You are on page 1of 2

10.1002/spepro.

003973
Better predictions of airfoil
strength
Yongbo Zhang, Huimin Fu, and Zhihua Wang
Combining the ber microbuckling and kinking models improves un-
derstanding of longitudinal compressive strength for unidirectional
laminated composites commonly used in the aerospace industry.
With the signicantly increasing use of carbon-ber-reinforced poly-
mer composite structures for load-bearing applications such as air-
foils, understanding the fundamental aspects of mechanical failure of
these structures is greatly needed. In most cases, the weakness under
compression of brous composites severely limits the structural ef-
ciency of the system and leads to under-utilization of the true material
properties. However, the mechanisms of compressive failure are not
well understood, and compressive strength is difcult to predict and
measure. The reason is that fracture of composites is usually instan-
taneous and catastrophic, and the buckling, in conjunction with com-
pressive loading, makes the structures highly sensitive to imperfections.
Therefore, there is much work to do for a clear understanding of the
compressive failure behavior and accurate prediction of the compres-
sive strength. We conducted a compression test of unidirectional (UD)
T300/QY8911 composite laminatesT300 is a high-tensile-strength
carbon ber and QY8911 is a bismaleimide resin made in China
and observed the post-failure surfaces of the specimens under scanning
electron microscopy to analyze the failure mechanism and determine
mechanical properties.
1
The most frequently considered failure modes in UD laminates are
microbuckling and kinking. Microbuckling is the buckling of bers
embedded in matrix foundation. For high-volume-fraction composites,
microbuckling is expected to be controlled by the matrix stiffness in
shear. It is strongly affected by the initial imperfections of the compos-
ite that were introduced during the manufacturing process including
associated defects such as ber misalignment, rich resin, and porosity.
Kinking is a term used to describe the deformation that is localized in a
band across the specimen. In the band, bers have signicantly rotated
and the matrix has undergone large shearing deformation. Important
parameters are the kink band width, w, the angle of the rotated ber, ,
and the angle of the kink band.
Previous studies considered kinking as a failure mechanism inde-
pendent of microbuckling, but our view is that kinking is the nal
Figure 1. (a) Typical overall failure mode of a UD T300/QY8911 com-
posite specimen and (b) a scanning electron micrograph illustrating the
failure mechanisms. : Angle of the kink band. EHT: Extra high ten-
sion. WD: Working distance. EHT and WD are the parameters of the
scanning electron microscope.
consequence of microbuckling; large bending stresses in the mi-
crobuckled ber lead to ber fracture. Figure 1 is a scanning electron
microscope image of the post-failure surface of a specimen. It suggests
that kinking is likely triggered by the local instability of bers embed-
ded in the matrix, which may be nucleated locally by ber waviness,
free edge region, resin rich region, and poor ber-matrix interfacial
bonding. This local instability leads to a locally initiated failure propa-
gating under incremental load through the laminate and thus creates a
kink band that nally causes the collapsing of the specimen.
According to our analysis of the failure mechanism, UD
T300/QY8911 composites fail due to microbuckling followed by
kinking. Therefore, we propose a combined mode model to predict
the compressive strength, based on the microbuckling model rst sug-
gested by Xu and Reifsnider
2
and the kinking model developed by
Budiansky.
3
We use our model to determine the fundamental stiffness
in the microbuckling model, thus achieving a new expression for the
compressive strength calculation, which considers both microbuckling
Continued on next page
10.1002/spepro.003973 Page 2/2
Figure 2. Predicted compressive strength using different models
compared with the measured result for different unidirectional com-
posite laminates.
4
HTS40 and IM7 are bers, and 977-2 and 8551-7
are resins.
5, 6
strength and kinking strength. It can be written as
o
c
D (1
f1
V
f
C1
m
(1 V
f
))
2
4
0.45

j1
m
r
f

p

1
m
V
f
C1
f1
(1 V
f
)
!
2=5
CG
m
3
5
C
t
n
t
y
;
n
;
y
C
0
(1)
where 1
f1
is the longitudinal Youngs modulus of the ber, V
f
is the
ber volume fraction, j represents the ber-matrix bonding conditions,
1
m
is Youngs modulus of the matrix, G
m
is the matrix shear modulus,
r
f
is the ber diameter, t
n
is the nal in-plane shear strength, t
y
is the
in-plane shear yield strength, ;
n
is the nal in-plane shear strain at
shear strength, ;
y
is the in-plane shear strain at yield, and
0
is the
initial ber misalignment.
We compared the compressive strengths predicted by three models
and the values determined experimentally for T300/QY8911 and also
HTS40/977-2, IM7/977-2, and IM7/8551-7, which are other UD com-
posite systems (see Figure 2). HTS40 and IM7 are bers and 977-2
and 8551-7 are resins. We found that the microbuckling model predicts
the lowest compressive strength of the three models, and its predic-
tions are generally 30% lower than the measured result. The reason is
that this model does not take plastic deformation into account. When
using the kinking model instead, the predictions for T300/QY8911
and HTS40/977-2 composites are generally 20% higher than the ex-
perimental data, and the predictions for IM7/977-2 (a graphite/epoxy
system) and IM7/8551-7 are generally 25% lower than the measured
result. However, when we use the combined model to predict the com-
pressive strength, the predictions show good agreement with the exper-
imental measurements. Except for IM7/8551-7, where the prediction
and measurement differ by 6.4%, differences between theoretical pre-
dictions and measurement for the other composite systems are less than
3%. We conclude that our combined model is suitable for predicting the
compressive strength of these four kinds of UD composite systems.
In summary, we have proposed and tested a combined model for lon-
gitudinal compressive strength for UD laminated composites. A novel
formula provides an efcient way to predict the compressive strength
prediction of UD composites laminates. As a next step, we will apply
our approach to estimate more complicated laminated structures.
Author Information
Yongbo Zhang, Huimin Fu, and Zhihua Wang
Department of Materials Science and Engineering
Beijing University of Aeronautics and Astronautics
Beijing, China
Yongbo Zhang is a postdoctoral fellow. He received his PhD in me-
chanical engineering at the Beijing University of Aeronautics and As-
tronautics, China.
References
1. Y. Zhang and H. Fu, On the longitudinal compressive strength prediction of unidi-
rectional laminated composites based on an improved model, Polym. Compos. 32,
pp. 18171826, 2011. doi:10.1002/pc.21214
2. Y. L. Xu and K. L. Reifsnider, Micromechanical modeling of composite compressive
strength, J. Compos. Mater. 27, pp. 57288, 1993.
3. B. Budiansky, Micromechanics, Comput. Struct. 16, pp. 312, 1983.
4. A. Jumahat, C. Soutis, F. R. Jones, and A. Hodzic, Fracture mechanisms and failure
analysis of carbon bre/toughened epoxy composite subjected to compressive loading,
Compos. Struct. 92, pp. 295305, 2010.
5. Resin product supplier. http://www.cytec.com. Accessed 24 November 2011
6. Carbon ber and composites supplier. http://www.hexcel.com. Accessed 24 November
2011.
c 2011 Society of Plastics Engineers (SPE)

You might also like