Professional Documents
Culture Documents
ISBN 978-1-56990-627-9
E-Book ISBN 978-1-56990-628-6
All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or me-
chanical, including photocopying or by any information storage and retrieval system, without permission in writing from the
publisher.
Short summary 6
Kurzzusammenfassung 7
1. Introduction 8
1.1. Motivation 8
1.1.1. References 8
2.1.3. Pultrusion 13
2.1.6. References 16
2.2.4. Preforms 19
2.2.6. References 21
1
2.3. Epoxy resin systems for fiber-reinforced composites 22
2.3.4 References 27
2.4.6. References 33
2.5.5. References 45
2
3.1.1. Abstract 48
3.1.2. Introduction 48
3.1.3. Discussion 51
3.1.4. Conclusions 62
3.1.5. References 63
3.2.1. Abstract 66
3.2.2. Introduction 66
3.2.3. Discussion 69
3.2.4. Conclusions 82
3
3.2.5. References 83
3.3.1. Abstract 86
3.3.2. Introduction 86
3.3.3. Discussion 89
4
3.5. Carbon fiber-reinforced composites with epoxy resin modified with
reactive liquid rubber and SiO2 nanoparticles 128
5. Bibliography 163
5
Short Summary
Combining this modification with the classic toughening concept using reactive
liquid rubbers or core-shell elastomers leads to so-called hybrid systems, which
are characterized by both high toughness and high stiffness. The fatigue
performance is improved further.
In own trials a fast curing epoxy resins system based on DGEBA/IPD/TMD was
employed to manufacture carbon fiber reinforced laminates; whose properties
were investigated. The cured bulk resin systems exhibit the expected property
improvements for the hybrid systems. However, the laminates based on hybrid
resin systems, modified with both reactive liquid rubber and SiO2 nanoparticles,
show no further improvements compared to the rubber-only modification but
rather slightly lower values for GIc and GIIc. The ILSS is comparable; the residual
strength after impact reduced.
In summary it can be stated that laminates made from hybrid epoxy resins are
tough and stiff and exhibit improved compressive strength as well as excellent
fatigue performance. This makes them especially suitable for highly stressed
composites parts like in automotive applications.
6
Kurzzusammenfassung
Die Eigenschaften von gehärteten Epoxidharzen können durch den Zusatz von
oberflächenmodifizierten SiO2-Nanopartikeln deutlich verbessert werden. Es
besteht für die meisten Härter ein annähernd linearer Zusammenhang zwischen
der Steigerung des Moduls bzw. der Bruchzähigkeit und der Zusatzmenge an
Nanopartikeln. Druckbeständigkeit und Ermüdungsverhalten können ebenfalls
verbessert werden.
Beim Einsatz dieser modifizierten Harze zur Herstellung von Laminaten können
an diesen ebenfalls Eigenschaftsverbesserungen beobachtet werden. Unabhängig
davon, ob Glas- oder Carbonfasern als Verstärkungsmaterial eingesetzt werden,
fallen diese Verbesserungen prozentual deutlich geringer aus. Es scheint eine
lineare Beziehung zwischen den prozentualen Steigerungen der Bruchzähigkeit
(GIc) der gehärteten Reinharzhybridsysteme und den daraus hergestellten
Laminaten zu existieren; mit einem Übertragungsfaktor von 0,18. Soll ein
Faserverbundbauteil um 100 % in der Zähigkeit gesteigert werden, so muss statt
des bisher verwendeten Harzsystems ein Hybridsystem mit einem 555 %
höheren GIc eingesetzt werden. Beim Ermüdungsverhalten der Laminate auf
Hybridharzbasis kann eine zehnfache Steigerung der zyklischen Belastung bis
zum Versagen erzielt werden.
7
1. Introduction
1.1. Motivation
1.1.1. References
8
1.2. Objectives and structure
First the modification of epoxy resins with SiO2 nanoparticles regarding the
achievable property improvements of cured bulk resins needs to be investigated
– this will be the subject of chapter 3.1. Secondly, the so-called hybrid epoxy
resins, that is epoxy resins modified with both a rubbery (i.e. elastomeric)
toughener and silica nanoparticles, will be evaluated with a focus on mechanical
properties in chapter 3.2. Thirdly, the use of nanosilica-modified epoxy resins in
fiber-reinforced composites based on glass or carbon fibers is the subject of
further investigations in chapter 3.3. One of the essential questions here is
whether property improvements found for modified bulk epoxy resins will be
found for fiber-reinforced composites as well. Fourthly, of course the same
question is even more important for fiber-reinforced composites made from
hybrid epoxy resins which will be investigated in chapter 3.4.
Finally, the insights obtained from the above studies will be verified by
investigating a bulk epoxy resin system and a composite made by using a
common resin system of major industrial importance in chapter 3.5. A carbon
fiber-reinforced system based on a fast-curing epoxy resin system, as typically
used for manufacturing automotive parts, was selected to perform this task.
To obtain a clear structure of this work it is divided in these five work packages
(chapters 3.1. - 3.5.). To ensure a correct interpretation of the many different
results from many individual sources as well as the results of my own work the
single work packages have been published up front in renowned scientific
journals.
As each work package has is a separate chapter and coherent in itself, the
necessity arose to use graphs or pictures several times. Examples are the SEM
picture of core shell elastomers used in chapters 2.4.3. and 3.2., or the SANS
curve for the particle size distribution presented in chapters 2.5.1. and 3.3., or
the TEM picture of the silica nanoparticles shown in chapters 2.5.1. and 3.2. This
recurrence is necessary to give a clear and readable structure. In the
bibliography the cross-references are given.
9
2. State-of-the-art of the science and technology
10
2.1.1. Prepreg methods
Here woven or nonwoven fabrics (the latter often being called non-crimp fabrics -
see chapters 2.2.2. and 2.2.3.) are impregnated with a blend of epoxy resin and
curing agent and precured to a certain degree of crosslinking. Using a film
transfer process a matrix film is first formed using the resin/hardener blend. In a
second step this film is transferred from the release paper to the fiber
architecture (i.e. the textile) which has been selected for the composite (Figure
2.1.2). The classic solvent-based solution route uses a solution of the
resin/hardener blend in a volatile solvent, e.g. methylethylketone. After
impregnating the fibers the solvent is evaporated in a vertical or horizontal dryer
(see Figure 2.1.2). Another variation is the hot-melt process, where the
resin/hardener blend is applied as a melt. In all these three processes the resin
systems are partially polymerized, to a relatively low degree, at an elevated
temperature (so-called B-stageing). The final prepreg is coiled up and stored in a
refrigerator.
11
Later, very often in a different facility, the prepregs are cut by hand or by robot
and laid into the form of the end-product. Then the epoxy resin is fully cured;
e.g. at 135 °C or 190 °C, depending on the choice of hardener. This can be done
in a vacuum, at room pressure, in a press, using vacuum bags or, typically for
aerospace parts, in an autoclave.
The flow behaviour of the precured epoxy resin at higher temperatures is very
critical for this process. Fillers are used rarely. Nanofillers, of course, can be
used.
The winding process is especially suitable for pipes, tubes used as ski poles or
masts of sailing boats as well as any other rotationally-symmetrical parts such as
tanks or pressure vessels for gases. The fiber rovings are passed through a
resin/hardener liquid blend and then wound around a mandrel; followed by being
cured at elevated temperatures.
2.1.3. Pultrusion
Most important are injection processes such as resin transfer moulding (RTM) or
vacuum assisted resin transfer moulding (VARTM). They are very suitable for the
semiautomatic production of relatively complex composite parts. A fiber textile,
from simple layers of fabric to complex preforms or layups with core materials
such as foams or honeycomb structures, is placed in a mold. Then the
resin/hardener blend is injected and cured at an elevated temperature. Figure
2.1.5 shows schematically the process. Of great importance in the RTM process
is the need for low resin viscosities of 100-500 mPas at the injection temperature
to obtain good fiber wetting and reasonably long flow distances for the injected
resin/hardener blend.
13
Fillers which would improve mechanical properties like compressive strength
cannot be used, as they would be filtered out at the first layer of fibers.
RTM, VARTM and similar processes, such as the wet-press process, are suitable
for the high volume production of fiber-reinforced composite parts. Even complex
geometries can be realized – since then, instead of simple layers of fabrics,
three-dimensional layups or preforms are placed in the mold. Figure 2.1.6 shows
a small series automotive part (sports car bracket):
14
Consequently injection processes are widespread in the composites industry. An
example of an automotive application is the convertible soft top cover of the VW
Eos. Also, the body of the electric BMW i3 is made entirely from composite parts,
which are made using RTM from preforms and assembled using structural
adhesives [2.1.13]. The use of preforms or very complex layups, which was
considered to be exotic a few years ago, is now an industrial standard practice.
Infusion processes, such as the vacuum assisted resin infusion process (VARI)
which is used to manufacture very large rotor blades for wind energy generators,
are of also great importance in the industry. The fabrics, unidirectional or
multiaxial textile layups are placed in an open mold, as can be seen in Figure
2.1.7. The fibers are covered with a so-called vacuum bag, which is sealed
hermetically at the edges. Subsequently the resin/hardener blend is sucked at
different entry points using a vacuum line into the textile and the resin/hardener
blend then impregnates the fibers. Cure can be undertaken at room temperature
or moderately elevated temperatures. Relatively low resin viscosities are
required. Thus, as for with above injection processes, fillers cannot be used in
the resin formulations.
Figure 2.1.7: Boat hull made from carbon fibers using VARI - picture courtesy of
NORCO GRP Ltd [2.1.14]
15
2.1.6. References
[2.1.4] Hillermeier R., Hasson T., Friedrich L., Ball C. (2013) Advanced
thermosetting resin matrix technology for next generation high
volume manufacture of automotive composite structures SAE
Technical Paper 2013-01-1176. doi:10.4271/2013-01-1176
[2.1.5] Ohlendorf, J.H., Rolbiecki, M., Schmohl, T.; Franke, J. Thoben, K.D.,
Ischtschuk, L. (2013) Innovationen in der Handhabungs- und Textil-
technik zur Rotorblattfertigung, Lightweight Design 6 (5), 50-57
16
2.2. Fiber reinforcements for epoxy resin systems
For both injection and infusion processes (as well as in the manufacture of
prepregs) fabrics and non-crimp fabrics, such as unidirectional or multiaxial
fabrics, are used. In a few applications fiber mats with randomly oriented glass
or carbon fibers are processed using epoxy resin systems.
17
Fabrics with different weavings can be made from glass and carbon fibers. The
weavings can be made in the form of, for example, plain or linen-type, twill,
satin and many others. Figure 2.2.1 shows a typical example. The different
weavings not only lead to different mechanical properties but can be especially
important for esthetic aspects in composites parts, where the reinforcement is
shown intentionally for decorative purposes.
In unidirectional fabrics (UD), the fibers are all oriented parallel to each other
and are fixed in place by an extremely thin sewing thread which is typically a
polyester yarn. In the example of a UD made from carbon fibers shown in Figure
2.2.2 this can be seen very well.
A typical application for unidirectional fabrics made from glass fibers are the
horizontal beams of rotor blades for wind turbines, as they have to carry the
maximum load. An automotive application is the leaf spring for light trucks, such
as the Mercedes Sprinter.
18
2.2.3. Multiaxial fabrics
The next step in the development of such textiles has been the multiaxial fabrics.
Here the orientation of the fiber layers to each other and the number of layers
can be chosen practically without any restriction. Figure 2.2.3 shows an example
of such a multiaxial textile layup. At the right side the stitching needles applying
the fixing thread can be seen.
2.2.4. Preforms
19
Figure 2.2.4: Braiding of a tubular preform using carbon fiber rovings [2.2.8]
Figure 2.2.5: Body in white of the BMW i3; manufactured by using preforms
which are impregnated subsequently in a HP-RTM process [2.2.9]
All commercial glass and carbon fibers are pretreated with a coating, the so-
called fiber sizing by the manufacturer. This coating has several different
functions. It is supposed to protect the fiber during processing, e.g. during the
weaving process.
20
Furthermore it should significantly increase the adhesion between the fiber and
the resin matrix. Sizings for epoxy resins typically contain epoxy functional
silanes as adhesion promoters. Most fiber manufacturers formulate their sizings
themselves and consider them as part of their core technology. Hence it is
difficult to obtain detailed information regarding their composition. Indeed, a
further detailed consideration would go beyond the scope of this dissertation.
However, as an introduction to this topic, I would like to recommend the recently
published review by Thomason [2.2.10].
2.2.6. References
[2.2.3] Market Research Report (2012) Global Glass Fiber Market 2012-
-2017: Trend, Forecast and Opportunity Analysis, Lucintel, Las
Colinas, TX, USA
[2.2.4] Jahn, B., Witten, E. (2013) Composites Market Report 2013. Carbon
Composites e.V., September 2013, Augsburg, Germany
[2.2.10] Thomason, J.L. (2013) Glass Fiber Sizings: A Review of the Scientific
Literature. ISBN 978-0-9573814-0-7, Create Space, UK
21
2.3. Epoxy resin systems for fiber-reinforced composites
Epoxy resins are one of the most important class of thermosetting polymers,
along with polyurethanes (PU), unsaturated polyester resins (UPE), vinyl ester
resins and phenolic resins. Since their introduction in the 1950s they have been
used for many very different industrial applications. For example, they are
employed as adhesives, electrical encapsulants, floor coatings, fiber-reinforced
composites, heavy duty marine coatings, repair pastes, abrasive systems and
have been used in aircraft, train and car manufacturing, electronic assembling
and jewelry - indeed the abundance of applications is almost endless. A
comparison of various matrix resin systems and their usability for fiber-
reinforced composites is given by Hasson [2.3.1].
In the year 2009 1.8 million tons of epoxy resin were used worldwide; 52.000
tons in fiber-reinforced composites. This equals a quite small share of only 2.9 %
[2.3.2]. For the years 2009 to 2014 an annual increase in epoxy resin demand of
5.8 % is expected so that the consumption in 2014 will be approximatively 2.4
million metric tons. The consumption in composite applications however is
growing much faster by 8-10 % per year and will grow even faster if the
automotive manufacturers develop new, large volume, composite parts. The
epoxy resins and associated hardeners relevant for fiber-reinforced composites
will be introduced briefly below. For a more detailed study, the work of Ellis
[2.3.3] is recommended.
The working horse of the epoxy resin industry is the diglycidylether of bisphenol
A (DGEBA). In composite applications it is by far the most commonly used type
of epoxy resin. Figure 2.3.1 gives an overview of the chemical structures of the
epoxy resins relevant for fiber-reinforced composites. Higher molecular weight
homologues of DGEBA, so-called solid epoxy resins, are frequently used in
prepreg manufacturing. The diglycidylether of bisphenol F (DGEBF) is used in
formulations were low viscosities are necessary. The higher homologues of
DGEBF, epoxidized novolac resins, are used in applications were increased
thermal and chemical stabilities are required. Due to their higher epoxy
functionality they form closer meshed networks which consequently possess a
higher glass transition temperature (Tg). Short chain aliphatic epoxy resins, such
as hexandiol diglycidylether (HDDGE) or butanediol diglycidylether (BDGE) are
introduced into formulations when viscosities need to be lowered. Aliphatic epoxy
resins with longer molecular chains, such as epoxidized polyether polyols or
epoxidized ethoxylated derivatives of trimethylolpropane are occasionally added
to flexibilize resin systems. They lower the crosslink density and consequently
the Tg decreases. Cycloaliphatic epoxy resins, such as hydrogenated DGEBA or
3,4-epoxycyclohexane carboxylic acid -3,4- epoxycyclohexane methylester (EEC),
are currently not relevant for fiber-reinforced composites applications. However,
very important are the higher functional resins, such as the triglycidyl-p-
aminophenol (TGpAP) and tetraglycidyl methylendianiline (TGMDA), and without
these resins most aerospace applications would not be possible.
22
Diglycidylether of bisphenol A Solid DGEBA resin (higher homologue, n=1)
(DGEBA)
The hardeners relevant for composites applications can be divided roughly into
amines and anhydrides. Catalytic curing agents such as BF3-complexes or UV
initiated cationic curing agents, are rarely used. Figure 2.3.2 gives an overview
of the chemical structures of the most important hardeners for fiber-reinforced
composites. For further reading the papers published by Hare [2.3.4], [2.3.5] are
to be recommended.
One of the most important amines is the cycloaliphatic isophorone diamine (IPD)
which is a major component for most infusion resin systems. It provides an
excellent balance between giving a relatively high crosslink density and a fast
reaction speed.
Linear polyether diamines are used as flexibilizers. Such amine formulations are
typically cured at relatively low temperatures – therefore these systems do not
exhibit very high glass transition temperatures (Tgs).
Aromatic amines are used for epoxy resin systems which require high glass
transition temperatures due to high service temperatures being experienced by
the composite parts, as in many aerospace applications. An important hardener
is the 4,4´-diaminodiphenylsulfone (DDS). It is used in prepregs as well as in
injection or infusion resin systems. Other important aromatic amines are the
4,4´-methylenbis-2,6-diethylaniline (M-DEA) and the 4,4´-Methylenbis-2,6-diiso-
propylaniline. A blend of both is used to formulate the benchmark for aerospace
injection resins, i.e. the famous RTM 6 resin/hardener blend.
24
Isophorondiamine (IPD) Polyetherdiamine
Dicyandiamide (DICY)
Figure 2.3.2: Chemical structures of the most relevant epoxy curing agents
25
2.3.3. Epoxy resin systems
Typically for filament winding processes, e.g. for ski pole manufacturing, the
combination of DGEBA with an anhydride curing agent is common. If the
performance requirements of the composite part increase, other resin systems
are used. For example, pressure vessels made by filament winding are made
from epoxidized novolac resins and anhydrides or DDS.
In RTM processes very often DGEBA is formulated with IPD/amine blends, e.g.
for automotive parts. We have used a similar system in our processes (see 4.5.).
For aerospace applications TGpAO or TGMDA together with aromatic amines such
as DDS are typical. It should be recalled that a relatively low viscosity resin
system is very important for the processability via RTM methods.
26
2.3.4. References
[2.3.3] Ellis, B., Hrsg. (1993) Chemistry and Technology of Epoxy Resins.
Springer Science+Business Media, Dordrecht, Netherlands
[2.3.4] Hare, C.J. (1994) Amine curing agents for epoxies. J. Prot. Coat.
Lin., 9, 77-103
[2.3.5] Hare, C.J. (1994) Epoxy curing agents II. J. Prot. Coat. Lin.,
10, 197-213
27
2.4. Toughening of epoxy resins
A disadvantage of cured epoxy resins is their extreme brittleness. The tighter the
close-meshed three-dimensional networks become, e.g. when using tetra-
functional epoxy resins, the greater is the observed brittleness. Since their
introduction in the 1950s, increasing the toughness of cured epoxy resins has
been the subject of many ongoing developments.
One has to distinguish between toughening and simply incressing the flexibility of
the cured epoxy resin.
For an increase in the flexibility of the cured epoxy resin, the network density of
the epoxy resin is reduced on purpose by adding flexible, long chain molecules,
which often are monofunctional. Thus, the failure strain may be increased but
the modulus and glass transition temperature are consequently lowered. As a
result the toughness is often somewhat increased. However, the tensile strength
is also reduced in most cases. Examples of such molecules are epoxidized
polyethers such as polypropylene glycol diglycidylether, C12-C14-glycidylether,
epoxidized soja bean oils, epoxidized cashew nut shell oil or dimeric fatty acid
esters.
When a cured epoxy resin is significantly toughened, the associated aims are not
to greatly reduce the modulus, glass transition temperature or tensile strength.
Thus, the usual approach here is to deliberately create a two-phase structure.
Therefore, in the brittle epoxy resin matrix exists a second discontinuous phase
in the form of particles or agglomerated/aggregated molecules.
One of the first modifications to try and toughen cured epoxy resins, which still
today is state-of-the-art and the industrial benchmark, was introduced into the
market in the 1980s – i.e. the modification of epoxy resins with reactive liquid
rubbers. Figure 2.4.1 shows the chemical structure of such a carboxy-terminated
butadiene-acrylonitrile copolymer (CTBN rubber):
28
However, carboxy-functional nitrile-butadiene copolymers are incompatible with
epoxy resins and when blended into an epoxy resin they phase-separate within
hours. Nevertheless, if the carboxy groups are first reacted with an excess of
epoxy resin at elevated temperatures then epoxy-rubber-epoxy block copolymers
are formed, i.e. termed rubber adducts. These can now be blended with epoxy
resins and do not phase separate. However, they have impart a relatively high
viscosity. Upon formation of the three-dimensional network during cure, then
phase separation of the rubber adduct molecules occurs. Being distributed
homogenously before the cure, they now form particles or domains in which the
long rubber molecules are entangled with each other and are bound covalently to
the epoxy matrix via the former carboxy groups.
After the extensive research in the 1980s and 1990s the mechanisms of rubber-
toughening are well understood [2.4.1], [2.4.2], [2.4.3], [2.4.4]. Recently a
comprehensive review was published by Pearson et al. [2.4.5].
29
Epoxy resins modified with reactive liquid rubbers are used in some composite
manufacturing processes, especially in prepregs and in filament winding.
However, since this type of modification considerably increases the resin
viscosity, such systems cannot be readily used in injection or infusion processes.
Another drawback of this type of modification is the fact that a small fraction of
the long, flexible rubber molecules do not participate in the phase separation
process but crosslink randomly into the three-dimensional molecular network of
the epoxy resin. Thus the network density is lowered, with an accompanying
decrease in the modulus and Tg of the cured epoxy resin.
(1)
(2)
30
Well-defined elastomeric particles are synthesized, which have a shell compatible
with epoxy resins around the rubbery core. The shell possesses reactive groups,
e.g. epoxy groups, to ensure covalent bonding with the resin matrix. The core
might be a polybutadiene, a polyacrylate or a polysiloxane. Depending on the
manufacturer, the core-shell particles have different particle sizes from
approximatively 110 nm to 300-400 nm up to 500-700 nm [2.4.8], [2.4.9],
[2.4.10], [2.4.11], [2.4.12].
32
2.4.6. References
[2.4.2] Kinloch, A.J., Shaw, S.J., Tod, D.A., Hunston D.L. (1983)
Deformation and fracture behaviour of a rubber-toughened epoxy:
1. Microstructure and fracture studies. Polymer, 24, 1341-1354
2. Failure criteria. Polymer, 24, 1355-1363
[2.4.10] Block, H., Pyrlik, M. (1988) Silicones are the Key: Modifying
Thermosetting Resins with Silicone Elastomers. Kunstst. Ger. Plast.
78, 1192-1196
[2.4.11] Lai, M., Friedrich, K., Botsis, J., Burkhart, T. (2010) Evaluation of
residual strains in epoxy with different nano/micro-fillers using
embedded fiber Bragg grating sensor. Compos. Sci. Technol. 70,
2168-2175
[2.4.12] Chen, J., Kinloch, A.J., Sprenger, S., Taylor, A.C. (2013) The
mechanical properties and toughening mechanisms of an epoxy
polymer modified with polysiloxane-based core-shell particles.
Polymer 54, 4276-4289
33
[2.4.13] Unpublished research report of nanoresins AG, Geesthacht,
Germany
[2.4.14] Yoon, T.H., Priddy, D.B., Lyle, G.D., McGrath, J.E. (1995)
Mechanical and morphological investigations of reactive polysulfone
toughened epoxy networks. Macromol. Symp. 98, 673-686
34
2.5. Modification of epoxy resins with SiO2 nanoparticles
In the years 2002 and 2003 the first commercial surface-modified silica
nanoparticles were introduced into the market. The particles are manufactured
using a modified sol-gel process, surface-modified and are supplied as
concentrates in a range of epoxy resins.
Figure 2.5.1 shows the molecular model of a 1.2 nm particle without surface
modification.
35
Figure 2.5.2: Particle size distribution of commercial SiO2 nanoparticles [2.5.2]
The particles are very well dispersed in the epoxy resin, as can be seen in Figure
2.5.3. Due to their small size they do not interact with the visible light; hence
cured bulk resins containing them seem to be transparent. This makes resins
modified with SiO2 nanoparticles suitable for use in visible composite parts
where, for instance, the appearance of the carbon fiber itself will be visible for
aesthetic reasons.
Figure 2.5.4 shows the viscosity increase of a DGEBA epoxy resin with increasing
nanosilica content [2.5.4]. The silica nanoparticles are introduced into the DGEBA
resin by blending with two different commercial nanosilica
concentrates.
