Silicone insulators of power transmission lines with a variable

inorganic load concentration: Electrical and physiochemical

analyses
Rodolfo Cardoso
a
, Ana Claudia Balestro
b
, Adriano Dellallibera
b
, Eduardo Coelho M. Costa
c,
,
Joceli Maria G. Angelini
d
, Lucia Helena Innocentini Mei
a
a
University of Campinas Unicamp, Faculty of Chemical Engineering, Campinas, SP, Brazil
b
Industria Eletromecnica Balestro Ltda., Mogi Mirim, SP, Brazil
c
Unesp Univ. Estadual Paulista, Faculdade de Engenharia de Ilha Solteira, Depto. de Engenharia Eltrica, Ilha Solteira, SP, Brazil
d
Centro de Pesquisa e Desenvolvimento em Telecomunicaes CPqD, Campinas, SP, Brazil
a r t i c l e i n f o
Article history:
Received 10 July 2013
Received in revised form 17 December 2013
Accepted 23 December 2013
Available online 7 January 2014
Keywords:
Silicone insulators
Electric tracking
Erosion test
Physiochemical measurements
Scanning electronic microscopy
Dispersive spectroscopy X-ray
a b s t r a c t
Polymeric insulation is an increasing tendency in projects and maintenance of electrical
networks for power distribution and transmission. Electrical power devices (e.g., insulators
and surge arresters) developed by using polymeric insulation presents many advantages
compared to the prior power components using ceramic insulation, such as: a better per-
formance under high pollution environment; high hydrophobicity; high resistance to
mechanical, electrical and chemical stresses. The practice with silicone insulators in pol-
luted environments has shown that the ideal performance is directly related to insulator
design and polymer formulation. One of the most common misunderstandings in the
design of silicone compounds for insulators is the amount of inorganic load used in their
formulation. This paper attempts to clarify how the variation of the inorganic load amount
affects physicochemical characteristics of different silicone compounds. The physicochem-
ical evaluation is performed from several measurements, such as: density, hardness, elon-
gation, tensile strength. In addition, the evaluation of the physicochemical structure is
carried out using infrared test and scanning electronic microscopy (SEM). The electrical
analysis is performed from the electric tracking wheel and erosion test, in agreement with
the recommendation of the International Electrotechnical Commission (IEC).
2014 Elsevier Ltd. All rights reserved.
1. Introduction
In the past thirty years, glass and porcelain insulators
have been substituted by a relatively new class of insulator
composed of polymeric compounds [1]. This technology
has been used in distribution and transmission systems.
The rst transmission line with composite insulators was
installed in Germany in 1969 [2]. Currently, the polymer
insulators represent more than 40% of the new insulation
market in the United States [3].
In Brazil, the replacement of ceramic to polymer insula-
tors began in 1987, because of the limited experience on
the electrical performance of these polymeric compounds
as function of the aging and physicochemical degradation
caused by environmental agents [3,4]. However, currently,
there is more than 50,000 km of transmission and distribu-
tion overhead lines, corresponding approximately to
400,000 new insulators [5].
Silicone, Terpolymer Ethylene and Propylene Diene
(EPDM) are materials used in the manufacturing of great
part of the polymeric insulators. However, the use of sili-
cone in the insulator housing has increased in the last
few years even more, mainly because of the great perfor-
mance in the hydrophobicity recovery [2,3,6,7].
0263-2241/$ - see front matter 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.measurement.2013.12.030

Corresponding author. Tel.: +55 1791080608.

E-mail address: [email protected] (E.C.M. Costa).
Measurement 50 (2014) 6373
Contents lists available at ScienceDirect
Measurement
j our nal homepage: www. el sevi er. com/ l ocat e/ measur ement
The analysis of the insulation coordination for overhead
transmission lines is not a trivial task. Specic issues
related to the environment characteristics, weather and
pollution should be taken into account. In regions with
high pollution levels, silicone insulators are more appropri-
ate (and also more expensive) than their similar-based
EPDM. The silicone insulators are also more appropriate
in areas with specic climatic conditions, e.g., sea coast
locations and tropical weather conditions [3].
