Floating Body Modelling Detail |
 
|
Floating Body Modelling Detail |
|
There are varying levels of detail with which you can physically model the floating structure. The simplest possible option is to place finite element nodes only at key points of interest, such as the centres of gravity and buoyancy, and to model all the applied loads as concentrated forces at these points. Conversely, it is also possible to model the floating body in explicit detail, including finite elements to model the various pontoons, decks, braces etc. which comprise the real-world structure. Typically the simple approach is adopted by most users, but there are some advantages associated with creating a more detailed model. Both methods are now briefly summarised and contrasted.
The preceding article on Diffraction-Radiation Theory & Morison's Equation is related to this one. It discusses the main differences between these hydrodynamic modelling approaches.
Finite element nodes are placed at key locations only, typically corresponding to...
•Platform centre of gravity (CoG)
•Platform centre of buoyancy (CoB)
•Platform RAO reference point (CoF)
•Platform centre of drag (CoD)
•Fairlead positions of attached mooring lines.
This effectively serves as a framework upon which the various constituents may be applied. All of the Applied Loads are concentrated at these key locations. The elements are defined using the Flexible Format as this affords greatest control over the assigned properties.
•The floating structure is assumed to act as a perfectly rigid body, so the nodes are connected together via elements of relatively high stiffness (e.g. 1.0E+10 N.m2).
•The inertia of the floating body is concentrated at the CoG node, so the finite elements are assigned zero Mass per Unit Length. The inertial terms are specified via the Floating Body Inertia Inputs.
•The buoyancy force on the floating body is concentrated at the CoB node, so the finite elements are assigned zero Buoyancy Diameter. The effects of buoyancy are modelled using the Floating Body Hydrodynamic Stiffness Inputs.
•Regarding the previous two points, it is important to note that gravitational and mean buoyancy forces, both static, are not explicitly simulated. It is assumed that the initial position and orientation of floating body represents the static equilibrium configuration. Assuming equilibrium conditions, the static forces cancel each other out. The dynamic inertia of the floating body is correctly modelled in a dynamic simulation via the inclusion of the inertial terms in the global mass matrix. Similarly, the dynamically changing buoyancy force is modelled correctly via the hydrostatic stiffness terms which simulate changes in buoyancy force due to displacements and rotations of the floating body away from its mean position.
•Generally speaking, the wave excitation forces on the floating structure are determined using a diffraction solver. These forces are produced using a frequency domain solution technique and are provided in a concentrated form at a single point on the body. Diffraction codes readily provide the total loads so this is the most convenient means of transferring loads to the structural model. This location is identified as the RAO Reference Node in the Flexcom model.
•The overall drag force on the floating body is concentrated at the CoD node, so the finite elements are assigned zero Drag Diameter. The effects of viscous drag are modelled using the Floating Body Viscous Drag Inputs.
•As the element assembly looks rather skeleton-like, and does not remotely resemble the real world structure, the addition of an auxiliary Vessel Profile is advisable. While this does not have any structural function, it greatly enhances the visual appeal of the model, and assists in the understanding of floating body motions post-simulation.
The floating body is modelled in explicit detail using an assemblage of finite element nodes and elements which approximate the physical structural profile. The coordinates of the nodes, and the connectivity of the structural model, would have a similar level of detail to that used to create the optional auxiliary profile for the simple model described above. Having this detailed framework in place facilitates the application of distributed rather than concentrated loads, for some or all of the constituent terms. Again it seems logical to define the elements using the Flexible Format as this affords greatest control over the assigned properties. The following are some advantages of this approach.
•Elasticity/flexibility of the floating body may be captured. So the elements should be assigned appropriate structural stiffness inputs rather than arbitrarily high stiffness values (particularly in bending). In practice however, most floating bodies are effectively rigid so the choice of inputs is unlikely to have any major impact on the overall solution. Note also that hydrodynamic loading would need to be applied in a distributed manner. Assuming potential flow theory is being used, the wave excitation forces come from a diffraction solver. Wave pressures would need to be integrated over the hydrodynamic panels and the derived forces mapped to relevant locations in Flexcom model. An alternative approach would be to use Morison’s Equation rather than the diffraction approach. This is much easier to do, but the theory is not well suited to large structures like FOWTs where wave field disturbance is a feature. Refer to Diffraction-Radiation Theory and Morison's Equation for some background information.
•Structural mass can be modelled in a distributed manner. In this case, the elements should be assigned appropriate Mass per Unit Length terms, and the inertia of the floating body itself (specified via the Floating Body Inertia Inputs) should be set to zero. In a similar vein to the previous point, the floating body is effectively rigid, so there will be little or no difference between concentrated or distributed mass modelling. Note that if a distributed mass approach is adopted, then a distributed buoyancy approach (discussed in the next point) should also be adopted for consistency. This will ensure that both aspects inherently include static/mean and dynamic terms and avoid any complications arising from force imbalances.
•Buoyancy forces can be modelled in a distributed manner. This is a more accurate modelling technique which allows non-linear variations in buoyancy (e.g. due to significant pitch motions) to be simulated. In this case, the elements should be assigned appropriate Buoyancy Diameter terms, and the buoyancy associated with the floating body itself (specified via the Floating Body Hydrodynamic Stiffness Inputs) should be set to zero. Note also that if a concentrated mass approach is being used in conjunction with a distributed buoyancy approach, then the gravitational (static) force due to the floating body must be applied separately as a Point Load at the CoG node. This is necessary in order to ensure a balance of forces and static equilibrium.
•Viscous drag forces may be applied in a distributed manner. This is a more accurate modelling technique which allows non-linear variations in drag (e.g. due to intermittent submergence) to be simulated. In this case, the elements should be assigned appropriate Drag Diameter terms (with loading to computed via Morison's Equation), and the drag associated with the floating body itself (specified via the Floating Body Viscous Drag Inputs) should be set to zero.