You are on page 1of 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271965590

Generation of Synthetic FRP Microstructures Based on Experimentally Observed


Microstructures

Conference Paper · March 2015


DOI: 10.1002/9781119093466.ch126

CITATIONS READS
2 319

2 authors:

Seyed Hamid Reza Sanei Ray S Fertig III


Penn State University, Pennsylvania, USA University of Wyoming
32 PUBLICATIONS   293 CITATIONS    73 PUBLICATIONS   638 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Effect of Microstructure on the Heat Transfer of Nanofluids View project

Aeroelastic Optimization for Improved Durability View project

All content following this page was uploaded by Seyed Hamid Reza Sanei on 19 November 2017.

The user has requested enhancement of the downloaded file.


GENERATION OF SYNTHETIC FRP MICROSTRUCTURES BASED ON
EXPERIMENTALLY OBSERVED MICROSTRUCTURES

Seyed Hamid Reza Sanei, Ray S. Fertig, III

Dept. of Mechanical Engineering, University of Wyoming, Laramie, WY, 82071, USA

Keywords: FRP, Synthetic microstructures, Microstructural features

Abstract

One of the key defects in composite materials is the large variability in mechanical properties. To
capture the variability of strength in FRPs, random microstructures have to be analyzed.
Developing a realistic model for generation of random microstructures required first imaging a
carbon reinforced epoxy and then quantifying prominent microstructural features.
Microstructures were synthetically generated including experimentally observed microstructural
features such as elliptical fibers, alignment fibers, voids, and resin seams. Material periodicity of
microstructures was considered to facilitate the application of displacement periodic boundary
condition for later finite element analysis.

Introduction

The high strength to weight ratio of fiber reinforced polymers (FRPs) and their ability to produce
properties tailored for specific applications have been the driving forces in their widespread
applications. But despite the merits associated with FRPs, the large variability in their
mechanical properties[1] have hindered their optimal utilization. Garnich et al. demonstrated
how random fiber microstructural variation could lead to fatigue life variability of more than an
order of magnitude [2]. Models of ideal microstructures can successfully predict average
behavior of composite materials, but for prediction of scatter in properties, microstructural
variability need to be considered. Characterizing actual rather than ideal microstructures is a
necessary step but not sufficient because for a fixed region size of interest, each microstructure is
different from its adjacent microstructures. Therefore, random microstructures need to be
analyzed in stochastic modeling of composite materials.
Image analysis techniques have been implemented for quantification of FRP microstructures [3-
12] and several studies have been conducted to generate synthetic microstructures [2, 13-19], but
random microstructures are mostly generated based on ideal circular fibers with a single radius
and potential ellipsoidal nature of fibers is neglected. Several recent studies have focused on
linking the findings of image analysis of actual microstructures to synthetic generation of
microstructures. Vajari et al. modeled synthetic microstructures including inter-fiber and matrix
voids[17]. They studied the location and geometry of voids on transverse response. Willot et al.

1
generated 3D systems of non-overlapping, wavy fibrous composites based on a specific
orientation distributions. They optimized their model by modifying fiber radius, orientation and
length based on statistical parameters of actual microstructures [18]. Melro et al. proposed an
algorithm for generating random distribution of fibers based on actual fiber distribution in the
transverse direction. They investigated the influence of fiber radius on effective engineering
properties [16]. While generation of voids have been the focus of few studies [17, 20], the
synthetic generation of microstructures including resin seam and alignment fibers has never been
conducted, however, our experimental results have shown that such defects play significant role
in the structural strength.
In this study, actual microstructures are statistically characterized and used as a model for
generation of synthetic microstructures. Scanning electron microscope (SEM) images of carbon
reinforced epoxy were taken and prominent features such as non-uniformity of fiber distribution,
elliptical geometry of fibers, presence of resin seams, voids and alignment fibers were quantified.
The findings are used for synthetic generation of microstructures.

Microstructural characterization and synthetic microstructures

To quantify actual microstructural features, the cross sectional image of a carbon reinforced
epoxy with fibers perpendicular to the surface was imaged and the prominent features were
characterized. Microstructural features such as void, alignment fiber and resin seams are depicted
in Fig.1. Alignment fibers are thermoplastic fibers that are used in prepreg sheet to avoid
distortion, orienting the plies in stacking a laminate and to aid air removal during the curing
process to prevent the formation of air pockets.

