Abstract
A study of the effect of tethered chains at the fiber-matrix interface on the performance of a particulate composite at very high strain rate was conducted. The materials system studied was composed of glass beads, polysulfone matrix, and polysulfone tethered chains. The entanglement of the tethered chains with the matrix in the composite specimen was confirmed by physico-chemical analysis. Although the tethered chains had been shown to have a beneficial effect on the interface in a single-fiber, model composite, they were found to have no effect whatsoever on the macroscopic behavior of the particulate composite at very high strain rate (150/s).
| Original language | English |
|---|---|
| Pages (from-to) | 175-186 |
| Number of pages | 12 |
| Journal | Journal of Composite Materials |
| Volume | 35 |
| Issue number | 2 |
| DOIs | |
| State | Published - 2001 |
Bibliographical note
Funding Information:Greenfield M. J. Hunter T. F. Kalika D. S. Penn L. S. Department of Chemical and Materials Engineering, University of Kentucky, 177 Anderson Hall, Lexington, KY 40506-0046 01 2001 35 2 175 186 A study of the effect of tethered chains at the fiber-matrix interface on the performance of a particulate composite at very high strain rate was conducted. The materials system studied was composed of glass beads, polysulfone matrix, and polysulfone tethered chains. The entanglement of the tethered chains with the matrix in the composite specimen was confirmed by physico-chemical analysis. Although the tethered chains had been shown to have a beneficial effect on the interface in a single-fiber, model composite, they were found to have no effect whatsoever on the macroscopic behavior of the particulate composite at very high strain rate (150/s). interface tethered chains high rate impact resistance thermoplastic composite energy-dissipation mechanisms compression test particulate composite sagemeta-type Journal Article search-text Improvement of Interfaces by Tethered Polymer Chains. Part II: Evaluation of a Particulate Composite at Very High Strain Rate M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN* Department of Chemical and Materials Engineering University of Kentucky 177 Anderson Hall Lexington, KY 40506-0046 (Received May 28, 1999) (Revised April 1, 2000) ABSTRACT: A study of the effect of tethered chains at the fiber-matrix interface on the performance of a particulate composite at very high strain rate was conducted. The materi- als system studied was composed of glass beads, polysulfone matrix, and polysulfone teth- ered chains. The entanglement of the tethered chains with the matrix in the composite spec- imen was confirmed by physico-chemical analysis. Although the tethered chains had been shown to have a beneficial effect on the interface in a single-fiber, model composite, they were found to have no effect whatsoever on the macroscopic behavior of the particulate composite at very high strain rate (150/s). KEY WORDS: interface, tethered chains, high rate, impact resistance, thermoplastic composite, energy-dissipation mechanisms, compression test, particulate composite INTRODUCTION N THIS PAPER, Part II of a pair, we describe the extension of an investigation of the tethered chain strategy from the single-fiber, model composite described in Part I [1] to a more realistic composite containing multiple, uniformly distributed interfaces. As described in Part I, tethered chains were found to improve the inter- face in a single-fiber composite at low strain rate and also under impact, at a strain rate approximately three orders of magnitude higher. These findings were taken as 175 Journal of COMPOSITE MATERIALS, Vol. 35, No. 02/2001 0021-9983/01/02 017512 $10.00/0 DOI: 10.1106/N8H5-724T-LBMB-F3GK 2001 Technomic Publishing Co., Inc. *Author to whom correspondence should be addressed. evidence that the molecular-level, energy-dissipation mechanisms provided by the tethered chains at low strain rate (long time scales) were still operative at very high rate (short time scales). The goal of the work described below was to deter- mine whether the beneficial effect of polymer chains tethered to the interface could be observed in the macroscopic response of a more realistic composite at very high strain rate. The obvious choice for the test specimen in this phase of the investigation would have been a unidirectional fiber composite, the multi-fiber analogue to the sin- gle-fiber composite used in Part I. Unfortunately, practical problems made the use of a unidirectional fiber composite impossible. The main problem was that accu- mulation of sufficient fiber for preparation of fiber composite specimens would have required an excessively long time, since our chemical procedures allowed only small amounts of fiber to be subjected to the tethering chemistry at one time. Two other problems arose from the high temperatures and