Article

Cure kinetics of a fast-cure epoxy with silica nanoparticles

Abstract and Figures

This article describes how the strong exothermic reaction of a fast-cure epoxy can be better controlled with the addition of non-reactive silica nanoparticles, further enhancing the benefits of a fast process, the toughness and the ability to understand the properties of thicker composite parts. F ast-cure epoxy systems are used for the mass production of composite parts with cycle times in the minute range. However , the use of fast-cure resins results in new challenges regarding the understanding of their properties and processing behaviour. One of the major difficulties with fast-cure resins is their strong exother-mic reaction during cure, which may result in a significant temperature overshoot and large temperature (and therefore T g , shrinkage and residual stress) gradients over the thickness. Further, the resin viscosity evolution strongly affects the impregnation process. Cure cycle optimisation through modelling of heat transfer and flow helps to find more suitable process parameters to reduce the exothermic reaction. However, in some cases it is also necessary to reduce the cure temperature with a negative influence on the cycle time. A possible approach is to reduce the exothermic mass of the resin itself, which can be achieved by adding non-reactive particles to the resin. In composite manufacturing, any such particles must be small enough that they are not filtered out during infusion processes , for example silica nanoparticles with a diameter of 20 nm [1]. The addition of silica nanoparticles up to 20 wt.% has been shown to increase stiffness and fracture energy [2] without having a strong influence on the viscosity of the epoxy matrix [3]. This article focuses on a fast-cure resin for automotive applications, XB3585/XB3458, supplied by Huntsman, Switzerland. This epoxy has a cure time of 5 min at 100°C. The results are compared to Hex-Flow RTM6, a commonly used RTM epoxy. Nanopox F400 from Evonik Hanse, Germany, was supplied as a masterbatch containing 40 wt.% of silica nanoparticles predispersed in a diglycidyl ether of bisphenol A (DGEBA) epoxy resin. These particles were then mixed with the XB3585 resin and the masterbatch resin in order to produce formulations including 10 wt.% and 20 wt.% of silica nanoparticles. Reaction kinetics Cure kinetics of a fast-cure epoxy with silica nanoparticles Software Structural health of polymer composites: non-destructive diagnosis using a hybrid NDT approach 59 62 N o
Content may be subject to copyright.
reaction kinetics TECHNOLOGY
TECHNOLOGY
Cure kinetics of a fast-cure epoxy
with silica nanoparticles
This article describes how the strong exothermic reaction of
a fast-cure epoxy can be better controlled with the addition of
non-reactive silica nanoparticles, further enhancing the benefits
of a fast process, the toughness and the ability to understand the
properties of thicker composite parts.
Fast-cure epoxy systems are used for the mass production of
composite parts with cycle times in the minute range. Howev-
er, the use of fast-cure resins results in new challenges regarding
the understanding of their properties and processing behaviour. One
of the major diculties with fast-cure resins is their strong exother-
mic reaction during cure, which may result in a signicant temper-
ature overshoot and large temperature (and therefore Tg, shrinkage
and residual stress) gradients over the thickness. Further, the resin
viscosity evolution strongly aects the impregnation process.
Cure cycle optimisation through modelling of heat transfer and
ow helps to nd more suitable process parameters to reduce the
exothermic reaction. However, in some cases it is also necessary to
reduce the cure temperature with a negative inuence on the cycle
time. A possible approach is to reduce the exothermic mass of the
resin itself, which can be achieved by adding non-reactive particles
to the resin. In composite manufacturing, any such particles must
be small enough that they are not ltered out during infusion pro-
cesses, for example silica nanoparticles with a diameter of 20 nm [1].
e addition of silica nanoparticles up to 20 wt.% has been shown
to increase stiness and fracture energy [2] without having a strong
inuence on the viscosity of the epoxy matrix [3].
is article focuses on a fast-cure resin for automotive applications,
XB3585/XB3458, supplied by Huntsman, Switzerland. is epoxy
has a cure time of 5 min at 100°C. e results are compared to Hex-
Flow RTM6, a commonly used RTM epoxy. Nanopox F400 from
Evonik Hanse, Germany, was supplied as a masterbatch containing
40 wt.% of silica nanoparticles predispersed in a diglycidyl ether
of bisphenol A (DGEBA) epoxy resin. ese particles were then
mixed with the XB3585 resin and the masterbatch resin in order
to produce formulations including 10 wt.% and 20 wt.% of silica
nanoparticles.
