THE EVALUATION AND DEVELOPMENT OF THE QUICKSTEP OUT-OF-AUTOCLAVE COMPOSITES PROCESSING METHOD

DAVID BOND AND ALAN NESBITT University of Manchester School of Mechanical, Aerospace and Civil Engineering PO Box 88 Manchester M60 1QD UK d.bond@manchester.ac.uk VICTORIA COENEN Hurel-Hispano Ltd Bancroft Rd, Burnley Lancashire BB10 2TQ UK DALE BROSIUS Quickstep Technologies Pty. Ltd. 152 Vulcan Rd, Canning Vale Western Australia 6155 AUSTRALIA

ABSTRACT
Quickstep is a polymer composite manufacturing technique for out-of-autoclave processing of high-quality, low-cost, components. The Quickstep process uses a liquid to transfer heat to the uncured laminate stack, enabling precise control of the stack temperature and a considerable reduction of cure-cycle times. Plant and tool structural requirements are significantly reduced compared to those of an autoclave process by eliminating the need for high consolidation pressures. This paper reports on the installation of a Quickstep product development plant at the University of Manchester for industrial use in the evaluation of the process. An assessment of the suitability of the Quickstep method for the processing of a typical aerospace composite material, 914/40/G703 carbon epoxy prepreg, is also presented. Physical and mechanical properties of laminate test specimens manufactured using various Quickstep process cycles and a conventional autoclave cycle are compared. The Quickstep processed material has similar or slightly improved properties over those of the autoclave processed material. The Quickstep process also achieves a significant reduction in the overall cure-cycle time and estimated manufacturing costs.

INTRODUCTION
Advanced fibre-reinforced polymer composite materials have not been as widely used within manufacturing industry applications as would be predicted from their structural performance characteristics. The main reasons for this aversion are the complexity of their processing techniques and the associated high-costs. The use of advanced composites has therefore,

largely been restricted to industries with already high product costs and low volume production, viz the aerospace and high-performance automotive industries. Currently, most high-performance composite components are produced by hand or machine lay-up of prepreg laminates followed by curing in an autoclave at elevated temperature and pressure in order to obtain the required structural properties. This manufacturing process is expensive and time consuming, with a typical process cycle time of up to 16 hours and an estimated cost per kilogram part of $640 [1]. There has been an increase in the development of alternative composite manufacturing technologies, with the aim of producing materials of similar quality to aerospace grade composites, but in a shorter processing time and at a lower cost e.g. RTM, RFI, VARTM [2, 3]. Although many of these techniques benefit from a shorter process cycle time than autoclave curing, they tend to produce material with inferior mechanical properties whilst still retaining a high manufacturing cost [1]. Quickstep Technologies Pty Ltd has developed a balanced pressure, heated mould process, that can be utilised for the out-of-autoclave curing of advanced composite materials [4, 5]. The process benefits from versatile production facilities, fast cure cycles and reduced capital, tooling and operational costs [6, 7]. The technology enables the fabrication of large composite components, together with the flexibility to co-cure or meld parts to produce complex components with superior structural integration. As part of an initiative to encourage development and pan-industrial exploitation of composites technology in the UK, the University of Manchester has, with Northern Aerospace Technology Exploitation Centre (NATEC) and local industrial funding, installed and commissioned a Quickstep process plant for the purposes of industrial evaluation, prototyping and product development.

THE QUICKSTEP PROCESS An overview

The Quickstep process utilises a fluid-heated, balanced-pressure, floating mould for the curing, partial curing and joining of composite materials. The process works by rapidly applying heat to an uncured laminate stack that is moulded to a rigid (or semi-rigid) tool floating in a Heat Transfer Fluid (HTF). The mould and laminate stack are separated from the circulating HTF by a flexible membrane. The temperature and pressure of the HTF behind the mould and flexible membrane stay the same.The process uses vacuum, combined with vibration, to evacuate air and volatiles from the laminate as well as to compact, heat and cure the part. The laminate may be thermoset or thermoplastic prepreg or a wet resin/dry fibre combination. Typically, parts manufactured using this process have equivalent, and in many instances superior, strength, stiffness, surface finish and appearance when compared to autoclave cured components. The laminate stack is assembled on a single-sided tool using conventional lay-up, sealed in a vacuum bag and then installed in a low-pressure chamber containing a glycol based HTF. The tool and laminate are supported between two flexible membranes in the pressure chamber (see Figure 1).

