Академический Документы
Профессиональный Документы
Культура Документы
entitled
Negative Poissons Ratio Composites - Finite Element Modeling and Experiments
by
Sharmila Jayanty
Submitted to the Graduate Faculty as partial fulfillment of the requirements for the
Master of Science degree in Mechanical Engineering
_______________________________________
Dr. Lesley M. Berhan, Committee Chair
________________________________________
Dr. Maria Coleman , Committee Member
________________________________________
Dr. Yong X. Gan, Committee Member
________________________________________
Dr. Patricia R. Komuniecki, Dean
College of Graduate Studies
An Abstract of
Negative Poissons Ratio Composites - Finite Element Modeling and Experiments
by
Sharmila Jayanty
Submitted to the Graduate Faculty in partial fulfillment of the requirement for the
Master of Science Degree in Mechanical Engineering
The University of Toledo
December 2010
Auxetic or negative Poissons ratio materials are of great research interest as they
exhibit improved mechanical properties such as impact resistance, fracture toughness,
hardness and shear modulus over conventional materials with a positive Poissons ratio
and same stiffness. Most auxetic materials manufactured to date are porous, and the
major hindrance of using these materials in structural applications is their low strength. In
this research a new approach for manufacturing auxetic fiber-reinforced composites by
embedding an auxetic fibrous network in a conventional polymer matrix is investigated
using numerical and experimental methods.
iii
In this project, a finite element model was developed to investigate the proposed
hypothesis of manufacturing an auxetic composite. Compressed mats of sintered stainless
steel fibers are known to have negative Poissons ratio out-of-plane. It is hypothesized
that their auxetic behavior is from the arrangement of fibers within the matrix, relative
stiffness of the matrix and network of reinforcement. First, tension tests were performed
on the compressed sintered mats and the experimental results have shown that Poissons
ratio for all metal mat specimens is negative in transverse direction. Next, mechanical
tests were performed and microscopic images were taken of the composite samples made
by embedding the stainless steel mats with silanol terminated polydimethoxysiloxane
(PDMS) polymer. Finally, polymer nanocomposites were made through the inclusion of
high loading carbon nanofibers in the polymer matrix (PDMS). The experimental results
of both the composite mats and nanocomposites confirm the feasibility of manufacturing
an auxetic composite by embedding an auxetic network in a conventional polymer. The
auxetic behavior in the composite mats is due to the arrangement of fibers in the network
and compression of stainless steel mats during fabrication process. It was assumed that
the auxetic behavior of the nanocomposites is due to effective network formation of
nanofibers and the compression of the sample during preparation process.
iv
Acknowledgements
I would like to express my sincere gratitude to my advisor, Dr. Lesley M. Berhan,
for her invaluable support, guidance and encouragement during the course of this thesis
project and throughout my study at The University of Toledo. Her moral support and
continuous guidance enabled me to complete my work successfully. I am thankful to her
for giving me the opportunity to work in this project.
I would like to thank my committee members Dr. Maria R. Coleman and Dr.
Yong Gan, for their valuable suggestions and support throughout this study. I sincerely
thank Mehmet Tatlier and Jason Crowe for their help in the research lab. I truly
appreciate all the time and suggestions given by them. I would like to specially
acknowledge Nima Hakim-elahis for his help in membrane materials characterization
lab.
I sincerely thank all the faculty members in the Department of Mechanical,
Industrial and Manufacturing Engineering for their help during my Masters program. I
would also like to express my gratitude to all the staff members at the College of
Engineering. I would like to thank all my lab mates Azita Belashi, Julie Matheny, Greg
Tilton and Carrie Baker. I would like to thank my roommates and all my friends for their
help and support during my stay in Toledo.
vi
Contents
ix
List of Tables
Table 2.1: Different Cell Geometries................................................................................ 15
Table 2.2: Effect of strain rate on Poissons ratio ............................................................. 19
Table 2.3: Effect of Youngs modulus on Poissons ratio ................................................ 21
Table 3.1: Average values of Youngs modulus and Poissons ratio for sintered mats ... 34
Table 4.1: Mixture of chemicals ....................................................................................... 38
Table 4.2: Weight fraction of fiber and polymer in composite mat for different porosities
........................................................................................................................................... 51
Table 4.3: Average values of Youngs modulus and Poissons ratio for plain PDMS ..... 51
Table 4.4: Comparison of Youngs modulus and Poissons ratio for composite and
sintered mats ..................................................................................................................... 53
Table 4.5: Variation in thickness for sintered and composite mats .................................. 53
Table 5.1: Fiber loading ratios .......................................................................................... 63
Table 5.2: Youngs modulus and Poissons ratio for 15% fiber loading of pristine fibers
........................................................................................................................................... 68
Table 5.3: Youngs modulus and Poissons ratio for 20% fiber loading of pristine fibers
........................................................................................................................................... 68
Table 5.4: Youngs modulus and Poissons ratio for 25% fiber loading of pristine fibers
........................................................................................................................................... 69
Table 5.5: Youngs modulus and Poissons ratio for 15% fiber loading of oxidized fibers
........................................................................................................................................... 69
Table 5.6: Youngs modulus and Poissons ratio for 20% fiber loading of oxidized fibers
........................................................................................................................................... 70
Table 5.7: Youngs modulus and Poissons ratio for 25% fiber loading of oxidized fibers
........................................................................................................................................... 70
Table 5.8: Youngs modulus and Poissons ratio for 20% fiber loading of functionalized