37
Figure 2.5.5: TEM picture of a GFRC with 15 wt% SiO2 nanoparticles [2.5.5]
Rosso et al. [2.5.7] investigated this phenomenon with fumed silica in cured bulk
epoxy resin and found that with an addition level of 5 wt% an increase in
modulus of 20 %, with no change in the tensile strength, could be achieved. The
fracture toughness (KIc) was improved by 70 % and the fracture energy (GIc) by
140 %.
Zhang et al. [2.5.9] modified DGEBA with pyrogenic silica and compared this
system with the commercially available nanosilica. However, using more than 6
vol% of fumed silica was impossible due to the extreme increase in the viscosity
of the epoxy resin system.
38
At 6 vol% both fumed silica and nanosilica increased the modulus by 17 %. KIc
was increased by the pyrogenic silica by 49 % and by 29 % by using nanosilica.
At 3 vol% (approx. 5 wt%) the modulus was increased by 9 %, KIc by 24 % and
GIc by 43 %. Zhang et al. used an anhydride curing agent, whereas Rosso et al.
employed an amine curing agent. Although the epoxy resin molecules and
hardener molecules were much smaller than the 20 nm silica particles, being
approximately 2 nm for DGEBA [2.5.10] and 0.7 nm for DETDA [2.5.11], the
hardener used seems to have had an influence on the improvements found.
Different curing agents caused different crosslink densities of the cured bulk
resins, and hence different values of the toughness and the glass transition
temperature were recorded. The improvements upon nanosilica addition as a
function of the curing agent needs to be examined in detail (see chapter 3.1.),
since formulation guidelines would be of great help to the industrial formulator.
Figure 2.5.6: TEM picture of a cured epoxy resin modified with CTBN and
nanosilica – unknown addition levels for both modifications [2.5.15]
39
Tsai et al. [2.5.16] investigated the system DGEBA/isophorone diamine. They
found that the highest value of GIc was for the rubber-toughened system and
that the hybrid had a lower toughness. However, the loss in modulus due to the
reactive liquid rubber was nearly compensated for by the nanosilica addition. No
synergy was found between the two modifications.
Figure 2.5.7: Bulk GIc of a DGEBA modified with CTBN and SiO2 nanoparticles
[2.5.17]
The combination of core-shell particles with nanosilica has also been subject of
several investigations. An extensive study conducted by researchers of the IVW
[2.5.18] revealed a homogenous distribution of both the micron-sized core-shell
and the 20 nm silica particles. They found that the modulus was lowered only
marginally by the core-shell particles. The addition of nanosilica up to 8 vol%
increased the modulus and KIc by 25-50 %. The improvements found for this
amine-cured epoxy resin system were, however, minor.
In contrast Liu et al. [2.5.19] reported an increase in GIc by nearly 800 % for a
combination of 10 wt% core-shell particles and 10 wt% nanosilica for their
amine-cured system.
40
2.5.4. Property improvements of fiber-reinforced composites
One of the most important questions for an industrial formulator is whether the
improvements achieved for cured bulk resin systems will be found in a fiber-
reinforced composite based upon the same epoxy resin system as a matrix.
Mahrholz et al. [2.5.20] reported very early improvements for a glass fiber-
reinforced laminate based on an anhydride cured DGEBA by the addition of
nanoparticles. The bidirectional layup (+45°/-45°) was used in an injection
process.
Tsai et al. [2.5.16], [2.5.25] made laminates from an amine cured epoxy and UD
glass fibers. The addition of 10 wt% SiO2 nanoparticles increased GIc by only 8
%. Although the GIc of the modified bulk resin was improved by 47 %, the tensile
strength and compressive strength of the laminates did not increase.
Tang et al. [2.5.26] used piperidine as the hardener and UD carbon fiber textiles
to make laminates by the VARTM process. The addition of 10 wt% of nanosilica
increased the bulk GIc of the epoxy resin matrix by 93 % and 20wt % of
nanosilica increased the bulk GIc by 180 %. For the composites, however, the
improvements in GIc were only 1 % and 21 %, respectively. Also, the
interlaminar shear strength (ILSS) was reduced by the nanosilica addition to the
epoxy matrix.
41
2.5.8 the delaminated areas from the falling dart test are shown and there is no
visible difference.
Figure 2.5.8: GFRC laminates after falling dart test with 30 J impact without and
with nanosilica [2.5.27]
To study these contradictions, and then to expand and complement them with
other research results and eventually to identify basic mechanisms is the subject
of chapter 3.3. Another aim is to undertake a detailed examination of the
influence of the curing agent used in the epoxy resin system in combination with
nanosilica on composite properties, as silica nanoparticles are slightly acidic,
which might have an influence on amine curing agents. Since, many actual
composite parts made in industry today contain SiO2 nanoparticles: be it
aerospace components, machine parts, automotive parts or sporting equipment.
Figure 2.5.9: GIc of a CFRC based on DGEBA modified with CTBN and SiO2
nanoparticles [2.5.17]
43
A resin modification with 9 wt% CTBN and 10 wt% silica nanoparticles increased
the composite laminate GIc from 330 J/m2 to 860 J/m2; which is an improvement
of 160 %. The composite mode II fracture energy (GIIc) of the hybrid was 46 %
higher compared to the unmodified control.
Tsai et al. [2.5.16], [2.5.26] formulated hybrid resin systems with CTBN as well
as with core-shell particles. The loss in modulus was in both cases reduced
signifcantly by the addition of 10 wt% nanosilica. The bulk GIc of the hybrid was
improved by modification with 10 wt% CTBN by 390 % and with 10 wt% core-
shell particles by 442 %. However, the composite GIc of a glass fiber-reinforced
laminate was improved only by 48 % for CTBN and by 82 % for the core-shell
particles system.
The very interesting question how much of a property improvement of the bulk
resin may be transferred into the composite laminate has already been studied in
the past for different rubber-toughened systems [2.5.30]. In Figure 2.5.10 the
relationship between bulk GIc and laminate GIc is given.
Figure 2.5.10: Bulk GIc of cured epoxy resins versus laminate GIc [2.5.30]
It is obvious that the increase in laminate fracture energy with increasing cured
bulk epoxy resin fracture energy is not proprtional.
44
First there is a significant increase in laminate GIc up to a neat resin fracture
energy of approximately 250 J/m2, then the slope of the curve flattens. It then
takes considerable increases in bulk GIc to improve the laminate fracture
toughness just by a small amount.
According to Altstädt these two different areas are the result of the constriction
of the plastic deformation zone in the epoxy resin matrix between the
reinforcement [2.5.30]. Tougher resins have a more extensive deformation zone
ahead of the crack tip than brittle resins.
Of course the question arises, will similar relationships be found to exist for the
hybrid epoxy resins with silica nanoparticles present, and this will be discussed
extensively in chapter 3.4.
2.5.5. References
[2.5.1] Odegard, G.M., Clancy, T.C., Gates, T.S. (2005) Modeling of the
mechanical properties of nanoparticle/polymer composites. Polymer
46, 553-563
[2.5.2] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. (2003) Nano-
toughening of Epoxies. Proceedings of Stick! Conference April 9th,
Nürnberg, Germany, Vincentz Verlag 2003
[2.5.5] Sprenger S., Kinloch A.J., Taylor, A.C., Mohammed, R.D. (2005)
Rubber-toughened GFRCs optimized by nanoparticles. JEC Compos.
Magazine 21, 66-69
[2.5.6] Eger C., Schultz, P. (2005) Reinforcing epoxy resins with silica nano-
particles. In Proceedings of "High Performance Fillers 2005", March
8-9, Köln, Germany
[2.5.7] Rosso, P., Ye, L., Friedrich, K., Sprenger, S. (2006) A toughened
epoxy resin by silica nanoparticle reinforcement. J. Appl. Polym. Sci.
100, 1849-1855
45
[2.5.9] Liu, S., Zhang, H., Zhang, Z., Zhang, T., Sprenger, S. (2008)
Tailoring the mechanical performance of epoxy resin by various
nanoparticles. Polymers & Polymer Composites Vol. 16, No. 8, 471-
477
[2.5.11] Gou, J., Fan, B., Song, G., Khan, A. (2006) Study of affinities
between single-walled nanotube and epoxy resin using molecular
dynamic simulation. Int. J. Nanosci. 5, 131-144
[2.5.13] Sprenger, S., Eger, C. Kinloch, A.J., Lee, J.H., Taylor, A.C., Egan, D.
(2003) Nanoadhesives: Toughness and high strength. Adhäsion,
Kleben & Dichten 03/2003, 24-30
[2.5.16] Tsai, J.-L., Huang, B.-H. Cheng, Y.-L. (2009) Enhancing Fracture
Toughness of Glass/Epoxy Composites by Using Rubber Particles
Together with Silica Nanoparticles. J. Comp. Mater. 43, 3107-3123
[2.5.17] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D., Eger, C.
(2005) Rubber-toughened FRCs optimized by nanoparticles. JEC
Comp. Mag. No 19, 73–76
[2.5.19] Liu, H.Y., Wang, G.T., Mai, Y.W. Zeng, Y. (2011) On fracture
toughness of nanoparticle modified epoxy. Composites: Part B 42,
2170-2175
46
[2.5.23] Kinloch, A.J., Masania, K., Taylor, A.C., Agarwal, R., Sprenger, S.,
Egan, D. (2008) The fracture of glass-fibre-reinforced epoxy
composites using nanoparticle-modified matrices. J. Mater. Sci.
Letters 43, 1151-1154
[2.5.25] Tsai, J.L., Huang, B-H., Cheng, Y.L. (2011) Enhancing Fracture
Toughness of Glass/Epoxy Composites for Wind Blades Using Silica
Nanoparticles and Rubber Particles. Procedia Engineering 14, 1982-
1987
[2.5.26] Tang, Y., Ye, L., Zhang, D., Deng, S. (2011) Characterization of
transverse tensile, interlaminar shear and interlaminate fracture in
CF/EP laminates with 10 wt% and 20 wt% silica nanoparticles in
matrix resins. Composites: Part A, 42, 1943-1950
[2.5.27] Sprenger, S., Eger, C., Kinloch, A., Mohammed, R., Taylor, A.
(2005) Rubber-toughening and Nanoparticles in Epoxies: Synergies
in FRC. Proceedings SAMPE, Paris, France, April 5 – 7, 2005
[2.5.28] Brandt, J., Drechsler, K., Schmidtke, K. (2004) Composites für den
Flugzeugbau. Kunststoffe 10, 290-294
47
3. Results and Discussion
In this chapter first all property improvements of cured epoxy resins achievable
by the addition of surface-modified SiO2 nanoparticles will be investigated. The
influence of the particle modification on the cure performance of epoxy resins is
of interest as well. Furthermore it will be investigated if there are differences or
commonalities for the different resin/hardener combinations. The mechanisms
relevant for property improvements will be highlighted and the question
regarding the optimum addition level will be answered. With the gained
knowledge of the behavior of silica nanoparticles in bulk epoxy resin systems the
behavior in fiber-reinforced laminates can be studied later on (see chapter 3.3.).
3.1.1. Abstract
3.1.2. Introduction
Epoxy resins are very versatile raw materials for industrial products, from
windmill blades to highly sophisticated aerospace parts such as wings or fuselage
to coatings and adhesives for construction. They are used in large volumes in
generator encapsulations and in microlectronics and UV-cured electronic
adhesives. Structural adhesives for automotive or aerospace applications,
shipbuilding or windmill blade construction are mainly based on epoxy resins.
Therefore, quite a range of different epoxy resins is available: from low viscous
short-chain aliphatics such as the diglycidyl ether of hexanediole to high
performance, multifunctional aromatic resins such as the triglycidyl ether of
aminophenole or tetraglycidyl ether of methyl dianiline (TGMDA). Of course, the
diglycidyl ether of bisphenol F (DGEBF) and its higher molecular weight
variations, epoxidized novolac resins, are of great importance. However, by far,
the biggest volume of epoxy resins is produced as diglycidyl ether of bisphenol A
(DGEBA), the workhorse of the epoxy industry.
With regard to the many different applications and their substantially different
property profiles for the materials used, a big variety of hardeners is used in
industrial applications as well. An excellent and very comprehensive overview of
the different hardeners used in the industry, their chemical natures and their
network formation was published by Hare some years ago [3.1.1], [3.1.2.].
48
Sterically hindered aromatic amines are especially suitable for densely
crosslinked, high-glass-transition-temperature (Tg) aerospace formulations; they
are typically used in combination with trifunctional or tetrafunctional epoxy
resins.
Acid anhydrides, being part of the formulation in almost equal amounts to epoxy
resins and exhibiting a very low viscosity, are very useful in highly filled
encapsulation systems and composites made by filament winding as they lower
the viscosities significantly. On curing, they form medium crosslinked resin
systems.
In quite a few epoxy resin formulations, fillers are necessary. Fillers improve
mechanical properties such as strength, stiffness and modulus. However, they
have a negative impact on the viscosity of the resin, which forbids their use in
some applications. Furthermore, they are filtered out by the fabric when the
resin formulation containing the filler is subjected to injection manufacturing
methods for fiber-reinforced composites. Thus, many applications whose
performance could be improved by the use of a filler do not permit the use of
classical micrometer-sized fillers, and of course, fillers cannot be used in
transparent applications.
49
Figure 3.1.1: Transmission electron microscopy picture of cured epoxy resin with
5 wt % nanosilica
In the years 2002 and 2003, the first commercial-grade surface-modified silica
nanoparticles were introduced into the market. They were manufactured in situ
directly in the epoxy resin by a modified sol-gel process and had an average
particle size of 20 nm and a very narrow particle size distribution. Odegard et al.
[3.1.3.] showed in their article molecular models of such particles and the huge
amount of hydroxyl groups on the particle surface. The industrially manufactured
particles were surface coated. The hydroxyl groups were reacted with silanes to
prevent agglomeration and to compatibilize the particles with the resin.
Nevertheless, there were still some remaining free hydroxyl groups, and thus,
the particles were slightly acidic. One needs to keep this in mind when looking at
the different additives for epoxy resins. The industrial material is very close to
the model: isolated spherical particles as concentrates in epoxy resins with an
average size of approximately 20 nm. Figure 3.1.1 shows a cured epoxy resin
with such silica nanoparticles (5 wt %).
Consequently, they are a very attractive raw material for epoxy resin
formulators.
50
Today, 10 years later, they are used in many industrial formulations, including
encapsulating resins, adhesives, and composites such as automotive parts and
machine parts. They improve various properties, including strength, modulus,
stiffness, toughness, and scratch resistance. Significant improvements in the
fatigue performance were reported when the epoxy resin was modified with
nanosilica.
3.1.3. Discussion
The dispersion of the nanosilica was investigated by all researchers and was
always found to be homogenous. Agglomerates or areas with different silica
nanoparticle concentrations were not observed.
This behaviour was mainly due to the surface coating of the particles as uncoated
particles tend to agglomerate. In some rare cases, when nanosilica-containing
epoxy resins were cured with amine-functional reactive liquid rubbers, some
agglomeration was found. Such exceptions are described in another article
currently under preparation.
Tg was not influenced in most cases and sometimes decreased by 1-2 °C at very
high addition levels of nanosilica. Sanctuary at al. [3.1.4] investigated the
complex specific heat capacity and reported that a blend of DGEBA and surface-
modified nanosilica behaved in a neutral manner with regard to the glass
transition dynamics of the resin matrix, just like a mixture.
In a continuation of this work, Wetzel et al. [3.1.6] explored the fracture and
toughening mechanisms using Al2O3 and TiO2 nanoparticles of similar sizes (ca.
20 nm) in 4,4'-methylene bis(2-methylcyclohexyl-amine) cured DGEBA.
51
They identified crack deflection processes, crack pinning, and energy dissipation
rather than debonding at the particle-resin interface as reasons for the
toughness improvements. A very detailed description is given in Wetzel´s Ph.D.
thesis [3.1.7].
The curing kinetics of a modified DGEBA cured with piperidine were studied by
Rosso and Ye [3.1.9]. The addition of 1-5 vol % nanosilica led to a higher
reactivity in curing and an alteration in crosslinking. They suggested the
formation of an amino-rich interphase region around the silica nanoparticles,
which could have been responsible for the property improvements.
Haupert et al. [3.1.10] looked into the tribological properties of DGEBA cured
with an aliphatic amine. They found improvements at nanosilica addition levels
above 2 vol % and a maximum at 5.5 vol %. The wear resistance was improved
by 30%.
DGEBA and DGEBF cured with an accelerated aminoethyl piperazine were the
subject of studies by Dittanet [3.1.11]. She reported an increase in the modulus
of 108% for DGEBA and 90% for DGEBF at a 30 wt % addition level of
nanosilica. A significant reduction in coefficient of thermal expansion (CTE) was
reported as well.
52
This study was continued by Philipp et al. [3.1.15] and the generalized Cauchy
relation was investigated. It seems that the cured epoxy resins with different
amounts of nanosilica incorporated behaved similarly to porous silica glasses,
and this indicated a perfect distribution of monodisperse silica nanoparticles.
Ye et al. [3.1.18] reported increases in the modulus from 2.9 to 3.3 GPa (with 10
wt % nanosilica) and 3.6 GPa (with 20 wt % nanosilica) for a DGEBA resin cured
with piperidine. GIc was increased from 238 to 458 and 666 J/m2 (improvements
of 92 and 180%, respectively).
Liu et al. [3.1.19] looked further into KIc of piperidine-cured DGEBA. They found
an increase in the modulus and KIc with increasing loading level: 22% increase in
the modulus and 304% increase in GIc at 20 wt % nanosilica.
In another study, Liu et al. [3.1.20] examined cyclic fatigue crack propagation
and reported significant improvements in the fatigue lifetimes for 6 and 12 wt %
nanosilica. They discussed extensively the contribution of the different
toughening mechanisms identified at high and low loading levels.
In continuation of earlier work, Dittanet and Pearson [3.1.21] tried to identify the
influence of the nanoparticle size on the toughening of a piperidine-cured DGEBA
epoxy resin. They used particles with average sizes of 23, 74 and 170 nm from a
small-scale manufacturer and reported improved properties with increasing
addition levels of nanosilica. The modulus was increased by approximately 60%
at 30 vol % nanosilica addition regardless of the particle size. GIc was improved
by 221% for the 23 nm particles, 239% by the 170 nm particles and 317% by
the 74 nm particles. Interesting was the reduction of the CTE size dependence as
well; the 23 nm particles performed best.
Mechanisms for the toughening effect of nanosilica were discussed as well, and
the model from Kinloch et al. [3.1.22] was confirmed; see the next two sections
in this article. Matrix shear banding was the dominant mechanism, matrix void
growth was secondary, and the debonding of silica nanoparticles had only a
minor effect.
Table 3.1.1 gives an overview of the increases in modulus and KIc (at RT) versus
addition levels of nanosilica. The same particles were used together with
piperidine as a hardener, and identical cure conditions were used.
As a short summary, I concluded that the tensile strength remains more or less
unchanged by the addition of silica nanoparticles. At very high addition levels,
there might by a slight increase [3.1.19]. The modulus increases with increasing
concentrations of silica nanoparticles.
53
However, toughness and fatigue improvements have been either reported to
increase steadily or have a maximum at 5-6% loading level.
Table 3.1.1: Properties of piperidine cured epoxy resins with various nanosilica
contents
Kinloch et al. [3.1.22] investigated DGEBA and a DGEBA/DGEBF blend cured with
a difunctional short-chain poly(ether amine). They reported only very small
increases in the modulus (17 and 10%, respectively) with 20 wt % nanosilica.
Toughness by means of GIc was improved in both cases by approximately 280%.
A linear increase with increasing addition level was found. The toughening
mechanisms were investigated and compared with theoretical predictions.
Localized plastic shear bands initiated by the stress concentrations around the
periphery of the silica nanoparticles were the main contributor to the increase in
toughness. The debonding of the nanoparticles seemed to be less important, as
only approximately 15% of the nanoparticles were found to debond.
54
However, the plastic void growth following the debonding contributed to the
toughness increase.
DGEBA modified with various amounts of nanosilica and cured with the
difunctional short-chain poly(ether amine) was studied by Tsai et al. [3.1.16] as
well. They found exactly the same 17% improvement in the modulus at 20 wt %
nanosilica like Kinloch et al. [3.1.22] and a 40% improvement at a 40 wt %
loading level. The strength was slightly improved at the 40 wt % level. Three-
point bending tests showed an improvement in the flexural strength with
increasing addition of silica nanoparticles up to 16%. The toughness increase was
found to be very small because of the fact that KIc of the unmodified resin was
quite high. Improvements reached a plateau at approximately 10 wt %
nanosilica. One has to take into account the fact that the curing conditions were
different.
The work of Jajam and Tippur [3.1.24] focused on a DGEBA blended with 15% n-
butyl glycidyl ether cured with a commercial hardener formulation consisting of
poly(ether amine), trimethyl hexane diamine, benzene-1,3-dimethane amine,
nonyl phenol, and substituted phenol. In addition to nanosilica, they tested
micrometer-sized spherical glass particles with a mean diameter of 35 µm. They
found a linear increase of KIc with increasing addition level for both particles. At
10 vol %, the nanosilica provided a 78% enhancement relative to the 35 µm
glass particles. In dynamic fracture tests, both materials showed improved
dynamic KIc values with increasing loading levels. However, the nanosilica
showed only minor improvement of 34% at a 10 vol % addition level. In another
study [3.1.25], it was confirmed that the addition of nanosilica did not
necessarily improve the toughness when a fast impact occured. Nevertheless, it
was shown that quite significant improvements could be achieved when
commercial resin systems were used, with the hardeners typically being complex
amine blends. The effects found for nanosilica modification of epoxy resins have
not been limited to model systems.
Kinloch et al. [3.1.22] also investigated a high performance, high Tg epoxy resin
system similar to the industrial benchmark RTM6. Nanosilica filled TGMDA was
cured with a blend of 4,4´-methylenbis(2,6-diethylaniline) and 4,4´-
methylenebis(2,6-diisopropylaniline). At 10 wt % nanosilica loading level, the
modulus was increased by 26%; and GIc was increased by 146%, although it was
still at a very low level of 172 J/m2.
Ma et al. [3.1.23] studied DGEBA cured with 4,4´-DDS. The modulus was
increased by 18% at a 10 wt % addition level of nanosilica. Doubling the
addition level to 20% increased the modulus by 40% compared to the neat
epoxy resin. GIc was improved by 49 and 81%, respectively.
55
Transmission electron microscopy showed some dilatation in the propagated
crack-tip area and some nanovoid formation.
The research of Gurung [3.1.27] was based on a DGEBA cured with 4,4´-DDS as
well. An industrially available nanosilica epoxy masterbatch was compared to
silica nanopowder modified with aminopropyl triethoxysilane. Gurung reported an
acceleration of curing at the beginning of curing caused by the industrial
nanosilica as well. A significant drop in Tg from 173 to 130°C at approximately
33 wt % nanosilica was found in both thermomechanical analysis and differential
scanning calorimetry studies. The modulus was improved by 59%, but the stress
at break was reduced by 12%. The industrial material performed better than the
"homemade" nanosilica, and this was attributed to a better particle dispersion.
It is interesting to see the effects of the different network densities deriving from
the two different DDS molecules when they were modified with the same
nanoparticles with regard to the reduction of Tg.
By far, most researchers have worked with anhydride curing agents for various
reasons, including their low viscosity and easily controlled curing cycle. The side
reactions can also be controlled and can be suppressed with a well-defined curing
schedule. Thus, a tremendous amount of test results is available in this field and
are described hereafter. All of the anhydride curing agents described were
accelerated with very small amounts (e.g., 1%) of ternary amines.
Eger and Schultz [3.1.28] reported a significant reduction in the viscosity when
nanosilica was used in combination with a DGEBA epoxy resin in place of fumed
silica. An amount of 25 wt % increased the modulus by 37% and KIc by 72%. A
slight increase in tensile strength was reported as well.
Kinloch et al. [3.1.30] looked into the fatigue performance of such systems. The
modulus was increased by 30% and KIc by 73% at 20.2 wt % nanosilica.
Cyclic fatigue testing was applied to the compact tension test specimen, and it
was found that the addition of nanosilica clearly and significantly improved the
fatigue performance: the more, the better.
56
Zhang et al. [3.1.31] reported an almost linear increase in the modulus with
increasing silica nanoparticle content, from 2.75 to 3.95 GPa at a 14 vol %
loading level (an improvement of 44%). KIc was improved by more than 50%.