Heat Transfer Fluid Pressure Chamber

Vibration Source Diaphragm Flexible Bladder

Flexible Bladder sealing Mould tool floating and mould to pressure chamber supported in HTF

Composite part to be moulded

Figure 1. Schematic of Quickstep process Temperature control is maintained by circulating the HTF through the pressure chamber. As fluids have a heat energy per unit volume capacity much greater than that of a gas, the heat transfer rate between the HTF and laminate is much higher than that achievable in an an autoclave. This allows rapid heating/cooling rates to be achieved and provides precise control of the resin viscosity especially during the early consolidation phase of the cure. The high heat transfer rate allows a lower vicosity to be obtained in the laminate than that achievable with a slower autoclave heat-up rate. This is made possible by the laminate reaching consolidation dwell temperatures with reduced chemical cross-linking having occurred within the matrix (see Figure 2). By achieving lower viscosities in the laminate, excellent consolidation is obtainable at low applied pressures (typically vacuum, plus 10 kPa externally from the fluid). The HTF also acts as a large thermal sink, removing any excess heat generated in an exothermic reaction, thus a constant cure temperature may more easily be maintained, even for thick laminates. On top of the mean static pressure, an alternating pressure, generated by a vibrating diaphragm, may also be applied to the stack via the HTF in order to facilitate laminate compaction and minimise void content.
100000

Viscosity (poise)

10000 1000 100 10 1 0 10 20

2.8C/min 11C/min 22C/min

30

40

Time (min)
Figure 2. Variation of viscosity in Toray G83C Prepreg for different heating rates (Quickstep: 22 and 11C/min, Autoclave: 2.8C/min)

The HTF is stored in three separate tanks, a hot tank, medium tank and cold tank (See Figure 3). Heat energy can be efficiently accumulated over many hours and stored in the tanks and used repeatedly for many cure cycles unlike the consumables (eg. pressurised nitrogen) in autoclave curing. The pumping and recycling of the HTF to the curing chambers can take place over minutes thus allowing for rapid heat-up and cool-down of the part. With the high level of available heat-energy the cross-linking takes place rapidly and in a fraction of the time achievable within an autoclave. As a result of such rapid curing, the final laminate can be markedly different from those produced in an autoclave and higher glass transition temperatures and material properties have been demonstrated for some materials [8].

(a)

(b)

Figure 3. (a) Quickstep QS5 evaluation and prototyping plant and (b) small curing chamber

Industrial advantages of Quickstep process

Cycle times for the Quickstep process are typically between 40 and 60 minutes for current generation pre-preg materials. This is a significant time-saving over the 6-8 hours typically required in autoclave curing processes. While less parts are manufactured per Quickstep cycle, significant savings are available to industry from reduced scrap rates for interrupted cure cycles, reduced tooling requirements and increased flexibility in part manufacture order. An important and novel aspect of the Quickstep process is the ability to stop the cure reaction over any area of a curing laminate, at any time in the cure cycle. This is achieved by returning the hot HTF to the tank and using the cooling HTF to quench the reaction. In this way, the whole part or a section of a part can be held uncured or partly cured. As a consequence, it then becomes possible to join such partially cured material to other uncured or partially-cured material in a secondary cure-cycle. This process, referred to as melding, allows multiple laminates to be joined, without secondary adhesive bondlines, into larger parts and offers an opportunity for novel new product designs to be produced.

RESULTS OF STUDY TO DATE

Prior to the decision to invest in a UK based plant, The University of Manchester, Quickstep Pty. Ltd. and a UK aerospace consortium conducted a process feasibility study, utilising the Quickstep development plant at Deakin University in Geelong, Australia. The objectives of this study were to: (a) Develop a Quickstep cure cycle for the carbon fibre reinforced epoxy prepreg 914/40/G703 that would minimise cycle time without adversely affecting laminate chemical and physical properties; (b) Compare the chemical and physical properties of the Quickstep cured material with laminates produced using a conventional (manufacturer recommended) autoclave cure process; (c) Estimate the manufacturing cost-savings achievable through use of the Quickstep process for simple aircraft components.

MATERIALS & PROCESSING

The material used in the present study was carbon fibre reinforced epoxy prepreg 914/40/G703, which is used extensively in high temperature resistant, primary aerospace structures. The prepreg is a 0 / 90 (5 harness satin) cloth and was provided by HurelHispano (Burnley, UK) from stock acquired directly from Hexcel UK. Quickstep specimens were manufactured by Quickstep Technologies at Deakin University, while conventional autoclave specimens were manufactured at Hurel-Hispano. In order to minimise variation in the laminates compared in the study, all specimens were made using the same batch of prepreg and were laid-up and debulked (every four plies) at Hurel-Hispano. Specimens to be cured using the Quickstep process were vacuum-bagged under a light pressure and airfreight couriered to Deakin University, with a typical transit period of one week. All consumables were consistent in the two processes. Quickstep specimens, once cured, were shipped back to Hurel Hispano where both Quickstep and Autoclave cured specimens were post-cured for 4 hours at 190C.

