Thesis--Auburn University, 2007. / Abstract. Vita. Includes bibliographic references (p.192-193).

This thesis presents results of an experimental investigation that characterizes the wall vibration of a pipe with turbulent flow passing through it. Specifically, experiments were conducted using a water flow loop to address three general phenomena. The topics of investigation were: 1) How does the pipe wall vibration depend on the average flow speed, pipe diameter, and pipe thickness for an unsupported pipe? 2) How does the behavior change if the pipe is clamp supported at various clamping lengths? 3) What influence does turbulence generation caused by holed baffle plates exert on the pipe response? A single pipe material (PVC) was used with a range of internal diameters from 5.08 cm to 10.16 cm and diameter to thickness ratios ranging from 8.90 to 16.94. The average flow speed that the experiments were conducted at ranged from 0 to 11.5 m/s. Pipe vibrations were characterized by accelerometers mounted on the pipe wall at several locations along the pipe length. Rms values of the pipe wall acceleration and velocity time series were measured at various flow speeds. Power spectral densities of the accelerometer data were computed and analyzed. Concurrent wall pressure fluctuation measurements were also obtained. The results show that for a fully developed turbulent flow, the rms of the wall pressure fluctuations is proportional to the rms of the wall acceleration and each scale nominally as the square of the average fluid velocity. Also, the rms of the pipe wall acceleration increases with decreasing pipe wall thickness. When changes were made in the pipe support length, it was observed that, in general, pipe support length exercises little influence on the pipe wall acceleration. The influence of pipe support length on the pipe wall velocity is much more pronounced. A non-dimensional parameter describing the pipe wall acceleration is defined and its dependence on relevant independent non-dimensional parameters is presented. Turbulence was induced using baffle plates with various sizes (2.54 cm to 0.159 cm) and numbers of holes drilled through them to provide a constant through area of 35.48 cm2 for each plate. Cavitation exists at high speeds for the largest holed baffle plates and this significantly increases the rms of the pipe wall acceleration. As the baffle plate hole size decreases, vibration levels were observed to return to levels that were observed when no baffle plate was employed. Power spectral densities of the accelerometer data from each baffle plate scenario were also computed and analyzed.