Wind of change

UK engineers are close to achieving a world first, by building a magnetic levitation system for aerodynamic testing of Formula One racing cars.


If successful, the team at Durham University will be the first to develop the technology for large-scale wind tunnels, which could significantly improve testing by eliminating the interference caused by physical supports.


The project is part of a drive to develop ever more sophisticated aerodynamic experiments for racing cars, passenger vehicles and even buildings.


The use of maglev technology in aerodynamic experiments has been attempted before, most notably by NASA, but with limited success. In the 1980s NASA built a system with the aim of using the technology in space shuttle testing, but its prototype was only able to support models in a tiny wind tunnel 150mm across, while consuming a whopping 8MW of power.


The Durham team, led by Dr David Sims-Williams, is assembling prototype systems to be installed in the university’s 40 per cent wind tunnel.


On the road, cars move through still air and over a stationary surface. In conventional wind tunnel tests this is reversed — the car is held still while the air moves past. As racing cars are so close to the ground the interaction between the floor and the underside of the car is very important.


So to generate a more accurate simulation, F1 teams also make the ground under the car move, using a rolling road. However, the use of a moving belt means the car model has to be supported by a strut, usually from above, and this interferes with the airflow around the vehicle.


The strut wake particularly affects the sensitive aerodynamics at the rear of the vehicle. So the Durham researchers have been developing the maglev system to eliminate this interference, which can have significant effects on the measured drag and downforce.


‘You do away with all these external struts, so you get a truer airflow around the car, and you effectively support the model invisibly using magnetic levitation,’ said Sims-Williams.


The technology is proving more suitable for use with cars than aircraft or spacecraft, as the magnetic airgap between the device and the underside of the vehicle is quite small. Aircraft obviously have to be positioned further away from the walls of the wind tunnel, and magnetic forces drop off very quickly as the size of the airgap increases.


As NASA found to its cost, this means that the magnets needed to support the aircraft must be very powerful. The researchers have been investigating two options for the system, and have built prototypes of each, both based on high-intensity, rare-Earth permanent magnets and electromagnets.


The first approach combines these with an electromagnetic feedback system to stabilise the car. ‘If you try to support the car over straight magnets it will always be unstable and skid off to the side. We have a feedback system where we measure where the car is and then use electromagnets to modify the magnetic field to manipulate the car horizontally, so we’re stabilising what would otherwise be an unstable system. It’s a bit like trying to balance a pencil on the end of your finger,’ he said.


However, with an active control system any local power cut would mean a loss of levitation stability, causing the model to crash into the moving belt, which can each cost up to £6m.


So the team is also developing a system based on stable superconducting levitation, to eliminate the need for active control. The superconductors, which would continue to levitate the model without any external power, operate at 77 Kelvin, or around –200º so require cooling with liquid nitrogen.


The aerodynamic forces acting on the model are measured using a conventional under-floor strain gauge balance, but the researchers are also considering using the currents applied to the electromagnets and the relative position of the model and supporting magnets to take these measurements.


The two prototypes are due to be completed by the end of this month, when the researchers will demonstrate them to F1 teams for their feedback, before deciding whether to commercialise the system or carry out further research work first. Although the technology is initially likely to be most attractive to F1 teams, who use the most advanced wind tunnel techniques, it could also be used in the production of standard cars, where developments to the underside of the vehicle could lead to significant aerodynamic improvements without any impact on styling.


In the field of aerodynamics research, F1 teams and conventional car makers have invested considerable sums in computational fluid dynamic (CFD) techniques over the past 15 years.


According to CFD specialist Fluent, Sauber has recently been using a supercomputer to model two F1 cars overtaking each other.


Companies are beginning to see the fruits of this investment, particularly in areas such as cooling flows, or the flow through engine compartments and brake systems, said Dr Martin Passmore of the aerodynamics research group at Loughborough University. But for predicting the aerodynamics of the whole vehicle, computer models do not yet deliver the goods, he added.


