Uncovering Flame Physics with Machine Learning: Application to the Reaction Rate in Hydrogen Flames

License: See LICENSE file · Updated 2026-09-26
My rating

The work targets combustion physics in lean premixed hydrogen flames. Convolutional neural networks are applied as analytical tools, rather than as surrogate models, to spatially resolved fields generated by direct numerical simulations. The analysis focuses on the relationship between the progress variable field and the local chemical source term, with the goal of investigating reaction-rate physics.

Tags: