LES and RANS Modeling of a Hydrogen Low-Pressure Direct-Injection Injector

Hydrogen direct injection is a promising strategy for enabling high-efficiency, low-emission powertrains. However, challenges related to mixture stratification and jet modeling persist, particularly under engine representative conditions. This study numerically investigates a simplified injector model, focusing on the downstream hydrogen jet behavior from of a hydrogen low-pressure direct-injection jet-forming cap under both constant-volume chamber (CVC) and engine conditions. The primary objective is to evaluate numerical methodologies and explore model simplification strategies that remain computationally feasible while preserving physical fidelity—particularly relevant for early-stage hydrogen injector development. Experimental data serve as validation benchmarks across operating regimes. In the CVC platform, large eddy simulations (LES) provide turbulence-resolving insights that inform the refinement of Reynolds-averaged Navier–Stokes (RANS) models. RANS simulations are then extended to engine representative conditions to examine dominant mixing mechanisms and assess the cap geometry's influence on mixture formation. The results highlight that adjusting the RANS turbulence model constant C_ϵ1 enhances radial momentum transport and reduces jet tip penetration, aligning with experiments. Notably, simulations incorporating a hypothetical poppet valve inside the injector cap show that internal flow disturbances can physically induce similar jet spreading, reinforcing the rationale behind the turbulence model adjustment. These findings support the development of simplified, yet predictive, modeling practices for hydrogen direct injection systems.