FetchCFD

Here’s a cool fluid simulation made with NumericalAI (https://numericalai.net/)! It shows a two-dimensional Kelvin–Helmholtz instability, a process that happens when two layers of fluid moving at different speeds start to mix. The smooth interface between them ripples, twists, and turns into beautiful wave-like patterns and swirling vortices.

This kind of motion shows up all over nature from the waves you see in clouds and the ocean to patterns in planetary atmospheres and even space plasmas. It’s a fascinating reminder that the same physics can shape both our skies and the stars.

This simulation depicts a three-dimensional Taylor–Green vortex (TGV) computed using NumericalAI (https://numericalai.net/) in a triply periodic cubic domain. The flow is initialized with the classical Taylor–Green configuration and evolved at a Reynolds number of Re = 1600 with viscosity enabled, a well-established benchmark for studying the transition from coherent vortical motion to broadband, decaying turbulence. The visualization shows the normalized velocity magnitude, highlighting relative spatial variations and revealing the multi-scale organization of energetic structures, even in regions of lower absolute velocity. Together, these results illustrate the solver’s capability to accurately capture vortex stretching, breakdown, and the emergence of complex turbulent structures characteristic of the TGV benchmark.

Simulation of low-Mach Gresho vortex computed with NumericalAI (https://numericalai.net/). The image shows pressure contours, highlighting the characteristic low-pressure center and surrounding high-pressure ring of the Gresho vortex.

2D shock-bubble cfd simulation with NumericalAI (https://numericalai.net/). The image shows contours of 𝛼2 (volume fraction of fluid 2), clearly delineating the bubble interface as it begins to distort under shock loading.