As it can be observed from Fig.2, the morphology of fibers is far from a perfect circular shape
and an ellipse better describes the geometry of fibers. To quantify fiber geometry, small, and
large fiber radii of 168 fibers were individually measured using imdistline command of
MATLAB which gives the distance between two points in an image to 0.01 of a pixel. To
minimize the measuring error, only fibers in focus with clear boundaries were chosen to be
measured. The obtained distributions show that a Weibull distribution best fits the small radius
distribution and a Beta distribution best fit the ratio of small to large fiber radius. The
distributions are shown in Figs 3 & 4, respectively.

2
Alignment
fiber

Void
Resin
seams

Fig.1. Scanning electron image of FRP Fig.2. Scanning electron image of FRP
microstructure illustrating alignment fiber, void microstructure illustrating non circular geometry
and resin seams. of fibers.

0.9 Actual distribution


Actual distribution
Normal
0.8 6 Beta Distribution
LogNormal
0.7 Weibull
5
Probability Density

Probability Density

0.6
4
0.5

0.4 3

0.3
2
0.2
1
0.1

0 0
5 5.5 6 6.5 7 0.7 0.8 0.9 1
Small Fiber Diameter Large to small fiber radii ratio

Fig.3. Small fiber radius distribution Fig.4. Distributions of Small to large


of carbon fibers fitted with Weibull fiber radii ratio of carbon fibers fitted
distribution. by Beta distribution.
3
Prior to generation of fibers, the image dimension is specified, and the number of pixels
associated with the image size is determined according to the prescribed resolution. The fiber
centers are randomly generated such that the fiber centers can have coordinates one radius
beyond the edges of the image to model the incomplete fibers on the edges to ensure periodicity
of the structure for later FE simulation.
In generation of elliptical fibers, as large and small
fiber radii are correlated, therefore, they must not be
assigned independently. Hence, small fiber radius is
randomly selected from a Weibull distribution and the
large radius is determined as function of small radius
divided by a radius ratio picked from a Beta
distribution. The tails of the distributions were
truncated by setting minimum and maximum values
identical to values obtained from actual fiber geometry
to avoid generation of unrealistic fiber configurations. Fig.5. Schematic geometry of
In addition to fiber radii and the center locations, the fibers where 𝑅1 and 𝑅2 are large
fiber cross sectional angles are randomly assigned. The and small fiber radii, respectively
schematic configuration of a considered elliptical fiber
and 𝛼 is fiber cross section angle.
is depicted in Fig.5. Each generated fiber is compared
with existing fibers to avoid generation of overlapping
fibers.

The material periodicity of the microstructure which facilitates the application of displacement
periodic boundary conditions in simulation is implemented such that, a condition is set so as,
upon generation of fibers on an edge, a new fiber is generated on the opposite edge to enforce
periodicity. An image of synthetic microstructure with material periodicity is illustrated in Fig.6.

Fig.6. Synthetic microstructures of Fig.7. Synthetic microstructures


elliptical fibers with material including microstructural features such
periodicity. as alignment fiber, void and matrix
seams. 4
Common FRP’s defects such as resin seam, alignment fiber and voids are also generated. They
are modeled as ellipses with random locations and angles. Their radii is selected based on
quantification of actual microstructures. Fiber volume fraction is evaluated at each step and fiber
generation is terminated when the fiber volume fraction reaches the prescribed value. The
generated microstructure including all features is shown in Fig. 7. Green ellipses denote fibers,
black background is the matric phase, blue and red ellipses are the alignment fiber and void,
respectively. Figure 8 depicts two synthetic microstructures in comparison to actual
microstructure to confirm the accuracy of generated microstructures.

Fig.8. Comparison of actual microstructure (left) with two synthetic microstructures (middle and
right).

The developed method for generation of synthetic microstructure is a powerful tool for
investigation of various microstructural defects on the structural reliability and strength. A
parametric study on the effect of void size, location and morphology on the strength under
various loading conditions can be conducted. The effect of resin seams on the overall strength of
the structure and their interaction with other defects can be an interesting topic of future studies.
Furthermore, to strengthen the model, the correlation between the occurrences of defects need to
be considered, for instance, voids are prone to form near alignment fibers.

