On the analysis of sloshing of water in rectangular containers: Numerical study and experimental validation

In this work the analysis of sloshing of water in rectangular open tanks has been extensively carried out. Two mathematical models are employed, respectively the Reynolds Averaged Navier Stokes Equations (RANSE) and the Shallow Water Equations (SWE). The RANSE are solved using a modified form of the well established MAC method (SIMAC) able to treat both the free surface motion and the viscous stresses over the rigid walls accurately. The Shallow Water Equations are solved by means of a simple and powerful algorithm (CE-SE) able to deal with large impacting waves over the tank walls. Successively, in order to validate the mentioned algorithms and for a better understanding of the sloshing phenomenon, experimental tests have been carried out using a 0.5 m breadth rectangular tank in periodic roll motion. It has been shown that RANSE provide more accurate solutions than SWE for small or moderate amplitudes of excitation. In particular in this paper it is proved that the shallow water approximation can be efficiently adopted within liquid depth to tank breadth ratio = 0.15, when examining the sloshing problem. By increasing the water level inside the tank, results by SWE show large qualitative and quantitative disagreement with experiments. Nevertheless, in the case of large amplitude excitation, when sprays and large breaking waves are expected, SWE provide a fairly good estimate of the sloshing induced waves. Finally a simple baffle configuration inside the tank has been considered. By the analysis of numerical results, it has been observed that the presence of a vertical baffle at the middle of the tank dramatically changes the sloshing response compared to the unbaffled configuration. It produces a jump-like effect, resulting in a weak magnification of the dynamic loads on the vertical walls out of resonance, and a strong reduction of the dynamic loads in the resonance condition. Copyright (C) 1996 Elsevier Science Ltd