The adequate physicochemical formulation, based on
polymer and other components, combined with an opti-
mized physical structure of the insulator body can signi-
cantly improve the insulation levels of transmission and
distribution networks. The variation in the polymeric
housing formulation results in 20% up to 80% on the weight
of the material. Some additives included in the polymeric
housing of insulators reduce manufacturer costs, facilitate
the processing of materials and also improve the perfor-
mance of these insulators [1,3].
Some studies have been conducted to evaluate the
aging and degradation process on polymeric insulators, in
order to establish standards for aging/degradation tests
in a reduced-time scale. The aging/degradation tests on
polymeric devices are possible by means of an environ-
ment with temperature, moisture and pollution controlled,
in order to induce an accelerated aging and degradation on
the samples [3,8,9]. Based on the technical literature, the
degradation caused by aging is responsible for approxi-
mately 65% of failures in polymeric insulators. This means
that the remaining 35% are caused from mechanical and
electrical stresses [3,4].
One of the most classical tests for assessing the perfor-
mance of insulators on severe environmental conditions
and polluted areas is expensive and long, i.e., approxi-
mately 5000 h [10]. However, a new test was proposed in
2004, recommended by the IEC 62217 [11], representing
a test procedure less expensive and faster than the prior
procedure with 5000 h. This test is known as roll of electri-
cal routing or electric tracking wheel.
The roll of electrical routing is a relatively new test,
which allows to compare the performance of different
insulators keeping the same material or to evaluate differ-
ent polymeric compounds keeping the same project.
This paper proposes a complete investigation of
silicone compounds with different concentrations of
inorganic loads using the roll of electrical routing. Physio-
chemical measurements were carried out to evaluate
several electrical and mechanical variations among the
silicone compounds. Therefore, results and conclusions
presented from the test procedure proposed in this paper
represent a great database and reference for further
researches and manufacturing processes of polymeric
insulators.
2. Silicone insulators
The silicone rubber compounds evaluated in the
proposed study are ve types, varying according to the
inorganic load concentration:
1. (S1) silicone rubber compound with HTV (High Tem-
perature Vulcanizing) fumed Silica and Alumina Trihy-
drate (ATH) zero parts.
2. (S2) S1 + 25 parties of fumed silica and ATH.
3. (S3) S1 + 50 parties of fumed silica and ATH.
4. (S4) S1 + 75 parties of fumed silica and ATH.
5. (S5) S1 + 100 parties of fumed silica and ATH.
The compounds used in this research were manufac-
tured by Bluestar Silicones Brasil Ltda.
3. Electric tracking wheel
The IEC 62217 describes two or four insulators charac-
terized by the same compound with creepage distance
between 500 mm and 800 mm, which are tested simulta-
neously in the stem routing. Before starting the test, insu-
lators are cleaned with deionized water. In sequence, the
samples are arranges in four positions, as illustrated in
Fig. 1. The rotation process consists in each sample stays
stopped around 40 s at each position. Each 90 rotation
step takes approximately 8 s [11].
The rst step of the cycle (Immersion Time, as described
in Fig. 1), the sample is immersed in a saline solution. The
following step (Following Time, in Fig. 1), the sample
receives an excess of saline ow, ensuring that slight mois-
ture on the surface induces discharges through the dry
bands formed in the third step. The Energized Time or third
step, the sample is energized with industrial-frequency
voltage for 40 s. In the last step (Cooling Time, in Fig. 1),
the sample is cooled after the overheating, resulted from
the surface discharges in the Energized Time.
Through the test procedure, the saline solution was
replaced during the weekly interruptions for inspection.
These interruptions along the test time did not spend more
Fig. 1. Schematic illustration of the electric tracking wheel.
64 R. Cardoso et al. / Measurement 50 (2014) 6373
than one hour. Interruption periods were not computed in
the total time of test.
The voltage considered in the electrical test was set by
dividing the leakage distance measure in the insulator by
28.6 mm. The concentration of NaCl in deionized water
was 2.80 kg/m
3
0.06 kg/m
3
. The chamber temperature
was set as 20 5 C and the total duration of the test was
15,000 cycles (800 h). The IEC 62217 determines that the
amount of NaCl for the saline solution must be 1.4 kg/m
3
and duration of the trial should be 30,000 cycles [11]. How-
ever, the proposed test was carried out with half of cycles
suggested by the IEC and the concentration of NaCl in
water was two times major than suggested by the norms.
