An experimental and computational leakage investigation of labyrinth seals with rub grooves of actual size and shape

Abstract

A large scale water test facility and a commercial CFD computer program were used to investigate labyrinth seals with rub grooves of actual size and shape found in aircraft engines. The 2-D test rig cases focused on the effect of tooth position and operating condition for the standard geometry. The computed cases considered tooth axial and radial position, different operating conditions, and several geometric dimensions. This investigation also compares the leakage of the standard geometry to that of a modified convex wall geometry. The test facility is a 33 times enlargement of the actual seal. The pressure drop leakage rate and flow visualization digital images for the standard geometry seal were measured at various Reynolds numbers and at nine different tooth positions. The discharge coefficient and a dimensionless pressure drop number were used to plot the leakage data to make it easier for seal designers to predict the leakage of labyrinth seals. The experimental visualization results show for a given Reynolds number that the closer the labryinth tooth gets to the step the deeper the throughflow jet penetrated into the seal cavity. The decrease of the tooth tip clearance also has a similar effect. Specifically the smaller the tooth tip clearance the deeper the flow path penetrated into the seal cavity. The experimental measurements show that the tooth tip axial position, as well as the minimum-tooth clearance, affect the leakage. A significant improvement in leakage was generally observed when the minimum-distance tooth clearance occurs across the entire tip of the tooth. This occurs only at the most upstream tooth position tested. Similarly, the computed results show that the tooth axial position affects the seal leakage. It was also found that the leakage of the modified convex wall geometry was significantly less than that of the standard geometry.