July 27th, 2017

Droplets, bubbles and particles

        Many industrial applications, as well as many common situations in our daily lives, are characterised by the presence of gas and liquid interacting with each other. This is the case for example, with  rain droplets, ocean waves, fuel being injected in an engine, and with aerosols inhaled by patients   in hospital. The shape and the volume occupied by the two phases can be described through the evolution of the interface. The movement of the interface is quite a fascinating  problem, and involves several aspects of fluid motion, from inertia and compressibility forces, that can be described at the “macroscopic” level, with the molecular interactions at the interface resulting in surface tension forces and curvature. If the interface develops on a solid surface, the forces at the contact line between the solid, gas and liquid, become a critical aspect  for the overall evolution of the gas and liquid phases, determining the wettability of the surface. This is a relevant topic for atomising systems, and in particular the prefilming airblast injectors typically used in aero-engines applications. In these injectors, the quality of the fuel preparation largely depends on the capability of placing the fuel in the right position. All  aspects related to the surface tension and contact angle are fundamental to determine the evolution of the liquid film over the prefilming surface. The breakup of the liquid into tiny droplets is eventually achieved through the interaction with gas streams. Both the velocity of the air flows and the surface tension of the liquid are relevant parameters for the final size distribution of the cloud of droplets that is generated.

          The atomisation of a liquid is not peculiar to aero-engines only. The design of a proper atomising system was found necessary in many applications to control the evaporation time and the penetration of the liquid. A typical example is the nozzle used in painting systems. In the above examples we have  considered liquid droplets immersed in a gaseous medium.  Similarly, we can have gas bubbles in a liquid, for example due to cavitation or air blobs in the blood, where again the behaviour of the two phases involves the evolution of the interface.
       The design and understanding of systems operating with two-phase flows requires tools which are able to predict the behaviour of the interface and the interaction between the different phases. In this context, numerical simulation can provide a reliable framework to analyse the behaviour of the system. Let us consider the specific case of liquid atomisation. Several numerical methods have been developed so far for the prediction of two-phase flows, ranging from approaches able to resolve the interface between the gas and the liquid, to simpler methods where the liquid is assumed to be in the form of small droplets for which a material point approximation is used. The large improvements in computational capability of the last decade are making methods based on the solution of the interface more affordable for real scale problems. However, a comprehensive model able to include high speed flows, evaporation and compressibility effects in a realistic domain is not available yet.  Applied research is moving towards the use of hybrid methods where only the so called “dense” region is solved with methods able to track the interface whereas the evolution of the droplets, detached from the bulk of the liquid, is predicted with a particle tracking method.
          As always, the most suitable tool should be chosen on the basis of the relevant physics to be captured, as well as the target of the numerical simulations. For example, if the atomisation and secondary breakup of a liquid sheet has to be computed, methods able to capture all the details of the interface and the forces acting on it, including the surface tension, should be used. On the contrary, if the focus is more on the big picture and system behaviour, approaches based on tracking the statistically representative population of droplets might be a better choice. Following the same approach, these tools can also be extended to the simulation of particle motion in both liquids and gases. The behaviour of solid particles in a cyclone as well as the tracking of pieces of grass in a lawnmower can be predicted by numerical simulations.
      For each specific problem involving droplets, bubbles or solid particles, numerical simulations can offer a reliable framework for the prediction of system behaviour, and  there are many applications where computational fluid dynamics has probably never been used. This is an opportunity for both companies and researchers, since, as it often happens in engineering, the research is driven by the problems and requirements of practical systems. I would like to invite all of you to spend just few minutes to think about all the problems involving droplets, bubbles and particles, or a liquid-gas interface, you may have found in your daily work activity. I am sure we will find a lot of interesting and unexplored configurations for which somebody in XCores network of numerical simulators  may already have a solution.