8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
1/15
Composite Interfaces 18 (2011) 309322brill.nl/ci
A Fiber-Bundle Pull-out Test for Surface-Modified Glass
Fibers in GF/Polyester Composite
V. Cech a,,P. Janecek a,T. Lasota b and J. Bursa b
a Institute of Materials Chemistry, Brno University of Technology, CZ-612 00 Brno, Czech Republicb
Institute of Solid Mechanics, Mechatronics and Biomechanics, Brno University of Technology,
CZ-616 69 Brno, Czech Republic
Received 12 November 2010; accepted 16 March 2011
Abstract
The development of high-performance polymer composites is tightly bound with the functional surface
modification of reinforcements. A new method, based on the principle of the fiber-bundle pull-out test, is
proposed to analyze the interfacial properties between the long fibers in the form of a bundle and the polymer
matrix. Specimen geometry and a test fixture were designed using finite element analysis. The method
was verified for unsized and sized glass fibers embedded in polyester resin to demonstrate its applicability
for a wide range of adhesion between fibers and the polymer matrix. The pull-out test can be used for a
relative comparison of different surface modifications if the bundle geometry is unknown. The results ofhigh reproducibility and sensitivity for interfacial properties make the method attractive.
Koninklijke Brill NV, Leiden, 2011
Keywords
Polymermatrix composites (PMCs), interface/interphase, finite element analysis (FEA), electron mi-
croscopy
1. Introduction
Fiber-reinforced composites combine the high strength of rigid reinforcements with
the high toughness of a flexible matrix. As a result of the interphase region between
the reinforcement and the matrix, the synergism of the two phases produces me-
chanical properties of the composite material that cannot be achieved with either of
the constituents alone [13]. The development of high-performance fiber-reinforced
plastics (FRP) is linked to an effort to improve the properties of reinforcing fibers
and polymer matrices. The effort is directed at special materials, such as high-
modulus or high-strength fibers of stable mechanical properties and polymers of
high thermal or chemical resistance. Another way to improve composite perfor-
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
2/15
310 V. Cech et al./ Composite Interfaces 18 (2011) 309322
mance is advancement in engineered interphases, through surface modification of
fiber reinforcements, which better transfer the stress from the matrix to the fiber
reinforcement [46].
The rapid development of surface modifications of fiber reinforcements de-mands appropriate testing of adhesion between the fiber and the matrix. The sur-
face modification of fibers including surface activation, roughening, incorporation
of functional groups and grafting of molecules, together with fiber coating, can
be classified according to the technology used, e.g., wet chemical processes [7],
electrochemical [8] and ozone [9] treatments, high-energy irradiation (from a UV
lamp [10] or UV excimer laser [11] to gamma-ray irradiation [12]), low pres-
sure [13] and atmospheric pressure [14] plasmas. Plasma polymer films instead
of commercial sizing have been successfully tested for glass fibers [15].
The range of techniques for surface modification of fibers sketched in the lastparagraph together with the testing of new reinforcements and matrices gives an in-
dication of the huge quantity of samples to be analyzed for composite performance.
Many methods have been developed to measure fibermatrix adhesion in model
composites [16], where only one filament is embedded into the polymer matrix, or
high fiber volume fraction composites [17]. Generally, fibers are surface modified in
the form of bundles or even fabrics rather than single filaments. Shadowing effects
can be expected in central parts of bundles or 3D structures mainly if irradiation or
plasma technologies are employed. A different level of modification of filaments
extracted from a bundle or fabric may result in wide dispersion of measured valuesif single-fiber methods are used for the characterization of interfacial properties.
Batch rather than continuous surface modification is used in laboratories, which
means that there is often a lack of modified bundles that could be used to fabricate
composite lamina.
In this paper, we offer a simple method using a fiber bundle of short length to
characterize interfacial properties (fibermatrix adhesion). The method could be
classified as a fiber-bundle pull-out test and is a good analogy to the microbond
technique (fiber pull-out method) [18]. Specimen geometry and test conditions were
supported by modeling using finite element analysis, and examples of good and bad
adhesion between glass fibers (GF) and polyester resin are discussed. The short-
beam shear (ASTM D 2344) [19], as a representative of composite lamina methods,
was chosen to characterize the interfacial properties of the GF/polyester composite
and the results were compared with those of the fiber-bundle pull-out test.
