Massively Parallel High-Fidelity Simulation of Spray Atomization

Project Description

Detailed understanding of the complicated multiphase flow in liquid atomization processes is important for improving the efficiency and stability of the spray combustion occurring in many industrial combustors including those in aerospace engines. As a result, the research topic has attracted substantial attention from the multiphase flow scientific community. The overall objective of this project is to leverage the Oakridge Leadership Computing Facility (OLCF) high performance computational (HPC) resources and expertise to enhance the efficiency of a UTRC CFD code for simulating the atomization processes at high resolutions, and to demonstrate the effectiveness of the code in capturing the flow details and revealing the underlying physics. The work is aimed at developing a state-of-art HPC methodology for improving the engineering design of injection devices, and simultaneously exploring the fundamental physics that will impact a broader scientific community.

Experimental studies of fuel atomization have been rare due to limited optical access afforded in the near-field dense spray region and the physical challenges encountered by intrusive measurements there. In the recent past, high fidelity simulation has shown promise as an alternative approach for quantitatively predicting fuel atomization. Computationally, it remains challenging to resolve the wide range of spatial and temporal flow scales and to properly account for the large variation of density across the liquid-gas interface. As a result, the studies using this approach have typically remained qualitative.

In this project we challenge this paradigm and conduct high fidelity simulations of liquid jet in cross-flow atomization and for the first time demonstrate the simulations to be quantitatively predictive via detailed validation against experimental measurements. We leverage high performance computing to manage the cost of the high resolution simulations which were performed using over 5000 processors at the Oakridge Leadership Computing Facility of the US Department of Energy under the Advanced Scientific Computing Research (ASCR) program Leadership Computing Challenge (ALCC) project. The detailed simulation results at different Weber numbers are validated with experimental measurements of surface wavelength, breakup location and column trajectory. The physics of the size, velocity and mass rate of droplets formed along the jet column are studied and also compared with experimental measurements. The effects of increasing Weber number on jet breakup and aerodynamic flow are explored and will be discussed in this report. The high resolution simulations not only quantitatively capture detailed physics, but also establish a critical baseline case for comparing with other cost-reduction techniques, which is also an important contribution to the computational science community.