role-20-interface-1990

role-20-interface-1990 - u‘u $11444£4_ 13M ‘ (0L4-...

Info iconThis preview shows pages 1–14. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: u‘u $11444£4_ 13M ‘ (0L4- Cote am ROLE OF THE INTERFACE IN COMPOSITE MATERIALS DURING WATER AGEING C.BASTIOLI,* M.CASCIOLA,** AND G.ROMANO* *Ist. Donegani, Via Fauser 4, 28100 Novara, Italy; **Dipartimento di Chimica Inorganica, Universita' di Perugia, via Elce di Sctto 8, 06100 Perugia,Italy ABSTRACT The environmental aspects in composite and filled materials have to be carefully considered. Strong changes of their performances, in fact, may be observed, during water ageing, depending on: .reinforcement (nature, impurities, electro- kinetic potential, superficial area, etc): .surface treatment of the reinforcement; .matrix. In this work, a general model has been suggested to explain the hydrothermal behaviour of composite materials. Water sorptioh kinetics and dynamic dielectric analysis, performed on filled materials, with different reinforcement nature and composite structure, constitute the base for such a model. INTRODUCTION Composite materials, expecially thermosetting, attracted wide interest, as materials possessing a unique combination of properties, such as ease of fabrication, chemical inertness and excellent adhesion properties. The combination of these properties with those of fibrous materials, such as glass, carbon and kevlar aramide fibres resulted in structural materials. However, when composite and filled materials are used the environmental aspects have to be carefully considered. Strong changes of their performances, in fact, may be observed depending on the nature -of the composite or filled material. Water ageing may strongly affect the matrix behaviour. A lot of studies have been performed on these aspects. In the case of thermosetting resins the following phenomena have been extensivelly studied: — plastification effects due to water adsorption [1], - leaching of not bound substances and or of hydrolysis products. A not uniform distribution of these substances igside the matrix may also induce osmotic phenomena [ - Io - Mechanical and dynamic mechanical changes tied to reversible and irreversible modifications [5,6]. Previous work, for example, demonstrated how equilibrium moisture levels in an epoxy system can be related to thermodynamic states associated with: ."equilibrium moisture“ content of the compact—cured resin, ."excess moisture“ stored in the system as a result of hydrothermally induced microscopic network changes [7]. OleyElufia-Bdmhbfiahhgcm,he Gmmmdkmnhwnmammflhmuuhh HunmImMaEfius 569 570 Also thermoplastic matrices may exhibit, on water ageing, beside plastificution [8,9], irreversible changes such as recrystallization phenomena, reduction of molecular weight, etc.. These modifications may generate stress concentration and the embrlttlement of the matrix [9]. The behaviour of fibre/rosin composites in hot, wet environments has been intensively studied in recent years [10-13]. Interest was focused on such topics as: the state of aggregation and the mobility of water within the composite [14,15]; the reversibility of the wet/dry cycle [16-18]; the damage induced by the absorption process [19-28]; the effect of this damage on the later stages of the absorption process and on subsequent cycles [17,29,301; the effect of imposed stresses on the absorption of water [31-34], etc.. The experimental results on filled materials, presented in this work, have suggested a model to explain the hydrothermal behaviour of different filled and composites materials, based on the reinforcement nature and the composite structure. Specifically their dynamic dielectric properties have been related to their hydrothermal history considering interfacial and bulk matrix degradation mechanisms as semi—indipendent aspects of a general environmental degradation process in composites. EXPERIMENTAL In this work 4 kinds of filled materials have been analyzed containing 70% w/w of different fillers: 1) Quart: (Supervent. Cominder) 2) a-alumina (ART—HP Sumitomo) 3) Quartz (Cominder) treated by 3- methacryloxypropyl- trimethoxy silane (3HPS) 