2. Modeling
The specimen character was designed in an analogy of the microbond technique but
a fiber bundle was used instead of a single filament and a polymer droplet was sub-stituted by a polymer disk. The geometry of the specimen is given in Fig. 1(a). The
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
3/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 311
(a) (b)
Figure 1.(a) Specimen geometry (front, ground and profile view) and test fixture, (b) detail of the test
fixture. All dimensions are in millimeters.
shear strength at the bundle/disk interface. The specimen is then subjected to a ten-
sile test, where the dumbbell-shaped part of the specimen is held by grips and the
disk is held by a test fixture equipped with the edge, as shown in detail in Fig. 1(b).
The experiment consists of an increasing normal force applied to the fiber bundle
in order to pull it from the polymer disk.
The design of the test fixture, made of stainless steel, was optimized by finite
element analysis (Ansys 11.0 [20]) with respect to the specimen geometry and
mechanical properties of the materials used. The Youngs modulus, Poisson ratio,tensile strength () and yield strength (y) of the glass fiber were 70 GPa, 0.22
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
4/15
312 V. Cech et al./ Composite Interfaces 18 (2011) 309322
Figure 2.Influence of disk height on shear stress distribution along bundle/disk interface. Distance is
measured from the top of the disk.
of fibers impregnated by resin, in the narrow parallel part of the specimen between
the dumbbell and the disk, was 0.5 based on experimental results.
First, the model simulation was aimed at optimization of the height and diameter
of the polymer disk using deformation loading of the specimen at a displacement upto 3 mm. We followed the force in the fiber, which increased with increased height
of the disk, starting the simulation at a height of 2.0 mm; the disk diameter was
10.0 mm. The increase in force was weak for heights of 5.0 mm and over. Similar
results were obtained by following the shear stress along bundle/disk interface as a
function of disk height (Fig. 2) using the force loading that ranged up to 1500 N.
The maximum shear stress is found very close to the top surface of the disk for all
the distributions. Again we can see (Fig. 2) that the distribution of shear stress in
the polymer disk changed only slightly when the height was 5.0 mm. This means
that slight variations of disk height, as a result of disk fabrication, will not influencethe test character significantly. Our calculations showed that the force in the fiber
was not influenced by disk diameter variations from 10.0 to 16.0 mm. The suitable
height of the polymer disk seems to be 5.0 mm or above. Larger values would in-
crease the load necessary to overcome the adhesion between the fiber bundle and
the polymer matrix in the disk and could result in a fracture of the specimen outside
the disk.
Next, in order to optimize the edge of the test fixture, we conducted simulations
for a polymer disk at a height of 5.0 mm and a diameter of 10.0 mm (Fig. 1(b)). The
deformation loading of the specimen at a displacement up to 3 mm was applied tocalculate the force in the fiber as a function of the inner diameter (1.21.8 mm) of
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
5/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 313
Figure 3.Distribution of von Mises stress in a specimen. The bar values are in MPa.
shear stress along the bundle/disk interface was also not influenced by the edge
when a height of 2.0 mm and an inner radius of the rounded edge of 1.5 mm were
applied for simulations. Then the locus of maximum von Mises stress is localized at
a meniscus (radius 0.2 mm) formed at the top surface of the polymer disk and the
impregnated fiber bundle, as shown in Fig. 3 (see arrow); this is thus the locus of a
possible shear failure. The contact pressure under the rounded edge was minimized
(Fig. 4) as a result of the optimized edge rounding to prevent the formation of
possible flaws.
Finally, then, we can state on the basis of model simulations that the polymer
disk has to be of a height 5.0 mm and a diameter 10.0 mm. The edge of a test
fixture with an inner diameter of 1.5 mm and height of 2.0 mm has to be rounded
with a 1.5 mm inner radius and 1.0 mm outer radius (Fig. 1).
