Composites Manufacturing
Throughout the last four decades the exploitation of fibre-reinforced plastics (FRP) in engineering structures has been steadily diversifying from sports equipment and high performance racing cars, to helicopters and most recently commercial aeroplanes. Composite materials are essentially a combination of two or more dissimilar materials that are used together in order to combine best properties, or impart a new set of characteristics that neither of the constituent materials could achieve on their own. Engineering composites are typically built-up from individual plies that take the form of continuous, straight fibres (eg. carbon, glass, aramid etc.) embedded in a host polymer matrix (eg. phenolic, polyester, epoxy etc.), which are laminated layer-by-layer in order to built up the final material/structure.
In terms of manufacturing advanced fibre-reinforced composites the single most important aspect to recognize is that the material and the structure are created at the same time. Consequently any defects that are induced during the manufacturing process directly influence the strength and stiffness of the material and structure. Every little detail is important.
A large number of composite manufacturing processes have been developed over the last 40 years including: contact moulding, compression moulding, vacuum bag/autoclave moulding, rotational moulding, resin transfer moulding (RTM), tape wrapping, filament winding, pultrusion, expanding bladder moulding etc. All these processes have several characteristics in common; the reinforcements are brought into the required shape in a tool or mould, resin and fibres are brought together possibly under elevated temperature and pressure to cure the resin, and the moulding stripped from the part once the resin has cured. The different fabrication techniques can either be classified as direct processes (eg. RTM, pultrusion, contact moulding) that use separate fibres and resin brought together at the point of moulding or indirect processes that use fibres pre-impregnated with resin (eg. vacumm bag/autoclave moulding, compression moulding).
The selection of the manufacturing process will naturally have a great effect on the quality, the mechanical properties and fabrication cost of the component. According to Potter (1996) an ideal process can be defined as having:
- High Productivity – short cycle times, low labour contents etc.
- Minimum materials cost – low value added materials, low material storage and handling cost
- Maximum geometrical flexibility – shape complexity and size of component
- Maximum property flexibility – range of matrices, range of reinforcement types, ability to control mechanical properties and tailor characteristics
- Minimum finishing requirements – net shape manufacturing
- Reliable and high quality manufacture – low reject rates, low variability etc.
No manufacturing process exists that can simultaneously fulfill all these requirements; most importantly some of these requirements may be mutually exclusive. A comparison of the 5 most common processes is shown below.
Contact Moulding
This is the oldest and most primitive manufacturing process but also the most widely used around the world. In contact moulding resin is manually applied to a dry reinforcement placed onto a tool surface and can be compared to glueing wall paper with a brush. The tool and fabric are then enclosed by a vacuum bag and the air under the bag removed in order to cure the laminate under atmospheric pressure. However, since the applied pressure is relatively low and cure typically occurs at room temperature the volume fraction of reinforcement is limited to the natural packing density. Furthermore, the quality is totally dependent on the skill of the workforce and due to the difficulty in reliably guaranteeing high-quality laminates it is almost impossible to qualify contact moulded structural components for commercial aircraft. Finally, due to the limited external pressure voidage is difficult to control, which has a great effect on the variability in the thickness of laminates.
On the other hand the process is highly flexible, ideal for one-off-production and requires minimal infrastructure. While contact moulding is process of choice for very large structures the geometrical flexibility is more constrained in terms of creating parts with fine details, corner radii, etc. For this reason the process is extensively used in glassfibre/polyester resin shipbuilding and for gliders.
