Shape optimization of hydraulic turbine diffuser under swirling flow conditions to mitigate vortex rope instability

Shape optimization of hydraulic turbine diffuser under swirling flow conditions to mitigate vortex rope instability

Paper in proceedings (conference paper)

Vortex rope is a typical undesired phenomenon occuring in draft tube of Francis turbines operated under off-design operating conditions. It is characterized by pressure pulsations (synchronous and asynchronous), which influence dynamics of the hydraulic turbine and often cause limiting of the turbine operation. Because vortex rope is result of flow instability, there is a chance to influence the instability onset by modification of boundary conditions. Present contribution is focusing on mitigation by proper shaping of the draft tube walls. The approach is based on automatic procedure using coupling of CFD simulations and mathematical optimization methods. Efficient, yet robust, combination of stochastic optimization PSO algorithm with direct search Nelder-Mead method makes up sufficiently global, but relatively fast optimization algorithm. To make the whole optimization loop feasible, only 2d axisymmetric domain is assumed and the goal is minimization of the backflow region, which is an indivisible feature of the vortex breakdown instability.

Vortex rope is a typical undesired phenomenon occuring in draft tube of Francis turbines operated under off-design operating conditions. It is characterized by pressure pulsations (synchronous and asynchronous), which influence dynamics of the hydraulic turbine and often cause limiting of the turbine operation. Because vortex rope is result of flow instability, there is a chance to influence the instability onset by modification of boundary conditions. Present contribution is focusing on mitigation by proper shaping of the draft tube walls. The approach is based on automatic procedure using coupling of CFD simulations and mathematical optimization methods. Efficient, yet robust, combination of stochastic optimization PSO algorithm with direct search Nelder-Mead method makes up sufficiently global, but relatively fast optimization algorithm. To make the whole optimization loop feasible, only 2d axisymmetric domain is assumed and the goal is minimization of the backflow region, which is an indivisible feature of the vortex breakdown instability.

vortex rope; draft tube; francis turbine;optimization;PSO;CFD

vortex rope; draft tube; francis turbine;optimization;PSO;CFD

RUDOLF, P.; OBERTA, B.; ŠTEFAN, D.

07.04.2025

 Institute of Physics

IOP Conference Series: Earth and Environmental Science

1755-1307

IOP Conference Series: Earth and Environmental Science

c

1

United Kingdom of Great Britain and Northern Ireland

1

7

7

https://iopscience.iop.org/article/10.1088/1755-1315/1483/1/012040