Experiment-Based Prediction of Cable and Riser VIV

The development of empirically-based riser VIV software has reached a critical point. These codes, which are based on experimental data and calibration against measured responses, provide reasonable predictions using theoretical and experimental results that have provided considerable understanding of the VIV phenomenon. At the same time it is known that certain un-modeled dynamics and physical mechanisms, such as in-line riser motions, Reynolds number, and shear flow parameters, have a quantitative effect that must be taken into account if predictions are to become more accurate. To take advantage of all the developments in the field so far and incorporate the information that is becoming increasingly available from field data and targeted experiments, a new approach is taken: Make a few critical parameters inside the program and in the experimental database adjustable, to account for the new information as it becomes available, and provide parameter-identification procedures to extract this information.

The program VIVA was built to incorporate a dense hydrodynamic database obtained from rigid cylinders of various shapes, oscillating in cross-flow. Incorporating this data has provided significant understanding of the influence of certain parameters. The database is admittedly the most comprehensive experimental database available in the industry today, especially after including high Reynolds number data. As field data have become available, the need arises for a more flexible program that allows the identification of hydrodynamic properties in three-dimensional flows, and their implementation on VIVA, while retaining the significant experience gained so far.

The hydrodynamic quantities that control the VIV response are:

(*) The coefficient in phase with velocity, which depends on the Reynolds number, the geometric shape of the riser cross-section and the shear in the flow.
(*) The correlation length.
(*) In a shear flow, the spreading of vortex energy among the various frequencies.

The recent discovery that the in-line motions can cause substantial changes to the amplitude of response and the frequency of the lift forces by generating large higher harmonics makes it imperative to change the fatigue calculation to account for these effects.

The new VIVA will start by providing a multi-frequency response in accordance with the method of the existing VIVA, but using a new, much simplified experimental database. Then, the simplified database controlling the clv will be adjusted until the frequency content and rms response match the full scale data. The procedure will follow these steps:

(*) First, the frequency width of the plot of clv versus frequency will be adjusted until the bandwidth of the response matches;
(*) Then the height and slopes of the clv curve (versus A/D) will be adjusted until the rms matches.