We report additional tests of our extended Lee model for calibrating turbine meters. The model accounts for 1) Reynolds number (Re) dependent drag and lift, 2) bearing static drag and 3) bearing viscous drag. Initially, we tested this model using a dual−rotor, 2.5−cm−diameter turbine meter and flow measurements spanning a 200:1 range (50 Re 109,000) with liquid mixtures spanning a 42:1 kinematic viscosity range (1.2 × 106 m2 / s ν 50 × 106 m2 / s). The model correlated the volumetric flow data within 3.6 % over the entire Re range. The same data had a maximum deviation of 17 % from the commonly used Strouhal versus Roshko (or Re) correlation. In this work, we tested the model using three different single−rotor turbine meters with diameters of 2.5 cm, 1.6 cm, and 1.9 cm and flow measurements spanning a 75:1 range (140 < Re < 102,000) with liquid mixtures spanning a 12:1 kinematic viscosity range (1.2 × 106 m2 / s ν 14 × 106 m2 / s). The model correlates the flow data within 2.3 % for all three meters over the entire Re range. The same data had a maximum deviation of 4.76 % from the commonly used Strouhal versus Roshko (or Re) correlation. Therefore, the model works well for single−rotor and dual−rotor meters. The model shows that static bearing drag is responsible for fanning (or non−convergence) of multiple−ν calibration curves. However, as with the dual−rotor meter, the model begins to fail at low Re numbers where the bearing drags dominate the rotors behavior causing corrections as large as 26 % of the calibration factor.
Conference Dates: June 20-22, 2012
Conference Location: Colorado Springs, CO
Conference Title: 8th International Symposium on Fluid Flow Measurement
Pub Type: Conferences
Turbine Meter, Kinematic Viscosity, Propylene Glycol