Professional Documents
Culture Documents
net/publication/254112443
CITATIONS READS
12 1,061
5 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Rodney D. Averett on 16 May 2018.
Abstract
The mechanical and fatigue behavior of neat poly(lactic acid) (PLA) films and PLA films reinforced with 5 wt% nanoclay
particles has been examined using various analytical procedures. The results showed that for the films tested in this
study, PLA-5 wt% samples were more susceptible to crazing at the same maximum fatigue stress as the neat PLA samples,
as evidenced by results from light transmission experiments. Optical microscopy results confirmed this observation. In
addition, under fatigue loading conditions, the neat PLA samples displayed almost the same fatigue resistance (number of
cycles to failure) at 3 and 30 Hz, while the PLA-5 wt% samples showed a shift in the number of cycles to failure to higher
values at a frequency of 30 Hz. Using the linear regression curves from the S–N data (stress vs. number of cycles to
failure), time-to-failure curves were generated to show the difference between the neat PLA and PLA-5 wt% samples
when tested at frequencies of 3 and 30 Hz. Based on these results, it is known that the nanoclay particles served to
increase the fatigue resistance at the higher frequency of 30 Hz, when compared to the neat PLA sample.
Keywords
Films, nanocomposite, fatigue, poly(lactic acid), nanoclay
in PLA as a viable engineering and packaging material, The uniaxial tensile tests for characterization were per-
there should be more studies directed at the mechanical formed at an elongation rate of 5 mm/min on a standard
and fatigue behavior of the materials. This study seeks Instron machine and the fatigue tests at frequencies of
to mitigate this research gap by identifying the modes 3 and 30 Hz. The Enduratec has a maximum load of
and mechanisms of failure for PLA and PLA-5 wt% 225 N and a maximum frequency of 400 Hz. A set of low
samples with nanoparticle loaded under uniaxial ten- mass grips, Model GRP-TC-DMA450N from BOSE
sion and uniaxial fatigue loading. ELectroForceÕ (Eden Prairie, MN, USA), were used to
conduct the fatigue tests. The ELF 3200 measures dis-
placements via a Capacitec 100 mm displacement trans-
Experimental ducer (Model HPC-40/4101) used as a feedback for the
control loop. The Q800 dynamic mechanical analysis
Materials and film preparation
(DMA) machine from TA Instruments (New Castle,
PLA 2002D (LMW) was provided in pure and filled DE, USA) was used to evaluate and characterize the
(5 wt% Cloisite 30B organoclay) pellet form from dynamic behavior of the samples as a function of
NatureWorksÕ company. Cloisite 30B (C30B) is an frequency and temperature. The specimen size used for
organically modified montmorillonite (MMT) with the DMA tests was 10 mm wide and 25 mm long. All
alkyl ammonium as the clay modifier. Before the cast- experiments were conducted at room temperature, labo-
ing took place, the pellets were dried in a vacuum oven ratory air. The typical humidity of the laboratory air was
overnight at a temperature of 80 C. A 1.500 Prodex about 50%.
single screw extruder with a chill roll casting unit was
used to cast amorphous films of neat and filled PLA
2002D (LMW). The casting conditions are presented in Light transmission experiments
Table 1. Cast films were prepared with the single screw
extruder (melt temperature 200 C, L/D 32:1), which Light transmission experiments were conducted in
was equipped with a 200 mm wide T-die and a take- the visible wavelength spectrum (360–750 nm) on the
up system. The films were cast onto a chilled roll (tem- UltraScan XE spectrophotometer from HunterLab
perature 30 C) using the screw extruder. After process- (Reston, VA). Due to the small specimen width, an
ing, a gel permeation chromatograph machine was used aperture size of 0.37500 was attached to the instrument
to measure the molecular weights of PLA. The weight– to conduct the experiments. Light transmission experi-
average molecular weight (MW) of the 2002D film ments were conducted in the visible wavelength
grade was determined to be 137,000 g/mol. spectrum (360–750 nm) to determine the effects of
mechanical loading on the opacity (reduction in trans-
parency) of the PLA and PLA-5 wt% nanocomposite
Mechanical testing samples. The samples were loaded into the machine
The sample specimen size used for the fatigue tests as close to the fracture surface as possible, so as to
was 9.5 mm wide and 30 mm long. The thickness of the capture the effects of stress whitening close to the fail-
neat PLA and PLA-5 wt% samples was 0.26 0.02 and ure region.
