Open jjonkman opened 5 years ago
The improvements to HydroDyn related to floating platform structural flexibility and member-level loads mentioned above have now been completed and released in OpenFAST v2.6 (https://github.com/OpenFAST/openfast/releases/tag/v2.6.0). Included in that effort was an overhaul of how the water ballasting functionality was implemented. So, I would hope that this bug has been fixed, but this should be verified.
The HydroDyn hydrodynamics module of OpenFAST permits strip-theory members to be fluid-filled, meaning that they are flooded and/or ballasted e.g. with water ballast. The fluid-filled members introduce interior buoyancy that subtracts from the exterior buoyancy and a mass / inertia. Through the new linearization functionality in OpenFAST under development, it has been found that when fluid-filled members are present in the HydroDyn model of a semisubmersible with offset columns, that sizeable platform-yaw angles may induce an unphysical asymmetric body-mass inertia matrix matrix. That is, the Ixy associated with fluid-filled members in HydroDyn may not equal to Iyx for large platform-yaw angles. This unphysical asymmetry can introduce negative damping and instability of the time-domain solution. We have not yet tracked down what in HydroDyn is causing this asymmetry, but for now, we recommend that you remove fluid-filled members in HydroDyn, and instead, place the equivalent mass and inertia in the structural module (ElastoDyn).
NREL is currently working on improvements to HydroDyn related to floating platform structural flexibility and member-level loads and we will look further into and (hopefully) resolve this bug during that effort.