The tensile strength was improved only slightly. They reported an observation of
a polymer shell around the inorganic particle, as suggested by Wetzel et al.
[3.1.6] for other nanoparticles and Zhang et al. [3.1.32] for SiO2-nanoparticles.
Table 3.1.2 gives an overview of the modulus and KIc (at RT) values versus the
addition level of nanosilica. The same particles were used in DGEBA as epoxy
resin and were cured with methyl hexahaydrophthalic acid anhydride. The curing
conditions varied slightly.
57
SiO2 Modulus (GPa) KIc (MPam1/2)
(wt
%)
0 2.96 2.75 2.96 3.01 0.59 0.55 0.51 0.46
± 0.05 ± 0.05 ± 0.12 ± 0.04
2 2.79 0.57
± 0.09 ± 0.11
4 3.20 3.20 1.03 0.64
In short, I concluded that the tensile strength was slightly improved by the
addition of silica nanoparticles at high loading levels. The modulus and toughness
increased with increasing concentrations of silica nanoparticles. There seems to
be no maximum addition level.
58
Gu et al. [3.1.38] examined the mechanical and tribological aspects of modified
DGEBF. The hardness and modulus increased nearly linearly with increasing
nanosilica content. When 15 vol % nanoparticles were used, the modulus was
improved by 40% and the hardness was improved by 33%. Plasticity index
decreased first, showed a minimum at 8 vol % nanosilica, and then increased.
The friction coefficient showed similar behaviour.
Zhang et al. [3.1.32] investigated EEC as well. With a 22.7 wt % silica content,
the modulus was found to increase from 3.05 to 4.18 GPa or by 37%. KIc was
improved by 76%. Interestingly, the impact strength (by Charpy impact testing)
went from 25.6 to 31.4 J/m2 at 5.3 wt% nanosilica content to 23.7 J/m2 at 22.7
wt % loading. Apparently, a maximum existed. The strength was not affected by
the modifications. As a main toughening mechanism, the energy dissipation
caused by nanoparticle-induced dimples was claimed.
They found a slight increase in tensile strength and an increase in the modulus
with increasing nanosilica content. At a 25 wt % loading level, an improvement
of 36% was found. The impact strength, determined by a Charpy test, showed a
maximum improvement at a content of 15 wt % nanoparticles and decreased
upon further addition of nanosilica. Shrinkage was reduced by 19% and the
thermal conductivity increased with increasing nanosilica content.
59
Methylnadic Acid Anhydride as a Hardener
In another study, Hodzic et al. [3.1.43] employed both cylindric and prismatic
test specimens. In both cases, the test specimen with the highest addition level
of nanosilica showed the best performance. Enhanced shear deformation of the
matrix and the formation of shear bands that influenced the crack propagation
were identified as mechanisms of action of the nanoparticles.
The Tg´s reported for the DGEBA/isophorone diamine system were around 80°C.
Cured DGEBA/piperidine systems exhibited Tg´s between 80 and 100°C. Short-
chain aliphatic poly(ether amine)s as curing agents achieved Tg´s around 80°C
as well. For DGEBA crosslinked with methylhexahydrophthalic acid anhydride, the
Tg´s were found to range between 150 and 164°C. A 4,4´-DDS hardener yielded
in combination with DGEBA Tg´s between 163 and 173°C. DDS or other aromatic
amines used in combination with tetrafunctional epoxy resin (TGMDA) can
achieve Tg´s of up to 260°C.
In Figure 3.1.2, the modulus data from Tables 3.1.1 and 3.1.2 and from some
literature are given [3.1.16], [3.1.22], [3.1.23], and a linear interpolation is
applied. The epoxy resin used for all systems was DGEBA. The increase in
modulus with increasing nanosilica content was larger for hardeners that form a
close meshed network. Piperidine led to polymers with the lowest network
density. This was followed by the difunctional poly(ether amine) (here called D
230), methylhexahydrophthalic acid anhydride (MHHPA), and finally 4,4´-DDS.
60
5
D 230
4,5 Piperidine
Modulus (GPa)
MHHPSA
4
DDS
Linear (DDS)
3,5
Linear (MHHPSA)
3 Linear (D 230)
Linear (Piperidine)
2,5
0 5 10 15 20 25
wt% nanosilica
Figure 3.1.2: Moduli of epoxy resins with various contents of nanosilica cured
with different hardeners
Figure 3.1.3 shows the KIc as function of the nanosilica content for these different
hardeners. It is evident that hardeners forming a closer meshed network and
therefore yielding cured resins with a higher modulus tend to be more brittle and
thus having a lower toughness. Again, the toughness increases more or less
linearly with increased addition level of silica nanoparticles. The improvements at
the same loading level are bigger for systems that already exhibit a higher
toughness. This is consistent with the fact that toughening mechanisms are more
efficient for polymer matrices with a higher ductility.
2,5
Fracture toughness K1c (MPam0,5)
D 230
2 Piperidine
MHHPSA
DDS
1,5
Linear (DDS)
Linear (MHHPSA)
1
Linear (D 230)
Linear (Piperidine)
0,5
0 5 10 15 20 25
wt% nanosilica
Figure 3.1.3: KIc of epoxy resins with various contents of nanosilica cured with
different hardeners
As shown in Figure 3.1.3, the linear approach did not work for the rather flexible,
short-chain poly(ether amine) (D 230) as a curing agent.
61
Here, there seemed to be a maximum; this means that an optimal addition level
of nanosilica existed, apparently at lower addition levels. This was consistent
with other studies where such a curing agent was used in combination with
flexible amine-functional reactive liquid rubbers [3.1.45].
3.1.4. Conclusions
Taking into account the data gathered and published over the last 10 years, I
drew the following conclusions:
2. The particles behave like a filler and do not significantly change the curing
characteristics of epoxy resin/hardener blend. Toward the end of the
curing cycle, curing is slowed down to a certain extent.
6. KIc increases with increasing level of silica nanoparticles but not always in
a linear function. At 10 wt % nanosilica, an increase of approximately 50%
can be expected.
7. The toughening mechanisms are the debonding of the epoxy polymer from
the silica nanoparticles followed by plastic void growth. Localised plastic
shear banding contributes as well. Crack deflection does not seem to play
a significant role.
62
3.1.5. References
[3.1.3] Odegard, G.M.; Clancy, T.; Gates, T.S. Polymer (2005), 46, 553–
563.
[3.1.4] Sanctuary, R.; Baller, J.; Krüger, J.-K.; Schäfer, D; Wetzel, B.;
Possart, W.; Alnot, P. Thermochimica Acta (2006), 445, No. 2, 111
–115.
[3.1.5] Rosso, P.; Ye, L.; Friedrich, K.; Sprenger, S. J. Appl. Polym. Sci.
(2006), 100, 1849-1855.
[3.1.8] Deng, S.; Ye, L.; Friedrich, K. J. Mater. Sci. (2007), 42, No. 8, 2766
-2774.
[3.1.9] Rosso, P.; Ye, L. Macromol. Rapid Commun. (2007), 28, 121-126.
[3.1.12] Wang, G.-T.; Liu, H.-Y.; Saintier, N.; Mai, Y.-W. Eng. Failure
Analysis (2009), 16, 2635-2645.
[3.1.14] Baller, J.; Becker, N.; Ziehmer, M.; Thomassey, M.; Zielinski, B.;
Müller, U.; Sanctuary, R. Polymer (2009), 50, 3211-3219.
[3.1.15] Philipp, M.; Müller, U.; Jiménez Riobóo, R.J.; Baller, J.; Sanctuary,
R.; Possart, W.; Krüger, J.K. New Journal of Physics (2009), 11,
023015.
[3.1.16] Tsai, J.-L.; Huang, B.-H.; Cheng, Y.-L. J. Comp. Materials (2010),
4/44, 505-524.
[3.1.17] Tsai, J.-L.; Chang, N.-R. J. Comp. Materials (2011), 2/45, 2157-
2164.
63
[3.1.18] Tang, Y.; Ye, L.; Zhang, D.; Deng, S. Composites: Part A (2011),
42, 1943-1950.
[3.1.19] Liu, H.Y.; Wang, G.T.; Mai, Y.W.; Zeng, Y. Composites: Part B
(2011), 42, 2170-2175.
[3.1.20] Liu, H.Y.; Wang, G.T.; Mai, Y.W. Comp. Sci. Tech. (2012), 72, 1530
-1538.
[3.1.22] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Taylor, A.C.; Sprenger, S.
Polymer (2010), 51, 6284-6294.
[3.1.23] Ma, J.; Mo, M.-S.; Du, X.-S.; Rosso, P.; Friedrich, K.; Kuan, H.-C.
Polymer (2008), 49, 3510-3523.
[3.1.25] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D. JEC
Composites Magazine (2005), No 21, 66–69.
[3.1.26] Pethrick, R.; Miller, C.; Rhoney, I. Polym. Int. (2010), 59, 236–
241.
[3.1.30] Blackman, B.R.K.; Kinloch, A.J.; Sohn Lee, J.; Taylor, A.C.; Agarwal,
R.; Schueneman, G.; Sprenger, S. J. Mater. Sci. Letters (2007), 42,
7049–7051.
[3.1.31] Liu, S.; Zhang, H.; Zhang, Z.; Sprenger, S. J. Nanosci. Nanotechnol.
(2008), Vol.8, 1-6.
[3.1.32] Zjang, H.; Zhang, Z.; Friedrich, K.; Eger, C. Acta Materialica (2006),
54, 1833-1842.
[3.1.33] Liu, S.; Zhang, H.; Zhang, Z.; Zhang, T.; Sprenger, S. Polymers &
Polymer Composites (2008), Vol. 16, No. 8, 471-477.
[3.1.34] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Sohn Lee, J.; Taylor, A.C.;
Sprenger, S. J. Mater. Sci. (2010), 45, 1193-1210.
64
[3.1.35] Hsieh, T.H.; Kinloch, A.J.; Taylor, A.C.; Sprenger, S. J. Applied
Polym. Sci. (2011), 119, 2135-2142.
[3.1.36] Tang, L.-C.; Zhang, H.; Sprenger, S.; Ye, L.; Zhang, Z. Composites
Science and Technology (2012), 72, issue 5, 558-565.
[3.1.37] Zhang, H.; Tang, L.-C.; Zhang, Z.; Friedrich, K.; Sprenger, S.
Polymer (2008), 49, 3816-3825.
[3.1.38] Wang, Z.,Z.; Gu, P.; Zhang, Z.; Gu, L.; Xu, Y.Z. Tribol. Lett. (2011),
42, 185-191.
[3.1.39] Zhang, Y.-F.; Bai, S.-L.; Li, X.-K.; Zhang, Z. J. Polym. Sci. Part B:
Polym. Phys. (2009), 47, 1030-1038.
[3.1.41] Duwe, S.; Arlt, C.; Aranda, S.; Riedel, U.; Ziegmann, G. Comp. Sci.
Tech. (2012), 72; 1324-1330.
[3.1.42] Jumahat, A.; Soutis, C.; Jones, F.; Hodzic, A. "The effects of
nanosilica contents on thermal and mechanical properties of epoxy
polymers"; Proceedings of 10th SAMPE Europe conference, 12-14
April, Paris, France (2010)
[3.1.43] Jumahat, A.; Soutis, C.; Jones, F.; Hodzic, A. J Mater. Sci. (2010),
45, No. 21, 5973-5983.
[3.1.45] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Hsieh, T.-H. Adhesion,
Adhesives & Sealants (2009), 10, 8–11
65
3.2. Property improvements of epoxy resins modified with
SiO2 nanoparticles and elastomers (hybrid systems)
3.2.1. Abstract
Epoxy resins are inherently brittle. Thus they are toughened with reactive liquid
rubbers or core-shell elastomer. Surface-modified silica nanoparticles, 20 nm in
diameter and with a very narrow particle size distribution, are available as
concentrates in epoxy resins in industrial quantities since 10 years. Some of the
drawbacks of toughening, like lower modulus or a loss in strength can be
overcompensated when using nanosilica together with these tougheners.
Apparently there exists a synergy as toughness and fatigue performance are
increased significantly. In this article the literature published in the last decade is
studied with a focus on mechanical properties. Results are compared and the
mechanisms responsible for the property improvements are discussed.
3.2.2. Introduction
Epoxy resins are used for many years in a multitude of industrial products, like
structural automotive adhesives, high performance fibre reinforced composites,
electrical and electronic applications, heavy duty protective coatings and many
more. However, they are very brittle and therefore in most commercial
formulations tougheners are used.
Since the seventies and eighties of last century the use of reactive liquid rubbers
as tougheners for epoxy resins became industrial standard. Carboxy terminated
butadiene acrylonitrile copolymers (CTBNs) are reacted with an excess of epoxy
resin to form a so-called adduct, an epoxy-rubber-epoxy terpolymer. These are
soluble in epoxy resins, whereas the pure rubber is not. Upon cure and
subsequent formation of the three-dimensional network, the rubber molecules
become insoluble again, phase separate and form small rubber domains or
particles within the cured polymer matrix. These rubber particles are chemically
linked to the polymer. Since the very early work of Kinloch and his team [3.2.1],
[3.2.2] the mechanisms of rubber toughening have been the subject of intensive
research and are well understood. An excellent review was published recently
[3.2.3].
However, the phase separation and domain formation depend on the cure speed,
cure temperature and the curing agent itself. The acrylonitrile content of the
copolymer has an influence on the particle size as well.
66
To overcome these disadvantages core-shell elastomers (CSRs) have been
developed in the 1980s-1990s. Instead of forming a second phase upon cure the
rubber particles were added from the beginning. They consist of an elastomeric,
rubber-like core of approx. 90 nm; typically a butadiene homopolymer or a
butadiene-styrene copolymer with a random copolymer shell of 10-20 nm which
is compatible with the epoxy resin [3.2.4], [3.2.5]. They will be referred to as
CSR Type I. Others are based on polyacrylate cores and have a diameter in the
range of 300- 400 nm [3.2.6]. They will be referred to as CSR Type II. If these
core-shell particles are dispersed in epoxy resins, the viscosities of modified
resins are much lower compared to epoxy resins modified with reactive liquid
rubbers. The toughening effects are independent from the curing agent and the
cure schedule. Sometimes strength and modulus are lowered, but not to the
same extent as with reactive liquid rubbers. The use in high temperature
applications however is limited, as the shell tends to soften at higher
temperatures followed by a drastic loss in strength and modulus of the cured
polymer.
In the years 2002/2003 the first commercial grades of surface modified silica
nanoparticles were introduced into the market. They are manufactured in situ
directly in the epoxy resin by a modified sol-gel process and have an average
diameter of 20 nm as well as a very narrow particle diameter distribution.
These particles are completely monodisperse and do increase the resin viscosity
only slightly at higher concentrations. In contrast to fumed silica they exhibit no
thixotropic properties but behave like a Newtonian liquid. Due to their size they
are transparent and can easily penetrate even close meshed fabrics in composite
manufacturing when being injected. The property improvements which can be
achieved by modifying epoxy resins with these silica nanoparticles, like modulus,
toughness and fatigue performance, have been the subject of intensive research
in the last decade [3.2.10].
The mechanisms how nanoscaled spherical fillers can improve epoxy polymer
properties have been identified; however the contribution of each one might be
of a different proportion depending on the hardeners used to form the three-
dimensional network upon cure. Figure 3.2.2 shows the excellent dispersion and
the very narrow particle diameter distribution of the spherical nanosilica.
67
Figure 3.2.1: SEM image of an epoxy polymer with 5.5 wt% CSR Type III
Figure 3.2.2: TEM image of an epoxy polymer with approx. 20 wt% silica
nanoparticles [3.2.11]
68
3.2.3. Discussion
If not mentioned otherwise, the researchers cited used commercial 40% (by
weight) concentrated masterbatches of surface-modified nanosilica in DGEBA
from one supplier. These particles have an average diameter of 20 nm and a
very narrow particle diameter distribution. They were then diluted down using
commercial epoxy resins to vary the nanosilica concentrations.
At first the silica nanoparticles, after being commercially available, had been
added to epoxy formulations containing CTBNs to reduce the loss in strength and
modulus caused by the rubber modification without increasing the viscosity. Very
soon in some applications a synergy between elastomeric tougheners and silica
nanoparticles was discovered and patented consequently [3.2.12].
However, as will be shown in this article, the synergy is not necessarily related to
morphology and sometimes only found for one polymer property or not at all.
One of the first industrial applications where nanosilica was used together with
reactive liquid rubbers were structural epoxy adhesives. We found an increase in
adhesive lap shear strength at low addition levels of nanosilica (<2 wt%) of a
one-component, heat-curing adhesive [3.2.13]. The toughness seemed not to be
increased further compared to the formulation without nanosilica. This might be
due to the fact that the curing agent of this adhesive, dicyandiamide, forms very
close-meshed molecular networks. The CTBN rubber in the formulation had 26 %
acrylonitrile in the copolymer.
Caccavale [3.2.15] studied epoxy resin systems modified with the same rubber
and cured with another commercial polyamine hardener. The loss in modulus due
to a 7.3 wt% rubber modification was only partially compensated by the addition
of 3.7 wt% nanosilica and still approx. 6% lower than the control. Tg was lowered
significantly by 19 °C. KIc of the hybrid however was increased by 136% (dry
conditions) respectively 154% (wet conditions). Microscopical investigations
revealed a good dispersion of both silica nanoparticles and rubber domains
formed upon cure. Moisture absorption of the hybrid system was higher than for
the unmodified epoxy and this could be assigned to the CTBN modification.
69
Tsai et al. [3.2.16] based their research upon the same rubber with 18 %
acrylonitrile in the copolymer and isophorone diamine as a hardener. They
reported a fair dispersion of both rubber domains formed upon cure and silica
nanoparticles. The modulus of the unmodified system was lowered by 10 wt% of
CTBN from 3.25 GPa to 2.63 GPa compared to the control and brought back to
3.18 GPa by the addition of 10 wt% nanosilica. Similar behaviour was found for
the tensile strength - a loss of approx. 20% for the rubber-modified polymer and
approx. 3% for the hybrid. Fracture toughness by means of GIc was increased by
516% to 1170 J/m2 by the addition of the rubber. The hybrid system achieved
only 930 J/m2. This indicates a tough and stiff system, but no synergistic effects.
In the basic study they published further details [3.2.18]. The rubber used
contained 18% acrylonitrile in the copolymer. Piperidine was the curing agent of
choice. Compressive moduli are improved only slightly compared to the
unmodified control when nanosilica is added; with the values being lower for the
20 nm particles than for the 80 nm particles. The 80 nm particles seem to have a
smaller effect on the toughness; the addition level of the rubber dominates. The
20 nm particles, however, do increase the fracture toughness up to a maximum
at approx. 6 vol% nanosilica. Looking at the fracture energy of 3250 J/m2 for the
unmodified system containing approx. 21 vol% of CTBN, the 80 nm particles
increase the fracture energy up to 4310 J/m2 at 1.5 vol% addition level. The 20
nm particles show with 5720 J/m2 a maximum at 3.1% addition level. A similar
toughness was found with 12 and 17 vol% rubber. Pearson observed
agglomerate formation (clustering) of the nanosilica particles at rubber
concentrations above 12 vol%.
Robinette and his group [3.2.19] used two commercial epoxy resin formulations
which contain a CTBN adduct in the resin part. They modified the resin part with
nanosilica and adapted the amount of amine-based hardener accordingly. The
tougher, low Tg system called SC-15 showed an increase in modulus from 2.37
GPa to 2.96 GPa at 15 wt% addition level of nanosilica.
70
This equals an improvement of 25%. Tg seemed to be nearly unaffected. GIc
however, increased from 900 J/m2 to 1020 J/m2 at 1 wt% nanosilica.
For higher loading levels the GIc was more or less at 800 J/m2. In contrast, the
high Tg system SC-79 showed no increase in modulus at first, then a somewhat
higher modulus at 7.5 and 10 wt% addition level, then a drop in modulus again.
The glass transition temperature was lowered with increasing nanosilica
concentration from 180 °C to 162 °C at 15 wt% nanoparticles. GIc was increased
from 180 J/m2 to 380 J/m2 at 10 wt%, then drops again to 250 J/m2 at 15 wt%.
Apparently both systems have an optimum nanosilica addition level, however at
totally different concentrations. Figure 3.2.3 shows the good distribution of the
nanosilica and the typical rubber domains formed upon cure. It is interesting to
see that no silica nanoparticles are trapped in the rubber domains upon their
formation during cure.
Figure 3.2.3: TEM picture of epoxy polymer modified with reactive liquid rubber
(CTBN) and nanosilica [3.2.20]
Sun et al. [3.2.21] worked with the SC-79 epoxy resin system as well. Modulus
increased with increasing nanosilica addition level up to approx. +40% for a 15
wt% loading level. Flexural strength and strain at break showed a similar
behaviour. Additionally they investigated combinations with other nanoparticles
like alumina or carbon nanofibers but could not find further improvements.
Another team investigating the SC-79 two part epoxy system was the group of
Renukappa et al [3.2.22]. They focused on tribological and electrical properties
and reported a maximum for strength at 10 wt% nanosilica. Wear resistance was
increased significantly and showed a maximum at 10 wt% as well.
71
Dielectric strength, arc and track resistance were improved first with increasing
the nanosilica content, showed a maximum at 15 wt% and decreased with 20
wt% below the level of the unmodified control.
Tetrafunctional epoxy resins, cured with sterically hindered aromatic amines are
used in high performance aerospace applications. We examined such a system
and observed a further increase in toughness by the addition of 10 wt%
nanosilica to the reactive liquid rubber modified system [3.2.24]. The loss in
modulus due to the rubber modification was overcompensated by the nanosilica.
Thus a tough and stiff resin system could be formulated. Dry Tg was not affected
by the modifications; however the wet Tg (after two weeks at 70 °C and 100 %
relative humidity) was the lowest for the hybrid system.
3.2.3.3. Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, anhydride cured
In an early investigation we used the reactive liquid rubber with 18% acrylonitrile
in the copolymer and methyl hexahydrophthalic acid anhydride (MHHPA)
accelerated with a small amount of a ternary amine [3.2.26]. Modulus was
lowered by the rubber modification and increased again by the nanosilica
addition to the level of the unmodified system. This system had a fracture energy
(GIc) of 103 J/m2, which was increased to 406 J/m2 by the addition of 9 wt%
CTBN. Adding 4.5 wt% nanosilica pushed the GIc to 917 J/m2; 9 wt% nanosilica
to 973 J/m2. Compared to the control this is an increase of approx. 850%. The
glass transition temperature dropped by the rubber modification from 143 °C to
133 °C. The nanosilica addition had no influence on Tg.
Manjunatha et al. [3.2.27], working with the same system, investigated the cyclic
fatigue behaviour. Tensile strength and modulus of the polymer were lowered by
the rubber addition and increased by the nanosilica addition. The system with 9
wt% CTBN and 10 wt% nanosilica had approx. 10 % lower strength and
modulus.
72
The cyclic loading tests at different stress levels showed a clear shift towards
much more cycles before failure for both the rubber and the nanosilica modified
polymer. However, the hybrid performed best and gave a 6-10 times
enhancement of fatigue life.
Kinloch, Taylor et al. [3.2.28] used the same system and reported significant
improvements in toughness. Though the modulus of the hybrid system was
slightly lower than for the unmodified system (-4%), the fracture energy GIc was
significantly increased by the addition of 9 wt% CTBN and further increased by
the nanosilica addition. A maximum was found at 15 wt% nanosilica: 965 J/m2
compared to 77 J/m2 for the control which equals an improvement by 1150%. At
20 wt% nanosilica the toughness shifts to lower values. Interestingly the
formation of small silica agglomerates was observed, but did not seem to affect
the rubber domain formation nor the performance of the cured resin.
Zhang, Tang et al. [3.2.29] used a powdered CTBN rubber for their study,
together with accelerated methyl hexahydrophthalic acid anhydride. They
reported well dispersed rubber particles and well dispersed silica nanoparticles.
