| description abstract | Abstract. This study introduces the first application of a sparsity-promoting linear regression technique to analyze the flow behavior within the rim seal region between the rotor and stator disks in a highly instrumented engine-representative turbine setup. This approach allows decomposing pressure fluctuations in the rim seal into multiple rotating structures, providing a more precise and detailed interpretation of data collected by flush-mounted, fast-response pressure transducers (Kulite XCQ-062). Many sources of periodic fluctuations, such as the HP rotor, the LP rotor, and the rotor–rotor interaction, characterize the flow in such a complex environment as a two-spool rig. Therefore, the decomposition provided by the linear regression algorithm is beneficial in identifying further periodic structures, such as nonblade passing-related cavity flow modes. The calculated angular speed of the identified structures is qualitatively validated by comparing it with the measured swirl ratio in the cavity at the same radial height. The research was conducted at the Institute of Thermal Turbomachinery at Graz University of Technology, specifically in the Transonic Test Turbine Facility (TTTF). The TTTF is a state-of-the-art facility that replicates engine-relevant flow conditions. It consists of the last high-pressure turbine stage, the turbine intermediate duct, and the first low-pressure turbine (LPT) rotor. Purge flows were supplied to all cavities to simulate engine-like conditions. The fast-response pressure transducers were installed in the downstream hub cavity (DHC), located between the high-pressure stage and the intermediate turbine duct. The measurements were performed across different purge flow rates (PFR) and different stator-clocking positions, producing a comprehensive dataset. The analysis uncovered low-intensity, nonblade passing-related flow modes within the cavity, which could not have been easily identified using other methodologies. | |
