This paper reports the first phase of an investigation aiming to determine the validity of using a CO2 marker in cold rig experiments to characterize the thermal performances of turbine rim seals under actual engine operating conditions. For comparison purposes, simulations are carried out for two sets of operating conditions, namely cold rig (with uniform low temperature) and real turbine thermal conditions (high temperature gaspath and cold purge flow). Sealing effectiveness based on the CO2 diagnostic under cold rig operating conditions is compared to sealing effectiveness based on the computed temperature field under real engine temperature conditions. Unsteady RANS simulations with different purge flow rates are performed. Tested geometries include a 180° domain presenting a simplified rim seal geometry with no vanes nor blades in the gaspath, and a 24° sector of a complete turbine stage including 3 vanes and 4 blades. Three-dimensional flow structures known to affect ingestion are found with both geometries but appear to be sensitive to the differences in operating conditions. Indeed, their circumferential number and strength differ between the two scenarios of conditions. Furthermore, it is found that the cold rig predictor tends to slightly overestimate the sealing effectiveness, while providing nonetheless the right trends and reasonably accurate average values in levels of actual sealing. At this stage of the investigation, we conclude that it seems adequate to use a passive tracer in cold rig experiments to compare performances of rim seal designs.

This content is only available via PDF.
You do not currently have access to this content.