This paper focuses on the modelling of compressor data for use in real-time engine models where speed of execution is critical. The objective of this investigation was to develop a new technique for interpolation and extrapolation of manufacturer data to create extended, denser maps.
Dynamic simulation and emulation methods are increasingly being employed due to the escalating difficultly in matching a turbocharger configuration to an engine, with today's ultra-low levels of regulated emissions. The turbocharger sub-model is a vital part of the of the complete engine model, therefore it is important that it is reliable and accurate. Turbocharger maps are typically presented in graphical form and do not cover the entire operating range of the engine, leading to difficulties for inclusion in simulation.
Out of the techniques studied, a combination of spline and parametric fitting modelled the data most successfully. Comparison against experimental data proved it to be accurate but also illustrated some of its limitations. Benchmark tests showed a significant execution time benefit of the new map created by the technique over the use of the original map.