Analysis and Optimization of the

Silicone Molding Process Based

on Numerical Simulations and
Experiments
LUKASZ MATYSIAK
ABB Corporate Research, Starowislna 13a, 31-038 Krakow, Poland

XAVIER KORNMANN
ABB Corporate Research, Segelhof 1, CH-5405 Baden-Daettwil, Switzerland

PIOTR SAJ, ROBERT SEKULA

ABB Corporate Research, Starowislna 13a, 31-038 Krakow, Poland
Received: October 13, 2011
Accepted: February 14, 2012

ABSTRACT: Injection molding of liquid silicone rubbers (LSR) is widely

utilized in the production and insulation of many medium- and high-voltage
electrical components. It is dictated mainly by the excellent properties of LSR,
ensuring high quality of the final product. However, this also requires a
well-established and controlled manufacturing process. Typically, experimental
methods are used to get a detailed knowledge regarding the process and to gain
the necessary experience. Unfortunately, such an approach is expensive and time
consuming, and minimal marketing time is one of the most important
requirements for being competitive. In this connection, an advanced silicone
molding simulation tool was successfully developed and validated
experimentally. Furthermore, computer simulations were applied to analyze and

Correspondence to: Lukasz Matysiak; e-mail: lukasz.matysiak@

pl.abb.com.

Advances in Polymer Technology, Vol. 32, No. S1, E258–E273 (2013)


C 2012 Wiley Periodicals, Inc.
 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

optimize the process for a selected product, avoiding trial and error methods with
the only cost connected to the computational time. An accurate representation of
the process required a mathematical approach, among others, consisting of
viscosity, pVT, and curing kinetics models that were built up based on rheological,
pressure, and calorimetric measurements conducted for a considered silicone
material.  C 2012 Wiley Periodicals, Inc. Adv Polym Techn 32: E258–E273, 2013;

View this article online at wileyonlinelibrary.com. DOI 10.1002/adv.21272

KEY WORDS: Calorimetry, Injection molding, Rheology, Silicones,

Simulations

is based on computational fluid dynamics (CFD) and

Introduction can be applied to analyze the process parameters as
well as product and mold design with the help of

I njection molding of liquid silicone rubbers (LSR,

also called silicone molding or LSR molding) is
a common technology used for the production of an
computer simulations. This allows one to find an
optimized solution without the financial expenses
characterizing trial and error methods, because the
only cost is the computational time; consequently, a
outdoor insulation for medium- and high-voltage
shorter time is needed to prepare and to perform the
electrical components. Surge arresters, bushings, cir-
real experiments.
cuit breakers, instrument transformers, and capaci-
The complicated nature of the silicone molding
tors can be quoted as examples.1 It is dictated mainly
process also affects its numerical modeling. In this
by the excellent properties of LSR ensuring a high
connection, a mathematical model applied in com-
quality of the final product. Among the most impor-
puter simulations must take into account many as-
tant features of this elastomer, a very high chemical
pects such as multiphase mass flow, conjugate heat
and thermal stability should be emphasized. This
transfer, or pressure build-up due to thermal expan-
manifests in material hydrophobicity, UV stability,
sion of the silicone material. In addition to that, sil-
as well as flashover and erosion resistance.1,2 An ad-
icone is known by its specific rheological properties
ditional advantage of silicone insulation is its low
manifest in a strong influence of the temperature and
weight and nonbrittle character, which reduces the
shear rate on the material viscosity. Furthermore,
risk of shed damage during transport, installation,
silicone rubber belongs to the group of thermosets
and service and minimizes the possibility of scatter-
and, hence, the kinetics of polymerization reaction
ing of pieces that may be dangerous in the case of
has to be determined to enable tracing of the course
explosion.
of a cross-linking phenomenon. Consequently, one
The above-mentioned properties of LSR are not
has to conduct pressure, rheological, and calorimet-
the only factors affecting the quality of insulation
ric measurements to supplement the mathematical
and consequently of the final product. The proper
model consisting of Navier–Stokes equations (mo-
course of the silicone molding process is also cru-
mentum, continuity, and energy equation), a multi-
cial. In this connection, much attention must be paid
phase (volume of fluid, VOF)4 model with viscous,
to the choice of the most effective process parameters
pressure–volume–temperature (pVT), and curing ki-
both from the product reliability and manufacturing
netics models to also reliably represent these specific
economy point of view. Typically, such optimization
silicone properties in numerical simulations.
involves a deep knowledge and experience regard-
ing the process itself that can be gained in real ex-
periments. However, they are very often expensive
and time consuming, and minimal marketing time
is one of the most important requirements for being
competitive.
Silicone Molding Technology
In this connection, an advanced simulation tool
capable of predicting the course of a complex LSR LSR molding is a multistep process3 and is pre-
molding process was successfully developed and sented schematically in Fig. 1. The very first op-
validated experimentally.3 The proposed approach eration is the preparation of a mixture usually

