The roll damping coefficient is a crucial parameter for several design and operational aspects of FPSOs. The accurate prediction of the coefficient is not a trivial task and generally performed experimentally. A polynomial linearization of the decay test data has been widely applied in the offshore industry. However, research has indicated that for FPSOs with rectangular cross section and attached bilge keels, this methodology may lead to inaccurate damping coefficients.
This paper presents a study on the experimental determination of roll damping coefficients for FPSOs, obtained by free decay tests. For this purpose model tests are executed in the towing tank of the Marine Hydrodynamic Laboratory at Newcastle University. The model is based on the design of a purposely build FPSO, as typically applied in the central North Sea sector. The cross section of the FPSO is boxed shaped with a characteristic knuckle shaped bilge. The tests are conducted using three different bilge keel arrangements. The parametric change in bilge keel size results in the variation of the flow characteristics around the bilge knuckle. The damping coefficients are then established from the decay test data using a polynomial approach, a bi-linear approach and a hyperbolic approach.
A comparison between the damping evolutions obtained with the different methodologies is performed for each bilge keel configuration. Further, a numerical model of the FPSO is created using DNVs Sesam software. With the established damping coefficients, damping matrices are manually defined as an input to Sesam and roll transfer functions are numerically established. The computational determined transfer functions are then compared against the RAOs obtained from the model tests in regular waves to determine the most appropriate methodology.
The damping coefficient for the bare hull is well established by all three proposed methodologies. However, with the attached bilge keels the bi-linear and the hyperbolic methodologies produce damping coefficients reflecting the experimental results more accurately than the polynomial approach, indicating that the recently developed hyperbolic method is a valid alternative, and in certain cases, is more suitable to determine the roll damping coefficient. The experimental measurements could serve as a benchmark for further research and contribute to the practical application of FPSO roll damping determination.