Date of Award:

10-2011

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Civil and Environmental Engineering

Advisor/Chair:

Steven L. Barfuss

Abstract

In the arid west, water is among the most valuable of resources. Typically, in order to successfully and accurately measure a flow rate using most types of flow meters, it is recommended that a straight section of pipe be installed immediately upstream from the flow meter to avoid distorted flow patterns and extreme turbulence at the metering location. In many field piping situations, however, such flow conditions are impossible to achieve.

The performance of ultrasonic flow meters in non-ideal piping scenarios was studied using laboratory experiments and numerical Computational Fluid Dynamics models. A Fuji PortaflowX clamp-on transit-time ultrasonic flow meter was calibrated in the laboratory and performed well within the manufacturer’s stated level of accuracy of ±1% for straight pipe. The ultrasonic flow meter was also used downstream of a single 12-inch long-radius elbow and the error in flow measurement resulting from the flow disturbance was measured. These tests were performed at four locations downstream of the elbow, at three orientations on the circumference of the pipe, and for three flow rates with Reynolds numbers ranging from 250,000 to 750,000. Using the same physical conditions, numerical models were built that utilized the realizable k-ε eddy viscosity model for turbulent flows. The resulting downstream velocity profiles from the numerical models were used to integrate the velocity of the flow across the ultrasonic signal path. The resulting velocity profiles were compared to similar profiles for fully developed flow to determine the error in flow measurement. The ultrasonic flow meter measurement errors downstream of the elbow were always negative, and were found to be as great as -16%. In general, the ultrasonic flow meter performed less accurately closer to the elbow due to the turbulent fluctuations and flow separation created by the elbow.

A correction curve was applied to the results of this study, which provided a method to correct the ultrasonic flow measurement. By applying the correction equation to the laboratory ultrasonic flow meter measurements downstream of the elbow, the measurement error was reduced by nearly 90%. Computational Fluid Dynamics appears to be a viable tool for studying flow disturbances and making appropriate corrections for ultrasonic flow measurement in pipes.

Comments

This work made publicly available electronically on November 21, 2011.