Reaction kinetics
Cure kinetics of a fast-cure epoxy
with silica nanoparticles
Software
Structural health of polymer composites:
non-destructive diagnosis using a hybrid NDT approach
59
62
No107 August - September 2016 / jec composites magazine 59
Andre Keller,
Research Assistant,
Prof. Clemens drAnsfeld,
Head of Institute of Polymer
Engineering, Institute of Polymer Engineering, FHNW
University of Applied Sciences and Arts Northwestern
Switzerland,
dr Ambrose C. TAylor,
Reader in M aterials Enginee-
ring, Department of Mechanical Engineering, Imperial
College London,
dr.dr.-Ing. sTePhAn sPrenger,
Senior Market Development
Manager Composites & Lightweight Construction
Evonik Hanse GmbH
KlAus rITTer,
Head Centre of Excellence Composites
Huntsman Advanced Materials
dr KunAl mAsAnIA,
Senior Scientist
Complex Materials Group, Department of Materials,
ETH Zürich
Fig. 4: Comparison
of the maximum
heat flow of the
epoxy systems
under isothermal
conditions [5]
Cure kinetics of fast-cure epoxy
As the epoxy cure reaction is exothermic, the resin temperature can
overshoot the mould temperature. When manufacturing a composite
plate in a mould, the magnitude of the overshoot depends on the mass
of resin and the plate thickness. e heat ow in the resin is limited by
the low thermal conductivity (0.2 W/(m °C)) of the epoxy, meaning
that heat will not readily dissipate once produced.
e exothermic reaction during cure can be characterised by meas-
uring the heat ow using dierential scanning calorimetry (DSC).
e severity of the exothermic reaction depends on the total heat of
reaction (integral of the heat ow curve) and the time span in which
the heat is released. Isothermal DSC measurements of the XB3585/
XB3458 and RTM6 systems are shown in Fig. 2. Two important
dierences between these two epoxies explain why the XB3585/
XB3458 system has a much higher tendency to overshoot the mould
temperature than RTM6:
i) e total heat of reaction derived from dynamic DSC measure-
ments is in the same order of magnitude for both materials with a
slightly higher value of 494 ±4 J/g for the fast-cure resin and 457 J/g
for RTM6 [4]. However, the release of a similar amount of energy
within a much shorter time span results in a higher heat ow. e heat
ow of the XB3585/XB3458 system at 100°C is a factor of 20 higher
than RTM6 at 180°C, their respective recommended processing tem-
peratures.
ii) e heat ow of the XB3585/XB3458 system peaks very early in
the cure process, indicating that a large amount of heat is released at
cure reaction initiation. In contrast, a slow, steady increase of the heat
ow can be measured with RTM6.
A temperature progression measured during cure is shown in Fig. 3
for a 3-mm-thick bulk epoxy plate manufactured using an aluminium
mould at 100°C. e temperature increased from 100°C to 176°C
within 20 seconds. Typically, the temperature distribution during
such a strong overshoot is not homogeneous but rather shows a sig-
nicant variation with the maximum temperature in the centre of the
mould, and only a small overshoot close to the mould surface. ese
phenomena aect the part during processing and can result in defects
of the nal part such as material decomposition or internal stress.
Influence of silica nanoparticles
Dynamic and isothermal DSC measurements were conducted and
the maximum heat ow was extracted, as shown in Fig. 4. e heat
ow was found to be inversely proportional to the wt.% of silica nan-
oparticles. e glass transition temperature, Tg, remained unaected
with a constant value of 121 ±1°C. ese results indicate that the par-
ticles do not inuence the cure reaction adversely, but do reduce the
exothermic mass.
e temperature overshoot was reduced from 176°C for the unmod-
ied epoxy to 166°C and 157°C with the addition of 10 wt.% and 20
wt.% of silica nanoparticles respectively. e importance of this tem-
perature reduction for the modelled case can be shown by comparing
Fig. 1: Part of
a convertible roof
cover manufactured
with a toughened,
fast-cure resin
TECHNOLOGY reaction kinetics
60 jec composites magazine / No107 August - September 2016
Fig. 2: Isothermal DSC
measurements of the
XB3585/XB3458 system
compared to RTM6 epoxy
resin
Fig. 3: Temperature
progression during
cure of the XB3585/
XB3458 epoxy resin
[3]
the glass transition temperature (Tg) distributions over the thickness,
as shown in Fig. 5. In the case of the unmodied epoxy, the dierence
between the Tg at the edge and in the middle is 11°C. e addition
of 20 wt.% of silica nanoparticles results in a dierence of only 5.5°C.
e temperature distribution over the thickness was not completely
dismissed with the addition of silica nanoparticles but signicantly re-
duced, leading to more uniform properties over the thickness.