There are no standard correlations between the techniques
presented in IEC 62217 and the proposed procedure; how-
ever, the results and measurements obtained in this re-
search are in agreement with those obtained following
the norms.
The conclusions obtained from this research were based
on measurements of density, hardness, elongation and ten-
sile of the silicone compounds of the insulators. The elec-
tric tracking and erosion test, infrared and scanning
electronic microscopy raised several issues related to the
physiochemical degradation of the silicone compounds.
4. Results and discussions
In this section, the methodology adopted for each test
procedure and following measurements are discussed sep-
arately as well as the evaluation criteria for each
measurement.
4.1. Wheel of electric tracking
One way to assess electric tracking in an insulator is to
submit it to a number of cycles to failure (Table 1). It is as-
sumed in this test that the device will not fail until the end
of the total 15,000 cycles. However, the methodology of
the IEC 62217 is not totally suitable to the proposed test,
considering the number of cycles, NaCl concentration and
also that the IEC standard is recommended for commercial
materials, which is not the case of the silicone compounds
studied in this research.
Table 1 shows the number of cycles before failure of
each insulator sample (14) and the average of number
of cycles for each compound (S1S5).
Best electrical performances were observed for the
compounds with a major concentration of ATH, as ob-
served in the average obtained from the number of cycles
of each compound, in Table 1. Otherwise, the worse perfor-
mance was observed for the insulators manufactured using
the compound S1, without ATH.
4.2. Density
The test to determine the density was carried out in tri-
ples. The mass and volume of three pycnometers were
measured in an analytical balance, using thermometer
and deionized water. In each pycnometer was introduced
20 g of well minced silicone rubber and the pycnometer
volume was lled with a liquid of well-dened density.
The microair bubbles in the material were removed with
a vacuum pump and a dessicator for few hours.
In sequence, the pycnometer volume was again ad-
justed with the same liquid and the temperature of the
system was xed at 25 C. After that, the total mass of
the pycnometer lled with rubber and uid was measured
again. Thus, calculating the difference in the masses of pyc-
nometer with rubber and vacuum and the pycnometer
with rubber and uid, the bulk density of the liquid leads
precisely to the rubber density.
Therefore, based on the given measurement process,
Fig. 2 shows the density values of the silicone compounds
S1S5.
Fig. 2 shows that the bulk densities increase with the
major concentration of loads in the silicone compounds.
The behavior observed in Fig. 2 was expected taken into
account that the ATH density is major than the Poly-
dimethylsiloxane (PDMS) which is the main constituent
the silicone rubber. Thus, it is natural that increasing the
amount of denser material in the compound, the bulk den-
sity of the material will be major as well.
4.3. Hardness
The hardness measures were obtained with a durome-
ter type Shore A, model Kori KR 14A in triplicate. The spec-
imens were 6 mm thick and vulcanized at 180 C for
10 min.
Fig. 3 shows the results of Shore A hardness test for sil-
icone compounds with different ATH quantities.
These results indicate that major quantities of load in
the compound results in major values of hardness. This
behavior was expected for polymer composites. However,
this measurement shows the exact value of the hardness
Table 1
Electric tracking cycles per insulator sample and per compound (S1S5).
1 2 3 4 Avg.
S1 5600 4327 6498 6498 5730
S2 5351 15,009 10,370 7250 9495
S3 9230 12,288 12,288 5532 9834
S4 10,350 16,531 10,350 11,060 12,070
S5 13,353 13,353 11,670 13,353 12,930
Density
1,09
1,36
1,54
1,67
1,77
0,0
0,5
1,0
1,5
2,0
S1 S2 S3 S4 S5
Formulations
D
e
n
s
i
t
y
(
g
/
c
m
3
)
Fig. 2. Roughening scheme of the specimens of the sub stage II-D.
R. Cardoso et al. / Measurement 50 (2014) 6373 65
for the specic compounds given by S1S5, according to
the amount of ATH in the formulations.
4.4. Maximum elongation and tensile strength
These tests were conducted under recommendations of
the ASTM D 412 standard [12]. The test specimens were
molded with 2 mm thick and vulcanized at 180 C for a
period of 10 min.
