Library Subscription: Guest

ISSN Online: 2688-7231

ISBN Flash Drive: 978-1-56700-497-7

ISBN Online: 978-1-56700-496-0

Proceedings of the 25th National and 3rd International ISHMT-ASTFE Heat and Mass Transfer Conference (IHMTC-2019)
December, 28-31, 2019, IIT Roorkee, Roorkee, India

POD analysis of turbulent swirling flow in draft tube of a high-head Francis turbine model at part load operation

Get access (open in a dialog) DOI: 10.1615/IHMTC-2019.1580
pages 941-946

Abstract

A reaction turbine when operates at part load operating condition, the flow field inside its draft tube changes considerably. A large scale vortical structure, i.e., rotating vortex rope (RVR), is formed as the result of the adverse pressure gradient and the runner outlet swirling flow. Even with the use of advanced experimental instruments like particle image velocimetry (PIV) and its data processing tools, the information available about these flow instabilities is still incomplete. This paper presents the application of proper orthogonal decomposition (POD) method on the flow field inside the draft tube cone at part load (PL) operation of high-head model Francis turbine. The POD analysis is carried out on 250 PIV snapshots containing the axial and radial velocities. The investigation of a vortex rope structure has been attempted via decomposed POD modes. The instantaneous velocity field has been reconstructed using the eigenfunction values and POD coefficients. The results show that the first eight modes contain more than 95% of the total kinetic energy (KE) of the flow field and are associated with the organized motion of the flow. The first mode contains more than 50% of the total energy and the axial velocity profile reconstructed with the first mode is identical to the actual mean axial velocity flow field. The direction of flow in modes 2 is opposite to the runner rotational direction, and mode 3 reveals the organized symmetrical vortical motion in pairs. The contribution of mode 2 and 3 is found around 39% of the total energy. The energy of modes 4 and 5 is very close to each other and contributes more than 5% of the total energy. The modes 9 to 250 are associated with randomness of the flow.