Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 57 (2016) 241 246

49th CIRP Conference on Manufacturing Systems (CIRP-CMS 2016)

Autoclave cycle optimization

for high performance composite parts manufacturing
Luigi Nelea, Alessandra Caggianob,c,*, Roberto Tetia,c
a
Dept. of Chemical, Materials and Industrial Production Engineering, University of Naples Federico II, P.le Tecchio 80, 80125 Naples, Italy
b
Dept. of Industrial Engineering, University of Naples Federico II, P.le Tecchio 80, 80125 Naples, Italy
c
Fraunhofer Joint Laboratory of Excellence on Advanced Production Technology (Fh-J_LEAPT Naples), P.le Tecchio 80, 80125 Naples, Italy

* Corresponding author. Tel.: +39-081-7682371; fax: +39-081-7682362. E-mail address: alessandra.caggiano@unina.it

Abstract

In aeronautical production, autoclave curing of composite parts must be performed according to a specified diagram of temperature
and pressure vs time. Part-tool assembly thermal inertia and shape have a large influence on the heating and cooling rate, and
therefore on the dwell time within the target temperature range. When simultaneously curing diverse composite parts, the total
autoclave cycle time is driven by the part-tool assembly with the lower heating and cooling rates. With the aim to minimize the
autoclave cycle time and energy consumption improving the manufacturing system resource efficiency, a new parameter was
defined to characterize the part-tool assembly thermal and geometric properties. This parameter was applied to determine the
optimal positioning of the parts on the autoclave charge floor.
2016
2015TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier B.V.is an open access article under the CC BY-NC-ND license
B.V. This
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of Scientific committee of the 49th CIRP Conference on Manufacturing Systems (CIRP-CMS 2016).
Peer-review under responsibility of the scientific committee of the 49th CIRP Conference on Manufacturing Systems
Keywords: Autoclave curing; Composite parts; Optimization; Resource efficiency

1. Introduction time, which depends on the specific composite material.

In the aeronautical industry, small numbers of different
composite material parts need to be manufactured with high
quality and performance requirements. Autoclave vacuum bag
manufacturing technology, which entails applying carefully
controlled levels of heat and pressure to parts and assemblies,
is particularly suitable and largely employed for this purpose
[1].
In autoclave vacuum bag manufacturing, the quality of
advanced composite material parts strongly depends on the
thermal and pressure cycles imposed to the composite material
during the curing process [2,3].
Modeling and control of an autoclave cure cycle during the
processing of thermoset matrix composites have always been
challenging problems for manufacturers of high performance
composite materials, particularly in the case of high thickness
of the parts to be manufactured, as referred in Soo et al [4] and
Martinez [5]. The curing of each part must be performed
following a specified diagram of temperature and pressure vs
Once the thermal and pressure cycles have been selected
according to the composite material to be cured, the
manufacturing setup must ensure that, during the process, cycle
parameters such as part heating rate, cooling rate and dwell time
at constant temperature are respected [6].
As a consequence, when simultaneously curing diverse
composite parts having different thermal and/or geometric
properties, the total autoclave cycle time is driven by the part
with the lower heating and cooling rates. The resulting long
curing times negatively affect the manufacturing system
productivity and determine a very low energy efficiency.
In current aeronautical industry practice, planning of
autoclave manufacturing processes is still carried out by
grouping the composite parts to be cured together according to
their polymerization specifications (i.e. parts having the same
curing temperature), with very limited consideration for the
shape, size and material of which the parts are made, and no
consideration for the tools on which they are mounted.
In order to achieve higher resource efficiency and shorter

2212-8271 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 49th CIRP Conference on Manufacturing Systems
doi:10.1016/j.procir.2016.11.042
242 Luigi Nele et al. / Procedia CIRP 57 (2016) 241 246

cycle times, a new criterion for the effective aggregation of

different parts to be cured simultaneously and for the selection
of the optimal strategy for positioning the part-tool assemblies
on the autoclave charge floor is proposed in this work.
For this purpose, a new parameter, S, which allows to
characterize the geometry and thermal inertia of the part-tool
assemblies, is defined. Based on the value of the S parameter
for each part-tool assembly, it is possible to group all the
assemblies according to their thermal behavior in order to
simultaneously process in the autoclave harmonized groups
with similar parameter values. Moreover, the value of the S
parameter can be used to select the optimal positioning of the
parts on the autoclave charge floor. Fig. 1. Temperature/pressure vs time for an epoxy resin based prepreg.

