Numerical-Experimental Study of Global Buckling in Catenary Risers

Abstract

Abstract Risers are critical structures for the offshore oil and gas industry connecting floating production platforms to seabed equipment for production, injection and export functions, often through catenary configurations. The effect of external flow induced vibrations (VIV) and the occurrence of buckling are critical factors to lifespan of these structures. Therefore, the consistent evaluation of these factors is a strategic issue. Software for riser structural analysis usually employs Morison’s equation [1] to evaluate hydrodynamic forces along riser structure. Although this methodology is well established, their results are potentially conservative due to simplifications, besides not including lift forces. The present work employs an alternative methodology to calculate hydrodynamic loading on risers, based on the discrete vortex method (DVM) [2]. The DVM uses the Lagrangian approach in the vortex modeling, for incompressible, two-dimensional flows with regions of high vorticity and with dominant convective effect over the viscous one. The method creates and moves vortices along the riser wall perimeter, updates wake vortices at every time step considering the Biot-Savart law and calculates circulation by imposing the zero normal velocity condition on the riser wall. The structural analysis software, based on finite element method (FEM), Anflex [3], drives the DVM algorithm. Anflex applies an implicit time integration algorithm based on Newmark method together with three-dimensional nonlinear Euler beam elements with large displacements. This coupling occurs by modeling the flow in two-dimensional domains, called DVM planes, associated with each structural finite element along riser structure. The flow through DVM planes is responsible for the hydrodynamic forces on the structure, which in turn interferes with the fluid flow by structure displacements, performing a two-way coupling process [4]. This work presents experimental model results of a free catenary riser subjected to top-end displacements in still water, and compares them to numerical results obtained by Anflex/DVM. The experimental riser model is 41.45 m long under a 14.5 m water depth. Two-dimensional AMTI load cells [5] measured the catenary top force values and MCS Qualisys camera system [6] captured displacements on 40 points along the riser model. LabOceano hydrodynamic tank [7] performed the tests. The results comprised out-of-plane catenary vibration, catenary top force, and deformed configuration with global buckling at touch down point region (TDP). The experimental results showed significant self-induced vibrations due to the vertical movements applied at top connection, which indicated that the global buckling at TDP is highly influenced by these vibrations. Numerical analysis using Anflex/DVM showed good agreement with the experimental results. Anflex/DVM satisfactorily captured TDP buckling, which did not occur for the model based on Morison’s equation. These results indicated that the DVM-based method leads to more realistic dynamic responses, when compared to Morison’s equation. This paper defines global buckling as the dynamic wavy configuration that takes place at the TDP region of the riser under compression loading due to the vertical movement imposed at the riser top.