Database of model-scale sloshing experiment for LNG tank and application of artificial neural network for sloshing load prediction

Abstract

Seoul National University has conducted a considerable number of six degree-of-freedom irregular small-scale sloshing model tests 1/70–1/25 scales, particularly focusing on the tanks of liquefied natural gas (LNG) carriers. An experimental database has been created to provide information of sloshing load severity, which are obtained from a lot of the post-processed experimental results. In this paper, the summary of the database is described. The artificial neural network is trained based on the database to predict sloshing load severity. Various attributes that affect experimental results are considered. Management of these attributes and the machine learning architecture are illustrated. The prediction results are validated for several experiments that are not included in the training process. Further possibilities of using the database for model test planning and cargo hold design are discussed.