An Extensive Parametric Study on Diesel Spray Simulation and Verification with Experimental Data

Diesel spray simulation with a Lagrangian approach for the dispersed phase and a Eulerian approach for the continuum phase is known to be sensitive to mesh resolution and its structure. Inaddition a dependency to turbulent length scale at nozzle exit has been reported in the computational literature. The aim of this work is to quantify these sensitivities and verify computational results for an extensive series of parameters on the basis of detailed shadowgraphy and PDA experiments in a constant volume bomb. The analysis consists of temporal and spatial resolution studies, initial turbulent length scale variation, investigation of the sensitivity to two different atomization models and the influence of the injection direction to the mesh orientation. This study has been done both for non-evaporating and evaporating diesel sprays. The calculations showed a high mesh sensitivity on spray penetration. To achieve an asymptotical behavior of the spray penetration, an orifice resolving mesh is needed. A high sensitivity on initial turbulent length scale and injection profile has also been detected. Usually the turbulent length scale in constant volume bomb is initialized to high and has a unphysical meaning.

The sensitivity study was also extended to engine calculations. The calculated turbulent length scale at SOI are using the proposed correlation physical reasonable. The influence of the injection profile on spray penetration was also detected. But the compared and relevant quantities are different for constant volume calculations then for engine calculations. For constant volume calculations the penetration and spray angle is important but the important quantities for engine calculation are the heat release and the pressure curve, respectively. For these quantities the influence of the investigated parameters is small.