Simulations numériques de l'écoulement turbulent dans un aspirateur de turbine hydraulique

Authors: Beaubien, Carl-Anthony
Advisor: Dumas, Guy
Abstract: The work carried throughout this thesis has for objective to enhance losses predictions in hydraulic turbines draft tube. In order to acheive this, the flow in a draft tube charaterized by a sharp drop in the pressure recovery coefficient near the best efficiency point was studied. Detached Eddy Simulation (DES), an advanced turbulence modeling approach, was put to the test, in order to asses the gain attributable to a finer and more precise description of turbulent motions in this component. The numerical methods required associated to this approach, especially regarding the inlet boundary condition, were investigated. It was shown that the radial velocity profile specified at the inlet of the computational domain alters significantly the flow downstream and the predicted performance. With the measured radial velocity profile specified at the inlet of the draft tube, reasonnable agreement was found between URANS numerical results and experimental measurements of pressure recovery. However, some aspects of the numerical simulations does not agree well with experimental data. It is the case for flow imbalance between the two outlet bays. It was established that rotating flow structures underneath the runner blades require extremely fine grid and time step resolution to avoid their premature diffusion underneath the inlet plane. Nevertheless, at the studied operating point, their influence on draft tube performance was found to be very limited. DES and URANS simulations of the draft tube where axisymmetric inlet boundary conditions were imposed predicted similar pressure recovery. However, DES enables to simulate much more complex and rich turbulent motions, at a computational cost similar to the one of a URANS simulatation and with much less influence from the modeled turbulent quantities specified at the inlet plane.
Document Type: Mémoire de maîtrise
Issue Date: 2013
Open Access Date: 19 April 2018
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
29942.pdfTexte46.99 MBAdobe PDFThumbnail
All documents in CorpusUL are protected by Copyright Act of Canada.