2. Thermal behavior during autoclave curing of diverse

composite material parts

Autoclave curing of composite material parts must be

performed according to a specified diagram of temperature and
pressure vs time, which depends on the specific composite
material. In general, most aeronautical epoxy-based prepreg
systems require curing temperatures around 121 C or 180 C,
and the application of adequate pressure assisting the
consolidation, eliminating the excess resin and helping suppress
voids in the laminates. Fig. 1 shows the autoclave curing cycle
for an epoxy resin based prepreg.
For a successful curing process, the manufacturing setup
must ensure that the following cycle parameters are controlled
during the process [6]: heating rate and cooling rate, to be kept Fig. 2. Thermal profile of part-tool assemblies (coloured lines) and autoclave
(black line) during a curing cycle.
lower than critical values, and dwell time within the curing
temperature range (e.g. 179.4 5.6 C in Fig. 1), that should be
The interval of time elapsed from the instant at which the
at least equal to the recommended duration.
autoclave environment reaches the cure temperature and the
When simultaneously curing diverse composite parts having
instant at which the last part-tool assembly in order of time
different thermal and/or geometric properties, in order to satisfy
reaches the curing temperature can be defined as the heating
the process requirements for all parts, the total autoclave cycle
delay of the entire autoclave charge, RH (Fig. 2).
time must be fixed on the basis of the behavior of the part-tool
Likewise, the interval of time elapsed from the instant at
assembly with the lower heating and cooling rates.
which the autoclave environment reaches the end cycle
In fact, the curing process is properly performed only if each
temperature and the instant at which the latest component
composite part-tool assembly remains at the curing temperature
reaches the end cycle temperature can be defined as the cooling
for the defined time required by the part prepreg resin.
delay of the entire charge, RC (Fig. 2).
Therefore, the cooling phase, carried out by the autoclave
As mentioned before, the overall duration of the autoclave
forced cooling system, may only start when the last part in order
cure cycle depends on the delay with which the slowest part
of time has spent the required dwell time at curing temperature.
reaches the cure temperature and the end of cycle temperature.
This generates long curing times which negatively affect the
A number of studies on autoclave curing showed that heating
system productivity and determine a very low efficiency in
and cooling delay for each part depend on the material of the
terms of energy consumption and resource utilization.
parts to be cured as well as of the tools on which they are
In order to observe the thermal behavior of diverse
mounted in the autoclave charging floor [7,8].
composite material parts during simultaneous autoclave curing,
However, in this paper additional factors are taken into
monitoring of the thermal profiles of the diverse part-tool
account. As a matter of fact, beyond the thermal properties
assemblies during an industrial autoclave curing process is
related to the material of the part-tool assembly, also their
performed by measuring the in-process temperature with
geometry should be taken into account, as it has a large
thermocouples inserted within the material to be cured.
influence on the heating and cooling rate, and therefore on the
Fig. 2 shows the thermal profiles of part-tool assemblies and
dwell time within the target temperature range.
autoclave during an industrial curing cycle: each assembly
Moreover, the position of each part-tool assembly on the
reaches the minimum value of the curing temperature range at
autoclave charging floor (the scheme of the industrial autoclave
a time different from those of the other assemblies and of the
employed in this work is shown in Fig. 3), e.g. close to/far from
autoclave ambient.
the autoclave front door, significantly affects the heating and
cooling rates.
 Luigi Nele et al. / Procedia CIRP 57 (2016) 241 246
(a)
(b)
(c)
Fig. 3. (a) Industrial autoclave scheme (OP-PANINI model 0030-164); (b)
autoclave front door view; (b) part-tool assemblies positioned on the
autoclave charge floor.
Nevertheless, planning of autoclave manufacturing
processes in current aeronautical industry practice is still
carried out by grouping the composite parts to be cured together
according to their polymerization specifications (i.e. parts
having the same curing temperature), with very limited
consideration for the shape, size and material of which the parts
are made, and no consideration for the tools on which they are
mounted or the position on the charging floor.
In order to achieve higher resource efficiency and shorter
cycle times, a new criterion for the effective aggregation of
different parts to be cured simultaneously and for the selection
243
of the optimal strategy for positioning the part-tool assemblies
on the autoclave charge floor is proposed in this work.
For this purpose, a new parameter, S, which allows to
characterize both the geometry and thermal inertia of the part-
tool assemblies, is defined. Based on the value of the S
parameter for each part-tool assembly, it is possible to group all
the assemblies according to their and to select the optimal
positioning of the parts on the autoclave charge floor allowing
a reduction of the heating and cooling delays, Rh and Rc, and
consequently a lower autoclave cycle time.
3. Improved S parameter definition
With the aim of supporting decision making on autoclave
curing strategies, such as grouping of homogeneous part-tool
assemblies and optimal positioning of assemblies on the
autoclave charging floor, the newly defined S parameter should
be as comprehensive as possible so as to effectively represent
the part-tool assembly properties of relevance to the autoclave
curing process.
It has been shown that heat exchange between autoclave air
and prepreg is hardly influenced by the material of the part and
of the tool on which the prepreg is layed-up [7, 8].
While the prepreg layup thickness is very small (few
millimetres), tool thickness can vary according to the part
shape, usually ranging from about 3 mm to over 50 mm.
Consequently, the thermal inertia of the part-tool assembly is
significantly influenced by the thermal inertia of the tool.
Process analysis has shown that a high heating rate is typical of
parts which are laid on small thickness aluminum tools, while a
low heating rate can be determined by large thickness steel tools
or tools made of materials of low thermal conductivity.
Since tools do not always consist of simple flat surfaces, in
most cases their thickness cannot be defined in a
straightforward manner. Therefore, a kind of average thickness
has been considered, i.e. the ratio V/S, where V is the mould
volume and S is the prepreg covered area. As reported in [7]
this ratio cannot be assumed as grouping thermal parameter,
because it does not consider how heat exchange actually
proceeds. As a matter of fact, volume V should be referred to
the entire heat exchange area and not only to the prepreg
covered area.
The heat is transmitted from the environment to the part-tool
assembly through two surfaces: the part surface, Ap, and the
tool surface At. Ap is covered by the bag and other ancillaries;
At can be free or covered by the bag.
In order to consider this aspect, the thermal parameter
assumes the expression reported in [8]:
= (1)
+
ZKHUH!PHDQVWKDW$t is more influential than Ap, e.g. when
At is free from ancillaries and the charge floor of the autoclave
is perforated (aerated).
The above mentioned thermal parameter T only takes into
account the thermal inertia of the part-tool assemblies,
neglecting the role of the geometry of the part-tool assembly.
244 Luigi Nele et al. / Procedia CIRP 57 (2016) 241 246