Advances in Polymer Technology DOI 10.1002/adv E259

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 2. Exemplary clamping machine and mold for

silicone molding.

uum conditions. The injection time depends on the

injection pressure determined by the product to be
manufactured and the silicone properties like vis-
cosity depending on both a shear rate arising during
the fluid flow and temperature.
Once the mold is filled with silicone, the mold
temperature is increased to accelerate the cross-
FIGURE 1. Schematic course of silicone molding linking process of silicone, resulting in the mate-
process. rial transformation from liquid to solid state—it is
the so-called curing stage. The time of the curing
stage is specific for a given product. It is another
essential processing step, because the material over-
consisting of two silicone components, pigment, and heating and significant pressure increase inside the
optional additives (e.g., filler), depending on the ap- mold can be observed as a result of the exother-
plication. Within this step, a two-stage mixture ho- mic effect of a curing reaction and high thermal ex-
mogenization is conducted and can be combined pansion of silicone. The final operation is an initial
with the temperature stabilization system for a bet- product cooling inside the mold being followed by
ter control of silicone temperature (ambient temper- a product demolding and final cooling in ambient
ature or silicone preheating). In the meantime, inter- conditions.
nal parts of a product (so-called inserts), which are It is worth stressing that all process parame-
embedded in the silicone coating after the injection ters mentioned above, that is, mixture composition,
process, are placed in a mold. These inserts can be temperature of inserts, mold, and injected silicone,
either kept in ambient conditions prior to injection clamping force, injection pressure, curing time, lo-
or preheated. The mold is usually heated by water cation of injection and ventilation channels and/
flowing inside the mold or by electrical heaters. In or injection pressure, as well as silicone flow prop-
addition, a cold runner system can be used to pre- erties (the mentioned dependence between viscos-
vent the premature curing of inflowing silicone in ity and shear rate and temperature) strongly af-
the injection channel or its vicinity. In the subsequent fect the course of silicone molding. Consequently,
step, the mold is closed by a clamping machine as all these aspects can be subjected to optimization
presented in Fig. 2, and the injection process starts. It to ensure the shortest production cycle time and
is conducted either under atmospheric pressure (an simultaneously keeping the highest final product
efficient ventilation system is required) or in vac- quality.

