Experimental Investigation of the Total Temperature Increase and Swirl Development in Rotating Labyrinth Seals

Abstract

Labyrinth seals are widely used as reliable components in many areas of turbo machines, e.g. the cooling air system in gas turbines. While the discharge behavior is generally well predictable, the uncertainty predicting the exit circumferential velocity (exit-swirl) and the total temperature increase due to internal losses (windage heating) is comparably large. In order to evaluate analytical correlations and for the validation of numerical simulations convergent and divergent stepped labyrinth seals were investigated experimentally. The change in total temperature across the labyrinth seal was measured in a test rig capable to establish different rotational speeds, pressure ratios and various inlet swirls. In an engine, honeycomb abrasive liners on the stator protect the seal fins. To simulate real engine conditions honeycombs were applied in the test setup, too and the influence of these liners on the windage heating was compared to smooth stator configurations. Detailed velocity profiles within the seal chambers were determined using a 2D Laser-Doppler-Velocimeter. Additionally, the ability of axisymmetric numerical k-ε simulations to predict the data was evaluated. The present study provides important data for the design of future turbo machines, because the exact knowledge of the labyrinth seal exit swirl and temperature is expected to further improve the design of downstream components such as the pre-swirl system. Additionally, more accurate boundary conditions for the thermal analysis will be available and the rotor dynamic stability of the seal can be estimated better.

Validation of a CFD Model of a Labyrinth Seal for Low Pressure Turbines Using a Fluid-Thermal Tool Tuned Through Experimental Measurements

“…CFD analyses, concerning labyrinth seals in both smooth and honeycomb configuration, are presented also by Yan et al [7][8][9][10][11]. The outcomes of their numerical analyses were compared with the experimental data available in the papers by Denecke et al [6,12].…”

Experimental and numerical validation study of the labyrinth seal configurations

“…In their experiments most researchers focus on stationary configurations, where the motion of the rotor with the labyrinth seal relative to the casing is ignored. In experimental works (Denecke et al, 2005), (Paolio et al, 2007) it is proved that this simplification may be assumed if the flow velocity over the labyrinth seal fin is higher than the rotor tangential velocity -in this case stationary test rigs simulate the flow behaviour correctly and the condition u/c ax < 1 is satisfied. However, if the abovementioned condition u/c ax < 1 is not met, a substantial increase occurs in the mass flow difference between the stationary model and the rotating one.…”

A Novel Design of Cooling Air Supply System with Dual Row Pre-Swirl Nozzles

“…As discussed above, the theoretical equations make the assumption that the swirl ratio at receiver hole is equal to 1, and the friction torque on the surface of the stator disk is ignored, however, in practice, the deviation of receiver hole swirl ratio from 1 is far, leading to a huge deviation in total temperature reduction coefficient, besides, the friction torque on the surface of stator disk is mainly influenced by rotational Reynolds number, angular velocity and average swirl ratio in the rotor-stator cavity (Denecke et al 2005) and non-ignorable in high rotational Reynolds numbers. These two are the main cause of deviation between CFD and theoretical results.…”