pressures needed to make void-free unidirectional composites from a high-performance thermoplastic matrix. First, the temperature of at least 350C needed to get the polysulfone ma- trix to flow is high enough to disconnect the tethered chains by decomposing the aliphatic linkages connecting the chains to the surface of the glass fiber. While these linkages can survive temperatures up to 300C for moderate lengths of time, they are degraded almost immediately at 350C. Second, the flow under high pres- sure of a viscous matrix such as polysulfone, even at high temperature, produces significant fiber misalignment. In our case, this could not be tolerated, because we were committed to compression testing (see below) in this part of the research, and compression tests are extremely sensitive to fiber misalignment. The problems just described were avoided by the selection of glass beads in- stead of glass fiber for this phase of the investigation. Not only could a much greater mass of glass beads than glass fiber be subjected at one time to the tethering chemistry, but a lower volume percent of glass beads than glass fiber could be used to make a realistic composite. Furthermore, with beads it was possible to develop a multi-step procedure (described later) for making uniform, void-free composites under much milder time-temperature exposures than would have been required for making fiber composites. Finally, the spherical geometry of glass beads, in addi- tion to making fiber misalignment irrelevant, would result in compression of the interface at the poles and tension at the equator of each bead, a situation that would be expected to provide an excellent test of the tethered chain strategy. Composite specimens with and without tethered chains at the glass/polysulfone interface were tested on a special apparatus at the Army Research Laboratory in Aberdeen, Maryland [2]. The apparatus was selected because of its ability to apply extremely high and constant strain rate (150/s) and to provide an accurate record of specimen response throughout the extremely short duration of the test. This appa- ratus operated in compression mode only and required small, cylindrical speci- mens. 176M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN EXPERIMENTAL Silanization of Solid Glass Beads Solid glass beads (average diameter = 10 m) were obtained from Potters Indus- tries, Carlstadt, NJ. Slightly more than 100 g of beads were washed in 0.15 M HCl for 24 hours and were rinsed thoroughly with distilled water. They were then dried and stored in a vacuum desiccator. The procedure used to create epoxide reactive sites on the glass bead surface was the same as that used on the glass fiber in Part I, but was carried out on a larger scale. A 500-ml, rather than a 100-ml, round-bottomed flask was fitted with a condenser, glass stopper, and rubber septum. Glassware was flame-dried and flowing, dry N2 was introduced. To the flask, 100 g of glass beads were added, fol- lowed by 75 mL of anhydrous toluene and 0.2714 g of 3-glycidoxypropyl trimethoxy silane (Aldrich, Minneapolis, MN). The reaction mixture was heated at reflux for 16 h under dry N2. After this, the silanized beads were separated from the reaction mixture by filtration through a 3-m membrane filter (Poretics, Livermore, CA). The beads were rinsed three times with toluene and were dried at 110C under vacuum. As described in Part I, this procedure resulted in 2.71 0.24 epoxide sites/nm2 of glass surface [1]. Some batches of these beads were reserved for use in the control sample and other batches were subjected to the tethering reac- tion described below. Tethering of Polysulfone Chains to Solid Glass Beads A 500-ml, 3-necked, round-bottomed flask, was equipped and dried as de- scribed above. Then 100 g of solid glass beads (silanized) and 1.0 g of amine func- tional-ended polysulfone (Mn = 14,100) were transferred to the flask through the stoppered neck. The polysulfone was dissolved with addition of 75 ml of anhy- drous tetrahydrofuran by syringe through the septum. The reaction mixture was heated at reflux for 48 h under dry N2. The reacted beads were separated from the cooled reaction mixture by filtration through a 3-m membrane filter. Then the beads were subjected to Soxhlet extraction for 48 h in methylene chloride, after which they were dried at 110C under vacuum for 24 hours. They were stored un- der dry conditions until needed. As described in Part I, this treatment resulted in 0.281 chains/nm2 of surface [1]. Preparation of Molding Compound Preparation of a molding compound comprised all but the final step of a multi-step process developed specifically for fabrication of cylindrical composite test specimens. This multi-step process, in which the temperature never exceeded Improvement of Interfaces by Tethered Polymer Chains, Part II177 270C, was a substitute for a single-step, injection