PhD: Numerical study of the physics of atomisation to control its efficiency

Starting: 2017

Liquid sprays (i.e, assemblies of fine droplets) play a key role in many environmental flows (e.g., breaking waves) and engineering devices (e.g., in food processing, coating, printing or fire safety). In combustion applications, a widespread fuel injection strategy consists in assisting liquid fuel atomisation using a fast surrounding air stream able to break up the continuous liquid jet into droplets (see image below).


A major objective of combustion engineers is to optimise the efficiency of such systems in order to increase fuel efficiency and limit pollutant emissions. A promising but under-exploited route for achieving this optimisation is to modify the dynamics of liquid jet atomisation via the control of inflow conditions. But the mechanisms involved in air-assisted atomisation are very complex, mostly due to the interaction of two phases with large viscosity and density differences, leading to discontinuous pressure and velocity gradients, but also because of the complex exchange of kinetic energy at the liquid-gas interface via surface tension. Moreover, spray formation involves frequent topology changes, usually through the formation and rupture of liquid ligaments and sheets, rendering the description of the behaviour very challenging.


As a consequence, the detailed understanding of liquid atomisation physics that would be needed to enable spray control does not exist today, and the goal of this project is to deepen our understanding of the mechanisms leading to air-blast liquid jet atomisation to pave the way towards control.

Numerical simulation of coaxial air-blast liquid atomizer. Low velocity liquid is issued from the round middle pipe, while a high
velocity gaseous annular co-flow is injected around it, promoting liquid break-up. The image illustrates the type of behavior
expected from the liquid jet.

Thanks to recently advanced numerical methods specifically dedicated to the simulation of multiphase flows associated to high performance computing, simulations can now deal with large viscosity / density ratio, complex discontinuities and a wide range of scales, to give an accurate description of the atomisation process.


The scientific approach of this project is to employ high-fidelity simulations in a canonical geometry in order to rapidly improve our understanding of the dynamics of air-blast liquid destabilisation (i.e. the interplay between nozzle flow conditions, the growth of interfacial instabilities, and the evolution of downstream flow structures). In the project, two types of inflow conditions are simulated and the associated regimes of air-blast liquid jet dynamics are studied.

  • The first inflow condition consists of an analytical mean velocity gas profile to which perturbations with small amplitude are added to mimic turbulence.
  • The second type of inflow is generated from coupled simulations of straight long pipes, in order to generate more realistic, fully turbulent inflow conditions.

The simulations will be coupled with experimental diagnostics and theoretical stability analysis, to provide additional guidance and opportunities for validation of the numerical observations.


This project involves a collaboration between the LEGI and the Sibley School of Mechanical and Aerospace Engineering at Cornell University.


PI: Guillaume Balarac; Co-PI: Olivier Desjardins; PhD Student: Cyril Bozonnet



<< back