CURE CYCLE OPTIMISATION

The effect of varying Quickstep cure cycle parameters for 914/40/G703 carbon epoxy prepreg was investigated in order to minimize the process cycle time, whilst retaining laminate properties (e.g. mechanical, void/volume fraction, consolidation, Tg) equivalent to autoclave cured laminates. The process cycle is shown in Figure 4 and the cycle parameters (ramp times, dwell times, dwell temperatures) together with the total cycle time (including cooling time) are listed in Table 1. The conventional manufacturer recommended autoclave cure cycle (A1) and three Quickstep cure cycles (Q1-3) have been compared. Differential scanning calorimetry (DSC), dynamic mechanical thermal analysis (DMTA) and optical microscopy were performed at The University of Manchester. The average thickness of the 8 ply specimens, the degree of cure calculated from DSC and the glass transition temperature (Tg) determined from DMTA on the post-cured materials for the cure cycles A1 and Q1-3 are reported in Table 1.

1st Dwell 2nd Dwell Temp Temp

Temperature

2nd Dwell Time 2nd Ramp Time 1st Dwell Time 1st Ramp Time Cooling Time

Room Temp

Processing Time (min)

Figure 4. Processing cycle
1st Dwell 2nd Dwell DSC 1st 2nd Total Applied Average Ramp Ramp Cycle Pressure Thickness Time Time Time Time Time % T ( C) T ( C) (kPa) (mm) (min) (min) (min) (min) (min) cure 55 10 10 8 135 175 135 110 30 0.02 35 30 20 5 5 7 175 125 175 180 60 120 60 60 260 145 120 115 700 10 10 10 2.24 0.05 99.7 2.50 0.05 59.6 2.32 0.05 84.0 2.43 0.05 80.3 DMTA Tg ( C) 200 208 207 207

Cure Cycle A1 Q1 Q2 Q3

Table 1: Comparison of autoclave and Quickstep cure cycles The cure cycle optimization process did not look to minimize specimen thickness and only required that the final laminte thickness was within an acceptable manufacturing tolerance (0.2mm) of the nominal autoclave thickness (2.25mm). Associated work with other bleeder and breather material combinations has produced laminates of comparable thickness to that of the autoclave process. DSC measurements of the Quickstep cured specimens (Figure 3) show two features; a reaction exotherm for the 914 epoxy between 200C and 300C; and an endothermic peak between 70C and 100C. The latter is most likely due to a phase change in either the excess curing agent (DICY) or the thermoplastic toughener included in the pre-preg. The autoclave sample (A1) is almost fully cured after its cure cycle alone, whilst the Quickstep specimens (Q1-3) are significantly less so. The Q1 cycle produced a material that was significantly unreacted and which had only just reached its gel point and was barely vitrified. Some residual cure was also observed for the Q2 and Q3 cycles but at a much reduced level to that of the Q1 specimens. After post-curing, all specimens were fully cured and DSC measurements showed only the low temperature phase change peak (Figure 4). Note that the different appearance of the Q3 data in Figures 3 and 4 is a result of regular maintenance on the DSC having been carried out between testing of the A1/Q1-2 specimens and the Q3

specimens.
0.4 0.4 Q3

Heat Flow (W/g)

0.0 -0.2 -0.4 -0.6 -0.8 -100 Q1 Q2 A1 0 100 200 300 400

Heat Flow (W/g)

0.2

0.2 0.0 -0.2 -0.4 -0.6 -0.8 -100

Q3

Q2 Q1 A1

100

200

300

400

Temperature (C)

Temperature (C)

Figure 3. DSC of cured specimens

Figure 4. DSC of post-cured specimens

Optical microscopy revealed significant void content in the Q1-2 specimens compared to that found in the autoclave processed laminate. The Q3 cycle produced a void content comparable to that of the A1 specimens (see Figure 5).
A1 Q1

Q2

Q3

Figure 5. Optical micrographs of composite samples manufactured using A1, Q1, Q2 and Q3 cure cycles.