molding process, which would have required temperatures of at least 350C. As mentioned in the introduction, at 350C the aliphatic linkages connecting the functional-ended polysulfone chains to the glass surface would be thermally decomposed, disconnecting the tethered chains. A mixture of 12.5 g of treated glass beads, 24.8 g of polysulfone (Grade P1700, Amoco, Alpharetta, GA), and 75 ml of dimethyl acetamide (Sigma, St. Louis, MO) was prepared and degassed for 15 min under vacuum. Films of the mixture were cast onto thick glass plates (120 mm 120 mm 6.35 mm); a doctor blade was used to control the film thickness to 2 mm. The solvent was allowed to evapo- rate slowly, over 5 days at room temperature, in a dry box. Residual solvent was re- moved from the films by exposure to the following sequence of elevated tempera- tures: 100C for 1 h, 200C for 1 h, and 250C for 1 h. After this, the dried, brittle films were floated off the glass plates in a water bath at room temperature and were air dried. The relative amounts of glass and polymer used were intended to yield films with 20% by volume of glass after removal of solvent; however, a small resi- due of beads left on the glass plates during removal of the films led to slightly lower glass bead content. Despite previous degassing, these solid films contained air bubbles. The air bubbles were removed by the following pressing procedure: dried films were bro- ken into pieces manually, were placed within a picture-frame mold, and were pressed into thin sheets with a laboratory press (Carver, Menomonee Falls, WI) at 260C for 2 min. The resultant sheets, now free of air bubbles, were fragmented into oat-sized chips with a rotary mill, and were reserved until needed for compres- sion molding. Compression Molding of Cylinders Compression molding was the final step of the multi-step process developed specifically for fabrication of the composite test specimens. In this step, a cus- tom-made steel mold, consisting of a cavity and a plunger, was used to compres- sion-mold the molding compound into cylindrical specimens. The mold, previ- ously sprayed with mold release, was brought to 270C and held at temperature by a band heater. The plunger was removed, 6.75 g of molding compound was poured into the cavity, and the plunger was replaced. The mold was immediately placed between the heated platens (270C) of the Carver press, and a load of 4 metric tons (gauge) was applied. After 2 min, the band heater and the platen heaters were switched off, and the load was reduced to 2 metric tons (gauge). The mold and the platens were allowed to cool together to room temperature. After this, the mold was removed from the press, and the composite cylinder, 10 mm in diameter 30 mm long, was removed from the mold. The glass bead content was determined gravimetrically to be 17% by volume. 178M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN Recovery and Analysis of Beads from Cylindrical Composite Specimens Solid glass beads were recovered from the cylindrical composite specimens and analyzed for the presence of tethered chains. For this, a composite specimen 10 mm in diameter and 30 mm long, prepared by compression molding as described above, was placed in a 100-ml round-bottomed flask fitted with a condenser. After addition of 50 ml of tetrahydrofuran (THF) to the flask, the mixture was stirred at reflux for 24 h, during which time the polysulfone matrix completely dissolved and released the beads. The supernatant was decanted from the beads and dis- carded. The recovered beads were thoroughly rinsed in refluxing THF several times and were dried under vacuum at 110C. These beads were then subjected to thermogravimetric analysis/mass spectrometry (TGA/MS). For comparison, beads without tethered chains were also fabricated into a cylindrical composite specimen, and were recovered from the specimen for TGA/MS. The TGA/MS technique is an analysis method in which the volatile decomposi- tion products from TGA are channeled into a mass spectrometer for analysis. The MS analysis provides chemical identification of the decomposition products com- ing off the sample at each temperature. In our case, bead samples of approximately 5 mg were heated from room temperature to 850C at 25C per minute under he- lium atmosphere. One complete mass spectrum of the volatile decomposition products was collected over each minute of the TGA run. High-Strain-Rate Compression Testing of Cylinders High-rate compression tests were performed at the Army Research Laboratory at Aberdeen Proving Ground, MD, on a special servohydraulic mechanical test ap- paratus capable of applying very high, constant deformation rates [2]. Figure 1 shows the apparatus. Prior to testing, each cylinder was trimmed with a dou- ble-bladed diamond wheel saw so that the end surfaces were flat, perfectly parallel to each other, and perpendicular to the specimen