Positive reinforcement: 21st century 3D weaving

The art of weaving by loom could hold the key to producing stronger and more complex carbon-fibre structures.

What does a textile loom invented more than 200 years ago have to do with 21st-century aircraft engines and high-performance cars?

It turns out the centuries-old art of weaving by loom may hold the key to producing much stronger and more complex carbon-fibre composite structures.

Composites are traditionally produced by placing layers of material on top of one another and infusing them with resin.

Looming prospect: weaving could result in an impact-resistant 3D composite material

However, this means that there is nothing holding the layers of material together, other than the resin itself. As a result, composites can start to come apart when a load is applied to them, in a process known as delamination, said Dr Jody Turner, a research engineer at Sheffield University’s Advanced Manufacturing Research Centre (AMRC) with Boeing. “It can result in a problem called inter-laminar shear stress, which is where the plies start to separate,” she said.

By using a loom to weave the material, engineers can introduce a reinforcing fibre running through the thickness of the fabric, to connect the layers together. This results in a 3D composite material that is better able to cope with impacts, or loads such as bending, without delaminating.

Research by Ulster University and spinout 3D weaving company Axis Composites has found that the materials exhibit fracture toughness up to 20 times higher than conventional epoxy laminates.

Everything old is new: weaving machines are still based on the two-centuries old fabric loom

Distinctive properties can also be woven into different areas of a single composite structure, said Dr Edward Archer, a lecturer in the Engineering Research Institute at Ulster University and technical director of Axis Composites. This means the properties of the material can be altered at different points to best react to the loads acting upon it, he added.

So, for example, the stresses acting upon a material at curves in a structure will be in a different direction to those on flat sections. But by adding more reinforcing fibres in the direction of these stresses, it can counteract the load, he said.

What’s more, 3D weaving also allows you to produce more geometrically complex composite structures, said Andrew Long, professor of the mechanics of materials at Nottingham University, and director of the EPSRC-funded Centre for Innovative Manufacturing in Composites (CIMComp).

“So, for example, you could weave a flat piece of material on the loom, with the layers all connected together, but then somewhere your material might split into two layers that sit on top of each other,” he said.

This can be done by ensuring the reinforcing fibres do not reach all the way through the thickness of the material, he said. “Then when you take it off the loom you can open it up to produce something like a T shape or an I beam – very important structural elements that are regularly used for stiffening in aircraft wings,” he said.

Although relatively new to the composites industry, the weaving process – in which a weft thread is inserted through a warp thread at a 90º angle within a loom – has remained relatively unchanged for hundreds of years.

“With a 3D weaving machine, it is just a matter of controlling how you lift the warp yarns, and how many weft yarns you insert before you move the fabric along,” said Long. “So you can insert a number of weft yarns, one on top of another, and control the warp yarns so that some of them form through-thickness reinforcement fibres.”

While some companies are developing their own looms, many use conventional Jacquard machines, which are widely employed in the textile industry to weave multilayered patterns, such as brocade bedspreads. Invented in 1801 by Joseph Marie Jacquard, when punched cards were used to guide their operation, the looms are now controlled by electronics.

“You have a lot of design freedom, in terms of the pattern you can make your yarns follow within your woven structure,” said Long. “You can vary the patterns of your through-thickness reinforcing fibres in lots of different ways.”

Much of the interest in 3D weaving to date has come from the aerospace industry. The landing gear on Boeing’s 787 Dreamliner, manufactured by Messier-Bugatti-Dowty, is equipped with fibre-reinforced composite braces manufactured using a 3D weaving process.

A traditional laminated composite would not be able to withstand bird impact or runway debris, or cope with the complex axial and shear loads that landing gear face during the course of a flight, according to Albany International, based in Rochester, New Hampshire, which wove the braces for Messier.


Albany is also weaving 3D composite fan blades for the LEAP turbofan aero-engine, which is being built by CFM International, a joint venture between Snecma and GE. The 3D composite structures, or preforms, are woven on a Jacquard-type loom, and then cut to the desired shape using waterjets. They are then injected with resin in a mould.

The 3D woven fan blades and casings will reduce the weight of each aircraft by around 500kg, according to a spokesperson for Snecma. The LEAP engine is due to enter service on the Airbus A320neo in 2016, and the Boeing 737MAX in 2017.

The technology is also beginning to find its way into the automotive industry. Earlier this year, Albany announced a partnership with Ricardo to explore the use of its 3D composites in automotive applications. Meanwhile, Toyota is using a loom to weave a carbon-fibre front crash box for the Lexus LFA supercar. It has also developed a circular machine that can braid individual fibres together using an automated template, to produce the side rail for the car’s front windscreen.

But despite this progress, engineers may still just be scratching the surface of what these looms can do, in terms of composite development, said Long. “At the moment, manufacturers tend to restrict themselves to two or three styles of 3D woven material, broadly speaking, because they don’t really have any data on any other patterns on which to base design decisions,” he said.

The LEAP engine has 3D woven fan-blades from Albany and will enter service in 2016

To this end, researchers at CIMComp are modelling the performance of woven composites, with the aim of being able to modify the design of the weave for different applications, to ensure the most significant weight savings, for example.

Dr Xuesen Zeng, a research fellow at Nottingham University, is using a genetic algorithm to find the most suitable weave for a given application – in this case aircraft landing gear braces – using the process of natural selection. The algorithm mutates and then breads the ‘fittest’ weave patterns from a random selection to find the optimal design.

Computer analysis of his selected design has shown it is able to withstand buckling twice as well as conventional laminated composites, and 50 per cent better than a 3D material with a commercially available orthogonal weave, Zeng said.

Working with researchers at Manchester University, and with composite manufacturer Sigmatex, based in Runcorn in Cheshire, Zeng ultimately plans to produce a 3D weave of his selected design, to demonstrate that it has the improved load-bearing properties he predicts.

Meanwhile, Turner and her colleagues at the AMRC, which has recently purchased its own loom, are hoping to soon begin weaving their own material in a bid to learn more about the resin transfer moulding (RTM) process, in which resin is injected into the fibres in a closed mould.

At the moment, researchers cannot be sure exactly what happens once the resin is injected into the mould, according to Turner. “To enable you to predict where the resin will go, you’ve got to understand what is going on inside the mould – and at the moment we just don’t,” she said.

By controlling the position of the weft and warp threads during the weaving process, Turner hopes to introduce deliberate errors into the material that will allow her to determine if these then affect the resin flow through the fibres, as she expects. “In this way we can start to build up our understanding of what is happening and why,” she said.

This feature originally appeared in our sister magazine MWP Advanced Manufacturing