MPFBench: A Large Scale Dataset for SciML of Multi-Phase-Flows: Droplet and Bubble Dynamics

TL;DR

This paper introduces MPFBench, a large-scale dataset for scientific machine learning of multiphase fluid dynamics, enabling more accurate and efficient modeling of complex droplet and bubble behaviors.

Contribution

It provides a comprehensive dataset from 11,000 simulations for training neural operators and foundation models in multiphase flow modeling.

Findings

Machine learning models effectively capture transient multiphase dynamics.

Neural operators show promise for accurate, efficient simulations.

The dataset facilitates future SciML research in multiphase fluid dynamics.

Abstract

Multiphase fluid dynamics, such as falling droplets and rising bubbles, are critical to many industrial applications. However, simulating these phenomena efficiently is challenging due to the complexity of instabilities, wave patterns, and bubble breakup. This paper investigates the potential of scientific machine learning (SciML) to model these dynamics using neural operators and foundation models. We apply sequence-to-sequence techniques on a comprehensive dataset generated from 11,000 simulations, comprising 1 million time snapshots, produced with a well-validated Lattice Boltzmann method (LBM) framework. The results demonstrate the ability of machine learning models to capture transient dynamics and intricate fluid interactions, paving the way for more accurate and computationally efficient SciML-based solvers for multiphase applications.