axis. Trimmed cylinders were 17.5 mm long. Specimens were tested at a strain rate of 150/s. The apparatus pro- duced complete stress-strain curves for each specimen until the test was stopped at about 50% compressive strain. Six replicate experimental specimens (with teth- ered chains) and six replicate control specimens (silanized only, no tethered chains) were tested. In addition, three cylinders of polysulfone containing no glass beads were tested under the same conditions. Microscopy of Tested Specimens The fracture surfaces of compression-tested cylindrical specimens were exam- ined with a scanning electron microscope (Hitachi 3200N) at voltages of 20 keV and 5 keV, with the lower voltage providing higher resolution. To prevent charge Improvement of Interfaces by Tethered Polymer Chains, Part II179 buildup on the specimens in the beam a 1-nm layer of gold was sputter-coated onto the surface of each specimen. The interiors of compression-tested specimens were examined by optical microscopy. Interior planes were exposed and prepared by standard metallographic techniques as follows. Tested cylinders were potted in room- temperature-curable epoxy resin, which was a stoichiometric mixture of diglycidyl ether of bisphenol A (Epon 828 from Shell, Houston, TX) and a liquid aliphatic amine (T403 from Huntsman, Salt Lake City, UT). After cure, a diamond wheel saw was used to cut the potted specimens through the middle. The exposed interior planes were ground with a progression of silicon carbide papers (180 to 4000 grit), and then were polished with successively finer particles of alumina. This careful procedure, water-cooled at every step, produced planes polished to optical flatness without introduction of damage to the specimens. 180M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN Figure 1. Diagram of high rate compression testing apparatus. RESULTS AND DISCUSSION Results of TGA/MS Analysis of Recovered Glass Beads The TGA/MS analysis of the beads recovered from the control cylinder (no teth- ered chains) showed no evidence whatsoever of polysulfone. Of course, this was the expected result, since only polysulfone chains chemically reacted by their functional end-groups to the glass surface would have been expected to survive the rigorous rinsing in organic solvent. On the other hand, TGA/MS analysis of the beads recovered from the experimental cylinder did show evidence of polysulfone, indicating that the tethered chains had remained connected to the glass surface throughout the multi-step procedure used to fabricate the cylinders. In other words, the aliphatic linkages connecting the functional-ended polysulfone chains to the glass surface were not degraded by the temperatures of the fabrication procedure. Even more interestingly, the beads recovered from the experimental cylinder showed significantly larger percent mass loss than could be accounted for by tethered chains alone. We suspected that the extra mass was from chains of matrix polymer so well intermingled with the tethered chains that they were not dissolved away by the rinsing procedures used in the recovery of the beads from the cylinder. Our suspicion was confirmed after these beads were sub- jected to extensive, additional rinsing in THF at reflux, with the mass on their sur- faces gradually decreasing to a level consistent with that of tethered chains alone. In sum, the tethered chains in the fabricated composite were found to be not only attached to the surface of the glass beads but also intermingled with the matrix chains. This is highly significant because these are the necessary conditions for tethered chains to provide molecular-level, energy-dissipation mechanisms at the interface. High Rate Compression Test Results High-rate compression testing caused the cylindrical specimens to flatten to about 60% of their original length and to expand radially to about twice their origi- nal diameter. Viewed from the top, each tested specimen had a central region that was flattened but free of fractures. This central region was bounded laterally by material that had barrelled and fractured radially. The occurrence of fractures in rapidly deforming specimens is not surprising, since there would not be time for the specimen to fully accommodate such a drastic change in shape by plastic defor- mation alone. Figure 2 shows the stress-strain curves of the cylindrical specimens tested in compression at high rate, and Table 1 lists values of key engineering properties ob- tained from the tests. Comparison of the values of these properties for the Improvement of Interfaces by Tethered Polymer Chains, Part II181 bead-filled specimens (experimental and control) with the values for the unfilled specimens (reference) reveals the expected changes due to the presence of beads: higher initial modulus, higher yield stress, and lower strain at yield. The unexpected result was the