The project "GMMC--Combustion Engine Optimisation" is a cooperation between GMMC, Volvo Car Corporation and Volvo Powertrain. The goal of the project is to develop best practice and a software demonstrator for combustion engine optimisation based on computational fluid dynamics.
It is nowadays possible to simulate the physical and chemical processes inside combustion engines by appropriate software and high performance computers. These simulations can predict, for example, fuel consumption and emission of soot and nitrogen oxides. By varying the design parameters of the engine, different configurations of these can be simulated and their performances compared. In this project we use global and local optimisation algorithms in combination with the simulations to find optimal design parameters with respect to a goal function.
Evidently, it is favourable for combustion engines to have a low fuel consumption as well as a low emission of soot and nitrogen oxides. The goal function, defined by the designer, combines these different goals. Since the engine must work well in certain ranges of load and speed conditions, it is necessary to take several such conditions into account in the optimisation. Moreover, since an engine simulation can consume several days of computer time, it is clear that very efficient optimisation algorithms are required.
The figure shows a sector calculation of the cylinder of a combustion engine. The droplets consist of fuel and their colors indicate their temperature. The grey object is an isosurface for the concentration of soot and the plane with a coloured contour plot illustrates the fuel concentration.