Deterministic validation of a time domain panel code for parametric roll

Abstract

Validation of simulation methods for dynamic stability is hampered by the fact that dynamic stability phenomena can be quite rare. In order to obtain sufficient statistical confidence in both experimental data and simulation results long duration time histories are required for a range of operational conditions. This is at most times not feasible from a practical point of view. One way of validating time domain simulation methods for dynamic stability phenomena is by deterministic validation. This means that the simulation is run in the same wave sequence as used during the model experiments. Ideally, a one to one comparison between experiments and simulations is then possible. A difficulty in such an approach is, in case of irregular waves, the reconstruction of the experimental wave train in the simulation tool. Even if this were successful, the encountered wave train in the simulations will deviate from the experimental one because it is inevitable that the position in the horizontal plane will differ from the experimental one after some time. The paper describes the deterministic validation of a non-linear, 6-DoF time domain panel code for parametric roll. The paper explains the method for reconstructing the experimental wave train in the simulation method and how to circumvent the problem of the deviation in horizontal position. Finally, comparisons between experimental and simulated time traces are given for the motions in the vertical plane.