MECHANICAL TESTING
Mechanical testing was performed at the University of Manchester. Specimens were manufactured using the conventional autoclave cure-cycle A1 and the Quickstep cure cycle Q3 (4 post-cured specimens for each test and process combination). The mechanical

property test-matrix was based upon the Hurel-Hispano batch acceptance test matrix. The relevant industry test specifications are summarised in Table 2 (note that where test standards are for unidirectional laminates they have been used for guidance on the procedures used in this study). Test panels comprising 8 ply laminates were prepared with the 3 fibre configurations listed (note that the angles refer to a single axis within each woven cloth prepreg lamina). Property Tensile Strength (MPa) Standard NF EN 2561 Configuration 8 plies at 0 Autoclave 600 50 240 10 226 4 75 4 Quickstep 660 30 270 10 229 4 83 2

8 plies quasi-isotropic Open Hole Tensile NF EN 2561 Strength (MPa) (0/45/90) In-Plane Shear Strength (MPa) NF EN 2561 8 plies at 45 8 plies at 0

Interlaminar Shear NF EN 2563 Strength (MPa)

Table 2: Mechanical property comparison for A1 and Q3 processing cyles

ESTIMATED PROCESS COST SAVINGS

An estimate for the cost-savings achievable through use of the Quickstep process for manufacture of an aircraft gas turbine engine thrust reverser blocker door was made using the following assumptions: (i) material, cutting, debulking and laminating costs would be the same regardless of curing process and (ii) two components could be made using the Quickstep process in the same time as one batch of autoclave components - reducing tool requirements by half. These assumptions combined with an estimated cost saving of 65% for a Quickstep tool over that of an autoclave tool meant that an 82% cost-saving could be made on tooling alone. Note that these figures are an extremely conservative estimate of the cost-benfit offered by Quickstep as they ignore the significantly higher capital and maintenance costs of an autoclave compared to those required for the Quickstep process. For a more accurate estimate of the total savings possible through use of the Quickstep plant, the lower investment and recurring costs will need to be considered as well as the lower energy consumption for heating/cooling (estimated to be in the order of 75-90% when compared to autoclave processing), and reduced consumable costs.

CONCLUDING REMARKS
Quickstep is an advanced composite component manufacturing process for out of autoclave processing of high-quality, low-cost, low-weight components using faster cure-cycles. This process offers distinct advantages to companies looking to introduce composite technology to their product range. A process evaluation and product development plant has been

installed at the University of Manchester for use by UK industry. This plant has been installed on the basis of extremely promising results from an initial feasibility study conducted by the University and a local industrial consortium. In that study laminate samples of 914/40/G703 epoxy carbon prepreg have been successfully manufactured using the Quickstep technique, a novel out-of-autoclave process. A considerable reduction in the overall process cycle time (50%), compared with conventional autoclave curing, was achieved due to the precise temperature control and high heating/cooling rates (8C/min) that are possible with Quickstep. The Quickstep cured specimens exhibited a higher glass transition temperature and were on average 8% thicker than the autoclave samples. The thickness/volume fraction effect may be explained by the lower pressure applied during the Quickstep cure cycle resulting in reduced resin bleed out. Comparable void content and mechanical test data were obtained for the Quickstep and autoclave specimens. The tooling cost for a simple aircraft component was estimated to be 82% cheaper for the Quickstep process compared to that for an autoclave process.

REFERENCES
[1] Bader MG (2002), Selection of composite materials and manufacturing routes for costeffective performance, Composites Part A, 33, 913-934. [2] USDoD (2002), The Composite Materials Handbook, MIL-HDBK-17, US DoD & FAA. [3] FAA (1997), Handbook: Manufacturing Advanced Composite Components for Airframes, DOT/FAA/AR-96/75, FAA. [4] Griffiths B and Noble N (2004), Process and tooling for low cost, rapid curing of composite structures, SAMPE Journal, 40(1), 41-46. [5] Hodgkin JH and Rabu N (2000), A new development in high-speed composite fabrication, Proceedings of 45th Int. SAMPE Symposium, Longbeach CA, 2274-2282. [6] Anon (2004), Quickstep process slashes time from auto hood, Performance Materials, 19(9), 4. [7] Bader MG and Noble N (2004), A comparison of the costs of manufacture of a laminated composite component by conventional autoclave and a non-autoclave process, Proceedings of SAMPE-EUROPE, Paris. [8] Coenen V, Hatrick M, Law H, Brosius D, Nesbitt A and Bond D (2005), A Feasibility Study of Quickstep Processing of an Aerospace Composite Material , Proceedings of SAMPEEUROPE, Paris.