Researchers look to simplify dynamic systems design

The five-year project has received £4.2m funding from the EPSRC.

The current problem, explained project director Prof David Wagg, head of structural dynamics at Bristol’s mechanical engineering department, is that design tools — even advanced finite-element analysis systems — can only deal with structures that react linearly to the forces that act on them. ‘But in the real world, things are more complex,’ he said. ‘For example, if you’re dealing with a large wind turbine rotor, the blades will bend under their own weight, even when they’re static. When they’re moving, their behaviour becomes much harder to model, and you have to avoid a situation where they’d flex so much that they could actually hit the supporting tower.’

Similar problems occur with very large buildings, Wagg said. ‘We can model a one-storey building, of course. But large buildings with complex geometry can move and flex in many modes, and if you have to model the effect of large forces, such as earthquakes, on them, then they quickly start to behave in a non-linear manner.’

The problems also extend to very small objects, such as the probes of scanning atomic-force microscopes, which use a sharp point mounted onto a cantilever to scan across a surface and provide an image of its texture. These cantilevers can vibrate in a non-linear manner, distorting the image, Wagg said.

‘You can model these systems to a degree with FEM systems,’ Wagg said. ‘But it’s very difficult to interpret the results from these simulations, and they often don’t agree with experiment.’

The research project, which also involves the universities of Cambridge, Sheffield, Southampton and Swansea, will combine mathematical modelling with experimental work including wind-tunnel testing, Wagg said. Industrial partners Stirling Dynamics, EDF Energy, Airbus UK, GL Garrad Hassan, Rolls-Royce, Romax Technology, Agusta Westland and the ESI Group will also contribute.

‘We’re looking specifically at technologies such as wind and tidal turbines, as well as aircraft and some medical devices, such as cochlear implants, which are a particular interest of Steve Elliott of Southampton University,’ Wagg said. ‘In some cases, we’ll be working with software from our industrial partners; for others, we’ll generate our own codes.’

The final goal of the five-year project is to develop software that engineers can use to model dynamic, non-linear behaviour. ‘You could, for example, use this to tailor the properties of a composite component to give it more strength and flexibility in the places where it’s needed,’ Wagg said. ‘Ultimately, this might become a standard module in FEM or other computer-aided design packages.’