0.30 0.04 mm, respectively. The BOSEÕ ELectroForceÕ The equation that appositely describes the amount
(ELFÕ ) 3200 testing machine (Enduratec)) was used to of light that is transmitted through the sample can be
conduct the mechanical experiments in uniaxial fatigue. determined by considering the Beer–Lambert Law,7, 8
considering that the fractional intensity that is lost can
be represented as:
Table 1. Processing conditions used to fabricate PLA nano-
composite film specimens dI
¼ aT dh ð1Þ
I
Feed temperature (zone 1) 180 C
Compression section temperature (zone 2) 190 C
where dI is the fractional loss in intensity, I the intensity
Metering temperature (zone 3) 200 C of light, aT the extinction coefficient, and dh the infin-
Adapter temperature 200 C itesimal thickness of the sample. If this equation is inte-
Die temperature 200 C grated, one obtains the following for the general form
Chill roll temperature 52 C of the Beer–Lambert Law:
Screw speed 33 rpm
Take-up speed 70 rpm Z h Z h
dI
Die gap thickness 0.4 mm ¼ aT dh ð2Þ
0 I 0
Averett et al. 2719
The intensity as a function of the extinction coefficient wt% sample was approximately 5% more than the
and thickness of the sample is expressed as the Beer– neat PLA sample. Finally, the maximum stress of
Lambert Law:7 the neat PLA sample was actually 18% higher than
the 5 wt% sample. These results are not in accordance
I ¼ I0 eaT h ð3Þ with similar studies that have evaluated the mechanical
behavior of PLA with the inclusion of clay nanoparti-
The transmission of light through the sample can be cles. Jiang et al.5 noted an increase in the tensile
expressed as: strength of the PLA nanocomposites up to 5 wt%,
whereas in this study, a decrease in the tensile strength
I P was observed.
T¼ ð4Þ
I0 P0 These results are somewhat counterintuitive, since
one would envision that the addition of rigid platelet
where I and I0 represent the transmitted intensity and clay nanoparticles would significantly enhance all the
the incident intensity, respectively, and P and P0 the mechanical properties of the PLA samples. However,
transmitted power and incident power, respectively. the intimacy of the bonding between the matrix and
nanoparticle was low, since the addition of functional
groups or other bonding agents was not incorporated
Optical microscopy experiments into the fabrication process. This led to a lower inter-
Optical microscopy of the unfilled and filled samples facial strength of the nanocomposite specimens. Due to
was performed on a Leitz Laborlux 12 Pol S micro- the low interfacial adhesion between the nanoparticles
scope in unpolarized light configuration. Optical pic- and PLA matrix, debonding occurred at the site of the
tures were taken by a digital camera, Micropublisher interface, leading to lower fracture strengths and higher
5.0 RTV, from Qimaging connected to the microscope strains at break under uniaxial conditions. In fact, it
and collected using Image Pro plus 6.2 software. has been observed in literature that systems with low
interfacial adhesion between the matrix and polymer
exhibit debonding at a stress lower than that of the
Results and discussion neat matrix,5 leading to lower yield strengths and
higher strains at break (due to massive crazing) with
Uniaxial tensile test results
increasing nanoparticle concentration. Also, due to
The results from the uniaxial tensile tests indicate that the platelet shape of the nanoparticles, stress concen-
the neat PLA and PLA-5 wt% films exhibited similar tration sites were engendered, causing the debonding to
mechanical properties. The results for these experi- occur more easily. Argon and Cohen9 have illustrated
ments are presented in Tables 2 and 3. The modulus how the yield stress in a polymeric sample decreases
of the neat film (3.1 GPa) was very close to that of the with the addition of rigid platelet CaCO3 nanoparticles
PLA-5 wt% nanocomposite film (3.2 GPa). On aver- (Figure 1). In essence, they9 explained that the reduc-
age, the maximum strain of the neat film sample was tion in yield strength was due to the fact that the par-
very close to that of the nanocomposite sample (0.081 ticles debonded from the matrix before overall yield.
compared to 0.10). The tensile energy of the PLA-5 Based on these results, it is believed that under uniaxial
Table 3. Uniaxial tensile test results for PLA films with 5 wt% nanoclay particles
0.3 9000
tan delta
Storage modulus 8000
0.25
5 wt% sample 7000
tan delta
5000
0.15
4000
0.1 3000
Neat
sample 2000
0.05
1000
0 0
0 50 100 150 200 250
frequency (Hz)
Figure 3. Tan and storage modulus vs. frequency for neat PLA
and PLA-5 wt% samples.
