Simulation is an increasingly common aid in the development of advanced products like cars, lorries and even aeroplanes. Computer modelling is significantly more efficient and cheaper than building prototypes to be tested in a wind tunnel or actual flight test. The ambition of the aviation industry is therefore to run more computer tests, and today’s ever increasing computing capacity is making this possible.
Being able to calculate flight properties, engine performance, vibrations and noise at an early stage has both technical and financial advantages. Obviously there is a need for simulation tools and well-designed mathematical models. So far it has been possible to produce reliable results with normal, “steady” flow. But anyone who has ever sat in an aeroplane knows about turbulence – airflows that twist and turn – and how it is not possible to predict its strength or direction.
Investment i efficient modelsVinnova has decided to invest SEK 7.5 million (ca EUR 860,000) over three years into developing efficient computational models for turbulent flows. The recipient of this funding is SAAB, who are also contributing to the research carried out by Linköping University; the Swedish Defence Research Agency; and Chalmers and the KTH Royal Institute of Technology in close collaboration with SAAB and Creo Dynamics, an innovation and consultancy company in Linköping specialising in acoustics, aero acoustics and aerodynamics.
Two PhD students and seven senior researchers will take part, as will one industrial PhD student from Chalmers and one PhD student under Jan Nordström, professor of Computational Mathematics at the LiU Department of Mathematics.
“There are methods today with the help of which we can compute turbulence, but we need greater computational power and they don’t work for large angles of incidence, such as when an aircraft is taking off or landing. But in four years, when the project is over, it will be possible to calculate the flow for very steep angles of attack,” Nordström says.
Refining modelsWhat Nordström and one newly-recruited PhD student will do is refine the computational models currently being used. Instead of making calculations at a large number of points, we should be able to calculate more precisely at a smaller number of points by using higher-order methods.
“We do this by solving advanced equation systems effectively, and we believe we can do this for higher-order methods too, something we are good at here,” Nordström says.
Out of these funds, SEK 1.8 million (ca Euro 200,000) will come to Linköping University, if the project goes according to plan. The coordinator and cluster leader is Per Weinerfelt of Saab Aeronautics, who is also adjunct professor of Computational Sciences at LiU.
Research in the project will be carried out in three areas: models for simulating turbulence (Chalmers and the Swedish Defence Research Agency), improved and more precise numerical methods (LiU, the Swedish Defence Research Agency and Saab), and aeroacoustics as well as improved opportunities for simulating noise and vibrations (Creo Dynamics and the KTH Royal Institute of Technology).
“Chalmers and Saab, with the help of an industrial PhD student, will develop a simulation model which will then be used by our refined computational model. The same technology may then be used to calculate sound and noise, something that Creo Dynamics is interested in,” Nordström says.
Over the long term, this research will strengthen the competitiveness of the Swedish aviation industry, but the models are so general that they can just as well be used in the motor vehicle industry and in the development of wind or water power.
Picture: Center for Turbulence Research, Stanford University.