No agglomerate formation could be observed. The elastic modulus of all hybrid
formulations was lower than for the unmodified system with a modulus of 3.01
GPa; except for the formulation with 7 vol% nanosilica and 2 vol% rubber which
achieved a modulus of 3.25 GPa. GIc of the control was 62 J/m2 and increased to
186 J/m2 for 7 vol% nanosilica and 2 vol% rubber. 2 vol% nanosilica and 7 vol%
rubber gave a further increase to 371 J/m2. However, the best overall
performance was shown for the system with 4.5 vol% silica and 4.5 vol% rubber
with a GIc of 306 J/m2 (equals an improvement of approx. 394%), a slightly
lower modulus (-4%) and the highest impact energy of all systems tested. 31.29
kJ/m2 compared to 16.2 equals an improvement of approx. 93%. The glass
transition temperature was reported to be unaffected by the modification which
indicates a particulate nature of the rubber similar to core shell materials.
3.2.3.4. Epoxy resins modified with silica nanoparticles, cured with amines
and amino-functional reactive liquid rubbers (ATBNs)
Totally different systems can be achieved when amine functional reactive liquid
rubbers are used to modify an epoxy resin. They are used as part of the
hardener blend in 2 part epoxy formulations, e.g. structural adhesives.
73
Compared to epoxy resins modified with CTBN adducts a much smaller
percentage of the rubber molecules phase separate and form rubber domains. A
large proportion is crosslinked randomly into the polymer and lowers the
crosslink density significantly - with the known effects of a much lower Tg, lower
modulus, increased elongation at break, increased toughness etc.
In continuation of this work we varied the rubber level of the formulations and
found similar behaviour: there was always a significant property improvement at
low addition levels of nanosilica (1-4 wt%) [3.2.32]. The same was found for
roller peel tests at different temperatures (substrate chromic acid etched
aluminium) as well as for wedge impact test with adhesive joints of degreased
steel. Most striking were the results of the adhesive fracture energy test where
the GIc was doubled from 1200 J/m2 to 2400 J/m2 by the addition of approx. 1
wt% nanosilica.
74
3.2.3.5. Epoxy resins modified with core-shell elastomers (CSRs) and silica
nanoparticles, amine cured
CSR Type I
Mai et al. [3.2.34] studied the cyclic fatigue behaviour of an epoxy resin system
modified with a core-shell elastomer characterized by a particle diameter of
approx. 110 nm. Piperidine was used as a curing agent. They reported a
reduction in strength by the addition of the core-shell rubber, which was not
improved by the addition of the 20 nm silica nanoparticles. The loss in modulus
however was compensated; it dropped from 2.86 GPa to 2.25 GPa at 10 wt%
elastomer and was brought back to 2.81 GPa with 10 wt% nanosilica. The silica
nanoparticles increased the fatigue life whereas the core-shell particles
decreased it. Thus the hybrid system did not perform better than the neat
polymer. Toughness was increased from 277 J/m2 to 1250 J/m2.
This work was continued by Liu et al. [3.2.35]. They looked further into
toughness and toughening mechanisms. Core-shell particles and silica
nanoparticles were well distributed, as TEM microscopy revealed. GIc for the
unmodified resin was 277 J/m2. The elastomer addition increased GIc with raising
addition level. At 10 wt% core-shell the GIc was 1930 J/m2. An increase in
toughness with the addition of nanosilica was found as well. At 10 wt% nanosilica
GIc was 690 J/m2. The hybrid system with 10 wt% of both modifications had the
highest toughness of 2480 J/m2. This seems to imply an additive behaviour, no
synergistic effects. However a good balance between toughness and modulus
could be achieved. As major toughness contributors nano-silica debonding and
bridging before pullout and core-shell rubber cavitation and matrix plastic
shearing were identified. They claimed that nanosilica and core-shell elastomer
act independently during crack growth.
CSR Type II
Tsai et al. [3.2.16] used the larger Type II particles and isophorone diamine as a
hardener. TEM microscopy showed a good dispersion of both silica nanoparticles
and elastomer particles as well. The modulus was reduced from 3.25 GPa to 2.73
GPa by the addition of 10 wt% core shell particles and brought back to 2.97 GPa
by further addition of 10 wt% silica nanoparticles. The tensile strength was
reduced by approx. 27 % by the elastomer and this reduction was not changed
by the nanosilica addition - though the system modified only with nanosilica
exhibited no loss in strength. The fracture toughness of the epoxy resin was
determined to be 190 J/m2. It was increased by a 10 wt% core-shell modification
to 1420 J/m2. A modification with 10 wt% of nanosilica achieved 280 J/m2. The
hybrid system with 10 wt% of both modifications achieved only 1030 J/m2. A
good balance between toughness and modulus, but no synergistic effects.
75
In continuation of this work Tsai et al. [3.2.6] investigated the damping
properties at room temperatures and found a 42% improvement of the loss
factor for the hybrid system.
Figure 3.2.4: TEM picture of epoxy polymer modified with 8 vol% Type III core-
shell elastomer and 5.5 vol% silica nanoparticles [3.2.11]
3.2.3.6. Epoxy resins modified with core-shell elastomers (CSRs) and silica
nanoparticles, anhydride cured
CSR Type I
Figure 3.2.6: TEM picture of epoxy polymer modified with 9 wt% Type I core-
shell rubber and 6 wt% silica nanoparticles
78
Figure 3.2.7: TEM picture of epoxy polymer modified with 9 wt% Type I core-
shell rubber and 6 wt% silica nanoparticles
Having discussed all research results in detail, Table 3.2.1 gives an overview of
the results of some selected references; regardless of the resin/hardener
combination.
As can be seen, the modulus of the hybrids is a few percent lower than for the
unmodified systems; especially for higher rubber or elastomer levels.
All systems are tough and stiff, which means that the best of both modifications
is found in the hybrid systems.
79
Resin / Modifications Modulus Improv. GIc Improv. Ref.
Hardener (GPa) versus (J/m2) versus
control control
DGEBA 4.6 wt% CTBN* 2.89 -4% 241 + 289 % [3.2.29]
MHHPA 8.5 wt%
nanosilica*
DGEBA 9 wt% CTBN 2.77 -6% 683 + 787 % [3.2.28]
MHHPA 9 wt%
nanosilica
DGEBA 9 wt% CTBN 2.85 -4% 965 + 1153 % [3.2.28]
MHHPA 15 wt%
nanosilica
DGEBA unknown, but 3.25 - 21 % 380 + 111 % [3.2.19]
commercial low % CTBN
amine I 10 wt%
nanosilica
DGEBA 11.3 wt% CTBN 2.68 - 22 % 1843 + 203 % [3.2.14]
commercial 5.7 wt%
amine II nanosilica
TGMDA 8 wt% CTBN 3.37 +4% 708 + 302 % [3.2.24]
aromatic 10 wt%
amines nanosilica
DGEBA 10 wt% CSR I 2.78 -3% 2480 + 795 % [3.2.35]
Piperidine 10 wt%
nanosilica
DGEBA 10 wt% CSR II 2.97 -9% 1030 + 442 % [3.2.16]
Isophorone 10 wt%
diamine nanosilica
DGEBA 5.1 wt% CSR 3.50 +9% 1851 + 300 % [3.2.38]
commercial III*
amine 11 wt%
nanosilica*
*recalculated from vol%
If Figures 3.2.3 to 3.2.7 are compared, the morphology of the fracture surfaces
looks identical, though the elastomer particles have different sizes. However,
what looks the same, behaves differently regarding micro-mechanics.
We found in our investigations with the Type III core-shell rubber no synergy
between core-shell particles and silica nanoparticles. Toughness increases
compared to the control were additive. Similar findings have been reported by
Liu et al. [3.2.35], [3.2.36] for Type I core-shell rubbers except for fatigue
performance where a synergy seems to exist. Tsai et al. [3.2.16] could not
confirm the existence of a synergy for Type II core shell particles either.
80
No agglomeration was reported for all systems containing both core-shell rubber
particles and silica nanoparticles.
Pearson et al. [3.2.41] investigated two different sizes of nanosilica (20 nm and
80 nm) in piperidine cured epoxy resin. They reported a neglible effect of the
particle size on fracture toughness and compressive properties. Thus it seems to
be a fair assumption to consider these small agglomerates or clusters as
somewhat larger nanoparticles which do not or nearly not affect the performance
of the cured epoxy polymer.
Recently Pearson [3.2.18] studied the hybrids based on the two different particle
sizes and reactive liquid rubber. He found no considerable effect of particle size
on the toughening of the hybrid nor could he confirm a synergy.
A correlation between very high, synergistic values of GIc and the observation of
small nanosilica agglommerations or clusters could not be established.
It is well known that the GIc of rubber toughened epoxy polymers increases
dramatically with a lower crosslink density [3.2.42]. This is the reason why many
industrial formulations contain the CTBN rubber adduct along with a small
amount of a solid DGEBA with a higher molecular weight. The long chain
molecules of the solid epoxy resin lower the crosslink density to a certain extent
and make the resin matrix more ductile as well as the rubber toughening more
efficient.
When silica nanoparticles are added to an epoxy resin modified with CTBN, the
proportion of rubber molecules not participating in the phase separation upon
cure increases with increasing addition level of nanosilica [3.2.28]. If more long
flexible rubber molecules are crosslinked randomly into the polymer matrix, the
crosslink density is lowered and the ductility of the matrix is increased - thus
making the rubber toughening more efficient. This could be the explanation of
the synergistic improvement of GIc and fatigue performance for hybrid systems
based upon the combination CTBNs and nanosilica.
81
As indication for the lower crosslink density can be seen the development of the
glass transition temperature [3.2.28]: The addition of nanosilica does not lower
the Tg compared to the unmodified epoxy polymer system. This is different for
the CTBN modified system. At a 9 wt% rubber addition level the Tg is 150 °C.
Further addition of silica nanoparticles lowers the Tg to 145°C at 10 wt% and to
140 °C at 20 wt%.
Though the difference in Tg might seem small, this is not unusual as minor
changes in crosslink density can have a big impact on toughness. Pearson
[3.2.42] reported for two rubber-toughened epoxies with a small difference in
crosslink density and a difference in Tg of only 5 °C an increase in fracture
toughness by factor three.
Still there is a need for additional research to explain the synergy, like the nature
of interactions at the tip of a forming crack.
3.2.4. Conclusions
Taking into account all the data accumulated, a couple of conclusions can be
drawn:
82
3.2.5. References
[3.2.1] Kinloch, A.J.; Shaw, S.J.; Tod, D.A. Polymer (1983), 24, 1341–
1354.
[3.2.3] Bagheri, R.; Marouf, B.T.; Pearson, R.A. J. Macromol. Sci. Part C:
Polymer Reviews (2009), 49, 201-225.
[3.2.4] Giannakopoulos, G.; Masania, K.; Taylor, A.C. J. Mater. Sci. (2011),
46, 327-338.
[3.2.5] Lai, M.; Friedrich, K.; Botsis, J.; Burkhart, T. Composites Sci. Tech.
(2010), 70, 2168-2175.
[3.2.6] Tsai, J.-L.; Chang, N.-R. Journal of Composite Materials (2011), 45,
10/2011, 2157–2164.
[3.2.8] Block, H.; Pyrlik, M. Kunststoffe German Plastics (1988), 78, 29.
[3.2.9] Chen, J.; Kinloch, A.J.; Sprenger, S.; Taylor, A.C. Polymer accepted
for publication May (2013)
[3.2.13] Sprenger, S.; Eger, C.; Kinloch, A.J.; Lee, J.H.; Taylor, A.C.; Egan,
D. Adhäsion, Kleben & Dichten (2003), 03/2003, 24-30.
[3.2.14] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D. JEC
Composites Magazine (2005), No 21, 66–69.
83
[3.2.20] Picture with permission of J. Robinette, Army Research Laboratory
(see [3.2.19] as well)
[3.2.21] Uddin, M.F.; Sun, C.T. Composites Sci. and Tech. (2010), 70, 223-
230.
[3.2.24] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D. JEC
Composites Magazine (2007), No 30 January-February, 54–57.
[3.2.26] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D.; Eger, C.
JEC Composites Magazine (2005), No 19 August/September, 73–
76.
[3.2.28] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Sohn Lee, J.; Taylor, A.C.;
Sprenger, S. J. Mater. Sci. (2010), 45, 1193-1210.
[3.2.29] Tang, L.-C.; Zhang, H.; Pei, Y.-B.; Chen, L.-M.; Song, P.; Zhang, Z.
Composites: Part A (2013), in press
[3.2.30] Tang, L.-C.; Zhang, H.; Sprenger, S.; Ye, L.; Zhang, Z. Composites
Science and Technology (2012), 72, issue 5, 558-565.
[3.2.31] Sprenger, S.; Eger, C.; Kinloch, A.J; Lee, J.H.; Taylor, A.C.; Egan,
D. Journal of Adhesion (2003), 79, 867-873.
[3.2.32] Sprenger, S.; Eger, C.; Kinloch, A.J.; Lee, J.H.; Taylor, A.C.; Egan,
D. Adhäsion, Kleben & Dichten (2004), 03/2004, 17-21.
[3.2.33] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Hsieh, T.-H. Adhesion,
Adhesives & Sealants (2009), 10/2009, 8-11.
[3.2.34] Wang, G.-T.; Liu, H.-Y.; Saintier, N.; Mai, Y.-W. Eng. Failure
Analysis (2009), 16, 2635-2645.
[3.2.35] Liu, H.Y.; Wang, G.T.; Mai, Y.W.; Zeng, Y. Composites: Part B
(2011), 42, 2170-2175.
[3.2.36] Liu, H.-Y.; Wang, G.; Mai, Y.-W. Composites Sci Tech. (2012), 72,
1530-1538.
84
[3.2.37] Palinsky, A. Composites Processing (2006), 27 April 2006,
Merseyside, UK
[3.2.42] Pearson, R.A.; Yee, A.F. J. Mat. Sci. (1989), 24, 2571–2580.
85
3.3. Property improvements of fiber-reinforced composites
based on epoxy resins modified with SiO2 nanoparticles
3.3.1. Abstract
3.3.2. Introduction
Mineral fillers like calcium carbonate or quartz powder can improve mechanical
properties of epoxy resins like strength, stiffness and modulus. However, they do
increase the viscosity of the resin significantly.
86
Due to their size, typically microns, they cannot be used in injection
manufacturing methods because they are filtered out by the fabric upon
injection. In prepreg manufacturing they tend to sediment in the process. Thus in
most composites applications no mineral fillers are used.
Used in resin systems for composite applications, they offer several advantages:
being 20 nm small and completely monodisperse, they do increase the resin
viscosity only slightly at higher addition levels. In contrast to fumed silica they do
not exhibit thixotropic properties but behave like a Newtonian liquid. Due to their
size they can easily penetrate even close meshed fabrics in composites
manufacturing when being injected.
Figures 3.3.2 and 3.3.3 show TEM pictures of glass fiber reinforced composites
(GFRC) made by using an epoxy resin containing silica nanoparticles [3.3.3]. The
laminates have been made by using VARTM.
87
The uniform dispersion of the particles is evident and the difference in size
between fiber and particle is obvious - thus it is self-explaining why there is no
gradient in nanosilica distribution even after quite a distance from the point of
resin injection.
As silica nanoparticles are a very attractive raw material for epoxy resin
formulators, they are used today, 10 years later, in many industrial composites
applications. They improve various properties like tensile strength, tensile
modulus, flexural stiffness, toughness and scratch resistance.
88
Nevertheless, it is a difficult task to determine the optimum addition level in
function of resin and hardener used. Very often the improvements found for a
modified bulk resin [3.3.2] are not translated into the fiber-reinforced composite
made from this resin.
The aim of this review is to give a comprehensive overview regarding the actual
state of research, to understand the mechanisms of property improvements of
bulk resins as well as laminates and to provide formulating guidelines. First
results for laminates with nanosilica at the fiber/matrix interface are presented.
3.3.3. Discussion
In continuation of this work they investigated in-plane shear strength and off-
axis compressive strength [3.3.6]. DGEBA and a short-chain polyether diamine
as curing agent were blended to prepare thin laminates (1.45 mm, 5 layers of
UD, 10° off axis). Again hand lay-up and cure under pressure and vacuum were
used. The in-plane shear strength was increased from 23.7 MPa to 27.5 MPa at 5
wt% silica nanoparticles to 30.1 MPa at 10 wt% and 29.6 MPa at 20 wt% loading
level. This equals improvements of 16%, 27% and 25%, respectively.
To test the off-axis compressive strength thick laminates (6.15 mm, 22 layers of
UD, 0°, 5°, 10°, 15° and 90° off axis) were prepared. The unmodified control
was compared to a system with 20% of silica nanoparticles. The improvements
were different for the different fiber orientations: + 7% at 0°, + 17% at 5°, +
6% at 10°, +6 % at 15° and + 2% at 90°. For 0° the failure mechanism was
dominated by fiber splitting. At 90° the out-of-plane shear failure mechanism
was observed. For 5°, 10° and 15° the main failure mode was microbuckling
where improvements can be expected to be found as the silica nanoparticles do
increase the modulus of the epoxy resin matrix.
Tate et al. [3.3.7] used a commercial epoxy resin system, aliphatic amine cured,
which they modified with silica nanoparticles. They used a -45°/90°/+45°
bidiagonal stitched bonded fabric (E-Glass) and VARTM to prepare the laminates
(5.95 mm, 8 layers of fabric). Tensile strength was improved from 92.1 MPa to
122.4 MPa at 6 wt% silica nanoparticles, which equals an improvement of 33%.
At 7 wt% 109.5 MPa and at 8 wt% 101.4 MPa could be achieved. A similar
behaviour was found for the modulus which increased from 10.3 GPa to 13.2 GPa
at 6 wt% nanosilica, again an improvement of 29%.
7 wt% and 8 wt% achieved only 12.8 GPa and 11.7 GPa, respectively; which
might indicate the existence of a plateau in performance. Or, there might be
further improvements found at 10 wt% or 15 wt% addition levels – which have
not been investigated.
90
The interlaminar shear strength (ILSS: ASTM D 2344) showed the same
tendency: it went up from 24.5 MPa to 31.0 MPa (+27%) with 6 wt% silica
nanoparticle modification. With 7 wt% and 8 wt% 30.1 MPa and 29.8 MPa were
obtained. A slightly different picture was found for flexural strength and flexural
modulus. Strength increased from 146 MPa to 177 MPa to 181 MPa and 180 MPa.
Modulus was improved from 25.0 GPa to 29.5 GPa to 30.5 GPa to 30.7 GPa for 6
wt%, 7 wt% and 8 wt% nanosilica.
If not mentioned otherwise, the systems investigated are all based on DGEBA,
cured with methyl hexahydrophthalic acid anhydride; accelerated with a very
small amount of a ternary amine.
Mahrholz et al. [3.3.8], [3.3.9] published very early promising results. They
made laminates with a bidirectional glass fiber at +45°/-45° orientation using
the SLI technique. At a 20 wt% addition level of silica nanoparticles, the tensile
strength was improved by 25%, tensile modulus was increased by 64% and the
shear modulus was increased by 75%.
Sprenger et al. [3.3.12] tested similar laminates with a falling dart impact test
(30 J) and could not detect any improvements. The delaminated area had exactly
the same size with or without silica nanoparticles present in the resin - see
Figure 3.3.4. Unfortunately, the compression after impact (CAI) test results have
not been released for publication.
Kuehn et al. [3.3.13] found slightly improved CAI properties for laminates with
nanosilica as will be shown in the chapter carbon fiber-reinforced composites.
The question which mechanisms of toughness improvements by silica
nanoparticles apply to fiber-reinforced composites and wether they work at high
speed impact will be discussed later.
91
Manjunatha et al. [3.3.14] investigated the tensile fatigue behaviour. They used
a fabric with a +45°/-45° pattern, laid up 16 ply in a quasi-isotropic sequence
(+45°/-45° alternating with 90°/0°). Infusion temperature was 50 °C. The
tensile strength was increased from 365 MPa to 381 MPa (+4%); modulus was
increased from 17.5 GPa to 18.8 GPa (+7%). The 10 wt% silica nanoparticles
improved significantly the fatigue performance of the laminate. At a very high
stress (225 MPa) the number of cycles until failure was more than doubled.
Figure 3.3.4: Photograph of GFRC after impact testing without and with
nanosilica [3.3.12]
92
300
Neat matrix
(MPa)
Modified matrix
250
max
Maximum Stress, s
200
150
GFRP composite
R = 0.1
n = 1 to 3 Hz
RT, Lab air
100
50
1.E+02 1.E+03 1.E+04 1.E+05 1.E+06
No. of cycles, N f
Figure 3.3.5: Stress versus lifetime for a GFRC with epoxy resin matrix;
unmodified and with 10 wt% nanosilica [3.3.15]
Kinloch, Taylor and their team prepared UD GFRCs, 12 ply, 6-mm thick and
investigated the toughness [3.3.16]. 10 wt% silica nanoparticles increased the
GIc from 330 J/m2 to 1015 J/m2, this equals an improvement of more than 200%.
The bulk resin GIc was improved by 130% at 11 wt% nanosilica. They reported
small agglomerates of silica nanoparticles for the bulk resin, which seemed to
have no negative influence on performance. Quasi-isotropic laminates, 8 ply, 4-
mm thick by using a biaxial stitched fabric laid up to give a 0°/0° interface
across the fracture plane were prepared. Interestingly the GIc was lowered by the
nanosilica modification from 718 J/m2 to 626 J/m2 (-13%). This is very probably
due to the different fibre architecture.
93
Figure 3.3.6 compares the fracture surfaces of laminates without and with silica
nanoparticles. Crack propagation was in the direction of the fibers; from left to
right. Disregarding the different magnifications, there seems to be no big
difference for the fracture surface; both exhibit brittle failure with little plastic
failure.
In her thesis more details were given [3.3.18]. Cyclic fatigue performance was
not improved either by the silica nanoparticles addition. Improvements of the
matrix properties could not be transformed into improvements of the laminate.
The conclusion drawn was that these properties are fiber-dominated and not
improved by the nanosilica addition. However, it should rather be said that the
nanosilica had no influence on toughness and fatigue performance in this case.
Or, eventually higher loading levels of nanosilica are necessary to see an effect.
Tang et al. [3.3.19] used piperidine as curing agent and vacuum-assisted resin
infusion (VARI) to prepare laminates. The unidirectional carbon fiber fabric used
was UT70-20 (24 layers for ILSS testing and 20 layers for mode I and mode II
testing) and the resulting fiber volume fraction was 65 %. The tensile strength of
the bulk epoxy was increased from 64.8 MPa to 69.4 MPa at 10 wt% nanosilica
and to 72.4 MPa at 20 wt% loading. Tensile modulus increased from 2.9 GPa to
3.3 GPa to 3.6 GPa. GIc increased from 237.9 J/m2 to 457.9 J/m2 at 10 wt%
nanosilica to 666.4 J/m2 at 20 wt% nanosilica. These are improvements of 93%
and 180%, respectively.
The laminates with the modified resins were tested extensively. The G Ic of the
laminate (at crack propagation) was improved from 995 J/m2 to 1007 J/m2 at 10
wt% nanosilica and to 1203 J/m2 at 20 wt%; an improvement of 1% and 21%
only. The ILSS was lowered by the nanosilica addition by 3% and 12%. GIIc was
lowered as well – from 969 J/m2 to 929 J/m2 and to 750 J/m2 (-23 %). The
conclusion drawn was that especially for the high concentration of 20 wt% of
silica nanoparticles the shear strength of the matrix was reduced.
Liu et al. [3.3.20] investigated laminates made from plain woven carbon fibers,
18 ply, with a fiber volume fraction of 60%. Piperidine was used as curing agent,
VARI as manufacturing method.
94
GIc of the laminate was improved by approx. 24 % for 4 wt%, 6 wt%, 8 wt% and
12 wt% silica nanoparticles, whereas the GIc for the bulk resin is increased
linearly with increasing nanosilica level for more than 100%. This could indicate a
plateau in performance improvements already at 4 wt% nanosilica. Of course it
has to be taken into account that woven fabrics have a big influence on GIc;
which might “hide” some effects of the silica nanoparticles.
A completely different test specimen and totally different properties were studied
by Zhang and his team [3.3.21]. They prepared short carbon fiber reinforced
epoxy resins; cured with a cycloaliphatic amine. Tenax A-385 short carbon fibers
with an average diameter of 7 µm and a length of 40 -70 µm were added at 10
vol% as well as 8 vol% graphite flakes. Silica nanoparticles were introduced in
addition levels of 1–5 vol%. The distribution of the nanosilica was uniform;
strength and modulus were increased. However, the main topic of the
investigation was related to the wear properties on both rough and mirror-
polished surfaces. The coefficient of friction was lowered significantly by the
spherical nanosilica particles. The wear rate was improved by approximately
50%. A maximum in performance improvements was found for 3 vol% silica
nanoparticles which equals approx. 5 wt%.