Seven specimens were obtained using Die C form
according to the standard [12]; after that, they were tested
in a universal testing machine Kratos, model K2001. The
load cell used in the test was 200 kgf and the speed was
500 mm/min.
To register very low elongation values, a strength gage
produced by Kratos was used.
The elongation of silicone compounds with different
loads is presented in Fig. 4. It is observed that the elonga-
tion varies inversely to the amount of ller in the compos-
ite. The interaction of ller with the rubber produces a
largest hardness, on the other hand, the same interaction
difcult the molecule rubber elongation.
Another important mechanical property of polymer
composites is their tensile strength. These measurements
are presented in Fig. 5.
From the results in Fig. 5, the tensile measurements for
the ve compounds did not show a linear behavior, as
observed in the prior analyses. A better performance was
observer for the specimen S2, with 25 parties of fumed
silica and ATH. The tensile strength performances of S3
S5 decrease with the increase of amount of silica and ATH.
Another observation is that the results of tensile
strength are inversely proportional to the hardness mea-
surements presented in Fig. 3. This means that with a
greater hardness, the compound presents a minor rupture
tensile strength.
4.5. Tear resistance
This trial was conducted under recommendations of the
ASTM D 624 standards [13]. The test specimens were
molded with 2 mm thick and vulcanized at 180 C for a
period of 10 min, such as for the elongation and tensile
tests.
Seven specimens were obtained using Die C form
according to the ASTM D 624 standards and tested using
a machine testing mechanical Kratos model K200, based
on the same conditions described in the elongation.
The tear resistance is one of the most important proper-
ties of the composites in insulators. Fig. 6 shows the
behavior of tear depending of the ATH amount in the
compounds. Similarly to the maximum elongation behav-
ior (Fig. 4), the tear resistance decreases with major
concentrations of ATH in the compound formulation.
Hardness - Shore A
48,3
61,5
73,0
82,3
87,3
0
20
40
60
80
100
S1 S2 S3 S4 S5
Formulations
H
a
r
d
n
e
s
s
(
S
h
o
r
e
A
)
Fig. 3. Shore A hardness of rubber with different ATH concentrations.
Stretching
450,57
383,94
198,00
112,57
69,66
0
100
200
300
400
500
S1 S2 S3 S4 S5
Formulations
S
t
r
e
t
c
h
i
n
g
(
%
)
Fig. 4. Maximum elongation of silicone rubbers with different ATH
concentrations.
Tensile Strength
56
72
67
59
45
0
20
40
60
80
S1 S2 S3 S4 S5
Formulations
T
e
n
s
i
l
e
S
t
r
e
n
g
t
h
(
k
g
f
/
c
m
2
)
Fig. 5. Tensile strength variation with ATH ller concentration for
composites rubber based.
Tear Strength
16
15
11
10
9
0
5
10
15
20
S1 S2 S3 S4 S5
Formulations
T
e
a
r
S
t
r
e
n
g
t
h
(
k
g
f
/
c
m
2
)
Fig. 6. Tear resistance of rubbers with different ATH concentrations.
66 R. Cardoso et al. / Measurement 50 (2014) 6373
4.6. Electrical tracking and erosion
The equipment used for this test was manufactured by
Balestro Ltda. The specimens were 6 mm thick and were
vulcanized at 180 C for a period of 10 min.
The specimens were sanded and tested according to
NBR 10296 standards [9].
The pollutant solution, with a resistivity of 4.05 Xm,
was prepared at a concentration of 0.1% of Ammonium
Chloride (NH
4
Cl) and 0.02% of wetting agent Isooctil
Phenoxypolietoxietanol.
The test started with a tension of 1 kV and was in-
creased 0.25 kV per hour [9]. The weights of the specimens
were measured before and after tests.
The trials of electrical routing and short period of ero-
sion in inclined plane provided many details regarding
the formulations studied in this work. The voltage values
for each compound versus the respective electrical routing
resistances are described in Fig. 7.
The results show that the increase of the ATH concen-
tration is proportional to the increase of the resistance of
electrical tracking. These results were previously observed
in reference [3]. The difference between this work was the
reference based on the NBR 10296 [9], whereas reference
[3] adopted the standards presented in ASTM 2303[14].
