Reconstructing Turbine Non-Uniform Flow Including Hot Streaks Using Spatially Under-Sampled Measurements

Xu Zhu,Fangyuan Lou,Kai Zhou,Xinqian Zheng
DOI: https://doi.org/10.1115/gt2022-83509
2022-01-01
Abstract:The demand for higher thermal efficiency of gas turbines has been pushing the limit of turbine inlet temperature. Modern high-pressure turbine (HPT) blades are cooled with a combination of film cooling, impingement cooling, and passage cooling. However, for a given material and cooling strategy, the turbine inlet temperatures are limited by the performance penalty caused by the required cooling air typically extracted from the high-pressure compressor. Although it is well recognized that the HPT inflow is highly unsteady and non-uniform, it is still generally the case that turbines are designed for uniform inlet temperature-often the predicted peak gas temperature due to the vacancy of a reliable methodology for estimating the non-uniform temperature field at HPT inlet. This paper introduces a new approach to predicting the turbine inlet temperature, including the magnitude of non-uniformity, based on a few temperature measurements at the HPT exit to shed light on this topic. The technique includes three steps. First, investigate the hot streaks migration across the HPT and establish a projection of hot-streaks-related magnitude from HPT inlet to exit. Second, resolve the hot-streaks-related extent at HPT exit using a multi-wavelet approximation method based on a few discrete measurements. Lastly, estimate the HPT inlet mean temperature and magnitude of hot streaks. The method's effectiveness is examined using the two-stage high-pressure turbine from Energy Efficiency Engine (E3). An engine representative HPT inlet temperature distribution available in the open literature is applied at the HPT inlet. The true temperature distribution at HPT exit is obtained from full-annulus unsteady simulations. The estimated HPT inlet temperature distribution using the new approach agrees with the true values, with an error of less than 0.3% in the mean temperature and uncertainty on the order of 20% in the magnitude of hot streaks.
What problem does this paper attempt to address?