抄録
Fluidelastic instability remains the most important vibration excitation mechanism in nuclear steam generators (SGs). Design guidelines, aimed at eliminating the possibility of fluidelastic instability, have been developed over the past 40 years. The design guidelines, based on the Connors equation, depend on a large database on cross-flow fluidelastic instability i.e. instability in the direction transverse to the flow. Past experience had shown that for an axi-symmetrically flexible tube, instability generally occurred in the transverse direction, at least at first. Although often not explicitly stated, there has been an implicit assumption that the in-plane direction was either stable, or would only suffer instability at velocities significantly higher than the transverse direction. This explains why SGs are fitted with anti-vibrations bars (AVBs) to mitigate transverse (out-of-plane) vibrations with no equivalent consideration for potential in-plane instability. This 'oversight' recently came to a head when SG at the San-Onofre NPP suffered in-plane fluidelastic instability. The present paper addresses the question of in-plane fluidelastic instability in large SGs. A historical review is presented to explain why this potential problem was left unresolved (or ignored) over the past 40+ years, and why engineers got away with it - at least until recently. Following the review, some recent work on in-plane fluidelastic instability modeling, using the quasi-steady model is presented. It is shown that in-plane fluidelastic instability can be fully modelled using this approach. The model results are used to propose some changes to existing design guidelines to cover the case of in-plane fluidelastic instability.
