2002 Fiscal Year Final Research Report Summary
Development of parallel computational method for granular mixtures transported by turbulent flows
Project/Area Number |
13650568
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
水工水理学
|
Research Institution | Kyoto university |
Principal Investigator |
USHIJIMA Satoru Kyoto University, Faculty of Engineering, associate professor, 工学研究科, 助教授 (70324655)
|
Co-Investigator(Kenkyū-buntansha) |
ONITSUKA Kouki Kyusyu Institute of Technology, Faculty of Engineering, assistant, 工学部, 助手 (20293904)
|
Project Period (FY) |
2001 – 2002
|
Keywords | parallel computation / granular mixture / segregation / turbulent transportation / particle pathline / numerical visualization |
Research Abstract |
It is important to understand the physical process on the segregation of the non-uniform bed-load particles, which consist of non-uniform diameters and densities. The purpose of this study is to develop a computational method to predict turbulence transportation of the non-uniform particles and their segregation. The key techniques of this study are distinct element method (DEM) to treat contact forces among particles, liquid-solid two-way turbulence model and a parallel computation method to deal with large number of particles The developed computational method is characterized by the following items : (1) The employment of DEM allows us to predict 3D translational and rotational motions of particles as well as. their contacts. (2) The turbulent liquid flows including solid. particles are modeled with a liquid-solid turbulence model. (3) The interaction between liquid and solid phases is taken into account in the liquid-solid turbulence model. (4) A parallel computational method based on a domain decomposition method enables us to deal with a large number of particles. (5) The curvilinear coordinates are utilized to represent the geometry of each subdomain and it is applicable to complicated-shaped boundaries. (6) The numerical visualization methods, such as a particle pathline techniques, allows us to understand the flow patterns of each component of particles. The developed computational method was applied various problems, such as a granular mixture in a rotating cylinder, non-uniform particles under oscillating liquid flows, particle segregation arising in front of hydraulics structures Generally, it was confirmed that the predicted particle distributions agree with experimental results and that computational efficiency is satisfactory due to the parallel computation.
|
Research Products
(15 results)