The Effect of Sampling Rate on the Statistics of Impact Pressure

Sloshing model tests have been performed to estimate the sloshing loads for design of LNG containment systems. The experiments have revealed that the sloshing phenomenon is highly stochastic and impact pressure varies significantly even for a simple harmonic excitation in one direction. It is important to select an appropriate sampling rate and duration to capture the true pressure peaks in order to obtain a reasonable statistical estimation. In this pursuit, experiments have been performed on a 2D model scale tank with sway motion for duration of 30 minutes at different sampling rates of 20, 40, 60, 80 and 100 kHz. Comparison of statistical quantities like maximum pressure, rise time, decay time and impulse for various sampling frequencies are presented. Exceedence probability is also evaluated for each case and reported. The high sampling rate runs are down sampled to see the effect on the magnitude of the pressure peaks. Also the 30 minutes runs are split into a set of three 10 minute runs to see how the statistics change for each segment. The paper makes recommendations on required sampling rate and test duration for model scale to capture the various local effects such as breaking waves and spray, pronounced during the liquid sloshing impact.