Long-term efforts have been made to understand loss generation and its reduction in the field of axial turbomachines. The traditional approach to losses for an isolated blade row considers the profile and the secondary losses as a result of viscous flow. The additional kinds of losses in the stage are connected with the shear stress in the mixing process. These losses result from the mixing of the main stream flow with 1) the stator leakage injected through the root axial gap and 2) the return of the tip leakage over the bucket shroud. This article focuses on the first type of mixing losses. The leakage to the main stream flow ratio and the root reaction are the two key parameters investigated in this study. The primary data source for this study is the experiment. An experimental single stage air turbine was modified to set and precisely measure the stator leakage flow. Three configurations of the single-stage test rig with different reaction levels were tested. The second data source for this study is CFD computation. These computations are applied to different geometries and conditions from the experiment; they are derived from real steam turbine stages designed in DSPW. The computations simulate multistage configuration and real steam is considered as the working fluid. CFD computations were performed in the commercial software ANSYS CFX. Each configuration task was computed in three iterative steps. Each step takes the distribution of the flow parameters on the boundary domains from the previous iteration. The final results from this ‘repeating boundary conditions’ approach better correspond with the real expansion in a multistage configuration. The two data sources are not directly comparable. The experiment is used for validation of the trends. The computations provide the possibility of a multi-parametric study. The multi-parametric study is necessary to obtain a more general loss model which can be used during turbine design. The evaluation of the experimental and numerical parts focuses on a comparison of the overall stage performance. Stage efficiency and reaction are presented in relation to the ratio between leakage and main stream flow.

We present a educational poster supporting the subject „Mechanics of fluids I“, which the students evaluate to be difficult mainly due to abstractness. Our goal is to show in vivo the behavior, especially the non-linearity, of various flows transiting into turbulence. The fluid motion is visualized by using the rheoscopic fluid, which consist of water and the dust of mica, whose particles are longitudinal and shiny resulting into easily observable reflections, when the particles coherently orient along the maximum stress. This happens mainly in shear layers, e.g. at the boundary between vortex core and envelope. An example of flow transiting into turbulence is the Taylor-Couette flow between two concentric cylinders, which with increasing Taylor number passes through various regimes from fully laminar bearing flow through the Taylor vortex flow (TVF) and later Wavy vortex flow (WVF) up to Turbulent Taylor vortices regime (TTV) and, finally, the regime of featureless turbulence.

This paper focuses on the influence of shaft labyrinth seal flow on full stage performance. Experimental data are studied, expected design conditions and experimental results are compared and discussed and a losses breakdown for the design procedure is presented. The experimental investigation was performed in VZLU’s air test turbine which is a part of a closed-loop system equipped with a radial compressor. The test turbine configuration simulated the real drum-stage geometry of an axial steam turbine. The geometry of the turbine represents a typical mid-pressure stage of a steam turbine. The configuration of the test rig was adapted in order to easily change the shaft labyrinth seal geometry. The study covered a wide range of seal clearances from very small to extremely large clearances, reaching a maximum relative mass flow approximately 10% of the stator blade flow. Different types of seal feed were also tested to compare internal feed (the flow obtained from the stator flow by the hub-gap just in front of the stator) and external feed realized by additional piping with external regulation. Three stage reactions were tested in this work — Low Reaction, Mid Reaction and Full Reaction. The stator of the stages was the same in all cases, thus the reaction was changed by implementing three different rotor geometries. The influence of the labyrinth seal clearance was investigated by overall performance measurement and by detailed investigation of the flow field. The turbine stage was loaded by a hydraulic dynamometer used for regulating the rotational speed and a flange torquemeter was used to determine the stage efficiency. The total mass flow was measured using an orifice plate. Each seal geometry configuration was calibrated to compute the seal mass flow. The turbine stage and seal were equipped with a number of static pressure taps, and miniature pressure probes were used for measuring the flow field parameters in detail. The discussion of the results is divided into two areas. Firstly, the influence of the degree of reaction on axial steam turbine stage performance in the configuration without the seal flow is presented. Then, a combination of various degrees of reaction is studied as a function of mass flow through the shaft labyrinth seal. The measured data are evaluated by a breakdown of loss sources. The decomposition of the total loss into row losses, leakage losses and mixing losses is highly advantageous. This total loss analysis is carried out for all three stages and both off-design performance and ratios of the shaft seal flow to nozzle blade flow are measured. The post-processing of measured data through this loss breakdown and the comparison with the design is used to validate the design process.