Explainable deep learning reveals the physical mechanisms behind the turbulent kinetic energy equation

Francisco Alcántara-Ávila, Andrés Cremades, Sergio Hoyas, Ricardo Vinuesa · 2026

Read Explainable deep learning reveals the physical mechanisms behind the turbulent kinetic energy equation by Francisco Alcántara-Ávila, Andrés Cremades, Sergio Hoyas, Ricardo Vinuesa (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-20052.txt.

Description

Francisco Alcántara-Ávila, Andrés Cremades, Sergio Hoyas, and Ricardo Vinuesa use explainable deep learning to open up turbulence physics. The turbulent kinetic energy equation tracks how the energy of swirling motion is produced, moved, and destroyed. Which physical mechanisms govern each term is partly a matter of modelling taste. The authors let a learning model point at the structures that matter.

Their explainable model, built on Shapley-style attributions, identifies the structures most important for each budget term. The test case is turbulent channel flow at a modest friction Reynolds number. Structures the method flags can be checked against physical intuition. Explainable AI here acts as a microscope for theory, not just a predictor. The findings may guide leaner turbulence closures for engineering. The flow is a plain pipe flow. It is a test bed, not a wild storm.

Catalog record

CREATORFrancisco Alcántara-Ávila, Andrés Cremades, Sergio Hoyas, Ricardo Vinuesa
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-20052

Browse the full library · About the Tower of Babel