Morphology and properties of an epoxy alloy system containing thermoplastics and a reactive rubber.
Thermosetting epoxies exhibit many desirable properties, such as high tensile strength and modulus, excellent chemical and solvent resistance, dimensional and thermal stability, good creep resistance, and excellent fatigue properties. These characteristics make them ideal candidates as matrices for high-performance carbon fiber reinforced composites. However, the epoxy resins are also generally brittle because of high crosslink densities.
One conventional way to toughen thermosetting epoxy matrices is by incorporating liquid reactive rubbers. The most common reactive rubber modifiers are based on butadiene acrylonitrile copolymers terminated with carboxyl (-COOH) or amine (-N[H.sub.2]) end groups, which are called CTBNs and ATBNs, respectively. Toughening epoxy matrices using liquid reactive rubbers has been widely reported (1-5). However, toughness improvements in most rubber-modified thermosetting systems usually result in a significant decrease in [T.sub.g]s of the cured thermosetting resins. The [T.sub.g] depression of the modified epoxy matrices becomes severe especially when the reactive rubbers are used at levels high enough to noticeably enhance the fracture toughness.
More recent studies have demonstrated that polymeric thermoplastics, such as polysulfones, polyethersulfones, polyetherimides, and polyimides, can enhance fracture toughness without sacrificing [T.sub.g]s or other desirable properties of thermosetting resin systems (6-10). Thermoplastics incorporation into the epoxy matrices, however, significantly increases the viscosities after blending, and decreases the solvent resistances of the modified epoxy networks after cure. At typical loading of thermoplastics in the epoxy resins, improvement in the fracture toughness is usually moderate. More recently, some studies (11-14) claimed that using oligomeric thermoplastics as toughness modifiers for epoxies provided processing advantages over high-molecular-weight thermoplastics as toughness modifiers in that the viscosity of the modified resin systems is lower, and thus the modified systems are easier for prepregging or other processing necessary in manufacturing fiber composites from the modified epoxy resin systems. Additionally, reactive oligomeric thermoplastics can promote good bonding between the precipitated phase and the continuous phase domains. Thermoplastics such as polysulfone or polyethersulfone with reactive functional groups have been the recent focus of toughness improvement for epoxy (15, 16). Substantial improvements have also been reported for a carbon fiber composite system based on a polysulfone-modified bismaleimide resin (17). Microwave radiation has been applied to enhance the rate of network formation in a poly (arylene ether sulfone)-modified epoxy and bismaleimide system (18). Although it has long been recognized that phase-separated morphology is an essential means of achieving fracture toughness improvement, effectiveness in enhancing the fracture toughness is not the same for all systems. Successful applications rarely result from utilizing only one of the above approaches by itself.
The objective of this study was to explore a novel toughness improvement approach by incorporating both thermoplastics and liquid reactive rubbers into epoxy systems. A mechanism of toughness improvement was proposed and explained in detail for this new toughening approach. The methodology of how to effectively utilize the combination of dual toughness modifiers and to minimize compromises in processability or other properties of the epoxy resins is discussed.
The epoxy used was based on diglycidyl ether of bisphenol-A (DGEBA) epoxy resin (Shell Epon Resin 828), which was actually a mixture of the DGEBA monomer and its polyhydroxyl ether oligomer with the same epoxy functional groups as those in DGEBA. The average degree of polymerization of DGEBA in this epoxy resin was between 1 and 2. A commercial-grade curing agent, 4,4[prime]-diaminodiphenyl sulfone (DDS, Ciba-Geigy HT 976), was used at 33 to 40 phr for curing the DGEBA epoxy resin. The thermoplastic used was a commercial-grade polysulfone (PSu, Amoco Chemicals Udel P1700). The polysulfone molecules were randomly terminated with chloro- and phenol groups. The PSu was used at 20 parts per hundred (20 phr) of the epoxy resin. The liquid reactive rubber used was a carboxyl (-COOH) terminated butadiene acrylonitrile with an acrylonitrile content of 18% (BFGoodrich Co., Hycar CTBN 1300X8). The liquid reactive rubber was used at a low level of 5 phr. Figure 1 shows the chemical structures of the epoxy resin, curing agent, polysulfone, and liquid reactive rubber used in this study. For simplicity, Fig. 1 shows only the chemical structure for the DGEBA epoxy resin in the monomeric form.
