Coupled CFD simulation of the response of a calm buoy in waves

This paper describes work done within the EU FP5 Project EXPRO-CFD to develop a system to couple commercial CFD software to existing hydromechanics tools to allow prediction of the response of floating structures in waves and currents, including viscous effects. Its focus is the use of this system to improve the prediction of CALM buoy response in waves. The Atkins EXPRO-CFD system is made up from the CFD code CFX, coupled to the AQWA-LINE and AQWA-NAUT hydromechanics codes. In this system, CFD provides the complete set of hydrodynamic forces and moments at each time step in the motions simulation, with the dynamics of the floating structure, its moorings and riser/export lines modeled in the AQWA-NAUT software. AQWA-NAUT returns the structure's displacements and velocities to the CFD model and a moving grid algorithm uses these to couple motions and fluid flow in an accurate and stable manner. The motions of a CALM buoy were studied to test the capabilities of the system. The CALM buoy geometry (based on current designs by SBM) is 23m in diameter with a 2m wide skirt attached 1m above the keel; the effects of flow separation off this skirt and the associated viscous damping on the motions of the buoy were expected to be significant, especially around its natural period. A series of 1:40 model scale tests were carried out by Sirenha using a simplified mooring system with no risers. The results from the model tests, from AQWA-NAUT alone (carried out by SBM), and from the coupled EXPRO model were compared directly at model scale. The same AQWA-NAUT model was used in both the AQWA-NAUT-only simulations and the coupled simulations, allowing direct comparison between the results. The EXPRO-system simulations were carried out 'blind', i.e. without access to the experimental data. The three sets of RAOs showed reasonable agreement in long or short waves (within the limits of the specification of the mooring system). However, around the natural period, the AQWA-NAUT-only model significantly over-predicted the response in heave, and in particular in pitch. Although the EXPRO-CFD system slightly over predicted the heave and pitch responses, the results were close to the experimental measurements throughout. Further tests indicate that the weakness in the potential flow approach appears to be in the formulation of added viscous damping rather than the choice of model values for drag coefficients.