In turbomachinery, annular seals are used to reduce leakage between regions of high and low pressure. Many configurations of annular seals have been developed and studied in the literature including plain, labyrinth, pocket-damper, honeycomb, and hole-pattern. In machines experiencing stability issues, honeycomb and hole-pattern type seals have been used to replace labyrinth seals. Bulk-flow models are typically used to predict the leakage and dynamic coefficients of hole-pattern seals, relying on empirically derived friction factor coefficients. Previous experimental studies have shown that, for hole-pattern seals, the leakage and stator friction factor are strongly influenced by hole-depth. However, this behavior is not a monotonic function of hole-depth, a fact that might reduce confidence in future bulk-flow model predictions if not properly accounted for. A recent numerical study has highlighted the role of vortex formation in the holes which has a strong influence on the flow in the clearance region. Depending on the shape of the vortex, the flow in the hole can act much like a pinch valve, reducing the effective clearance of the jet flow.
In this paper, computational fluid dynamics simulations of several hole-pattern seal configurations have been performed to study the effect of hole-aspect ratio (depth versus diameter) on the leakage and friction factors. The Reynolds Averaged Navier Stokes (RANS) equations with k-ε turbulence model were solved using ANSYS CFX. It was found that the shape of the hole influences the vortex formation within the hole, effecting the jet flow in the clearance region and the seal leakage. The results show that the leakage is heavily dependent on the hole diameter in addition to the hole depth. The relationship between the friction factors and the geometry of the seal was found to be non-monotonic. It is therefore difficult to develop a friction factor model that will accurately encompass all configurations and it is recommended that friction factor data be interpolated from experimental or numerical results.