In this paper, an improved parameter, S, has been developed
to include in the formulation not only the thermal inertia but
also the geometric characteristics of the part-tool assemblies.
During the autoclave curing process, the thermal controlled air
flux can be deviated by the parts to be cured: the deviation
depends on the shape of the part-tool assemblies and on the
cross section of the assemblies.
The analysis of the role played by the shape of the part-tool
assembly highlights that assemblies with larger height shield
the ones with smaller height preventing proper heat transfer to
the parts in the "shadow" of the air flux. Consequently, the
shielded parts are characterized by low heating rates, as they are
not reached by the full hot air flux. This effect is accentuated if
their height and cross section are low and their thermal inertia
is high.
The improved S parameter is therefore obtained by adapting
the formulation of the T parameter to include a relative
geometric feature of the part-tool assemblies:
= ( ) (2)
where hi = height of the ith part-tool assembly; hmax and hmin =
maximum and minimum height among all part-tool assemblies.
The S parameter is utilised in the present work to determine
the optimal positioning of multiple part-tool assemblies on the
autoclave charging floor based on both their thermal and
relative geometric characteristics.
5000 mm x 2610 mm and twelve vacuum lines was employed
(Fig. 3). Twelve thermocouples were utilized to acquire the
temperature profiles during the curing process.
At first, an experimental curing process was carried out by
positioning the parts according to the common industrial
practice (Fig. 4). In order to measure the part-tool assembly
temperature profiles, two thermocouples were located at two
diagonally opposed extremities of each part-tool assembly (Fig.
4). Moreover, three thermocouples were utilized to measure the
temperature profiles of the autoclave ambient.
In Fig. 5, the temperature profiles recorded during the non-
optimized autoclave curing process are reported. The autoclave
temperature remained within the required temperature range of
179.4 5.6 C for 217 min. The dwell time of each part-tool
assembly within the required temperature range depended on its
thermal inertia and position on the autoclave charging floor. In
particular, part-tool assembly # 2 displayed the longest heating
delay time, equal to 74 min with respect to the time instant when
the autoclave ambient reached the minimum value of the
allowed curing temperature range (173.8 C).
With the aim to test the new part-tool assembly positioning
strategy based on the S parameter, the S value for each part-tool
assembly was calculated as in eq. 2 and reported in Table 1.

