Application of a High Order LES Approach to the Redistribution of Inlet Temperature Distortion in a Turbine

The demand of an increase in the cycle performance of today’s gas turbines creates severe heat loads in the first turbine stage, since higher operating temperatures are required. The mean flow temperature is usually well above the limit supported by the surrounding material. Cooling of both end-walls and the blades of the first stage is thus usually necessary. Consequently, mid-span streaks of hot gas pass through the first stator row and become hot jets of fluid. Also, the exit flow from a gas turbine combustor entering a turbine stage can have a wide variation in temperature. These variations may be both spatial and temporal. The implementation of cooling method requires a clear understanding of the aerodynamics involved. Both qualitative and quantitative assessments of the redistribution of inlet temperature distortions can be used to considerable advantage by the turbine designer. Experimentally it has been demonstrated that the rotor actually separates the hotter and cooler streams of fluid so that a hotter fluid migrates toward the pressure surface and cooler fluid migrates towards the suction surface. The main purpose of this study is to test the performance of a high-order LES model in terms of predicting this type of highly complicated unsteady flow and heat transfer phenomena. This work describes the performance of a high-order Large Eddy Simulation (LES) turbulent model (developed by the first author) related to the prediction of above mentioned redistribution of inlet temperature distortion in an experimental turbine. Because the understanding of the physical phenomena associated with this temperature redistribution behavior is a very challenging computational fluid dynamic problem. If the numerical method could predict the precisely measured data satisfactorily, then the fluid dynamic variables which are difficult to measure (but obtained as computed results) could be used to visualize the flow characteristics. This technique will also help to get rid off indirect measurement techniques with large measurement uncertainty. In our study emphasis is put to predict the unsteady turbulence characteristics. In this work 3-D unsteady Navier-Stokes analysis of a turbine stage (satisfying the experimental stator-rotor blade ratio) is carried out to study the above mentioned phenomena. The numerical results predicted the experimentally observed phenomena very well. The fact that the streamlines in the stator row remain unaffected was demonstrated by the numerical results. The measured characteristics of the streamline patterns in the rotor row resulted from the secondary flow effect and consequently the inlet temperature distortion effect is also very well predicted.