Vortex-Induced Vibrations Of Submarine Pipeline Spans

Submarine pipeline spans can arise due to uneven seabed or scouring for the pipelines being laid on the cohesionless soil. Vortex-induced vibrations (VIVs) may occur when the spans are subjective to the effects of currents and cause unacceptable fatigue damage in the pipeline. Model tests and numerical simulations were performed to improve the understanding of the VIV phenomenon and predict response under various conditions in order to provide theoretical foundation and scientific basis for disaster process, damage mechanism and safe assessment of submarine pipelines in an offshore environmental condition.Flexible pipe tests were performed to investigate the effect of the proximity to seabed on VIV and the mode transition in VIV as the flow velocity varies. Two pipe models,16mm in diameter,2.6m in length and with mass ratios (mass/displaced mass) of 2.62 and 4.30 respectively were tested in a wave current tank. The pipe models were designed as bending stiffness dominated beams which are generally chosed to represent scaled models of pipeline spans, and were installed with universal joints at each end. The middle of the models had initial sags of about 14mm and 50mm respectively in water. The reduced velocity based on the first natural frequency is in the range of 0-16.7. Due to the restriction of the initial sags, the gap ratios at the pipe ends for the pipe model with the mass ratio of 2.62 were 2.0,4.0,6.0 and 8.0, and those for the pipe model with the mass ratio of 4.30 were 4.0,6.0 and 8.0. The response of the models was measured using fiber optic strain gauges. Modal analysis was applied to analyze the strain data from the tests, and the final results were the response amplitudes, response frequencies and the dominant modes. The effect of the proximity to seabed on VIV and the coupling of in-line and cross flow VIV were discussed. The results show that, as the gap ratio decreased, the shift in the dominant mode took place at a higher reduced velocity, and the dramatic increase in response frequency appears with the shift in dominant mode from the first mode to the second one; the pipe undergoes a common 8-shape motion for e/D>2.0, but a teardrop shape motion for e/D=2.0. Furthermore, the mode transition as the flow velocity increases is analyzed. The characteristics of VIV mode transition from the first mode to the second one for different gap ratios are revealed.A second set of experiments were conducted to investigate the effect of the pipe-seabed-interaction at the span shoulders on the VIVs of submarine pipeline spans. The pipe model with the mass ratio of 4.30 was tested in the wave current tank. The pipe was laid horizontally on the soil, and had a free span in length of 2.138m. The gap ratio at the pipe ends was maintained at 6.0. The tests in both still water and a current were conducted. The flow velocity was in the range of 0-0.60m/s. The frequency responses and the time-domain tracing of cross-flow strain responses are presented and analyzed. The experimental results exhibit several valuable features:the natural frequency of the model has a fuzzy property, but the response frequency in a current has not and increases linearly with the increase of flow velocity; the existence of soil support does not affect the harmonic characteristic of VIV.Combining the VIVANA model with a flexible beam model, a computer analysis program is developed for cross-flow VIV. The Newton-Raphson method is used to solve the dynamic equilibrium equation. Calculations are carried out using two different groups of parameter curves from Larsen et al. (2002,2004). The calculated results are compared with experimental results from both previous and present laboratory model tests. It is shown that Curve 1 (Larsen et al.,2002) where the maximum response amplitude for zero lift (i.e. CL=0) is from free vibration tests is more suitable for the VIV analysis of submarine free span pipelines. The agreement is in general good for the vortex-induced resonance at the first mode, and the VIV response model is able to identify the shift in the dominant mode.Finally, inverse force analysis is performed using the response data from the flexible pipe tests to investigate the effect of the proximity to seabed on the hydrodynamic force coefficients and predict the response with wall interface. The force contour plots and coefficients are generated. Response calculations are carried out using force coefficients from the inverse force analysis and the calculated results are compared with experimental data.