fibers ................................................................................................................................. 71
Table 5.9: Volume fraction of fiber and polymer for 15% fiber loading ......................... 75
Table 5.10: Volume fraction of fiber and polymer for 20% fiber loading ....................... 75
Table 5.11: Volume fraction of fiber and polymer for 25% fiber loading ....................... 75
xi
List of Figures
Figure 1-1: Schematic diagram of positive Poissons ratio deformation when a tensile
load is applied in the longitudinal direction, initially undeformed material (thin line)
undergoes longitudinal extension and lateral contraction (solid line) ................................ 2
Figure 1-2: Schematic diagram of negative Poissons ratio deformation when a tensile
load is applied in the longitudinal direction, initially undeformed material (thin line)
undergoes longitudinal and lateral extension (solid line) ................................................... 2
Figure 1-3: Deformation mechanisms in re-entrant and conventional honeycomb ............ 7
Figure 1-4: Pore structures of non auxetic and auxetic foam ............................................. 8
Figure 2-1: Fiber network model similar to random fiber network model before
compression ...................................................................................................................... 13
Figure 2-2: Fiber network model similar to random fiber network model after
compression ..................................................................................................................... 13
Figure 2-3: Dimensions of a re-entrant honey comb structure (unit cell)......................... 15
Figure 2-4: Boundary conditions for plane stress and beam elements ............................. 16
Figure 2-5: Deformed fibrous network model .................................................................. 17
Figure 2-6: Poissons ratio versus strain for re-entrant honeycomb network ................... 20
Figure 2-7: Poissons ratio versus effective stiffness ratio for re-entrant honeycomb ..... 21
Figure 2-8: Compliant matrix ........................................................................................... 23
Figure 2-9: Stiff matrix ..................................................................................................... 24
xii
Figure 4-14: Lateral strain vs. longitudinal strain graphs for composite mats ................ 54
Figure 5-1: Sonication process.......................................................................................... 60
Figure 5-2: Experimental setup for purification of oxidized carbon nanofibers .............. 61
Figure 5-3: Filtration process ............................................................................................ 62
Figure 5-4:: Schematic procedure for nanocomposite preparation ................................... 65
Figure 5-5: OCNF- 2 roll mill procedure (20% fiber loading) ......................................... 66
Figure 5-6: OCNF- manual press with weight during curing procedure (20% fiber
loading) ............................................................................................................................. 66
Figure 5-7: OCNF- manual press with weight during curing procedure(20% fiber
loading) ............................................................................................................................. 67
Figure 5-8: Stress vs. strain graphs for different nanocomposites of 20% fiber loading.. 72
Figure 5-9: Stress vs. strain graphs for different nanocomposites of 20% fiber loading
(elastic range) .................................................................................................................... 72
Figure 5-10: Lateral strain vs. longitudinal strain graphs for nanocomposite samples at
20% fiber loading .............................................................................................................. 73
Figure 5-11: Lateral strain vs. longitudinal strain graphs for different fiber loading of
pristine fibers .................................................................................................................... 74
Figure 5-12: Stress vs. strain curves for nanocomposite samples with pristine nanofibers
........................................................................................................................................... 74
Figure 5-13: Lateral strain vs. longitudinal strain for 20% fiber loading of pristine fibers
........................................................................................................................................... 77
xiv
List of Symbols
Poissons ratio
x ..Transverse strain
y ...Longitudinal strain
E.Youngs modulus
K.Bulk modulus
H.Hardness
P..Load
A0.Initial cross section
l ...Change in length
l0.Initial Length
.........Stress
t .....Change in thickness
t0...Initial thickness
xv
Chapter 1
Introduction
xy
x
(x / x)
,
y
(y / y )
(1)
where
= Poissons ratio
x = Transverse strain
1
y = Longitudinal strain
In Equation (1), there is a negative sign so that the Poissons ratio is positive as most of
the materials contract in the transverse direction due to an applied tensile load. Figure 1-1
shows a schematic diagram of positive Poissons ratio materials.
Lakes first synthesized auxetic polyurethane (PU) foams with re-entrant structures
in 1987 [1]. Later Evens designated these negative Poissons ratio materials as
Auxetics in 1991 [2]. Negative Poissons ratio materials are also known as anti-rubber
because when stretched they become fatter in cross-sectional area, while rubber becomes
thinner [3]. These materials are also known as dilatational materials as they undergo
volumetric change instead of shape change [4].
Generally for an isotropic material, the four elastic constants used to describe the
materials properties are Youngs modulus (E), Shear modulus (G), Bulk modulus
(K) and Poissons ratio (). The first three constants are a measure of stiffness,
rigidity and compressibility of a material. The equations relating these four
constants are given as follows [9].
E
2(1 v)
(2)
E
31 2
(3)
According to Equations (2) and (3), its clear that when Poissons ratio
approaches -1, the shear modulus approaches infinity. Therefore a very high shear
3
E
,
H
2
1
(4)
2
for
3
Auxetic materials also show remarkable damping and absorption properties. The
damping capacity of high compressive pre-strained auxetic foam is greater than
that of the conventional foam. This effect is due to the abrasion inwardly buckled
cell ribs of the auxetic foam [10]. It is reported that the absorption of sound at all
frequencies by auxetic foam is superior to conventional foams [11, 12].
Auxetic materials have been proposed for potential military applications like
bullet-proof or projectile resistant vests, body armors, crash helmets, and
bandages or wound pressure pads [16].
Auxetic materials could be used for the construction of wear resistant machine
components as they have improved hardness. For instance press-fit fasteners with
negative Poissons ratio during compression, undergo lateral contraction which
assists the insertion of the fastener [17, 18].
