Optimization of the last stage of gas-steam turbine using a hybrid method

Authors

Paweł Ziółkowski - Gdańsk University of Technology
Łukasz Witanowski - Institute of Fluid-Flow Machinery Polish Academy of Sciences
Piotr Klonowicz - Institute of Fluid-Flow Machinery Polish Academy of Sciences
Stanisław Głuch - Gdańsk University of Technology

Abstract

Research regarding blade design and analysis of flow have been conducted for over a century. Meanwhile new concepts and design approaches were created and improved. Advancements in information technologies allowed to introduce computational fluid dynamics (CFD) and computational flow mechanics (CFM). CFM calculations, presented in previous articles, included a design solution for a novel thermodynamic cycle with three new devices—namely, a wet combustion chamber, a steam-gas turbine and a spray-ejector condenser. In the gas-steam turbine cycle, high temperature occurs in the combustion chamber because of the burning of the fuel in oxygen. As a consequence of the chemical reaction and open water cooling, a mixture of H2O and CO2 is created. The obtaining working medium expands in one turbine, which combines the advantages of gas turbines (high turbine inlet temperatures) and steam turbines (full expansion to vacuum).  Currently a combination of CFM and CFD methods is used for design of turbine blades. This paper relates to the CFD calculation of a new type turbine which is in the phase of theoretical analysis, because the working fluid is a mixture of steam and gas generated in wet combustion chamber. The first, this articles concentrates on a possibility of streamlining the flow efficiency of a last stage of axial turbine working on gas-steam mixture using several optimization algorithms such as a simplex method of Nelder-Mead and a hybrid of particle swarm optimization  with the Nelder-Mead method called HNMPSO. The second, this article aim is redesign and analysis of flow characteristics of the last stage of gas-steam turbine. Values of the maximized objective function, which is the isentropic efficiency of the turbine stage, are found from 3D RANS computation of the flowpath geometry changing during the improving scheme. Among the optimized variables are stator and rotor profile geometric parameters, rotor blade twist angle, circumferential lean and axial sweep angles, as well as parameters characterising the shape of endwall contours within the stator and rotor domain. The presented approach of optimization can significantly reduce the secondary, tip leakage flow losses as well as boundary-layer, separation and leaving energy losses and as a consequence improve flow efficiency of turbine stages.