ABSTRACT

A numerical code based on a particle method, which is for violent free surface simulations, is validated through comparison of numerical simulations with experimental results of violent sloshing problems in a rectangular tank. The result shows good qualitative agreement and acceptable quantitative agreement.

INTRODUCTION

In the field of naval and ocean engineering, reasonable estimation of impulsive loads from fluid is an important information for safety of ships and marine structures. The large deformation of free surface is one of the most difficult problems for numerical simulations of fluid motion. There are many kinds of numerical techniques to treat the problem. Most of them have tried to capture the free surface using by some scalar function like a density function on the grid system. On the other hand, there is a different approach to the problem. It is called particle methods that use moving particles as a proxy for grid or mesh system. The most popular particle method is the SPH (Smoothed Particle Hydrodynamics) method. A technique to apply it to incompressible flow was initiated by Monaghan (1994). In the field of marine hydrodynamics, there are some applications of SPH method (Fontaine et al., 2000, Landrini et al., 2002, Souto and Pavon, 2003). Its capability to treat violent free-surface flow has been already shown.

In this paper, the method is based on not the SPH method but the MPS (Moving Particle Semi-implicit) method that is initiated by Koshizuka in the field of nuclear engineering in order to simulate violent motions of a boundary (1996). There are the preceding researches of MPS method by Goto et.al (2001). This method is quite different from the SPH method about spatial discretization and velocity-pressure computing scheme.

You can access this article if you purchase or spend a download.