Two dimensional silicon rubber or aluminum honey combs were the large scale
auxetic cellular structures first developed in 1982. The deformation mechanism in these
structures is shown in Figure 1-3 which is due to flexure of diagonal ribs under applied
tensile loads [25, 26]. These honeycomb structures are used in filtration media as they
sustain pressure across the filter when load is applied. However, to take full advantage of
these structures the cell size must be reduced below 1mm. Techniques used for this
purpose are molecular self-assembly [27], LIGA technology, silicon surface
micromachining techniques [28], laser stereo lithography [27] and nanomaterials
fabrication processes [29]. Honeycomb structures reduced to a scale of 50m have
potential applications in nano-technology and MEMS [30].
angle-ply reinforced laminates and the development of composites which have one or
more auxetic phases are the two primary techniques used in the design of auxetic
composites.
One approach to manufacturing auxetic composites by laminated angle ply
method was proposed by G. Milton in 1992 [45]. Later this approach was proposed and
validated through experiments by stacking unidirectional layers of carbon fiber reinforced
epoxy in a special sequence which gives rise to negative Poissons ratio in-plane and outof-plane direction [46-49]. This type of development has predicted an improvement in
mechanical properties such as static indentation resistance and low velocity impact
testing [50].
A second approach to manufacturing auxetic composites is using one or more
auxetic components as reinforcement material or the matrix. For instance, re-entrant
copper foam was used as matrix material with viscoelastic polymer, solder and, indium as
filler material in the fabrication of auxetic composite [51]. Recent developments made in
using different auxetic fibers as reinforcement in conventional matrices motivated interest
in fiber reinforced composites [ 52,53,54]. Use of auxetic fibers is proposed to enhance
pull-out performance as these auxetic fibers maintain an interface between the matrix and
the fiber. Experimental results reported that the energy used to pull out the auxetic fiber is
three times greater than a conventional fiber.
1.5 Objectives
Deformation mechanisms in re-entrant honey comb structures, auxetic foams,
liquid crystal polymers are well understood because of their common periodic micro
10
structures but there is limited knowledge of how the auxetic effect arises in non-periodic
microstructures. In order to understand the auxetic effect in these disordered and random
fiber networks, linear 2D and 3D fiber network models were developed by Tatlier and
Berhan [55]. They investigated the effect of compression ratio, fiber density, anisotropy
of fiber mats, applied strain, and aspect ratio on mechanical properties such as Poissons
ratio and effective modulus. Simulation results showed that compression of random fiber
networks is one of the key parameters in determination of auxetic behavior.
The goal of the project is divided into two parts. First, to experimentally validate
the numerical model developed by Tatlier and Berhan by conducting tensile tests on
compressed mats of sintered stainless steel fibers which are known to exhibit negative
Poissons ratio out-of-plane. Second, to test the hypothesis whether an auxetic composite
can be manufactured by embedding a random fibrous network with a negative Poissons
ratio in a conventional matrix material provided the modulus of the reinforcing fibers is
sufficiently high compared to the matrix material. To validate the hypothesis, two
dimensional finite element analysis and experiments on composite mats (sintered
stainless steel mats as random fibers and PDMS polymer as conventional matrix) and
nanocomposites (high loading carbon nanofibers in PDMS polymer matrix) were
performed.
11
Chapter 2
Mechanism of the auxetic behavior in fibrous network was studied by Tatlier and
Berhan [56] which showed resemblance between re-entrant honey comb or inverted cell
structures and random fiber networks. Their study proposed that compression is the
critical factor which resulted in similarities between the deformation mechanism of reentrant honeycomb structure and random fiber networks. Figure 2-1 shows the fiber
network model which is similar to random fiber network before compression where the
contact points are considered to be rigid and Figure 2-2 shows fiber network model which
is similar to random fiber network after compression. From these figures it is clear that
compression on the random fiber mats resulted in a geometrical structure equivalent to
geometry of re-entrant honeycombs.
12
Figure 2-1: Fiber network model similar to random fiber network model before
compression[56]
Figure 2-2: Fiber network model similar to random fiber network model after
compression[56]
13
3.8
9.2
7.8
9.2
12.2
10.6
3.2
9.6
8.4
0.6
1.2
0.6
1.2
0.6
0.4
0.8
0.4
1.2
1.2
0.2
0.8
0.2
0.8
0.6
0.6
15
Figure 2-4: Boundary conditions for plane stress and beam elements
16
18
Strain
Poisson's ratio
0.001
0.0001
0.12
0.005
0.0006
0.12
0.009
0.0011
0.12
0.03
0.0037
0.12
0.07
0.0087
0.12
0.1
0.0125
0.12
0.5
0.0625
0.12
0.9
0.1125
0.12
0.125
0.12
1.5
0.1875
0.12
1.9
0.2375
0.12
19
Figure 2-6: Poissons ratio versus strain for re-entrant honeycomb network
From the above results, it can be observed that the variation in strain i.e. displacement on
the top edge, does not affect the Poissons ratio. It was observed that Poissons ratio was
not affected significantly either by change in cell geometry or number of cells for the
various combinations listed in Table 2.1.
20
Poissons ratio
2.97E-08
0.33
2.97E-07
0.3299
2.97E-05
0.3196
2.97E-04
0.127
2.97E-02
-1.3799
2.97E+00
-1.55099
Figure 2-7: Poissons ratio versus effective stiffness ratio for re-entrant honeycomb
A significant change in the Poissons ratio was observed with the variation in
Youngs modulus. The deformation of auxetic network (beam elements embedded in
21
compliant plane stress elements) is shown in Figure 2-8 where the lateral dimension
increases with the application of longitudinal displacement. In Figure 2-9 conventional
network (beam elements were embedded in stiffer plane stress elements) is shown where
the lateral dimension decreases with longitudinal displacement. Stiffness ratio of beam to
plane stress elements for compliant matrix was 5.93, for stiffer matrix the ratio was 5.93
x 10-8. Thus the effective Poissons ratio of the network changes from negative to
positive when the matrix material (plane stress elements) changes from a compliant
matrix to a stiffer matrix as shown in Figure 2-10. The obtained finite element simulation
results validate the hypothesis of manufacturing an auxetic composite by embedding an
auxetic network in a conventional matrix where the auxetic network is much stiffer than
the conventional matrix.
