The formation and penetration characteristics of underwater shaped charge jets at varying depths: experiment, simulation, and machine learning

Chao Cao, Jingwen Xia, Kui Tang, Jinxiang Wang, Ming Yang, Yangchen Gu, Minhui Gu, Hanxin Gong, Yuanbo Li,
The formation and penetration characteristics of underwater shaped charge jets at varying depths: experiment, simulation, and machine learning,
Ocean Engineering,
Volume 339, Part 2,
2025,
122164,
ISSN 0029-8018,
https://doi.org/10.1016/j.oceaneng.2025.122164.
(https://www.sciencedirect.com/science/article/pii/S0029801825018487)
Abstract: Underwater shaped charges are widely utilized in marine engineering. However, as humans explore deeper into the ocean, studying how shaped charge jets work at different water depths remains a challenge. A combination of experiments, numerical simulations, and machine learning was applied in this study. After confirming the accuracy of the simulation, a machine learning method was used to understand the relationship between shaped charge parameters and jet performance. Then, the influence of deep water was considered, and the formation characteristics of jets at different water depths were obtained, as well as the evolution mechanisms. The results indicate that accurate jet characteristics can be predicted using machine learning, with a maximum error of 3.2 %. Continuous velocity curves were obtained, revealing the influence of underwater shaped charge structural parameters on jet performance. Furthermore, the variation patterns of jet characteristics in the depth of 0∼6000 m were determined using machine learning. As water depth increases, the jet performance exhibits nonlinear growth, which is attributed to the energy release characteristics of explosive and the liner strain. The liner strain is the primary factor driving this nonlinear behavior. This study provides guidance for the design and application of shaped charges in deep-water environments.
Keywords: Shaped charge; Machine learning; Formation characteristic; Deep water