4) c—alumina (Sumitomo) treated by 3MPS. A new' thermosetting resin (Poly di-trifluoriden methyl isopropyliden bis (1,4-phenoxy-2 methyl-1,3-propy1en methacrylate) (P—BIShI—MPMA)) has been used as a matrix. The resin has been polymerized in presence of 1% of henzoyl peroxide at 80°C for 2 hours and at 120°C for 30 minutes. The double bond conversion degree was abOut 94% [35]. Composites for water sorption experiments have been chemically polymerized in presence of about 0.5% w/w of N,N—dihydroxyethyl—p-toluidlne and 1% of benzoyl peroxide at 40°C for 15 minutes. The double bond conversion degree was about 87% [35]. Composites for dynamic—dielectric analysis (DDA) have been also thermally polymerized in the same conditions adopted for the reSin. Water sorption kinetics have been followed by means of gravimetric measurements using a Sartorius analytical balance with an accuracy of 0.01 mg. DDA tests have been performed at 20°C in the range of frequency between 100 Hz and 10 MHz, using the Hewlett-Packard 4192A Impedance Analyzer. The matrix alone has been also analyzed in comparison with its composites to demonstrate the possibility to separate the interfacial contribution to the hydrothermal ageing. RESULTS AND DISCUSSION Water sorption Water aorption kinetics of p—BISAF-MPHA resin show a fickian behaviour [40] up to 60°C, ' furthermore, the equilibrium water uptake decreases with increasing temperature (fig.1). The deviation observed at 90°C can be attributed to microcavitation phenomena as already observed on the literature for other thermo- setting resins [2,5,7]. Figg.2,3 and 4 show wuLcr sorption kinetic curves in the range of temperature between 37 and 90°C for the treated a-alumina composite and at 90°C for the untreated one.- A first water sorption cycle at 60°C was performed on the samples before ageing at 37°, 48° and 60°C. Both sileneted and not treated u-alumina composites, at curve concomitant effects of water sorption and leach out present a similar first cycle monomers . The second cycle curves present effects are shown over the The equilibrium graphs (figg. calculated starting from at 90°C (£ig.1). The the a plateau water uptakes 2,3) together with experimental data 571 0 K)- 20 Wn (flrzmano' I FIG.1. Water sorption kinetics of p-BISAF—MPMA at different temperatures. 90°C, resulting by of residual (figg.2,3) and no debonding time investigated. reported in the values (Mth) curve t0 0!: range of (M0) are the theoretical p—BISAF-MPMA kinetic comparable are lower than the theoretical equilibrium uptakes. “0'0171 Mm-fi2l: 30 40 éo fin (fitcmn10‘3 FIG.2. Water sorption kinetics 0 10 of the silanated a—alumina composite at 90°C. Men-0207. th' 0.211. 20 30 40 fin tVEIcmmo'3 FIG.3. Water sorption kinetics of the untreated a—alumina composite at 90°C. 572 In the investigated range of temperature, therefore, the water sorption kinetics of a—alumina composites are perfectly consistent with the matrix trend. The presence of 3MPS does not play any role on the water ageing process [36], due to its low c—alumina coverage degree [36]. A more complex behaviour is found for quartz composite (figg.4,5). As in the previous case a first water sorption cycle was performed before ageing at 37°C, 48°C and 60°C. Single step 0 curves have- been observed up to 48°C for silanated quartz composites: such a 20 40 'VTIlu/Elcmnlo‘3 behaviour is in PIG.4. Sorption kinetic curves of reasonable accord with silanated e—alumina(—e) and silanated that expected on the guartz(-)composites at different base of the resin temperatures. experimentalbehaviour. However, at 60°C and 90°C, double step kinetic curves can be observed for both treated and untreated quartz composites. 0 1o 20 30 40 VII-ll (Elem 31:10.3 FIG.5 Water sorption kinetics of quartz composites at 90°C. a) Silanated quartz/I cycle; b) Silanated quartz/II cycle; c) Untreated quartz/I cycle. Such a phenomenon, not observed in the a-alumina composites, depends greatly on the filler surface treatment. Possible explanations of the above mentioned phenomenon have been proposed for glass fiber epoxy resins [10,32], crosslinked polyesters filled with glass beads [11] and glass fiber reinforced polyesters and vinylester composites [2] based on filler/matrix debonding with consequent formation of liquid films around the filler surface. As shown in fig. 5, a higher sorption level and a faster sorption process can be observed for the not treated quartz composite. .