22
23
24
25
Chapter 3
P
A0
26
(7)
Engineering strain ( ) is defined as the ratio of change in length ( l ) to the initial length
of the specimen ( l0 )
l
l0
(8)
27
The specimen was clamped vertically to the top grip and allowed to hang freely
for better alignment and then clamped to the bottom grip. Vertical alignment of the
specimen was important to avoid side loading or bending moments in the specimen. After
mounting the specimen, a small amount of load was applied with the help of jog up
switch in the Instron to ensure that it is stiff enough in between the grips. The laser
extensometer was adjusted horizontally such that the laser beam showed the minimum
thickness of the specimen in the Bluehill2 software.
The initial input parameters were gauge length, width and thicknesses of the
narrow cross sectional area. The gauge length and the width of the specimen were
measured with the help of Vernier calipers (Figure 3-6) and thickness was obtained from
the laser reading. The experiments were controlled at a constant strain rate of 1 mm/min
and the intervals at which the raw data points were collected were fixed by the system
using two channels load and time. The experimental procedure was performed on five
samples for each type of mat.
The raw data file obtained from the software after testing comprises of load,
tensile extension, tensile stress, tensile strain, and thickness value (laser reading).
Youngs modulus and Poissons ratio were calculated with the help of test results.
Youngs modulus was calculated as the slope of the line in the elastic region where stress
is proportional to strain obtained from the raw data files.
E /
(9)
E is the Youngs modulus, is the tensile stress and is the tensile strain. Poissons
ratio is the ratio of lateral strain to longitudinal strain with a negative sign. Since most
32
of the materials contract in the direction perpendicular to the applied load, negative sign
is included in the formula to make sure that the ratio is positive.
3
1
(10)
1 is longitudinal strain in the direction of the applied load and 3 is the lateral strain in
the direction perpendicular to the applied load. Later strain is calculated using laser
extensometer reading. Longitudinal strain is calculated from equation 8. l is the
longitudinal displacement of the specimen obtained from the software and l 0 is the initial
gauge length of the specimen. Lateral strain is calculated as the ratio of change in
thickness to the original thickness.
t
t0
(11)
t is the change in thickness of the specimen calculated from the laser extensometer
33
Table 3.1: Average values of Youngs modulus and Poissons ratio for sintered mats
Sintered mat
Young's modulus
Poissons ratio
(%Porosity)
(MPa)
80
923.51
-5.26
70
1197.12
-7.54
60
1283.52
-17.81
Figure 3-7: Stress vs. strain graphs for sintered mats of different porosities
34
Figure 3-8: Lateral strain vs. longitudinal strain graphs for sintered mats of
different porosities
3.5 Discussion
Sintered stainless steel fiber mats purchased from Bekaert are random metallic
fibrous networks. The porosity of the fiber mat is restrained by the application of pressure
either by vacuum molding or pressurized viscous flow as described in Section 3.1.
Compression of the fibrous mat is related to its porosity, as compression on the mat
increases porosity decreases. The experimental tests conducted by F. Delannay on
sintered mats of compressed steel fibers exhibited negative Poissons ratio out-of-plane
[61, 62]. The auxetic behavior in compressed random fibers was explained by the
deformation mechanism put forward by Tatlier and Berhan as discussed in Chapter 2 i.e.,
compression of random metal fibers results in a geometrical structure equivalent to
geometry of re-entrant honeycombs. Thus compressed sintered stainless steel mats were
35
36
Chapter 4
Composite Mats
Sintered stainless steel mats are highly porous in nature and are known to exhibit
negative Poissons ratio. The main limitation of these mats is their low strength because
of the porosity. Finite element model developed using Comsol Multiphysics discussed in
Chapter 2 validates the hypothesis of manufacturing an auxetic composite by embedding
an auxetic fibrous network with higher stiffness in a conventional matrix with lower
stiffness. To investigate the feasibility of manufacturing the hypothetical composite
model as real materials, experiments were conducted by infusing auxetic network
(compressed sintered stainless steel mats) in a polymer matrix whose stiffness is
significantly lower than that of the auxetic network. DMS- S21 Silanol-terminated
polydimethylsiloxane elastomer was used as the polymer matrix with PSI 021- Poly
(Diethoxysiloxane) as the cross-linker and SND-3260 Di-n-Butyldilauryl Tin as the
catalyst. PDMS is viscoelastic, acts like a viscous liquid at high temperatures, and like an
elastic solid at low temperatures. It is optically clear, and considered to be inert, nontoxic and non-flammable. The chemical formula of PDMS is CH3[Si(CH3)2O]nSi(CH3)3.
37
Oligomer
Cross linker
Type -1
2.7
0.3
38
(a) Monomer
(c) Catalyst
40
41
42
43
44
45
46
47
48
49
spaces in between the fibers decrease which result in decrease in amount of polymer
deposited on the fibrous mats. Deposition of polymer is least on 60% porosity composite
mat and highest on 80% porosity composite mat.