4. Optimal part-tool assembly positioning strategy

Fig. 4. Part-tool assembly positioning in the common industrial practice.
Once the part-tool assemblies to be cured simultaneously
have been determined, to reduce the time of the curing cycle to
the minimum value, it is vital to determine the optimal part-tool
assembly positioning on the autoclave charge floor. The
optimal positioning is obtained based on the criterion that the
value of the S parameter of each part-tool assembly must
decrease going from the front to the rear of the autoclave
interior. The aim is to reduce the shield effect, in particular
for part-tool assemblies with high thermal inertia. This strategy
was applied to a real industrial case of autoclave curing of
multiple composite material parts to verify the advantages
offered by the proposed optimal part-tool assembly positioning
strategy based on the S parameter value.

5. Experimental results
Fig. 5. Part-tool assemblies temperature profiles recorded during the standard
industrial curing process; the black curve refers to the autoclave temperature.
To perform the validation of the developed model,
experimental curing processes were performed by monitoring Table 1. Values of the S parameter for the six part-tool assemblies and
the thermal profiles of the part-tool assemblies and of the consequent optimized position on the autoclave charging floor (1 = close to
autoclave through the use of thermocouples. The objective was the front door, 5 = close to the back of the autoclave).
to compare the temperature profiles recorded in the common Part-tool assembly S parameter value Optimal position
industrial process to the temperature profiles recorded in the
#1 5.70 5
new experimental process with optimal positioning of part-tool
#2 6.64 3
assemblies on the charge floor.
Six Carbon Fiber Reinforced Plastic (CFRP) parts with #3 6.09 4
different geometry, mounted on dedicated tools, were #4 8.02 1
considered for simultaneous autoclave curing. An industrial #5 7.01 2
OP-PANINI autoclave model 0030-164 with a charge floor of #6 7.61 2
 Luigi Nele et al. / Procedia CIRP 57 (2016) 241 246
Figure 6. Part-tool assembly positioning in the optimized curing process.
Fig. 7. Part-tool assemblies temperature profiles recorded during the
optimized curing cycle process; the black curve is the autoclave temperature.
Accordingly, the part-tool assemblies were positioned with
decreasing S values going from the front to the rear of the
autoclave interior (Table 1 and Fig. 6).
An experimental curing cycle process was carried out using
this optimized part-tool assembly positioning and the
corresponding temperature profiles are shown in Fig. 7.
The autoclave temperature remained within the required
range 179.4 5.6 C for 206 min. This dwell time is shorter
than for the non optimised curing process by (217 min 206
min) = 11 min.
This time saving was achieved through the S parameter
based position optimization that determined a 20% reduction of
the slowest heating delay time: from 74 min for part-tool
assembly # 2 in the case of non optimised curing process (Fig.
5), to 59 min for part-tool assembly # 4 in the case of optimised
curing cycle (Fig. 7), i.e. (74 min 59 min) = 15 min.
It is worth mentioning that the time saving should not be
calculated in absolute terms but with reference to the autoclave
temperature profile. As a matter of fact, the same autoclave
curing cycle process carried out in different moments may be
characterized by different heating rates of the autoclave interior
due to external factors such as the plant ambient temperature,
the simultaneous operation of several machines/autoclaves in
the same plant, etc. Thus, the autoclave curing cycle duration
should be evaluated from the time instant when the autoclave
ambient reaches the target curing temperature range rather than
from the start of the autoclave heating.
From Figs. 5 and 7, it can be noted that during the heating
phase the temperature profiles of the part-tool assemblies are
significantly spread out in the case of non optimised curing
process, displaying a heating delay at 173.8 C between the
fastest and the slowest curve equal to 60 min (Fig. 5). In the
case of the optimised curing process, this same heating delay is
notably reduced to 39 min (Fig. 7). The more homogeneous
heating realised with the optimised curing cycle is beneficial in
245
terms of more proper resin curing for all the part-tool
assemblies in the charge.
As regards the forced cooling phase, in the case of non
optimised curing cycle (Fig. 5) part-tool assembly # 2 displayed
the longest cooling delay time, equal to 68 min evaluated with
