fig. 6e. On the other hand, the rubber domains reduce the volume of the plastic deformation region of the matrix in the plastic deformation band at the crack tip, which has a negative contribution to the improvement of fracture toughness. Considering the reduction of fracture toughness beyond the optimum rubber content (shown in Fig. 4] and significant increase in volume fraction of dispersed rubber domains at the 15% initial rubber content, one can expect that the rapid increase in volume fraction of dispersed rubber domains after the 15% initial rubber content is responsible for the reduction of the fracture toughness of the modified epoxies. The optimum rubber content for the present modified epoxies results from two opposite contributions to the toughness, i.e., one is a positive contribut ion from the dissolved rubber, the other is a negative contribution from the phase-separated rubber.
Figure 12 presents DMA spectra of epoxy-rich continuous phases with the initial rubber content. It is clearly seen that the glass transition temperature, defined by the position of the [alpha] relaxation peak (hereafter called the epoxy peak), of the modified epoxies shifts to lower temperatures with increasing rubber content, indicating that some of the added rubber is dissolved in the epoxy domain. The peaks also broaden on the low temperature side, showing the appearance of new and faster relaxation modes, which are probably due to the epoxy residing in the dispersed rubber domains relaxing in the lower temperature region of the [alpha] relaxation. This relaxation may be due to the covalently bonded CTBN blocks or to a lower degree of crosslinking density because of a nonstoichiometric segregation of epoxy and amine functionalities from the matrix. 
Figure 13 gives the variation of glass transition temperature, [T.sub.g], of the rubber-rich phase, epoxy-rich phase and core-shell-rubber (CSR)-modified epoxies  with the initial rubber content, obtained from the positions of the rubber peak, epoxy peak and the corresponding [alpha] relaxation peak of CSR-modified epoxies, respectively. The glass transition temperature of the rubber-rich phase clearly increases with the initial rubber content, while that of the epoxy-rich phase decreases. [T.sub.g] of the rubber-rich phase is raised by about 21[degrees]C for the 30% rubber-modified epoxy. However, the [T.sub.g] of the CSR-modified epoxy remains almost unchanged, which means that the shifting of the a relaxation peak of the rubber-modified epoxies in the present study is caused by the dissolved rubber in the epoxy. [T.sub.g] can be a good indicator of the amount of the dissolved phases. The weight percent of rubber dissolved in the epoxy phase can be quantitatively estimated by Gordon-Taylor equation .
[w.sub.R] = [T.sub.gE(0)] - [T.sub.gE(r)]/[k([T.sub.gE(r)] - [T.sub.gR(0)] + ([T.sub.gE(0)] - [T.sub.gE(r)]] (1)
where [w.sub.R] is the weight fraction of the rubber dissolved in the epoxy phase. [T.sub.gE(0)] and [T.sub.gR(0)] are the glass transition temperatures of the unplasticized epoxy and the pure rubber, respectively, and [T.sub.gE(r)] is the glass transition temperature of the epoxy-rich phases. k is a constant and is set to be 2 . The glass transition temperature, [T.sub.gR(0)], of CTBN 1800 X 13 rubber is 215 K, obtained from literature .
The weight fraction of epoxy dissolved in the rubber can also be estimated in a similar manner using the above equation. The results are presented in Fig. 14. The content of rubber that dissolves into the epoxy increases with the initial rubber content, and is roughly proportional to the initial rubber content. The linear relationship between the dissolved rubber and added rubber was also found by several other investigators [24, 35, 38]. It is interesting to note that the content of epoxy dissolved in rubber domains is twice as much as that of rubber dissolved into the epoxy matrix. This large dissolubility of epoxy into rubber domains seems to be responsible for the suppression of rubber cavitation. On the one hand, the critical stress to trigger rubber cavitation increases owing to the dissolution of the epoxy in the rubber domains. However, the stress to induce plastic deformation of the epoxy matrix is significantly reduced owing to the existence of rubber in the epoxy. As a result, plastic deformation b anding proceeds at the front of the crack tip without rubber cavitation. The model developed by Bucknall et al.  examined effects of size and properties of rubber particles on the cavitation. It was pointed out that a small change in shear modulus of rubber particles has a dramatical influence on the capability of rubber particles to cavitate, and the increase of shear modulus of rubber particles can suppress cavitation. This is consistent with the present experimental results since the shear modulus of rubber domains can be greatly increased by the dissolution of the epoxy, considering the fact that shear modulus of the epoxy (about 1.2 GPa) is five orders of magnitude larger than that of the rubber phase (0.4 MPa).
