Please use this identifier to cite or link to this item: https://scidar.kg.ac.rs/handle/123456789/17716
Title: BOUNDARY IDENTIFICATION AND WEAK PERIODIC CONDITION APPLICATION IN DEM METHOD
Authors: Zivkovic, Miroslav
Topalović, Marko
Blagojevic, Milan
Aleksandar, Nikolic
Milovanović, Vladimir
Mesarovic, Sinisa
Padbidri, Jagan
Issue Date: 2014
Abstract: Meshless character of the DEM method brings forth certain problems that must be solved in a computer implementation, such as the determination of the neighbouring particles, determining the boundaries of the model and the definition of boundary conditions. During large deformations of the model a particular internal particle may become boundary particle and the other way round so it is necessary to re-determine the boundary layer after a given number of steps in the analysis. In this paper, two approaches for defining the boundary layer are shown, whose effectiveness depends on the current positions of the particles, and imperfections in their arrangement. Due to the large computer requirements of the DEM method, only a small part of the real problem is modelled and continuity can be achieved by using periodic boundary conditions. Since the DEM particles have arbitrary radius and the boundary is often irregular, it is not possible to achieve the classic periodic boundary conditions in which the particle that leaves the model on one side enters the model on the other side, but it is possible to apply a soft periodic boundary conditions that transmit only the force from one to another end of the model. Previously mentioned routines are applied to the modelling of shear in the critical layer of granular material.
URI: https://scidar.kg.ac.rs/handle/123456789/17716
Type: conferenceObject
Appears in Collections:Faculty of Engineering, Kragujevac

Page views(s)

390

Downloads(s)

39

Files in This Item:
File Description SizeFormat 
M33-12.pdf13.84 MBAdobe PDFThumbnail
View/Open


Items in SCIDAR are protected by copyright, with all rights reserved, unless otherwise indicated.