The space race - nanotechnology for metal matrix composites
Nanotechnology has opened the door on a new generation of metal matrix composites. Raphael Addinall from German research organisation Fraunhofer looks at its potential applications in aerospace industries and beyond.
The space race may be over, but the marathon for metal matrix composites (MMCs) for use in space has just begun. As early as the 1960s, the use of MMCs with continuous fibre reinforcement was a trend no one could ignore. In the ‘80s, the adoption of incorporating discontinuous fibres gained pace. But it was not until the early ‘80s that MMCs made their debut within aerospace applications, including a boron–aluminium composite for the truss tubes used in the mid-fuselage of the space shuttle, and the graphite–aluminium composite for the antennae boom of the Hubble telescope.
The era of MMCs was blossoming – not only for aerospace applications, but also in the automotive, construction and electronics industries. Its use in so many applications was fortified with companies such as Alcan, 3M company, Ameco-B and others investing millions into research, manufacture optimisation and process development. Large factories were built around the world, enabling mass production for a cost-reduced product.
Unfortunately, the heyday of discontinuous reinforced MMCs would come to an end in the mid-nineties. This was partly due to the strong competition of fibre/particulate reinforced polymer composites, but also the difficulty for particulate/discontinuous reinforced MMCs to compete with the well established, cost-effective and excellent properties of fibre-reinforced MMCs. Hence from both a material (polymer) point of view as well as a manufacturing perspective, the need, interest and investment required were not enough to keep particulate/discontinuous reinforced MMCs in the race. Interest will only start to reappear if scientists, engineers and factories can show an existing manufacturing infrastructure to make parts in commercial quantities. For this to become a reality, companies as well as individuals must start ordering MMCs in bulk to generate the interest and investment needed.
With the development of nanotechnology – including the discovery of fullerenes in 1985 and, more importantly, Professor Lijina’s invention of the method for producing reproducible carbon nanotubes (CNT) out of transition metals catalysts in 1993 – the door for a new generation of MMCs was opened once again. Considering the formidable attributes of CNTs, their incorporation into MMCs seemed like an ideal choice. The mechanical (tensile strength 13–53GPa, E-Modulus ~1TPa) and thermal (conductivity up to 6,000Wm/K, stable up to 2,800°C in vacuum) properties for single-wall CNT speak for themselves. As a result, investment in MMC research and development greatly increased in Japan, the USA and Europe.
Of course, theoretical and experimental values are two very different points. As such, the aims, method and conclusions brought about by synthesis, functionalisation and dispersion of CNTs within MMCs – especially aluminium – need to be discussed. This, along with further emphasis on the applications of these composites, may also go some way to generating inspiration when it comes to the future of MMCs and their application within the aerospace industry.
It has been well documented and acknowledged by experts in the field that for efficient stress-strain transfer from fibres (in this case, CNT), good interfacial bonding between the filler and the matrix material must be accomplished. To generate the best possible properties in a reproducible way, a homogeneous dispersion of filler within the matrix material is required. For CNTs this is extremely hard to accomplish, mainly due to the high aspect ratio (1:1,000,000) leading to re-agglomeration through Van-der-Waals forces of attraction. In addition, the problem of interfacial interaction is a difficult task to take on due to the naturally inert nature of the CNT surface. Only through functionalisation will these problems be eliminated. However, to either oxidise or attach functional groups to CNT surfaces, carbon-carbon bonds will have to be broken, a procedure detrimental to the properties of the tubes. It is therefore logical to use multiwall carbon nanotubes (MWCNT) where only the outermost parts of the tubes are damaged, the inner tubes staying intact. The use of MWCNT has a further advantage of being more economically feasible, at an average price of US$300 per kilogramme, compared to US$20,000 for single-wall CNTs.
The aim in creating MMCs for aeronautical applications lies not only in producing parts with improved properties, but also from a good manufacturing process. The science of scaling-up from lab conditions is vast and complicated – in most cases, cost and reproducibility are considered more important than having the best possible mechanical properties. It goes without saying that a 5-10% decrease in mechanical and thermal properties can achieve a 50% reduction in production cost, and depending on the scale of things this path will be chosen.