However, in fact, the NBR 10296 and ASTM D 2303 are
practically similar.
In Fig. 8, the maximum and minimum values show a
major variation for the compound S3, while the more
homogeneous result was veried for the compound S5.
Fig. 9 shows the mass lost average at the maximum and
minimum erosion, for each compound. It is possible to ver-
ify that the mass lost average is almost close to the average
value between the extremes of maximum and minimum,
except for the compound S5. The compound S5 showed a
deviation of the minimum mass loss, what demonstrates
that the mass loss was not similar for the ve specimens.
Fig. 10 shows the volume lost by erosion in the electri-
cal tracking test.
According to Fig. 10, the loss of volume by material ero-
sion was similar to the mass loss (Fig. 9). This behavior can
be explained from the same arguments to explain the mass
loss. These last two results show that only the voltage dis-
charge (when it occurs routing) is not enough for a reliable
evaluation of insulator compound. It is more appropriate to
combine the electrical tracking and erosion test with other
tests for more accurate conclusions, such as the mechani-
cal tests of elongation and tensile strength. However, even
so, the volume or/and mass erosion are important param-
eters for evaluation of silicone compounds.
4.7. Infrared FTIR
The spectra were carried out using an infrared spectro-
photometer Nicolet, model Magna-560, scanning range
of 4004000 cm
1
(infrared middle) and resolution of
4 cm
1
. The measures were conducted by diffuse reec-
tance (DRIFT) on the surface of the polymer. The best
denition was obtained with 128 scans.
Tracking
2,50
2,75 2,75
3,75
5,75
0
2
4
6
8
S1 S2 S3 S4 S5
Formulations
T
r
a
c
k
i
n
g
V
o
l
t
a
g
e
(
k
V
)
Fig. 7. Values of electrical tracking voltage and erosion according to NBR
10296 standard.
Tracking
2
3
4
5
6
7
S1 S2 S3 S4 S5
Formulations
T
r
a
c
k
i
n
g
V
o
l
t
a
g
e
(
k
V
)
Fig. 8. Minimum and maximum electrical tracking voltage.
Erosion - Mass Lost
1,137
2,175
1,747
3,326
0,831
0
1
2
3
4
5
S1 S2 S3 S4 S5
Formulations
M
a
s
s
L
o
s
t
(
g
)
Fig. 9. Mass lost by erosion in the electrical tracking test.
Erosion - Volume Lost
1,025
1,586
1,127
1,986
0,466
0
1
2
3
S1 S2 S3 S4 S5
Formulations
V
o
l
u
m
e
L
o
s
t
(
m
L
)
Fig. 10. Volume lost by erosion in the test of electrical tracking.
R. Cardoso et al. / Measurement 50 (2014) 6373 67
The infrared responses for the ve compounds are
described in Fig. 11, into the wavenumber range from
4000 to 400 cm
1
.
Reference [15] allocated bands composed of silicone ob-
served in commercial insulators. The absorption bands at
2960 cm
1
and 1270 cm
1
are related to CH and SiCH
3
400 800 1200 1600 2000 2400 2800 3200 3600 4000
Wavenumber (cm
-1
)
A
b
s
o
r
b
a
n
c
e
(
u
.
a
.
)S1
S2
S3
S4
S5
Fig. 11. Infrared spectra of rubbers with different ATH concentrations.
S1 S2
S3 S4
S5
Fig. 12. SEM micrographs of the new insulator compounds surfaces S1S5 (1000).
68 R. Cardoso et al. / Measurement 50 (2014) 6373
stretching, respectively, which are associated with CH
3
side groups. The absorption band at 1020 cm
1
is attrib-
uted to the SiOSi group of the polymer main chain.
The absorption range between 3700 and 3200 cm
1
is
related to OH groups, which are associated with the Alu-
mina Trihydrate ller (ATH).
Fig. 11 shows a more intense absorption at 3700 and
3200 cm
1
, which means a major concentration of ATH l-
ler in the compound. The absorption bands at 1270 cm
1
and 1020 cm
1
are more intense in the compound with
less ATH ller.
The variations observed around 800 cm
1
are associ-
ated with the SiC vibration stretching and CH
3
rocking
[15,16].
