Large Eddy Simulation of atomisation process using the Eulerian Stochastic Fields method

Abstract

The aim of this work is the application of the Stochastic Fields method to investigate the atomisation process.

The thesis illustrates the details of the method and reports its implementation into OpenFOAM libraries in the finite volume framework. The phase tracking and surface density implementations have been validated against two different benchmark cases: a bubble rising and a turbulent liquid jet.

Then, it is showed the application of the method in complex configurations. First, the method has been used to investigate the ECN Spray A test case as representative of diesel injection at high Reynolds and Weber numbers. For this configuration, the solver has been tested also with the classical ELSA formulation. The results have been compared to experimental data. From the comparison has emerged the benefit of the stochastic fields implementation, particularly in the droplet size prediction. The improvement of the sub-grid description provides a better agreement with the experimental data, with the presence of large droplets filtered out as the number of fields employed increases.

The method has been then used to investigate the multi-hole gasoline direct injector ECN Spray G. The investigation focuses on the primary break-up, for both the whole geometry and each single hole. The agreement with numerical and experimental data proves the capability of the solver for complex geometries.

Finally, the implemented solver has been employed to investigate a different break-up mechanism, consisting of a liquid jet into a gaseous crossflow. After a first comparison with DNS results, which shows the reliability of the predicted spray, the influence of the density ratio on such configuration has been observed. The density ratio has been modified using the same Reynolds and Weber numbers. This influences the break-up mechanisms with the presence or not of lateral deformation. The droplet distributions is affected as well, with shape and sizes influenced by the density ratio.

To summarise, the present investigations are in good agreement with experimental and numerical data from previous studies. The adoption of the Stochastic Fields method shows its advantage in the sub-grid description, providing foundations for its future development in the two-phase flows modelling.