wrap>Yahoo! is not affiliated with the authors of this page or responsible for its content.
Fatigue crack propagation in microcapsule-toughened epoxy
Abstract
The addition of liquid-lled urea-formalde-
hyde (UF) microcapsules to an epoxy matrix leads to
signicant reduction in fatigue crack growth rate and
corresponding increase in fatigue life. Mode-I fatigue
crack propagation is measured using a tapered double-
cantilever beam (TDCB) specimen for a range of
microcapsule concentrations and sizes: 0, 5, 10, and
20% by weight and 50, 180, and 460 lm diameter.
Cyclic crack growth in both the neat epoxy and epoxy
lled with microcapsules obeys the Paris power law.
Above a transition value of the applied stress intensity
factor DK
T
, which corresponds to loading conditions
where the size of the plastic zone approaches the size
of the embedded microcapsules, the Paris law expo-
nent decreases with increasing content of microcap-
sules, ranging from 9.7 for neat epoxy to approximately
4.5 for concentrations above 10 wt% microcapsules.
Improved resistance to fatigue crack propagation,
indicated by both the decreased crack growth rates and
increased cyclic stress intensity for the onset of unsta-
ble fatigue-crack growth, is attributed to toughening
mechanisms induced by the embedded microcapsules
as well as crack shielding due to the release of uid as
the capsules are ruptured. In addition to increasing the
inherent fatigue life of epoxy, embedded microcapsules
lled with an appropriate healing agent provide
a potential mechanism for self-healing of fatigue
damage.
Introduction
Highly crosslinked epoxy resins have low strain-
to-failure and exhibit poor resistance to crack propa-
gation. Fatigue loading is particularly problematic,
causing small cracks to initiate and grow rapidly. These
cracks often lead to catastrophic failure. An extensive
body of work exists for the general area of fatigue of
polymers [13], which focuses on understanding the
mechanisms of fatigue and predicting the rates of
fatigue-crack growth.
Fatigue crack propagation studies are performed
with the cyclic-crack-tip stress state varying over a
range dened by DK
I
(K
max
) K
min
). Dependence
of the fatigue-crack-growth rate da/dN on the applied
range of stress intensity factors DK
I
is generally
described by the empirical Paris law equation [4]
da
dN ¼ C
0
D
K
n
I
;
ð1Þ
E. N. Brown (
&) Æ N. R. Sottos
Department of Theoretical and Applied Mechanics
and the Beckman Institute for Advanced Science and
Technology, University of Illinois at Urbana-Champaign,
216 Talbot Laboratory, 104 S. Wright St., Urbana, IL 61801,
USA
e-mail: en_brown@lanl.gov
N. R. Sottos
e-mail: n-sottos@uiuc.edu
Present Address:
E. N. Brown
Materials Science and Technology Division, MS G-755, Los
Alamos National Laboratory, Los Alamos, NM 87545, USA
S. R. White
Department of Aerospace Engineering and the Beckman
Institute for Advanced Science and Technology, University
of Illinois at Urbana-Champaign, 306 Talbot Laboratory,
104 S. Wright St., Urbana, IL 61801, USA
e-mail: swhite@uiuc.edu
J Mater Sci (2006) 41:62666273
DOI 10.1007/s10853-006-0512-y
123
Fatigue crack propagation in microcapsule-toughened epoxy
E. N. Brown Æ S. R. White Æ N. R. Sottos
Received: 4 May 2004 / Accepted: 1 September 2005 / Published online: 12 August 2006
Springer Science+Business Media, LLC 2006 where C
0
and n are material constants that depend on
the ratio of applied stress intensity R K
min
/K
max
,
the loading frequency f, and the testing environment.
The typical crack growth behavior described by Eq. (1)
yields a linear loglog plot that is bounded by a
threshold stress intensity range DK
th
below which a
crack ceases to propagate, and the critical stress
intensity K
IC
above which crack growth in unstable.
Several researchers [57] have successfully mea-
sured fatigue-crack propagation in epoxy resins and
obtained values of the Paris law exponent n on the
order of 10. Incorporation of either a rubbery second
phase [811] or solid particles [7, 12, 13] signicantly
improves the resistance to fatigue-crack propagation.
Several of these studies [5, 8, 9, 11, 13, 14] suggest that
improvements in the resistance to fatigue crack prop-
agation behavior are also associated with increased
toughness in monotonic fracture [1, 2, 14, 15].