The mixing procedures are as follows. The epoxy resin (100 g) was first heated in an oil bath to 150 to 170 [degrees] C. Subsequently, a predetermined quantity (20 parts per hundred parts of epoxy) of polysulfone powder was added slowly into the heated epoxy resin. The mixture was stirred continuously for about 30 to 60 min with a mechanical stirrer until all polysulfone powder was dissolved and a homogeneous mixture with the epoxy resin was formed. After the polysulfone was dissolved, the viscosity of the mixture increased substantially because of incorporation of the high-molecular-weight thermoplastic PSu. Subsequently, the temperature of the oil bath was lowered from 160 [degrees] C to 135 [degrees] C, and the curing agent (DDS in powder form) was slowly added into the resin-PSu mixture with continuous stirring. The mixture was then stirred at a high speed again until all DDS was dissolved in the mixture. The rubber was added and the mixture was stirred again until a homogeneous mixture was obtained. The mixed resin was then thoroughly degassed in a vacuum oven at a temperature of 100 to 130 [degrees] C for 20 to 30 min until the mixture was visually bubble-free and clearly transparent. The properly blended and degassed resin mixture was then cast into a mold (glass or steel treated with a release agent) and cured with a standard cure cycle at 177 [degrees] C for 2 h. Typically, optimal properties for carbon fiber epoxy composites, especially for structural applications in the aero-space industry, require no post-curing at higher temperatures in order to maintain a balance in [T.sub.g], modulus, and fracture toughness.
Differential Scanning Calorimetry
Differential scanning calorimetry (DuPont DSC-910 with TA 2000) was used to measure the extent of residual curing reactions, and also to determine the glass transition temperatures of the cured matrix systems. The heating rate used was 5 [degrees] C/min from the ambient temperature to 300 [degrees] C in all DSC tests.
Dynamic Mechanical Analysis
Dynamic mechanical analysis was performed by using a DuPont DMA 983 coupled to a DuPont TA 2000 computer system for data acquisition and analysis. The phase behavior as related to morphological structure was investigated by determining the relaxation transitions of the thermoplastic-modified epoxy matrices. Dynamic mechanical analysis was used to measure the storage and loss moduli and transition temperatures. The dimensions of the samples for DMA were approximately 5.0 x 1.5 x 0.30 cm. The storage and loss moduli measured by the DMA 983 were the fiexural moduli, E[prime] and E[double prime], respectively. The heating rate was 5 [degrees]d C/min from -- 100 to 300 [degrees] C, and a fixed frequency of 1 Hz was used.
The morphology of the cured resins was examined using a scanning electron microscope (ISI mini-SEM). The thermoplastic-modified epoxy samples were fractured at ambient temperature, while those samples containing the liquid reactive rubber were fractured at the liquid nitrogen temperature to ensure brittle failure. The fractured surfaces of the samples were coated with gold by vapor deposition using a vacuum sputterer.
Fracture Toughness Measurements
Fracture properties of the unmodified and modified epoxies were determined by using compact tension (CT) specimens according to the specifications of ASTM E399-83. The dimensions of the specimens were approximately 3.0 x 3.0 x 0.35 cm. The specimens were first notched by using a diamond saw blade. A starter crack was then initiated using a sharp razor blade. The specimens were tested utilizing a universal mechanical testing machine (Instron) at a constant crosshead speed. The reported values of fracture toughness were averages from the results on three to five specimens. The critical stress intensity factor for Mode I crack was calculated using the following expression (19):
[K.sub.Ic]=[P.sub.c]Y [a.sup.1/2]/(BW) (1)
Y = 29.6 - 185.5 (a/w) + 655.7 [(a/w).sup.2] - 1017 [(a/w).sup.3] + 638.9 (a/[w.sup.4]),
P = load at crack initiation,
B = thickness of specimen (0.35),
w = width, and
a = crack length.
The values of critical stress intensity factor could be converted to corresponding critical strain energy, [G.sub.Ic], by the following well-established relation for the plane-strain conditions:
[Mathematical Expression Omitted] (2)
where E is the modulus and [Nu] is the Poisson ratio (= 0.35) of the epoxy.