respect to the time instant when the autoclave ambient reached
75 C (this temperature level has been selected for comparison
purposes). In the case of optimised curing cycle (Fig. 7), the
longest cooling delay time is verified for part-tool assembly # 3
with a value of 36 min, yielding a time saving of (68 min 36
min) = 32 min i.e. 47% cooling time reduction in comparison
with the non optimised case.
Also during the cooling phase, the part-tool assemblies
temperature profiles are highly outspread for the non optimised
case, with a cooling delay at 75 C between the fastest and the
slowest curve equal to 60 min (Fig. 5). For the optimised case,
this cooling delay decreases to as low as 22 min (Fig. 7). As for
the heating phase, the more uniform cooling in the optimised
curing cycle is again beneficial in terms of correct resin curing
for all the part-tool assemblies in the charge.
The time reduction of the heating and cooling phases has a
twofold advantage: on the one hand, it enhances the autoclave
productivity by making the resource available much sooner for
subsequent batch fabrication, and, on the other hand, it
improves the resource energy efficiency by largely reducing the
energy consumption through the shorter operation of the
heating system as well as of the forced cooling system.
6. Conclusions
A new methodology for autoclave charge planning aimed at
the reduction of autoclave curing cycle time for multiple
composite material parts was proposed.
The methodology is based on the employment of a newly
defined S parameter taking into account both the thermal inertia
and the relative geometry of the part-tool assemblies.
With the aim to minimize the autoclave dwell time within
the target temperature range, a new strategy based on the values
of the S parameter to determine the optimal position of the part-
tool assemblies on the autoclave charge floor was defined.
The proposed methodology for the optimal positioning of
the part-tool assemblies was implemented with reference to a
real industrial case of simultaneous autoclave curing of multiple
carbon fiber reinforced plastic components.
The proposed methodology for autoclave charge planning
based on the newly defined S parameter allowed to reduce the
heating delay time, i.e. the delay in reaching the curing
temperature range, up to 20% with a time saving of 15 minutes.
Moreover, the cooling delay time in the optimized case was
decreased by 47% in comparison with the non optimised case,
with a time saving of 32 min.
The time reduction of the heating and cooling phases
allowed to enhance the autoclave productivity by reducing the
overall autoclave cycle time and to improve the resource energy
efficiency by significantly reducing the energy consumption
related to the operation of the heating system and of the forced
cooling system.
246 Luigi Nele et al. / Procedia CIRP 57 (2016) 241 246
Acknowledgements
CAM Compositi Avanzati Meridionali Srl, Paolisi (BN),
Italy, is gratefully acknowledged for its technical support.
The Fraunhofer Joint Laboratory of Excellence on
Advanced Production Technology (Fh-J_LEAPT Naples) at
the Dept. of Chemical, Materials and Industrial Production
Engineering, University of Naples Federico II, is gratefully
acknowledged for its support to this research activity.
References
[1] Campbell FC. Manufacturing technology for aerospace structural
materials. London: Elsevier; 2006
[2] Ciriscioli P, Wang Q, Springer G. Autoclave curing-comparison of model
and test results. Journal of Composite Materials 1992;26(1):90-102.
[3] Blest DC, Duffy BR, McKee S, Zulkifle AK. Curing simulation of
thermoset composites. Composites 1999;Part A(30):1289-1309.
[4] Kim JS, Lee DG. Development of an Autoclave Cure Cycle with Cooling
and Reheating Steps for Thick Thermoset Composite Laminates. Journal
of Composite Materials 1997; 31(22):2264-2282.
[5] Martinez GM. Fast Cures for Thick Laminated Organic Matrix
Composites. Chemical Engineering Science 1991; 46(2):439-450.
[6] Li M, Tucker C L. Optimal Curing for Thermoset Matrix Composites:
Thermochemical and Consolidation Considerations. Polymer Composites
2002; 23(5):739-757.
[7] Park HC, Lee SW. Cure simulation of Thick Composite structures Using
the Finite Element Method. Journal of Composites Materials 2001;35(03):
188-201.
[8] Hubert P, Johnston A, Vaziri R, Poursartip A. A two dimensional finite
element processing model for FRP composite components. In:
Proceedings of ICCM-10. Whistler; 1995. p. 149-156.
[9] Criscuolo Gaito V, Langella A, Nele L. Composite autoclave curing.
Advancing with Composites 2005, Naples 11-14 October 2005, CD rom
(31b.pdf) pag. 8, Amme-Asmeccanica.