LAUSR

Abstract The start of injection (SOI) transition processes present challenges for diesel spray combustion simulations by computational fluid dynamics (CFD). In the previous studies, claims are usually placed on the experimental uncertainties or numerical methods such as grid size selection and sub-model comparison. Until now, few studies have been attempted to revise the fuel injection rate (FIR) profile used as the fuel input boundary condition for numerical models. It this paper, considering the difference of in-nozzle flow between the Bosch long-tube method and actual injection, the ''Velocity Revised'' and ''Mass Revised'' FIR profiles are formulated to keep the fuel injection velocity and fuel injection mass consistent with the actual injection, respectively. After a brief description of the experimental and numerical methods, the effectiveness of the newly developed two FIR profiles and the uncorrected and the Engine Combustion Network (ECN) FIR profiles used as boundary conditions for CFD simulation is validated against the experimental data under both the non-reacting and burning conditions. The simulations with the uncorrected FIR profile show remarkably lower spray and flame penetrations at the initial stage of injection, while the simulations with the newly developed two FIR profiles exhibit almost identical spray and flame penetrations with the experimental data. The accuracy of the ECN FIR profile simulation is between the above two cases. These results are believed to be valuable reference for researchers to evaluate whether the potential shortcoming in diesel spray combustion simulations is from the variation in boundary condition or the code itself.