E260 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

CURING KINETICS MODEL

Mathematical Modeling of the
Another extremely important property of a sil-
Silicone Molding Process icone is its thermosetting nature. This means that
the material undergoes irreversible transformation
Computer simulations have been recognized as from the liquid phase to the solid state due to a
a powerful tool for analysis of many processes, in- chemical cross-linking reaction initiated and deter-
cluding industrial problems. The recent progress in mined by heat. One of threats related to this pro-
computer hardware and in numerical algorithms al- cess is an inseparable exothermic effect that might
lows one to simulate in three dimensions even the lead to local silicone overheating and consequently
most complex issues like silicone molding with an to the degradation of material properties. The proper
acceptable computational time. curing front propagation is also of vital importance.
The approach for numerical modeling of the sil- Both these issues can be analyzed by means of nu-
icone molding process was developed on the basis merical modeling; however, it requires an accurate
of a commercial CFD package ANSYS Fluent uti- curing kinetics model. One of the most universal ex-
lizing the finite volume method to solve numeri- pressions describing reliably the change in the state
cally the governing equations describing the prob- of matter of thermosetting materials such as silicones
lem under consideration.4 The core of a numerical and epoxy resins is Kamal’s equation7,8 :
model constitutes Navier–Stokes equations, includ-
ing continuity, momentum, and the energy equation dα/dt = A1 exp(−E 1 /RT)
together with the multiphase VOF model, which are
+ A2 exp(−E 2 /RT) α m (1 − α)n
comprehensively described, for example, in Ref. 3 or
4. These substantial equations were supplemented
with additional user-defined formulas to reflect the where dα/dt represents the curing rate, A1 and A2
specific silicone properties and in this way to repre- are the preexponential factors, E 1 and E 2 are the
sent the process in a more reliable way. Among the activation energies, R is the universal gas constant, m
implemented expressions, a viscous model, curing and n are the model constants, and α is the degree of
kinetics model, and pVT model can be found. curing informing about how advanced is the curing
process (α = 0 for liquid material, and α = 1 means
solid state).
VISCOUS MODEL The degree of curing is calculated according to
the following equation:
The viscous model was taken into account, be-
cause a non-Newtonian character of a silicone and
strong dependence between its viscosity, tempera- α = H(t)/H
ture, and the shear rate was reported by several
authors.5,6 Moreover, this relationship can influence where H(t) is the amount of heat released during
directly the silicone molding process, because low the curing reaction up to time t, and H is the total
viscosity together with high shear rate can lead to amount of heat released during the curing reaction.
void formation. The proposed viscous model allows It is worth stressing that all parameters of Kamal’s
one to analyze the influence of this silicone prop- equation, that is, A1 , A2 , E 1 , E 2 , m, and n, together
erty on processability and to eliminate the potential with the total amount of heat generated as an effect
problems already at the simulation stage. The math- of the curing process (H ), were determined experi-
ematical function expressing this dependence was mentally by using differential scanning calorimetry.
elaborated on the basis of rheological measurements In turn, tracing of the curing progress in tran-
done for different materials and at temperatures and sient simulations was possible thanks to an addi-
shear rates. According to the viscous model, the sil- tional unsteady-state conservation equation solved
icone viscosity μ is calculated as follows: for the degree of curing:

μ = exp(a 1 + a 2 ln γ + a 3 ln(T)2 )
∂(ρα)/∂t + ∇(ρwα) = Sa
where γ stands for the shear rate, T is temperature;
a 1 , a 2 , and a 3 are the model parameters characteristic where ρ is the fluid density, w is the vector of velocity,
for the given material and determined experimen- and Sa is the source term corresponding to the curing
tally. rate dα/dt.

Advances in Polymer Technology DOI 10.1002/adv E261

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

Finally, the exothermic effect of the curing reac- TABLE I

tion was considered by adding the source term ST Coefficients of the Viscosity Model Determined
into the energy equation: Experimentally
Coefficient Value
ST = Sa H
a1 5.765
a2 −0.354
pVT MODEL a3 −0.079
LSR are also known for their high thermal ex-
pansion, which results in a significant pressure
build-up during the curing stage of the silicone
molding process, when the mold is closed and the
material is heated up to enhance the polymerization
phenomenon, also producing additional heat. The
information about the pressure increase during the
process can be useful when deciding about the nec-
essary clamping forces. The effect of pressure rise
was taken into account in simulations by introduc-
ing the pVT model, coupling pressure p with volume
V, and temperature T, on the basis of the equation
FIGURE 3. Experimental stand used in the viscosity
of state. The expression describing the relationship
measurements.
between all three state variables was developed on
the basis of the pressure measurements performed
for the silicone sample closed in a heated mold, en-
suring constant material volume. A simple polyno- used in these measurements is depicted in Fig. 3. A
mial function turned out to fit the experimental data rheometer equipped with a torque measuring sys-
accurately: tem was applied to sense the resistance to the rota-
tion of a cone caused by the presence of silicone in
p = b1 T 2 + b2 T + b3 between the rotating cone and a fixed plate. This re-
sistance produced a torque proportional to the shear
where b 1 , b 2 , and b 3 are the model constants. rate in fluid, whereas the viscosity was calculated
from the known geometric constants of the cone,
the rotational speed, and shear stress-related torque.
Different temperatures were ensured by heating or
Rheological, Pressure, and cooling of the fixed plate.
The obtained results confirmed a strong correla-
Calorimetric Measurements tion between the silicone viscosity and shear rate
and temperature as shown in Fig. 4. It can be ob-
One can notice that all additional models dis- served that viscosity tends to have high values for
cussed previously, that is, viscous, pVT, and cur- low shear rates and for higher shear rates lower vis-
ing kinetics models, include coefficients that have to cosity should be expected. Similar dependence was
be determined experimentally for a specific mate- detected in the case of the temperature, that is, at the
rial under consideration. Consequently, three inde- higher temperature the lower viscosity is observed.
pendent experiments were conducted to investigate Table I presents the viscous model coefficients that
properties of silicone. were calculated on the basis of the performed
measurements.