Vac. Bag/Autoclave
In advanced composites autoclave processes are by far the most widely used and autoclave moulding is the process of choice for the aerospace industry. These processes use pre-impregnated uni-directional plies or woven cloths, which have been partially cured or beta-staged. One disadvantage is that pre-preg has to be kept in a freezer in order to prevent the resin from going-off. Multiple prepreg plies are laid down onto a tool surface with the pre-defined fibre orientations, to build up the required thickness, and then covered with a release film, breather fabric and a vacuum bag or silicon pressure bag. The air is drawn out from the bag to create a vacuum and the tool heated under elevated temperature and pressure to cure the resin. In principle multiple demoulding cycles are performed by covering the laminate and applying a vacuum after every 3-4 ply layers in order to remove any excess air between layers. This reduces the bulk factor and helps to prevent delaminations between plies and controls the thickness dimension. Regular demoulding cycles and sufficient hydrostatic pressure on the part during curing are the two basic requirements for achieving good mouldings. The productivity of autoclave moulding is generally quite low since the manual lay-up, bagging and demoulding cycles consume significant labour and time. Furtermore, the capital expenditure of autoclaves are enourmous, which constrains its use to larger structures where these expendictures are justified. Since, pre-preg is no longer in a low-value added state the material costs are also higher.
Geometrical flexibility in both shape and size are better than for most processes. Recently it has been possible to manufacture the entire floor of a helicopter in one piece, which would not be possible with a metallic approach. Autoclave mouldings are often used in conjunction with honeycomb cores such that very lightweight components can be manufactured. This is one of the reasons why the dominance of autoclave mouldings seems very likely to continue in the near future, at least in the aerospace environment.
Filament Winding
In filament winding a tow of fibres is passed through a bath of resin and wound onto a revolving mandrel by traversing longitudinally along the axis of the rotating mandrel. Unless tacky pre-impregnated fibre tows are used the path followed by the tow must closely follow a geodesic path (fibre paths that do not cause fibres to slip if tensioned). Any simple helical path on a cylinder is defined to be a geodesic path but once curvature in two directions is introduced (e.g. a globe) the number of possible paths becomes very limited. For this reason property flexibility is rather constrained such that filament winding is typically used for manufacturing pipework, pressure vessels and rockets motors. Especially, pressure vessels are conducive to filament winding since they have two clearly defined stress-directions (the hoop and longitudinal stresses) that can be accommodated by the winding direction.
One disadvantage of filament winding is that the mandrel is often enclosed within the winding. If a liner of metal or polymer is used as a mandrel it may form a permanent part of the structure but it is more common that the winding is slit-off at the ends to demould the part. The geometrical flexibility is also constrained by having to wind around circular or prismatic mouldings. One major advantage is that the process lends itself to automation such that cycle times and labour costs can be kept low with high reliability and quality. This latter aspect is one of the reasons why efforts are being made to widen the process’ geometrical limits and possible applications.
Resin Transfer Moulding (RTM)
RTM can not be considered as a single process but is better regarded as a “manufacturing philosophy in which the resin and fibres are held apart until the very last moment” (Potter, 1996). However, all process variations have the common features of holding unresinated fibres within a closed tool cavity with a differential pressure applied to a supply of resin such that the resin permeates into the reinforcement. The tool may be rigid or contain flexible elements. The consolidation pressure on the tool is applied by means of mechanical clamps, a tooling press or the use of internal vacuum and defines the achieved volume fraction of fibre with respect to resin. RTM has been used since the 1970s to build radomes as well as aeroengine compressor blades. The main driver behind further developing RTM processes is to devise fabrication methods that can overcome the geometrical complexity limitations imposed by autoclave mouldings. In terms of productivity cycles times are lower than most other processes and in the automotive industry small components are manufactured within minutes.
A major advantage of RTM is the use of low added value materials (dry fibres and low viscosity resins) which do not have to be stored in freezers, thus driving down material and handling costs. The major advantages of RTM however lie within their geometrical and property flexibility. RTM can be used with UD stitched cloths, woven fabrics and 3D fabrics, and the resin injection can be varied to control the volume fraction and therefore the stiffness and strength of the component. Furthermore, small components with very fine details are manufactured on rigid metal tooling while larger components can be produced on flexible moulds. Finally, with a closely controlled process it is possible to create net-shape mouldings with minimal finishing requirements. However, all this comes at the cost at a slightly trickier production technique. In order to guarantee high-quality components the resin injection and resin flow has to be closely controlled such that all of the reinforcement is equally wetted-out. This requires quite advanced fluid dynamics simulations and extensive testing in order to come up with a mould shape that allows even resin flow to all parts of the component.