These results are important because demonstrate a
semi-quantitative nature of the infrared technique.
4.8. Scanning electronic microscopy SEM
The scanning of the compound by electronic micros-
copy shows several details on the molecular structure of
the ATH with the PDMS. Based on this information, the sil-
icone rubber quality can be evaluated as well as the valida-
tion of the results obtained from the physiochemical testes
presented in the previous sections.
The SEM of the cryogenic fracture is an important
procedure to observe the morphology and molecular struc-
ture into the insulator compound. Otherwise, only the
surface of the specimens could be evaluated, without a
proper analysis of the load concentration into the
specimens.
The cryogenic fracture process consists to immerse the
specimens in liquid nitrogen, which are cooled to the glass
transition temperature. Thus, the specimens can be frac-
tured without damages of their internal structure. Other-
wise, a fracture or cut at ambient temperature, using a
conventional cutting tool, deforms the internal structure
of the compound. This means that the load distribution
in the polymeric compound will be modied because of
the deformation resulted from the mechanical action of
the cutting tool, which makes the results obtained from
the SEM unreliable.
S1 S2
S3 S4
S5
Fig. 13. SEM micrographs of the cryogenic fractures of S1S5 (1000).
R. Cardoso et al. / Measurement 50 (2014) 6373 69
The morphology of new and aged rubber insulators was
tested using a Leica scanning electronic microscope (SEM),
model LEO 440i. The top surface and the internal surface
(cryogenic fracture) of the specimens were observed and
analyzed as follows.
In addition, the specimens were also analyzed using
Energy Dispersive Spectroscopy X-ray (EDX) in order to de-
tect other elements on the insulators surfaces.
The micrographs of external and internal surfaces and
the EDX spectra are shown in this section.
Fig. 12 shows the micrographs of the surface of new
insulators composed of S1S5, before the aging at the
wheel of routing. The micrograins composed of inorganic
llers (white points) are more abundant in compounds for-
mulated with higher concentrations of ATH. The inorganic
llers are also homogeneously distributed on the external
surface of the specimens, which proves the good quality
in the molding process of the specimens.
Fig. 13 shows the micrographs of the surfaces obtained
from the cryogenic fractures of the insulators before aging
at the wheel of routing.
The llers dispersion is more dened in the micro-
graphs in Fig. 13 than in Fig. 12. This dispersion is resulted
from the action of the liquid nitrogen on the compound
surfaces.
The images, in Figs. 12 and 13, show a major load con-
centration on the internal surfaces (by cryogenic fracture)
than on the external surfaces. This behavior shows that
the PDMS tends to involve the load llers, which results
a major load concentration into the silicone-rubber
insulators.
Fig. 13 shows that a major load amount results a minor
PDMS phase. However, results show that the progressive
load increase did not result a saturation or dissolution of
the compounds with major load concentrations, even in
S4 and S5.
The micrographs surfaces of the compounds S1S5 after
aging by tracking wheel are observed in Fig. 14.
In Fig. 14, the compounds S1, S2 and S5 did not present
signicant variations on the surface texture. On the other
hand, surface textures of S3 and S4 were degraded by ac-
tion of electrical discharges, which are generated by partial
formation of dry bands during the wheel test. Based on
these images and other images of S3 and S4, degradation
or structural damage were not veried for the ve
compounds.
S1
S2
S3
S4
S5
Fig. 14. SEM micrographs of the insulator compounds surfaces S1S5 after the wheel of electric tracking test (1000).
70 R. Cardoso et al. / Measurement 50 (2014) 6373
The micrographs of the aged surfaces, from the cryo-
genic fractures, are shown in Fig. 15. The micro-images
describe the same behaviors as in Fig. 14, especially for
compounds S3S5. Deep degradation of the polymer ma-
trix was observed, which exposed the inorganic loads of
compounds S3, S4 and S5.
In addition, the analyses from the EDX spectrum of new
compounds proved a signicant amount of Si atoms, as
described in Fig. 16.
The EDX spectra of the aging compounds by electric
tracking wheel are described in Fig. 17(a) for compound
S1 and (b) for S2S5, with different concentrations of
fumed silica and ATH.
