Virtual Assisted Atomisation – VAA

The quick desintegration of compact lumps of liquid (liquid film or jet atomisation, drop breakup, spray formation on top of waves) is a process commonly encountered in numerous applications. These very complex flows that combine interface dynamics, instabilities, turbulence... challenge the modellers. The main issues related with such phenomena are statistical in nature : what are the initial size-velocity distributions of the drops ? How these characteristics do evolve in space further away from their birth location ?

Our project adresses configurations involving a quick gas stream co-flowing on top of a liquid at a lower velocity. This so-called assisted atomisation is notably encountered in aeronautic and space propulsion systems. Although some partial answers have been identified from experiments, the involved physical mechanisms are far from being completely understood. Further, and although international as well as national research teams (including the participants to the ANR DYNAA 2005-2008) have made important progress in the simulations, there is presently no model able to predict the size distribution of the atomised droplets from the knowledge of the injector geometry and of the flow conditions. That limitation especially holds for the flow conditions prevailing in turboreactors, that correspond to a high density ratio (about 1000) and a strong velocity difference between the two phases (10 to 100m/s).

Very recent (2009) progress have been made on direct numerical simulations that path the way to a modeling breakthrough, even for high density ratios. In addition, new instability scenarii have been proposed that more adequately represent the actual phenomena. The French teams involved in the present project are leaders in all these fields. Their objectives are twofold:
- to understand and master all the physical mechanisms driving the drop formation in assisted atomisation,
- to built the first reliable and validated direct numerical simulation prototypes for the assisted atomisation.

The choosen methodology is grounded on systematic comparisons between experiments and simulations at every stage of the project. Three of the involved teams bring in their specific and complementary skills and numerical tools, the fourth one being a specialist of refined experimental investigations. All partners will collaborate to analyse the results gathered. The flow conditions will be close to those arising in turboreactors, yet with slightly simplified geometries compared with industrial injectors.

The fundamental work program proposed here should contribute to help the aeronautical and space industry along two directions. On one hand, the good understanding of primary atomisation processes will be exploited to define efficient strategies for atomisation control. On another hand, the reliable and validated direct numerical simulations wil be integrated in multi-scales multi-physics numerical simulator of entire combustion chambers. Others sectors of industry should also benefit from the outputs of this project.