Point-Cloud Deep Learning for Prediction of Fluid Flow Fields on Irregular Geometries (Supervised Learning)
Authors: Ali Kashefi (kashefi@stanford.edu) and Davis Rempe (drempe@stanford.edu)
Description: Implementation of PointNet for supervised learning of computational mechanics on domains with irregular geometries
Version: 1.0
Guidance: We recommend opening and running the code on Google Colab as a first try.
Citation
If you use the code, please cite the following journal paper:
A point-cloud deep learning framework for prediction of fluid flow fields on irregular geometries
@article{kashefi2021PointNetCFD,
author = {Kashefi, Ali and Rempe, Davis and Guibas, Leonidas J.},
title = {A point-cloud deep learning framework for prediction of fluid flow fields on irregular geometries},
journal = {Physics of Fluids},
volume = {33},
number = {2},
pages = {027104},
year = {2021},
doi = {10.1063/5.0033376},}
1. Introduction
We provide the implementation of PointNet for the prediction of quantities of interest in the area of computational mechanics on domains with irregular geometries. Specifically, we present the example of flow past a cylinder with various shapes for its cross sections. We hope that this simple example motivates other researchers to use PointNet for different areas of computational mechanics and physics such as compressible flows, solid mechanics, etc.
To make the code usable for everyone (even with a moderate knowledge of deep learning), we implement the code using Keras. We explain the procedure step by step.
We strongly recommend users read the journal paper "A point-cloud deep learning framework for prediction of fluid flow fields on irregular geometries" (https://aip.scitation.org/doi/full/10.1063/5.0033376).
You might also find free versions of this article on arXiv or ResearchGates.
2. Google Colab
We strongly recommend running the code on Google Colab as a first try. Here is a link to Colab https://research.google.com/colaboratory. In this way, you will not need to install different libraries. Moreover, you do not need to be worried about matching the required libraries.
3. Flow past a cylinder
We consider laminar steady-state flow past a cylinder with different shapes for its cross sections. In fact, we consider the same example discussed in the journal paper. Please see Figure 2 and Figure 3 of the journal paper for the generated meshes and flow fields. For geometries of the cross-section of the cylinder, we use those geometries described in Table 1 of the journal paper. However, due to reducing the data size and making it possible to run the code in a reasonable amount of time on Google Colab, we only consider "circle," "equilateral hexagon," "equilateral pentagon," "square," and "equilateral triangle." For the fluid and flow properties such as density, viscosity, and the magnitude of free stream velocity, please see the section of Governing equations of fluid dynamics of the journal paper for further details.
Please read the rest of this documentation by opening the "PointNetCFD.ipynb" file on Google Colab.
Questions?
If you have any questions or need assistance, please do not hesitate to contact Ali Kashefi (kashefi@stanford.edu) or Davis Rempe (drempe@stanford.edu) via email.
About the Author
Please see the author's website: Ali Kashefi