The mechanical properties of liquid rubber-toughened epoxies are found to be strongly dependent on the rubber concentration and crosshead rates. The yield stress decreases with increasing rubber content but it increases with increasing crosshead rate. Both yield stress and Young's modulus show a linear variation with logarithm of the crosshead rate. The deviation of Young's modulus from the linear relationship with added rubber content at low initial rubber concentration is ascribed to the fact that at the very low initial rubber content, rubber mainly dissolved into the epoxy matrix and the rubber phase separation process was largely suppressed in this stage.
From SEM observation and the nonlinear change of the intensity of the rubber-rich phase on the DMA spectra, it is concluded that volume fraction of phase-separated rubber domains dramatically increases after the 15% rubber content. However, the phase-separated rubber domains, containing a fraction of dissolved epoxy, have a negative effect on the improvement of fracture toughness of the materials. The glass transition temperature of the epoxy-rich phase decrease with addition of rubber to the pure epoxy, showing dissolution of rubber into the epoxy matrix.
Fracture toughness of rubber-modified epoxy changes with the initial rubber content. An optimum rubber concentration exists at around 15% at a crosshead rate of 0.5 mm/min, and it slightly shifts to about 20% at a crosshead rate of 500 mm/min. The optimum rubber content for the fracture toughness is attributed to the mechanisms associated with the facts of epoxy residing in the rubber phase and rubber dissolving in the epoxy continuous phase.
Rubber cavitation, as an energy dissipation process and trigger of localized shear deformation of epoxy matrix around the rubber particles, cannot be found in the present study, while plastic deformation banding at the precrack tip, formed as a result of the stable crack propagation, is the major toughening mechanism. Rubber dissolved in the epoxy matrix plays a critical role for the improvement of fracture toughness in the rubber-modified epoxies by increasing the capability of plastic deformation of the epoxy matrix. Absence of rubber particle cavitation is ascribed to dissolution of epoxy into the rubber phase, which dramatically raises the shear modulus of the rubber-rich phase, resulting in an increase of resistance of rubber particles to cavitation.
The authors wish to express their gratitude to the Electron Microscope Unit of Sydney University for the access to its facilities. Keqin Xiao is supported by an Overseas Postgraduate Research Scholarship (OPRS). a University Postgraduate Research Award (UPRA) and a Supplementary Scholarship from the Department of Mechanical and Mechatronic Engineering, Sydney University.
(*.) To whom correspondence should be addressed.
(1.) C. K. Riew, E. H. Rowe, and A. R. Siebert, Adv. Chem. Ser., 154, 326 (1976).
(2.) A. R. Siebert, Adv. Chem. Ser., 208, 179 (1984).
(3.) J. N. Sultan and R. C. McGarry, Polym. Eng. Sci, 13, 29 (1973).
(4.) S. Kunz-Douglass, P. W. R. Beaumont, and M. F. Ashby, J. Mater. Sci, 15, 1109 (1980).
(5.) A. J. Kinloch, S. J. Shaw, D. A. Tad, and D. L. Hunston, Polymer, 24, 1341 (1983).
(6.) A. F. Yee and R. A. Pearson, J. Mater. Sci, 21, 2462 (1986).
(7.) R. A. Pearson and A. F. Yee, J. Mater. Sci, 21, 2475 (1986).
(8.) A. F. Yee, D. Li. and X. Li, J. Mater. Sci., 28, 6392 (1993).
(9.) Y. Huang and A. J. Kinloch, J. Mater. Sci Lett, 11, 484 (1992).
(10.) Y. Huang and A. J. Kinloch, J. Mater. Sci, 27, 2763 (1992).
(11.) Y. Huang and A. J. Kinloch, Polymer, 33, 5338 (1992).
(12.) F. J. Guild and R. J. Young, J. Mater. Sci, 24, 2454 (1989).
