Abstract
The numerical methods for iterative solving of discretized governing equations often require special treatment for the purpose of achieving not only sufficiently accurate and reliable results, but stable and gradual convergence of the solution too. The general remedy for such challenge, for a certain case, is to use a fine mesh to a certain level and/or to slow down the numerical procedure, a two useful strategies by which numerical instabilities will be avoided on the account of a greater CPU load. This paper presents the employment of these two strategies by conducting a grid dependency analysis for a 2D model of the stay and guide vanes of a hydraulic Francis turbine and furthering the solution to iteration procedure adjustment for a 3D representation of the same model. The ultimate accent is placed on how to deal with a particular numerical instability problem in a pure mathematical fashion without getting into the experimental validation of the results and calibration of the method. (Author)
Iliev, Igor;
Markov, Zoran
[1]
- Faculty of Mechanical Engineering, 'Ss. Cyril and Methodius' University, Skopje (Macedonia, The Former Yugoslav Republic of)
Citation Formats
Iliev, Igor, and Markov, Zoran.
Grid dependency and relaxation of an iteration procedure for flow calculations in stationary hydraulic turbine parts.
Macedonia, The Former Yugoslav Republic of: N. p.,
2014.
Web.
Iliev, Igor, & Markov, Zoran.
Grid dependency and relaxation of an iteration procedure for flow calculations in stationary hydraulic turbine parts.
Macedonia, The Former Yugoslav Republic of.
Iliev, Igor, and Markov, Zoran.
2014.
"Grid dependency and relaxation of an iteration procedure for flow calculations in stationary hydraulic turbine parts."
Macedonia, The Former Yugoslav Republic of.
@misc{etde_22390252,
title = {Grid dependency and relaxation of an iteration procedure for flow calculations in stationary hydraulic turbine parts}
author = {Iliev, Igor, and Markov, Zoran}
abstractNote = {The numerical methods for iterative solving of discretized governing equations often require special treatment for the purpose of achieving not only sufficiently accurate and reliable results, but stable and gradual convergence of the solution too. The general remedy for such challenge, for a certain case, is to use a fine mesh to a certain level and/or to slow down the numerical procedure, a two useful strategies by which numerical instabilities will be avoided on the account of a greater CPU load. This paper presents the employment of these two strategies by conducting a grid dependency analysis for a 2D model of the stay and guide vanes of a hydraulic Francis turbine and furthering the solution to iteration procedure adjustment for a 3D representation of the same model. The ultimate accent is placed on how to deal with a particular numerical instability problem in a pure mathematical fashion without getting into the experimental validation of the results and calibration of the method. (Author)}
journal = []
issue = {2}
volume = {32}
journal type = {AC}
place = {Macedonia, The Former Yugoslav Republic of}
year = {2014}
month = {Jul}
}
title = {Grid dependency and relaxation of an iteration procedure for flow calculations in stationary hydraulic turbine parts}
author = {Iliev, Igor, and Markov, Zoran}
abstractNote = {The numerical methods for iterative solving of discretized governing equations often require special treatment for the purpose of achieving not only sufficiently accurate and reliable results, but stable and gradual convergence of the solution too. The general remedy for such challenge, for a certain case, is to use a fine mesh to a certain level and/or to slow down the numerical procedure, a two useful strategies by which numerical instabilities will be avoided on the account of a greater CPU load. This paper presents the employment of these two strategies by conducting a grid dependency analysis for a 2D model of the stay and guide vanes of a hydraulic Francis turbine and furthering the solution to iteration procedure adjustment for a 3D representation of the same model. The ultimate accent is placed on how to deal with a particular numerical instability problem in a pure mathematical fashion without getting into the experimental validation of the results and calibration of the method. (Author)}
journal = []
issue = {2}
volume = {32}
journal type = {AC}
place = {Macedonia, The Former Yugoslav Republic of}
year = {2014}
month = {Jul}
}