Material gains

Posted on January 9, 2009 at 10:59 am by Greg Corke

1831 0

Military aircraft programs have been the proving ground for carbon composite materials. Usage has now crossed over into commercial aircraft, and other areas, where these materials offer significant benefits over their metal counterparts. Corrosion and fatigue of aluminium airframes and surfaces, for example, makes maintaining these aircraft increasingly expensive. Composite structures, on the other hand, remain chemically inert, whilst being stronger and lighter than metal.

The Boeing 787 Dreamliner will have 50% of composite materials by weight, compared to just 10% in the Boeing 777; and the Airbus A350 XWB will have 52% by weight, compared to 23% in the A380.

So what are these new materials and techniques? In general composites consist of a fibre structure inside a matrix of resin, where the resin protects the structure and holds it in place; loads are transferred to the fibre structure which acts as a reinforcement. Glass is still probably the most-used fibre, but increasingly graphite (carbon) and aramid, known by the trade-name Kevlar or Twaron, are used. Resins typically comprise two types: thermosets (which cure permanently when heated) and thermoplastics (which soften when re-heated).

Exactly half of Boeing’s 787 Dreamliner is made out of composites materials. Images courtesy of the Boeing Company

Material is laid-up on a tool or mould to form the composite part using various processes. There are four main composite manufacturing processes: Laminate Ply, Tape Laying, Fibre Placement and Resin-Transfer Moulding. The part produced is then cured or hardened in an autoclave or industrial oven; some parts require placing in a vacuum bag inside the autoclave to ensure proper resin transfer across the whole part. The main difference between these processes and traditional production methods is that they add, rather than take away, material. Currently, much of the lay-up process is undertaken by hand and processes like Resin-Transfer Moulding and Laminate Ply are still labour-intensive.

Uptake of composites in automotive is increasing, but it will be at least five to ten years before they are more widely used in production models

Advertisement

Advertisement

The software challenge

This produces new challenges for software suppliers who supply design and manufacture systems for composites. The major vendors, such as Siemens PLM Software and Dassault Systèmes rely on close cooperation with Airbus and Boeing to develop design solutions, often in association with small developers close to the production facilities. Vistagy, a Waltham, MA-based company, is a recognised leader in Composite Design environments, which integrate with all the leading CAD/PLM systems. MATERIAL, a Belgian developer, is a leader in Filament Winding software with CADWIND. Tooling, such as wing panels and fuselage sections, upon which material is laid-up, is similar to that used in conventional mould and die and flat-pattern development. Manufacturing software suppliers like Siemens PLM Software, Delcam and JetCAM are already developing their business in this area. Specialised cutting, drilling, fixtures and tooling are still required to suit the particular type of composite being prepared.

Driving Composites machines

Many of the machine-tool suppliers provide their own programming solutions to drive composite machines; these can be very large (tens of metres long and wide) and operate on up to 7-axes, to apply material at the correct angle and thickness. There are three main types of machine for material placement: Automated Tape Laying (ATL), Automated Fibre Placement (AFP) and Filament winding (see box out). Other machines are used for cutting and preparing the material for lay-up and for trimming and drilling the finished part.

NC machines for material placement

Automated Tape Laying (ATL) – A numerically-controlled (NC) machine which is mainly used for flat, or near-flat, parts with uniform, or near-uniform, thickness. This limits the class of parts produced, but is quick and cost-effective (compared to manual methods). Tape, with unidirectional fibres, rather than the bi-directional fibres of woven tape, is usually pre-impregnated with resin and applied wet, though in some cases, dry-fibre, without the resin, is used. The tape is applied by a delivery head over the tool surface in a repetitive pattern to form a ply. Plies are usually built from alternate crossing patterns to produce a finished laminate with homogenous strength. Standard tape widths are 3”, 6” and 12”. Wings are an example usage.

Automated Fibre Placement (AFP) – A numerically controlled machine which is similar to an ATL but which uses a strand comprising up to 32 composite “tows”, instead of a tape. Each “tow” is a pre-calculated bundle of fibres, usually 0.125” to 0.250” wide. The main advantage of fibre placement over tape is having the ability to produce parts that have a higher degree of contour and complexity, such as aircraft fuselage panels, than can be produced using an ATL. Individual tows can be dropped and started separately, to eliminate buckles and wrinkles and adjust the width of the tow to fit into a tapered shape.