40
PLA 5wt% Sample increased as a function of the nanoclay content for
all frequencies. In this study, for both the neat PLA
30 and the PLA-5 wt% samples, the critical frequency at
which the tan function displayed its maximum
20
peak was approximately 100 Hz. The results from the
10 storage modulus vs. temperature show a similar result
(Figure 4), where the mechanical behavior of the films
0
0 0.02 0.04 0.06 0.08 0.1 0.12 from room temperature up to 120 C was virtually iden-
strain tical. The results (Figure 5) show that for PLA-5 wt%
PLA Neat Sample PLA 5wt% sample
samples, the glass transition temperature was slightly
Figure 2. Example stress–strain curves of neat PLA and PLA-5 shifted to a lower temperature. The Tg of the neat
wt% samples under uniaxial tensile loading conditions. PLA samples was 77.1 C while that of the PLA-5
wt% samples was 74.7 C.
10000
Storage modulus PLA neat
PLA 5wt%
1000
Loss modulus
100
Modulus (MPa)
10
1
0 20 40 60 80 100 120 140
0.1
0.01
Temperature (deg C)
Figure 4. Storage and loss modulus vs. temperature for neat PLA and PLA-5 wt% samples.
3.5
rigid particles were responsible for preventing light
from passing through the film.
3
PLA Neat The light transmission of PLA-5 wt% vs. neat PLA
PLA 5wt% samples can be seen more clearly from Figure 8, where
2.5
the light transmission at 420 nm (critical wavelength) of
tan delta
Figure 6. Uniaxial tensile test results: example light transmission vs. wavelength in the visible spectrum (360–750 nm) for PLA-5 wt%
sample tested under uniaxial loading to failure.
Figure 7. Fatigue results: light transmission vs. wavelength for neat PLA and PLA-5 wt% nanocomposite samples.
opacity due to stress whitening was observed for the tested in the study (both neat and 5 wt% nanocompo-
PLA-5 wt% nanoclay films. site). The experiments conducted below the critical
glass transition frequency (below 100 Hz) at a fre-
quency of 3 Hz indicate a result that is concomitant
Results from fatigue
with the results from the uniaxial tensile tests. At
Fatigue tests were conducted on the PLA and PLA-5 3 Hz, the unreinforced PLA samples exhibited a
wt% samples at various levels of maximum stress to higher fatigue resistance than the PLA samples rein-
determine the fatigue resistance at frequencies of 3 forced with 5 wt% nanoclay particles, as shown in
and 30 Hz. Maximum stress vs. number of cycles to Figure 9. However, at 30 Hz, the neat PLA and PLA-
failure (S–N) curves were generated for these samples 5 wt% samples exhibited almost the same fatigue
to determine their resistance to a repeated cyclic load- behavior. This indicates that the nanoclay particles
ing for a fixed maximum stress. Maximum fatigue stress served to enhance the fatigue resistance of the samples
values in the range of approximately 25% to 75% of at 30 Hz. Based on the regression results shown in
the maximum uniaxial stress were used to conduct the Figure 9, it is known that both neat PLA and PLA-5
fatigue experiments. Approximately 54 samples were wt% samples displayed a linear semi-log behavior with
Averett et al. 2723
a functional form that relates the maximum fatigue where max represents the maximum fatigue stress, m
stress to the number of cycles to failure as: the slope of the regression curve, Nf the number of
cycles to failure, and b the y-intercept. Equation (5),
which relates the maximum stress to the number of
cycles to failure can be rewritten in a modified form,
expressing the maximum stress as a function of time to
max ¼ m lnðNf Þ þ b ð5Þ
failure, tf, and frequency, f, as:
max b
e ð m Þ
75 tf ¼ ð7Þ
f
PLA 5 wt% samples
70
For the neat PLA and PLA-5 wt% samples con-
ducted at 3 and 30 Hz, one would expect a similar
65 shift in the number of cycles that yields an equivalent
time to failure shift, since (1) the results from DMA
60 indicate that the storage modulus and tan functions
0 10 20 30 40 50
σmax(MPa)
were virtually constant in this frequency range and (2)
the uniaxial characterization results indicate that the
Figure 8. Percentage light transmission vs. maximum stress for neat PLA sample is superior to that of the PLA-5
neat PLA and PLA-5 wt% samples as a function of maximum wt% sample in terms of yield strength and fracture
fatigue stress. strength. However, the neat PLA and PLA-5 wt%
45
PLA neat
sample
40
PLA neat sample
f =30Hz
35 y = -8.1Ln(x) + 101.6
PLA 5wt%
sample
30 f = 30Hz
Maximum Stress (MPa)
y = -0.9Ln(x) + 26.4
15
runout
y = -1.1Ln(x) + 26.6
10
y = -6.0Ln(x) + 80.7
0
100 1000 10000 100000 1000000
# Cycles to Failure
Figure 9. Maximum stress vs. number of cycles to failure (S–N curve) for neat PLA and PLA-5 wt% samples conducted at 3 and
