Measured Leakage and Rotordynamic Force Coefficients for Two Liquid Annular Seal Configurations: Smooth-Rotor/Grooved-Stator Versus Grooved-Rotor/Smooth-Stator

This paper reports and compares the experimental results of leakage and dynamic force coefficients for two liquid annular pressure seals, one having a smooth-rotor/circumferentially-grooved stator (SR/GS), the other one with a circumferentially-grooved rotor/smooth-stator (GR/SS). Differing only in the grooves’ location, the GR/SS seal’s geometry and operating conditions are representative of those in electrical submersible pumps (ESPs) used for oil recovery. Supplied with an ISO VG2 oil at 46 °C, both seals have the same diameter D = 102 mm, length-to-diameter ratio L/D = 0.5, and nominal land clearance Cr = 0.203 mm. The seals have 15 circumferential grooves with groove and land lengths equal to 1.52 mm. Test variable ranges include: (a) shaft speeds (ω) ranging from 2 to 8 krpm (shaft surface speed ∼ 43 m/s), and (b) pressure differences (ΔP) from 2 to 8 bar. Upstream of the test seals, three separate pre-rotation rings generate a range of inlet circumferential velocities (entrance swirl). Under all conditions, the GR/SS seal leaks about 10% less than the SR/GS seal. For both seals, the direct stiffnesses (KXX, KYY) have low magnitudes that drop with increasing ω; in some cases, they turn negative at 6 krpm. The GR/SS seal produces cross-coupled stiffnesses (KXY, KYX) that are ∼1.5 times larger than those for the SR/GS seal. Under the same conditions, the SR/GS seal is more stabilizing as its direct damping and added mass coefficients are ∼ 20% larger than those for the GR/SS seal. Instability issues are likely to arise with either seal geometry because negative KXX and KYY drop a pump critical speed, aggravating the well-known destabilizing coefficients KXY and KYX. The whirl frequency ratio (WFR) combines the effects of the cross-coupled stiffness, direct damping and cross-coupled mass terms, thus providing a good basis for comparing two seals’ stability characteristics. Overall, the WFR magnitudes for the GR/SS seal are about three times higher than those for the SR/GS seal. Note that, irrespective of the inlet swirl condition, the GR/SS approaches a WFR ∼ 0.50 for operating shaft speeds greater than 4 krpm. At the lowest shaft speed (2 krpm), the WFR << 0.5 for the low inlet pre-swirl ring, whereas WFR > 0.8 for the medium and high pre-swirl rings. For the SR/GS seal, the WFR ∼ 0.2 at the highest shaft speed of 6 krpm, and not affected by the inlet pre-swirl condition. On the other hand, at the lowest shaft speed, the WFR ranges from 0.5 to 0.75 for the SR/GS seal with medium and high pre-swirl rings. At this speed, the WFR = 0 with the low inlet pre-swirl ring. Hence, to enhance the operational stability, an effective swirl brake that could drop the inlet pre-swirl ratio upstream of a seal is helpful for the GR/SS seal out to 4 krpm and for the SR/GS seal out to 6 krpm.