Optimization of the transition piece applying genetic algorithms

Abstract

Abstract The geometric optimization of the gas turbine transition piece, applying the Computational Fluid Dynamics (CFD) and Genetic Algorithms (GA), to obtain a uniform temperature and velocity distribution in the center of the outlet section, is presented. The study is carried out considering different dimensions of the transition piece which were defined by the genetic algorithm analysis. In each case, the temperature and velocity contours of the combustion gases are examined along the transition piece and outlet section, comparing the radial and circumferential velocity and temperature profiles. In this study, it is important to know the temperature and velocity profiles at the outlet of the transition piece, because they have impact on the vanes and blades of the first stage of the gas turbine. The study considers a steady state three-dimensional model, using FLUENT® version 6.3.26 and a code of genetic algorithms where the geometric parameters were analyzed. To obtain the results, a structured grid about 5.1 million cells with second-order upwind scheme and coupled solver was applied. The results show the effect of the geometry of the transition piece in the velocity and temperature profiles at the outlet area, where the change of the cross sectional area influences the contours. A smooth change in the cross sectional section allows a uniform temperature and velocity distribution at the outlet section of the transition piece. The peak value of the temperature is decreased about 5.5% (11.4 K) in the radial direction whereas the velocity is decreased 3.62% (4.27 m/s). The peak value of the temperature in the circumferential direction is decreased about 3.33% (45.51 K) whereas velocity is decreased 3.63% (4.171 m/s). The new geometry of the transition piece leads to a diminution of the average turbine inlet temperature (TIT) of about 2.32% (28.29 K) and the average velocity of about 7.73% (4.217 m/s).