Vortex-Induced Motion of Deep-Draft Semisubmersibles: A CFD-Based Parametric Study

Deep Draft Column Stabilized Floaters (DDCSFs) provide a viable dry-tree platform for ultra-deep water applications. A particular area of concern for this concept is its susceptibility to Vortex-Induced-Motions (VIM) due to its deeper draft that results in higher column slenderness ratios than conventional semisubmersibles. The VIM characteristics of platforms have traditionally been assessed through experimental measurements but in recent years Computational Fluid Dynamics (CFD) has been used alongside experiments to predict VIM. Besides solving problems that cannot be addressed by other analysis methods, CFD can be seamlessly integrated into the overall concept design cycle where different geometric parameters can be varied to optimize the performance of the platform. While tank experiments give little fluid flow information without extensive instrumentation, CFD solutions provide flow details that can be used to improve design. CFD also provides a cost-effective solution given the high cost of experiments and facility scheduling constraints. One such DDCSF concept is the Paired-Column semi-submersible (PC-Semi) and the design has a pair of columns instead of one at each of the four sides of the platform. Several design parameters of the columns, like the inter-column spacing, the cross-sectional areas, and the shape of the respective columns can be tuned to minimize the heave response. This study takes a detailed look at the effect of the change in the column-dependent parameters on the overall VIM characteristics of the PC semisubmersible. A two-step approach is presented in which the first step is to identify aspects of the physics that are unique to VIM and then formulate a methodology to predict it within the framework of a set of commercially available CFD tools. The methodology is verified further through a comparison with available experimental data. The second step is to apply the verified methodology and CFD tools to predict the VIM performance of a range of PC-Semi concepts obtained through a systematic variation of the column-dependent design parameters in model scale. Through this exercise the critical design parameters that can either improve or prove detrimental to the VIM performance of the PC-Semi design are identified.