Teaching the Incompressible Navier-Stokes Equations to Fast Neural Surrogate Models in 3D

TL;DR

This paper introduces an efficient neural network framework for real-time 3D fluid simulation that generalizes well across different fluid phenomena and geometries, reducing training data needs and computational costs.

Contribution

It extends a 2D neural fluid model to 3D with an architecture that handles high-dimensional data efficiently and incorporates fluid properties for versatile flow simulations.

Findings

Achieves real-time 3D fluid simulation on 128x64x64 grids.

Models both laminar and turbulent flows using additional fluid properties.

Outperforms existing neural models in accuracy, speed, and generalization.

Abstract

Physically plausible fluid simulations play an important role in modern computer graphics and engineering. However, in order to achieve real-time performance, computational speed needs to be traded-off with physical accuracy. Surrogate fluid models based on neural networks have the potential to achieve both, fast fluid simulations and high physical accuracy. However, these approaches rely on massive amounts of training data, require complex pipelines for training and inference or do not generalize to new fluid domains. In this work, we present significant extensions to a recently proposed deep learning framework, which addresses the aforementioned challenges in 2D. We go from 2D to 3D and propose an efficient architecture to cope with the high demands of 3D grids in terms of memory and computational complexity. Furthermore, we condition the neural fluid model on additional information about the fluid's viscosity and density which allows simulating laminar as well as turbulent flows based on the same surrogate model. Our method allows to train fluid models without requiring fluid simulation data beforehand. Inference is fast and simple, as the fluid model directly maps a fluid state and boundary conditions at a moment t to a subsequent fluid state at t+dt. We obtain real-time fluid simulations on a 128x64x64 grid that include various fluid phenomena such as the Magnus effect or Karman vortex streets and generalize to domain geometries not considered during training. Our method indicates strong improvements in terms of accuracy, speed and generalization capabilities over current 3D NN-based fluid models.