For the calculation of fluid pressure, linear and nonlinear panel models and a set of slamming sections are prepared. In the linear panel model for 3-D Rankine panel method, 500

panels and 3000 panels are distributed on the mean body and free surface. 4000 panels are distributed on the whole body surface in the nonlinear panel model for weakly nonlinear approach. The ship is sliced into 40 slamming sections in the longitudinal direction for evaluation of slamming loads. The sections are perpendicular to the free surface of the calm water. The wet mode natural frequencies of the numerical models are obtained through a hammering test, which is shown in Table 10. The hammering test is conducted by applying an impulsive load in the coupled-analysis without waves, which is same with that in the model test. A difference of about 10% between the numerical models and the experimental Roxadustat cost model is observed in the natural frequencies of 2-node torsion and 2-node vertical bending. It is not clear what causes the difference because the experimental data is insufficient. In order to simulate springing in the same frequency conditions as the model test, the rigidity of the backbone is adjusted according to the experimental result. The wires in the experiment are modeled by the soft spring in order to

prevent mTOR inhibitor drift motions. The natural periods of surge, sway, and yaw motions are about 30 s in the numerical models. The Pregnenolone soft spring is stiffer than the wires but its effect is negligible because its frequency is much lower than the encounter frequency. It is also confirmed by tests with various stiffness of soft springs in the computation. It should be noted that all the numerical inputs and parameters are determined by the convergence and hammering tests. Structural damping ratios of the numerical models are deduced from the total damping ratios of the experimental model since they cannot be measured directly. In the modal superposition method, damping ratios of all modes are separately handled using modal damping ratio. However, in direct integration, it is hard to model structural damping

based on nodal velocity. Rayleigh damping is often used for modeling of structural damping. It can handle damping ratios of two natural modes directly but inevitably induces unwanted damping on rigid body motions because the damping matrix includes a nodal mass matrix which is not formulated in generalized coordinate system. By excluding the portion of the mass matrix from the Rayleigh damping matrix, the unwanted damping can be removed, but the number of controllable modes decreases to one. Linear simulations in oblique seas are performed on the three models, and the results are compared with the experiment. Linear motions including both rigid and flexible motions at the center of mass are compared with the experimental data in Fig. 26.