Computational optimization of a hydrogen direct-injection compression-ignition engine for jet mixing dominated nonpremixed combustion

Hydrogen (H2) nonpremixed combustion has been showcased as a potentially viable and preferable strategy for direct-injection compression-ignition (DICI) engines for its ability to deliver high heat release rates and low heat transfer losses, in addition to potentially zero CO2 emissions. However, this concept requires a different optimization strategy compared to conventional diesel engines, prioritizing a combustion mode dominated by free turbulent jet mixing. In the present work, this optimization strategy is realized and studied computationally using the CONVERGE CFD solver. It involves adopting wide piston bowl designs with shapes adapted to the H2 jets, altered injector umbrella angle, and an increased number of nozzle orifices with either smaller orifice diameter or reduced injection pressure to maintain constant injector flow rate capacity. This work shows that these modifications are effective at maximizing free-jet mixing, thus enabling more favorable heat release profiles, reducing wall heat transfer by 35%, and improving indicated efficiency by 2.2 percentage points. However, they also caused elevated incomplete combustion losses at low excess air ratios, which may be eliminated by implementing a moderate swirl, small post-injections, and further optimized jet momentum and piston design. Noise emissions with the optimized DICI H2 combustion are shown to be comparable to those from conventional diesel engines. Finally, it is demonstrated that modern engine concepts, such as the double compression-expansion engine, may achieve around 56% brake thermal efficiency with the DICI H2 combustion, which is 1.1 percentage point higher than with diesel fuel. Thus, this work contributes to the knowledge base required for future improvements in H2 engine efficiency.