Emergency gate closing in a Kaplan turbine intake for runaway condition: CFD transient study for two-phase flow and experimental validation
In order to prevent a turbine to reach its runaway speed when load rejection occurs, an emergency closing system must be devised in case the regulation system fails. For Kaplan turbines, fixed wheel gates located in the turbine intake or in the draft tube outlet are usually employed. Gates of this t...
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Articulo |
| Lenguaje: | Inglés |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/127962 |
| Aporte de: |
| Sumario: | In order to prevent a turbine to reach its runaway speed when load rejection occurs, an emergency closing system must be devised in case the regulation system fails. For Kaplan turbines, fixed wheel gates located in the turbine intake or in the draft tube outlet are usually employed. Gates of this type are move by gravity and the closing velocity is controlled by gantry cranes. The closing maneuver is complex due to the high flow rates inherent to runaway conditions and the rotational deceleration during the gate's closing time. Research on this topic is scarce and limited, and numerical studies are usually clouded by uncertainties concerning the setting of proper boundary conditions. In this work, the closing maneuver of the emergency fixed wheel gates at the intake of a Kaplan turbine was studied with CFD two-phase transient simulation. The software used was ANSYS CFX, that solves unsteady Navier-Stokes equations (URANS) by means of the finite volume method. The simulated domain includes a 2D case from one of the span of the semi-spiral casing and a 3D case of a complete span. Two types of simulation were considered, namely: quasi-steady state, where the position of the gate is fixed; and full transient state, where the gate movement was modelled by an immersed solid model. In search of the optimum model layout with its set of boundary conditions, numerical results were compared and validated against experiments performed on a physical scale model in accordance with IEC 60193 norms for several turbulence models. Results show that the pull-up force on the gate increases as it is being closed. Analysis of pressure fluctuation at different points of the gate suggest that the main frequency component is the vortex shedding of the gate lip. |
|---|