Evaluation of the Interaction Losses in a Transonic Turbine HP Rotor/LP Vane Configuration

Abstract

Transonic turbine rotors produce shock waves, wakes, tip leakage flows, and other secondary flows that the downstream stators have to ingest. While the physics of wake ingestion and shock interaction have been studied quite extensively, few ideas for reducing the aerodynamic interaction losses have been forthcoming. This paper aims to extend previously reported work performed by GE Aircraft Engines in this area. It reports on both average-passage (steady) and unsteady three-dimensional numerical simulations of a candidate design to shed light on the interaction loss mechanisms and evaluate the design. The results from these simulations are first shown against test data for a baseline configuration to engender confidence in the numerical approach. Simulations with the proposed newly designed rotor are then performed to show the trade-offs that are being made in such designs. The new rotor does improve the overall efficiency of the group and physical explanations are presented based on examining entropy production.

Reducing Shock Interactions in Transonic Turbine via Three-Dimensional Aerodynamic Shaping

“…Fortunately for the airfoil designer, improvements in gas-turbine computational fluid dynamics (CFD) have allowed for ever more sophisticated flowfield predictions (e.g., Dunn [21] and Adamczyk [22]), and it is now possible to predict both the time-averaged and time-resolved pressure loadings on transonic airfoils with good accuracy within constraints consistent with appropriate code validation (e.g., Rao et al [23], Busby et al [24], Hilditch et al [25], and many others [26][27][28][29]). Design-optimization systems have previously been used in conjunction with steady-state flow solvers with beneficial effects on transonic turbine airfoils (e.g., Jennions and Adamczyk [30] and Clark et al [20]). Further development of predictive methods such as multigrid techniques and implicit dual time stepping, coupled with the parallelization of codes [31] have created the possibility for using 3D, unsteady Navier-Stokes analyses throughout the design cycle and even within an optimization loop.…”

Analysis and Judgment of the Local Loss Based on the Entropy Gradient

“…This method pay attention on the whole research object, can only get the total loss of the system, can't reflect the local loss or the loss of one point. Another method is local loss method developed in recent years, such as the method proposed by Jennions [1] , VanZante [2] and SUN Yutao [3] ,the former two methods made time and space average to get local loss, local loss characterization is not obvious, the method of SUN requires complex treatment. In the light of this situation this paper deduces a new method--the irreversibility method to express the loss based on the first and second law of thermodynamics, it also can judge whether the reflux happening through the positive or negative of the loss.…”

Blade Row Interaction Effects on the Performance of a Moderately Loaded NASA Transonic Compressor Stage

“…The disadvantages are: 1. CFD codes can give spurious entropy sources and numerical errors (the first law efficiency is also impacted by these errors), 2. entropy production can be computed directly from the viscous dissipation function and heat transfer across a temperature gradient [25] however, the information necessary to compute entropy production directly is not typically available in standard CFD simulation output. As a tool for quickly assessing losses in a design environment the advantages of using irreversibility outweigh the disadvantages.…”