RHEOLOGICAL EXPERIMENTS
CALORIMETRIC EXPERIMENTS
Viscosity measurements were done for a chosen
silicone material at several temperatures and shear The measurements of curing kinetics were con-
rates. This enables one to find the relationship de- ducted by using DSC, which is commonly used for
scribing the influence of the shear rate and tempera- the thermal investigation of a broad range of mate-
ture on the silicone viscosity. The experimental stand rials, including thermosetting silicones. One of the

E262 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 4. Results of the viscosity measurements.

DSC applications is curing kinetic analysis that can from the ones elaborated on the basis of isothermal
be done either by using a dynamic or isothermal DSC. A deeper study of these results revealed that
method. In dynamic DSC, the sample is usually the correlation between the experimental data (solid
heated starting from ambient conditions with dif- lines in Fig. 6) and the data predicted by the cur-
ferent heating rates, for example, 2, 5, or 10 K/min. ing kinetic models (dotted lines in Fig. 6) was better
In turn, in isothermal DSC several experimental runs for dynamic measurements. In addition to that, the
are performed and the sample temperature is kept dynamic measurements tend to be closer to the re-
constant at different levels (e.g., 100, 120, or 140◦ C) ality, because the material is heated up likewise in
during each such run.9 In both DSC methods, the the production process. All this led to the conclusion
heat flow into or from a sample is recorded as de- that the results of dynamic measurements are valid
picted in Fig. 5. ones.
The results of dynamic and isothermal calorimet-
ric measurements performed for a selected silicone
PRESSURE EXPERIMENTS
are presented in Fig. 6. Generally, it can be concluded
that the course of the curing reaction is strongly de- The mentioned pVT model was developed on the
pendent on the heating rate and temperature. The basis of pressure measurements performed for a sil-
coefficients of the curing kinetics model were calcu- icone sample closed in a tight steel mold. The mold
lated by using nonlinear regression separately based was placed in a laboratory oven as illustrated in
on dynamic and on isothermal runs. Their values Fig. 7 and heated up to force the thermal expansion
are gathered in Table II. One can notice that the cur- of silicone. Expanding silicone raised the pressure
ing kinetics model and the total heat of the reaction inside the mold, which was recorded by the assem-
obtained by using dynamic DSC data are different bled pressure sensor.

FIGURE 5. Concept of DSC measurements.

Advances in Polymer Technology DOI 10.1002/adv E263

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 6. Results of the DSC measurements.

TABLE II TABLE III

Coefficients of the Curing Kinetics Model Determined Coefficients of the pVT Model Determined
Experimentally Experimentally
Value Coefficient Value

Coefficient Unit Dynamic Isothermal b1 −1424.87

b2 1,664,054.35
A1 1/s 2.30E+11 0.00 b3 379,492,473.03
A2 1/s 5.82E+13 1.48E+20
E1 K 13024.26 0.00
E2 K 13024.26 18659.01
The effect of pressure build-up during the de-
m – 1.00 0.75
n – 1.23 0.48
scribed experiment is presented in Fig. 8. It can
H J/kg 10029.44 6656.33 be seen that the maximum measured pressure was
above 300 bars. Further measurements turned out to
be troublesome due to leakages, which are reflected
in Fig. 8 in the final pressure. Nevertheless, the per-
formed experiments confirmed high thermal expan-
sion of silicone and resulted in pressure increase dur-
ing the curing stage of the silicone molding process.
Moreover, the experimental data allowed to deter-
mine the coefficients of the pressure model as shown
in Table III.