‘Ill ‘ .__ “many-«vu- ‘ «aw-M- Eilnnnte u-ulumina cumyulee FIG.E. Elnc+rnn micrograph: of compmfiitc fracture surfncca Ina-fore {a} and after {by 9 day-a of water ageing. 574 The higher water sorption (Ht) values obtained in the second cycle take into account the loss of soluble species, occurring during the first water sorption cycle. such a weight loss resulted to be 0.11% for the treated quartz composites and 0.24% for the untreated quartz composites. -The weight loss of soluble species was practically absent during the water sorption cycles of the unfilled p—BISAF-MPMA at 90°C, because of the thermal polymerization instead of the chemical one. On the base of the results obtained for the resin alone or in the presence of m—alumina, the double step kinetic curves cannot be accounted for by resin contribution. The generation of significant free volume, due to relaxation of interfacial stresses, caused by the resin shrinkage during the crosslinking step, has also to be excluded. If present, in fact, it should be evident also in a-alumina composites. Two, therefore, can be the possible causes for the quartz composites sorption behaviour at temperatures higher than 48°C: 1) High filler concentration. From this point of view u-alumina with its high density occupies a significantly lower volume (43.3%) in comparison with quartz (57.3%). It implies quartz composites may show with major probability macrocavities and/or regions of direct filler-filler contact. SEM micrographs of thermally polymerized materials confirm the presence of macrocavities at different extent in quartz composites (fig.6). The 3MPS filler treatment enhances the filler wettabiiity with an expected and observed (fig.6) reduction of microcavities. Such a phenomenon, however, can explain only at a minimum extent the abnorm water sorption. Otherwise, the sorption kinetics at 37° and 48°C should have been significantly different if compared with the a—alumina composite ones (fig. 4); this difference is, on the contrary, really small. 2) Soluble species at the interface able to generate strong osmotic effects [28]. Such an hypothesis could explain the different behaviour of quartz and alumina composites only if quartz was able to produce soluble species starting from 60°C. The osmotic phenomenon already underlined [21] for E glassIpolyethylentherephtalate composites, can explain the double step water sorption kinetic and the progressive loss of mechanical characteristics of the samples. The high sorption level achieved, in fact, may imply crack growth on water ageing, with the consequent modification of the liquid films which, insulated at the beginning, become progressively more interconnected. In quartz composites the presence of 3MPS coupling agent plays a role more tied to the protection of the filler surface than to a real possibility to create chemical bonds between resin and filler. 3HPS, in fact, in this case has lost the acrylic insaturation before the crosslinking step [36]. gynamic Dielectric Analysis D can less In the range of frequency between lKHz and lOMHz the dielectric properties of p-BISAF-MPMA, containing the equilibrium water uptake of 0.61%, do not differ significantly from dry p—BISAF-hPMA. 575 In the dry composites the filler presence promotes an increase of E', explained on the base of the higher dielectric constant of the filler compared with the matrix. Both dry matrix and composites exhibit E' values practically constant in the whole investigated frequency range. P—BISAF-MPMA a—alumina com osites BISALS : BISALS, aged at different times in water at 90"C, shows a circunference arc as Cole—Cole diagram [3?] (fig.7). The E'm values, extrapolated at high frequencies, coincide with the E'm of the dry materials. The semicircular shape of the Cole-Cole plot implies the'presence of a maximum in tho E'llgf curve. The frequency corresponding to e'max (fig.8) progressively