RESULTS AND DISCUSSION
Figure 2 shows the glass transition temperatures, as revealed by the DSC traces, of four cured epoxy resins: unmodified (A-0-0), 5 phr CTBN-modified (A-0-5), 20 phr polysulfone-modified (A-20-0), and 20 phr polysulfone/5 phr CTBN-modified (A-20-5). [T.sub.g]s of the basic and 5 phr CTBN-modified epoxies were 185 [degrees] C and 190 [degrees] C, respectively. At the low concentration level of 5 phr CTBN in the epoxy, [T.sub.g] depression by the rubber modifier was found to be negligible. As a matter of fact, the [T.sub.g] of the liquid rubber modified epoxy may be slightly higher because of the catalytic effect of the -OH group on epoxy curing. For sample A-20-0 (the epoxy modified with 20 phr of polysulfone), the [T.sub.g] of the sample was found to decrease by approximately 20 [degrees] C to about 165 [degrees] C. This [T.sub.g] decrease might be caused by a dilution of the DDS concentration resulting from the presence of the thermoplastic modifier, rather than by plasticization of the epoxy network by the thermoplastic component. To compensate for the dilution of the curing agent concentration as a result of incorporation of the thermoplastic and rubber modifiers, a higher concentration of the curing agent could be used. For the A-20-5 sample (the epoxy modified with 20 phr PSu and 5 phr CTBN), the [T.sub.g] (175 [degrees] C) of the modified epoxy was raised back to the [T.sub.g] of the unmodified epoxy (185 [degrees] C) by the use of a higher concentration of DDS (37 phr instead of 33 phr).
The network structures of the basic and modified epoxy samples were examined using the dynamic mechanical analysis technique. Figure 3 shows the DMA results of the cured resin A-0-5 (the formulation containing 5 phr CTBN as the sole modifier). The storage modulus (E[prime]) of the modified epoxy sample exhibited two relaxation peaks. The higher-temperature relaxation peak at 190 [degrees] C is apparently the [T.sub.g] of the epoxy phase, which is at a slightly lower temperature than the [T.sub.g] of the unmodified epoxy (200 [degrees] C). The broad low-temperature relaxation from -- 100 [degrees] C to 50 [degrees] C is associated with the [Beta] relaxation of the epoxy possibly in superposition with the glass transition of the rubber component. Since the [T.sub.g] of the rubber (- 60 [degrees] C to - 50 [degrees] C) superimposes with the quite broad [Beta]-relaxation peaks of the epoxy component (- 70 [degrees] C), it is difficult to elucidate the phase behavior of the rubber component in the cured epoxy network from the DMA result. Direct observation of morphology using SEM is discussed later.
Figure 4 shows the DMA results of sample A-20-5, i.e., the modified epoxy containing both 20 phr PSu and 5 phr CTBN as the modifiers. Again, the broad low-temperature relaxation peak at - 80 [degrees] C may actually be a superposition of the [Beta]-relaxation of the epoxy component and the [T.sub.g] of the rubber component. These two peaks, however, could not be resolved under the experimental conditions used. Interestingly, the high-temperature [T.sub.g] relaxation was observed also as a single sharp peak at 190 [degrees] C. This might seem contrary to the observation of an opaque appearance of the cured resin plaque, which suggested a heterogeneous morphology in the modified matrix resin. However, since the [T.sub.g] of the polysulfone (195 [degrees] C) is close to the [T.sub.g] of the epoxy phase (190 [degrees] C), it was possible that the single relaxation peak was actually two overlapped relaxation peaks, one being associated with the [T.sub.g] of PSu and the other associated with the [T.sub.g] of DGEBA epoxy.
To confirm that the two relaxation peaks actually overlapped each other in the observed single relaxation peak, attempts were made to move the epoxy [T.sub.g] to a higher temperature by replacing part of the di-functional DGEBA epoxy with a tetra-functional epoxy. A sample was prepared in which the epoxy consisted of 75% of the DGEBA and 25% of a tetraglycidyl 4,4[prime]-diaminodiphenyl methane (TGDDM) epoxy. This new DGEBA/TGDDM mixed epoxy resin was then modified with 20 phr PSu and 5 phr of CTBN and was cured with the same concentration of DDS (37 phr), using exactly the same curing conditions as those used for sample A-20-5. This sample was designated C-20-5.
