U-net architectures for fast prediction of incompressible laminar flows

TL;DR

This paper explores the use of U-net neural network architectures to rapidly predict 2D velocity and pressure fields in incompressible laminar flows around arbitrary shapes, leveraging a dataset generated via CFD simulations.

Contribution

It introduces a dataset of random shapes with corresponding flow fields and evaluates U-net models for fast, accurate flow prediction in laminar regimes.

Findings

U-net models achieve accurate flow predictions on unseen shapes.

The approach significantly reduces computational time compared to traditional CFD methods.

U-net architectures are effective for physics-informed flow prediction tasks.

Abstract

Machine learning is a popular tool that is being applied to many domains, from computer vision to natural language processing. It is not long ago that its use was extended to physics, but its capabilities remain to be accurately contoured. In this paper, we are interested in the prediction of 2D velocity and pressure fields around arbitrary shapes in laminar flows using supervised neural networks. To this end, a dataset composed of random shapes is built using Bezier curves, each shape being labeled with its pressure and velocity fields by solving Navier-Stokes equations using a CFD solver. Then, several U-net architectures are trained on the latter dataset, and their predictive efficiency is assessed on unseen shapes, using ad hoc error functions.