(13.) A. C. Garg and Yiu-Wing Mat, Composites Science and Technology, 31, 179 (1988).
(14.) A. Lazzeri and C. B. Bucknall, J. Mater. Sci., 28, 6799 (1993).
(15.) P. Beahan, A. Thomas, and M. Bevis, J. Mater. Sci., 11, 1207 (1976).
(16.) H. J. Sue and E. I. Garcia-Meitin, J. Appl Polym. Sci Part B: Poly. Phys., 31, 595 (1993).
(17.) W. N. Kim, C. E. Park, and C. M. Burns, J. Appl Polym. Sci, 50, 1951 (1993).
(18.) L. T. Manzione, J. K. Gillham, and C. A. McPherson, J. Appl Polym. Sci, 26, 889 (1981).
(19.) L. T. Manzione, J. K. Gillham, and C. A. McPherson, J. Appl Polym. Sci, 26, 907 (1981).
(20.) P. Bussi and H. Ishida, J. Appl Polym. Sci., 53, 441 (1994).
(21.) K. Xiao, Y, S. Kwok, and L. Ye, J. Mater. Sci, 33, 2831 (1998).
(22.) D. Li, A. F. Yee, I. W. Chen, S. C. Chang, and K. Takahashi, J. Mater. Sci, 29, 2205 (1994).
(23.) C. B. Bucknall, Toughened Plastics, Applied Science, London (1977).
(24.) P. Bussi and H. Ishida, J. Polym. Sci., Part B: Poly. Phys., 32, 647 (1994).
(25.) D. Verchere, J. P. Pascault, H. Sautereau, S. M. Moschiar, C. C. Riccardi, and R J. J. Williams, J. Appl Polym. Sci, 43, 293 (1991),
(26.) A. Tcharkhtchi, S. Faivre, L. E. Roy, J. P. Trotignon, and J. Verdu, J. Mater. Sci, 31, 2687 (1996).
(27.) W. D. Bascom, R. L. Cottington, R. L. Jones, and P. Peyser, J. Appl Polym. Sci, 19, 2545 (1975).
(28.) M. E. J. Dekkers, S. Y. Hobbs, and V. H. Watkins, J. Mater. Sci, 23, 1225 (1988).
(29.) G. Cigna, J. Appl Polym. Sci, 14, 1781 (1970).
(30.) H. Keskkula, S. G. Turley, and R. F. Bayer, J. Appl Pal yin. Sci, 15, 351 (1971).
(31.) C. B. Bucknall, P. Davis, and I. K. Partridge, J. Mater. Sci, 21, 307 (1986).
(32.) W. A. Romanchik, J. E. Sohn, and J. F. Geibel, Adv. Chem. Ser., 221, 85 (1983),
(33.) L. T. Manzione, J. K. Gillham, and C. A. McPherson, J. Appl. Polym. Sci, 26, 889 (1981).
(34.) R. H. Beck, S. Gratch, S. Newman, and K. C. Rusch, Polym. Lett., 6, 707 (1968).
(35.) D. Verchere, J. P. Pascault, H. Sautereau, S. M. Moschiar, C. C. Riccardi, and R. J. J. Williams, J. Appl Polym. Sci, 42, 701 (1991).
(36.) K. Xiao, "Fracture Toughness of Rubber-Modified Epoxies and Their Carbon Fibre Composites," PhD thesis, University of Sydney, NSW 2006, Australia (2000).
(37.) P. Bussi and H. Ishida, Polymer, 35, 956 (1994).
(38.) D. Verchere, J. P. Pascault, H. Sautereau, S. M. Moschiar, C. C. Riccardi, and H. J. J. Williams, J. Appl. Polym. Sci, 41, 467 (1990).
(39.) C. B. Bucknall, A. Karpodinis, and X. C. Zhang, J. Mater. Sci, 29, 3377 (1994).
|Printer friendly Cite/link Email Feedback|
|Author:||XIAO, KEQIN; YE, LIN|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 2000|
|Previous Article:||Long Chain Branching and Polydispersity Effects on the Rheological Properties of Polyethylenes.|
|Next Article:||Preparation and Characterization of Biaxially Oriented Films From Polybutylene Terephthalate Based Thermoplastic Elastomer Block Copolymers.|