moves to higher values on water 4 ageing and achieves a £0 maximum at 21 days. For longer times the frequency shifts again to lower values. The dielectric relaxation of the hydrated 2 composite cannot be attributed to the water molecules at the matrix—filler interface. In fact, even if the amount of 0 water present at the interface (<0.02%) behaved as free water, its contribution to 6' could be PIG.7. Cole—Cole plot of BISALS estimated lower than the 6' samples aged in water at 90°C. measurement error and, a) 5h: b) 27h: C)49d- therefore, not singled out in the Cole-Cole diagrams. It is well known [38,39] that the region in the vicinity of oxide/water interface may show a discontinuity in the electrical properties termed electrical double layer. Consequently the dielectric 5b relaxation can be explained supposing the a—alumina particle surface may charge itself, in the presence of water molecules, so that the liquid films around the filler particles become conductive. The Contiguous surfaces of the particles may be considered as armors of capacitors connected by 40 the liquid film resistances. t(days) In terms of equivalent circuit each couple of PIG.8. Trend of the frequency contiguous particles can be corresponding to E"max of represented, to a first BISALS on water ageing at 90°C. approximation, by a series combination of a resistance (r) and a capacitance (c) with time costant t=rc. In an ideal composite, 3 Log kcquencyiflr) % M '0! O 20 576 Z“‘=k(iw—)n C(O' C(m) FIG.9. Equivalent circuits for an ideal (a) and a real (b) composite. log le2) F1G.10. Simulation BISALS permittivity data on the base of the circuit of fig.9b. K=8.3540x1036 (nHz-n); n90.6163;C(0)=1.7787xlflE—9(F); C(m)=6.9182 x 10E-10 (F). of the constituted by spaced filler equal dimensions, r-c elements characterized by same time constant. The corresponding equivalent circuit is shown in fig.9a, where R and C(0) represent the resultant resistance and capacitance (RxC(0)=t). C(w) is the capacitance limit value of the hydrated composite extrapolated at high frequencies. In a real case the particles could be different in size and irregularly spacsd, thus originating a distribution of time constants. In this case, as shown by Cole—Cole [37], R has to be substituted by a constant phase angle element whose impedance is z*=x(iw)-n (fig.9b). The Cole-Cole diagram of such a circuit is constituted by an arc of a not distorted circle whose center is under the 6' axis. The permittivity data of the hydrated BISALS have been simulated on the base of the circuit of fig .9b by means of non linear least squares fitting, obtaining a good agreement between observed and calculated values (fig.10). The translation of the frequency corresponding to E“max towards higher values is in accord with the matrix regularly particles of all the are the relaxation around the fillers and with the activation of a—alumina surface by means of water molecules with a consequent, progressive increase of the generated charge mobility. The counterions of the hypothetic electric double layer at the a-alumina surface may only generate small osmotic effects on ageing since counterions and fixed charges cannot be separated 577 at infinite distance due to their reciprocal attraction. Such an effect is confirmed by the water sorption kinetic, which, in this case, does not move away significantly from the theoretical value calculated on the base of the resin volume fraction. The electron micrographs of the fracture surface for the thermally polymerized composite, before and after 9 days of water ageing, are reported in fig.6. The smooth surface of the fillers after ageing confirms the relaxation phenomena of the matrix and the liquid film formation around the fillers. The progressive reduction of the E'max frequency after 21 days of water ageing could be well explained by the proposed model, considering an increase of the resistance of the liquid films around the filler particles. However further investigation is needed to better understand the above mentioned phenomenon. P-BISAF—HPMA silanated uartz com osites BIS ZS : The permittivity real part of the hot water aged BISOZS samples shows