3 E i t l
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
6/15
314 V. Cech et al./ Composite Interfaces 18 (2011) 309322
Figure 4.Contact pressure under the rounded edge. The bar values are in MPa.
developed for polyester resin and bundles unsized (water sized) were supplied bySaint-Gobain Vertex, Czech Rep. The matrix material used in this study was un-
saturated polyester resin VIAPAL VUP 4649 E(M) on a base of isophthalic acid
(Vianova Kunstharz, Austria) with a density (m) of 1.13 g/cm3. The resin was
mixed with hardener (styrene), a low-temperature initiator, Perkadox 16 (Akzo
Nobel Chemicals, The Netherlands), a high-temperature initiator, Norpol No. 62
(Norac Andos, Sweden), a UV stabilizer (3V Sigma S.p.A., Italy) and an internal
lubricant INT-PUL24 (AXEL, USA).
Sufficiently stirred resin with the curing components was evacuated for 5 min
to avoid air entrapment. First, a bundle of fibers was impregnated with the resinand extra resin carefully wiped from the bundle. The impregnated bundle was po-
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
7/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 315
dumbbell-shaped part of the specimen was formed using the rubber mold and cur-
ing process. Fibers protruding from the disk were cut and the bottom part of the
polymer disk polished using emery paper (220 m particles) without water. A set
of five specimens was made for each fiber surface modification.The pull-out test was carried out in a universal testing machine Zwick Z010/
TH2A at room temperature. The specimen was clamped using screw grips at the
top (dumbbell) and loaded against a supporting test fixture, placed in a frame
clamped by screw grips, which held the bottom part (polymer disk). The load
displacement curves were recorded during the tests using a set of five specimens of
specific surface modification. The experiments were conducted at a constant cross-
head speed of either 1 mm/min or 10 mm/min in order to study a possible influence
of the strain rate on the pull-out behavior. The fiber bundle pulled out of the poly-
mer disk was examined using a scanning electron microscope (SEM) (Philips XL30/EDAX/Microspec).
Hand lay-up composites, reinforced with unsized and sized fibers, were fabri-
cated for measurement by short-beam shear test. Five specimens, 10 mm wide,
6 mm thick and 35 mm long, were prepared for each fiber surface modification
and cured at 140C. The beams were polished with emery paper (220 m particle
size) without water and stored in a desiccator before further testing. Short-beam
composites were tested in a three-point bending test according to the configuration
designated in ASTM D 2344 [19], where R1=6.0 mm and R
2=3.0 mm. The
span length to specimen thickness ratio was 4. Five specimens were tested for each
type of fiber reinforcement. In the short-beam shear test, the maximum value max(short-beam strength) of the shear stress along the thickness direction is related to
the maximum applied load, Pmax, specimen width b and thickness t, according to
the short-beam shear relationship:
max =3Pmax/(4bt). (1)
The test speed was set at a cross-head movement of 1 mm/min and the force applied
on the specimens was monitored using a universal testing machine.
4. Results and Discussion
The typical loaddisplacement curves corresponding to sized GF are shown in
Fig. 5. The curves were measured at a cross-head speed of 1 mm/min and were
of linear character except for a number of kinks due to a break of some fibers as
a result of the stress distribution in the bundle. The slope and the maximum ap-
plied load were very similar for all the dependences. The maximum applied load
is given in Table 1; the average maximum load was estimated as 611 20 N at amean displacement of 2.3 mm. The maximum applied load,Pmax, corresponds to a
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
8/15
316 V. Cech et al./ Composite Interfaces 18 (2011) 309322
Figure 5.Loaddisplacement curves for glass fibers surface modified by silane-based sizing.
Table 1.
Interfacial properties of sized glass fibers in polyester matrix using the fiber-bundle pull-out test
Sample no. Disk height Maximum Specific maximum IFSS
(mm) load (N) load (N/mm) (MPa)
1 5.07 598 118 41.8
2 5.15 634 123 43.5
3 5.00 625 125 44.2
4 4.94 588 119 42.1
5 5.09 595 117 41.6
Mean SD 5.05 0.08 611 20 121 3 42.9 1.1
crobond technique, the interfacial shear strength can be calculated as the maximum
applied load divided by the contact area using the relation:
int =Pmax/(hd). (2)
However, the bundle diameter can be evaluated only as an approximate value, in
contrast to the fiber diameter in the microbond technique. The specific maximum
load can be used to eliminate a different disk height, due to disk polishing, if the
maximum load is divided by the disk height as given in Table 1. In most cases
the standard deviation (SD) of the disk height is proportionally smaller than the
SD of the maximum load, and calculation of the specific maximum load is thusnot necessary. A bundle diameter of 0.9 mm was used to estimate the interfacial
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
9/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 317
decreased with greater fabrication experience and the ten sets of five specimens
resulted in a 3% SD as given in Table 1.