As, typically, the addition of particles can increase the viscosity, the
eect of particle concentration on the initial viscosity and cure cy-
cle evolution was also studied. e eect of silica nanoparticles was
found to be relatively small for silica contents up to 20 wt.% (Fig. 6),
whereas a higher wt.% results in an exponential viscosity increase. e
inuence of silica nanoparticles on the viscosity of the fast-cure resin
becomes negligible at about 10 Pas as the cure reaction kinetics dom-
inate the evolution of the viscosity, explaining why no eect on the
gelation time was noticed.
Modelling
e temperature distribution during cure can be modelled by incor-
porating the cure kinetics into the heat transfer equation [3]. e pre-
diction is shown in Fig. 3. Modelling the temperature makes it possi-
ble to calculate the viscosity, the degree of cure and the glass transition
temperature (more details are available in a previous study [3]). e
T
g gradient over the thickness of a 3 mm bulk epoxy plate without
and with 10 and 20 wt.% silica nanoparticles is shown in Fig. 5. ese
models may be implemented into ow simulations to predict the ow
front during bre impregnation in liquid composite moulding (LCM)
processes.
Conclusions
e addition of silica nanoparticles is a suitable method to reduce ex-
othermic heat generation during cure without inuencing the curing
reaction itself. is allows the use of fast cure cycles, makes the pro-
cess more controllable and leads to more uniform properties over
the thickness of the part. Since the viscosity is only slightly increased
with contents up to 20 wt.%, the silica nanoparticles are well suited for
liquid composite moulding (LCM) processes. Modelling of the cure
kinetics, heat transfer and viscosity combined with ow simulations
may help to nd the optimum process parameters during preform im-
pregnation.
More information: www.fhnw.ch/ikt
Contact: andre.keller@fhnw.ch
References
[1] Kinloch, A. J., Masania, K., Taylor, A. C., Sprenger, S., & Egan, D.
(2008). The fracture of glass-fibre-reinforced epoxy composites using
nanoparticle-modified matrices. Journal of Materials Science, 43(3),
1151-1154.
[2] Hsieh, T. H., Kinloch, A. J., Masania, K., Sohn Lee, J., Taylor, A. C.,
& Sprenger, S. (2010). The toughness of epoxy polymers and fibre
composites modified with rubber microparticles and silica nanoparti-
cles. Journal of Materials Science, 45(5), 1193-1210.
[3] Keller, A., Masania, K., Taylor, A. C., & Dransfeld, C. (2016).
Fast-curing epoxy polymers with silica nanoparticles: properties and
rheo-kinetic modelling. Journal of Materials Science, 51(1), 236-251.
[4] Studer, J., Dransfeld, C., & Masania, K. (2016). An analytical model
for B-stage joining and co-curing of carbon fibre epoxy composites.
Composites Part A: Applied Science and Manufacturing, 87, 282-289.
[5] Keller, A., Masania, K., Taylor, A. C., & Dransfeld, C. (2015). Mod-
elling characterisation of a fast curing silica nanoparticle modified
epoxy. International Conference on Composite Materials (ICCM) 2016,
Copenhagen.
Cure kinetics of a fast-cure epoxy with silica nanoparticles
No107 August - September 2016 / jec composites magazine 61
Fig. 5: Modelled
glass transition
temperature gra-
dient over the plate
thickness during
cure [5]
Fig. 6: Initial visco-
sity of the XB3585/
XB3458 epoxy resins
[3]
... The modification of the epoxy resins used with silica nanoparticles does not only improve the mechanical performance, but can reduce the exotherm. Taylor et al. were able to show that the curing reaction itself was not influenced but the total heat of reaction was reduced [22,23]. Furthermore, they were able to develop a kinetic and rheological model that enables them to predict and control fast cure cycles and/or manufacture thick parts. ...
