Learning Interface Breakup: A Geometry-Conditioned Latent Surrogate for Spray Formation
2026-06-15 • Artificial Intelligence
Artificial IntelligenceMachine Learning
AI summaryⓘ
The authors developed a fast method to predict how liquid sprays break up from nozzles, which is usually very slow to simulate in detail. Instead of tracking every detail of the complex fluid flow, they focus on the changing pattern of the simulation's mesh (the grid used to calculate the flow), which highlights important areas. By training a model on many simulations, they can quickly approximate how the spray evolves and capture the key liquid-gas interface changes. Their approach is much faster than traditional simulations while still being accurate for new nozzle shapes.
spray nozzletwo-phase flowvolume-of-fluid (VOF)adaptive mesh refinement (AMR)surrogate modellatent representationfluid dynamics simulationtransient flowmesh adaptationcomputational fluid dynamics (CFD)
Authors
Julius H Ramlau, Friedrich Hastedt, Tolga Birdal, Ehecatl-Antonio del Río Chanona, Nausheen S Basha, Omar K Matar
Abstract
Designing spray nozzles requires predicting how geometry shapes transient two-phase breakup, but high-fidelity volume-of-fluid (VOF) simulations with adaptive mesh refinement (AMR) are too expensive for iterative design exploration. Standard surrogate models are also challenged by this setting because both the liquid--gas interface and the underlying adaptive discretization evolve across time and geometries. We introduce a geometry-conditioned latent surrogate trained on 797 two-phase nozzle simulations that addresses this by encoding the AMR cell-density field, rather than the full multi-channel flow state, as a compact proxy for where the solver concentrates resolution. From this representation, the model reconstructs transient density evolution and nozzle geometry, and a lightweight second stage recovers the remaining flow variables. On held-out simulations, the method accurately captures key interface dynamics while reducing inference time to 0.045 seconds per trajectory, corresponding to a speed-up of more than $6\times10^4$ relative to Basilisk CFD. These results suggest that AMR refinement structure can serve as a compact and learnable representation for geometry-conditioned surrogate modeling of transient two-phase flows.