The vibratory response of a long slender riser, made of composite materials and subject to an ocean current, is examined for a range of conditions. A major focus of this study is the performance of composite materials when used for risers. The influence of the number of modes of vibration is studied, as is the influence of the mass ratio and the value of the damping coefficient. The flow past the riser is represented by a shear flow, ranging from Re=8000 at the lower end of the riser to Re=10,000 at the upper end of the riser. The riser vibration is treated as a coupled fluid-flow/vibration problem. The fluid-flow equations are represented by a large eddy simulation model for the wake turbulence present in the flow. Strip theory is used to represent different forcing locations along the length of the riser. Since the composite riser has a material damping that is frequency dependent (it decreases with increasing frequency), its response is different from, say, a steel riser with a constant material damping. The composite riser, with variable damping, has a larger rms displacement than a riser with constant damping, primarily because of the smaller mass ratio. The vibration amplitude is found to increase with an increase in the number of modes.

The vibratory response of a long slender riser, made of composite materials and subject to an ocean current, is studied for a range of conditions. The influence of the number of modes of vibration is studied as is the influence of the mass ratio and the value of the damping coefficient. The flow past the riser is represented by a shear flow, ranging from Re = 8000 at the lower end of the riser to Re = 10,000 at the upper end of the riser. The riser vibration is treated as a coupled fluid-flow/vibration problem. The fluid-flow equations are represented by a Large Eddy Simulation model for the wake turbulence present in the flow. Strip theory is used to represent different forcing locations along the length of the riser. Since the composite riser has a variable damping coefficient, which decreases with increasing frequency, its response is different from, say, a steel riser with a constant damping coefficient. The composite riser, with variable damping, has a larger RMS displacement than a riser with constant damping. The vibration amplitude is found to increase with an increase in number of modes.

Vortex Induced Vibration (VIV) analysis of risers using time domain Finite Element Analysis (FEA) captures nonlinearity in the system. Such non-linearity may be in the form of boundary conditions or the imposed loadings. Riser VIV analysis using a frequency domain code has to linearize the system, such as imposing linearized boundary conditions. Non-linear boundary conditions for steel catenary riser (SCR) could be in the SCR-soil interaction and in the hang-off system. An SCR is often supported by a flex joint that has non-linear rotational stiffness. The effect of these non-linear boundary conditions in SCR VIV response is investigated in this paper. In addition, the effect of the directions of current profiles with respect to the riser on the SCR VIV response is also investigated. Comparison to standard industry practice, that is applying currents either in the plane or normal to the plane of the SCR, is presented. Time domain VIV code SimVIV coupled with FEA code ABAQUS was used in the analysis.

Substructure modeling using pipe-in-pipe (PiP) elements in a finite element program allows representation of dynamic interaction between riser components. This modeling technique is especially useful when it comes to the design of a complex riser system in deepwater applications. In this paper, the ABAQUS finite element program was used to illustrate the substructure models and the results for dynamic analysis of a classic Spar top tension riser (TTR) system in the Gulf of Mexico subjected to a given Hurricane Rita sea state. Nonlinear contacts between the buoyancy can and compliant guides are represented by two different substructure models: compliant guide surface model with friction and frictionless compliant guide spring model. The effects of centerwell hydrodynamic forces were considered. ABAQUS dynamic results were compared between the PiP substructure model and a conventional structure model treating the buoyancy can and the riser inside as a composite beam. The PiP guide friction surface model with centerwell hydrodynamics theoretically is the most representative model for riser analysis. However, the PiP guide spring model is more computationally efficient. It generates comparable guide loads but produces lower riser fatigue damage than the PiP guide friction surface model. The composite beam model leads to guide loads comparable to the PiP model, but cannot be used to determine the spacer loads between the buoyancy can and riser. The composite model also could underestimate riser stresses and riser fatigue damage within the buoyancy can region. The riser guide loads and riser damages from the calculation models without centerwell hydrodynamics are generally higher than those by the same calculation models with such consideration. It was concluded the PiP guide spring model can be used for riser design in lieu of the PiP guide surface model. The additional fatigue damage contribution from axial tension variation due to guide surface friction could be accounted for by adding a damage factor to the total fatigue damage along the riser.