absence of a distinct effect of tethered chains on performance. In Figure 2, there is no apparent difference between the experimen- tal (with tethered chains) and control (no tethered chains) samples. The small but statistically significant difference (two-tailed Student's t test at = 0.10) shown in Table 1 between experimental and control samples for stress at yield and strain at yield is not of practical importance. Even the total energy absorbed (determined 182M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN Figure 2. Stress-strain curves of cylindrical specimens tested in compression at high rate. Table 1. Salient mechanical features in high rate compression tests (mean std. dev.). Sample, N = Number of SpecimensModulus, GPa (Mean 1 Std. Dev.)Stress at Yield, MPa (Mean 1 Std. Dev.)Strain at Yield, % (Mean 1 Std. Dev.) Experimental (beads, tethered chains) N = 63.27 0.36131.72 1.916.64 0.43 Control (beads, silanized) N = 63.44 0.23138.67 1.636.35 0.09 Reference (no beads) N = 32.02 0.10119.54 3.227.68 0.60 from areas under the curves in Figure 2) was no greater for the experimental sam- ple than for the control. In light of the results presented in Part I, i.e., tethered chains produced a distinct positive effect on the interface in the single-fiber com- posite at high strain rate, the absence of effect in the particulate composite was dis- appointing. Below, we consider two possible explanations. The first possible explanation for the absence of any observed effect of tethered chains in the particulate composite is based on energy considerations. If the total energy absorbed by the specimen in a compression test were much larger than the energy consumed in rupture of the interfaces within the specimen, then the effect of the tethered chains would be too small, relative to the total energy, to be noticed. This possibility can be evaluated by consideration of the various energies in- volved. The literature provides values for debond energies of model polymer/glass interfaces with and without tethered chains [3]. These values, obtained at very low test rates, are about 1 J/m2 for an interface with no tethered chains and about 45 J/m2 for an interface with moderately dense tethered chains. A debond energy of 90 J/m2 was reported for an interface in which the tethered chains were attached at the maximum achievable density [4]. Since the reaction conditions we used for tethering chains to glass beads corresponded to those for the maximum achievable chain density, we adopt the value of 90 J/m2 for this discussion. (None of the quoted values contain the effect of increasing test rate, which would act to increase the debond energy between a rigid substrate and a viscoelastic matrix.) Estimates of energy consumed per specimen by interfacial debonding can be made from specimen geometry: the total interfacial area within a cylindrical specimen 10 cm in diameter, 17.5 cm in length, and containing 17% by volume of beads (each 10 m in diameter) is 0.14 m2. Thus, if all the interfaces in a specimen were debonded during the high-rate compression test, the difference between a specimen with tethered chains (experimental) and one without (control) would be at least (90 J/m2 1 J/m2) 0.14 m2 = 12.5 J. From the area under the stress-strain curve in Figure 2, the total energy absorbed by a typical bead-filled cylinder during com- pression to 60% of its original length is about 88 J. Since an additional 12.5 J or more would have been readily observed in the instrumented compression tests we conducted, the explanation that the tethered chains could not provide a noticeable effect is not the correct one. The other possible explanation for the absence of an observed effect is that only a small portion, or even none, of the interfaces within the specimen were stressed to rupture, i.e., to a point where the molecular-level, energy-dissipation mecha- nisms of the tethered chains were engaged. If few of the interfaces within the spec- imen were so stressed, then the energy difference discussed above would be much reduced, or even zero, and the stress-strain curves of the experimental and control samples would be indistinguishable. Electron microscopy of the fracture surfaces and optical microscopy of the unfractured interiors of the compression-tested specimens were used to evaluate this possibility. Improvement of Interfaces by Tethered Polymer Chains, Part II183 Microscopy Results Scanning electron micrographs of typical fracture surfaces are presented in Fig- ures 3 and 4 for an experimental and a control specimen, respectively. The fracture surfaces of the experimental and the control look the same; both show consider- able plastic deformation of the matrix and the complete debonding of the glass beads from the matrix. The complete separation of the beads from the matrix is proof that the glass bead/polysulfone interfaces in the fracture path