Machine learning models can simulate complex fluid turbulence using only boundary data
In short: Researchers using physics-informed neural networks found they could simulate incompressible turbulent flows without relying on traditional turbulence models. By using only data from the boundaries of a domain, the machine learning approach successfully predicted flow patterns and Reynolds-stress components across various complex scenarios, including flows over an airfoil.
This approach is significant because it bypasses the need for specific assumptions about turbulence, which often complicate fluid dynamics simulations. The researchers tested the method on laminar flows and four distinct turbulent cases, such as flows over a NACA4412 airfoil. However, the study was conducted through computational simulations rather than physical experiments.
Why it matters: Machine learning can potentially simplify complex fluid simulations by removing the need for predefined turbulence models.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Source: Research: Eivazi et al. (2022). Physics-informed neural networks for solving Reynolds-averaged Navier–Stokes equations. Physics of Fluids. · cc-by
To cite this card, link to https://intellectplus.co.uk/app/science-research-machine-learning-models-can-simulate-complex-fluid-turbulenc.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Physics · More in Science · Latest · A–Z · All topics