Process Optimization in
Computer Simulations

The complexity of the silicone molding process

has already been discussed. There are a number of
parameters that can be modified to find the optimum
setup of the process from the manufacturability
viewpoint. The prevailing part of aspects affecting
the final product quality and production cycle time
(reflected in cost and production capacity) can be
analyzed and optimized in computer simulations.
FIGURE 7. Experimental stand used in the pressure Table IV gathers all process scenarios that were
measurements. investigated by the application of numerical

E264 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 8. Results of the pressure measurements.

modeling in ANSYS Fluent for an exemplary prod- and to accelerate in this way the numerical compu-
uct manufactured by using silicone injection mold- tations. The hybrid numerical mesh (illustrated in
ing technology. The considered topics were number Fig. 10) was applied during the geometry discretiza-
and location of injection and ventilation channels, tion into the finite number of elements. In total, it
temperature of injected silicone, temperature of in- consisted of ∼1.4 million tetrahedral and hexahe-
ternal product parts prior to injection, temperature dral elements. For the sake of simplicity, all required
of mold during filling and curing stage, and finally material properties (except the silicone viscosity cal-
injection time. culated according to the mentioned viscous model)
were assumed to be constant as presented in Table V.
PRODUCT UNDER INVESTIGATION
A simplified hollow-core insulator enclosed in a
SIMULATION RESULTS
steel mold was selected as a subject of optimization.
The geometrical model of the product is presented The starting point for the optimization was the
in Fig. 9. One can notice that the product symmetry results obtained for the basic process configuration,
enables one to consider only half of the geometry that is, Case 01. It can be observed in Fig. 11 that the

FIGURE 9. Simplified geometry of the hollow core insulator.

Advances in Polymer Technology DOI 10.1002/adv E265

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

during Curing Stage

Mold Temperature

130
130
130
130
130
140
during Injection Stage
Mold Temperature

100
80
80
80
80
80 FIGURE 10. Numerical mesh applied in computer
simulations.
Injected Silicone
Temperature of

simulation revealed a potential problem with the air

gap formation during the filling stage. Such voids
25
25
25
25
60
60

in the silicone insulation are unacceptable, because

Different Scenarios of the Silicone Molding Process Investigated in the Numerical Simulations

they could cause partial discharges during the oper-

ation of an electrical device. In this connection, it was
decided to modify the ventilation system by intro-
Inserts Temperature

ducing two additional channels in the troublesome

Prior to Injection

areas—Case 02. This change avoided the described

problem.
25
25
25
70
70
70

In the next step, the possibility of shortening of

the filling stage was analyzed. For this purpose, the
injection system was developed by two additional
injection channels (Case 03) with the same injection
parameters as in the case of the main injection chan-
Injection

104.16
104.16
34.38
34.38
34.38
34.38

nel in Case 01. Such a multiinjection system affected

Time

the silicone flow pattern as can be seen in Fig. 12,

but the filling process remained trouble free and the
filling time was reduced by 66%.
Injection Channels

The next aspect under consideration was the opti-

Number of

mization of the curing process. As it was mentioned,

the curing reaction is driven by the amount of heat.
1
1
3
3
3
3

Consequently, several methods that increased the

energy delivered to the silicone and accelerated the
polymerization process were investigated. Among
them, the following issues were considered in the
Ventilation Channels

subsequent steps: inserts preheating (Case 04), sil-

icone preheating prior to injection (Case 05), and
Number of

higher mold temperature during the filling and cur-

20
22
22
22
22
22

ing stage (Case 06). Temperature and degree of cur-

ing were observed both in the whole fluid domain
as well as at several monitoring points depicted in
Fig. 9. The comparison of results obtained for all
TABLE IV

investigated scenarios is presented in Figs. 13–21.