an approximately linear decrease on the frequency logarithm; it always remains higher than the E'm value of the dry sample (fig.11). The imaginary part is constant in almost all the range of the explored frequencies and increases in the region of the low and high frequencies (£19.12). This behaviour indicates the overlapping of, at least, tWO' relaxation phenomena at frequencies higher than lOHHz and lower than lkflz. The presence of a relaxation phenomenon at high frequencies becomes evident considering the Cole—Cole diagrams (fig.13). It is possible to explain the low frequency relaxation on the base of the BISALS proposed model. The large volume fraction of fillers, with the consequent high probability of particle aggregates, should give C(O) values larger than those of BISALS. The high quartz coverage degree by 3MPS coupling agent should avoid quartz impurity solubilization. In agreement with the last consideration, the electrical conductivity of 50cc of water in presence of log of silanated quartz, measured after 30 days of ageing at 90°C, gave no significant differences from the demineralized water conductivity. The presence of liquid films of high resistance, combined with high interparticle capacitances is consistent with the low frequency relaxation. The high frequency relaxation phenomenon cannot be explained on the base of the model applied to the a—alumina composite; it should be necessary, in fact, to admit the presence of r-c elements with time constants three order of magnitude lower than that observed for BISALS. It can he, therefore, attributed to a relaxation of the water molecules present at the matrix/filler interface and inside the micro and macrocavities of the composite material. Its microstructure, after and before ageing in water, is reported in fig.6. The Cole-Cole diagram permits only a rough evaluation of £>E'. An accurate calculation is not possible for the complexity of the system; if AE' is due to the water relaxation, it cannot be higher than the product between the water dielectric constant (80) and the water volume fraction of the composite. 0n the base around 1.5, higher than the value extrapolated by the Cole-Cole diagram (AE'al). 578 30 SB 45 EA 62 10 Log Iroqucncy Wt) FIG.11. Real part of the £16.12. Imaginary part of the permittivity of 318035 permittivity of 313025 samples, aged in water at samples. aged in water at 90°C, vs. lgf. a) Dry; b) 1h; 90°C, vs. lgf. a) Dry; b) 1n, c) 1d: d) 16d. c) 1d; (1) 16d. E“ Log trequencymu' FIG.13. Cole-Cole plots of £16.14. Imaginary part of the BISQZS samples aged in water permittivity of BISQZ samples, at 90°C. a) Dry; b) 1h; c) 1d; aged in water at 90°C, vs. (1) 16d. lgf. a) Dry; b) 211; :3) 2h; d) 6h. 160 La 80 _ O 10 22 3A 4B 58 IO 20 16 132 188 24A 300 . Log lroquencytflzl' I E, FIG.15. Real part of the FIG.16. Cole-Cole plots of permittivity 0f BISQZ Samples: BISQZ samples aged in water at aged in water at 90°C, vs. 90°C. a) 2h; b) an; c) 6h. égf. a) Dry; b) 2h; c) 4h; d) ,-__ 579 P—BI AF-MPMA uartz com site BIS z : It is possible to distinguish. in this case, three different trends during water ageing: ' 1.After 2 hours 6“ increases with the frequency (£19.14) and 6' assumes values higher than that of the dry material (fig.15). This behaviour is analogous to the BISQZS one (fig.11) and it can be attributed to the water molecule relaxation at the matrix/filler interface. 2.8etween 4 and 6 hours of ageing, in the region of the low and medium frequencies, the Cole—Cole diagram (fig.16) shows a depressed circumference arc, attributed to r-c elements, due to liquid conductive paths at the filler/matrix interface, as demonstrated for aged BISALS. In this case, however, the behaviour is related to the solubilization of_conductive impurities. This aspect has been demonstrated by measuring the electric conductivity of 50cc. of water in presence of log. of quartz, after 30 days at 90°C. It was three times the conductivity of the demineralized water. If the water/filler ratio of this experiment is reported to that of the composite, the conductivity of the liquid films in BISQZ turns out to be about two orders of magnitude higher than in 315925. The Cole-Cole arc is larger than the BISALS one, in accordance with the proposed model of fig.9b. 31502, in fact, has to exhibit a higher C(O) associated with the less spaced particles, due to the quartz larger volume fraction and to the consequent filler aggregates. Differently from BISALS the Cole-Cole diagram cannot be interpolated by a circunference arc in the whole range of frequencies; at the high frequencies, in fact. there is an overlapping of two relaxation phenomena due respectively to the filler conductive surfaces and to water. 