The modulus of the bulk resin was increased with increasing nanosilica addition
level by more than 100% at 45 wt% nanosilica. GIc was increased by approx.
50%. Bulk strength remained unchanged up to 35 wt% and then increased
slightly. The compressive strength of the laminate was increased from 1.8 GPa to
2.0 GPa at 45 wt% nanosilica which equals an improvement of 11%. At 15 wt%
1.8 GPa or an improvement of 3% were found.
95
Flexural modulus was found unchanged even at very high addition levels of silica
nanoparticles and is apparently fiber-dominated. The flexural strength was
increased from 1.5 GPa to 1.9 GPa at 45 wt% (an improvement of 26%). At 15
wt% the improvement found was approximately 10%.
In recent papers they revealed more details [3.3.25], [3.3.26]: imidazole was
used as hardener. The prepregs were prepared using T300-3k and T700-12k 2x2
twill carbon fabrics. Four types of laminates were made, three with T300: 8 ply
0° for compression tests, 14 ply 0° for flexural testing, 36 ply +-45° for in-plane
shear. The fourth was made from 1 layer T300 followed by 6 respectively 8
layers of T700 and finally a last layer of T300. Nanoindentation tests showed an
increased strength and hardness. Vickers hardness tests revealed similar
improvements.
96
The compressive strength increased quickly from 825.9 MPa to 1150.6 MPa at
2.1 vol%, then to 1268 MPa at 5.5 vol% (54% improvement) and finally dropped
to 1172 MPa at 11.7 vol % silica nanoparticles. Strain at break did not change
significantly. It is clear that the presence of nanosilica in the resin improves the
compressive properties of a laminate significantly. This finding is used in quite a
few industrial applications, like composite leaf springs.
Hsieh et al. [3.3.16] investigated the same system. The GIc of the bulk resin was
increased from 77 J/m2 with increasing silica addition level to 212 J/m2 at 20
wt% nanosilica which represents an improvement of 175%. Modulus was
improved from 2.96 GPa to 3.85 GPa at 20 wt% nanosilica. Laminates were
prepared by stacking a linen woven carbon fiber fabric (0°/90° weave) and using
VARTM. The modified system contained 12 wt% silica nanoparticles. GIc was
increased from 439 J/m2 to 489 J/m2, a minor improvement of only 11%. Figure
4.3.7 shows the fracture surfaces of the laminates with the unmodified and
modified epoxy resin matrix. Apparently the presence of silica nanoparticles has
no significant influence on the fracture behaviour in this fiber-reinforced system.
The research of Liu and his team [3.3.31] was based on DGEBA as well and used
a blend of methylhexahydrophthalic acid anhydride and sebacic anhydride as
hardener. The tensile strength of the bulk resin was increased from 74.0 MPa to
79.7 MPa at 2 vol% nanosilica and to 86.7 MPa at 14 vol% nanosilica.
The elastic modulus was increased from 2.6 GPa to 2.8 GPa (at 2 vol% SiO2)
respectively 3.8 GPa at 14 vol% SiO2. Single T-300 carbon fibers were embedded
into the center of the test specimen mold. Single-fiber fragmentation tests were
performed.
97
Figure 3.3.7: Scanning electron micrographs of fracture surfaces of CFRC with an
epoxy matrix: (a) unmodified; (b) with 12 wt% nanosilica [3.3.16]
The critical break density (number of breaks / mm at failure) of the fiber was
improved by 19% at 2 vol% and by 27% at 14 vol% nanosilica. The interfacial
shear strength was calculated and showed improvements of 21% and 30%
respectively. Taking into account the results of micro-Raman spectroscopy and
SEM microscopy, they concluded, two effects to be responsible: The stress
transfer from matrix to fiber is made more efficient by the nanosilica particles
and the nanoparticles increase the interfacial frictional stress between fiber and
matrix.
By the addition of silica nanoparticles strength and modulus of cured bulk epoxy
resins were increased, of course, depending on the epoxy resin/hardener
combination, by different magnitudes.
Unfortunately, very often either bulk or laminate properties were studied by the
researchers, but not both. Few report GIIc data.
98
Thus Table 3.3.1 gives an overview of the GIc data provided by the researchers
who investigated both bulk and laminate properties. It becomes evident, that the
significant improvements found for the modified bulk resins are not reflected in
the fiber-reinforced laminates.
This raises the question regarding the toughening mechanisms which apply for
the fiber-reinforced materials and how they differ from the toughening
mechanisms of cured bulk resins.
The mechanisms for the property improvements of bulk epoxy resins have been
well investigated and are fully understood [3.3.2]. Two main toughening
mechanism have been identified for silica nanoparticles: Localised shear-bands
initiated by the stress concentrations around the periphery of the nanoparticles
and the debonding of the silica nanoparticles followed by subsequent plastic void
growth.
This might explain why the improvements in toughness found are relatively small
compared to bulk resins, as the major failure “source” is still the failure at the
interface between fiber and resin matrix.
99
Another issue to take into account is the effect of strain rate – GIc testing versus
falling dart test. There is abundant literature available which to discuss would go
beyond the scope of this paper. Jacob et al. [3.3.32] published a comprehensive
overview for various fiber reinforced composites.
As the modification of the matrix resin apparently has a rather small effect on
fiber matrix adhesion, this could explain why we found no difference in
delaminated area between control and modified system for the falling dart tests.
3.3.3.5. How much silica nanoparticles do you really need – and where?
A very interesting and promising aspect was researched by Gao et al. [3.3.33].
They investigated changes in laminate properties when the silica nanoparticles
are only present in the interphase between fiber and resin. Fiber sizings
containing 1% of 22 nm silica nanoparticles without surface modification from a
different source were prepared and applied to the unsized glass fibers. An
increase in fiber surface roughness was observed. They reported superior
interphase performance compared to the unsized control: strength, debonding
and total energy absorption were improved. Failure modes were studied by SEM
and it was found that the crack was deflected which apparently increased energy
consumption.
An increase in ILSS was reported: +14% compared to unsized fibers and +5%
compared to a sizing without nanosilica. They concluded that the interfacial
adhesion between the fiber and the matrix is increased. The crack propagation
path in the sizing layer was found to be changed by the presence of the
nanosilica.
The first tests performed were fracture toughness tests. Figure 3.3.8 shows the
improvements in GIc of the laminates made from pretreated UD. An addition level
of 1.5 wt% aqueous emulsion equals 0.36 wt% silica nanoparticles on fiber
weight; 4 wt% equal 1.12 wt% silica nanoparticles on fiber weight.
100
Additionally, an increase in the adhesion between fibers and matrix was indicated
by the presence of a significant amount of resin bonded on the fibers when the
fracture surfaces of the modified samples were observed, a feature that was not
observed in the case of control fracture surfaces.
Figure 3.3.8: GIc of pretreated glass fiber reinforced composites made from
commercial unmodified epoxy resin
If more data will confirm that for substantial improvements of the laminates it
will be sufficient to apply thin layers on the fabric, this will be of huge
commercial interest. Calculated on total it would mean that far less than 1 wt%
silica nanoparticles would be enough, which is negligible from a cost aspect –
compared to modify the whole resin matrix.
Of course this field needs further future research to fully understand such
systems.
3.3.4. Conclusions
101
2. Not taking into account the reinforcing fibers, their volume fraction and
orientation respectively the layup used and disregarding the hardeners
used a general outline can be given: Typical improvements found for fiber
reinforced composites with an epoxy resin matrix modified with 10 wt%-
20 wt% of silica nanoparticles are increases of 5–25% in fracture
toughness.
3. Most significant improvements are found for the compressive properties
and for the fatigue performance.
4. Consequently many industrial composite parts today are based on epoxy
resins modified with silica nanoparticles – glass fiber-reinforced as well as
carbon fiber-reinforced.
5. Improving laminate properties by using silica nanoparticles in the sizing
seems to be a very cost-efficient route, which needs further evaluation.
6. To push the performance levels of fiber-reinforced systems further,
especially regarding toughness and fast impact, the combination of
classical toughening methods with the silica nanoparticle modification is
the next logical step in technology development [3.3.35]. This will be the
subject of future investigation.
3.3.5. References
[3.3.4] Tsai JL, Huang BH and Cheng YL. Enhancing fracture toughness of
glass/epoxy composites by using rubber particles together with
silica. J. of Compos. Mater. (2009) 43, 25, 3107.
[3.3.5] Tsai JL, Huang BH and Cheng YL. Enhancing fracture toughness of
glass/epoxy composites for wind blades using silica nanoparticles
and rubber particles. Procedia Eng. (2011), 14, 1982.
102
[3.3.9] Herbeck L, Mahrholz T, Mosch J, et al. Faserverstärkte Nano-
composites – Stand der Technik und Perspektiven. GAK (2005), 58,
161.
[3.3.10] Kinloch AJ, Masania K, Taylor AC, et al. The fracture of glass-fibre-
reinforced epoxy composites using nanoparticle-modified matrices.
J. Mater. Sci. Letters (2008), 43, 1151.
[3.3.11] Kinloch AJ, Masania K, Taylor AC, et al. The fracture of nanosilica
and rubber toughened epoxy fibre composites. In: Proceedings of
the Composites & Polycon (2009), American Composites
Manufacturers Association, Tampa, Florida, USA, 15-17 Jan. 2009
[3.3.14] Manjunatha CM, Taylor AC, Kinloch AJ, et al. The effect of rubber
microparticles and silica nanoparticles on the tensile fatigue
behavior of a glass-fibre epoxy composite. J. Mater. Sci. Lett.
(2009), 44, 342.
[3.3.15] Manjunatha CM, Taylor AC, Kinloch AJ, et al. The tensile fatigue
behaviour of a silica nanoparticle-modified glass fibre reinforced
epoxy composite. Compos. Sci. Technol. (2010), 70, 193.
[3.3.16] Hsieh TH, Kinloch AJ, Masania K, et al. The toughness of epoxy
polymers and fibre composites modified with rubber microparticles
and silica nanoparticles.; J. Mater. Sci. (2010), 45, 1193.
[3.3.20] Zeng Y, Liu HY, Mai YW, et al. Improving interlaminar fracture
toughness of carbon fibre/epoxy laminates by incorporation of
nanoparticles. Compos. Part B (2012), 43, 90.
103
[3.3.21] Zhang G, Sebastian R, Burkhart T, et al. Role of monodispersed
nanoparticles on the tribological behaviour of conventional epoxy
composites filled with carbon fibers and graphite lubricants, Wear
(2012), 292-293, 176.
[3.3.23] Hackett SH, Nelson JM, Hine AM, et al. The effect of nanosilica
concentration on the enhancement of epoxy matrix resins for
prepreg composites. In: Proceedings SAMPE (2010), Salt Lake City,
UT, U.S.A. 11–14 Oct., 2010
[3.3.24] Nelson JM, Hackett SC, Goetz DP, et al. Development of nanosilica-
thermoset matrix resins for prerpreg composites. NSTI-Nanotech
(2012), ISBN 978-1-4665-6274-5 Vol. 1
[3.3.25] Nelson JM, Hine AM, Goetz DP, et al. Nanosilica-modified epoxy
matrix resin for prepreg composite tooling applications. In:
Proceedings SAMPE (2012), Baltimore, MD, U.S.A. 21–24 May,
2012
[3.3.26] Nelson M, Hine AM, Goetz DP, et al. Properties and applications of
nanosilica-modified tooling prepregs. SAMPE J. (2013), 49, No. 1, 7.
[3.3.29] Thunhorst K, Goetz DP, Hine AM, et al. The effect of nanosilica
matrix modification on the improvement of the pultrusion process
and mechanical properties of pultruded epoxy carbon fiber
composites. In: Proceedings Composites (2011), Ft. Lauderdale,
FL, U.S.A., 2–4 Feb., 2011
[3.3.31] Liu LQ, Li L, Gao Y, et al. Single carbon fiber fracture embedded in
an epoxy matrix modified by nanoparticles. Compos. Sci.Technol.
(2013), 77, 101.
[3.3.32] Jacob GC, Starbuck JM, Fellers JF, et al. Strain rate effects on the
mechanical properties of polymer composite materials. J. Appl.
Polym. Sci. (2004), 94, 296.
104
[3.3.33] Gao X, Jensen RE, McKnight SH, et al. Effect of colloidal silica on the
strength and energy absorption of glass fiber/epoxy interphases.
Composites: Part A (2011), 42, 1738.
105
3.4. Property improvements of fiber-reinforced composites
based on epoxy resins modified with SiO2 nanoparticles
and elastomers (hybrid systems)
3.4.1. Abstract
3.4.2. Introduction
In many composite applications, epoxy resins are used as matrix materials, and
the diglycidyl ether of bisphenol A (DGEBA) is the most important resin used by
the industry. The diglycidyl ether of bisphenol F (DGEBF) and the higher
molecular weight variations, epoxidized novolac resins, are popular raw materials
as well. For high-performance applications like aerospace parts, multifunctional
epoxy resins are used, mainly the tetraglycidyl ether of methyldianiline (TGMDA).
The combination of epoxy resin and hardener defines the matrix material and the
various possible combinations allow to formulate according to all possible
property requirements [3.4.1], [3.4.2].
106
The variety of epoxy curing agents used in the industry and their network
formation was described by Hare quite a few years ago [3.4.3], [3.4.4]. In low
viscosity infusion resins typically for rotor blade manufacturing, mainly DGEBA is
used as base resin, blended with low viscosity aliphatic epoxy resins like
hexanediole diglycidylether (HDDGE). As hardeners cycloaliphatic amines like
isophorone diamine (IPD) are the material of choice; often in combination with
low molecular weight polyetheramines. Medium performance injection resins are
based on DGEBA with aromatic amines like diaminodiphenylsulfone (DDS) as
curing agents.
Standard prepreg resins systems are often formulated with DGEBA and
dicyandiamide (DICY) as curing agent. In filament winding applications acid
anhydrides like methyl hexahydrophthalic acid anhydride (MHHPA) are used. Due
to new EU regulations, MHHPA will probably be phased out, but new high
performance dianhydrides like 4,4´carbonyl diphthalic acid anhydride (BTDA) are
available and enable to achieve extremely high glass transition temperatures (Tg).
Composite parts manufactured for the automotive industry have to match the
production cycles of the industry of only a few minutes. This limits of course the
manufacturing methods and the epoxy resin/hardener combinations which are
suitable [3.4.6].
However, epoxy resins are inherently brittle. Therefore many epoxy resin
formulations are toughened. Very often, rubber adducts are used. Introduced in
the 1970s, these carboxyl terminated nitrile butadienes (CTBNs) are immiscible
with epoxies but become miscible after a pre-reaction with an excess of epoxy
resin. They form soluble adducts which phase separate again upon cure and form
rubber domains in the epoxy resin. The mechanisms of rubber toughening are
well known [3.4.7], [3.4.8], [3.4.9].
Rubber toughening has several drawbacks like lowering the modulus and
reducing the glass transition temperature. This is caused by some rubber
molecules not participating in the phase separation, but crosslinking randomly
into the matrix. Another issue is the relatively high viscosity of epoxy resins
containing reactive liquid rubbers. Nevertheless they are used in composites
manufacturing.
107
Figure 3.4.1 shows the scanning electron microscopy (SEM) picture of a carbon
fiber-reinforced composite (CFRC) made by RTM, the epoxy resin matrix
toughened with 6.9 wt% of a reactive liquid rubber (CTBN). The spherical rubber
domains, which have formed upon cure, are clearly visible in the resin matrix
and are well dispersed. Thus rubber toughened epoxy resins can be used for
infusion-manufacturing methods.
In the 1990s CSR Type III have been developed [3.4.13], [3.4.14]. They affect
strength, modulus, or glass transition temperature less. This was achieved by
reducing the thickness of the shell which consequently cannot soften anymore.
These epoxy-functional particles with a polysiloxane core have an average
diameter of 500-700 nm.
108
These spherical particles appear transparent, are monodisperse and do not
increase resin viscosity up to quite high addition levels [3.4.15]. They do not
exhibit thixotropic properties like fumed silicas but behave like a Newtonian liquid.
particle density
0 20 40 60 80 100
particle size [nm]
109
This has been subject of intensive research in the last ten years [3.4.16]. The
mechanisms responsible for the property improvements are rather well
understood today.
Due to their size and the unchanged viscosity of the resin, they can be used in
injection or infusion methods for composite manufacturing where they can
penetrate even close meshed fabrics easily.
The combination of reactive liquid rubbers or CSRs and the use of silica
nanoparticles as additional modifier in epoxy resin systems yields additive and
sometimes synergistic property improvements [3.4.17]. It is possible to
formulate tough and stiff materials. These resin systems and their synergies in
various applications, including fiber-reinforced composites were patented
consequently [3.4.18].
Today such hybrid systems are used in many industrial epoxy formulations.
Figure 3.4.3 shows the transmission electron microscopy (TEM) picture of a
cured hybrid epoxy resin. Both the core shell particles Type III (8 vol%) and the
silica nanoparticles (5.5 vol%) are well distributed.
The property improvements found for bulk epoxy resins modified with silica
nanoparticles were not found in fiber-reinforced composites made with such
resins or the improvements were much smaller in scale [3.4.16], [3.4.20].
3.4.3. Discussion
These were then diluted down using commercial epoxy resins to vary the
nanosilica concentrations. All investigations revealed good dispersion of the
nanosilica particles, sometimes small agglomerates; typically determined using
TEM.
Though, quite some research was done with hybrid epoxy resins, only a part of
the researchers used the commercial nanosilica. Furthermore were the results of
several research projects transferred directly into industrial products and not
published at all.
110
3.4.3.1. Glass fiber-reinforced epoxy resin composites
Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, amine cured
Sprenger et al. [3.4.21] prepared their samples from a linen-type fabric made
from 200 g/m2 glass fibers; eight layers of fabric had been placed in a closed
mold and impregnated by using VARTM. Standard DGEBA and a commercial
amine curing agent were used as resin matrix. They reported a decrease in
Youngs modulus of the bulk epoxy resin upon the CTBN modification and an
increase for the nanosilica addition. The hybrid resin exhibited a modulus of the
bulk resin slightly below the control.
An increase in toughness for the cured bulk resin was observed with increasing
addition levels of both rubber and nanosilica. However, the corresponding
laminates showed a different behaviour at 4.1 wt% rubber and 2 wt% nanosilica;
the fracture toughness fracture energy (GIc) of the bulk resin was more than
doubled from 609 J/m2 to 1576 J/m2; an improvement of 159 %, the GIc of the
laminate made using this system did not change much from 1175 J/m2 to 1205
J/m2. At 6.9 wt% rubber and 3.5 wt% nanosilica, the increase for GIc of the bulk
resin was + 174 %; for the laminate + 30 %. The GIc for the laminate showed a
maximum at this level and decreased on further addition of both rubber and
nanosilica, whereas the GIc of the bulk resin was further increased.
Tsai et al. [3.4.24], [3.4.25] modified a DGEBA resin with CTBN and silica
nanoparticles and cured it with a modified isophorone diamine. 12 layers of UD
E-glass were stacked in a modified hand lay-up process to prepare the laminates.
At 10 wt% rubber and 10 wt% nanosilica the GIc of the bulk resin was increased
from 190 J/m2 to 930 J/m2 – this represents an improvement of 390 %. The
laminate based on the unmodified resin had a GIc of 840 J/m2 and the laminate
made from the hybrid resin 1230 J/m2.
111
This equals an improvement of only 48 %. The Youngs modulus of the bulk
hybrid resin was reported to be only 2 % lower than for the control.
A commercial DGEBA resin formulation, modified with a CTBN adduct and silica
nanoparticles, cured with a commercial aliphatic amine blend were investigated
by Schultz et al. [3.4.26]. The hybrid epoxy resin formulation contained 6.5 wt%
CTBN and 8.1 wt% nanosilica. 8-ply of a bidiagonal ±45° stitch bonded glass
fiber fabric were processed by VARTM. The fiber volume fraction of the control
laminate was 0.51 and of the hybrid 0.54. Low-velocity impact tests were
performed by a drop test with different heights and weights corresponding to
different impact energies. For each energy level the hybrid samples showed
visibly more matrix cracking and ply delamination than the laminates made with
the control resin. It was concluded that the hybrid resin system has a better
impact energy dissipation mechanism than the unmodified epoxy resin.
Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, anhydride cured
Kinloch et al. [3.4.28] investigated DGEBA, modified with 9 wt% CTBN and 10
wt% nanosilica. They prepared the laminates by using 16 plies of unidirectional
glass fiber fabric. The fiber volume fraction was determined to be approximately
57 %. The modulus of the laminate made with the hybrid resin was 40.6 GPa, 2
% higher than for the control; the Tg was slightly lower. Fracture toughness tests
were performed with promising results. The GIc of the control panel was 330
J/m2, and the hybrid was improved by 160 % to 860 J/m2. GIIc of the control was
found to be 1300 J/m2 and 1895 J/m2 for the hybrid, again an improvement of
46 %.
In continuation of this work Charpy impact tests have been performed and
showed improvements of approx. 40 % in impact energy [3.4.29]. Ballistic
impact tests with different impact energies showed slightly higher values for the
laminates made with the hybrid epoxy system.
Non-crimp glass fiber fabric arranged in ±45° pattern was laid up in a quasi-
isotropic sequence [(+45/-45,90/0)8]2 and infused with DGEBA epoxy resin
modified with 9 wt% rubber and 10 wt% silica nanoparticles by Manjunatha et al.
[3.4.30]. Fatigue testing showed improvements for laminates for both rubber-
modified and nanosilica-modified epoxy matrices in the same range.
112
The hybrid system with both modifications clearly performed best, as can be
seen in Figure 3.4.4. Regardless of the stress applied during the cyclic testing,
the number of cycles before failure was increased by up to 10 times.
Kinloch, Taylor et al. [3.4.31] continued this study and had a closer look at the
mechanisms of toughening involved. They observed minor agglomerations of the
silica nanoparticles, which apparently had no negative effect on mechanical
properties. The mechanisms of toughening for the bulk resin were debonding of
the epoxy polymer from the silica nanoparticles, followed by plastic void growth
of the epoxy. Localised shear-banding in the polymer was observed as well. The
synergy between rubber and nanosilica (as described by Sprenger [3.4.17] for
bulk epoxy resins) was confirmed. The increase in toughness of the bulk hybrid
resin was transferred into the fiber composite. Typical toughening mechanisms
for fiber reinforced materials like fiber bridging, fiber debonding, and fiber pullout
came into play as well. They extended an existing model to predict laminate
toughness and found good agreement with the experimental data.
113
The modification obviously suppresses matrix cracking and reduces the crack
growth rate. The cyclic fatigue life is 6–10 times higher for the hybrid than for
the control; which is quite a substantial improvement regarding the lifespan of
composite parts used in high stress applications (e.g., rotor blades for wind
energy installations).
Tsai et al. [3.4.24], [3.4.25] modified a DGEBA epoxy resin with CSR Type II
particles with an average particle size of 300–400 nm and silica nanoparticles. A
modified isophorone diamine was used as curing agent. The laminates were
made of 12 layers of UD E-glass. Mode I fracture tests with the laminates from
the unmodified epoxy resin system resulted in a GIc of 830 J/m2. The fiber
reinforced composite made from a resin modified with 10 wt% CSR and 10 wt%
silica nanoparticles showed a GIc of 1510 J/m2; which equals an improvement of
82 %. However, they reported a reduced modulus of the bulk resin by 8.6 %.
When comparing 10 wt% CSR Type II to 10 wt% CTBN, they concluded that the
property profile of the CTBN/nanosilica hybrid is superior. Keeping in mind that
for bulk resins, CSR and nanosilica are not synergistic compared to CTBN and
nanosilica this could be expected for fiber reinforced composites too [3.4.17].
As the core-shell technology is younger than the epoxy resin modification with
reactive liquid rubbers and less established in composites manufacturing, less
research was performed in this field. As an example, no data is published
regarding the combination of CSR Type I and nanosilica in amine cured systems.
Thorough literature search did not yield any data regarding glass fiber-reinforced
composites made from epoxy resins modified with CSRs and industrial 20 nm
silica nanoparticles, which were anhydride cured.
However, the recent increases in using core-shell toughening for fiber reinforced
composites started to push research projects with these materials.
