Assessment of modeling methods for predicting load resulting from hydrogen-air detonation

Abstract

As hydrogen becomes an increasingly vital component in the transition toward a sustainable energy system, its flammable and detonable properties necessitate a comprehensive understanding of its explosive characteristics. This study evaluated the accuracy and computational efficiency of an innovative numerical approach that integrates CESE compressible CFD solver, chemistry reaction model, and structural FEM solver within LS-DYNA to predict hydrogen detonation loads. Comparisons were made with the commonly used energy equivalent methods, i.e., the TNT equivalent method, and the high-pressure volume method, which utilizes multi-material ALE techniques. Hydrogen detonation test results from open-air space, open-air space with a blast wall, and semi-confined space were compared against numerical simulations. The results revealed that, for scaled distance exceeding 0.79 m/kg1/3, all three methods accurately predicted the peak overpressure. The TNT equivalent method exhibited an unexpectedly high energy efficiency factor exceeding 0.51, significantly surpassing the recommended range of 0.01–0.1 for typical vapor cloud accidents. The CESE-chemistry coupling method exceled in capturing overpressure duration and structural response due to its consideration of chemical kinetics. As the scaled distance reduced to 0.37 m/kg1/3, the CESE-chemistry coupling method maintained its proficiency in modelling pressure waves, while the TNT equivalent method overestimated peak pressure by 494%. Conversely, the high-pressure volume method underestimated the peak pressures within or near the H2-air cloud. Nevertheless, the CESE-chemistry coupling method required significantly higher computational costs, with 15–20 times more computational time compared to the other two methods, and 60–70% of the total computation time was spent solving chemical kinetics. It is concluded from the current study that for scenarios involving close scaled distances (less than 0.37 m/kg1/3) or where the structure locates inside or near the gas cloud, the CESE-chemistry coupling method may be preferred despite its higher computational demands. Conversely, for simulations prioritizing computational efficiency and larger scaled distances, the TNT equivalent method or high-pressure volume method is recommended. These findings offer guidelines for researchers and engineering professionals engaged in assessing and mitigating risks associated with hydrogen explosion accidents in the pursuit of safe and sustainable hydrogen utilization.