【摘要】 针对高压氢气泄漏扩散风险的高精度预测需求，基于多组经典试验数据，系统验证了商业软件ANSYS FLUENT中的五种虚喷嘴模型与五种湍流模型的耦合适用性。通过对比分析虚喷嘴模型，发现Yuceil2002模型对远场区域(≥5 m)氢气摩尔分数分布的预测与试验值的吻合度最优，而Birch1987与Schefer2007模型在出口区域(<1 m)精度较高。针对湍流模型，标准k-ε与k-ω模型在射流初始阶段(<5 m)计算效率与精度较优，而RNG k-ε模型在远场(≥5 m)预测更接近试验值。通过耦合RNG k-ε湍流模型与Yuceil2002虚喷嘴模型，验证了氢气摩尔分数稀释至安全阈值(4%)的距离与泄漏压力呈线性正相关，并提出多尺度泄漏场景的模型选择准则：短距离(<5 m)推荐Birch1987与RNG k-ω组合，长距离(≥5 m)采用Yuceil2002与RNG k-ε组合。研究成果揭示了高压氢气泄漏扩散的仿真模型适用边界，为氢能设施的安全间距设计、泄漏应急响应及风险评估的高精度仿真研究提供了支撑。更多还原

【Abstract】 To meet the high-precision requirements for predicting risks associated with high-pressure hydrogen leakage and diffusion, this study systematically validates the coupling applicability between five virtual nozzle models and five turbulence models. The validation is grounded in multiple sets of classical experimental data to ensure reliability and accuracy. Through comparative analysis, the results indicate that among the virtual nozzle models, the Yuceil2002 model demonstrates the best agreement with experimental results in the far-field regions(distances greater than 5 m), accurately capturing the hydrogen concentration distribution during the dilution process. In contrast, the Birch1987 and Schefer2007 models demonstrate superior performance in the near-field region(less than 1 m), where jet behavior is primarily influenced by initial momentum and complex mixing processes. When verifying and analyzing turbulence models, the standard k-ε and k-ω models exhibit high computational efficiency while maintaining good accuracy during the early stages of jet development(within 5 m). In the far-field region, the RNG k-ε turbulence model yields simulation results that align more closely with experimental observations, particularly in predicting hydrogen concentration gradients over longer distances. By coupling the Yuceil2002 virtual nozzle model with the RNG k-ε turbulence model, this study establishes a linear positive correlation between the distance required for hydrogen concentration to dilute to the safety threshold(4%) and the leakage pressure. Based on these insights, we propose model selection criteria for different leakage scales: for short-range scenarios(less than 5 m), the combination of the Birch1987 model and the RNG k-ω model is recommended, while for long-range scenarios(greater than 5 m), the Yuceil2002-RNG k-ε combination is deemed optimal. These findings clearly delineate the applicability boundaries of commonly used simulation models and provide a high-precision numerical tool for safety distance design, emergency response planning, and comprehensive risk assessment in hydrogen energy facility operations.更多还原