Modeling Fuel-Air Mixing, Combustion and Soot Formation with Ducted Fuel Injection Using Tabulated Kinetics

Ducted Fuel Injection (DFI) has the potential to reduce soot emissions in Diesel engines thanks to the enhanced mixing rate resulting from the liquid fuel flow through a small cylindrical pipe located at a certain distance from the nozzle injector hole. A consolidated set of experiments in constant-volume vessel and engine allowed to understand the effects of ambient conditions, duct geometry and shape on fuel-air mixing, combustion and soot formation. However, implementation of this promising technology in compression-ignition engines requires predictive numerical models that can properly support the design of combustion systems in a wide range of operating conditions.

This work presents a computational methodology to predict fuel-air mixing and combustion with ducted fuel injection. Attention is mainly focused on turbulence and combustion modelling. The first is mainly responsible for the mixture formation process in presence of large velocity gradients and flow recirculations, while the second must include detailed kinetics and turbulence chemistry-interaction to correctly predict ignition delay and flame structure. Literature experimental data were used for model assessment and validation under different ambient conditions considering both free-spray and ducted fuel injection configurations. Two different RANS turbulence models were tested (k − ε and k− ω −SST) to evaluate how they describe the flow in the duct region and the air/fuel mixing occurring downstream. Afterwards, combustion simulations were carried out using a tabulated flamelet progress variable model based on auto-ignition calculations of diffusion flames using detailed kinetics. Experimental data of ignition delay, flame lift-off and soot mass evolution were used to validate the proposed approach.