Similar results were obtained for a cross-head speed of 10 mm/min. The slope of
the loaddisplacement curves was the same but the maximum applied load was 11%higher and the SD was also higher (6%) than with a cross-head speed of 1 mm/min.
Thus with respect to the lower SD, the cross-head speed of 1 mm/min seemed to be
more suitable and was used for all other measurements. A force loading of 1500 N
and deformation loading of 3 mm used in the model simulations are above the
experimental values.
The above results on sized GF came not solely from mechanical bonding and
van der Waals forces but predominantly from chemical bonding between the
glass reinforcements and polyester matrix. Silane-based sizing P707 is one of the
organofunctional silane coupling agents [21] recommended for surface modifica-tion of glass fibers used as reinforcements for unsaturated polyester resin. The
monomer molecule (RSiX3) is a multifunctional one, which reacts at one end with
the glass surface and at the other with the polymer matrix. The X group repre-
sents a hydrolyzable unit, and thus the silane is hydrolyzed to the corresponding
silanol in the aqueous solution to which the glass fibers are exposed. These silanol
molecules compete with water molecules to form hydrogen bonds with the hydroxyl
groups present at the glass surface. When the glass fiber is dried, the free water is
driven off and condensation reactions then occur, both at the silanol/glass interface
and between neighboring silanol molecules. The result is a polysiloxane monolayer
bonded to the glass surface, presenting an array of R groups oriented outwards. If
the matrix is to be a thermosetting resin, such as unsaturated polyester resin, then
the vinyl group reacts with the resin during polymerization [21]. In principle, such
a polysiloxane monolayer should provide a strong bond between the reinforcement
and the polymer matrix. However, in practice the polycondensed film is hetero-
geneous [4] with respect to uniformity and film thickness [22]. Thus the surface
roughness could measure hundreds of nanometers or even several microns. The
molecules of silane coupling agents have a tendency to self-condensation, formingsiloxane oligomers [23, 24] rather than complete bonding with the glass surface,
which results in a lower density of siloxane bonds with the glass surface. It is for
this reason that new ways are being developed to find a more effective sizing, with
many techniques employed for surface modification of reinforcements, not only
glass fibers.
The maximum load or the specific maximum load can be used for comparison
of interfacial properties of fibers of different surface modification if a bundle of the
same number of fibers and with fibers of the same diameter is used. A bundle of
unsized GF means an opposite extreme, where the maximum load should be of alower value due to mechanical bonding and van der Waals forces. The bundle of
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
10/15
318 V. Cech et al./ Composite Interfaces 18 (2011) 309322
Figure 6.Loaddisplacement curves for unsized (water-sized) glass fibers.
Table 2.
Interfacial properties of unsized glass fibers in polyester matrix using the fiber-bundle pull-out test
Sample no. Disk height Maximum Specific maximum IFSS
(mm) load (N) load (N/mm) (MPa)
1 5.02 257 51.2 18.1
2 4.90 269 54.9 19.4
3 4.98 259 52.0 18.4
4 5.11 269 52.6 18.6
5 4.89 257 52.6 18.6
Mean SD 4.98 0.09 264 6 52.7 1.4 18.6 0.5
sized GF measured by optical polarizing microscope (BX-P 50 Olympus). Typicalloaddisplacement curves for unsized GF are shown in Fig. 6; a cross-head speed of
1 mm/min was used. The slope of plotted data is very similar to that for sized fibers,
but the average maximum load of 264 6 N (Table 2) is lower by 57% compared
with sized GF. The maximum load was a reproducible parameter and the standard
deviation was again only 3%. The interfacial shear strength can be evaluated as
19 MPa if a bundle diameter of 0.9 mm is used. The lower maximum load for
unsized GF resulted mainly from a weaker interface but we suspect that there could
be some specimen imperfections. It can be envisioned that the resin will not readily
wet the fibers due to high surface tension of the glass and this could lead to poorinfiltration of the bundle creating voids that would serve as stress concentrators.