Table 4.2: Weight fraction of fiber and polymer in composite mat for different
porosities
Composite mat
Weight fraction
Porosity
Fiber
Polymer
60%
0.81
0.19
70%
0.77
0.22
80%
0.75
0.24
Table 4.3: Average values of Youngs modulus and Poissons ratio for plain PDMS
Plain PDMS
Young's modulus(MPa)
Poissons ratio
Type-1
5.49
1.12
51
Figure 4-13: Lateral strain vs. longitudinal strain graphs for plain PDMS
52
Mats
Composite Mats
Young's
modulus(MPa)
Composite
Mats Poissons
ratio
Sintered
Mats
Young's
modulus
(MPa)
60%
1371.7
-9.2
923.51
-5.26
70%
1325
-9.2
1197.12
-7.54
80%
1272.8
-4
1283.52
-17.81
Sintered
Mats
Poissons
ratio
Mat porosity
Sintered mat
thickness(mm)
Composite mat
thickness(mm)
Thickness
compressed
(mm)
60%
0.38
0.33
0.05
70%
0.42
0.35
0.07
80%
0.54
0.41
0.13
53
Figure 4-14: Lateral strain vs. longitudinal strain graphs for composite mats
4.3 Discussion
A considerable amount of deposited polymer was seen both from SEM image
analysis and weight measurement analysis as described in previous sections. From
Figure- 4-14 it can be observed that the Poissons ratio remains negative after auxetic
network is embedded in the polymer matrix. This can be associated with the fact that the
stiffness of the plain polymer (of order 5 MPa) as shown in Figure 4-12 is much lower
compared to the sintered metal fiber mat (of order 1100 MPa).
Experimental results from Table 4.4 exhibit a significant increase in stiffness of
the composite mats, formed by embedding an auxetic network in a polymer matrix,
compared to sintered metal mats. This increase of stiffness in composite mats can be
54
attributed to the compression when passed over 2-roll mill but not because of presence of
polymer whose stiffness is much lower than that of fibrous mats. Table 4.5 shows the
variation in thickness of the mat before and after passing through 2-roll mill. From the
table one can observe that 80% porosity composite mats were more compressed when
compared to 70% and 60% because these mats are highly porous and thicker.
Composite mats exhibit negative Poissons ratio which is shown in Table 4.4.
From this we can conclude that reinforcement of an auxetic network (sintered metal mat)
in a conventional matrix (polymer matrix) resulted in an auxetic composite. This
confirms the feasibility of the hypothetical composite discussed in Chapter 2. However,
the increase of stiffness observed in composite mats was mainly due to the compression
of mats during preparation process.
55
Chapter 5
Nanocomposites
56
5.2 Introduction
In this section purification and preparation methods of high strength and flexible
carbon
nanofiber
elastomer
composites
are
discussed.
Silanol
terminated
58
5.3.1 Sonication
14 grams of oxidized fibers were taken and crushed in the mortar with the help of
a pestle to disperse agglomerated carbon fibers. They were weighed with the weight
balance and transferred to a round bottle flask. Nanofibers were mixed with diluted nitric
acid (For 110 ml of nitric acid 590 ml of distilled water was used for dilution) and
sonicated in the sonicator for about 2 hours. Sonication, the process of applying ultra
sound waves to agitate carbon fiber particles in the liquid solvent, also helps in
evacuating dissolved gases from liquid under vacuum. The sonication process is shown in
Figure 5-1.
59
5.3.2 Purification
Experimental setup for purification process is shown in the Figure 5-2. This
consists of an electric heater to maintain a constant temperature of 110 degrees centigrade
throughout the experiment. Heat is conducted from the electric heater by a sand bath
placed on it to the round bottle flask which contains sonicated carbon nanofibers. A
thermometer was placed in the sand bath for measurement of temperature. A refluxor was
placed on top of the round bottle flask to facilitate the back flow of heated nitric acid in to
the round bottle flask. Condenser was attached to the refluxor to help in cooling the acid
60
vapors. Two outlets were provided to the condenser for tap water flow throughout the
process. Entire experimental setup was placed in a hood and was heated for about 48
hours. Then heat was removed and the setup was cooled down to room temperature.
Carbon fibers were then separated from nitric acid solution by filtration procedure.
5.3.3 Filtration
Experimental setup for filtration process is shown in Figure 5-3. The apparatus
consists of a conical filtration flask, funnel and a vacuum pump. Filter papers were placed
in the funnel and then cooled carbon fibers, nitric acid solution were poured. Nitric acid
was collected in to the conical flask, with an outlet to the vacuum pump. Carbon fibers
were collected on the filter paper. Vacuum was applied to speed up the separation of
nitric acid from fibers with the help of vacuum pump. The fibers were washed about 8 to
10 times with distilled water to completely remove nitric acid. Moist carbon fibers were
carefully transformed into a plastic dish to dry at 60 degree centigrade overnight in the
oven. This lump of carbon fibers was crushed in a mortar and used for nanocomposite
preparation.
Carbon
nanofibers
(Grams)
0.45
Oligomer
(Grams)
2.295
Cross linker
(Grams)
0.255
Catalyst
(Grams)
0.255
20
0.6
2.16
0.24
0.24
25
0.75
2.025
0.225
0.225
Carbon nanofibers were taken into mortar and crushed with the help of pestle.
Desired amount of fibers (Table 5.1) were weighed in a plastic cup with weight balance.
Required amount of oligomer was added to the fibers with plastic filler. The plastic cup
was placed in the holder of the speed mixer. The purpose of the speed mixer is for better
dispersion of carbon fibers in the elastomer. The speed mixer was programmed to mix for
20 seconds at 3500 rpm. The process was repeated by adding cross linker and catalyst to
the mixture. Lump of carbon fiber and elastomer obtained was removed and placed in
63
between two wax sheets with a spatula. One of the three different methods was followed
in the curing process of the sample.
(a) In 2- roll mill procedure the roller was set to desired thickness and the mixture
obtained was run in between the rolls. The sample was flipped and rotated to 180
degrees. The process was repeated for about three to four times. Now the sample was
left to cure for about 24 hours. The 2- roll mill procedure gives a uniform thickness
throughout the sample.
(b)
In manual press with no weight procedure the lump of carbon fiber mixture was
pressed with the help of glass cap uniformly and left to cure for 24 hours.
