As a novel, clean, and efficient energy source, hydrogen has emerged as a cornerstone of sustainable transportation, facilitating the rapid development and deployment of hydrogen-powered trains. While current research mainly explores relatively idealized open environments or simple confined volumes, the unique challenges posed by semi-enclosed, longitudinal tunnel geometries remain insufficiently investigated. Specifically, the dispersion and transport mechanisms of hydrogen following high-pressure leakage in the tunnel scenario, along with the resulting hydrogen concentration distribution, remain to be investigated. Furthermore, there is a lack of connection and analysis between the evolution of the velocity field and the variations in momentum flux following high-pressure hydrogen leakage.

To investigate the evolution law of high-pressure hydrogen leakage and dispersion within tunnels, a numerical modeling approach was employed to establish physical geometry models of the tunnel and train. By integrating the turbulence model with the Molkov virtual nozzle model, the evolution of the velocity field, the variations in the hydrogen-leakage momentum flux, and the three-dimensional concentration distribution profiles were systematically analyzed.

After hydrogen leaked, its jet impinged on the tunnel ceiling, followed by rapid lateral dispersion and downward flow along the tunnel ceiling. During this process, the hydrogen momentum vector underwent multiple reorientations at the wall, which led to a rapid decrease in its momentum. Consequently, the momentum flux distribution exhibited a distinct gradient distribution along the tunnel ceiling. The velocity decayed significantly toward the tunnel exits, where density variations across the different zones remained negligible; the lateral momentum gradually homogenized. Hydrogen dispersion exhibited radial symmetry along the longitudinal axis of the tunnel ceiling. Furthermore, a high-concentration accumulation zone was identified within the flammable-hydrogen cloud surface layer, extending 0-60.0 m downstream from the leakage source. Laterally, the cloud expanded symmetrically under the influence of turbulent mixing, eventually spanning the entire width of the tunnel. Vertically, buoyancy-driven effects confined the hydrogen cloud accumulation to the upper 3.0-5.0 m of the tunnel, while concentrations in the lower strata remained consistently below the safety threshold. The initial momentum of the hydrogen jet significantly influenced its spatial distribution and dispersion. In the near-field region proximal to the leakage source, the high-velocity hydrogen jet impinged on the tunnel ceiling and was forced downward along the tunnel walls, thereby preventing hydrogen accumulation near the ceiling close to the wall. However, when the distance from the leakage source was relatively large, the velocity vector of hydrogen was lost, as hydrogen dispersion and transport at that time were mainly affected by buoyancy. Consequently, hydrogen rose from its previously downward-spread position and accumulated gradually on the tunnel ceiling. Therefore, a hydrogen-concentration peak appeared at a relatively distant position.

The findings of this study provide basic data support for hydrogen-related parameters for analyzing hydrogen leakage in hydrogen-powered trains in tunnel scenarios, which is conducive to the application and promotion of hydrogen energy in non-traditional enclosed scenarios, such as tunnels.