Moreland, James Alexander (2016-05). Influence of Pre-Swirl and Eccentricity in Smooth Stator/Grooved Rotor Liquid Annular Seals, Measured Static and Rotordynamic Characteristics. Master's Thesis.
Thesis
Circumferential grooves are machined into the rotor surface of liquid annular seals to reduce leakage. Analyses in the literature regarding rotordynamic coefficients for smooth stator/grooved rotor (SS/GR) seals are scarce, and measured results are absent despite the significant effect of these annular seals on pump rotordynamics. Furthermore, few annular seal test rigs can impose and measure pre-swirl. Consequently, this study aims to quantify the effect of pre-swirl and operating static eccentricity on measured static and dynamic characteristics of the SS/GR seal. The test seal has a nominal minimum radial clearance of Cr = 203.2 um (8 mil) and the length-to-diameter ratio of L/D = 0.5. The rotor surface has 15 square grooves with a depth of 1.524 mm (60 mils). This geometry is based on measured dimensions in an Electrical Submersible Pump. The author conducts tests at speeds of 2, 4, 6, and 8 krpm, axial pressure drops of 2.07, 4.14, 6.21, and 8.27 bar (30, 60, 90, and 120 psi), and eccentricity ratios up to ?0 = 0.80. Pre-swirl is imposed and varied over low, medium, and high values. The lubricant is ISO VG 2 oil at 46.10C (1150F) to ensure turbulent flow at all operating conditions. Measured grooved seal results are compared to results of a smooth seal of equal length, diameter, and nominal clearance. Note that no published predictions for this seal exist, precluding measurement-versus-prediction comparisons. Static measurements include leakage rate, applied static load, eccentric position, pre-swirl ratio (PSR), and outlet swirl ratio (OSR). At 2 krpm and 8.27 bar, the SS/GR seal leakage rate is lower than the smooth seal by a factor of 0.7 for a total reduction of 19 LPM. However, at high speeds and low pressures, grooved seal leakage is 0.8 to 0.9 times that of the smooth seal with reductions in leakage rate on the order of 2 LPM. Consequently, adding rotor grooves to restrict leakage is more advantageous at high pressure drops and low speeds. Outlet swirl is measured for the first time in a liquid-annular-seal test rig. Contrary to expectation, measured OSR is generally lower than 0.5, possibly due to the axial and radial location of the pitot tube within the outlet chamber. Dynamic measurements include rotor-stator relative displacement, stator acceleration, and applied dynamic excitations. The author calculates dynamic-stiffness values, rotordynamic coefficients, whirl frequency ratio, and effective damping. Most notable of the dynamic results is the negative direct stiffness of the SS/GR seal. Negative direct stiffness would have a detrimental effect on pump rotordynamics, lowering both the natural frequency and the onset speed of instability. Calculated whirl frequency ratio is generally high (>=0.5) with values up to 1.2 for the grooved seal. Finally, effective damping is lower for the grooved seal than the smooth seal indicating that the SS/GR seal has worse stability characteristics than a comparable SS/SR seal. Overall, increases in PSR have a detrimental effect on SS/GR seal rotordynamic performance leading to modest decreases in direct stiffness, increases in the magnitude of cross-coupled stiffness, and increases in whirl frequency ratio. Additionally, operation at high eccentricity ratios for the SS/GR seal has negative effects. Leakage increases, and direct stiffness decreases as eccentricity ratios approach 0.80.