Figure 5 shows the DMA results of the cured 75% DGEBA/25% TGDDM/20 PSu/5 CTBN resin (sample C-20-5). Though not completely resolved, two [T.sub.g] relaxation peaks may be discerned in the loss modulus (E[double prime]) curve, with one at 190 [degrees] C (the PSu phase peak) as the shoulder peak of another major peak at 220 [degrees] C (the epoxy phase peak). This clearly demonstrates that the polysulfone indeed formed a phase-separated domain from the epoxy phase.
Figure 6 shows the SEM micrograph of the 5 phr CTBN-modified epoxy (sample A-0-5). The 5 phr CTBN-modified epoxy network exhibited a small number of dispersed spherical rubber particles 1 [[micro]meter] in diameter surrounded by the continuous epoxy phase domain. The DSC results showed that the [T.sub.g] of the rubber-modified epoxy was lower than that of the unmodified epoxy. This fact suggests that some of the liquid reactive rubber might remain dissolved in the epoxy and plasticize the epoxy network. Only a fraction of the added CTBN was completely precipitated out as discrete particles after cure.
Figure 7 shows the drastically different morphology of the 20 phr PSu modified epoxy, which is more complicated than the morphology of the rubber-modified epoxy shown in Fig. 6. The morphology of the PSu-modified epoxy, as revealed by the SEM, showed that the fracture surface consisted of a continuous phase domain (the darker area) surrounding discrete "island" domains with aggregated spherical particles imbedded in the islands (the brighter areas with spherical particles). The discrete islands were quite large (50 to 100 [[micro]meter]) and exhibited irregular shapes. The darker area that surrounded the islands was the continuous phase and might be identified as the epoxy phase domain since it showed brittle fracture characteristic of the epoxy. The discrete brighter areas (the islands), by contrast, exhibited ductile shear yielding fracture characteristic of the PSu thermoplastic and were obviously polysulfone domains. However, there might be ambiguity in identifying the constituent nature of the spherical particles imbedded in the thermoplastic domains. Since there were only two components in this epoxy/thermoplastic matrix system, the precipitated spherical particles in the thermoplastic islands were clearly of epoxy nature. Therefore, these discrete PSu islands with spherical particles were a clearly phase-inverted domain in which the originally dissolved epoxy had been precipitated as spherical particles and were surrounded by the continuous thermoplastic PSu domain.
The above results suggested that the thermoplastic PSu-modified epoxy exhibited a complex "phase-in-phase" morphology. The majority of the epoxy component formed the main continuous phase in which the thermoplastic component along with a minor portion of the epoxy component was precipitated altogether to form islands of irregular shapes. The islands were the domains of a phase-inverted morphology in which a small portion of the epoxy formed discrete particles surrounded by the thermoplastic PSu component as the continuous phase.
A careful examination of the SEM micrograph of the PSu-modified epoxy sample revealed that there might be adhesion problems in the interfaces between the discrete PSu domains (the islands) and the continuous epoxy phase. Partial debonding was noticed between the phase domains in the fractured surface. Additionally. inside the islands, the same statements are also true for the adhesive bonding between the discrete spherical epoxy particles and the continuous thermoplastic phase domain that surrounded the particles. Good adhesive bonding between the thermoplastic and thermosetting phases did not exist owing to the fact that the molecules of the thermoplastic PSu component contained no reactive functional groups to react with the epoxy molecules.
Figures 8a and b show the SEM micrographs, at 1000 X and 3000 X, respectively, of the epoxy modified with both PSu and rubber (sample A-20-5). Although the morphology of this sample seemed somewhat similar to that of the epoxy modified with only PSu (sample A-20-0), a careful examination of the micrographs showed that the interfacial bonding between the thermoplastic PSu and thermosetting epoxy phases was much improved. Therefore, addition of 5 phr reactive liquid rubber to the PSu-modified epoxy might not result in significant changes in the size or size distribution of the particulate morphology; however, it could significantly improve the fracture toughness by promoting the interfacial bonding between the thermoplastic and thermosetting phase domains in the modified epoxy.