Article
Full-text available
Surface-modified silica nanoparticles are available as concentrates in epoxy resins in industrial quantities for nearly 20 years. Meanwhile, they are used in many epoxy resin formulations for various applications like fiber-reinforced composites, adhesives or electronic components; even in space vehicles like satellites. Some of the drawbacks of "classic" epoxy toughening using elastomers as a second phase, like lower modulus or a loss in strength can be compensated by using nanosilica together with such tougheners. Apparently, there exists a synergy as toughness and fatigue performance are increased significantly. This work intends to provide an overview regarding the possibilities of nanotoughening with silica, the industrial applications of such epoxy resin formulations and the most recent research results.
Conference Paper
Full-text available
Fast-curing epoxy polymers allow composite parts to be manufactured in minutes, but the curing reaction is highly exothermic with heat flows up to 20 times higher than conventional epoxies. The low thermal conductivity of the polymer causes the mechanical and kinetic properties of parts to vary through their thickness. In the present work, silica nanoparticles were used to reduce the exotherm, and hence improve the consistency of the manufactured parts. The kinetic properties were measured as a function of part thickness and it was noted that the exothermic heat of reaction can be significantly reduced with the addition of silica nanoparticles, which were well-dispersed in the epoxy. A model was developed to describe the increase in viscosity and degree of cure of the unmodified and the silica-modified epoxies. A heat transfer equation was used to predict the temperature and resulting properties through the thickness of a plate, as well as the effect of the addition of silica nanoparticles. No significant viscosity increase was found with the addition of up to 20 wt% of silica nanoparticles. The predictions were compared to the experimental data, and the agreement was found to be very good.
Article
Full-text available
The present paper investigates the effect of adding silica nanoparticles to an anhydride-cured epoxy polymer in bulk and when used as the matrix of carbon- and glass-fibre reinforced composites. The formation of ‘hybrid’ epoxy polymers, containing both silica nanoparticles and carboxyl-terminated butadiene-acrylonitrile (CTBN) rubber microparticles, is also discussed. The structure/property relationships are considered, with an emphasis on the toughness and the toughening mechanisms. The fracture energy of the bulk epoxy polymer was increased from 77 to 212J/m2 by the presence of 20wt% of silica nanoparticles. The observed toughening mechanisms that were operative were (a) plastic shear-yield bands, and (b) debonding of the matrix from the silica nanoparticles, followed by plastic void-growth of the epoxy. The largest increases in toughness observed were for the ‘hybrid’ materials. Here a maximum fracture energy of 965J/m2 was measured for a ‘hybrid’ epoxy polymer containing 9wt% and 15wt% of the rubber microparticles and silica nanoparticles, respectively. Most noteworthy was the observation that these increases in the toughness of the bulk polymers were found to be transferred to the fibre composites. Indeed, the interlaminar fracture energies for the fibre-composite materials were increased even further by a fibre-bridging toughening mechanism. The present work also extends an existing model to predict the toughening effect of the nanoparticles in a thermoset polymer. There was excellent agreement between the predictions and the experimental data for the epoxy containing the silica nanoparticles, and for epoxy polymers containing micrometre-sized glass particles. The latter, relatively large, glass particles were investigated to establish whether a ‘nano-effect’, with respect to increasing the toughness of the epoxy bulk polymers, did indeed exist.
Article
Full-text available
The use of silica nanoparticles and rubbery particles to synthesize hybrid-toughened epoxy polymers has been shown to give a range of novel matrices that can be used to produce composite laminates by a resin infusion under flexible tooling (RITF) process. The glass-fiber-reinforced polymer (GFRP) composite panels were manufactured by RIFT. A thin film of poly(tetrafluoroethylene) (PTFE) was inserted into the fabric prior to resin infusion to act as a starter crack for the fracture specimens. It was observed that the presence of the rubber and nanoparticles does not significantly decrease the flexural modulus or the glass transition temperature of the composite. These formulations exhibit significant increase in the interlaminar fracture energy in both mode I and mode II fracture of GFRP composites.
Article
The objective of this paper was to develop cure kinetic models to describe the B-stage curing and co-curing assembly of carbon fibre reinforced thermosetting polymer (CFRP) composites. Starting from the analytical model, temperature cycles and experimental procedures are developed to join a B-stage CFRP part to a reinforcing B-stage CFRP patch for local reinforcement. Our results show that by using the analytical model, one may precisely describe the cure reaction and join the composites without any additional adhesive. The co-cured composites were successfully manufactured with stable volume fractions and glass transition temperatures between the two sub-components. Additionally, merits of the process, such as modifying reinforcing areas locally, or formation of net shape detail are discussed.