30 Hz.
2724 Journal of Composite Materials 45(26)
nanocomposite samples behaved differently in terms of experiments in ‘Uniaxial tensile test results’ section and
the time to failure due to fatigue loading as a function the fatigue results of this section, it can be concluded
of maximum stress and frequency. In an effort to quan- that the nanoclay particles engendered two antithetical
tify the difference in behavior of the two samples (neat mechanical property attributes: (1) decreased maximum
vs. 5 wt%) as a function of frequency, time-to-failure stress (yield stress) and fracture stress under uniaxial
curves were generated based on the empirical expres- tension conditions and (2) increased time to failure
sions and are shown in Figure 10. The time-to-failure between the lower frequency of 3 Hz and higher fre-
curves for the neat PLA samples at 3 and 30 Hz show a quency of 30 Hz when compared to the neat PLA
very wide difference in time to failure than the PLA-5 sample (lower time-to-failure differential between 5
wt% samples. In terms of time, the neat PLA samples wt% sample at 3 and 30 Hz).
at 30 Hz failed much sooner than the neat PLA samples
at 3 Hz. At 30 Hz; the PLA-5 wt% samples exhibited
Optical microscopy results
the same behavior as the neat PLA samples, where they
failed sooner than the samples tested at 3 Hz. However, The results from the optical microscopy study indicated
there is one noticeable difference between the fatigue that the nanofiller particles may have served to prevent
behaviors of the two samples. In terms of time, it the formation of cracks in the PLA sample at 30 Hz.
must be noted that the PLA-5 wt% samples exhibited Figure 11 shows the pictures of neat PLA and PLA-5
a much smaller time differential between 3 and 30 Hz wt% samples at 30 Hz. This is to show that crack for-
than the neat PLA samples. After all, based on the mation in neat PLA samples at 30 Hz was more severe
mechanical characterization results, one would expect than the PLA-5 wt% samples. This is evidenced by the
that the fatigue resistance of the PLA-5 wt% samples at fact that the neat PLA samples showed more severe and
30 Hz would be lower than the neat PLA samples; how- long cracks than the PLA-5 wt% samples parallel to the
ever, this was not the case. This observation is interest- loading direction. In comparison, the PLA nanocom-
ing, because it indicates that the nanoclay particles posite samples did not show a large density of cracks, as
served to enhance the fatigue resistance of the PLA shown in Figure 11.
samples at 30 Hz, where it increased the time necessary In contrast, for the PLA-5 wt% samples tested at the
to cause failure. In addition, the neat PLA samples lower frequency of 3 Hz (Figure 12), the samples
fatigued at 30 Hz displayed a steeper decline slope in appeared to have developed longitudinal cracks along
the semi-log maximum stress vs. number of cycles to the direction of loading at a stress ratio lower than that
failure behavior. Thus, from the results of the uniaxial of the PLA-5 wt% samples tested at 30 Hz. This clearly
20000
18000
16000
14000
Time to failure (s)
12000
10000
PLA Neat, f=3Hz
8000
4000
PLA 5wt%, f=30Hz
2000
PLA Neat, f=30Hz
0
0 5 10 15 20 25 30 35 40 45
Maximum fatigue stress (MPa)
PLA 5wt%, f=3Hz PLA 5wt%, f=30Hz PLA neat sample, f=3Hz PLA neat sample, f=30Hz
Figure 10. Empirical curves of time to failure vs. maximum fatigue stress for neat PLA and PLA-5 wt% samples.
