Oct. 12, 2010 A new simulation model developed by researchers at the University of Bradford is set to significantly reduce the time and costs required to calibrate a new engine, enabling car manufacturers to get new vehicles to market much more quickly.
The simulation model will also be easily transferable to different engines within the same family, reducing development time and costs by up to 80 per cent.
Modern engines are fitted with engine control units (ECUs) which manage the fuel/air ratio to maximise performance while keeping fuel consumption and emissions to a minimum. As each driving situation -- such as urban or motorway driving -- requires a different input, so the ECU draws on data created specifically for that engine through mapping and calibration models.
Traditional 'black box' models use statistical calculations based on the output generated by a given input during engine bench testing, so any changes to the engine require a new model to be created, which is both costly and time-consuming.
The system, developed by the University's Centre for Automotive Engineering, is based on the physics within the engine responsible for creating the output. By taking into account what is actually happening within the engine to generate the desired output, the model is more precise, will generate new data more quickly and can more easily be adapted for different engine capacities or compression ratios.
"Physics-based models are fairly common in other areas of engineering but have never been adopted by the automotive industry for ECU programming," says Professor of Mechanical Engineering, Kambiz Ebrahimi, who led the research. "We believe that this is the way forward for more efficient and effective engine mapping and calibration and that these kinds of models also have the necessary flexibility to help the industry stay ahead of new emissions regulations."
European legislation currently covers emissions and fleet average economy at given speeds, but this is likely to be extended to cover acceleration or deceleration between different speeds -- known as 'transient behaviour'. The physics-based model can generate data for transient behaviour -- unlike statistics-based models -- so will enable manufacturers to ensure they meet new legislation without greatly increasing the time and costs required for mapping and calibration.
Professor Ebrahimi and his team are already speaking to car manufacturers about taking the model forward into industrial use.
Other social bookmarking and sharing tools:
Note: If no author is given, the source is cited instead.