Hydrogen direct injection internal combustion engines (H2 ICE) present a practical and cost-effective alternative for medium-term energy transition applications, leveraging principles similar to spark ignition engines (SI). However, optimizing H2 ICE requires addressing hydrogen’s unique properties, including low molecular weight, high diffusivity, and distinct combustion characteristics. Existing research on hydrogen combustion often fails to replicate in-engine conditions, particularly regarding thermodynamic states and turbulence levels. This study investigates the entire in-engine process, including hydrogen jet injection, mixing, combustion, and heat transfer, under conditions representative of H2 ICE. Experiments were conducted using an optically accessible high-pressure, high-temperature vessel, allowing independent variation of parameters such as chamber gas and injector temperature, chamber pressure, and hydrogen injection pressure. A prototype injector for heavy-duty applications (PHINIA DI CHG15) was used to inject hydrogen into a quiescent environment during 7.5 ms, achieving a global equivalence ratio of 0.39. A slide-shaped deflector preserved the jet’s kinetic energy, generating a transient tumble motion that enhanced turbulence and mixing. Combustion behavior was analyzed using schlieren imaging, negative laser-induced fluorescence, and in-chamber high-speed pressure measurements. The results reveal a strong correlation between turbulence levels, flame front speed, and heat flux. Turbulence generated during injection influenced combustion up to 150 ms post-injection, enhancing mixing and flame propagation. Delaying ignition timing from 40 to 150 ms reduced the heat release rate fivefold and wall heat flux from 3.5 to 1 MW/m 2 , with sensitivity diminishing beyond 150 ms. Hydrogen exhibited significantly higher flame front speeds, even under quiescent conditions, compared to conventional fuels like methane (CH4), due to thermo-diffusive instabilities. These findings highlight hydrogen’s distinct combustion dynamics and providing a quantitative database for numerical models validation.
Bardi et al. (Fri,) studied this question.