No abstract

This paper focuses on the further development of a previously published semi-empirical method for time domain simulation of vortex-induced vibrations (VIV). A new hydrodynamic damping formulation is given, and the necessary coefficients are found from experimental data. It is shown that the new model predicts the observed hydrodynamic damping in still water and for cross-flow oscillations in stationary incoming flow with high accuracy. Next, the excitation force model is optimized by simulating the VIV response of an elastic cylinder in a series of experiments with stationary flow. The optimization is performed by repeating the simulations until the best possible agreement with the experiments is found. The optimized model is then applied to simulate the cross-flow VIV of an elastic cylinder in oscillating flow, without introducing any changes to the hydrodynamic force modeling. By comparison with experiment, it is shown that the model predicts the frequency content, mode and amplitude of vibration with a high level of realism, and the amplitude modulations occurring at high Keulegan-Carpenter numbers are well captured. The model is also utilized to investigate the effect of increasing the maximum reduced velocity and the mass ratio of the elastic cylinder in oscillating flow. Simulations show that complex response patterns with multiple modes and frequencies appear when the maximum reduced velocity is increased. If, however, the mass ratio is increased by a factor of 5, a single mode dominates. This illustrates that, in oscillating flows, the mass ratio is important in determining the mode participation at high maximum reduced velocities.

A previously proposed hydrodynamic load model for time domain simulation of cross-flow vortex-induced vibrations (VIV) is modified and combined with Morison's equation. The resulting model includes added mass, drag and a cross-flow vortex shedding force which is able to synchronize with the cylinder motion within a specified range of non-dimensional frequencies. It is demonstrated that the hydrodynamic load model provides a realistic representation of the cross-flow energy transfer and added mass for different values of the non-dimensional frequency and amplitude. Furthermore, it gives a reasonable approximation of the experimentally observed drag amplification. The load model is combined with a non-linear finite element model to predict the cross-flow VIV of a steel catenary riser in two different conditions: VIV due to a stationary uniform flow and VIV caused by periodic oscillation of the riser top end. In the latter case, the prescribed motion leads to an oscillating relative flow around the riser, causing an irregular response. The simulation results are compared to experimental measurements, and it is found that the model provides highly realistic results in terms of r.m.s. values of strains and frequency content, although some discrepancies are seen.

No abstract

Free hanging deep water flexible risers are associated with high static top tension. In addition comes significant dynamic tension due to the tangential drag forces mobilised along the riser due to floater motions and wave loads. At the upper end fitting where conditions of direct metallic contact between tensile armours might occur, this may result in the fretting fatigue failure mode. This mechanism may be additionally triggered by the increase in longitudinal stress resulting from local bending at the end fitting fixation. At the touch down point, the pipe cross-section is exposed to high external pressures resulting in true wall compression possibly initiating local buckling modes in the tensile armour wires. The buckling process may be initiated by cyclic bending effects leading to a gradual reduction of each tendon’s capacity resulting in excessive transverse displacements and cross-section failure. The present paper presents a finite element formulation and analytical models addressing both of the above topics. A case study is further carried out to document the performance of the FE model and to investigate effects related to the transverse bending stress at the end fitting and under which conditions one single armour tendon will fail in different buckling modes.