Research On Characteristics Analysis Of Sloshing Impact Load Of LNG Tank Based On Experiment

In ocean oil-gas engineering field, the development of FLNG (Floating Liquefied Natural Gas) and LNG (Liquefied Natural Gas) ship attract more attention recently, which will play an important role in the LNG(-163?) marine exploitation and the global transportation. GTT Membrane tank is one of key components of FLNG for storing the LNG. However, the sloshing problem is considered as a crucial factor of LNG tank design when it is partially filled. Recently, the indoor sloshing model experiment is adopted widely by most researchers as the best reliable approach to study sloshing problem in LNG tank. So, full investigation are made about sloshing model experiment method based on Froude similarity law, scaling effect of slamming load, frequency-domain analysis and prediction of sloshing load. In the paper, considering some imperfection in current research work, indoor model experiment system is set up and three different-scales tanks are manufactured to study scaling effect. These testing results of impact loads in scaled model tank are scaled up to prototype results through Froude similarity law. Then, peaks statistics would be made by exceeding probability diagrams and fitting three parameters Weibull distributions would be performed. It shows that the sloshing loads obtained from model tank based on Froude similarity law are relatively conservative. Moreover, scaling effect deviations are quantified, and some compensation measures could be proposed.Considering strong nonlinearity of sloshing load time-history, HHT (Hilbert-Huang Transform)-EMD (Empirical Mode Decomposition) is introduced to research the frequency domain analysis, including signal scale characteristics and instantaneous frequency information based on data in the largest scale model tank. The results indicate that the pressure signal is very complicated, in which the highest order components represent the high frequency noise, higher order components carry the most energy of sloshing load and low order components represent trend of the sloshing liquid surface. Based on the above results, the pressure signal is reconstructed. Finally, the ARMA model (Auto Regressive Moving Average Model) is introduced to fit load peaks time series of sloshing slamming under irregular excitation with mathematical model, which is adopted forecast the pressure peak. The results show that the ARMA model could be able to fit sloshing peaks time series, but the predicting result is conservative.The topic of paper derives from National Science and Technology Major Project in the 12th Five-Year Plan:" Key Technology Research of Large FLNG/FLPG and FDPSO Vessel". The research achievement can provide useful reference for the sloshing problem of large-scale LNG tank.