114
3.4.3.2. Carbon fiber-reinforced epoxy resin composites
Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, amine cured
Caccavale et al. [3.4.36], [3.4.37] investigated DGEBA modified with 7.3 wt%
CTBN and 3.7 wt% silica nanoparticles cured with a commercial polyamine
hardener. The tensile testing of the cured bulk resin showed a 10 % reduction in
modulus for the hybrid resin system. 18 layers of T300 NCF carbon fiber were
laid up in [90°/0°]9s and VARTM was used to manufacture the test laminates with
a fiber volume content of 53 %. The fracture toughness (KIc) of the control was
increased from 0.91 MPam0.5 to 2.11 MPam0.5 which equals an improvement of
132 %. Fatigue performance was reported to be increased significantly.
In falling dart tests the laminates were impacted with 30 J and the laminate
based on the hybrid resin system performed best (lowest delamination after
impact). In the pictures given in Figure 3.4.5, it is obvious that the dart nearly
went through the control, whereas the hybrid shows less damage. This is
confirmed by the corresponding C-Scans where there can be seen less
delamination for the hybrid system.
115
Figure 3.4.5: Pictures of CFRC laminates after 30 J impact testing and
corresponding C-Scans [3.4.38]
The toughness of laminate made from the hybrid epoxy resin was 1017 J/m2,
which equals an improvement of 33 %. Dry Tg were identical, however, when
stored at 100 % relative humidity and 70 °C for two weeks and wet Tg is
determined, there are differences. With increasing levels of nanosilica, the wet Tg
is lowered as well as for the CTBN modification. Thus, the hybrid system has the
lowest wet Tg (which is still above 190 °C).
In a project under the Clean Sky program CFRCs made by using hybrid epoxy
resins have been investigated as well. Kowalik et al. [3.4.39] infused HexForce
G0926 D1304 TCT. The GIc of a DDS-cured DGEBA was increased from 724 to
1000 J/m2; an improvement of 38 %. A second system similar to RTM 6 was
increased from 662 to 980 J/m2; which equals an improvement of 48 %. Both
hybrid formulations contained approx. 5 wt% nanosilica and approx. 10 wt%
CTBN.
Microscopical studies revealed well dispersed rubber domains and well dispersed
silica nanoparticles. In some formulations, small agglomerates were found which
were evenly dispersed and seemed not to affect the performance of the
composites. In Figure 3.4.6, the TEM microscopical image of the carbon fiber-
reinforced laminate is given. The rubber domains, formed upon cure are clearly
visible between two carbon fibers. At this magnification, however, the nanosilica
particles cannot be seen.
Figure 3.4.6: TEM picture of a carbon fiber-reinforced laminate made from hybrid
epoxy resin [3.4.40]
116
Figure 3.4.7 shows the same specimen at higher magnification. The uniform
distribution of the nanosilica is confirmed again; though the particle
concentration at the interface rubber particle/epoxy resin seems somewhat
higher. This seems to have no negative effect on performance.
Figure 3.4.7: TEM picture of a carbon fiber-reinforced laminate made from hybrid
epoxy resin [3.4.40]
Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, anhydride cured
Sprenger et al. [3.4.41] stacked six plies carbon fiber fabric (linen type) to
manufacture 3-mm thick laminates by VARTM. The GIc of the laminate made with
the bulk resin was 439 J/m2. The hybrid laminate, modified with 9 wt% CTBN
and 5 wt% silica nanoparticles showed increased toughness with 1080 J/m2. The
system containing 9 wt% CTBN and 8 wt% nanosilica achieved a further increase
in GIc of 1110 J/m2. The hybrid laminate modified with 9 wt% CTBN and 11 wt%
nanosilica performed best; the GIc was found to be 1320 J/m2 which equals an
improvement of 200 % compared to the control. The modulus of the bulk resin
was found to be identical for control and hybrid.
117
Hsieh et al. [3.4.31] continued this investigation. They reported the presence of
small agglomerates which seemed not to affect the performance of the hybrid
resin systems. They looked into the toughening mechanisms and identified two
types of plastic deformation. First, localized shear-bands initiated by the stress
concentrations around the periphery of the silica nanoparticles. Second, the
debonding of the silica nanoparticles followed by subsequent plastic void growth
of the epoxy polymer. Of course the well-known mechanisms of the “classic”
rubber toughening apply to the hybrid systems as well.
18 layers of plain woven carbon fibers (203 g/m2) were used to prepare
laminates by the hand lay-up method by Zeng et al. [3.4.42]. Piperidine-cured
DGEBA was used as resin matrix. The elastomer chosen for the hybrid systems
was a CSR Type I. The addition of various amounts of silica nanoparticles
increased the GIc of the laminate by 20-30 %. The modification with the core
shell particles provided the biggest improvement of 150 % at 10 wt% CSR.
However, bulk resin modulus was lowered by approx. 21 %. The hybrid with 6
wt% silica nanoparticles and 6 wt% core shell elastomer showed an
improvement in GIc of approx. 125 % with a modulus only 6 % lower. Thus a
good balance between toughness and stiffness of the fiber-reinforced composite
could be achieved.
Brandt et al. [3.4.43] reported of a resin system similar to RTM 6, modified with
a core shell elastomer Type III and silica nanoparticles regarding aerospace
applications. Details regarding the fibers or fabrics used were not given. The
toughness of the laminate, already being improved by the presence of the core
shell particles to 155 J/m2 was further increased to 250 J/m2 by the addition of
nanosilica. Delamination after impact with 30 J was reduced significantly. Details
regarding addition levels were not given.
Similarly, as for the glass fiber-reinforced composites, the literature search did
not yield any data regarding CFRCs made from epoxy resins modified with CSRs
and industrial 20 nm silica nanoparticles, which were anhydride cured.
It may be due to the competitive nature of the composites industry, that though
such resin systems are used in industrial products, the research results were not
published.
The reactive liquid rubber domains formed upon cure or the core shell particles
toughen the epoxy by the well-known mechanisms of classical rubber
toughening. The silica nanoparticles increase toughness further by debonding
and subsequent plastic void growth. Localized plastic shear banding contributes
as well. Furthermore they do increase modulus (as a filler) and improve fatigue
performance [3.4.16].
118
The synergy between reactive liquid rubbers and silica nanoparticles as well as
the rather additive performance improvements of hybrids based on core shell
elastomers and nanosilica have been discussed earlier for bulk resins [3.4.17].
Of course, fiber reinforcement can increase toughness as well with the known
mechanisms of fiber bridging, fiber debonding, or fiber pullout. However, when
laminates made from unmodified resins, from resins modified with either
elastomer or nanosilica and from hybrid resin systems are compared, the fiber
(or fabric) related effects should be the same for all samples.
Tsai et al. [3.4.24], [3.4.25] found a synergy for a glass fiber-reinforced laminate
modified with 10 wt% CTBN and 10 wt% silica nanoparticles. The GIc was 1230
J/m2 compared to 1080 J/m2, when the improvements of rubber-only and
nanosilica-only modification are added. A modification with 10 wt% CSR Type II
and 10 wt% nanosilica showed no synergy.
3.4.3.4. The transfer of improved bulk resin properties into the fiber-
reinforced composite
The question of the transfer of the property improvements of the bulk hybrid
resin systems is of great interest. As the fiber properties usually dominate;
especially for high fiber volume fractions which for most composites are between
50 and 70 %, the toughness and the fatigue performance are of bigger interest
than resin properties like modulus.
119
Table 3.4.1 gives an overview of the GIc data from the different sources for the
bulk cured epoxy resins and the laminates made from these resins as well as the
improvement in %.
Ref. Resin / Hybrid GIc bulk GIc bulk Impro- Fiber / GIc composite GIc composite Impro-
Hardener modification resin hybrid vement Fabric from bulk from hybrid vement
(J/m2) (J/m2) (%) resin (J/m2) resin (J/m2) (%)
[3.4 DGEBA / 10 wt% 77 906 1008 Glass 330 860 160
.26] MHHPA nanosilica,
9 wt% CTBN
[3.4 DGEBA / 10 wt% 190 930 390 Glass 840 1230 48
.19] IPD nanosilica,
10 wt% CTBN
[3.4 DGEBA / 2 wt% 609 1576 159 Glass 1175 1205 3
.16] Amine nanosilica
4.1 wt% CTBN
[3.4 DGEBA / 4.1 wt% 609 1344 121 Glass 1175 1349 15
.16] Amine nanosilica,
4.1 wt% CTBN
[3.4 DGEBA / 3.5 wt% 609 1671 174 Glass 1175 1525 30
.16] Amine nanosilica,
6.9 wt% CTBN
[3.4 DGEBA / 10 wt% 190 1030 442 Glass 830 1510 82
.19] IPD nanosilica
10 wt% CSR
Type I
[3.4 DGEBA / 10 wt % 77 906 1008 Carbon 439 1316 200
.26] MHHPA nanosilica,
9 wt% CTBN
[3.4 TGMDA / 10 wt% 176 708 302 Carbon 766 1017 33
.33] aromat. nanosilica
Amines 8 wt% CTBN
[3.4 DGEBA / 4.7 wt% 103 917 790 Carbon 439 1078 146
.26] MHHPA nanosilica,
[3.4 9 wt% CTBN
.36]
[3.4 DGEBA / 7.2 wt% 103 973 845 Carbon 439 1106 152
.26] MHHPA nanosilica,
[3.4 9 wt% CTBN
.36]
[3.4 DGEBA / 11 wt% 103 1250 1114 Carbon 439 1320 200
.26] MHHPA nanosilica,
[3.4 9 wt% CTBN
.36]
Table 3.4.1: GIc data for cured bulk epoxy resins and laminates made thereof
In Figure 3.4.8 the GIc values of the cured bulk resin systems versus the
laminates are given. There seems to be a linear relationship at low-fracture
toughness up to approx. 500 J/m2. Above that value, the relationship is no more
linear and exhibits a much smaller gradient.
Hunston et al. [3.4.44] reported a very similar behaviour for the fracture
toughness of CFRCs. They distinguished an area of low GIc below 200 J/m2 where
a toughness increase of the cured bulk resin was transferred into the laminate.
At higher GIc values, the bulk fracture behaviour was not fully transferred into
the fiber-reinforced composite, the data points more scattered (eventually no
linear relationship anymore).
120
Flemming et al. [3.4.2] showed for a system based on T-300 carbon fibers a
linear increase at first, and then a slower increase of the GIc of the laminates
versus the bulk resins – very similar to Figure 3.4.8.
Jordan and Bradley [3.4.45] investigated CFRCs made from rubber toughened
epoxy resins. They observed a sharp drop in fracture toughness improvements
for the laminates above 700 J/m2 – again very similar to Figure 3.4.8. They
concluded that delamination fracture is often dominated by interfacial failure and
that more ductile resins will not necessarily enhance delamination fracture
toughness. This statement can be confirmed for the subject of this review,
toughening by the combination of elastomers and silica nanoparticles as well.
In Figure 3.4.9, the relative improvement of the GIc of the cured bulk resins
versus the relative improvement of the GIc of the laminates when using a hybrid
resin instead of the unmodified control is shown. The nature of the epoxy resin
and curing agent as well as the nature of the reinforcement are not taken into
account. A linear slope can be drawn for the data points. As the line does not go
through the origin, this means a threshold of minimal improvement of fracture
toughness exists. In other words, small improvements of the GIc of the cured
bulk resin like 30 % will not yield an improvement in GIc of the laminates.
The gradient could be defined as transfer factor and would be 0.18 for the data
used in Figure 3.4.9. (accuracy range ± 0.02).
121
This means that an improvement in GIc of 100 % for the cured bulk resin
achieved by a resin modification improves the fracture toughness of the laminate
by only 18 %.
Figure 3.4.9: Relative improvements of GIc of cured bulk resins versus laminates
If, example given, an improvement of the laminate GIc by 100 % is desired, then
the composite manufacturer will have to replace his resin system by a hybrid
which has a 555 % higher GIc. By simply measuring the GIc of the cured bulk
resin a suitable hybrid can be chosen without the necessity of making many
tedious trials with laminates from various resins.
Of course, further research will be necessary to gather more data to support and
enhance this theory. Eventually, it makes more sense to differentiate between
glass fiber reinforced and CFRCs. It is very likely that they will have different
transfer factors. Eventually, different epoxy resins used as matrix could be
distinguished as well.
122
Especially for fiber-reinforced composites modified with core shell particles and
nanosilica significantly more research is necessary – so far we have only one
data point for those. Though a similar linear relationship can be expected; it
could be well a different gradient than for hybrid resins with CTBN.
3.4.4. Conclusions
1. Bulk resin modulus of cured hybrid resins was found either unchanged or
up to 6 % lower as for the control at 6 wt% CSR Type I and 6 %
nanosilica. The CSR Type I seems to affect modulus, a resin containing 10
wt% rubber exhibits a 21 % lower modulus of the cured bulk system.
2. The modulus of both glass fiber reinforced and CFRCs made with resins
containing both elastomer particles and silica nanoparticles was reported
to be identical for UD and linen type reinforcements. A quasi-isotropic fiber
layup was reported to have a 10 % lower modulus [3.4.32].
3. The significant improvements in fracture toughness found for bulk epoxy
resins modified with both elastomers and silica nanoparticles were found
for the fiber-reinforced systems on a much smaller scale.
4. A nearly linear relationship exists between the relative GIc increase of the
cured bulk resin systems containing both CTBN and silica nanoparticles
and the laminates made thereof. The fiber reinforcement, be it glass or
carbon fiber, regardless of the three-dimensional orientation of the fibers,
seems to have little or no influence on this relationship.
5. Remarkable improvements of fatigue properties of fiber-reinforced
composites made from hybrid resins were found – up to 10 times of cyclic
loadings before laminate failure.
3.4.5. References
[3.4.3] Hare C.J. Amine curing agents for epoxies. J. Prot. Coat. Lin.
(1994), 9, 77 – 103.
[3.4.4] Hare C.J. Epoxy curing agents II. J. Prot. Coat. Lin. (1994),10, 197 -
213.
123
[3.4.7] Kinloch A.J., Shaw S.J., Tod D.A. Deformation and fracture behavior
of a rubber-toughened epoxy: 1. Microstructure and fracture
studies. Polymer (1983), 24, 1341 – 1354.
[3.4.10] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. Rubber-
toughened FRCs optimised by nano-particles IV. JEC Compos. Mag.
(2008), 38, 34 – 37.
[3.4.13] Lai M., Friedrich K., Botsis J., Burkhart T. Evaluation of residual
strains in epoxy with different nano/micro-fillers using embedded
fiber Bragg grating sensor. Composites Sci. Technol. (2010), 70,
2168 – 2175.
[3.4.15] Sprenger S., Eger C., Kinloch A.J., Taylor A.C. Nanotoughening of
Epoxies. Proceedings of Stick! conference, Vincentz Verlag,
Nürnberg, Germany, 9 April (2003).
124
[3.4.21] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. Rubber-
toughened GFRCs optimised by nanoparticles. JEC Compos.
Mag. (2005), 21, 66 – 69.
[3.4.24] Tsai J.L., Huang B.H., Cheng Y.L. Enhancing fracture toughness of
glass/epoxy composites by using rubber particles together with silica
nanoparticles. J. Compos. Mater. (2009), 43 (25), 3107 – 3123.
[3.4.25] Tsai J.L., Huang B.H., Cheng Y.L. Enhancing fracture toughness of
glass/epoxy composites for wind blades using silica nanoparticles
and rubber particles. Procedia Eng. (2011), 14, 1982 – 1987.
[3.4.26] Schultz R., Tate J.S., Gaikwad S., Trevino E., Jacobs C., Sprenger S.
Low velocity impact on epoxy glass composites modified with rubber
microparticles and silica nanoparticles. In: Proceedings of the SAMPE
TECH conference, Fort Worth, TX, USA SAMPE Covina, CA, U.S.A.
18-19 October (2011).
[3.4.27] Tate J., Trevino E., Gaikward S., Sprenger S., Rosas I., Andrews
M.J. Low velocity impact on epoxy glass composites modified with
rubber microparticles and silica nanoparticles. J. Nanosci., Nanoeng.
Appl. (2013), 3 (1), ISSN: 2231-1777.
[3.4.28] Kinloch A.J., Masania K., Taylor A.C., Agarwal R., Sprenger S., Egan
D. The fracture of glass-fibre-reinforced epoxy composites using
nanoparticle-modified matrices. J. Mater. Sci. Lett. (2008), 43, 1151
– 1154.
[3.4.29] Kinloch A.J., Masania K., Taylor A.C., Sprenger S. The fracture of
nanosilica and rubber toughened epoxy fibre composites. In:
Proceedings of the Composites & Polycon, American Composites
Manufacturers Association, Tampa, Florida, USA, 15-17 January
(2009).
[3.4.30] Manjunatha C.M., Taylor A.C., Kinloch A.J., Sprenger S. The effect of
rubber micro-particles and silica nano-particles on the tensile fatigue
behaviour of a glass-fibre epoxy composite. J. Mater. Sci. Lett.
(2009), 44, 342 – 345.
[3.4.31] Hsieh T.H., Kinloch A.J., Masania K., Sohn Lee J., Taylor A.C.,
Sprenger S. The toughness of epoxy polymers and fibre composites
modified with rubber microparticles and silica nanoparticles. J.
Mater. Sci. (2010), 45, 1193 – 1210.
125
[3.4.32] Manjunatha C.M., Taylor A.C., Kinloch A.J., Sprenger S. The tensile
fatigue behavior of a glass-fiber reinforced plastic composite using a
hybrid-toughened epoxy matrix. J. Compos. Mater. (2010), 44 (17),
2095 – 2109.
[3.4.34] Manjunatha C.M., Bojja R., Jagannathan N., Kinloch A.J., Taylor A.C.
Enhanced fatigue behavior of a glass fiber reinforced hybrid particles
modified epoxy nanocomposite under WISPERX spectrum load
sequence. Int. J. Fatigue (2013), 54, 25 – 31.
[3.4.38] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. (2007)
Rubber-toughened CFRCs optimized by nanoparticles – Part III. JEC
Compos. Mag. 2007, 30, 54 – 57.
[3.4.41] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D., Eger C.
Rubber Rubber-toughened FRCs optimised by nanoparticles. JEC
Compos. Mag. (2005), 19, 73 – 76.
[3.4.42] Zeng Y., Liu H.Y., Mai Y.W., Du, X.S. Improving interlaminar
fracture toughness of carbon fibre/epoxy laminates by incorporation
of nanoparticles. Compos. B (2012), 43, 90 – 94.
[3.4.43] Brandt J., Drechsler K., Schmidtke K. Composites für den Flugzeug-
bau. Kunststoffe (2004), 10, 290 – 294.
126
[3.4.44] Hunston D.L., Moulton R.J., Johnston N.J., Bascom W.D. Tough
composites, STP 937, ASTM, Philadelphia, U.S.A. (1987), pp 74 –
94.
127
3.5. Carbon fiber-reinforced composites with epoxy resin
modified with reactive liquid rubber and SiO2
nanoparticles
3.5.1. Abstract
Carbon fiber-reinforced composites are gaining importance and are about to play
a very prominent role in automotive applications. In structural applications
mainly epoxy resins are used as matrix materials. The properties of epoxy resins
and laminates made thereof can be improved by tougheners like reactive liquid
rubbers. Further improvements can be achieved by adding surface-modified silica
nanoparticles, with 20 nm in size and a very narrow particle size distribution.
In this study carbon fiber-reinforced laminates made from epoxy resins modified
with reactive liquid rubber and silica nanoparticles have been prepared and
investigated. A very fast amine cure of 15 minutes has been chosen to match
industrial needs. Mechanical properties for bulk resins and laminates are
compared and the mechanisms responsible for the property improvements are
discussed. Structure property relationships between the neat resin fracture
toughness and the interlaminar GIc and GIIc of the reactive liquid rubber and silica
nanoparticles modified resins were established. Tough laminates could be
prepared.
3.5.2. Introduction
Lightweight construction becomes the key technology to solve these issues which
currently drives a trend towards the increased use of fiber-reinforced composites
in automotive construction [3.5.2].
128
Besides unsaturated polyester resins mainly epoxy resins are used as
thermosetting matrices for automotive components. The diglycidylether of
bisphenol A (DGEBA) is the most commonly used epoxy resin in automotive
industry. The diglycidylether of bisphenol F (DGEBF) and the higher molecular
weight variations, epoxidized novolac resins, are only used for special
applications. So-called reactive diluents, short chain aliphatic epoxy resins like
hexanediole diglycidylether (HDDGE), are very often blended with the DGEBA in
order to reduce the viscosity. A low viscosity of the resin is very important as the
preferred manufacturing methods for automotive parts currently are injection or
infusion processes – which enable short cycle times [3.5.3].
The combination of epoxy resin and hardener defines the matrix material and the
various possible combinations enable to formulate according to all possible
property requirements [3.5.4]. Hare described the variety of epoxy curing agents
a few years ago [3.5.5], [3.5.6]. Most popular for automotive part manufacturing
are aliphatic or cycloaliphatic amines like isophorone diamine (IPD) due to their
low viscosity and fast cure at elevated temperatures. Various accelerators and
additives are typically part of the hardener formulation.
However, epoxy resins are inherently brittle. Therefore many epoxy resin
formulations are toughened. In the seventies of last century carboxy-terminated
nitrile butadienes (CTBNs) were introduced as tougheners. They are immiscible
with epoxies but become miscible after a pre-reaction with an excess of epoxy
resin. These soluble adducts phase separate again upon cure and form rubber
domains in the epoxy resin. The mechanisms of rubber toughening are well
known [3.5.7], [3.5.8]. Rubber toughening has several drawbacks like lowering
the modulus and reducing the glass transition temperature, due to some rubber
molecules not participating in the phase separation, but crosslinking randomly
into the matrix. Another issue is the relatively high viscosity of epoxy resins
containing reactive liquid rubbers, which can be critical for infusion processes.
Nevertheless they are used in composites manufacturing.
These spherical particles are transparent, monodisperse and just slightly increase
the resin viscosity up to quite high addition levels. Resins containing such
nanoparticles exhibit no thixotropic properties but behave like Newtonian liquids.
Figure 3.5.1 shows the uniform particle distribution in a cured bulk epoxy resin
by transmission electron microscopy (TEM).
129
Figure 3.5.1: Transmission electron microscopy (TEM) image of amine-cured
DGEBA with 12.2 vol% silica nanoparticles at different magnifications [3.5.10]
Due to their size and the only slight increase in viscosity of the resin they can be
used in injection or infusion methods for composite manufacturing where they
can penetrate even close meshed fabrics easily. It is well known that properties
like modulus, fracture toughness, compressive strength and especially fatigue
performance upon cyclic loading can be improved by modifying the epoxy resin
matrix with silica nanoparticles. Some properties, e.g. high speed impact
performance, were not improved. An extensive review regarding the potential of
silica nanoparticles to toughen fiber-reinforced composites was published
recently [3.5.13].
The combination of reactive liquid rubbers or core shell elastomers and the use
of silica nanoparticles as additional modifier in epoxy resin systems yields
synergistic property improvements in cured bulk epoxy resins [3.5.14]. It is
possible to formulate tough and stiff materials. These resin systems and their
synergies in various applications, including fiber-reinforced composites were
patented consequently at a very early stage [3.5.15].
Today such hybrid systems are used in many industrial epoxy formulations.
Figure 3.5.2 shows the transmission electron microscopy (TEM) picture of a
cured bulk hybrid epoxy resin. Obviously both the rubber particles which have
formed upon cure and the silica nanoparticles are well dispersed.
130
Figure 3.5.2: Transmission electron microscopy (TEM) image of amine-cured
hybrid epoxy resin [3.5.16]
The toughening mechanisms of such hybrid systems are well understood [3.5.14]
and described recently in a review article [3.5.17]. In summary, significant
property improvements reported for hybrid neat resins are transferred only
partially into improvements in the corresponding composites.
The results published so far have been obtained with resin systems cured for
very long times. Amine cured systems with cure cycles of 4-6 hours; sometimes
followed by a postcure.
Anhydride cured systems cured for 3-12 hours at temperatures above 100 °C.
Thus a hardener system with a relatively fast cure of 15 minutes, more realistic
for industrial applications, was selected.
131
3.5.3. Experimental
3.5.3.1. Materials
132
Carbon fiber Tenax-E HTS40 F13 12 K was arranged into a -45°/45° nonwoven
with 254 g/m2. This material was kindly provided by Saertex GmbH, Saerbeck,
Germany.
