Galloping 2D Tests of a Two Cylinder Bundle Available to Purchase

Elastic mounted noncircular cross section geometries can produce large displacements and/or rotations often denoted galloping or fluttering, see e.g. Blevins (1990) and Faltinsen (1990). For such a structure the fluid forces that act on the structure change with orientation to the flow. If the oscillating fluid force tends to increase vibration, the structure is dynamically unstable and very large-amplitude vibration, galloping, can result. Statoil ASA has commissioned a model test campaign at MARINTEK to study the behaviour of flow induced vibrations for a 2D model of two pipes tied together with a special emphasize on the possibility of galloping response. The results were used in a Statoil study on the behaviour of free-span pipelines near sea bottom. However, the test results are generic and can also be relevant for deep water risers and umbilicals. The short rigid pipe section model was elastically mounted and was free to move as a single degree of freedom (SDOF) in the cross-flow direction. In some tests it was also allowed to rotate around the cylinders axial axis (2-DOF). By adjusting the torsion stiffness the same natural frequencies in the two DOFs were obtained. The cylinder was constrained in in-line direction. The tests were done for several different headings and towing velocities. Altogether 189 towing tests were done. As key result relative large displacement-to-diameter ratios were observed in the reduced velocities range 1–20. By evaluation of the oscillating frequency the response was characterized. For the lower reduced velocity range (say up to 10) the response could easily be characterized as VIV and for the highest reduced velocities as galloping. In an intermediate reduced velocity range a mixed kind of behavior is observed. These results coincide with results reported of flow induced vibrations of bridge decks with rectangular cross-sections, confer Blevins (1990). Most galloping analysis utilizes quasi-steady fluid assumption. A fundamental assumption for the quasi-steady assumption is that the fluid forces are quasi-steady and the structure oscillates around its natural frequency. Oscillating vortex shedding forces are at much higher frequencies and do not matter in the galloping response. Since the galloping response frequencies overlap with the vortex-shedding frequency it was concluded that the quasi-steady fluid assumption is not applicable for the present system.