Development of Atomization Geometry Databases from Simulations

Atomization is a complex process with numerous industrial and environmental applications. To enhance our understanding, accurate prediction of the underlying physics is crucial. Recent advances in computational resources and numerical methods have made simulations a viable alternative to experimental techniques. However, simulations have yet to yield significant insights into atomization due to high costs and massive databases. The present study is focused on developing a tool that extracts the complex geometries of key events from the simulation in real-time, eliminating the need for post-processing the entire set of raw simulation data. Specifically, the tool extracts and saves the liquid structure and flow field geometries associated with breakup events into a dedicated database. This focused data analysis will provide valuable insights. Additionally, our work establishes a foundation for a reduced-order atomization model. By assigning topological data markers to extracted geometries, efficient database parsing can be achieved. These markers enable predictions of droplet behavior in simulations based on shape and flow field, reducing computational cost in large-scale simulations.