The thermoplastic/rubber-modified epoxy apparently showed a complex phase-in-phase morphology, with a continuous epoxy phase surrounding a discrete thermoplastic/epoxy phase domain. The discrete thermoplastic/epoxy phase domain exhibited a phase inverted morphology consisting of a continuous thermoplastic and dispersed epoxy particles. The reactive rubber played an interesting role in enhancing the interfacial adhesive bonding between the thermoplastic PSu and thermosetting epoxy phase domains. The rubber modifier served a purpose of linking the discrete phase domains with the surrounding epoxy continuous phase. Because of a closer match of the solubility parameters, the rubber molecules would have a favorable interaction with the PSu phase. Meanwhile, since the reactive functional groups (-COOH) in the carboxyl-terminated rubber (CTBN) were also capable of reacting with the expoxide functional groups in the epoxy continuous phase that surrounded the discrete PSu island domains, therefore, a coupling mechanism existed, with the main chains of the rubber molecules interacting with the PSu phase by physical bonds, and the end groups linking by chemical bonds with the epoxy phase.
This "coupling" mechanism of the liquid reactive rubber is analogous to the familiar coupling mechanism with which a coupling agent (usually silanes or organofunctional silicon compounds) is commonly used to improve the glass fiber/matrix interfacial strengths through physical interactions and chemical bonds (20). The importance of coupling agents in fiber/matrix composites is that the stress transfer between fibers and matrix in a composite is signficantly improved. Furthermore, in systems where the interfacial bonding is not properly improved, moisture or other solvents that may diffuse through the matrix resin can accumulate at the interfaces and cause significant premature deterioration in composites. Similarly, Yee, et al. (21), have attempted to add a styrene-maleic anhydride copolymer to a polyphenylene oxide/(DGEBA-peperidine) system, and have discovered the copolymer served as an emulsifying agent to improve the morphology of the modified epoxy.
For comparison, the same rubber/thermoplastic-modified epoxy sample was fractured at the ambient temperature instead of the liquid nitrogen temperature, and SEM characterization was performed. Figures 9a and b show the SEM micrographs, at 1000 X and 2500 X, respectively, of the ambient-temperature fractured sample A-20-5. Apparently, compared to the thermoplastic-modified epoxy sample (A-20-0), the shear yielding in the island domains was more pronounced for this rubber/thermoplastic-modified epoxy (A-20-5).
Figures 10a and b show the micrographs, at 1000 X and 3000 X, respectively, of the rubber and PSu-modified epoxy samples (A-20-5). Both samples had been etched with the tetrahydrofuran (THF) solvent prior to the SEM characterization. The micrographs show that the continuous phase in the islands is now absent, but the imbedded particles remain. Apparently, the continuous phase in the islands was of a thermoplastic nature and was more easily etched away by the THF solvent. Obviously, the spherical particles in the islands were of a thermosetting nature since they could not be dissolved by the THF solvent etching. Thus, these SEM results of the THF-etched epoxy samples confirmed that the continuous phase in the islands were of the thermoplastic nature while the spherical particles dispersed and confined in the islands were of the thermosetting epoxy nature. It should be pointed out that the whole islands were surrounded by a continuous epoxy phase domain. Therefore, the phase domains of the modified epoxy network actually exhibited a complex phase-in-phase morphology.
To identify the phase domain where the rubber component was located, transmission electron microscopy (TEM) on the stained specimens was attempted. Figure 11 is the TEM micrograph (12,000 X) of sample A-20-5, which had been stained with aqueous solution of Ru[O.sub.4]. Since Ru[O.sub.4] interacted selectively only with the unsaturated double bonds of the rubber molecules, the dark area in the TEM micrograph was identified with the domains where the rubber component was present. The TEM micrograph shows that the dark area coincided with where the thermoplastic PSu constituted the continuous phase within the precipitated island domains, which in turn were the dispersed phase domain surrounded by the continuous epoxy phase. This evidence supported the earlier explanation that the rubber component preferentially interacted with the PSu, and acted as a link between the epoxy and PSu phases to enhance the bonding between the continuous epoxy phase domain and the discrete phase-inverted islands. Similarly, inside the islands, which consisted of a continuous thermoplastic PSu and discrete thermosetting epoxy particles, the bonding between the continuous PSu and the discrete epoxy particles was also improved by the same "coupling" mechanism provided through the liquid reactive rubber.