(c) In manual press with heavy weight procedure the lump of carbon fiber mixture was
pressed with the help of glass cap uniformly and a heavy weight was placed on the
glass cap during the curing process of 24 hours.
After curing the sample, they were washed with ethanol to remove excess catalyst
formed on the surface of the nanocomposite and dried in an oven at 60 degree centigrade
for about 24 hours. The sample was later cooled down to room temperature and was
tested on the Instron machine to find strength and Poissons ratio. Scanning electron
microscope images of 20% fiber loading oxidized carbon nanocomposites prepared by all
three procedures were taken as shown in Figures 5-5, 5-6 and 5-7.
64
65
Figure 5-6: OCNF- manual press with weight during curing procedure
(20% fiber loading)
66
Figure 5-7: OCNF- manual press with weight during curing procedure
(20% fiber loading)
67
Poisson's ratio
R1
R2
R3
R1
R2
R3
2Roll mill
13.33
10.97
8.32
0.15
0.17
2.73
21.03
18.15
13.99
3.09
0.168
-0.2
Manual press-no
weight
11.36
10.77
12.24
-2.95
-0.58
-2.27
Table 5.3: Youngs modulus and Poissons ratio for 20% fiber loading of pristine
fibers
Young's modulus (MPa)
Poisson's ratio
R1
R2
R3
R1
R2
R3
2Roll mill
12.46
11.26
12.57
-0.74
3.44
0.83
11.41
12.37
8.75
-1.43
-0.32
-0.77
Manual press-no
weight
13.41
12.69
12.6
-2.96
-6.4
0.35
68
Table 5.4: Youngs modulus and Poissons ratio for 25% fiber loading of pristine
fibers
25% Fiber
loading
2Roll mill
Manual
press-no
weight
Poisson's ratio
R1
R2
R1
R2
5.09
5.02
-2.23
5.9
8.88
11.8
-1.67
-1.96
R1
R2
R3
R1
R2
R3
2Roll mill
12.68
13.48
11.27
-3.15
2.81
-2.13
11.78
9.78
10.11
0.38
-1.72
0.29
69
Table 5.6: Youngs modulus and Poissons ratio for 20% fiber loading of oxidized
fibers
Young's modulus
(MPa)
Poisson's ratio
R1
R2
R3
R1
R2
R3
22.27
18.21
18.31
-3.04
-5.58
-6.64
used
17.34
16
-3.29
0.43
15.85
19.67
-0.8
0.84
2Roll mill
Manual press- weight
15.99
-0.35
Table 5.7: Youngs modulus and Poissons ratio for 25% fiber loading of oxidized
fibers
Young's modulus
(MPa)
Poisson's ratio
R1
R2
R3
R1
R2
R3
2Roll mill
18.72
15.49
16.53
3.2
-4.9
-4.1
12.56
14.83
16.05
-0.39
1.27
4.03
70
Poissons ratio
R1
R2
R3
R1
R2
R3
36.26
34.47
24.6
-1.39
0.61
0.98
5.6 Discussion
5.6.1 Effect of Nanofiber Type
Nanocomposite made of different types of nanofibers, i.e. pristine, oxidized and
functionalized showed a significant increase in stiffness respectively. From Figures 5-8
and 5-9, it can be observed that PDMS functionalized fibers exhibit better stiffness
compared to oxidized and pristine fibers. The -NH2 link is highly reactive when
compared to OH resulting in good dispersion of CNF-PDMS-NH2 in PDMS matrix. The
functionalized fibers also improve wetting, as they are cross linked and covalently
attached to the bulk polymer, enhance interfacial adhesion. The pristine nanocomposites
have poor dispersion and wetting compared to oxidized and functionalized composites as
they are not attached with any functionalized group on their surface. Test results of 20%
fiber loading of pristine, oxidized and functionalized nanocomposites tabulated in Table
5.3, Table 5.6 and Table 5.8 respectively exhibited auxetic behavior for most of the runs.
From the Figure 5-10 it can be observed that there is no significant effect on Poissons
71
ratio based on surface treatment methods. They have similar values of Poissons ratio for
all types of fiber.
Figure 5-8: Stress vs. strain graphs for different nanocomposites of 20% fiber
loading
Figure 5-9: Stress vs. strain graphs for different nanocomposites of 20% fiber
loading (elastic range)
72
Figure 5-10: Lateral strain vs. longitudinal strain graphs for nanocomposite samples
at 20% fiber loading
variation in the slope of the elastic portion of lateral strain versus longitudinal strain as
shown in Figure 5-11. Figure 5-12 shows the elastic portion of stress versus strain graph
for pristine fibers. It can be observed from these results that as the fiber loading increases
stiffness decreases. This is because if fiber loading increases mixing and dispersion of
fibers in the polymer matrix become difficult. From Table 5.9, Table 5.10 and Table 5.11
one can observe that as fiber loading increases the amount of cross linker decreses which
effects the dispersion of these fibers in the polymer matrix.
73
Figure 5-11: Lateral strain vs. longitudinal strain graphs for different fiber loading
of pristine fibers
Figure 5-12: Stress vs. strain curves for nanocomposite samples with pristine
nanofibers
74
Table 5.9: Volume fraction of fiber and polymer for 15% fiber loading
Nanocomposite
sample
Weight
Density(g/cm )
percentage
Volume fraction
Fiber
1.95
15%
.0813
PDMS
0.97
77%
.8339
Cross linker
1.05-1.07
9%
.0848
Table 5.10: Volume fraction of fiber and polymer for 20% fiber loading
Nanocomposite
Density(g/cm3)
Weight percentage
Volume fraction
Fiber
1.95
20%
.1114
PDMS
0.97
72%
.8065
Cross linker
1.05-1.07
8%
.0820
sample
Table 5.11: Volume fraction of fiber and polymer for 25% fiber loading
Nanocomposite
Weight
3
sample
Density(g/cm )
percentage
Volume fraction
Fiber
1.95
25%
.1433
PDMS
0.97
68%
.7777
Cross linker
1.05-1.07
8%
.0791
75
76
Figure 5-13: Lateral strain vs. longitudinal strain for 20% fiber loading of pristine
fibers
Thus nanocomposite sheets prepared as discussed show auxetic behavior and
support the central hypothesis of the project and also validate the random fiber network
model where compression is the key parameter in auxetic behavior of both fiber networks
and their composites. However, auxetic behavior is not observed in nanocomposite if the
compression is too high.
