Date of Award


Degree Type


Degree Name

Master of Science in Ocean Engineering


Ocean Engineering

First Advisor

Jason M. Dahl


Vortex Induced Vibrations (VIV) are a critical problem in the offshore industry, where interaction between flexible immersed marine structures and natural currents result in large structural oscillations. These vibrations can result in fatigue life reductions, increased factors of safety, risk of unplanned failure, reduction in operational time, and may require costly mitigation strategies. Present methods of modeling VIV used largely in industry are limited to considering motion restricted to the transverse direction relative to flow. Semi-empirical prediction methods of VIV offer a good estimate for these vibrations, but expanding them to include inline body motion would create a prohibitive increase in the number of experiments required to properly characterise VIV, even at a single Reynolds number. This thesis documents the research, development, and implementation of a novel simulation method which combines VIV prediction with on demand experiments to significantly reduce the experimental effort. Previous semi-empirical prediction methods use large databases of hydrodynamic force coefficients, obtained from forced motion experiments to predict VIV. The new method developed in this thesis conducts experiments on-demand, using the Newton-Raphson method to select new experiment conditions, in order to obtain a prediction using significantly fewer experiments. On-demand experimentation with autonomous test runs in a fully integrated experimental tank inform the simulation at each step in the iteration. The system is demonstrated to reproduce observed free vibration VIV data for cases of data varying by Reynolds number, and varying mass-damping parameters. Results of the implementation of the system suggest a vast reduction in the time required to characterize VIV at unfamiliar Reynolds numbers, with the output verified in comparison to existing free vibration VIV data and prior forced motion experiments done at limited Reynolds numbers. In doing so, the reduction in time and complexity makes possible the desired future objective of adding in-line oscillations to the prediction method, without the burden of an unwieldy number of forced motion experiments to perform.