Figure 12 shows the stress intensity factor, [K.sub.q], as a function of modifier composition as represented by the four epoxy formulations: A-0-0, A-0-5, A-20-0, and A-20-5. Note that [K.sub.q] values are close or equal to [K.sub.Ic] values if the plane strain condition in the fracture surfaces is met in the tests. Examination of the fracture surfaces of the samples suggested that the plane-strain condition was reasonably satisfied. Thus, for practical purposes, the values of [K.sub.q] were used for [K.sub.Ic.] The unmodified epoxy resin (A-0-0) had a [K.sub.Ic] of 0.8 MPa.[m.sup.1/2] or a [G.sub.Ic] of 230 J/[m.sup.2]. The incorporation of 5 phr CTBN by itself into the epoxy matrix-modified (A-0-5) did not show any noticeable improvement in the fracture toughness. This was probably because 5 phr CTBN by itself in the epoxy had only a minimum effect on the morphology. Incorporation of 20 phr of PSu into the epoxy resin (sample A-20-0) moderately improved the fracture energy ([G.sub.Ic]) by approximately 50% to 350 J/[m.sup.2]. The improvement in [G.sub.Ic] is obviously related to multiphase morphology in the PSu-modified epoxy matrix system. However, with 5 phr CTBN rubber in addition to 20 phr polysulfone as modifiers, the modified epoxy resin (A-20-5) exhibited a [G.sub.Ic] of 700 J/[m.sup.2]. This was an impressive 300% improvement of [G.sub.Ic] over the unmodified epoxy resin sample (A-0-0).
The islands are actually a composite phase (continuous PSu and discrete epoxy particles) within a phase (continuous epoxy phase). For the PSu-epoxy systems investigated in this study, the islands were formed in-situ during cure and therefore were not of regular shape. In practical applications, if the precipitated epoxy-PSu composite phase domain (the islands) could be controlled to more regular spherical shapes, improvement of fracture toughness could be even more impressively improved. This suggests that, to achieve better spherical shapes, composite particles of thermoplastic-thermoset phases would have to be preferably formed separately in a more easily controlled environment, such as in a low-viscosity liquid medium, before being added as toughening modifiers into epoxy resin systems.
Modification of the epoxy with the combination of thermoplastic and rubber resulted in a complex phase-in-phase morphology after cure, where a precipitated thermoplastic-epoxy composite phase was surrounded by a continuous epoxy phase. By incorporating a low concentration (5 phr) of CTBN and 20 phr thermoplastic PSu, the fracture toughness of the cured resin showed an impressive increase of 300% in [G.sub.Ic] over the base unmodified DGEBA/DDS epoxy resin. Incorporation of 5 phr of CTBN in the formulation made the epoxy network to be more effectively toughened by the thermoplastic polysulfone by improving the interfacial characteristics of the epoxy and PSu phase domains. The liquid reactive rubber acted as as "coupling" between the epoxy and PSu phase domains in much the same way as silane coupling agents in enhancing the interfacial strengths of glass fiber/matrix composites. Since the liquid reactive rubber was used at a low concentration, compromise of the hot-wet properties of the modified epoxy matrices was kept to minimum.
The thermoplastic-modified epoxy exhibited an interesting morphology, where the PSu component phase-separated and precipitated in large domains (islands) surrounded by the continuous epoxy phase. Interestingly, inside the PSu domains (the islands), a phase-inverted morphology existed, i.e., the epoxy precipitated in discontinuous domains as spherical particles, which were surrounded by a continuous PSu phase domain. The fracture toughness of the modified matrices was found to increase more dramatically when the epoxy was modified by incorporating the liquid reactive rubber in conjunction with the thermoplastic. The liquid reactive rubber was found to act as a coupling link between the separated PSu and epoxy phase domains, and thus further improved the toughness of PSu-modified epoxy matrices.
