A novel domain decomposition based numerical method is introduced in this paper. The prime goal is to develop a numerical method so as to facilitate the simulation of a complex marine offshore problem in an accurate and efficient way. To date, near-far field coupling and upper-lower field coupling are the two main approaches in the domain decomposition studies. The current work follows the basic idea of nearfar field coupling in a sense that the computational domain is decomposed into near and far regions, according to their respective distances to the offshore structure of interest. The resulting regions are then modeled in a segregated manner by solving the viscous Navier-Stokes equations (NSE) and the Potential Laplacian equation (PLE). Solving the NSE could be time consuming but in return reveals comprehensive flow details. On the contrary, solving the PLE is relatively fast but only gives limited flow information. Overall, the domain decomposition strategy finds a good balance between the simulation time and results accuracy. The novelty of the reported method lies in two-folds. First, the derivation of the so called overlapped buffer zone eliminates the need of performing time costing iterative schemes in the non-overlapping domain decomposition methods so as to ensure the matching of free surface elevation at the domain boundaries. The concept of a buffer zone is borrowed from the relaxation zone technique which is commonly used near the inlet to ensure a stable wave generation or near the outlet to absorb the reflected waves in numerical wave simulations. Second, an in-house developed OVERSET method is adopted for the viscous domain solver to handle large object displacement in the case of an extreme event.
The proposed method has been implemented in the OpenFOAM platform. To validate the method, the published data of wave-in-deck loading for a single box will be employed. Next, the method is applied to simulate a plunging breaker on a moderately-slopped beach and the result is compared with the experimental measurement. Various numerical examples presented in the paper demonstrate the accuracy and efficiency of the proposed method.