The radiation monitoring scale is given in counts per
seconds (vertical scale). The horizontal scale is given in
keV.
The spectra of the Figs. 16(a) and 17(a) indicates the
presence of Si, O, C in the new compound S1. After the
wheel test; Si, O, Ca, C, Na and Cl were detected in S1. Thus,
the variation between the two spectra is the presence of
Na, Ca and Cl atoms which are only observed after the
aging procedure. This behavior is attested because of the
contamination of the compound by Sodium Chloride in
the immersion stage of the wheel process.
The EDX of the compounds S2S5, before and after the
electric tracking wheel test, are shown in Figs. 16(b) and
17(b), respectively. The presence of Al and Ca in the com-
pounds S2S5 was observed for the new and aged speci-
mens as well, because of the presence of ATH. The same
atoms are not present in the compound S1 because there
was not ATH in its composition.
In addition, Calcium is not observed in the new com-
pound S1. The presence of Calciumis characteristic in com-
pounds with llers containing Calcium Carbonate. This
means that Ca was not veried in S1 composition, since
Calcium Carbonate was not present before the aging pro-
cess on the compound S1, as described in Fig. 16(a).
By comparing the spectra in Figs. 16(b) and 17(b), after
the aging, the contamination by NaCl is expected. Other
components, Iron and Zinc (Fe and Zn), were detected from
the EDX test. The presence of Fe and Zn can be explained
because of the deterioration of the terminal insulators dur-
S1 S2
S3 S4
S5
Fig. 15. SEM micrographs of the cryogenic fracture of the aged compounds surfaces S1S2 after test by electric tracking wheel (1000).
R. Cardoso et al. / Measurement 50 (2014) 6373 71
ing the aging process. Zinc came probably from the electro-
plating and Iron from the steel.
5. Conclusions
From the physiochemical and electrical tests presented
in this research, the main conclusions are highlighted as
follows:
The analyses of density, hardness and mechanical tests
were efcient to characterize the different compositions
of silicone rubber, pointing important details on the
quality of the compounds.
The density is greater for compounds characterized by a
major amount of inorganic loads.
The hardness of the compound is higher with a major
concentration of ATH.
Fig. 16. EDX of the new compound S1 (a). EDX of the new compounds S2, S3, S4 and S5 (b).
Fig. 17. EDX of the compound S1 (a). EDX of the compounds S2, S3, S4 and S5 (b). After Aging process by electric tracking wheel.
72 R. Cardoso et al. / Measurement 50 (2014) 6373
The polymer elongation decreases with major ller con-
centrations, which means variations on the mechanical
properties of the insulator.
The resistance to tensile shows a nonlinear behavior as
function of the ller concentration, as veried for inter-
mediary quantities of llers, where the major tensile
resistances were measured.
Rip resistance is major for compositions with less ller.
The infrared test proved to be an efcient technique to
determine silicone rubber as well as semi-quantitative
technique to determine the amount of ATH llers.
The electric tracking test represents an important tool
to evaluate polymer compounds for tracking and
erosion.
According to the tracking test, rubber compounds with
major load concentrations show better electrical perfor-
mance than those formulated with less load amounts
(Table 1).
The SEM micrographs of the insulator surfaces and the
cryogenic fractures were effective to observe the mor-
phology and degradation regions of the compounds
after aging.
The EDX results showed that this analysis can deter-
mine qualitatively the presence of llers and also the
presence of contaminants, such as Sodium Chloride
and metals waste from the progressive degradation of
the polymer supported by metal accessories during
the test.
These conclusions are some of the general observations
obtained along the proposed physiochemical and electrical
methodology of test. Depending of the insulator design and
composition, specic observations and conclusions could
slightly differ of the results obtained from the insulators
evaluated in this paper. However, the general conclusions
on the procedure carried out in this research shows that
the use of electromechanical tests associated with the
chemical analysis using infrared, EDX and SEM techniques
represent an accurate methodology to evaluate the compo-
sition of silicone rubber insulators and their electrical
performances.