were indeed stressed to rupture. However, the interfaces located within a bead's diameter of the fracture surfaces are only a small percent of the total number of interfaces in the entire specimen, and the interfaces remote from the fracture path must also be ex- amined. The interior planes of compression-tested specimens were examined by optical microscopy. It should be noted that, in specimens prepared by the typical metallographic procedures we used, even the slightest gap between a hard phase and a soft phase appears as a dark ring or arc. Close scrutiny of the polished cross-sections revealed complete material continuity between the glass beads and the matrix, with no hint of local deformation. This was taken as evidence that the glass bead/polysulfone interfaces away from the fracture surfaces were still fully bonded. Is it possible that the local stresses at these interfaces, not high enough to produce even the slightest debonding, were also unable to engage the molecu- lar-level, energy-dissipation mechanisms offered by the tethered chains? We think yes, and that the interfaces in both experimental and control samples experienced 184M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN Figure 3. Scanning electron micrograph of the fracture surface of experimental (with tethered chains) cylinder after compression testing at high rate. Magnification is 500. the same uncritical stresses (except at the fracture surfaces), thereby displaying in- distinguishable macroscopic behavior. CONCLUSIONS The effect of tethered chains, found to be beneficial to the glass/polysulfone in- terface in a model, single-fiber composite at both low and high strain rates, was evaluated for a more realistic, particulate composite at very high rate. For this evaluation, a process was developed whereby composite specimens could be fab- ricated without damage to the aliphatic linkage connecting the functional-ended polysulfone chains to the surface of the glass reinforcement. Analysis verified that the tethered chains not only survived the composite processing procedure, but also were intermingled with matrix chains. Thus, connection and entanglement, the conditions that are necessary for the molecular-level, energy-dissipation mecha- nisms to operate, were met. However, the results of high-strain-rate (150/s) com- pression tests of the composite specimens showed no observable effect of tethered chains. This result, although disappointing, was explained by the finding that only a small percent of the interfaces, i.e., those located on the fracture surfaces, were actually stressed to rupture in the high-rate compression tests. This suggests that most of the interfaces in the particulate composite were not sufficiently stressed to engage the molecular-level, energy-dissipation mechanisms of the tethered chains, and thus experimental and control samples were indistinguishable. These Improvement of Interfaces by Tethered Polymer Chains, Part II185 Figure 4. Scanning electron micrograph of the fracture surface of control (without tethered chains) cylinder after compression testing at high rate. Magnification is 500. results underscore the complexity of energy absorption within a specimen tested at very high rate, even when load application and specimen geometry seem simple. ACKNOWLEDGEMENTS This work was supported in part by the Office of Army Research Contract No. DAAH04-94-G-0331. The authors gratefully acknowledge the help of Dr. R. L. Lieb and Mr. M. G. Leodore, Mechanics and Structures Branch, Army Research Laboratory, Aberdeen Proving Ground, Maryland (high rate testing); Mr. Larry Rice, University of Kentucky (electron microscopy); and Ms. Aurora Rudell, Cen- ter for Applied Energy Research, Lexington, Kentucky (TGA/MS). REFERENCES 1. Greenfield, M.J., Hunter, T.F., Kalika, D.S. and L.S. Penn. 2001. J. Compos. Mater., 35(2): 164174. 2. Lieb, R.J. and M.G. Leadore. 1992. In Proceedings 1992 JANNAF Propulsion Systems Hazards Subcommittee Meeting, CPIA Publication 582, Vol.I, 145151. 3. Smith, J.W., Kramer, E.J., Xiao, F., Hui, C.Y., Reichert, W. and H.R. Brown. 1993. J. Mater. Sci., 25: 42344244. 4. Creton, C., Kramer, E.J., Hui, C.Y. and H.R. Brown. 1992. Macromol., 25: 30753088. 186M. J. GREENFIELD, T. F. HUNTER, D. S. KALIKA AND L. S. PENN 1. Greenfield, M.J. , Hunter, T.F. , Kalika, D.S. and L.S. Penn . 2001 . J. Compos. Mater. , 35 ( 2 ): 164 – 174 . 2. Lieb, R.J. and M.G. Leadore . 1992. In Proceedings 1992 JANNAF Propulsion Systems Hazards Subcommittee Meeting , CPIA Publication 582, Vol.I, 145 – 151 . 3. Smith, J.W. , Kramer, E.J. , Xiao, F. , Hui, C.Y. , Reichert, W. and H.R. Brown . 1993 . J. Mater. Sci. , 25 : 4234 – 4244 . 4. Creton, C. , Kramer, E.J. , Hui, C.Y. and H.R. Brown . 1992 . Macromol. , 25 : 3075 – 3088 .
ASJC Scopus subject areas
- Ceramics and Composites
- Mechanics of Materials
- Mechanical Engineering
- Materials Chemistry