Averett et al. 2725
Figure 11. Optical microscope results of neat PLA and PLA-5 wt% samples at 30 Hz.
when tested at 3 and 30 Hz cyclic frequencies. Perhaps 2. Okuyama N, Wang CY, Rose EA, Rodgers KE, Pines E,
the results at 3 Hz are concomitant with the results from diZerega GS, et al. Reduction of retrosternal and pericar-
the neat PLA and PLA-5 wt% uniaxial (quasistatic) dial adhesions with rapidly resorbable polymer films.
experiments, where the lower rate of strain application Ann Thorac Surg 1999; 68: 913–918.
3. Ersoy E, Ozlurk V and Yazgan A. Ozdogan M and
at that frequency was highly similar to that of uniaxial
Gundogdu H. Effect of polylactic acid film barrier on
tension. However, at the higher frequency of 30 Hz, the intra-abdominal adhesion formation. J Surg Res 2008;
neat PLA samples may have experienced more damage, 147: 148–152.
since the fatigue life was not increased. Optical micros- 4. Strange M, Plackett D, Kaasgaard M and Krebs FC.
copy results revealed that the clay nanoparticles served Biodegradable polymer solar cells. Degradation and sta-
to aid in the retardation of crack formation at a fatigue bility of polymer and organic solar cells. Sol Energy
frequency of 30 Hz. Based upon the results from litera- Mater Sol Cells 2008; 92(7): 805–813.
ture and the neat PLA fatigue experiments, it is known 5. Jiang L, Zhang J and Wolcott MP. Comparison of
that the nanoclay particles, although not intimately polylactide/nano-sized calcium carbonate and poly-
bonded to the PLA polymer chain architecture, served lactide/montmorillonite composites: Reinforcing effects
to bolster the fatigue resistance of the samples. The and toughening mechanisms. Polymer 2007; 48(26):
7632–7644.
results of this study are important because they indicate
6. Lee YH, Lee JH, An I-G, Kim C, Lee DS, Lee YK, et al.
that the mechanical behavior of neat PLA and PLA
Electrospun dual-porosity structure and biodegrada-
nanoclay samples tested under fatigue loading condi- tion morphology of Montmorillonite reinforced PLLA
tions is different for different frequencies. These results nanocomposite scaffolds. Biomaterials 2005; 26(16):
can be used as a quantitative and qualitative measure 3165–3172.
for studies such as accelerated life testing and mechan- 7. Jenkins FA and White HE. Fundamental of Optics. New
ical performance. It would be interesting to ascertain York: McGraw-Hill, 1981, p. 231.
how these results compare to a study in which the tem- 8. Coelho JMP, Abreu MA and Rodrigues FC.
perature of the samples is decreased (similar to increase Methodologies for determining thermoplastic films opti-
in frequency for this study), and fatigue experiments cal parameters at 10.6 mm laser wavelength. Polym Test
conducted to determine if the time–temperature super- 2004; 23(3): 307–312.
9. Argon AS and Cohen RE. Toughenability of polymers.
position phenomenon is applicable to these polymer
Polymer 2003; 44(19): 6013–6032.
systems.
10. Pluta M, Jeszka JK and Boiteux G. Polylactide/montmo-
rillonite nanocomposites: Structure, dielectric, viscoelas-
Acknowledgments tic and thermal properties. Eur Polym J 2007; 43(7):
The authors thank Dr Fred Cook and Murari Gupta at the 2819–2835.
Georgia Institute of Technology for their help with the light 11. Ward IM. Mechanical behavior of poly(ethylene tere-
transmission experiments and many helpful discussions. phthalate). J Macromol Sci Part B Phys 1967; 1(4):
667–694.
Funding 12. Anderson TL. Fracture Mechanics: Fundamentals and
Applications, 2nd ed. Boca Raton, FL: CRC Press,
This research received no specific grant from any funding 1995, p. 688.
agency in the public, commercial, or not-for-profit sectors. 13. Kim GH and Lu H. Characteristics of accelerated lifetime
behavior of polycarbonate under athermal and high load-
References ing frequency conditions. Polym Test 2007; 26(7):
1. Chang YN, Yang PC, Jin Z, Chellapa H, Xing C, Mueller 839–845.
R, Anderson M, Clarke A and Iannotti E. Low molecular
weight polylactic acid for controlled release of active
agents. In: Proceedings of the Corn Utilization Conference
VI. St. Louis, MO, 4–6 June 1996, National Corn
Growers Association and National Corn Development
Association, pp. 1–4.