A coupled hydroelastic solver for floating wind turbines: Development, validation, and application

Most hydro-aero-servo-elastic (HASE) simulation tools for Floating Offshore Wind Turbines (FOWT) model the platform as a rigid body, limiting their ability to predict internal structural loads. To address this, we present teddy, a new hydroelastic solver coupling the hydrodynamic capabilities of the in-house DIEGO toolbox with the open-source Chrono multibody dynamics library. This integration enables the modeling of fully flexible floating platforms. A multibody Boundary Element Method (BEM) provides the frequency-domain load input for the time-domain simulation. These bodies represent hull parts of the floater and are connected to Finite Element Model (FEM) nodes. Consequently, potential-flow loads in the multibody framework are transferred to the FEM model, implemented within Project::Chrono, which supports large deformations and nonlinear elasticity via the Absolute Nodal Coordinate Formulation (ANCF). The resulting coupled tool is first verified against theoretical results and DIEGO rigid-body simulations, yielding errors below 1% and 0.2%, respectively. Subsequently, the code is validated against experimental measurements from a flexible spar-type Floating Offshore Wind Turbines (FOWT), showing close agreement with experimental strain and motions Response Amplitude Operator (RAO) (errors under 3.5%). A slightly higher discrepancy of 9% is found for the tower-base bending moment RAO, which is attributed to uncertainties in the mass distribution of the experimental load cell. Furthermore, an analysis of a generic tri-floater foundation is conducted with two different hydrodynamic load models. This generic tri-floater is extensively described to be reproducible, and the corresponding results are presented and analyzed. Specifically, several significant bending-moment peaks, with a normalized RAO reaching up to 20 MN m/m, are observed in the brace structures at specific wave frequencies, particularly at T = 6 s where column spacing matches the incident wavelength. This underscores the critical need for hydroelastic modeling in the early design stages.