Experience with Application of Laser Diagnostics Systems in Experimental Studies of Wet Steam Flows in the Flow Path of Turbomachines: Basic Results of Studies and System Development Trends

The findings in application of laser diagnostic systems for the investigation of wet steam flows in flow paths of steam turbines, specifically, in channels of various configurations, are reviewed. The experimental results accumulated over more than a decade enabled the authors to generalize and formulate the essential features of the movement of coarse erosion-hazardous droplets downstream of turbine cascades and in interblade channels. The regions with liquid phase particles are found using the data of visual analysis. The results of application of the particle image velocimetry/particle tracking velocity (PIV/PTV) methods, which implement algorithms for determining vector fields of particle velocities, are presented. The features of the distribution of kinematic characteristics of the liquid phase in various regions of turbine cascades, which affect the erosion wear rate, are discussed. Data are presented on the effect of the blade profile on the formation of trajectories of droplet flows, and methods for improving the separation ability of a blade operating in a wet stem flow are proposed. Using the experimental database on liquid phase velocity fields in turbine channels of various configurations, a semiempirical model of the flow of large erosive-dangerous droplets is formulated. It describes their movement in a cocurrent steam flow and interaction with the walls of the interblade channels. A review is presented of engineering solutions that were obtained on the basis of theoretical and experimental studies of wet steam flows using laser diagnostic systems. The concept of blade surface heating, heating steam jet injection, and optimization of the in-channel separation system is examined. The principles are formulated for the development of approaches based on laser flow diagnostic systems in the field of application of neural networks, which should considerably extend the capabilities of experimental studies since they offer the potential for increasing the amount of data obtained by implementing such experimental methods.