ADDRESSING SHELL MODE VIBRATION IN DUCTS IN REFINERY WITH COMPUTATIONAL MODELS AND FIELD DATA

摘要

This paper describes the work performed to study the shell mode vibration of a large cross-section flue gas duct. The work involved the collection of field vibration data, as well as predictive computational models associated with finite element analysis (FEA) and computational fluid dynamics (CFD). The goal of this work was to use predictive models to ascertain whether a proposed design change would reduce the vibration levels of the duct under similar operating conditions. The vibration observed in the duct was identified as a flow induced vibration (FIV) which excited the shell modes of the duct walls. This case study serves as an example of using predictive computational models (FEA and CFD), calibrated with vibration response data from field measurements, to represent the real world situation as closely as possible within specified budget and schedule constraints. Such calibrated models can be useful for forecasting the effectiveness of various proposed design changes.