Abstract
Computational Fluid Dynamics (CFD) simulations have a great potential to guide the optimization of fuel stratification strategies for internal combustion engines, but well-validated spray models are required. In this study, we aim to understand the current capability of CFD simulations, under conditions representative of partially stratified advanced compression ignition engines, to predict gasoline sprays through quantitative comparisons based on liquid measurements. A series of backlit extinction imaging is carried out in a constant volume vessel under simulated engine-like cold start conditions. As a test fuel, regular E10 gasoline is injected using a commercial gasoline direct-injection (GDI) injector at two different injection pressures of 50 bar and 100 bar. High-speed imaging datasets were used to obtain quantitative measurements of the liquid penetration, and local liquid volume fraction from line-of-sight integration and computed tomographic reconstruction. Additionally, geometric information by x-ray computed tomography is provided to set initial and boundary conditions for the CFD simulations. Multidimensional large-eddy simulations (LES) of the gasoline fuel spray in a constant volume chamber are presented and compared to the experimentally obtained fundamental validation data. A well-validated surrogate fuel for regular E10 gasoline, called PACE-20, was used in the simulations. Detailed comparison between experiments and simulations shows faster spray development and evaporation of LES studies, leading to slightly longer liquid penetration lengths. Further, the numerical simulations were able to capture the strong plume collapsing of the tested injector under these conditions and to properly reproduce the experimental trends and the effects of injection pressure on the liquid spray.