Previously, we investigated the effect of embedded
urea-formaldehyde (UF) microcapsules on the mono-
tonic fracture properties of a self-healing epoxy [15]. In
addition to providing an efcient mechanism for self-
healing [1618], the presence of liquid-lled micro-
capsules increased the virgin monotonicfracture
toughness of epoxy by up to 127% [15, 17]. The
increased toughening was correlated with a change in
the fracture plane morphology from mirror-like to
hackle markings with subsurface microcracking. The
inherent fracture toughness as well as the healing
efciency both depended strongly on the size and
concentration of microcapsules. In the current work,
we extend this investigation to examine the inuence
of microcapsules on the fatigue crack propagation
behavior of epoxy, with the effects of self-healing
precluded. Consistent with the monotonic fracture
studies, the addition of microcapsules to an epoxy
matrix signicantly increased the resistance to crack
growth under dynamic loading conditions.
Experimental procedure
Materials and sample preparation
Urea-formaldehyde microcapsules containing dicyclo-
pentadiene (DCPD) monomer were manufactured with
average diameters of 50, 180, and 460 lm using the
emulsion in situ polymerization microencapsulation
method outlined by Brown et al. [19]. Shell wall thick-
ness was 190 ± 30 nm for all batches. Tapered double-
cantilever beam specimens were cast from EPON 828
epoxy resin (DGEBA) and 12 pph Ancamine DETA
(diethylenetriamine) curing agent with a prescribed
concentration of microcapsules mixed into the resin.
The epoxy mixture was degassed, poured into a closed
silicone rubber mold and cured for 24 h at room tem-
perature, followed by 24 h at 30 C. Relevant physical
and mechanical properties of the microcapsules and neat
epoxy are listed in Table 1. The tensile modulus and
mode I critical stress intensity factor, K
IC
, of the
microcapsule toughened epoxy were measured as a
function of capsule concentration by Brown et al. [15,
20] and Rzeszutko et al. [21] and summarized in Table 2.
Mechanical testing
The fatigue-crack propagation behavior of the micro-
capsule-modied epoxy was investigated using the
tapered double-cantilever beam (TDCB) specimen
shown in Fig. 1. Side grooves ensured controlled crack
growth along the centerline of the brittle specimen.
The premise of the TDCB geometry, developed by
Mostovoy et al. [22], was to provide a crack-length-
independent
relationship
between
applied
stress
intensity factor K
I
and load P,
K
I
¼ aP;
ð2Þ
which only required knowledge of the coefcient a.
While no single TDCB geometry has been adopted as a
standard, the literature contains numerous TDCB
designs, most of which attempt to optimize the constant
K region of the sample (see for examples [2326]).
Recently Mostovoy [27] introduced a new TDCB
geometry and showed that for crack lengths of a/W (i.e.
the crack length normalized by the length of the spec-
imen) between 0.25 and 0.50 the error in this new
sample design is 3%, compared to 15% for the original
specimen [22]. However, more accurate TDCB designs
are available in the literature and we adopted the
geometry developed for fatigue crack growth experi-
ments by Beres et al. [26]. Beres presented a compre-
hensive 3D FEM study to determine the divergence in
Table 1 Properties of the constituent materials [15]
Properties
Epoxy
Urea-formaldehyde
microcapsules
Density (kg/m
3
)
1160
~1000
Diameter (lm) 50 ± 20
180 ± 40
460 ± 80
Wall thickness (nm) 190±30
K
IC
(MPa m
1/2
)
0.55 ± 0.04 Youngs modulus (GPa)
3.4 ± 0.1 Ultimate stress (MPa)
39 ± 4 J Mater Sci (2006) 41:62666273
6267
123 K with increasing crack length and included the effect
of varying side grooves. For crack lengths of a/W be-
tween 0.23 and 0.48 Beress geometry has an error of
less than 0.5% [26] which is supported by our own FEA,
quasistatic compliance measurements, and constant
crack growth rate fatigue experiments (methods and
results presented in [15, 17, 20, 28]).
The sample in Fig. 1 is a modied version of the
TDCB geometry developed and veried by Beres et al.
[26], for which a = 11.2 · 10
3
m
)3/2
was determined
experimentally [17]. An extensive analysis of the TDCB
geometry is provided in Brown [20]. Comparison of the
current sample with Mostovoys original reveals several
improvements in the current design, most importantly
the reduction of edge effects. A constant range of
Mode-I stress intensity factor DK
I
was achieved by
applying a constant range of load DP, independent of
crack length. The constant-K region of the TDCB
specimen enables crack-growth-rate measurements
over a range of cycles, rather than requiring use of the
modied secant formulation commonly employed for
changing DK
I
of a compact tension specimen [14].
Moreover, the constant-K region is of great importance
for observing the time-dependent effects of self-healing
during growth of a fatigue-crack [20].
Fatigue crack propagation studies were performed
using an Instron Dyno