A “next-generation” spray and atomization model for high-pressure diesel sprays has been developed and implemented into an engine computational fluid dynamics code (KIVA-3V), together with a nozzle flow model and an evaporation model for the Eulerian liquid phase. The new model is based on the assumption that high-pressure spray atomization under modern diesel engine conditions can be described by considering a single “fluid” to represent the turbulent mixing of a liquid jet with ambient gases. The governing Navier-Stokes equations for the liquid-gas mixture are solved and several previously proposed techniques are used to correct for vortex stretching and compressibility effects in high-speed free jets. To describe the dispersion of the liquid phase into a gaseous medium, transport equations based on the turbulent mixing assumption are also solved for the liquid mass fraction and the liquid surface density (liquid surface area per unit volume). A switch from the Eulerian approach to a Lagrangian drop approach is allowed in order to benefit from the advantages of the traditional Lagrangian droplet tracking methodology beyond the dense spray region near the nozzle. However, a complete Eulerian approach is optional if desired. As in the existing ELSA (Eulerian-Lagrangian Spray and Atomization) model, the drop size, drop number and drop distributions are determined using the local liquid mass fraction and local liquid surface density.
A three-dimensional homogeneous equilibrium model was developed to simulate the cavitating flows within diesel injector nozzle passages. The effects of nozzle passage geometry and injection conditions on the development of cavitation zones and nozzle discharge coefficients were investigated. The predicted flow quantities at the nozzle exit were applied to the downstream spray atomization modeling as inflow boundary conditions. Vaporization in the Eulerian liquid phase was accounted for with an equilibrium evaporation model. Finally, the present new models were used to predict diesel spray atomization processes, and the numerical results compared favorably with experimental data.