This research presented several specic technical char-
acteristics associated with most of silicone insulators, such
as: density, hardness, elongation, mass loss by erosion and
many other details that associate these physiochemical
prosperities to the load concentration in the compound
formulation. However, one of the major contributions pre-
sented in this paper is the test methodology for silicone
insulators. This methodology associates conclusions ob-
tained from mechanical tests (e.g. density, hardness, elon-
gation, etc.), physiochemical analyses (contamination
evaluation using Infrared, EDX tests and SEM images of
the microscopic structure of the silicone compounds) and
conventional electrical tests, such as the electric tracking
and erosion tests. All these test procedures follows the
most rigorous standards established by the IEC, the
American Society for Testing and Materials ASTM and
the Brazilian Association of Technical Standards ABNT.
The combination of these techniques results a complete
methodology to evaluate silicone rubber and polymer
insulators for research purposes as well as for applications
in power distribution and transmission networks.
Acknowledgement
This research received support of Industria Eletromec-
nica Balestro Ltda.
References
[1] J. Mackevich, M. Shah, Polymer outdoor insulating materials Part I:
comparison of porcelain and polymer electrical insulation, IEEE
Elect. Insulation Mag. 13 (3) (1997) 512.
[2] J. Kindersberger, A. Schtz, H.C. Krner, R.V.D. Huir, Service
performance, material design and applications of composite
insulators with silicone rubber housings, in: CIGR Session, SC 33
303, Paris, 1996.
[3] E.G.C. Gonzalez, Desenvolvimento de Compsitos de Silicone para
Uso em Isoladores de Linhas de Transmisso M. Sc. dissertation, Rio
de Janeiro, 1999.
[4] D.A. Silva, E.C.M. Costa, J.L. Franco, M. Antonionni, R.C. Jesus, S.R.
Abreu, K. Lahti, L.H.I. Mei, J. Pissolato, Reliability of directly-molded
polymer surge arresters: degradation by immersion test versus
electrical performance, Int. J. Elect. Power Energy Syst. 53 (2013)
488498.
[5] ELETROBRS, 1999, Programa decenal de transmisso. In: Plano
decenal de expanso, 19992008.
[6] D.A. Swift, Hydrophobicity of silicone rubber insulators: some
intriguiding experimental ndings, Colloquium CIGRE SC 33303,
3.2, Harare, Zimbabwe, 1995.
[7] G. Riquel, J.M. Formigue, D. Decker, R. Parraud, Natural and articial
ageing of non-ceramic insulators evaluation of diagnostic
techniques, in: 9th Inter. Symp. High Voltage Engineering, 31841/
4, Austria, 1995.
[8] N. Yoshimura, S. Kumagai, S. Nishimura, Electrical and
environmental aging of silicone rubber used in outdoor insulation,
IEEE Trans. Dielectrics Elect. Insulation 6 (1999) 632650.
[9] Norma NBR 10296, Material isolante eltrico avaliao de sua
resistncia ao trilhamento eltrico e eroso sob severas condies
ambientais, ABNT Associao Brasileira de Normas Tcnicas, So
Paulo, 1998.
[10] IEC 61109 Composite Insulators for AC overhead lines with a
nominal voltage greater than 1000V Denitions, test methods and
acceptance criteria, 1995.
[11] IEC 62217 Polymeric insulators for indoor and outdoor used with a
nominal voltage greater than 1000 V General denitions, test
methods and acceptance criteria, rst edition, 2005.
[12] ASTM D412 Standard Test Methods for Vulcanized Rubber and
Thermoplastic Elastomers-Tension, 2006.
[13] ASTM D624 Standard Test Method for Tear Strength of
Conventional Vulcanized Rubber and Thermoplastic Elastomers,
20002001.
[14] ASTM D 2303 Standard Test Methods for Liquid-Contaminant,
Inclined-Plane Tracking and Erosion of Insulating Materials, 2004.
[15] P.B. Blackmore, D. Birtwhistle, G.A. Cash, G.A. George, Condition
assessment of EPDM composite insulators using FTIR spectroscopy,
IEEE Trans. Dielectrics Electr. Insulation 5 (1) (1998) 132141.
[16] G.A. George, M. Celina, A.M. Vassallo, P.A. Cole-Clarke, Real-time
analysis of the thermal oxidation of polyolens by FTIR emission,
Polym. Degrad. Stab. 48 (1995) 199210.
R. Cardoso et al. / Measurement 50 (2014) 6373 73