Multi-objective shape optimization of runner blade for Kaplan turbine

27th IAHR Symposium on Hydraulic Machinery and Systems

Semenova, A.^1 ; Chirkov, D.^2 ; Lyutov, A.^2 ; Chemy, S.^2 ; Skorospelov, V.^3 ; Pylev, I.^1

OJSC Power Machines, LMZ, Saint Petersburg, Russia^1

Institute of Computational Technologies, SB, RAS, Novosibirsk, Russia^2

Institute of Mathematics, SB, RAS, Novosibirsk, Russia^3

关键词: High specific speed;    Hydraulic turbomachineries;    Multi-objective genetic algorithm;    Optimization problems;    Optimization techniques;    Three-dimensional turbulent flow computations;    Turbine efficiency;    Velocity profiles;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/22/1/012025/pdf
DOI  :  10.1088/1755-1315/22/1/012025

来源: IOP

【 摘 要 】

Automatic runner shape optimization based on extensive CFD analysis proved to be a useful design tool in hydraulic turbomachinery. Previously the authors developed an efficient method for Francis runner optimization. It was successfully applied to the design of several runners with different specific speeds. In present work this method is extended to the task of a Kaplan runner optimization. Despite of relatively simpler blade shape, Kaplan turbines have several features, complicating the optimization problem. First, Kaplan turbines normally operate in a wide range of discharges, thus CFD analysis of each variant of the runner should be carried out for several operation points. Next, due to a high specific speed, draft tube losses have a great impact on the overall turbine efficiency, and thus should be accurately evaluated. Then, the flow in blade tip and hub clearances significantly affects the velocity profile behind the runner and draft tube behavior. All these features are accounted in the present optimization technique. Parameterization of runner blade surface using 24 geometrical parameters is described in details. For each variant of runner geometry steady state three-dimensional turbulent flow computations are carried out in the domain, including wicket gate, runner, draft tube, blade tip and hub clearances. The objectives are maximization of efficiency in best efficiency and high discharge operation points, with simultaneous minimization of cavitation area on the suction side of the blade. Multiobjective genetic algorithm is used for the solution of optimization problem, requiring the analysis of several thousands of runner variants. The method is applied to optimization of runner shape for several Kaplan turbines with different heads.