‘Although they are spending money on it [CFD], plenty of people are still building new wind tunnels. Red Bull Racing is just completing a tunnel in Bedford and DaimlerChrysler recently spent $35m (£19.2m) on a new tunnel,’ he said.


As vehicle refinement — the art of controlling noise, vibration and handling — has become more advanced, car makers have increasingly focused on areas such as aeroacoustics and unsteadiness. Unsteady loads on vehicles are caused by travelling through disturbed air created by ambient cross winds or driving in the wake of another vehicle, reducing the vehicle’s refinement.


Car makers would like to be able to control unsteady loads, but researchers do not yet fully understand the relationship between conventional steady-state and unsteady loads. Unsteady flows also cause wind noise, and as other aspects of design have made vehicles quieter, such aeroacoustic affects have tended to intrude more.


But while car manufacturers would like to reduce wind noise, to do so requires sophisticated experimental facilities, said Passmore. ‘


As soon as you get into problems of unsteadiness, numerical techniques just aren’t sophisticated enough,’ he explained.


‘Although some of the basic modelling tools partly exist, they are not practical for a whole number of reasons; they take far too long to get a first calculation and are insufficiently accurate once they do produce a result. And of course car manufacturers aren’t interested in the details, they’re interested in the outcome — that’s something they can deal with much more effectively experimentally.’


Some people have been lulled into a false sense of security by CFD, he said. ‘Some computational methods give people a warm feeling because they produce fantastic animations and colour pictures, but they’re not well validated, particularly for automotive work.’


It is not just in-car development that researchers are attempting to build more advanced measurement techniques for carrying out physical experiments.


Researchers at Manchester and Kent universities are developing pressure- sensitive paints for wind tunnel testing, in a project involving Airbus UK.


The paints can be used to monitor flow phenomena such as pressure, heat transfer, temperature, and turbulence, as well as in engine research for studying the combustion process. The paint is illuminated with light of an appropriate wavelength. This excites the paint, causing it to emit radiation, usually in the infra-red region. This radiation is then measured and correlated to determine pressure, temperature, or skin friction.


The paints can be re-used and could improve wind tunnel testing by significantly reducing the number of mechanical sensors needed, which obstruct the flow around the model. By limiting the number of sensors the technique should also reduce the cost of the experiments.


Meanwhile, researchers at Oxford University are attempting to improve wind tunnel experiments by fooling the vehicle being tested into believing it is not attached to anything.


The team, led by Dr Marko Bacic, lecturer in control engineering at the university, is developing a system to create ‘invisible’ struts using a combination of hardware and computer simulation. Isis Innovation, the university’s technology transfer arm, is negotiating with various companies to commercialise the system.


The technology could be particularly useful in the development of unmanned aerial vehicles, and more specifically their flight control systems, to ensure that such craft can operate in all conditions without the costly and risky process of full test-flights.


‘We’re trying to simulate free flight in a wind tunnel, so that the aircraft doesn’t feel the difference between flying autonomously in free space and flying in a wind tunnel,’ he said. The system will use sensors to take measurements in the tunnel and use these to calculate how the aircraft would behave in free flight.


The ‘transparency’ of the simulation will also be measured, to ensure the aircraft performs as if it were unfettered, and to determine how closely the simulation replicates free flight. Although the technology is being developed mainly for aircraft simulation, with modifications it could also be used to examine the dynamic performance of cars as they go around bends at high speeds. This could help in the design of suspension control systems and in evaluating vehicles’ safety-critical components, said Bacic.


As aircraft and cars get more complicated, it is becoming increasingly difficult to model them mathematically, he added. ‘Once you start getting into more complex systems, which have thousands of parts, it is extremely difficult to find a full set of differential equations and it would be extremely computationally intensive.’


So researchers will continue to develop increasingly advanced and ambitious experiments, he said. ‘You try to extract the necessary information experimentally. This may not be the perfect detail of the system, but as long as you ensure you can control the system, to make it do what you want it to do, that is what’s important,’ said Bacic.