Pultrusion
In this process fibres are drawn from a creel board and passed through a resin bath to impregnate the fibres with resin. The impregnated fibres are then passed through a pre-die to remove any excess resin and to pre-form the approximate final shape. The curing die is then entered, which takes the shape of the final required cross-section of the pultruded part. The curing die applies heat to the component to consolidate the resin and the cured, shaped profile is pulled from the die under tension. This means that productivity can be very high in an ongoing production but will fall for lower production volumes that require changes to new cross-section dies. Since the operation is automated labour costs are low and the reliability and quality of components is high. The process is generally limited to constant cross-section components, which greatly restricts applications. Pultrusion has been used very little in aerospace environments but has found application in manufacturing standardized profile beams for civil engineering structures.
Automated Processes
The use of robotics in composite manufacturing is growing at a rapid rate and is probably the most promising technology for the future. Obvious advantages of automating the manufacturing process include reduced variability in dimensions and less manufacturing defects. Furthermore, the feed material can be used more efficiently and labour costs are reduced. One promising class of system are the so-called Automated Fibre Placement (AFP) machines which use a robotic fibre placement head that deposits multiple pre-impregnated tows of “slit-tape” allowing cutting, clamping and restarting of every single tow. While the robotic head follows a specific fibre path tows are heated shortly before deposition and then compacted onto the substrate using a special roller. Due to the high fidelity of current robot technology AFP machines can provide high productivity and handle complex geometries. Current applications include the manufacture of the Boeing 787 fuselage and winding of square boxes, that are then slit lengthwise to make two ‘C’ sections for wing spars. Integrated manufacturing systems as designed by companies like ElectroImpact offer exciting turnkey capabilities for future aircraft structures. These systems combine multiple manufacturing processes, for example fibre placement and additive manufacturing on one robot head, and therefore facilitate the production of blended and integrated structures with fewer joints and connections. These systems will also allow engineers to design more efficient structures, such as integrated orthogrid or isogrid composite panels, that are currently hard to manufacture economically on a large scale.
References
(1) Potter, Kevin (1996). An Introduction to Composite Products: Design, Development and Manufacture. Springer, 5th Ed. Chapman & Hall, London.
(2) http://www.tca2000.co.uk/wilton3small.jpg
(3) http://csmres.co.uk/cs.public.upd/article-images/nose-72668.jpg
Related Posts
5 Responses to Composites Manufacturing
Leave a Reply Cancel reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
Sign-up to the monthly Aerospaced newsletter
Recent Posts
- Podcast Ep. #49 – 9T Labs is Producing High-Performance Composite Materials Through 3D Printing
- Podcast Ep. #48 – Engineering Complex Systems for Harsh Environments with First Mode
- Podcast Ep. #47 – Möbius Aero and MμZ Motion: a Winning Team for Electric Air Racing
- Podcast Ep. #46 – Tow-Steered Composite Materials with iCOMAT
- Podcast Ep. #45 – Industrialising Rocket Science with Rocket Factory Augsburg
Topics
- 3D Printing (4)
- Aerodynamics (29)
- Aerospace Engineering (11)
- Air-to-Air Refuelling (1)
- Aircraft (16)
- Autonomy (2)
- Bio-mimicry (9)
- Case Studies (15)
- Composite Materials (25)
- Composites (7)
- Computational Fluid Dynamics (2)
- Contra-Rotation (1)
- Design (2)
- Digitisation (2)
- Drones (1)
- Education (1)
- Electric Aviation (11)
- Engineering (23)
- General Aerospace (28)
- Gliders (1)
- Helicopters (3)
- History (26)
- Jet Engines (4)
- Machine Learning (4)
- Manufacturing (12)
- Military (2)
- Modelling (2)
- Nanomaterials (2)
- NASA (2)
- New Space (11)
- News (3)
- Nonlinear Structures (1)
- Novel Materials/Tailored Structures (14)
- Personal Aviation (5)
- Podcast (45)
- Propulsion (9)
- Renewable Energy (2)
- Renewables (1)
- Rocket Science (17)
- Satellites (8)
- Shape Adaptation (1)
- Smart Materials (1)
- Space (12)
- Space Junk (1)
- Sport Airplanes (2)
- Startup (19)
- STOL (1)
- Structural Efficiency (5)
- Structural Mechanics (1)
- Superalloys (1)
- Supersonic Flight (2)
- Technology (18)
- UAVs (2)
- Virtual Reality (2)
- VTOL (3)
- Privacy & Cookies: This site uses cookies. By continuing to use this website, you agree to their use.
