Font Size: a A A

Research On The Computation Of Unsteady Turbulent Flow And The Performance Prediction Of A Francis Turbine

Posted on:2009-12-02Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y HuFull Text:PDF
GTID:1102330332478709Subject:Engineering Mechanics
Abstract/Summary:PDF Full Text Request
The flow in a hydraulic turbine is three-dimensional viscid, incompressible and unsteady turbulent flow. The research on a 3D unsteady turbulent flow is a challenging problem, which the study of the flow field in a turbine has to face. As hydraulic turbines are developed toward large-sized, the problem of vibration is becoming more prominent. The unsteady flow phenomena are one of the main reasons that cause vibrations. Therefore, it is of great importance for research on the unsteady turbulent flow in a turbine. Nowadays, in the field of research and development of a hydraulic turbine, computation fluid dynamics (CFD) has become an indispensable method for analyzing the flow patterns and predicting turbine performance. Turbine characteristics are closely related to the state of flow in a flow field. Only if the flow is well distributed, a turbine has excellent characteristics and performance. In all advanced flow research all over the world, the CFD analysis is applied to calculate and simulate a flow field, hence to predict the efficiency of a runner and the performance of cavitation.The numerical simulation on the unsteady turbulent flow characteristics of Francis turbines flow filed has a great importance for turbine research. This thesis focuses on the unsteady turbulent flow calculations and its analysis. The framework of theoretical studies and methods in this dissertation are as follows:(1) Unsteady turbulent flow simulation in a model Francis turbine runner. Based on the 3D time average N-S equations, the 3D unsteady turbulent flow in a model Francis turbine runner is numerically simulated using realizable k-εturbulent model and CFX-TASCflow software. The calculation has been proven by characteristic tests of the model turbine. The reasonable comparison of the calculation and experiment reveals that the inner flow in the model runner may be truly simulated.(2) The energy and cavitation performances are predicted by means of numerical simulation. Compared with the test results, it shows that the computed efficiency and cavitation coefficient agree with the experiment data. The numerical calculation methods has improved and enriched the hydraulic turbine design.(3) Unsteady turbulent calculation in an entire model Francis turbine. The 3D unsteady turbulent flow in an entire model Francis turbine is simulated using the RNG k-sturbulent model and CFX-TASCflow software. Transient flow fields are simulated in the spiral casing, including the entire stay vane ring, and the draft tube. Based on the standard k-εmodel and RNG k-εmodel, the whirl vortexes in the draft tube are simulated. In comparison with the experimental data, it concludes that the self-stimulated unsteady flow is simulated better with RNG k-εmodel than the standard k-εmodel. The results computed with RNG k-smodel are in good agreement with experiment data. We simulated not only the pressure fluctuation in front of the runner, but also the unsteady vortex in the draft tube by means of the unsteady flow computation in an entire model turbine. Thus, the realistic hydraulic unsteady phenomena occurring in the turbine was predicted. The result shows that this research provides the great value on calculating 3D unsteady turbulent flow in an entire turbine accurately.The results of this study fill up deficiencies on the theoretical research in the studying and developing fields of a hydraulic turbine. These provide significant theoretical guide and technical support to optimize the design of a Francis turbine, to increase the efficiency, to shorten the time of research and development, to accurately predict the hydraulic performance. It will be valuable in engineering application.
Keywords/Search Tags:Francis turbine, Unsteady turbulent flow, CFD, Performance Prediction
PDF Full Text Request
Related items