Professional Documents
Culture Documents
S0142-9418(13)00104-9
DOI:
10.1016/j.polymertesting.2013.05.007
Reference:
POTE 4062
To appear in:
Polymer Testing
Please cite this article as: W.N. dos Santos, J.A. de Sousa, R. Gregorio Jr., Thermal conductivity
behaviour of polymers around glass transition and crystalline melting temperatures, Polymer Testing
(2013), doi: 10.1016/j.polymertesting.2013.05.007.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Material Behaviour
THERMAL CONDUCTIVITY BEHAVIOUR OF POLYMERS AROUND
GLASS TRANSITION AND CRYSTALLINE MELTING TEMPERATURES
RI
PT
jasousa@ufscar.br gregorio@ufscar.br
M
AN
U
ABSTRACT
SC
dwns@ufscar.br
The thermal conductivity of five semi-crystalline and four amorphous polymers was
determined within a wide range of temperature, starting at room temperature and going
up to temperatures above the polymer melting point (Tm) for semi-crystalline polymers
or above the glass transition temperature (Tg) for amorphous polymers. Two transient
techniques were employed in the experimental investigation: the hot wire technique for
the group of amorphous polymers, and the laser flash technique for the semicrystalline
TE
D
EP
AC
C
ACCEPTED MANUSCRIPT
*Corresponding author.
INTRODUCTION
Modelling heat-energy transfer problems under steady state or transient
conditions is of fundamental importance in engineering design as well as for tailoring of
RI
PT
are needed for heat transfer calculations. Detailed knowledge of such properties
becomes crucial in many situations when heat is added or removed from a material.
SC
M
AN
U
transfer. This property assumes a critical role in the performance of materials in high
temperature applications. Low thermal conductivity values are required when the
purpose is to minimize heat losses. On the other hand, when heat transfer from one site
to another is desirable, materials with higher thermal conductivities must be chosen.
Therefore, reliable thermal conductivity values are essential for material selection
criteria in order to get the best performance of the material in a specific application.
However, it is one of the physical quantities that is difficult to measure and requires
TE
D
EP
AC
C
conduction. In the extrusion and injection moulding processes, knowledge of the precise
values of both the thermal conductivity and diffusivity and their temperature
dependence is essential for heat-energy transfer calculations in both heating and cooling
cycles of polymer processing operations. In cooling operations, thermal conductivity
variations can induce non-uniform cooling and shrinkage problems and, thereby, cause
frozen-in stresses, delamination, shrink voids in extrudate profiles and injection
mouldings [2, 3].
The specific heat (cp - heat capacity per unit mass) is a thermodynamic quantity
that determines the amount of heat necessary to raise by one degree the temperature of a
unit mass of a material and is, therefore, associated with the energy consumption in
ACCEPTED MANUSCRIPT
heating processes. In the polymer melting process, the value of cp increases substantially
due to the fact that the high heat absorbed by the semi-crystalline material is used up as
latent heat of fusion and does not contribute towards a temperature variation.
The anomalous thermal conductivity variations verified for both crystalline and
RI
PT
has seldom been analyzed in terms of the actual phonon heat conduction theory used for
ceramic materials [4]. Hence in this work, the influence of thermal transition
temperatures (Tg and Tm) on the behaviour of the thermal conductivity with
SC
temperature is determined and analysed in terms of the phonon conduction process. The
M
AN
U
TE
D
The equation relating the thermal conductivity (k, W/m K), thermal diffusivity
(, m2/s), specific heat at constant pressure (cp, J/kg K) and the bulk density (, kg/m3)
is given by:
(1)
EP
k = cp
AC
C
ACCEPTED MANUSCRIPT
1
k = Cp v
3
(2)
For pure crystalline ceramic material, when temperature increases, specific heat
also increases to an approximately constant value, and the phonons mean free path
RI
PT
decreases, since it is proportional to T-1. For temperatures not so low (T > -2200C), the
net effect is a decrease in the thermal conductivity. Experimental results indicate that,
A
+B
T
SC
k=
(3)
M
AN
U
where T is the absolute temperature and A and B are constants to be determined. The
temperature dependent term is interpreted as the intrinsic thermal conductivity due to
the Umklapp process (an anharmonic phonon-phonon scattering process), and the
temperature independent term as conductivities due to the collective contribution of all
point defect scattering terms [6]. The conductivity decreases with temperature increase
because the anharmonic Umklapp process becomes more frequent, and hence the
phonon mean free path decreases. Umklapp scattering is the dominant process for
TE
D
k = CT + D
(4)
AC
C
EP
where C and D are also experimental constants to be determined, with the same
k=
1
AT + B +
C
T
(5)
ACCEPTED MANUSCRIPT
Although the previous equations are also valid for polymers, since phonons are
mainly responsible for heat transfer, polymeric materials have some particularities
which make the structure of such materials different from that of ceramic structures.
