A stack installation experienced vortex-induced vibrations (VIV) while in-service. The magnitude of the vibrations was severe enough that cracking in the welds at the stack bases was experienced shortly after their installation. Initially, straight strakes were placed on the stacks, based on field serviceability, per API 560. The strakes proved ineffective and it was determined that the stacks would be uninstalled for repair. During the repair process, design steps were required to reduce or eliminate the VIV experienced in-service. Due to a flaw in the initial design and the ineffectiveness of the straight strake solution, the end client required verification of any proposed design changes before their implementation. Additionally, little time was allotted for the investigation of solutions.
Tuned mass dampers were initially explored for the design modification. It was determined that they could not be constructed of materials suitable for the environmental operational characteristics of the stacks. It was then agreed that aerodynamic modifications of the stacks should be explored to reduce VIV. ASME STS I specifies the design and installation of helical strakes on stacks, but does not indicate the magnitude of vibration reduction that can be expected . Therefore, numerical models were used to determine if the strakes would reduce or eliminate the service vibration.
A baseline analysis was first conducted to validate that the tools — a combination of computational fluid dynamics (CFD) and finite element (FE) methods — could capture the in-service behavior. To perform this analysis a CFD model was constructed of the as-built stack. Using DES methods, this model was analyzed at several wind speeds to determine the magnitude and frequency content of the VIV-forcing functions. This information was then used to perform a dynamic analysis using an FE model of the stack. This model accurately predicted the correct wind speed corresponding to VIV and the amplitude of the stack’s vibration. A second model was then constructed of a stack with helical strakes, using a novel modeling methodology. This model was analyzed over a variety of wind speeds using DES methodologies. The forcing functions predicted with the helical strake model were then used to determine the stack’s in-service response. This paper contains the complete methodologies and results associated with these analyses.