Visualization of vibration experienced in offshore platforms

Abstract

In this thesis, I design and evaluate methods to optimize the visualization of vortex-induced vibration (VIV) in marine risers. VIV is vibration experienced by marine risers in offshore drilling platforms due to ocean current flows, and appears to be perpendicular to the direction of such flows. VIV causes oil companies large capital losses, supply chain disruption, and environmental and brand name damage. For these reasons, both researchers and manufacturers try to improve their models of VIV, while creating risers more resilient to it. The first step to understanding VIV is rapid visualization, ie. the ability to efficiently visualize large amounts of simulated and field data. In this thesis, I evaluate high and low level heuristics that optimize the run-time performance of applications by taking advantage of 64 -bit machines with large memory stores. Such heuristics include the introduction of object-oriented programming (OOP) with classes, dynamic binary loading, and source code management. I demonstrate that using these techniques allows speedups of many orders of magnitude, depending on the type of optimization and the structure of the input data. Finally, I reengineer an existing collection of disparate visualizations to take advantage of these heuristics, and achieve a run-time speedup of two orders of magnitude in most visualizations.