The polymer thermal conductivity is also dependent on molecular weight distribution,
RI
PT
M
AN
U
SC
for this class of polymers, a singular behaviour of the thermal conductivity is observed
As stated before, two experimental techniques were employed in this work: the
hot wire technique for the amorphous polymers and the laser flash technique for the
semicrystalline polymers. Thermal conductivity results obtained by using both
techniques are perfectly compatible for polymeric materials, and discrepancies verified
lie within the accuracy of each technique, as reported elsewhere by the authors [9].
The hot wire technique permits direct measurement of thermal conductivity of
TE
D
materials. However, due to larger sample (~1.5 kg) requirements for this test, the
thermal equilibrium is achieved only after relatively long periods (~6 h) of polymer
exposure to heat, and also the measurements are severely hampered in the melting
temperature range of semicrystalline polymers. The high and rapid heat expansion of
EP
AC
C
severely hampers the stability of the measurements being made in experimental runs of
10 to 15 minutes duration. Furthermore, restrictions due to economic or other technical
aspects also restrict the use of this technique for some materials. However, the hot wire
technique permits precise measurements of thermal conductivity variations in the
temperature range around Tg of polymers. Details of the hot wire technique and samples
preparation were given in many occasions elsewhere [10 - 13].
In the laser flash technique, the use of much smaller samples (~50 mg) simplifies
sample preparation, and the thermal equilibrium is quickly attained (1-2 min),
permitting measurements both in the solid and molten states and even in the narrow
crystalline melting interval of semicrystalline polymers. However, this technique
ACCEPTED MANUSCRIPT
measures the thermal diffusivity and not the thermal conductivity, which can then be
derived from equation (1) shown earlier, provided that the specific heat and density data
of the sample are also determined as a function of temperature. Details of the laser flash
technique and sample preparation were given on many occasions elsewhere [14 - 17].
RI
PT
In this work, the hot wire technique was adopted for the amorphous polymers
and the thermal conductivity was directly determined in the temperature interval starting
from room temperature and going up to temperatures above the Tg of the analyzed
polymers. In the case of the semicrystalline polymers, the laser flash technique was
SC
temperatures above the crystalline melting point of these materials. The thermal
conductivity data was then derived using the experimental values of both the polymer
M
AN
U
TE
D
EP
prepared through extrusion starting from powder or pellets of the solid polymer, using a
special mould of stainless steel with inner dimensions of 220x100x65 mm in the shape
Before each measurement, a dwell time at each
AC
C
of a rectangular parallelepiped.
temperature is required in order to get the thermal equilibrium: 6 hours for the hot wire
technique (large samples) and 1-2 minutes for the laser flash technique (small samples).
In both techniques, the thermal property measurements were carried out in the heating
mode.