3.Till 45 days of ageing it is possible to observe the progressive development of a tail in the low frequency region of the Cole7Ccle diagram (fig.17). This phenomenon can be explained on the base of conductive liquid paths connecting different filler particles to form a continuous network between the two faces of the dielectric material. This hypothesis has been confirmed, not only by SEH analysis (fig.6) during water ageing, but also by the ageing of BISALS, BISQZS and BISQZ in In NaCl solution. The 31502 Cole-Cole plot, in fact, showed a further increase of the low frequency tail. No significant modifications have been observed in the BISALS and BISQZS permittivity diagram, demonstrating the two materials contain filler particles mainly isolated by means of the matrix. The presence of a network of liquid paths in BISQZ has been also demonstrated by the evolution of the imaginary part (M") of the complex modulus (M*=M'+iM")=1/E*), during water and NaCl ageing (fig. 18). In the present case, the use of the modulus instead of the permittivity allows the time constant change to be more evident. The relaxation phenomenon, observed in the M'llgf plot at high frequencies, moves to lower values at long water ageing, showing a reduction of the liquid film conductivity due to the migration of the solubilized impurities towards the water bath. When the water aged sample is stored in in NaCl solution the liquid film conductivity increases again with a consequent decrease of the time constant. 580 160 20 7.6 132 135 24.4 300 to 22 a4 4.6 as 7.9 I E' log tuquencymn FIG.11. Cole-Cole plots of FIG.18. Imaginary part of the BISQZ samples differently aged complex MOdUIUB Of BISQZ At 90°C. a) 1d in water; b) samples differently aged at 49d in water; a) 49d in water 90°C. a) id in water; b) 49d and 3d in In Nacl. in water: c) 49d in water and 3d in 1H NaCl. CONCLUSIONS For a better understanding of the composite water ageing at the reinforcement/matrix interface. filled materials with high reinforcement concentration (70% w/w) have been studied as a model. The filler nature (u—alumina, quartz) has been changed in presence or absence of coupling agent (3MPS). The new thermosetting resin (p—BISAF~MPMA) used as a matrix is characterized by low water equilibrium uptake in comparison with epoxy resins and by high double bond conversion degree. Mt(%) 2 \/-;/| (El FIG.19. Possible trends of FIG.20. Possible trends of the water sorption kinetics in complex permittivity in composite materials . composite materi a Is . Combined DDA and water sorption experiments performed on the different model composites brought to the following conclusions: .When the composite is ideal and the interface does not contribute and no cavities are present, it has to show a water sorption kinetic in line with the matrix behaviour (curve 58] 1 of £19.19). If the filler is not conductive, the permittivity diagram can be represented by a point (point 1 of £19.20) as the pure matrix or the dry composite. .When a composite contains high amount of cavities it can show a non fickian water sorption kinetic close to curve 2 of £19.19. The permittivity diagram is constituted by a circunference arc, in the region of the high frequencies (element 2 of fig.20), coming from extra water relaxation phenomena. Both BISQZS and BISQZ at low ageing time exhibited this contribution. .When a composite, during water ageing, gives liquid films at the reinforcement/matrix interface. with _charqe mobility, the permittivity diagram can be simulated by the equivalent circuit of fig.9b, and it is close to element 3 of fig.20. Such a behaviour has been