Simulation of deformable drops in turbulence

Simulation of deformable drops in turbulence


Davide Procacci

  Employee profile at NTNU

  Research Group: Environmental Engineering and Reactor Technology
Department of Chemical Engineering, Faculty of Natural Sciences

Photo of a man

Multiphase flows have a central role in environmental problems such as climate change and the different technological solutions that we can use to counteract it.

Recent years have revealed that the mass exchange at the interface between the ocean and the atmosphere contributes to the absorption of CO2 and has a major impact on the weather. The mass exchange between the ocean and the atmosphere is due to the breakage of the upper part of the waves which spread tiny droplets in the atmosphere while at the same time bubbles of air are entrained into the ocean. This phenomenon is an example of multiphase flow with two different phases, the water, and the atmosphere.

This example shows that we deal with multiphase flows every day both in normal life and industrial applications. In fact, dispersed bubbles-in-liquid flow systems are widespread in the chemical and biochemical process industry, such as wastewater treatment, fermentation reactors, and absorption columns. Moreover, fluid transport through pipes and channels accounts for 25% of the energy used in industry. Multiphase flow involves complex phenomena that are far from being fully understood. Further developments of both experimental and computational works are needed in this field to progress our understanding of these phenomena.

One of the critical developments that are needed is new formulations of closure models for turbulent multiphase computational dynamics. It is not feasible with the current experimental technology to obtain the required data for this development. The approach to the grand challenge of developing new closure models is to let Direct Numerical Simulation (DNS) cast considerable light on this gap in data. Although several numerical methods for multiphase DNS have been developed over the last two decades, these methods are still in an early stage. Simulations of turbulent multiphase flows by DNS have a high computational cost due to their multiscale nature. The range of processes that we want to describe goes from the macroscopic scale (1 m) to the molecular scale (10-9 m) hence the use of high-performance computing (HPC) is crucial for the resolution of this kind of problem. Even using HPC it is impossible to resolve all the relevant scales, so we consider the smallest scale of the turbulence (10-4 m) as the minimum scale that must be resolved. My project will contribute to further developments of the numerical methods applied to turbulent multiphase DNS.

Main supervisor: Jannike Solsvik
Project period: 2021-2024

Illustration of turbulent flows with drops and bubbles
Illustration taken from: G. Soligo, A. Roccon & A. Soldati, Turbulent Flows WithDrops and Bubbles: What Numerical Simulations Can Tell Us - Freeman Scholar Lecture, J. Fluids Eng. (2021).


PhD Projects