77
Chapter 6
6.1 Conclusion
The two dimensional finite element model created confirms the hypothesis of
manufacturing an auxetic composite by embedding an auxetic fibrous network in a
conventional matrix material. Experiments conducted on sintered stainless steel mats
validated the simulations of numerical model qualitatively but not quantitatively and
those on composite mats confirmed the feasibility of manufacturing auxetic composites.
The amount of compression during mat formation was found to be significant in
determination of auxetic behavior.
Experimental results confirm that the auxetic nanocomposite can be
manufactured. From the results, at high carbon fiber loading it is assumed that the
embedded nanofibers, even not fused or sintered, form an effectively connected network
within a polymer matrix. Compression during the preparation of nanocomposite sheets
plays a key role in determining auxetic behavior.
Poissons ratio is found to decrease with increasing compression for both sintered
mats and nanocomposites. However excessive compression seems to compromise the
78
auxetic network and also the auxetic behavior in nanocomposites is not seen if the
compression is too high.
79
References
1. Lakes, Roderic.
Foam structures with a negative Poisson's ratio. Science, v. 235
issue 4792, 1987, p. 1038.
2. Evans, K. S. Molec
ular network design. Nature, v. 353 issue 6340, 1991.
3. Glieck, The New York Times (1987)
4. Milton, Graeme Composite materials with poisson's ratios close to 1.
Journal of the Mechanics and Physics of Solids, v. 40, 1992, p. 1038.
5. Yang, Wei.
Review on auxetic materials. Journal of Materials Science, v. 39
issue 10, 2004, p. 3269.
Fracture Toughness of Re-entrant Foam Materials with a
6. Lakes, Roderic.
Negative Poisson's Ratio: Experiment and Analysis International Journal of
Fracture, v. 80, 1996, p. 3269.
7. Evans, Kenneth E.
Auxetic materials: functional materials and structures from
lateral thinking!. Advanced Materials, v. 12 issue 9, 2000, p. 617.
8. Smith, C. W.
Strain dependent densification during indentation in auxetic
foams. Cellular Polymers, v. 18 issue 2, 1999, p. 79.
9. McClintock, Mechanical Behavior of Materials,1966.
10. Martz, Erik O.
Hysteresis behavior and specific damping capacity of negative
Poisson's ratio foams. Cellular Polymers, v. 15 issue 5, 1996, p. 349.
80
21. Li. T
he anisotropic behavior of Poisson's ratio, Young's modulus, and shear
modulus in hexagonal materials. physica status solidi (a), v. 38 issue 1, 1976,
p. 171.
22. Grima.
Natrolite: A zeolite with negative Poissons ratios. Journal of Applied
Physics, v. 101 issue 8, 2007, p. 086102-86104.
23. Lees, C.
Poisson's ratio in skin. Bio-Medical Materials and Engineering, v. 1
issue 1, 1991, p. 19.
24. LaBarbera, Michael 1994. Poisson's ratio of a crossed fiber sheath: the skin of
aquatic salamanders. Journal of Zoology, v. 232, 1994, p. 231.
25. Gibson, L. J.
The mechanics of two-dimensional cellular materials. Proceedings
of the Royal Society of London, Series A: Mathematical, Physical and Engineering
Sciences, v. 382 issue 1782, 1982, p. 25.
26. Gibson, L.J Cellular solids: structure and properties London: Pergamon Press,
1988.
27. Aksay, I. A.
Smart materials systems through mesoscale patterning.
Proceedings of SPIE-The International Society for Optical Engineering,
v. 2716 issue Smart, 1996.
28. Larsen, U. D.
structures
with
negative
Poissons
ratio",
IEEE
Journal
of
p. 1466.
30. Jackman, Rebecca J..
Design and fabrication of topologically complex,
three-dimensional microstructures. Science, v. 280 issue 5372, 1998,
p. 2089-2091.
31. Friis, E. A.
Negative Poisson's ratio polymeric and metallic foams. Journal of
Materials Science, v. 23 issue 12, 1988, p. 4406.
32. Choi, J. B.
Non-linear properties of metallic cellular materials with a negative
Poisson's ratio. Journal of Materials Science, v. 27 issue 19, 1992, p. 5375.
33. Choi, J. B.
Nonlinear properties of polymer cellular materials with a negative
Poisson's ratio. Journal of Materials Science, v. 27 issue 17, 1992, p. 4678.
34. Elms, K. Indentability of Conventional and Negative Poisson's Ratio Foams
Journal of Composite Materials, v. 27, 1993, p.1193.
35. Chan, N. Indentation Resilience of Conventional and Auxetic Foams
Journal of Cellular Plastics, v. 34, 1998, p. 231.
36. Hansen, L., David Taylor Research Center, 1991.
37. Chen.
Micromechanical analysis of dynamic behavior of conventional and
negative Poisson's ratio foams. Journal of Engineering Materials and Technology,
v. 118 issue 3, 1996, p. 285.
