Optimization of the workflow of multistage axial turbines with platforms

The turbine workflow is described by a large number of variables that often have contradictory impact on the turbine operating parameters. Hand-operated variation of these variables is a difficult time-consuming task. Application of mathematic optimization techniques could provide a solution of the problem. The paper outlines the basic principles of the method of mathematical optimization of an axial turbine based on the joint use of the NUMECA computational fluid dynamics software and the IOSO program optimizer. The procedure developed was implemented for a four-stage turbine developed and tested by NASA. Design models including one, two, three and four stages have been developed for the turbine under consideration. The results of calculations performed with their help showed good agreement with the experimental results. The settings of the numerical model that provide good agreement of design and experimental data but do not require large computational resource have been chosen according to the results of the research conducted. Low computational cost is very important because it allows significant reduction of the time spent to obtain the optimal solution with repeated reference to the design model. Automatic search of the turbine blade configuration allowing a 0.7 % increase of the turbine efficiency with constant mass flow and rate of expansion of gas (accuracy 0.5 %) was carried out with the help of the model created.