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
11/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 319
(a) (b)
Figure 7. Micrograph of a pulled-out bundle of sized glass fibers: (a) multi-fiber view, (b) detailed
view.
(a) (b)
Figure 8.Micrograph of a pulled-out bundle of unsized glass fibers: (a) multi-fiber view, (b) detailed
view.
and 8(a)) revealed a lack of polyester resin among the unsized fibers, but some
unsized fibers were broken, which agrees with idea of stress concentrators. The un-
sized fibers in Fig. 8(b) are mostly bare without residues of adhered resin on thefiber surface; only a small amount of resin debris can be found on fibers indicating
a weak interface at the smooth fiber surface. However, the sized fiber in Fig. 7(b)
is still covered by sizing or polyester resin and the character together with the large
amount of debris on and among the fibers give evidence for the higher adhesion and
cohesive-like resin failure.
A good agreement between the interfacial shear strength (IFSS), measured by the
single fiber fragmentation test, and the interlaminar shear strength (ILSS), measured
by the short-beam shear test, was noted by Drzal et al. [16]. Unsized and sized
GF were used to fabricate short beams and the interlaminar shear strength wasevaluated. In Fig. 9, the results are compared with the interfacial shear strength
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
12/15
320 V. Cech et al./ Composite Interfaces 18 (2011) 309322
Figure 9. Interfacial shear strength (IFSS) and interlaminar shear strength (ILSS) for unsized andsized glass fibers.
Similar to the microbond technique, it is assumed that the interfacial shear
strength is uniformly distributed along the embedded length of the fiber bundle.
A state of stress along the embedded length of fiber in a microdroplet has been dis-
cussed many times for the microbond technique and a change in microdroplet size
together with the presence of a meniscus formed by the matrix at its point of con-
tact with the fiber have been mentioned as influencing interfacial stress distribution.
Uncertainty in the polymer disk size and shape together with the minimized menis-cus formed at the top surface of the polymer disk and the impregnated fiber bundle
have been eliminated significantly in the case of the fiber-bundle pull-out test.
5. Conclusion
Surface modification of reinforcements controls the performance of polymer com-
posites via the interface/interphase region. Recent progress in technologies for
surface modification of fiber reinforcements demands a simple and reliable tech-
nique to characterize adhesion between fibers and the polymer matrix.A new method enabling evaluation of interfacial properties is proposed and ap-
plied for glass fibers of two different surface modifications. The method is based
on the principle of the fiber-bundle pull-out test, where a bundle of 15 cm is suffi-
ciently long for specimen fabrication and even a local surface modification 10 mm
along the bundle is sufficient for proper characterization of interfacial properties.
Specimen geometry and a test fixture design were proposed on the basis of model
simulations using finite element analysis. The fiber-bundle pull-out test was verified
for unsized and sized glass fibers embedded in polyester resin. The fiber surface
modified by commercial silane-based sizing was used as an example of good ad-hesion due to strong chemical bonding. Conversely, the unsized glass fibers with
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
13/15
V. Cechet al./ Composite Interfaces 18 (2011) 309322 321
relative comparison of different surface modifications if a bundle of the same num-
ber of fibers and with fibers of the same diameter is used. The reproducibility of
the results was high with 3% standard deviation, and the interfacial properties can
be characterized with sufficient sensitivity. The proposed test can be extended forother polymer composites of a specific fibermatrix system.
Acknowledgements
This work was supported by the Czech Ministry of Education, grant No. ME09061.
The authors would also like to thank Dr. M. Sirovy (Saint-Gobain Vertex, Lito-
mysl) for providing glass fibers and Dr. J. Prokes (Prefa Kompozity, A.S., Brno) for
chemical substances.
References
1. D. R. Askeland, The Science and Engineering of Materials. Chapman and Hall, London, UK
(1984).
2. M. Ohring,Engineering Materials Science. Academic Press, San Diego, USA (1995).
3. P. K. Mallick, Fiber-Reinforced Composites. CRC Press, New York, USA (2008).
4. L. J. Broutman and B. D. Agarwal, A theoretical study of the effect of an interface on the properties
of composites,Polym. Engng Sci.14, 581588 (1974).