38. Caddock, B. D "Microporous materials with negative Poisson's ratio: I.
Microstructure and mechanical properties", Journal of Physics D: Applied Physics,
v. 22, 1989, p. 1877.
39. Evans, K E Tensile network microstructures exhibiting negative Poisson's ratios,
Journal of Physics D: Applied Physics, v. 22 issue 12, 1989, p. 1870.
83
40. Evans, K E Microporous materials with negative Poisson's ratios. II. Mechanisms
and interpretation, Journal of Physics D: Applied Physics, v. 22 issue 12, 1989,
p. 1883.
41. Alderson, K. L.
The fabrication of microporous polyethylene having a negative
Poisson's ratio. Polymer, v. 33 issue 20, 1992, p. 4435.
42. Alderson, K. L.
Auxetic polyethylene: the effect of a negative Poisson's ratio
on hardness. Acta Metallurgica et Materialia, v. 42 issue 7, 1994, p. 2261.
43. Alderson, K. L.
An experimental study of ultrasonic attenuation in microporous
polyethylene. Applied Acoustics, v. 50 issue 1, 1997, p. 23.
44. Gibson. F,
Principles of Composite Material Mechanics, McGraw-Hill,
Inc.2005
45. Milton. G., "Composite materials with Poisson's ratios close to -1", Journal of
Mechanics and Physics of Solids, v.40 issue, 1992, p.1105-1137.
The design, matching and manufacture of auxetic carbon fibre
46. Evans.
laminates. Journal of Composite Materials, v. 38 issue 2, 2004, p. 95.
How to make auxetic fiber reinforced composites. physica status
47. Alderson.
solidi (b), v. 242 issue 3, 2005, p. 509.
48. Clarke, J.F., Negative Poissons ratio in angle ply laminates: theory and
experiments., Composites, v. 25, issue 1994, p.863.
49. Herakovich, C. T., Composite laminates with negative Poissions ratio
materials. Journal of Composite Materials, v. 18, issue 8, 1984, p. 447.
50. Alderson.
The low velocity impact response of auxetic carbon fibre laminates.
physica status solidi (b), v. 245 issue 3, 2008, p. 489.
84
51. Chen, C. P.
Viscoelastic behavior of composite materials with conventional- or
negative-Poisson's-ratio foam as one phase. Journal of Materials Science, v. 28
issue 16, 1993, p. 4288.
52. Edwards. Auxeticity windows for composites. Physica A, v. 258 issue 1-2,
1998, p. 5.
Effective elastic properties of composites of ellipsoids (II). Nearly
53. Edwards.
disk- and needle-like inclusions. Physica A, v. 264 issue 3-4, 1999, p. 404.
54. Edwards.
Effective elastic properties of composites of ellipsoids (I). Nearly
spherical inclusions. Physica A, v. 264 issue 3-4, 1999, p. 388
55. Tatlier.
Modelling the negative Poisson's ratio of compressed fused fibre
networks. physica status solidi (b), v. 246 issue 9, 2009, p. 2018.
56. Tatlier. PhD thesis, University of Toledo, 2009
57. Gabro, P,
Forming sintered metal fiber porous mats, Technical Report
6558810, (Freeport, NY) issue 2003.
58. Walter C. Troy. High strength- variable porosity sintered metal fiber articles and
method of making the same, patent No. 492/007, issue 1955, p. 182.
59. Fisher,
Metal fiber reinforcement, patent No. 770/459, issue 1968, p. 208.
Uniform metal fiber sheet, patent No. 09/983, issue, 2002 p.863
60. Kinsley,
61. Delannay Francis, "Elastic model of an entangled network of interconnected
fibres accounting for
85
62. Delince.
Elastic anisotropy of a transversely isotropic random network of
interconnected fibers: non-triangulated network model. Acta Materialia, v. 52
issue 4, 2004, p. 1013.
63. Saito, R,
Physical Properties of Carbon Nanotubes, Imperial College Press,
London, UK, 1998.
Ultrastrong, stiff, and lightweight carbon-nanotube fibers. Advanced
64. Zhang.
Materials, v. 19 issue 23, 2007, p. 4198.
Carbon Nanotube-Based Nonvolatile Random Access
65. Rueckes, Thomas.
Memory for Molecular Computing. Science, v. 289 issue 5476, 2000, p. 94-97.
66. Zhen Yao.
Carbon nanotube intramolecular junctions. Nature, v. 402
issue 6759, 1999, p. 273.
67. Qian, D..
Load transfer and deformation mechanisms in carbon nanotubepolystyrene composites. Applied Physics Letters, v. 76 issue 20, 2000.
68. Ajayan, Pulickel M.
Single-walled carbon nanotube-polymer composites:
strength and weakness. Advanced Materials, v. 12 issue 10, 2000, p. 750.
69. Lozano, K.
Nanofiber-reinforced thermoplastic composites. I. Thermoanalytical
and mechanical analyses. Journal of Applied Polymer Science, v. 79 issue 1,
2000, p. 125.
70. Kishi.
Damping properties of thermoplastic-elastomer interleaved carbon fiberreinforced epoxy composites. Composites Science and Technology, v. 64
issue 16, 2004, p. 2517.
86
71. Ritchie, G. L. D.
Temperature dependence of electric field-gradient induced
birefringence (the Buckingham effect) in C6H6 and C6F6: comparison of electric
and magnetic properties of C6H6 and C6F6. Chemical Physics Letters, v. 322
issue 3, 2000, p. 143.
72. Evans.
Modelling negative poisson ratio effects in network-embedded
composites Acta Metallurgica et Materialia, v. 40 issue 9, 1992, p. 2463.
73. Hall Lee J.
Sign Change of Poisson's Ratio for Carbon Nanotube Sheets. Science
(New York, N.Y.), v. 320 issue 5875, 2008, p. 504.
87