5
References

1. Lekou, D.J. and T.P. Philippidis, Mechanical property variability in FRP laminates and
its effect on failure prediction. Composites Part B: Engineering, 2008. 39(7-8): p. 1247-
1256.
2. Garnich, M.R., R.S.F. III, and E.M. Anderson, Random Fiber Micromechanics of Fatigue
Damage, in 54th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and
Materials Conference. 2013, American Institute of Aeronautics and Astronautics.
3. Cann, M.T., D.O. Adams, and C.L. Schneider, Characterization of Fiber Volume
Fraction Gradients in Composite Laminates. Journal of Composite Materials, 2008.
4. Clarke, a.R., G. Archenhold, and N.C. Davidson, A novel technique for determining the
3D spatial distribution of glass fibres in polymer composites. Composites Science and
Technology, 1995. 55(1): p. 75-91.
5. Davidson, N.C., A.R. Clarke, and G. Archenhold, Large-area , high-resolution image
analysis of composite materials. 1997. 185(February): p. 233-242.
6. Guild, F.J. and J. Summerscales, Microstructural image analysis applied to fibre
composite materials: a review. Composites, 1993. 24(5): p. 383-393.
7. Santulli, C., A.C. Long, and M.J. Clifford, Void content measurements in commingled E-
glass/polypropylene composites using image analysis from optical micrographs Science
and engineering of composite materials, 2002. 10(2): p. 77-90.
8. Shi, D. and D. Winslow, Accuracy of a Volume Fraction Measurement Using Areal
Image Analysis. American Society for testing and materials 1990: p. 210-213.
9. Waterbury, M.C. and L.T. Drzal, Determination of Fiber Volume Fractions by Optical
Numeric Volume Fraction Analysis. Journal of Reinforced Plastics and Composites,
1989. 8(6): p. 627-636.
10. Wimolkiatisak, A.S., et al., Assessment of fiber arrangement and contiguity in composite
materials by image analysis. Polymer Composites, 1990. 11(5): p. 274-279.
11. Yang, H. and J.S. Colton, Quantitative Image Processing Analysis of Composite
Materials Polymer Composites, 1994. 15(1): p. 46-54.
12. Fertig, R.S., E.M. Jensen, and K.A. Malusare. Effect of Fiber Volume Fraction Variation
Across Multiple Length Scales on Composite Stress Variation : The Possibility of
Stochastic Multiscale Analysis. in American Institute of Aeronautics and Astronautics
2014.
13. Wongsto, a. and S. Li, Micromechanical FE analysis of UD fibre-reinforced composites
with fibres distributed at random over the transverse cross-section. Composites Part A:
Applied Science and Manufacturing, 2005. 36(9): p. 1246-1266.
14. Huang, Y., K.K. Jin, and S.K. Ha, Effects of Fiber Arrangement on Mechanical Behavior
of Unidirectional Composites. Journal of Composite Materials, 2008. 42(18): p. 1851-
1871.
15. González, C. and J. Llorca, Mechanical behavior of unidirectional fiber-reinforced
polymers under transverse compression: Microscopic mechanisms and modeling.
Composites Science and Technology, 2007. 67(13): p. 2795-2806.

6
16. Melro, a.R., P.P. Camanho, and S.T. Pinho, Influence of geometrical parameters on the
elastic response of unidirectional composite materials. Composite Structures, 2012.
94(11): p. 3223-3231.
17. Ashouri Vajari, D., et al., A numerical study of the influence of microvoids in the
transverse mechanical response of unidirectional composites. Composites Science and
Technology, 2014. 97: p. 46-54.
18. Altendorf, H., D. Jeulin, and F. Willot, Influence of the fiber geometry on the
macroscopic elastic and thermal properties. International Journal of Solids and
Structures, 2014. 51(23-24): p. 3807-3822.
19. Garnich, M.R., R.S.F. Iii, and E.M. Anderson, Random Fiber Micromechanics of Fatigue
Damage. 2013: p. 1-9.
20. Huang, H. and R. Talreja, Effects of void geometry on elastic properties of unidirectional
fiber reinforced composites. Composites Science and Technology, 2005. 65(13): p. 1964-
1981.

View publication stats

You might also like