ACCEPTED MANUSCRIPT
RI
PT
Table 1
Figures 1 to 5 show, respectively, the results obtained for the group of semicrystalline polymers studied: high density polyethylene (HDPE), low density
polyethylene (LDPE), polypropylene (PP), nylon 6 (PA6) and polyvinylidene fluoride
SC
(PVDF). The traced vertical lines were included in the figures in order to indicate the
M
AN
U
Figures 1 to 5
Figure 6 shows the experimental results obtained for the group of amorphous
polymers studied: polymethylmethacrilate (PMMA), high impact polystyrene (HIPS),
crystal polystyrene (PS) and polycarbonate (PC). In these figures, the Tg values
TE
D
Figure 6
EP
AC
C
be explained using equation (2), in terms of two mechanisms of opposite effects: the
specific heat increases slightly when temperature rises, but the thermal diffusivity
reduces due to the decrease in the phonon mean free path. The reduction in the polymer
density (bulk density) with temperature also contributes to k values reduction. The net
effect of these mechanisms is a smooth decrease of the thermal conductivity with
temperature rise, except for PA6. In this case, the steeper slope of cp curve with
increasing temperature contributes towards a slight rise in k values. In the beginning of
the melting process, the thermal conductivity exhibits an abrupt increase as temperature
rises; reaches a maximum and then decreases, following the behaviour of the specific
heat within the melting temperature interval.
ACCEPTED MANUSCRIPT
RI
PT
regions may be considered: in the glassy state before Tg and in the rubbery state after
Tg. In the glassy state region, the thermal conductivity increases with temperature as
predicted by equation (4). This is due to the specific heat increase effect, since phonon
mean free path is considered practically constant with temperature for amorphous
materials. Starting in the rubbery region, an abrupt decrease of the thermal conductivity
SC
is observed. As a result, a peak is attained close to the Tg. This sudden drop may be
M
AN
U
neighborhood of Tg the polymer passes from a rigid and brittle glassy state to the soft
rubbery state. In the rubbery state, the large-scale mobility of atoms and
macromolecules increases and chain segments undergo intensive thermal motion and
torsional rotations. This disorder introduced in the material structure contributes to the
scattering center phonons and, thereby, decreasing the thermal conductivity. A sudden
drop of the thermal diffusivity of amorphous PET in Tg's region was also observed by
as follows.
TE
D
Chen et al using the flash method [18]. Physically, this phenomenon may be explained
For amorphous materials, the phonon mean free path is practically independent
of the temperature. When the lateral dimensions are much less than the wavelength, an
extensional wave is propagated and the phonon mean speed (v) is proportional to the
AC
C
i.e.:
EP
square root of the ratio of the elasticity modulus (E) and the bulk density of the material,
(6)
Although the density decreases with temperature rise, the elasticity modulus
experiences a sudden and intense drop above Tg, as shown schematically in Figure 7.
This figure displays the sudden drop of E after the Tg for semi-crystalline and
amorphous polymers. This sudden drop in E is much more pronounced (2-3 orders of
magnitude) in amorphous than in semi-crystalline polymers. The net effect is a decrease
in the phonon speed, which according to equation (2) leads to a corresponding decrease
ACCEPTED MANUSCRIPT
in the thermal conductivity. After this sudden drop, the thermal conductivity displays a
smooth variation with temperature, as shown in Figure 6.
RI
PT
Figure 7
exhibits high molecular stiffness below Tg, with the presence of two p-phenylene
groups in its macromolecular skeleton, and thus contributing towards higher phonon
SC
conduction [19]. It is also possible to observe in Figure 6 that the sudden drop in
M
AN
U
lower than the temperature verified for the Tg of this polymer, as determined by the
conventional calorimetric technique (112 C). This fact may be explained having in
mind the long dwell time (up to 6h) before each measurement of thermal conductivity in
order to get the thermal equilibrium in the hot wire test. After this elapsed time, part of
the polymeric mass would reach the rubbery state, even at a temperature below the
classical Tg, resulting in a decrease of the thermal conductivity. In order to confirm this
hypothesis, two DSC runs with the PMMA samples are provided in Figure 8. Initially,
TE
D
the sample (8 mg) remained at 86 C for a short dwell time of 5 minutes and was then
heated up to 140 C, as shown on Figure 8 (2nd heating run). This curve was then
compared to the one obtained in the determination of Tg, from room temperature up to
