Prediction of Turbo Compressor Maps using CFD

Abstract

Continuous improvements are of great importance in the heavy vehicle industry. Computational fluid dynamics (CFD) can make the developing process easier and can be used for simulations of the compressor in a turbocharger. The turbocharger is used in the environment of heavy vehicle engines and is delivered with a turbo compressor map. What is wanted as an outcome from this thesis is a toolchain, as automated as possible, for predicting turbo compressor maps using CFD. The used software is STAR-CCM+ and an investigation of parasitic currents was done followed by a creation of a turbo compressor map from simulations. The result from the first part gave information about using a rotating reference frame in simulations. The overall results from using CFD to compute a turbo compressor map got close to experimental data available in the project. An automated toolchain was developed containing a similar approach as developing the turbo compressor maps in the project. The realizable k-epsilon turbulence model was used and other turbulence models would be interesting to investigate in the future. Even finer meshes than the used ones could be tested further as well as introducing more truck-like intake systems.