Abstract: Hypergolic propellants represent a predominant development direction in propulsion systems due to their ignition device-free characteristic, with ignition delay time serving as their primary performance metric. This study employed a custom-designed drop-test apparatus to examine the ignition characteristics between propellant fuel and high-concentration hydrogen peroxide oxidizer, while optimizing the experimental procedure through digital design. To overcome the challenge of processing extensive image data in ignition delay time measurement, we developed Python-based software incorporating the YOLOv8 deep learning algorithm. This implementation of image recognition technology enables precise analysis of propellant self-ignition delay time, facilitating efficient and accurate experimental data processing. Furthermore, integrating this innovative experimental approach into the “University Chemistry” curriculum has demonstrated significant enhancement in student engagement and learning motivation.

Key words: Hypergolic propellant, Droplet test, Ignition delay time, Deep Learning, Application in Teaching