Specialised cutting, drilling, fixtures and tooling are required to suit the particular type of composite being prepared. Image courtesy of Delcam and Crosby Machining

Filament Winding – A variation of fibre-placement, the delivery head (or mandrel) is cylindrical in shape and spins on a lathe during the winding process. The resulting fibre tow is dispensed by a pay-out eye, which travels back and forth whilst the head is rotating, and can also be oriented to lay the fibre at various angles. The disadvantage of this technique is that the head does not have the compaction of an ATL or AFP, relying on the contour of the part to apply tension to the tow. Hence, convex shapes are not suitable for filament winding. Examples of parts produced using this technique include aircraft fuselage skins and fuel tanks.

Leading ATL vendors in aerospace include MAG Cincinnati, Ingersoll, Forest Line, ElectroImpact and M Torres, plus smaller players like American GFM and Accudyne Systems. Some of these vendors also provide composite design and manufacture sub-contracting services, providing a proving-ground for new technologies. This reduces the risks of using composites for their customers and promotes the take-up by industry. Our estimates suggest an installed base of around 120 ATL machines worldwide.

There are fewer AFP machines installed – perhaps 75 – with the leading vendors, in this case MAG Cincinnati, Ingersoll and lastly, Forest Line, working closely with Boeing and Airbus. These vendors also supply the controllers and CNC programming systems in a total “Composites solution” for their customers. Since volumes are small, many machines are “customized” to meet specific manufacturing requirements; this is a challenge for manufacturers, since uptake of composite parts is inhibited by the lack of “standard” machines at reasonable prices – a large ATL or AFP can cost several million dollars. Growth in the number of ATLs and AFPs required depends heavily on the production ramp-up of the current and next generation of commercial aircraft.

Fibre lay-up at Airbus. Image courtesy of Forest-Line

Some machine tool suppliers, including Zund, Gerber and Eastman, focus on cutting of the ply, or Filament Winding, like Entec and EHA. Others, such as Automated Dynamics, produce delivery heads and systems, rather than the whole machine, aimed at flexible and smaller-scale production. The smaller vendors often look beyond aerospace into adjacent industries, where composite materials also have very obvious benefits. These applications include drilling operations under high pressure and temperature; new, high-speed train networks, where weight and physical performance reduce lifecycle costs; turbines for wind-farms, where the weight and performance of the blade are directly linked; and corrosion-free storage for volatile liquids and fuels, such as petrol tanks and gas-bottles.

Market potential

Automotive is an obvious market for composite parts – carbon-fibre is already used in F1 and high-end sports cars – as the associated weight savings go some way to increasing the overall energy efficiency of a vehicle. A number of issues need to be addressed before we see wider adoption, however, not least the availability of materials. Aerospace consumes almost all the carbon fibre and resin currently in production, paying a premium to ensure supply. This makes their prospective use in production cars even more distant so long as the cost of composite materials remains prohibitive. The relative complexity and smaller size of automotive parts also remain a challenge to full automation, as does the joining of parts and the interface of carbon to metal.

Regulatory pressure also has implications for future use in automotive; European (End of Live Vehicle, ELV) regulation requires 95% by weight of new cars to be recycleable. However, these recycleable materials are currently 30% more expensive than non-recycleable composites, limiting their potential usage at the current time. There are also concerns about the environmental impact of composite production methods, involving hazardous substances.

In the current economic climate it’s all too easy to put aside longer term investments in new processes and materials. Many have argued that companies that invest in the down cycle stand to benefit disproportionately on the upswing. Carbon composites are by no means mainstream, but the businesses that can harness these assets and establish themselves in the vanguard may well be positioned to reap the benefits when usage of these materials becomes mainstream, as it surely will.
www.cambashi.com

Composites – slow on the uptake

OpenUSD backed by big name alliance for 3D content standard

Learning to melt metal

Authentise Threads adds Plyable and the Google Docs integrations

aesthetica 3.3

Toolpath expands with sales arm for AI machining

Review: SolidWorks 2015

Bunkspeed HyperShot 10

ODA announces new SDK for open access to CAD formats

Material gains

Posted on August 21, 2015 at 8:13 am by DEVELOP3D

1836 0

Physically-based rendering is a hot topic in product development. Unlike the biased renderers used for Hollywood special effects, physically-based renderers are very easy to use. This makes them perfect for designers and engineers who want a simple way to accurately visualise how a product will look in the real world.

Ford Focus interior created using the physically based real-world materials available in Iray+

Materials play a critical role. The big advantage is that, as you can define them to be physically accurate, there’s no eyeballing, there’s no fiddling with controls. You create a material, say brushed aluminium, and, as long as it is set-up correctly, you don’t have to edit it at all, unless you want a very specific effect, such as an oxidisation layer on top.

In many cases the manufacturers of the physical materials also produce accurate rendering materials for designers to use. So rather than trying to make your own steel material, you specify the type of material you’re using, then download it off the manufacturer’s website — or at least the data you’d need to create it.