The resin systems were blended according to the compositions given in Table
3.5.1 and cured for 15 minutes at 120 °C followed by 2 hours postcuring at 160
°C. The aluminum mold was coated with a release agent (Loctite Frekote 770-
NC, Henkel, Germany). Plates with a defined thickness of 4 mm respectively 2
mm were obtained.
Laminates for compression after impact (CAI) testing were produced with a
[(90/0)/(-45/+45)]4s layup and 54 vol% fibers at a laminate thickness of 4.2
mm. The closed aluminum mold was put into a hot press (LZT 110 L,
Maschinenfabrik Langzauner GmbH, Austria) and evacuated (10 mbar). An
injection temperature of 70 °C was chosen for the resin systems in order to have
a reasonably fast filling of the mould. The injection pressure was set to 8 bars.
After 15 minutes curing at 120 °C the laminates were demoulded and postcured
in an oven for 2 hours at 160 °C.
The viscosities of the uncured resin systems were determined using a rotational
rheometer (Physica MCR 301, Anton Paar GmbH, Austria) equipped with a
disposable aluminium plate tool and an aluminium bowl. Testing parameters
were rotational mode with a shear rate of 10 s-1 at 70 °C and a gap size of 1
mm.
All test specimens were prepared using a circular saw (Diadisc 6200, MUTRONIC
Präzisionsgerätebau GmbH & Co. KG, Germany) equipped with a diamond saw
blade.
Glass transition temperature of the cured bulk resin was determined by dynamic
mechanical thermal analysis (DMTA) according to DIN EN ISO 6721-7 using an
Advanced Rheometric Expansion System (ARES I, Rheometric Scientific,
Germany) in torsional mode. A sinusoidal deformation of 0.1 % with a frequency
of 1 Hz was applied from 25 °C to 200 °C with a heating rate of 3 K/min. The
glass transition temperature (Tg) was determined as maximum of the loss factor,
tan δ.
133
The critical stress intensity factor (KIc) and the critical energy release rate (GIc)
were determined according to ISO 13586 using compact tension (CT) test
specimens. The test specimen width was w = 33 mm, the thickness d = 4 mm.
For each sample, a sharp crack was generated in the V-notch by tapping a new
razor blade. The tests were performed using a universal tensile testing machine
(Zwick BZ 2.5/TN1S, Zwick Roell, Germany) with a test speed of 10 mm/min.
The crack opening displacement was measured using a clip extensometer
(632.29F-30, MTS, Germany). The modulus of the bulk resin samples was
calculated according to the method of Saxena and Hudak [3.5.18].
The interlaminar shear strength (ILSS) was measured according to DIN EN 2563
using test specimen with the dimension 25x10x3 mm prepared from the
laminates made for the GIc testing. The tests were performed on a universal
tensile testing machine (Zwick 1475, Zwick Roell, Germany) with a test speed of
1 mm/min.
Compression after impact (CAI) data was gathered in accordance to DIN 65561
using 3 samples with dimensions of 150x100 mm. The specimens were impacted
with an energy of 30 J using a drop-weight impactor. Subsequently the
delaminated area was measured using an ultrasonic inspection method (AirTech
HFUS 2400, Ingenieurbüro Dr. Hilger Ultraschall-Prüftechnik, Germany). The
compression after impact strength was measured on a universal tensile testing
machine (Zwick 1485, Zwick Roell, Germany) with a test speed of 0.5 mm/min.
Scanning electron microscopy (SEM) analysis was carried out using a Zeiss 1530
(Carl Zeiss AG, Germany) posessing a field emission cathode for high-resolution
micrographs (acceleration voltage 1.5 kV).
134
3.5.4. Results and discussion
All other studies cited used the same commercially available epoxy resins with 40
wt% silica nanoparticles with a diameter of 20 nm. The reactive liquid rubber
used was always introduced into the formulations as epoxy resin adduct;
typically 40 wt% CTBN rubber with an acrylonitrile content of 18 wt% had been
prereacted with 60 wt% DGEBA.
First the viscosities of the formulations given in Table 3.5.1 were measured. At
70 °C they ranged from 200 mPas for the control to 8.000 mPas for 10R10N. It
has to be noted that processing resin systems with viscosities above 1.000 mPas
in an RTM setup is anything but trivial.
In Figures 3.5.3 and 3.5.4 the Tg´s and the moduli for the cured bulk resins are
given. As expected, the modification with 10 wt% reactive liquid rubber (system
10R) lowered both modulus and Tg by approx. 10 %. When epoxy resins
containing reactive liquid rubbers are cured, the rubber phase separates and
forms a secondary phase, the rubber domains.
Figure 3.5.3: Glass transition temperature (Tg) for cured bulk epoxy resin
systems
It is well known that some of the long rubber molecules do not participate in the
phase separation upon cure but crosslink randomly into the matrix. This lowers
the network density, which affects of course modulus and Tg.
135
Figure 3.5.4: E-modulus of the cured bulk epoxy resin systems
The addition of silica nanoparticles as a filler reinforces the matrix and thus
increases the modulus again. The epoxy resin system with 10 wt% rubber and
10 wt% nanosilica has the same modulus as the unmodified control. An increase
in modulus of 0.35 GPa for 10 wt% silica nanoparticle addition is similar to what
was found for other amine-cured DGEBA systems [3.5.11].
136
Figure 3.5.5: KIc of the cured bulk epoxy resin systems
137
As can be seen in Figures 3.5.5 and 3.5.6, the increase in toughness by the
rubber addition is significant; the further addition of silica nanoparticles does not
contribute to toughening markedly. At 10 wt% nanosilica however, the
toughness is further increased – beyond the increase deriving from the rubber
modification. An improvement in GIc of approx. 250 % is not untypical. In a
recently published review improvements in the range of 111-302 % were
reported for amine-cured, CTBN-modified DGEBA resins [3.5.14].
Figure 3.5.7 shows that the hybrid system with 10R10N is a tough and stiff
matrix resin, clearly outperforming the unmodified control. The Tg with 140 °C is
still acceptable for structural automotive applications. In Table 3.5.2 an overview
of all bulk resin properties is given.
Figure 3.5.7: Modulus versus GIc for cured bulk epoxy resin systems
138
Table 3.5.2: Overview of cured bulk epoxy resin properties
It is well known that very often property improvements of matrix resins do not
translate into improvements of fiber-reinforced composites made thereof.
Especially carbon fibers can be a very dominating reinforcement. The recently
published review from Ye et al. gives an excellent overview regarding the
improvements of interlaminar fracture toughness and compression after impact
strength by using nanoparticles in epoxy resin matrices for fiber-reinforced
composites [3.5.20]. They distinguished two different relationships between GIc
of the matrix and GIc of the laminate - depending if the matrix is rather brittle or
tough. Furthermore they concluded that a positive correlation between the GIc of
the matrix and the GIIc of the composite or the CAI strength is not clear.
Therefore they voiced the opinion that the research “… is at an immature
stage…”. Adding an additional component, elastomeric tougheners like reactive
liquid rubbers or core-shell elastomers makes the understanding of such systems
even more complex.
The first laminate properties investigated in this study were GIc and GIIc. The
results are given in Figures 3.5.8 and 3.5.9.
139
Figure 3.5.8: GIc of carbon fiber-reinforced laminates based on various epoxy
resins systems
As can be seen in Figure 3.5.8, despite the higher modulus of the hybrid matrix
systems, the laminate toughness was improved by approx. 50 % for the hybrids
and even more by the rubber-only modified system.
For carbon fiber-reinforced materials much less literature data are available than
for glass fiber-reinforced composites made from hybrid epoxy resins containing
both the reactive liquid rubber and the silica nanoparticles [3.5.17]. Nevertheless
in most studies with CFRC published until now, the fiber-reinforced composites
based on a hybrid system were found to have the highest GIc. Kinloch et al.
reported about a CFRC made from anhydride cured DGEBA [3.5.21], [3.5.22].
The control with 439 J/m2 was increased to 1050 J/m2 by a modification with 9
wt% reactive liquid rubber. Further addition of 10.5 wt% silica nanoparticles
increased the GIc of the laminate further to 1320 J/m2. Ye et al. investigated
piperidine-cured DGEBA [3.5.23]. The GIc could be increased from 741 J/m2 to
1256 J/m2 by a modification with 10 wt% rubber. The addition of 10 wt%
nanosilica further increased the GIc to 1323 J/m2.
However, other findings have been published as well: Sprenger et al. reported
about a CFRC based on tetrafunctional epoxy resin cured with aromatic amines;
a system similar to the aerospace industry benchmark Hexflow® RTM6 [3.5.26].
Though the cured hybrid bulk system had the highest GIc, the hybrid laminate
was lower in GIc than the system containing only reactive liquid rubber.
140
Nevertheless the modulus of the hybrid was higher than for the rubber-only
modification.
Figure 3.5.9 compares the improvement of the bulk GIc with the improvement of
the laminate GIc (in %). The average of laminate improvement versus bulk
improvement was 0.2 ± 0.07 - which is very close to the conversion factor of
0.18 ±0.02 recently proposed by Sprenger [3.5.17].
141
Figure 3.5.10: GIIc of carbon fiber-reinforced laminates based on various epoxy
resins systems
Not much GIIc data has been published for hybrid systems so far. The data shown
in Figure 3.5.10 indicates that once a crack has formed, the delamination under
mode II loading continues at a sligthly lower force. How this is related to the
hybrid modification is unclear. Further clarification of the fracture mechanisms is
expected from the scanning electron micrographs of the fracture surfaces of GIc
and GIIc test specimen.
The ILSS was tested and the results are given in Figure 3.5.11. As can be seen,
it is slightly lower for all modified systems, but still on a very high overall level.
Having a look at the error bars, there cannot be seen real differences between
the rubber-only modification and the hybrids with different nanosilica
concentrations. The very small differences in ILSS for all samples tested indicate
an excellent laminate quality.
Quaresimin et al. studied CFRC made from amine-cured DGEBA [3.5.25]. The
modification with 7.3 wt% reactive liquid rubber and 3.7 wt% silica nanoparticles
increased the bulk resin KIc from 0.91 MPam1/2 to 2.11 MPam1/2 or by 130 %. The
mode I interlaminar fracture toughness was increased by 74 %. However the
interlaminar shear strength was reduced from 55 MPa to 50 MPa - very similar to
our results.
142
Figure 3.5.11: Interlaminar shear strength of carbon fiber-reinforced laminates
based on various epoxy resins systems
Falling dart test and the CAI strength were investigated next. The results are
given in Figures 3.5.12 and 3.5.13. As can be seen, for higher nanosilica
contents the hybrid systems show more delamination and the compressive
strength is reduced. This is unexpected, as for a laminate made from a tougher
epoxy resin matrix one would expect an increase in CAI.
Kuehn et al. investigated carbon fiber reinforced laminates made from anhydride
cured DGEBA [3.5.26]. A modification with 10 wt% reactive liquid rubber reduced
the damage area (at 30 J) from 38 to 24 cm2 – the compressive strength after
impact was increased from 165 MPa to 205 MPa. A modification with 25 wt%
nanosilica did not reduce the delaminated area at all but increased the CAI to
175 MPa. Obviously the higher modulus of the systems modified with nanosilica
is improving the residual compressive strength. Unfortunately they did not
investigate hybrid systems.
143
They noticed that small agglomerates had formed which seemed not to affect the
performance of the nanosilica modification.
Impact 30 J
Impact 30 J
144
The comparision of the delaminated area between the rubber-modified system
10R and the hybrid 10R5N which contains additionally 5 % of nanosilica is very
interesting: they exhibit the same delaminated area, but the hybrid shows a
lower compression after impact strength.
GIc and GIIc of this hybrid are lower as well. This indicates a more brittle matrix
behaviour, which is confirmed by the bulk resin results for 10R and 10R5N.
Impact 30 J
145
Table 3.5.3: Overview of carbon fiber-reinforced laminate properties
In a future paper the fatigue properties will be studied, as they have been
reported best for hybrid systems [3.5.17].
Figure 3.5.15 reveals some information regarding the fracture mechanism. The
fracture surface of the unmodified control shows less residual epoxy resin than
the rubber-modified epoxy and the hybrids, which indicates a more interfacial
failure between fiber layer and matrix. A similar effect was found by Ye et al.
[3.5.23].
The difference becomes more evident on the fracture surfaces from GIIc test
specimen at a higher magnification in Figure 3.5.16. The matrix resin between
the carbon fibers of the control shows the typical angular structures of brittle
materials; designated "hackles".
146
Figure 3.5.15: SEM images of fracture surfaces from GIc test specimens. The
white arrows indicate the direction of crack propagation.
Figure 3.5.16: SEM images of fracture surfaces from GIIc test specimens. The
white arrows indicate the direction of crack propagation.
147
In the picture of the matrix containing 10 wt% reactive liquid rubber the rubber
domains, which have formed upon cure, are clearly visible. The matrix shows no
angular structures any more - it looks rather peeled or teared off. This indicates
a more ductile behavior. The hybrid matrix with 10 wt% rubber and 5 wt%
nanosilica looks similar; however at 10 wt% nanosilica the matrix does not look
ductile anymore. Apparently the nanosilica addition reduces the ductility gained
by the rubber modification. This would explain the increase in delaminated area
and subsequent decrease in compressive strength after impact with increasing
silica nanoparticle content.
Figure 3.5.17: SEM image of hybrid 10R5N (fracture surface of GIIc test
specimen).
The white arrow indicates the direction of crack propagation. A major aspect
comes into focus when looking closer. In Figure 3.5.17 the matrix resin of the
hybrid 10R5N with 10 wt% rubber and 5 wt% nanosilica can be seen between
the surfaces of the carbon fibers. The rubber domains with their typical size of
0.5 - 0.7 micron are clearly visible, and, unexpected, agglomerates of silica
nanoparticles are found. This triggered an additional microscopical investigation
of the cured bulk resins in order to understand these systems.
Despite the rather fast cure the rubber domains have formed perfectly and are
evenly distributed, as can be seen in Figure 3.5.18 for the system 10R. All
hybrids show agglomerates of silica nanoparticles. They increase in size with
increasing addition level in nanosilica. At 5 wt% the agglomerates seem to be
approximately 100 nm large; at 7.5 wt% 200 - 300 nm and at 10 wt% they
seem to exceed 300 nm in size.
148
Eventually the agglomerates have formed due to an incompatibility of the
surface-modified silica nanoparticles with the hardener system. Another
possibility would be the very fast cure. This will be discussed in the next chapter.
Figure 3.5.18: TEM images of cross sections of cured bulk epoxy resins systems
Hsieh et al. prepared laminates with carbon fibers and glass fibers using different
hybrid systems [3.5.29]. Scanning electron microscopy of the anhydride-cured
bulk DGEBA revealed the formation of agglomerates as well. The nanoparticles
appeared to cluster in necklace-like structures and increased in size with
increasing nanosilica content, however the GIc of the bulk resin and the laminates
was increased significantly – and was always higher for the hybrid than for the
rubber-only modified epoxy resin.
149
For the 10R10N system they reported agglomerate sizes of up to 2000 nm in
length and 100 nm in width. Therefore it was concluded that the formation of
agglomerates does not affect performance.
We started to examine the different components of our epoxy resin system and
found no incompatibility between silica nanoparticles and the curing system
consisting of IPD and TMD. If our system is cured without the presence of
reactive liquid rubber but with the same hardener/accelerator and the same fast
cure, no agglomeration occurs - as can be seen in Figure 3.5.19.
Figure 3.5.19: TEM image of cross section of cured epoxy resin with 5 wt%
nanosilica (0R5N) at different magnifications
Next we investigated the blending process with reactive liquid rubber adducts.
When epoxy resins containing reactive liquid rubber adducts and epoxy resins
containing silica nanoparticles are blended, these blends are transparent and no
agglomeration takes place, even if such a resin blend is stored over weeks and
months. Now being mixed with the hardener and accelerator at room
temperature, the resin system remains transparent - still no agglomeration
occurs. We froze the system 10R5N immediately after blending with the hardener
using liquid nitrogen and examined using TEM under cryogenic conditions (Figure
3.5.20).
Despite the presence of hardener and accelerator and their intensive mixing with
the resin part no agglomeration has occurred – the silica nanoparticles are well
dispersed. As the system is uncured, the rubber phase separation has not
occurred and no domains have formed.
150
Figure 3.5.20: Cryo-TEM image of uncured epoxy resin system with 10 wt %
CTBN and 5 wt% nanosilica (10R5N)
Eventually the relatively fast cure cycle of 15 minutes, inducing a rather quick
rubber phase separation of the CTBN, is the cause and forces the nanoparticles
to agglomerate. This would explain why other researchers using unaccelerated
isophoronediamine curing did not report the formation of agglomerates [3.5.14].
In order to confirm this assumption the hybrid system 10R5N was cured at room
temperature for 48 hours. The microscopical investigation was evident, as can be
seen in in Figure 3.5.21.
Though the room temperature cure was almost completed after 2 hours we
prepared the microscopy test specimen after 48 hours of RT cure. As can be
seen, even during this somewhat slower cure agglomerates were formed. They
seem to be slightly smaller, but this is only a subjective impression.
151
Figure 3.5.21: TEM image of cross section of RT-cured epoxy resin with 10 wt%
CTBN and 5 wt% nanosilica (10R5N) cured 48 hours at RT
Curing 10R5N without the accelerator TMD did not impede nanosilica
agglomeration either.
In the literature Pearson et al. reported about bulk hybrid systems containing
large agglomerates as well [3.5.30]. They claimed that the formation of
agglomerates occurs for higher concentrations of reactive liquid rubber; probably
above 12 vol%. The improvements in fracture toughness by nanosilica addition
became smaller for systems with agglomerates present. This would explain our
findings for bulk GIc.
152
The question if the performance of our laminates would even be better with no
agglomerates present could not be answered as the rubber-phase separation of
our system, probably due to the relatively fast curing system, leads to the
formulation of agglomerates. Eventually another hardener system might behave
different. However, this will be the subject of future work.
Another approach for future investigations will be to use core shell rubber
particles instead of reactive liquid rubber; therefore no phase separation during
cure will occur. Thus, theoretically no agglomerates should be formed during
cure.
Using a fast curing epoxy resin system based on DGEBA/IPD/TMD carbon fiber-
reinforced laminates suitable for automotive applications were made in a
reasonable short manufacturing cycle. The modification of the matrix with both
reactive liquid rubber and silica nanoparticles resulted in tough laminates.
1. The bulk resin modified with the reactive liquid rubber shows the expected
increase in toughness alongside the loss in modulus and Tg.
2. The hybrid bulk resins modified additionally with silica nanoparticles show
a reduced loss of the modulus.
3. The hybrid bulk resin with the highest addition level of nanosilica performs
best; exhibiting the highest toughness and modulus.
4. The laminate made from the rubber-only resin shows an increase in GIc, a
slight reduction in GIIc and ILSS as well as a reduction of the delaminated
area in impact testing alongside with an increase in CAI.
5. The hybrid laminates showed no further increase in toughness compared
to the rubber-toughened system, but rather a slight reduction of both GIc
and GIIc. ILSS was on a similar level; CAI was reduced further.
6. The reason for this unexpected lack of further increase in toughness could
be the formation of agglomerates of nanosilica during the cure.
7. The agglomeration found is caused by the presence of the reactive liquid
rubber. Very probably the fast phase separation of the rubber, pushed by
the fast cure, is the driving force of the agglomeration.
8. The conversion factor for the relative improvement in GIc versus the
relative improvement of GIc of the laminate was found to be 0.2 - in
excellent agreement with earlier findings.
153
3.5.6. References
[3.5.2] Estin & Co. Main dynamics of the composite industry for automotive
applications 2010 – 2015. Publ. by JEC Composites (2011); Paris;
France
[3.5.3] Hillermeier R., Hasson T., Friedrich L., Ball C. Advanced thermo-
setting resin matrix technology for next generation high volume
manufacture of automotive composite structures. SAE Technical
Paper (2013)-01-1176.; 2013; doi:10.4271/2013-01-1176
[3.5.5] Hare, C. Amine curing agents for epoxies. J. Prot. Coat. Lin. (1994),
9, 77 – 103.
[3.5.6] Hare, C. Epoxy curing agents II. J. Prot. Coat. Lin.(1994), 10, 197 –
213.
[3.5.7] Kinloch, A.J., Shaw, S.J., Tod, D.A. Deformation and fracture
behavior of a rubber-toughened epoxy: 1. Microstructure and
fracture studies. Polymer (1983), 24, 1341 – 1354.
[3.5.9] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. Nanotoughening of
Epoxies. Proceedings of Stick! Conference April 9th 2003, Nürnberg,
Germany, Vincentz Verlag (2003)
[3.5.19] Xu, S.A., Wang, G.T., Mai, Y.W. Effect of hybridization of liquid
rubber and nanosilica particles on the morphology, mechanical
properties and fracture toughness of epoxy composites. J. Mater.
Sci. (2013), 48, 3546 – 3556.
[3.5.20] Tang, Y., Ye, L., Zhang, Z., Friedrich, K. Interlaminar fracture
toughness and CAI strength of fibre-reinforced composites with
nanoparticles – a review. Comp. Sci. Tech. (2013), 86, 26 – 37.
[3.5.21] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D., Eger, C.
Rubber-toughened FRCs optimised by nano-Particles. JEC
Composites Magazine (2005), No 19, 73 – 76.
[3.5.22] Kinloch, A.J., Mohammed, R.D., Taylor, A.C., Sprenger, S., Egan, D.
The interlaminar toughness of carbon-fibre reinforced plastic
Composites using “hybrid-toughened” matrices. J. Mater. Sci.
(2006), 41, 5043 – 5046.
[3.5.23] Zhang, J., Deng, S., Ye, L., Zhang, Z. Interlaminar Fracture
Toughness and Fatigue Delamination Growth of CF/EP Composites
with Matrices modified by Nanosilica and CTBN rubber. Proc. 13th
Int. Conf. on Fract. (2013), June 16-21, Beijing, China
[3.5.24] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D. Rubber-
toughened CFRCs optimized by nanoparticles – Part III JEC
Composites Magazine (2007), No 30, 54 – 57.
[3.5.26] Kuehn, A., Mahrholz, T., Mosch, Matrix optimization of CFRP parts
concerning fire and impact properties with process acceleration. J.
CEAS Aeronaut J. (2013), 4, 191 – 201.
155
[3.5.27] Tate, J., Trevino, E., Gaikward, S., Sprenger, S., Rosas, I., Andrews,
M. J. Low velocity impact on epoxy glass composites modified with
Rubber micro-particles and silica nano-particles. Journal of
Nanoscience, Nanoengineering, and Applications (2013), vol. 3,
issue 1, ISSN: 2231 – 1777.
[3.5.29] Hsieh, T.H., Kinloch, A.J., Masania, K., Sohn Lee, J., Taylor, A.C.,
Sprenger, S. The toughness of epoxy polymers and fibre composites
modified with rubber microparticles and silica nanoparticles. J.
Mater. Sci. (2010), 45, 1193 – 1210.
156
4. Summary and outlook
A linear relationship between the increase in fracture toughness of the cured bulk
epoxy resin and the silica nanoparticle addition level exists as well. For an
addition of 10 wt% of SiO2 nanoparticles an increase in fracture toughness of
approximatively 50 % can be expected. The mechanisms of toughening are a
void formation at the interface particle/resin matrix, followed by a plastic void
growth. The formation of local plastic shear bands contributes as well. Crack
deflection or dissipation at the particles seems to play no or only a minor role for
toughness improvements.
Fatigue performance at cyclic loading is another property of the cured bulk epoxy
resin improved by the addition of nanosilica. A linear relationship between the
nanoparticle concentration and the fatigue performance seems not to exist;
however there seems to be a maximum which depends on the resin/hardener
system. An addition of 10 wt% SiO2 improves the fatigue performance by
approximately 50-60 %.
Very interesting is the finding that the addition of SiO2 nanoparticles improves
both laminate KIc and GIc values, but no laminate improvements are found for
fast impact tests like a falling dart test.
The classic toughening of epoxy resins employing reactive liquid rubbers or core-
shell rubbers reduces the modulus of cured bulk epoxy resin systems. By the
addition of silica nanoparticles to the toughened epoxy resin matrix this can
nearly be compensated. These so-called hybrid resin systems are characterized
by a very high toughness and stiffness. Depending on the hardener of the epoxy
resin system toughness increases of 100-1200 % can be achieved. Fatigue
performance of the cured bulk resin is improved further compared to a
nanosilica-only modification of the epoxy resin.