Forming a bond
With respect to improved mechanical properties through functionalisation and good dispersion of CNTs within metal matrixes, while also taking into consideration manufacturing process, a project involving aluminium alloy 6061 and CNTs was made with the aim of improving both thermal and mechanical properties. The alloy AA6061 was specifically chosen due to its good mechanical properties, heat treatability and existing use as an alloy within the aerospace industry.
The aluminium alloy 6061 used had an original particle size of 45μm, and the MWCNT (obtained from Nanocyl) measured 150μm in length and 15nm diameter. The aim was to understand the influence of varying additions of CNTs to weight addition and milling time. To efficiently mix the two materials together, a high-energy attrition milling machine was employed. This enabled the mechanical alloying of aluminium particles with MWCNT via plastic deformation, followed by fracturing (due to the high dislocation strains of the crystallographic planes), then cold welding of the oxide-free fractured surfaces.
The mechanism of deformation, fracture and cold-welding experienced by the particles during milling was repeated within the process until a fine particulate powder mixture of AA6061 and MWCNT was produced, with a typical particulate size of 15- 20μm. Using the classic MMC production routes, the powder was hot-press sintered (Dr Fritsch DSP515) to consolidate the powder via diffusion mechanisms into predefined shapes ready for extrusion.
The extrusion process to some extent resulted in alignment of CNTs, as well as densification of the MMC through severe plastic deformation. The extruded samples were tested for thermal expansion, impact, tensile strength and Young’s modulus. The results proved promising, with major increases in mechanical properties and moderate improvements with respect to thermal properties.
The impact tests were carried out using a Zwick Roell impact test machine, while the tensile tests were made on an Instron Corporation 4507 universal mechanical testing machine Series 4500.
Concerning the CNT behaviour within the matrix, it is evident the CNTs are well-dispersed throughout the material (below), with good interfacial bonding with the aluminium. As can be seen, the CNTs are protruding out of the failed surface at regions of high necking, which indicates good interfacial strength.
Looking at the results of the mechanical and thermal properties (see table right), it is possible to say that the addition has a positive effect. With respect to mechanical properties, the standard deviation of the results is too large and, therefore, further optimisation in the form of synthesis (direct growing of CNTs on particles) and functionalisation of the CNTs needs to be made.
The ability to control the process from the beginning will enable improvement of both the dispersion (via synthesis) and interfacial bonding (via functionalisation). One downfall with both these routes is the extra cost involved, and the increased difficulty in scaling up for the production of parts applicable for the aerospace industry could also prove problematic. Although models currently exist that enable us to identify many aspects of MMC properties, the essential need for reproducible production methods can only be made through full control of the filler material.
Rather than going into detail about the models used to describe the elastic modulus increase through CNT introduction, the above table outlines the assumptions made by the three main models used (Voight-Reuss, Cox and Halpin-Tsiai). Although these models have been used to determine polymer-based results, the same ensuing predictions have been made for MMCs.
Will we have lift-off?
The future of MMCs is difficult to predict, due to the immense scientific and technological developments being made. The pure material research development, new characterisation methods and ever-developing manufacturing capabilities have, and will, continue to ensure that the MMCs will play a strong role not only in the aerospace industry, but the wider sector of automotive, construction, packaging and even sports industries.
It appears that a new path has been created with the emergence of nanotechnology and its applications within metal matrixes. However, it is still in its early stages and in order to ensure interest from end users and manufacturers, scientists and engineers must show an ability to control filler interaction within the matrix material and be capable of transferring this to the manufacturing industry. Only then, through the production of cost-effective and reproducible parts, will the rise of MMC happen.
The author wishes to thank Carsten Glanz, Ivica Kolaric and Professor Bauernhansl (Fraunhofer IPA Stuttgart).