High-fidelity computational modelling of fluid-structure interaction for moored floating bodies

The development and implementation process of a complete numerical framework for high-fidelity Fluid–Structure Interaction (FSI) simulations of moored floating bodies using Computational Fluid Dynamics (CFD) with the Finite Element Method (FEM) is presented here. For this purpose, the following three main aspects are coupled together: Two-Phase Flow (TPF), Multibody Dynamics (MBD), and mooring dynamics. The fluid–structure problem is two-way and fully partitioned, allowing for high modularity of the coupling and computational efficiency. The Arbitrary Lagrangian–Eulerian (ALE) formulation is used for describing the motion of the mesh-conforming fluid–solid interface, and mesh deformation is achieved with linear elastostatics. Mooring dynamics is performed using gradient deficient Absolute Nodal Coordinate Formulation (ANCF) elements with a two-way mooring–structure coupling and a one-way fluid–mooring coupling. Hydrodynamic loads are applied accurately along mooring cables using the solution of the fluid velocity provided by the TPF solver. For this purpose, fluid mesh elements containing cable nodes that do not conform to the fluid mesh are located with a computationally efficient particle-localisation algorithm. As it is common for partitioned FSI simulations of solids moving within a relatively dense fluid to experience unconditional instability from the added mass effect in CFD, a non-iterative stabilisation scheme is developed here. This is achieved with an accurate and dynamic estimation of the added mass for arbitrarily shaped structures that is then applied as a penalty term to the equations of motion of the solid. It is shown that this stabilisation scheme ensures stability of FSI simulations that are otherwise prone to strong added mass effect without affecting the expected response of structures significantly, even when using fully partitioned fluid–structure coupling schemes. Thorough verification and validation for all aspects of the FSI framework ultimately show that the produced numerical results are in good agreement with experimental data and other inherently stable numerical models, even when complex nonlinear events occur such as vortices forming around sharp corners or extreme wave loads and overtopping on moving structures. It is also shown that the mooring dynamics model can successfully reproduce nonlinearities from high frequency fairlead motions and hydrodynamic loads. The large-scale 3D simulation of a floating semi-submersible structure moored with three catenary lines ties all the models and tools developed here together and shows the capability of the high-fidelity FSI framework to model complex systems robustly and accurately.