Scenario

Case 01
Case 02
Case 03
Case 04
Case 05
Case 06
Process

It can be concluded on the basis of the tempera-

ture data that the preheating of inserts and silicone
as well as higher mold temperature also increased

E266 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

TABLE V
Material Properties Defined in the Computer Simulations
Material Density (kg/m3 ) Heat Capacity (J/kgK) Thermal Conductivity (W/mK) Viscosity (Pas)

Air 1.225 1006.43 0.0242 1.7894E-5

Silicone 1060 1230 0.22 Viscous model
Steel 7750 460 28.5 NA
Aluminum 2680 963 150 NA
Plastic 2000 980 0.3 NA

FIGURE 11. Solution of the air gaps formation problem.

FIGURE 12. Flow pattern of silicone during the mold filling stage depending on the injection system.

Advances in Polymer Technology DOI 10.1002/adv E267

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

the silicone temperature during the molding pro-

cess. The highest temperatures were recorded for the
simultaneous inserts and silicone preheating with
an increased mold temperature. In addition to that,
it can be noticed that none of the investigated sce-
narios were affected by the problem of tempera-
ture hot spots, although a characteristic tempera-
ture peak related to the curing reaction was detected
and can be relatively easily recognized in Figs. 13–15
(at t =∼8 min for Cases 04 and 05 and at t = ∼5 min
for Case 06).
The degree of curing is directly connected with
the temperature data, because the curing reaction
is determined by heat. It can be seen in Figs. 17–21
that the fastest curing process was observed for the
solution combining higher mold temperature with
the inserts and silicone preheating (Case 06). Ac-
cording to Fig. 20, the minimum degree of curing
in the whole silicone domain was 95% (a conven-
tional limit, signalizing a fully cured material) for
Case 06, about 14 min earlier than the reference Case FIGURE 14. Temperature computed at point 7 for
different process scenarios.
01. It means that the process time could be reduced
by about 55% simply by modifying the process
parameters.
It should be emphasized that the presence of air
gaps and hot spots as well as the length of the cy-
cle time should not be the only factors influencing
the process assessment. Equally important is the
proper curing front propagation, because the pre-

FIGURE 15. Temperature computed at point 9 for

different process scenarios.

mature gelation could result in an incomplete filling

of the mold cavity with the injected material. In this
connection, the maximum degree of curing in the
silicone volume was also thoroughly studied on the
FIGURE 13. Temperature computed at point 5 for basis of Fig. 21. These results revealed that the fluid
different process scenarios. material transformed into the solid state after the

E268 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 16. Maximum temperature computed in the FIGURE 18. Degree of curing computed at point 7 for
fluid domain for all analyzed cases. different process scenarios.

FIGURE 19. Degree of curing computed at point 9 for

FIGURE 17. Degree of curing computed at point 5 for different process scenarios.
different process scenarios.

gions the material remains the longest in the fluid

end of the filling stage for all investigated cases, so state. Such a representation of results obtained for
the threat of incomplete mold filling was excluded. Case 06 can be found in Fig. 22. It can be noticed that
The progress of the curing reaction can also be the curing reaction began in the considered scenario
analyzed on the basis of three-dimensional (3D) vi- in the neighborhood of the silicone sheds due to the
sualization. This enables observing exactly which efficient heating of these regions by the mold and
silicone portions cure the fastest and in which re- then proceeded in the direction of the colder plastic

Advances in Polymer Technology DOI 10.1002/adv E269

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 20. Minimum degree of curing for different FIGURE 21. Maximum degree of curing for different
process scenarios. process scenarios.