Figure 8
AC
C
EP
140 C (1st heating run). In both cases the heating rate was 10 C/min.
It is evident from Figure 8 that the variation of the base line decreased after
PMMA sample was held at 86 C for five minutes. This behavior is an indication that
part of the polymeric mass had already attained the rubbery state. As a consequence, a
drop in its thermal conductivity is observed at this temperature, several degrees before
the Tg value determined by DSC. This fact may be explained having in mind that the
conventional Tg is a kinetic (rate-dependent) manifestation of an underlying
thermodynamic phenomenon, and corresponds to the temperature at which the forces
that keep the connected chain segments of a solid polymer are overcome by the
thermally induced movement within the experimental time scale. Hence, the value of
ACCEPTED MANUSCRIPT
such a property depends on the heating rate and the time that the material remains at a
fixed constant temperature. Figure 8 shows that, for PMMA, the actual glass transition
temperature in the hot wire thermal conductivity technique is attained at around 86 0C,
and not at 112 0C as determined by the classical differential scanning calorimetry, as
RI
PT
shown in Table 1. For the other amorphous polymers, this behavior was not detected.
The variation of the base line remains approximately constant even when such a
material remained for a time time below its classical Tg value, which indicates that this
time interval was not enough to lead the polymer to a rubbery state. This difference in
the behaviour between amorphous polymers is associated with the type and stiffness of
SC
each structure. The effects of number-average molecular weight, chain stiffness and
M
AN
U
CONCLUSIONS
TE
D
EP
AC
C
ACKNOWLEDGEMENTS
The authors acknowledge to FAPESP (Proc. 2011/02943-4) for the financial
support and to Prof. Elias Hage Jr. for the valuable discussions.
REFERENCES
1. H. Lobo. Thermal conductivity and diffusivity of polymers. Handbook of Plastics
Analysis, Marcel Dekker Inc. 2003.
2. C. Rauwendaal. Polymer extrusion, Carl Hanser Verlag. 2001.
ACCEPTED MANUSCRIPT
3. D. G. Baird and D. I. Collias. Polymer processing principles and design. WileyInterScience Publi. 1998.
4. W.D. Kingery and M.C. McQuarrie. Thermal conductivity: I - concepts of
measurement and factors affecting thermal conductivity of ceramic materials. Journal of
RI
PT
SC
M
AN
U
9. W.N. dos Santos, J.N. dos Santos, P. Mummery, A. Wallwork. Thermal diffusivity of
polymers by modified Angstrm method. Polymer Testing 2010; 29: 107.
10. A.L. Schieirmacher, Wiedemann. Ann Phys. 1888; 34: 38.
11. E.F.M. Van Der Held, F.G. Van Drunen. A method of measuring the thermal
conductivity of liquids. Physics, 1949; 15(10):865.
12. W.N. dos Santos. O mtodo do fio quente: tcnica em paralelo e tcnica de
superfcie. Cermica 2002; 48(306):86.
TE
D
13. W.N. dos Santos, Thermal properties of melt polymers by the hot wire technique.
Polymer Testing 2005; 24, (7): 932.
14. W.J. Parker, R.J. Jenkins, C.P. Butter, G.L. Abbot. Flash method of determining
1679.
EP
thermal diffusivity, heat capacity, and thermal conductivity. J. Appl. Phys. 1961; 32:
AC
C
ACCEPTED MANUSCRIPT
HDPE
Braskem
HI760UV
Crystalline
melting
temperature
Tm (oC)
136
LDPE
Braskem
PB608
103
PP
Braskem
H503
163
Nylon-6
BASF
Ultramid 8200HS
223
PVDF
Atochem
F4000HD
169
PMMA
Unigel
Acrigel
HIPS
BASF
PS495F
PS
BASF
PS145D
PC
Sabic
Lexan grade
Glass
transition
temperature
Tg (oC)
-135
60
28
-18
38
47
30
-35
49
112
93
90
145
SC
-125
M
AN
U
TE
D
EP
AC
C
Crystallinity
(%)
RI
PT
Resin
Supplier
Code
Material
ACCEPTED MANUSCRIPT
Figure Captions
RI
PT
SC
M
AN
U
Figure 6. Thermal conductivity as a function of temperature for PS, PC, PMMA and
HIPS (the arrows indicate the Tg values for each one of the polymers).
