Identification of dynamic force coefficients of a labyrinth and gas damper seal using impact load excitations

Abstract

Experiments to identify stiffness and damping force coefficients of a two bladed teeth-on-stator labyrinth seal and a gas damper seal, both of diverging clearance, are presented. Calibrated impact guns excite a housing holding the test seal, and the seal displacement and acceleration time responses in two orthogonal directions are measured. A frequency domain parameter identification procedure allows the determination of the seals' dynamic force coefficients over a frequency range. Tests are made for a centered seal condition without journal rotation and with rotation at 1,500 and 3,000 rpm. The pressure drop across the seal is controlled by increments in the inlet supply pressure to values three times the exit (ambient) pressure. The two bladed labyrinth seal becomes less stable with increasing pressure ratio across the seal. This seal exhibits positive direct stiffness and negative direct damping. Cross-coupling is observed but identification of cross-coupled stiffnesses is poor. Comparison with analytical predictions is good for mass flow and direct damping coefficients. Predicted values for direct stiffness are negative, contrary to identified values. The same labyrinth'seal is transformed into a four pocket damper seal and tested under identical conditions. The damper seal is shown to be more stable dynamically but can create a static instability due to excessive negative direct stiffness. In the test facility used, the seal housing does not have enough direct stiffness to support the damper seal at pressure ratios greater than 2.0. Cross-coupled stiffness is not detected in the measurements. The direct damping coefficients are positive and large in magnitude compared to those from the labyrinth seal measurements. Analytical predictions show similar trends for mass flow, direct stiffness and direct damping coefficients. tests but this is not identified from the seal response measurements.