To find out more, including how to control cookies, see here: Cookie Policy
Hello everyone, I have few questions about this article.
“Pultrusion has been used very little in aerospace environments”,will pultruded products be more and more used in aerospace sector ( I should say aeronautics in one hand and space in the other hand)? Because aircrafts are made of high performance composites and the trends could be having more and more (100%?) composite components. With pultrusion, production costs could be lower, could it be a driving force for pultrusion market in aerospace?
Hi, pultruded parts are used in the aerospace industry but generally only for beam like sections. Pultrusion has two drawbacks: you can really only manufacture beam-like structures and it is much harder to incorporate of axis fibre directions in the layup i.e. the fibres are all aligned along the beam axis. Most bigger aerospace structures require ±45° plies for impact protection and 90° plies for transverse loading. These are probably impossible to incorporate in a pultrusion process. One of the biggest advantages of composites are their tailorability i.e. the ability to optimise the layup and therefore the structure to the specific loads acting. Pultrusion does not allow you to take advantage of this. However, in smaller applications such as UAV’s pultrusion has been used for wings http://www.compositesworld.com/articles/pultruding-cost-out-of-aerospace-parts. For commercial airlines you will probably only see pultruded sections for smaller beam-like structures that do not compare to the cost of the major structures such as the fuselage and wings. I also doubt that an aircraft will ever be 100% composite. There are application where metals are simply the better choice (engine components, landing gear, lugs and hinges etc.) in terms of their structural capabilities and in terms of manufacturing costs. Researchers are also working hard to improve current aerospace metals in terms of specific strength and stiffness, material cost, corrosion resistance, fatigue life etc. So I doubt you will ever see a commercial airliner that is 100% carbon composite.
Thanks for reading!
The Boeing 787 fuselage sections are fiber or tape placed by Automatic Fiber Placement (AFP) machines. A plethora of prepreg slit tapes are placed directly against the mold surface under a compacting roller, thereby mitigating the geodesic winding penalty of filament winding. Each individual tape of the multi-tape bandwidth is controlled individually, enabling cutting and adding of tapes as required by geometry change. Filament winding creates too much build up of material on changing geometries where weight sensitivity is an issue. Most of the AFP machines in service on these fuselage barrel sections are supplied by Ingersoll and ElectroImpact.
Hi Martin, thanks for your comment. Yes you are absolutely right about AFP manufacturing in the 787 process. I have actually written about a specific application of AFP here https://aerospaceengineeringblog.com/variable-stiffness-composites/. In fact I have worked with researchers at NASA Langely on the ElectroImpact machine they have recently installed there http://www.nasa.gov/larc/robot-isaac-will-help-nasa-langley-speed-toward-innovation/
When I wrote this post I wasn’t aware of Boeing’s capability in detail. I’ll update the post. Thanks!
It appears we know some of the same folks. That EI system at NASA LaRC has great potential and a solid team involved!
Keep up the good work.
Best regards,
Jim