Figure 7. Schematic variation of modulus of elasticity as a function of temperature for
semicrystalline and amorphous polymers.
Figure 8. DSC curves for PMMA: heating rate 10 C/min; 2nd heating after the specimen
AC
C
EP
TE
D
ACCEPTED MANUSCRIPT
Figure 1
20000
15000
10000
5000
0
25
50
75
100
125
150
175
200
Temperature ( C)
-7
HDPE
2.5
2.0
1.5
1.0
0.5
25
50
75
100
125
150
175
200
M
AN
U
3.0
225
250
EP
1.5
1.0
0.5
0.0
AC
C
HDPE
2.0
TE
D
Temperature ( C)
2.5
25
50
75
100
125
150
0
Temperature ( C)
175
SC
RI
PT
HDPE
25000
200
225
ACCEPTED MANUSCRIPT
Figure 2
5000
4000
3000
2000
1000
40
60
80
100
120
140
160
180
200
LDPE
-7
1.6
1.4
1.2
1.0
0.8
0.6
25
50
75
100
125
150
175
0.4
LDPE
0.3
0.2
0.1
0.0
25
50
EP
200
TE
D
Temperature ( C)
M
AN
U
Temperature ( C)
75
100
125
150
AC
C
Temperature ( C)
175
SC
LDPE
6000
RI
PT
7000
200
ACCEPTED MANUSCRIPT
PP
8000
6000
4000
2000
75
100
125
150
175
200
M
AN
U
50
PP
Temperature ( C)
1,2
1,0
0,8
50
75
100
TE
D
0,6
25
125
150
175
200
Temperature ( C)
PP
EP
0,4
0,3
0,2
AC
C
0,5
0,1
25
50
75
100
125
150
0
Temperature ( C)
175
SC
10000
RI
PT
Figure 3
200
225
ACCEPTED MANUSCRIPT
9000
PA6
8000
7000
6000
5000
4000
3000
2000
1000
50
75
100
125
150
175
200
225
250
275
PA6
1,6
1,4
-7
M
AN
U
Temperature ( C)
1,2
1,0
0,8
0,6
50
100
TE
D
0,4
0
150
200
250
300
Temperature ( C)
0,60
PA6
0,50
EP
0,45
0,40
0,35
0,30
0,25
AC
C
0,55
0,20
0,15
50
100
150
200
0
Temperature ( C)
SC
RI
PT
Figure 4
250
300
ACCEPTED MANUSCRIPT
RI
PT
Figure 5
9000
8000
PVDF
6000
5000
4000
3000
SC
7000
2000
1000
0
50
75
100
125
150
175
PVDF
-7
1,2
200
M
AN
U
Temperature ( C)
1,0
0,8
25
50
75
100
TE
D
0,6
125
150
175
200
225
Temperature ( C)
PVDF
EP
0,6
0,5
0,4
0,3
AC
C
0,7
0,2
0,1
25
50
75
100
125
150
0
Temperature ( C)
175
200
225
ACCEPTED MANUSCRIPT
Figure 6
0.35
SC
0.30
0.25
0.20
0.15
0.10
20
M
AN
U
0.40
RI
PT
PS
PC
PMMA
HIPS
0.45
40
60
80
100
120
0
AC
C
EP
TE
D
Temperature( C)
140
160
180
ACCEPTED MANUSCRIPT
AC
C
EP
TE
D
M
AN
U
SC
RI
PT
Figure 7
ACCEPTED MANUSCRIPT
-1,50
86 C
st
1 run
nd
2 run
PMMA
-2,00
SC
-2,25
-2,50
-2,75
-3,00
-3,25
85
90
95
M
AN
U
-1,75
RI
PT
Figure 8
100
105
110
115
AC
C
EP
TE
D
Temperature ( C)
120
125
130