5. M. Labronici and H. Ishida, Toughening composites by fiber coating. A review, Compos. Inter-
faces2, 199234 (1994).
6. J. K. Kim and Y. W. Mai, Engineered Interfaces in Fiber Reinforced Composites. Elsevier, Ams-
terdam, The Netherlands (1998).
7. Z. Xu, L. Chen, Y. Huang, J. Li, X. Wu, X. Li and Y. Jiao, Wettability of carbon fibers modified
by acrylic acid and interface properties of carbon fiber/epoxy,Eur. Polym. J.44, 494503 (2008).
8. H. L. Cao, Y. D. Huang, Z. Q. Zhang and J. T. Sun, Uniform modification of carbon fibers surface
in 3-D fabrics using intermittent electrochemical treatment,Compos. Sci. Technol.65, 16551662
(2005).
9. S. J. Park and B. J. Kim, Roles of acidic functional groups of carbon fiber surfaces in enhancinginterfacial adhesion behavior,Mater. Sci. EngngA408, 269273 (2005).
10. E. Mader, S. Melcher, J. W. Liu, S. L. Gao, A. D. Bianchi, S. Zherlitsyn and J. Wosnitza, Adhesion
of PBO fiber in epoxy composites,J. Mater. Sci.42, 80478052 (2007).
11. J. Zeng and A. N. Netravali, Effects of XeCl excimer laser treatment of vectrano fibers and their
adhesion to epoxy resin, J. Adhes. Sci. Technol.20, 387409 (2006).
12. Z. Xu, Y. Huang, C. Zhang, L. Liu, Y. Zhang and L. Wang, Effect of gamma-ray irradiation
grafting on the carbon fibers and interfacial adhesion of epoxy composites,Compos. Sci. Technol.
67, 32613270 (2007).
13. M. A. Montes-Moran, F. W. J. van Hattum, J. P. Nunes, A. Martinez-Alonso, J. M. D. Tascon and
C. A. Bernardo, A study of the effect of plasma treatment on the interfacial properties of carbon
fibre-thermoplastic composites,Carbon43, 17951799 (2005).
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
14/15
322 V. Cech et al./ Composite Interfaces 18 (2011) 309322
15. V. Cech, Plasma-polymerized organosilicones as engineered interlayers in glass fiber/polyester
composites,Compos. Interfaces14, 321334 (2007).
16. L. T. Drzal, P. J. Herrera-Franco and H. Ho, in: Comprehensive Composite Materials, A. Kelly
and C. Zweben (Eds), Chapter 5.05. Elsevier, Amsterdam, The Netherlands (2000).17. D. F. Adams, in: Comprehensive Composite Materials, A. Kelly and C. Zweben (Eds), Chapter
5.06. Elsevier, Amsterdam, The Netherlands (2000).
18. A. Hodzic, S. Kalyanasundaram, A. Lowe and Z. H. Stachurski, The microdroplet test: experi-
mental and finite element analysis of the dependence of failure mode on droplet shape, Compos.
Interfaces6, 375389 (1999).
19. ASTM D2344, Standard test method for short-beam strength of polymer matrix composite mate-
rials and their laminates.
20. Software Ansys 11.0, Release 11.0, Documentation for ANSYS.
21. J. G. Marsden, in: Handbook of Adhesives, I. Skeist (Ed.), pp. 536548. Van Nostrand Reinhold,
New York, USA (1990).
22. D. W. Dwigh, in: Comprehensive Composite Materials, A. Kelly and C. Zweben (Eds), Chap-
ter 1.08. Elsevier, Amsterdam, The Netherlands (2000).
23. G. M. Nishioka, Interaction of organosilanes with glass fibers, J. Non-Cryst. Solids120, 102107
(1990).
24. W. Wang and A. T. DiBenedetto, A modified silane treatment for superior hydrolytic stability of
glass reinforced composites,J. Adhesion68, 183201 (1999).
8/9/2019 A Fiber-Bundle Pull-out Test for Surface-Modified Glass
15/15
Copyright of Composite Interfaces is the property of VSP International Science Publishers and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.