Measured materials

The interesting thing here is that it can extend beyond the visual properties of the material. You can encode all kinds of information into the material that CAD and PLM systems can reuse, such as tensile strength — all those non-optical properties.

You can also link it to finishing processes in the manufacturing stage so you could start with an aluminium with measured data, then specify the brushed finish, a clear coat — and represent that in the material as both an instruction to manufacturing as well as visualisation. It’s an interesting area, but it’s one that’s not been addressed yet.

I do think this will happen, but it needs to be driven by two sides. The first is that the manufacturer has to be willing to give out that data. The second, is that the CAD systems need to support it by being able to read these descriptions.

In Nvidia’s Iray and our own physically based rendering technology, Iray+ (which is built on Nvidia’s Iray) you’ve got materials that can represent the whole of the lighting equation in terms of properties. You can encode and create a material that looks exactly like it would in the real world.

The other approach is to measure the materials. Companies like X-rite, which produce huge machines that cost more than a semi detached house in Sheffield, measure the optical properties in full. You can then take that data and use it to generate your materials.

It’s an area we’re exploring at Lightworks as it gives you a highly accurate representation of reflectance properties so there’s absolutely no need for you to change the material at all.

The downside is that there is no standard at the moment for encoding that data, so each company that has a measurement machine to sell will have a proprietary format. They don’t see it being in their interest to have a standard, as the data would be interchangeable between systems and they’d lose their exclusivity.

Iray+ can take data from X-rite and represent it to give you a superb representation. At present there are not many people using this method.

They are mostly car manufacturers and they’ll use this for automotive interiors, so they can see exactly how much light is reflected inside the vehicle. They’ll also use it for paint finishes so they have that absolute representation for design review.

Measured material –
car paint – in Iray+

The additional Fresnel layer creates the reflections in this version

The inevitable caveats

There are some caveats to these technologies. The measurement machines are prone to errors. They measure the materials in a 180 degree arc, they shine a light on it, and measure the light reflected at each angle in that arc. The problem is that when you get to very shallow angles (one or two degrees) or at 90 degrees, you may get errors, and have to correct the recorded values. None of that has been standardised yet, but it’s going to be a big growth area.

To create an easy to use API in Iray+ we’ve added a layer on top of the Iray API and simplified a lot of standard things you need to make integration easy for software vendors.

The other big area is how materials are defined. We use MDL (Material Definition Language) from Nvidia. It is open and it’s an abstraction layer, so you define the material and the graphic system interprets it how it needs to — the idea being that you could use it in any system that supported it.

A number of key applications have already provided support for MDL and I think others will follow.

At Lightworks we’ve taken a novel approach to material editing. We have created a set of components that are standard and reusable blobs of MDL that represent different types of reflectance characteristics. So you can take a standard MDL component to create a glossy surface, sub surface scattering, a matt or steel for instance.

To give this some context, generally with graphics systems, developers provide huge shaders requiring an unwieldy interface, which has everything, including the kitchen sink.

The user simply ticks on the bit he or she wants, say an anisotropic reflection, where you get elongated highlights (such as brushed metal) and adds on some sub-surface scattering.

For Iray+ we’ve made things simple. You build up a material with components, rather than picking effects from a monolithic list. We’ve defined our components in terms of classes; geometry, base, coating and surface. The idea is to present the user with a real world mental model for material creation.

To create a car paint material for example, you might start off with a metal surface, add an undercoat layer, then a paint layer with perhaps some flake, then a clear coat layer on top.

You can then go in and add additional layers such as dust and dirt if you want to. So, instead of getting involved in huge shaders that use arcane language, you’re actually just building materials up as you would in the real world.

With designers and engineers primarily being in the business of designing and engineering, we see this as critical.

Visualisation should be a simple process from start to finish and by making it easy to work with materials, physically-based rendering can become even more tightly integrated into the product development workflow.

■ lightworks-iray.com

DEVELOP3D Visualisation For Design And Engineering Special Guide

Download our free guide to the latest viz software and hardware, plus exclusive pro tips and views on where everything is heading in the future. Not signed up to DEVELOP3D yet? It’s simple and free to do, and you can download our monthly magazine for product designers, engineers, makers and anyone else wanting to join the party – Sign up here.

About the author

Dave Coldron

As the Lightworks’ Product Director, Dave has specific responsibility for the development of Iray+ and its use in applications and plug-ins.

With over 20 years’ experience in developing integrated systems for the computer graphics industry Dave knows how to create applications that support the design workflow; focusing on the use of compelling digital content, interactive design and the user experience

The latest trends in materials and shaders for physically-based renderers
Default