This research delves into the intricate relationship between the 6-degree-of-freedom (DoF) motion of LNG-related vessels and the sloshing impact loads. Conducted by HD Hyundai Heavy Industries, over a decade of sloshing model tests at Seoul National University’s facility has led to a comprehensive study. The tests involve more than 30 ship models under 5,000 varied maritime conditions. A sophisticated array of over 100 Piezo electric sensors collect data on sloshing impact pressures, alongside the vessels’ 6-DoF irregular motions, across five-hour test periods at real-ship scale. Post-processing utilizes the three-parameter Weibull distribution to extract the three-hour most probable maximum peak pressures as the sloshing loads from the time series data. Subsequently, the study examines the correlation between 18 features of the ship’s 6-DoF motion — encompassing displacement, velocity, and acceleration — and the measured sloshing impact peak pressures. Machine learning techniques are pivotal in determining the significance of each feature, employing both linear and nonlinear analyses including Generalized Additive Models (GAMs). By using elimination methods based on the Light Gradient-Boosting Machine (LightGBM), the research identifies and elucidates the key parameters among the eighteen features that significantly influence sloshing loads. This comprehensive investigation not only deepens understanding of the intricate interplay between ship movements and sloshing impacts but also provides a useful dataset that could potentially inform future navigational strategies and contribute to enhancing safety and efficiency in maritime operations and hull form design.
No takes yet. Share an insight, caveat, or question.
Ahn et al. (2024) studied this question.