The Fluid Cube dataset contains 100 fluid dynamics simulations of a fluid block flowing inside a unit cube domain. For each simulation, the fluid block is set with different initial shape, position, velocity, and fluid viscosity.

These simulations were computed using the Smoothed Particle Hydrodynamics (SPH) formulation which discretizes a fluid as a group of points in space, referred to as particles.

For more on the SPH formulations, head to our recent blog post on the introduction to SPH.

To download the dataset, head to this link.

Dataset structure

In the dataset folder you will find a .csv file holding the metadata of each of the 100 simulations. This metadata includes the fluid properties of each simulation, such as the fluid viscosity, initial shape and velocity and the respective simulation ID.

The outputs of each simulation are stored inside a folder with the respective simulation ID. In that folder, you’ll find a .npy file holding the particle’s data for all time steps.

Besides the simulation data, inside each folder, there is also a video of the simulation rollout to allow for an easy comprehension of the data in each simulation.

A schema of this file structure is shown below:

├── sim_0000
│   ├── particle_data.npy
│   └── simulation_movie.mp4
├── ...
├── sim_0099
│   ├── particle_data.npy
│   └── simulation_movie.mp4
└── summary.csv

Along with the dataset, you can also find a jupyter notebook with instructions on how to load each simulation file and access the information on the particle’s positions and velocities.

Ideas on how to use this dataset

This dataset can be used for different purposes, such as:

  • Compute dynamical metrics (e.g. average particle velocity, force ratios, etc.) as a function of time or the other dataset variables (e.g. viscosity, volume, etc.).
  • Train machine learning models to learn the fluid dynamics of the presented scenarios. For inspiration on how to do this, take a look at one of our recent blog posts.

🌊 Have fun! 🌊

If you have any doubts regarding the dataset, don’t hesitate to reach out at datasets@inductiva.ai!

Also, in case you’re using this data on your projects, we would love to know how!

Besides this, don’t forget to check out the other cool datasets!

Recent posts from our blog

Sofia Guerreiro

Sofia Guerreiro

Cristiana Carpinteiro

Cristiana Carpinteiro

In this series of blog posts we will explore a specific case of the use of AI in the pharmaceutical industry - using Graph Neural Networks for predicting binding affinity. But for now, let’s start by understanding the problem of drug discovery and some fundamental concepts like binding affinity.

Ivan Pombo

Ivan Pombo

Luís Sarmento

Luís Sarmento

Coastal engineering projects protect the coast from erosion, flooding and other events. Due to their cost, the design phase of these projects is heavily based on computational simulations. In this blog post, we enlighten how Inductiva API allowed researchers to scale their coastal engineering simulations.