In summary, this study has identified a novel methodology to effectively toughen an epoxy matrix resin by incorporating simultaneously a moderate quantity of thermoplastic and a small quantity of a liquid reactive rubber. The combination of dual modifiers has resulted in maximum improvement of fracture toughness with minimal compromises in processability and [T.sub.g] depression by rubbers. In practical applications, if the epoxy-PSu composite phase domain is controlled to spherical particles, the fracture toughness can be more impressively improved.
The authors express their appreciation to Dr. K. Riew of BFGoodrich for helpful discussions and for coordinating some of the morphology characterization at BFGoodrich Co., and to C. Pederson and B. Hayes of the Polymeric Composites Laboratory for technical assistance. Financial support for this work was provided through project support to the Polymeric Composites Laboratory by the Shell Companies Foundation and Boeing Commercial Airplane Group of the Boeing Company. One of the authors (E. M. Woo) acknowledges financial support (NSC-82-0113-E006-395-T) provided by the National Science Council of the Republic of China to continue follow-up studies in completing this work.
1. C. B. Bucknall and T. Yoshii, Brit. Polym. J., 10, 53 (1978).
2. S.C. Kunz, J. A. Sayre, and R. A. Assink, Polymer, 23, 1897 (1982).
3. R. S. Bauer, Proc. 18th Intl. SAMPE Tech. Conf., 510 (1986).
4. R. A. Pearson and A. F. Yee, J. Mater. Sci., 21, 2475 (1986).
5. J. Diamant and R. J. Moulton, Proc. 29th Nat. SAMPE Symp., 422 (1984).
6. C. B. Bucknall and I. K. Partridge, Polymer, 24, 639 (1983).
7. C. B. Bucknall and A. H. Gilbert, Polymer, 30, 213 (1989).
8. C. L. Loechelt, E. M. Woo, and J. C. Seferis, SPE ANTEC Tech. Papers, 34, 1645 (1988).
9. E. M. Woo, C. L. Pederson, J.C. Seferis, and L. D. Bravenec, Proc. ICCM-IIV, p. 358, Guanzhou, China (1989).
10. E. M. Woo, D. A. Shimp, and J. C. Seferis, "Phase Structure and Toughening Mechanism Of A Thermoplastic-Modified Aryl Dicyanate," to appear in Polymer (1994).
11. J. L. Hedrick, I. Yilgor, G. L. Wilkes, and J. E. McGrath, Polym. Bull., 13, 20 (1985).
12. J. A. Cecere. J. L. Hedrick, and J. E. McGrath. 31st Intl. SAMPE Symp., 580 (1986).
13. D. Garcia, J. Polym. Sci., Part B, Polym. Phys., 25, 1581 (1987).
14. P. M. Hergenrother, J. Macro. Sci.-Rev., Macro Chem, C19(1), 1 (1980).
15. J. C. Hedrick, N. M. Patel, and J. E. McGrath, in ACS Advances in Chemistry Series, No. 233, p. 293, C. K. Riew and A. J. Kinloch, eds. (1993).
16. S. J. Pak, G. D. Lyle, R. Mercier, and J. E. McGrath, Polymer, 34, 885 (1993).
17. S. P. Wilkinson, S. C. Liptak, J. J. Lesko, D. A. Dillard, J. Morton, J. E. McGrath, and T. C. Ward, Proc. 6th Japan-U.S. Conf. on Comp. Mat., p. 241 (1992).
18. S. C. Liptak, S. P. Wilkinson, J. C. Hedrick, T. C. Ward, and J. E. McGrath, in ACS Symp. Series 475, Chap. 22, p. 36, R. L. Clough and S. W. Shalaby, eds. (1991).
19. S. Takeda and H. Kakiuchi, J. Appl. Polym. Sci., 35, 1351 (1988).
20. P. K. Mallick, Fiber-Reinforced Composites, Chap. 2, Marcel Dekker, Inc., New York (1988).
21. R. A. Pearson and A. F. Yee, J. Appl. Polym. Sci., 48, 1051 (1993).
|Printer friendly Cite/link Email Feedback|
|Author:||Woo, E.M.; Bravenec, L.D.; Seferis, J.C.|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 1994|
|Previous Article:||Poly(ethylene terephthalate)-solvent interaction: gelation and melting point depression.|
|Next Article:||Rubber toughening of poly(lactide) by blending and block copolymerization.|