tube. This again confirmed the strong correlation be- during the flow that make the process analysis even
tween the curing rate and temperature determined more problematic.
by the process parameters. A similar conclusion can In this connection, an advanced 3D approach for
be drawn on the basis of Fig. 23, which presents dif- numerical modeling of silicone molding was de-
ferent degrees of curing registered for the analyzed veloped, validated, and proposed as a tool for the
scenarios after around 10 min counting from the pro- analysis and optimization of this complex process.
cess start. For example, in Case 06 almost the whole The standard mathematical equations describing the
amount of silicone was cured after around 10 min of multiphase fluid flow and conjugate heat transfer
processing, whereas in Case 01 a significant portion were supplemented with pressure, viscous, and cur-
of the material was still in the liquid state at the same ing kinetic models to reflect the mentioned silicone
moment. properties and simulate the process in a more re-
alistic way. This required three independent experi-
ments, that is, rheological, pressure, and calorimetric
measurements, to capture the unique silicone char-
acteristics and to determine the material-dependent
Summary coefficients present in each of the models listed
above. A high-pressure build-up was observed dur-
This article deals with the silicone molding tech- ing the silicone heating in a closed volume, which
nology widely used on the industrial scale for the confirmed high thermal expansion of the silicone.
production of silicone insulation for electrical appli- The rheological experiment was conducted for dif-
cations. Many parameters affecting the final product ferent temperatures and shear rates, and the results
quality and the process complexity were discussed. revealed a strong correlation between these variables
Among them, the mold temperature, injection pres- and the silicone viscosity. Finally, the calorimetric
sure reflected in the mold filling time, silicone tem- measurements allowed recognition of the kinetics of
perature, inserts temperature or design of injection, the silicone cross-linking reaction and defined the
and ventilation systems can be found. In addition exothermic effect related to this process.
to that, it was pointed out that the silicone material The experimentally developed models were in-
is known for its high thermal expansion, exother- troduced by using the numerical simulations of
mic polymerization, and non-Newtonian behavior the silicone molding process. The goal of the

E270 Advances in Polymer Technology DOI 10.1002/adv

 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 22. Curing front propagation observed for Case 06.

computations was twofold: analysis and improve- recognition of regions with possible air void for-
ment of the silicone insulation forming process for mation, which is an undesirable effect in the con-
the selected product. For this purpose, a simpli- text of the final product quality and reliability. The
fied hollow core insulator was chosen and a series correction of the ventilation system avoided this
of simulations for different process configurations problem. Apart from that, the threat of material
were performed. These scenarios corresponded to overheating was excluded and the filling time was
the parameters strongly affecting the process course shortened, thanks to the application of additional
that are mentioned above (number and location of injection channels. Moreover, the simulation results
injection and ventilation channels, injection time, showed the possibility of curing stage improvement.
temperature of inserts, silicone, and mold). The re- The introduction of inserts and silicone preheating
sults confirmed that the computer simulations can together with a higher mold heating temperature led
be successfully applied for the investigation and op- to a significant reduction in the time needed to carry
timization of such a complex process like silicone out the silicone polymerization process without the
molding. For example, the simulation data allowed risk of premature gelation of the material.

Advances in Polymer Technology DOI 10.1002/adv E271

10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

DOI 10.1002/adv
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

FIGURE 23. Different progress in the curing reaction depending on the process parameters.

Advances in Polymer Technology

E272
 10982329, 2013, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adv.21272 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [22/07/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SILICONE MOLDING PROCESS BASED ON NUMERICAL SIMULATIONS AND EXPERIMENTS

4. ANSYS FLUENT 13.0 Documentation; ANSYS Inc., Canons-

References burg, PA, 2010.
5. Winter, H. J. In Proceedings of the 1st Hedrich World Congress,
Frankfurt, Germany, April 21–22, 2008.
1. Stenstrom, L.; Gustavsson, D.; Windmar, D.; Gutman, I.; 6. Grunwald, M. In Proceedings of the 1st Hedrich World
Vosloo, W. L. In Proceedings of the 40th CIGRE Session, Paris, Congress, Frankfurt, Germany, April 21–22, 2008.
France, 2004.
7. Kamal, M. R.; Sourour, S. Polym Eng Sci 1973 13(1), 59–64.
2. Remote plant plays key role in ABB insulator busi-
8. Kamal, M. R.; Sourour, S.; Ryan, M. In Proceedings of the 31st
ness. INMR—Insul News Mark Rep Quart Rev 2005, 13,
Annual Technical Conference, Quebec, Canada, 1973, pp. 187–
54–61.
191.
3. Matysiak, L.; Saj, P.; Sekula, R. In Proceedings of the 5th Euro-
9. Kasza, K.; Matysiak, L.; Malinowski, L. Adv Polym Technol
pean Conference on Computational Fluid Dynamics ECCO-
2009, 28(4), 267–275.
MAS CFD 2010, Lisbon, Portugal, June 14–17, 2010.

Advances in Polymer Technology DOI 10.1002/adv E273