Application of flow-induced vibration predictive techniques to operating steam generators

Analytical techniques for flow-induced vibration (FIV), such as those incorporated in available design tools, are routinely applied to process equipment at the initial design stage. Unfortunately, this does not always apply to the situation when problems, related to FIV, develop in crucial operating equipment, since design uses conservative methods, whereas in-service applications require more realistic assessments. Usually these problems appear in the form of severe through wall fret flaws or fatigue cracks that compromise the integrity of the tubes and possibly the complete unit. It is here where a somewhat different approach must be taken in the evaluation of tube response to FIV. Tube damage from fretting wear or fatigue crack growth must be estimated from actual in situ operating conditions. In this paper, an overview of the predictive methods used in the development and/or qualification of remedial measures for problems that occur in operating process equipment along with applications are described. The steps in the evaluation procedure, from the prediction of flow regimes, the development of the nonlinear computer models and associated fluid forcing functions through to the estimates of tube damage in operating heat exchangers and steam generators are presented. A probabilistic (i.e. Monte Carlo simulation) FIV approach that readily accommodates uncertainties associated with damage predictions is summarized. The efficacy of this approach comes from the fact that probabilistic methods facilitate the incorporation of field data, and that a large number of tubes and possible variations in geometry, process and support conditions, usually present in such equipment, can be addressed effectively. (author)