Abstract
To predict axisymmetrical fluid-pipe coupled vibrations, there have been proposed a variety of theories ranging from the sophisicated shell model to the simplest but convenient one-dimensional approximation neglecting Poisson's ratio. In contrast, experimental works to verify the theories are sparse as yet. The present paper intends to experimentally single out an appropriate model for longitudinal vibrations of a pipe containing viscous fluid, and also to estimate the internal friction of pipe metals, which is supposed to take part in energy dissipation. Frequency response tests have been carried out in oil hydraulic pipes with one end fixed and another free. The shell model neglecting rotary and transverse inertia proves to be most available, while the one-dimensional model is incorrect to describe the pipe behavior. It also turns out that the solid viscosities of the metals must be less than one-tenth of their conventionally obtained values.
