Yu Bi
Ansys Employee

Hi, as the model is a fully submerged body, all the elements of the hydrostatic stiffness matrix should theoretically zero; the non-zero values in the Hydrostatic Results report are mainly due to the coarse meshes and the numerical truncated error.  Please check if using refiner mesh could help. Besides, as it is a fully submerged body, the center fo floatation listed in the Cut Water Plane Properties data group has no meaning, no worry about these values.

The large RAOs in the low-frequency region could be understandable.  At the low frequency, the inertial force/moment due to structure mass and added mass are very small, and there is no hydrostatic stiffness, any small wave exciting force/moment could generate large linear RAOs of the rigid body motions.

It is suggested that the nonlinear time-domain (Hydrodynamic Response in Irregular/Regular waves) hydrodynamic response analysis should be carried out, in which the instantaneous wetted surface will be used at each time step to calculate the nonlinear Froude-Krylov force and hydrostatic force, the responses will be more accurate for a body close to the mean water surface. To do so, you may need to use either the soft mooring lines or the course control system (for example a PID approach) to keep the structure in the required heading. 

Besides, for the submerged surface that is very close to the mean water surface, it'd be better to have the element size smaller than the depth of the surface. Otherwise, there could be large numerical errors.