Parameter-Conditioned Sequential Generative Modeling of Fluid Flows

TL;DR

This paper presents a neural network-based generative model that efficiently simulates fluid flows across multiple conditions, significantly reducing computational costs while maintaining accuracy.

Contribution

The authors introduce a parameter-conditioned generative modeling approach for fluid flows, enabling fast and accurate simulations across diverse flow conditions.

Findings

Models accurately capture local and global flow properties.

Simulations are orders of magnitude faster than traditional CFD.

Effective for both 2D and 3D fluid flows.

Abstract

The computational cost associated with simulating fluid flows can make it infeasible to run many simulations across multiple flow conditions. Building upon concepts from generative modeling, we introduce a new method for learning neural network models capable of performing efficient parameterized simulations of fluid flows. Evaluated on their ability to simulate both two-dimensional and three-dimensional fluid flows, trained models are shown to capture local and global properties of the flow fields at a wide array of flow conditions. Furthermore, flow simulations generated by the trained models are shown to be orders of magnitude faster than the corresponding computational fluid dynamics simulations.