1 NUMERICAL SIMULATION OF TSUNAMI CURRENTS AROUND MOVING STRUCTURES Eizo Nakaza 1 , Tsunakiyo Iribe 2 and Muhammad Abdur Rouf 1 The paper aims to simulate Tsunami currents around moving and fixed structures using the moving-particle semi- implicit method. An open channel with four different sets of structures is employed in the numerical model. The simulation results for the case with one structure indicate that the flow around the moving structure is faster than that around the fixed structure. The flow becomes more complex for cases with additional structures. Keywords: numerical simulation; Tsunami currents; bore; moving structures INTRODUCTION Tsunami is a well known phenomenon that often strike coastal areas and drowned various types of structures. The height and distance of swashed debris from the coast as well as the strength and run- up height of tsunamis are usually estimated by researchers. Estimating the tsunami-induced currents around structures and as a consequence, the moving patterns of the structures are very complex. The estimation of such complex currents and behaviors of drowned structures are important for designing coastal structures with the view of making countermeasures against tsunami attacks. In order to clarify such complex currents around moving three-dimensional structures as well as the interactions between the currents and the moving structures, the moving-particle semi-implicit method (MPSM) is introduced herein. MPSM The moving particle semi-implicit method (MPSM) is a modified particle method developed by Koshizuka and Oka (1996) for solving an incompressible flow. This method is suitable for simulation of complex free surface motion because the method does not use grids. Furthermore, since this method is a Lagrangian method, the calculation of advection terms is not required. The MPSM has been successfully applied to wave breaking (Koshizuka et al. 1998), shipping water on the deck of a ship (Shibata et al. 2009), and validating pressure (Khayyer and Gotoh 2008) as coastal engineering problems. The MPSM can be described as follows. Discretization Method Weight function. The MPSM uses a model of interaction among particles for discretizing a differential operator. The interaction between particle i and its neighboring particle j involves a weight function w , as follows: () ⎪ ⎩ ⎪ ⎨ ⎧ < ≤ − = r r r r r r r w e e e 0 1 (1) In Eq. (1), ij r is the distance between particles i and j , and the radius of the interaction area is represented by parameter e r . The particle number density is defined as ( ) ∑ ≠ − = i j i j i r r w n r r (2) which is the sum of the weight function of neighboring particle j. In Eq. (2), the contribution from particle i is not considered. The initial particle number density 0 n is calculated for the initial particle positions, which are usually given as a square lattice. Gradient model in the MPS method. A gradient vector at particle i that has scalar quantity i φ is calculated using the following equation: 1 Faculty of Engineering, University of the Ryukyus, 1 Senbaru, Nishihara, Okinawa, 903-0213, Japan 2 DPRCIR, University of the Ryukyus, 1 Senbaru, Nishihara, Okinawa, 903-0213, Japan