Cascade flow simulation and measurement for the study of axial compressor loss mechanism.

摘要

The flow field of axial-flow turbomachines, such as compressors and turbines, can be characterized as complex and three-dimensional (3-D) in nature. The accurate prediction of stage efficiency is still one of the most challenging engineering problems in turbomachinery design, due primarily to the complexity of the associated flow field. Historically, major improvements in turbomachinery component efficiency have come at a cost since these turbomachinery units have often been tested on expensive full-scale rotating prototype models. The data obtained from these full-scale tests can be, on the other hand, quite difficult to analyze. These full-scale test data usually provide good empirical information, but are quite often difficult to use for the analytical prediction of design efficiency. An alternative approach for the turbomachinery flow field analysis is to use a “cascade” model. Cascade is a simplified planar model representing a turbomachinery main flow passage at a specified spanwise location in the blade airfoil. Using a cascade model, detailed flow field characteristics in two-dimensional (2-D) blade-to-blade flow of the turbomachinery main flow passage can be investigated by deliberately excluding the complex 3-D viscous flow phenomena. Utilizing experimental measurement from a newly developed water table cascade model as well as cascade numerical simulation of computational fluid dynamics (CFD), a systematic and detailed investigation into axial compressor loss mechanism has been conducted. In order to accurately predict the stage efficiency of 3-D turbomachines, it is quite important to understand the details of associated physical mechanism of loss in 2-D cascade models. Traditionally, turbomachinery losses in both cascade and actual turbomachines have often been quantified by the drop in total (or stagnation) pressure across the stage. This quantification has long been known as an inconvenient approach for design, since the coefficient of loss based on this “pressure loss” depends on Mach number. The more universal quantification of “energy loss” has recently received much research attention, and it is the major focus in the present study. In this relatively new concept, turbomachinery losses are quantified by “generation of entropy” due to irreversible thermodynamic processes associated with viscous flow phenomena. The advantage of this approach is that the loss coefficient based on entropy generation is independent of Mach number. (Abstract shortened by UMI.)