157
New is the finding that the toughness of cured bulk hybrids based on core-shell
rubbers and nanosilica behaves additively, whereas for the combination of
reactive liquid rubber and nanosilica a synergy seems to exist. The reason is very
probably an increase in toughening efficiency due to a more ductile matrix. It is
the result of a minor distortion in rubber phase separation during the cure
caused by the presence of the silica nanoparticles and increases with increasing
addition level of nanosilica.
Bulk resin systems with high toughness and stiffness as well as excellent fatigue
performance can be formulated by a modification of 5-10 wt% core-shell rubber
or 5-15 wt% reactive liquid rubber combined with 5-10 wt% SiO2 nanoparticles.
New is the finding that a nearly linear relationship exists between the relative
improvement of the GIc of the cured bulk epoxy resin and the relative
improvement of the GIc of the laminate made thereof. The reinforcing fibers,
regardless if glass or carbon fibers are used, as well as their three-dimensional
orientation in the composite, seem to have little or no influence on this
relationship. A "conversion factor" can be defined which is 0.18 (± 0.02) for the
systems investigated. Means, if the toughness of a composite part shall be
increased by 100 %, the epoxy resin matrix needs to be replaced by a hybrid
resin exhibiting a toughness 555 % higher than the original unmodified bulk
resin.
Of further interest is the fact, that the linear relationship found does not have a
line through origin. A threshold value must exist, below which the toughness of
the bulk resin is increased by the hybrid modification, but no improvements will
be found in the laminate using this hybrid resin system as matrix resin.
In own experiments laminates reinforced with carbon fibers, made by RTM and
based on a fast curing epoxy resin system (DGEBA/IPD/TMD) have been
investigated. As expected the cured bulk resin system with the highest addition
level of nanosilica exhibited the best properties with the highest values for
fracture toughness and modulus.
The laminates made from the rubber-toughened epoxy resin had a somewhat
higher GIc compared to the laminates made from the unmodified control. The
values for GIIc and the interlaminar shear strength (ILSS) were somewhat lower.
The delaminated area after a 30 J impact was smaller; the value for the
compressive strength after impact (CAI) therefore higher.
158
The laminates based on hybrid epoxy resin systems, containing both reactive
liquid rubber and SiO2 nanoparticles, showed no further increase in fracture
toughness compared to the laminates made from rubber-toughened epoxy resin,
but a slight reduction in values for GIc and GIIc. The values for the ILSS were
almost identical, compressive strength after impact was lower for the hybrid
system.
The reason for this unexpected behaviour is very probably the formation of
agglomerates of nanoparticles during the cure. It seems like the fast phase
separation of the reactive liquid rubber, forced by the very fast cure, is
responsible for the agglomerate formation.
The conversion factor between the relative improvement of the value of GIc of the
cured bulk resin and the relative improvement of the laminate GIc was
determined to be 0.2 – in excellent accordance with earlier results.
It can be concluded that composites based on hybrid epoxy resin systems are a
new class in laminate performance – tough and stiff with a excellent fatigue
performance. No wonder they are used in an increasing number of industrial
applications.
Die Eigenschaften von Epoxidharzen können durch den Zusatz von oberflächen-
modifizierten SiO2-Nanopartikeln deutlich verbessert werden. Auch bei höheren
Konzentrationen bleiben die Partikel in der Regel monodispers und es werden
keine Agglomerate gefunden. Die rheologischen Eigenschaften der Harze werden
durch den Nanopartikelzusatz nicht verändert; die Viskosität nur geringfügig
erhöht. Auch die Glasumwandlungstemperatur ändert sich in den meisten Fällen
nicht. Eine neue Erkenntnis ist der annähernd lineare Zusammenhang zwischen
der Erhöhung des Moduls und dem Zusatz an Nanopartikeln. Bei einem Zusatz
von 10 Gewichtsprozent SiO2-Nanopartikeln kann eine Steigerung des Moduls
von etwa 50 % erwartet werden. Druckfestigkeit und Druckmodul werden
ebenfalls erhöht. Verbesserungen von 10-30 % können bei einem Zusatz von 10
Gewichtsprozent SiO2-Nanopartikeln erzielt werden.
159
Die Mechanismen der Verbesserung der Schlagzähigkeit sind hauptsächlich ein
Ablösen des Epoxidharzes von den Nanopartikeln, gefolgt von einer plastischen
Hohlraumvergrößerung. Die Ausbildung lokaler plastischer Scherbänder trägt
ebenfalls zur höheren Schlagzähigkeit bei. Rissumleitung oder –streuung an den
Partikeln scheint hingegen keine oder nur eine untergeordnete Rolle zu spielen.
Die Eigenschaften von Laminaten werden durch eine Modifikation der Harzmatrix
mit SiO2-Nanopartikeln ebenfalls verbessert; so etwa die Bruchzähigkeit um 5-
25 % bei typischen Zusatzmengen von 10-20 Gew% Nanosilica. Die
Kompressionsbeständigkeit und das Ermüdungsverhalten werden signifikant
besser. Dies gilt sowohl für Glasfasern als auch für Carbonfasern als Verstärkung.
160
Das Ermüdungsverhalten von Laminaten auf Basis von Hybridharzsystemen ist
mit Abstand am besten – es sind bis zu zehnfache zyklische Belastungen möglich,
bevor die Laminate versagen.
Neu ist die Erkenntnis, dass eine nahezu lineare Beziehung zwischen den
prozentualen Steigerungen des GIc der gehärteten Reinharzhybridsysteme und
den daraus hergestellten Laminaten existiert. Die Verstärkungsfasern,
unabhängig davon ob Glas oder Carbon, als auch ihre dreidimensionale
Orientierung im Verbund, scheinen wenig oder gar keinen Einfluss auf diese
Beziehung zu haben. Ein “Übertragungsfaktor“ kann definiert werden und beträgt
für die untersuchten Systeme 0,18 (± 0,02). D.h., wenn ein Faserverbundbauteil
um 100 % in der Zähigkeit gesteigert werden soll, muss statt des bisher
verwendeten Harzsystems ein Hybridsystem mit einem 555 % höheren GIc
eingesetzt werden.
Ebenfalls von Interesse ist die Tatsache, dass die gefundene lineare Beziehung
keine Ursprungsgerade ist. Es muss also ein Schwellenwert existieren, bei dem
zwar die Zähigkeit des ausgehärteten Reinharzes durch die Modifikation mit
Elastomeren und Nanopartikeln gesteigert wird, im daraus hergestellten Laminat
aber keine Verbesserung im Vergleich zur unmodifizierten Kontrolle zu finden ist.
Die Laminate auf Basis Hybridharz, sowohl mit reaktivem Flüssigkautschuk als
auch mit SiO2-Nanopartikeln modifiziert zeigten jedoch keine weitere
Verbesserung der Bruchzähigkeit gegenüber der reinen Kautschukmodifikation,
sondern eher geringfügig niedrigere Werte für GIc und GIIc. Die ILSS war
vergleichbar; die Kompressionsbeständigkeit nach Impact jedoch verringert.
Der Grund für dieses unerwartete Verhalten ist möglicherweise in der Bildung
von Nanopartikel-Agglomeraten während der Härtung zu suchen. Wahrscheinlich
ist die rasche Phasentrennung des reaktiven Flüssigkautschukes, erzwungen
durch die rasche Vernetzung, verantwortlich für die Entstehung der Agglomerate.
161
Die Frage, ob die Agglomerate einen negativen Einfluss auf die Performance des
Laminates haben, konnte nicht endgültig beantwortet werden. Weitere
systematische Untersuchungen mit Hybridsystemen auf Basis reaktiver
Flüssigkautschuke unter Verwendung von Härtersystemen mit abgestuften
Härtungsgeschwindigkeiten werden diese Frage beantworten.
162
5. Bibliography
The literature cited in this work is numbered by chapter and then continuously
within the chapter. This was necessary in order to keep the integrity of the
published papers. If literature is cited in several chapters, all numbers are given
as cross-reference.
163
[2.1.10] Pulsstar brochure. (2013) Strongwell, Bristol, VA, USA
164
[2.3.1] Hasson, T. (2013) Neue Matrixtechnologie für die CFK-
Fertigung in der Automobilindustrie. Proceedings of Neue
Technologien im textilbasierten Faserverbund-Leichtbau, 18.
April, DLR Stuttgart, Germany
[2.3.4] Hare, C.J. (1994) Amine curing agents for epoxies. J. Prot.
[3.1.1] Coat. Lin., 9, 77-103
[3.4.3]
[3.5.5]
[2.3.5] Hare, C.J. (1994) Epoxy curing agents II. J. Prot. Coat. Lin.,
[3.1.2] 10, 197-213
[3.4.4]
[3.5.6]
[2.4.2] Kinloch, A.J., Shaw, S.J., Tod, D.A., Hunston D.L. (1983)
[3.2.1] Deformation and fracture behaviour of a rubber-toughened
[3.4.7] epoxy: 1. Microstructure and fracture studies. Polymer, 24,
[3.5.7] 1341–1354 2. Failure criteria. Polymer, 24, 1355-1363
165
[2.4.7.] Sprenger, S., Weber, C., Pulliam, L. (1997) Elastomer-
modified epoxy prepolymers - the new generations. European
Adhesives & Sealants, September 1997, 9-12
[2.4.10] Block, H., Pyrlik, M. (1988) Silicones are the Key: Modifying
[3.2.8] Thermosetting Resins with Silicone Elastomers. Kunstst. Ger.
Plast. 78, 1192-1196
[2.4.12] Chen, J., Kinloch, A.J., Sprenger, S., Taylor, A.C. (2013) The
[3.2.9] mechanical properties and toughening mechanisms of an
epoxy polymer modified with polysiloxane-based core-shell
particles. Polymer 54, 4276-4289
[2.4.14] Yoon, T.H., Priddy, D.B., Lyle, G.D., McGrath, J.E. (1995)
Mechanical and morphological investigations of reactive
polysulfone toughened epoxy networks. Macromol. Symp. 98,
673-686
[2.5.2] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. (2003)
[3.3.1] Nanotoughening of Epoxies. Proceedings of Stick! Conference
[3.4.15] April 9th, Nürnberg, Germany, Vincentz Verlag 2003
[3.5.9]
166
[2.5.3] Picture courtesy of IVW Kaiserslautern and
[3.2.11] Polymerservice Merseburg, Germany
[3.4.19]
[3.5.10]
[2.5.6] Eger C., Schultz, P. (2005) Reinforcing epoxy resins with silica
[3.1.28] nanoparticles. In Proceedings of "High Performance Fillers
2005", March 8-9, Köln, Germany
[2.5.9] Liu, S., Zhang, H., Zhang, Z., Zhang, T., Sprenger, S. (2008)
[3.1.33] Tailoring the mechanical performance of epoxy resin by
various nanoparticles. Polymers & Polymer Composites Vol.
16, No. 8, 471-477
[2.5.11] Gou, J., Fan, B., Song, G., Khan, A. (2006) Study of affinities
between single-walled nanotube and epoxy resin using
molecular dynamic simulation. Int. J. Nanosci. 5, 131-144
[2.5.13] Sprenger, S., Eger, C. Kinloch, A.J., Lee, J.H., Taylor, A.C.,
[3.2.13] Egan, D. (2003) Nanoadhesives: Toughness and high
strength. Adhäsion, Kleben & Dichten 03/2003, 24-30
167
[2.5.15] Picture courtesy of J. Robinette, Army Research
[3.2.20] Laboratory, MD, U.S.A.
[3.5.16]
[2.5.19] Liu, H.Y., Wang, G.T., Mai, Y.W. Zeng, Y. (2011) On fracture
[3.1.19] toughness of nanoparticle modified epoxy. Composites: Part
[3.2.35] B 42, 2170-2175
168
[2.5.26] Tang, Y., Ye, L., Zhang, D., Deng, S. (2011) Characterization
[3.1.18] of transverse tensile, interlaminar shear and interlaminate
[3.3.19] fracture in CF/EP laminates with 10 wt% and 20 wt% silica
nanoparticles in matrix resins. Composites: Part A, 42, 1943-
1950
[2.5.27] Sprenger, S., Eger, C., Kinloch, A., Mohammed, R., Taylor, A.
[3.3.12] (2005) Rubber-toughening and Nanoparticles in Epoxies:
Synergies in FRC. Proceedings SAMPE, Paris, France, April 5 –
7, 2005
[3.1.4] Sanctuary, R., Baller, J., Krüger, J.-K., Schäfer, D; Wetzel, B.,
Possart, W., Alnot, P. (2006) Complex specific heat capacity
of two nanocomposite systems. Thermochimica Acta 445, No.
2, 111-115
169
[3.1.11] Dittanet, P. (2008) The use of nanosilica in epoxy resins.
Master Thesis, Lehigh University, Bethlehem, Pennsylvania,
U.S.A.
[3.1.12] Wang, G.T., Liu, H.Y., Saintier, N., Mai, Y.W. (2009) Cyclic
[3.2.34] fatigue of polymer nanocomposites. Eng. Failure Analysis
16, 2635-2645
[3.1.14] Baller, J., Becker, N., Ziehmer, M., Thomassey, M., Zielinski,
B., Müller, U., Sanctuary, R. (2009) Interactions between
silica nanoparticles and an epoxy resin before and during
network formation. Polymer 50, 3211-3219
[3.1.15] Philipp, M., Müller, U., Jiménez Riobóo, R.J., Baller, J.,
Sanctuary, R., Possart, W., Krüger, J.K. (2009) Interphases,
gelation, vitrification, porous glasses and the generalized
Cauchy relation: epoxy/silica nanocomposites. New Journal of
Physics 11, 023015
[3.1.20] Liu, H.Y., Wang, G.T., Mai, Y.W. (2012) Cyclic fatigue crack
[3.2.36] propagation of nanoparticle modified epoxy. Comp. Sci. Tech.
72, 1530-1538
[3.1.22] Hsieh, T.H., Kinloch, A.J., Masania, K., Taylor, A.C., Sprenger,
S. (2010) The mechanisms and mechanics of the toughening
of epoxy polymers modified with silica nanoparticles. Polymer
51, 6284-6294
[3.1.23] Ma, J., Mo, M.S., Du, X.S., Rosso, P., Friedrich, K. Kuan, H.C.
(2008) Effect of inorganic nanoparticles on mechanical
property, fracture toughness and toughening mechanism of
two epoxy systems. Polymer 49, 3510-3523
170
[3.1.26] Pethrick, R., Miller, C., Rhoney, I. (2010) Influence of
nanosilica particles on the cure and physical properties of a
thermoset epoxy resin. Polym. Int. 59, 236–241
[3.1.30] Blackman, B.R.K., Kinloch, A.J., Sohn Lee, J., Taylor, A.C.,
Agarwal, R., Schueneman, G., Sprenger, S. (2007) The
fracture and fatigue behaviour of nano-modified epoxy
polymers. J. Mater. Sci. Letters 42, 7049-7051
[3.1.31] Liu, S., Zhang, H., Zhang, Z., Sprenger, S. (2008) Epoxy
Resin Filled with High Volume Content Nano-SiO2 Particles.
J. Nanosci. Nanotechnol. Vol. 8, 1-6
[3.1.32] Zjang, H., Zhang, Z., Friedrich, K., Eger, C. (2006) Property
improvements of in situ epoxy nanocomposites with reduced
interparticle distance at high nanosilica content. Acta Mater.
54, 1833-1842
[3.1.34] Hsieh, T.H., Kinloch, A.J., Masania, K., Sohn Lee, J., Taylor,
[3.2.28] A.C., Sprenger, S. (2010) The toughness of epoxy polymers
[3.3.16] and fibre composites modified with rubber microparticles and
[3.4.31] silica nanoparticles. J. Mater. Sci. 45, 1193-1210
[3.5.29]
[3.1.36] Tang, L.C., Zhang, H., Sprenger, S., Ye, L., Zhang, Z. (2012)
[3.2.30] Fracture mechanisms of epoxy-based ternary composites
filled with rigid-soft particles. Compos. Sci. Technol. 72, issue
5, 558–565
[3.1.38] Wang, Z.Z., Gu, P., Zhang, Z., Gu, L., Xu, Y.Z. (2011)
Mechanical and Tribological Behavior of Epoxy/Silica Nano-
composites at the Micro/Nano Scale. Tribol. Lett. 42, 185–
191
[3.1.39] Zhang, Y.F., Bai, S.L., Li, X.K., Zhang, Z. (2009) Viscoelastic
Properties of Nanosilica-Filled Epoxy Composites Investigated
by Dynamic Nanoindentation. J. Polym. Sci. Part B: Polym.
Phys. 47, 1030–1038
171
[3.1.40] Mahrholz, T., Stängle, J., Sinapius, M. (2009) Quantitation of
the reinforcement effect of silica nanoparticles in epoxy resins
used in liquid composites moulding processes. Composites:
Part A 40, 235-243
[3.1.41] Duwe, S., Arlt, C., Aranda, S., Riedel, U., Ziegmann, G.
(2012) A detailed thermal analysis of nanocomposites filled
with SiO2, AlN or boehmite at varied contents and a review of
selected rules of mixture. Comp. Sci. Tech. 72, 1324-1330
[3.1.42] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010) The
effects of nanosilica contents on thermal and mechanical
properties of epoxy polymers"; Proceedings of 10th SAMPE
Europe conference, 12-14 April, Paris, France
[3.1.43] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010) Effect of
silica nanoparticles on compressive properties of an epoxy
polymer. J Mater. Sci. 45, No. 21, 5973-5983
[3.1.45] Sprenger, S., Kinloch, A.J., Taylor, A.C., Hsieh, T.-H. (2009)
[3.2.33] Ultra-tough and fatigue resistant. Adhesion, Adhesives &
Sealants 10, 8–11
172
[3.2.23] Veena, M.G., Renukappa, N.M., Raj, J.M., Ranganathaiah, C.,
Shivakumar, K.N. (2011) Characterization of Nanosilica-Filled
Epoxy Composites for Electrical and Insulation Applications. J.
Appl. Polym. Sci. 121, 2752–2760
[3.2.29] Tang, L.C., Zhang, H., Pei, Y.B., Chen, L.M., Song, P.,
Zhang, Z. (2013) Improved impact resistance of epoxy resin
filled with rigid and soft nanoparticles. Composites: Part A
accepted for publication
[3.2.31] Sprenger, S., Eger, C., Kinloch, A.J., Lee, J.H., Taylor, A.C.,
Egan, D. (2003) Toughening structural adhesives via nano-
and micro-phase inclusions. Journal of Adhesion 79, 867–
873
[3.2.32] Sprenger, S., Eger, C., Kinloch, A.J., Lee, J.H., Taylor, A.C.,
Egan, D. (2004) Nano-modified ambient temperature curing
epoxy adhesives. Adhäsion, Kleben & Dichten 03/2004, 17–21
173
[3.3.11] Kinloch, A.J., Masania, K., Taylor, A.C., Sprenger, S. (2009)
[3.4.29] The Fracture of Nanosilica and Rubber Toughened Epoxy Fibre
Composites. Proceedings of the Composites & Polycon,
American Composites Manufacturers Association, January 15
17, Tampa, Florida, U.S.A.
[3.3.20] Zeng, Y., Liu, H.Y., Mai, Y.W., Du, X.S. (2012) Improving
interlaminar fracture toughness of carbon fibre/epoxy
laminates by incorporation of nanoparticles. Composites:
Part B 43, 90-94
[3.3.23] Hackett, S.H., Nelson, J.M., Hine, A.M., Sedgwick, P., Lowe,
R.H., Goetz, D.P., Schultz, W.J. (2010) The Effect of
Nanosilica Concentration on the Enhancement of Epoxy Matrix
Resins for Prepreg Composites. Proceedings SAMPE 2010,
Oct. 11-14, Salt Lake City, UT, U.S.A.
[3.3.24] Nelson, J.M., Hackett, S.C., Goetz, D.P., Hine, A.M., Schultz,
W.J. (2012) Development of Nanosilica-Thermoset Matrix
Resins for Prepreg Composites. NSTINanotech 2012, ISBN
978-1-4665-6274-5 Vol. 1
[3.3.25] Nelson, J.M., Hine, A.M., Goetz, D.P., Sedgwick, P., Lowe,
R.H., Rexeisen, E., King, R.E., Patz, N. (2012) Nanosilica-
modified Epoxy Matrix Resin for Prepreg Composite Tooling
Applications. Proceedings SAMPE 2012, May 21-24, Baltimore,
MD, U.S.A.
[3.3.26] Nelson, M., Hine, A.M., Goetz, D.P., Sedgwick, P., Lowe,
R.H., Rexeisen, E., King, R.E., Aitken, C., Pham, Q. (2013)
Properties and Applications of Nanosilica-modified Tooling
Prepregs. SAMPE Journal 49, No. 1, 7-17
[3.3.27] Chiu, K.R., Duenas, T., Dzenis, Y., Kaser, J., Bakis, C.E.,
Roberts, J.K., Carter, D. (2013) Comparative Study of Nano-
materials for Interlaminar Reinforcement of Fiber-Composite
Panels. Proc. of SPIE 2013, 8689, 86891D1
174
[3.3.28] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010)
Improved Compressive Properties of a Unidirectional CFRP
Laminate Using Nanosilica Particles. Advanced Composites
Letters 19 (6), 218-221
[3.3.31] Liu, L.Q., Li, L., Gao, Y., Tang, L., Zhang, Z. (2013) Single
carbon fiber fracture embedded in an epoxy matrix modified
by nanoparticles. Composites Science and Technology 77,
101-109
[3.3.33] Gao, X., Jensen, R.E., McKnight, S.H., Gillespie Jr., J.W.
(2011) Effect of colloidal silica on the strength and energy
absorption of glass fiber/epoxy interphases. Composites:
Part A 42, 1738-1747
[3.3.34] Yang, Y., Lu, C.X., Su, X.L., Wu, G.P., Wang, X.K. (2007)
Effect of nano-SiO2 modified emulsion sizing on the interfacial
adhesion of carbon fibers reinforced composites. Materials
Letters 61, 3601-3604
175
[3.4.23] Sun, C.T. (2012) Development of toughened and multi-
functional Nanocomposites for ship structures. Final Report
for Office of Naval Research, School of Aeronautics and
Astronautics, Purdue University, West Lafayette, IN 47907,
U.S.A.
[3.4.26] Schultz, R., Tate, J.S., Gaikwad, S., Trevino, E., Jacobs, C.,
Sprenger, S. (2011) Low velocity impact on epoxy glass
composites modified with rubber micro-particles and silica
nanoparticles. Proceedings of the SAMPE TECH Conference,
October 18-19, Fort Worth, TX, USA SAMPE Covina, CA,
U.S.A.
[3.4.27] Tate, J., Trevino, E., Gaikward, S., Sprenger, S., Rosas, I.,
[3.5.27] Andrews, M. J. (2013) Low velocity impact on epoxy glass
composites modified with rubber micro-particles and silica
nanoparticles. Journal of Nanoscience, Nanoengineering, and
Applications, vol. 3, issue 1, ISSN: 2231-1777
[3.4.42] Zeng, Y., Liu, H.Y., Mai, Y.W., Du, X.S. (2012) Improving
interlaminar fracture toughness of carbon fibre/epoxy
laminates by incorporation of nanoparticles. Composites: Part
B 43, 90–94
176
[3.4.45] Jordan, W.M., Bradley, W.L. (1988) The relationship between
resin mechanical properties and Mode I delamination fracture
toughness. J. Mat. Sci. Lett., 7, 1362–1364
[3.5.2] Estin & Co. (2011) Main dynamics of the composite industry
for automotive applications 2010-2015. Publ. by JEC
Composites, Paris, France
[3.5.19] Xu, S.A., Qang, G.T., Mai, Y.W. (2013) Effect of hybridization
of liquid rubber and nanosilica particles on the morphology,
mechanical properties and fracture toughness of epoxy
composites. J. Mater. Sci. 48, 3546-3556
[3.5.20] Tang, Y., Ye, L., Zhang, Z., Friedrich, K. (2013) Interlaminar
fracture toughness and CAI strength of fibre-reinforced
composites with nanoparticles - a review. Comp. Sci. Tech.
86, 26-37
[3.5.23] Zhang, J., Deng, S., Ye, L., Zhang, Z. (2013) Interlaminar
Fracture Toughness and Fatigue Delamination Growth of
CF/EP Composites with Matrices Modified by Nano-silica and
CTBN rubber. Proc. 13th Int. Conf. on Fract., June 16-21,
Beijing, China
177