Many engineering applications with annular- or leakage-flow over a finite length can be encountered especially in the power generation plants. For instance, heat exchanger tubes with gap supports in steam generator, UO2 fuel rods with spacer grids in fuel bundles and fuel assemblies in gas-cooled reactors during refueling, etc. Nonetheless, few articles can be found on this subject. In this study, therefore, the annular-flow-induced vibrations of a pinned-pinned cylinder with and without a finite-length narrow-gap diffuser are studied by analytical and experimental methods.;The stability of a simply-supported tube subjected to narrow annular flow in a finite-length gap support is experimentally and analytically investigated. For the experiment, a 2.5 m test section and several finite-length gap supports have been made considering different gap size and diffuser angles of the support. The tube was observed to lose stability by flutter. The critical flow velocity was strongly dependent on the annular gap size and the diffuser angle at the downstream end of the support. A solution for the perturbation pressure on the tube is analytically obtained considering the friction loss, the contraction loss at the entrance, and the pressure recovery at the exit of the support. In the analytical solution, the exit boundary condition for pressure recovery is found to be predominant for flutter instability. However, flutter instability does not materialize for lossless boundaries such as short-lossless inlet and free-discharge outlet. Based on the solution, a simple semi-analytical model to predict the critical flow velocity is proposed for the first mode instability. The prediction of the semi-analytical model agrees reasonably well with the experimental results. However, it is judged that the pressure recovery at the diffuser should be experimentally measured more accurately to have better prediction.;For the annular-flow-induced vibrations of a pinned-pinned cylinder, an analytical model is proposed based on three main assumptions; (1) small perturbations in flow components, (2) negligible radial flow to reduce the annular flow to two-dimensional flow, and axial flow only for reduction to one-dimensional flow, and (3) perturbation frictional loss depending on the variation of axial perturbation velocity in terms of space and time. In this study, it is concluded that (1) the difference in fluidelastic forces between two- and one-dimensional flow models depends mostly on cylinder radius, and on whether perturbation flow is mainly allowed in the axial or circumferential direction, (2) the one-dimensional flow model should be limited to 1-d.o.f vibration analysis or the case of a cylinder having a large radius-to-length ratio, and (3) perturbation assumption makes little change to the dynamics of annular-flow-induced vibrations, however, the critical flow velocity is diminished considerably.
展开▼
