Aircraft designers could make more efficient use of composite materials following the development of a computer model that predicts the material’s behaviour when damaged.
New codes developed at Imperial College London form the basis of the computer model that will let designers explore damage tolerance in much greater detail and bring advantages such as reduced development cycles and R&D costs, plus lighter, more fuel-efficient aircraft.
The project, undertaken with partners Airbus and funded by EPSRC, sought to predict how cracks propagate through a composite part when subjected to anomalous events such as bird strike or an extremely large turbulent flow.
‘It is true that composite materials have cracks within them with lengths in the order of micrometres, and those cracks are everywhere and aren’t something to be worried about,’ explained project leader Dr Silvestre Pinho. ‘The focus of this work was in situations where cracks actually grow so that we can predict how they grow.’
Carbon fibre reinforced composites are an attractive structural material because they are lighter than the metals they replace and possess favourable strength-to-weight and stiffness-to-weight ratios.
A quarter of the Airbus A380 and half of the Boeing 787 is made from composite materials but EPSRC note that aircraft designers have so far overcompensated for a lack of knowledge on how composites behave by over-reinforcing composite panels.
Pinho explained that the types of composites his project focused on are carbon fibre reinforced composites (CFRP), which are carbon fibres in a polymeric matrix. In these composites the carbon provides strength and stiffness and the softer matrix provides load transfer mechanism between the fibres.
‘The way they fail is so complex,’ he said. ‘In metals, for instance, you have one well-understood mechanism through which they break; with composites, you can have debonding between the fibres and the matrix, you can have breaking of the fibres, you can have yielding of the matrix, and you can have delamination between different plies that have fibres oriented in different directions.’
In order to understand failure mechanisms in the composite material the team designed specimens that would break in a specific manner. They then visualised those failure mechanisms using optical and scanning electron microscopes.
Pinho said: ‘With that we built physically-based models for those mechanisms. The physically-based models are really something which can be expressed in a set of equations.
‘Those physically-based failure models are then used as an input for numerical failure propagation models that we’ve developed, which are based on a numerical technique called finite elements (FE).
‘Our failure models sit inside FE software, which allows us to [make predictions] using homogenised finite element descriptions of the composite material. In reality, inside each finite element, the material is modelled at a lower scale analytically through those equations. In this way, we obtain a good compromise between having an accurate description of what physical processes lead to failure and, on the other hand, a model that can be applied at an acceptable computational cost in a reasonably large structure.
‘The way industry can use our models is for instance through commercial software packages, either using our sub-routines which we can supply, or, as in the case of LS-Dyna, a version of our material models is available directly with the package.’
Materials developers can benefit also from the physically-based models developed at Imperial as they relate performance of the composite to the properties of fibres, of the matrix and their interface.
Pinho believes a more thorough understanding of damage tolerance and materials development will have added environmental benefits too.
Up to eight per cent of jet fuel could be saved if less material were used to build large aircraft parts for aircraft. Pinho added in a statement that this amounts to approximately 20 million tonnes of fuel, 50 million tonnes of CO2, and £20bn saved annually.
Personally given our lack of experience with these materials in such critical areas I’d prefer the overcompensation.
When the model has got 20 years of real world experience in these applications behind it then I’ll be happy. Until then lets think like the Victorians with new innovations: over engineer it!
I can think of other areas of research where models have been used extensively to predict outcomes that haven’t thus far come true but which have cost society a lot of money.
I am looking forward to giving my ‘composites are NOT very thin flexible metals’ lecture again: The whole composition (sorry!) of composites does require a different mind-set to that necessary to sensibly design and stress a structure based upon crystalline materials. We, based in textiles (you know, that low technology industry that clothes, fashions, functions, covers, protects, supports, …the majority of the population of our world) look forward to assisting other newer, high-tech industries (just as our former guru Mr Jacquard did for computing) with their education.
As I have pointed out: in aircraft manufacture, you used to take large chunks of metal and cut and manipulate them to the shapes you need: in textiles we have always done this the other way round. We take small pieces of material and manipulate them together to get the shapes we want. There is much we can offer. Ask us!
Best
Mike B
Interesting point from Mike Blamey.
Would ‘sewing’ composites together rather than bonding allow some de-stressing like riveting metallic sheets?
Rob Watson’s question is highly relevant. In fact my lecture/talk was given at a well known aircraft manufacturer with HQ located not a million miles from Seattle: and came about because a university friend, was a production Engineer there during the major 747 build programme in the early 70s: and knew of my ‘textile’ background. he invited me to speak. One of the things I saw during my factory tour was the drilling of holes in bonded composite structures (in order to ‘fix’ them into place with rivets!) I pointed out that it was (and still is!) perfectly possible to ‘weave/lay’ holes into pieces for this purpose! I welcome any requests for additional assistance from Engineers involved in this area.
[No charge!] I do believe that we are all guilty of sometimes NOT thinking outside the ‘box’ -and only doing so in areas in which we are comfortable. But, as our Editor often points out, particularly in the training/education of our younger colleagues…(and notwithstanding the attention of HR departments who seem to seek graduate Engineers already ‘fully-formed’) it is the ability to look sideways (as well as forward) that sets us apart.
NSN again: “Engineers can (if they are encouraged and allowed to do so -MJB’s addition) do for 10p what any fool can do for £1.”
best
Mike B