3D proper orthogonal decomposition analysis of engine in-cylinder velocity fields

Abstract

Cycle-to-cycle variation of flow fields inside an internal combustion engine has a significant influence on the combustion efficiency and emission. Meanwhile, the velocity field inside the engine is very complex because of its turbulent and transient nature. Particle imaging velocimetry (PIV) has been widely used to investigate the instantaneous flow field within optical engines and various numerical tools were used to analyze the velocity fields. This method is limited since PIV is usually 2D or 3D for a small volume, which cannot reflect the overall flow field characteristics within the cylinder. This work utilized 3D computational fluid dynamics (CFD) simulations validated by sensor measurements and PIV measurements using an optical engine. Then, 3D proper orthogonal decomposition (POD) analysis was performed to analyze large scale and small-scale velocity characteristics, as well as vortex structures. This study then introduces an improved 2D POD scheme with multi-plane measurement and its performance is then compared against volumetric POD results. Finally, optimal plane parameters are found for practical experiments considerations and the multi-plane scheme was discussed for practical experimental investigations.