observed in all the examined composites. Water sorption kinetics, however, can show two different shapes depending on the charged species which can be solubilized or not in the liquid films. In the case of solubilized impurities the sorption kinetic is a double step one (31802 and BISQZS), while in the presence of fixed charges it can be very close to that of an ideal composite (BISALS and BISAL). The double step sorption kinetics are tied to osmotic effects which, if strong, generate a network of conductive liquid films, sometime able to connect the two faces of the dielectric material. In this last case it is possible to observe, on water ageing, a tail growth (contribute 4 of fig.20) in the permittivity diagram. This phenomenon has been clearly noticed in BISQZ samples after long water ageing. REFERENCES 1. A.Apice11a, L.Nicolais, G.Astarita, E.Drioli Polymer 20, 1143—1148 (1979) . A.Apicelle, C.Migliarasi, L.Nic01ais, L.Iaccarino, S.Roccote111 Composites 14,N°4,387-392(1983) H.P.Abeysinghe, H.Eduards, G.Pritchard, G.J.Swampillai polymer 23,1785-1790(1982) E.Walter, K.H.G.Ashbee Composites,365-368(0ctober 1982) . A.Apicella, C.Higliaresi, L.Nicolais, L.Iaccarino, S.Roccotelli, Composites,406—410(1982) 6. D.H.Kaelble, P.J.Dynes, J. Adhesion 8,195-212(1977) 1. W.J.Mikols, J.C.Seferis, A.Apice11a, L.Nicolais Polym. Composites 3,N°3,118-124(1982) B. M.Ashida, T.Noguchi, S.Mashimo J. Appl. Polym. Sci. 29,4101—4114(1984) 9. C.Bastioli, I.Guanella, G.Romano J. Eng. Sci. to be published in February 1990 10. B.Dewimille, A.R.Bunse11 Composites 14,35-4D(January 1983) 11. G.S.Springer, B.A.Sanders, R.W.Tung J.Composite Mat. 14, 213-232(1980) 12. V.N.Kanellopoulos, G.H.Wostenholm, B.Yates, R.C.Sanders J. Nat. Sci. 21,643—648(1986) 13. M.R.Piggott, P.W.K.Lam, J.T.Lim, M.S.Woo Comp. Sci. Tech. 23,247-262(1985) l4. P.Hoy, F.E.Karasz Polym. Eng. Sci. 20,315-319(1980) 15. Y.Diamant, G.Moron, L.J.Broutman J. Appl. Polym. Sci. 26,3015-3025(1981) 16. R.J.A.Sha1ash, E.A.Sarah in: Crosslinked Epoxies (Walter de Gruyter & Co.,Germany,1987) pp.459-470 17. 0.Ishai _pn1ym. Eng. Sci. 15,N°7,491—499(1975) 01-th 582 18. J.L.Illinger, N.S.Schneider Polym. Eng. Sci. 20,N°4,310- 314(1990) 19. B.Harris, P.W.R.Beaumont, E.M. de Ferrau J. Mat. Sci. 6.238(1971) 20. P.W.R.Besumont, B.Harris J. Comp. Mat. 1,1265(1972) 21. D.H.Kse1b1e et s1. J. Adhesion 1,25(1974) 22. T.A.Collings, D.L.Mead Composites 19,N°1,61-66(1988) 23. H.Aytsc, J.Renard, G.Verchery Comptes Rendus des Troisiemes Journees Nationales sur les Composites (JNC3)(1982) pp.141-149 24. K.w.Thomson, L.J.aroutmsn Polym. Composites 3,N°3, 113—117(1982) 25. I.Verpoest, G.S.Springer J. Reinforced Plastics and Composites 1,23432(1988) 26. I.Verpoest, G.S.Springer J. Reinforced Plastics and Composites 1,2-21(1988) 27. C.Bastioli, M.Casciola et a1. Free. of the Second International Conference of Composite Interfaces (H. Ishida, Editor,Elsevier,1988) pp.189-203 2B. K.H.G.Ashbee, R.C.Wyatt Proc. Roy. Soc. A132,558-564(1969) 29. G.Pritchard, S.D.Speske Composites 1B,N°3,221—232(1981) 30. G.S.Springer in: Developments in Reinforced Plastics,2 (Applied Science Publishers Ltd., London 1982)ch3,p.60 31. V;F.anas, R.L.Mccu110ugh Composites Science and Technology 29,293-315(1987) 32. K.A.Kasturiarschchi, G.Pritchard Composites 14,N°3,244— 250(1933) 33. halvena, A.R.Bunse11 Composites 19,N°5,355—357(1988) 34. O.Gillst, L.J.Broutman in: Advanced Composite Materials Environmental Effects, ASTH STP658 (ASTM, 1978) pp.61-83 35. C.Bsstioli, G.Romano to be published 36. F.Garbassi, E.0cehiello, C.Ba8tioli, G.Romano, .Brown Proc. of the First International Conference on Composite Interfaces (H.Ishida Editor, Elsevier,1986) pp.241-250 37. K.S.Cole, R.H.Cole J. Chem. Phys. 9,341(1941) 38. A.L.Smith in: Dispersion of Powders in Liquids (G.D.Parfitt Editor, Elsevier,1969)ch2,pp.39-80 39. M.D.Sacks, T.Y.Tseng J. Am. Ceramic Soc. 66,N°4,242—246 (1982) 40. J.Crank, G.S.Psrk in: Diffusion in Polymers (Academic Press, London,1968) ...
View Full Document

Page1 / 14

role-20-interface-1990 - u‘u $11444£4_ 13M ‘ (0L4-...

This preview shows document pages 1 - 14. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online