Speed-up of computing time for numerical analysis of particle charging process by using discrete element method

Abstract

The objective of this paper is to improve the computing time for numerical analysis of particle charging process by using Discrete Element Method. The rule for ignoring the calculations of contact forces and updating trajectories of unmoved particles were discussed. When the relative displacement of a particle within certain calculation steps became less than 0.1% of particle radius, this particle was determined to be unmoved and the calculations of this particle were ignored. The computing time was improved significantly when this new method was used, and its calculation speed was more than two times faster than that of original. It was found that this speed-up method is more useful for the cases that the particle becomes unmoved in short time or the height of charged bed is large. The simulation of charging process in an industrial-scale surge hopper was studied by using new method, the calculation speed became 2.88 times faster than that of original, and the quite similar particle size segregation between original and new methods was given. This new method for speed-up of the charging process in DEM is very useful, and the charging processes of the industrial scale storages can be simulated by using this method.

Discrete particle simulation of solid flow in a three-dimensional blast furnace sector model

“…The application of DEM to BF has been made by some researchers on various aspects, including burden distribution at the top [12][13][14][15], gas-solid flow in BF shaft [16][17][18][19][20][21] and raceway [22][23][24][25], and solid behaviour in the hearth [26,27]. Most of those studies are mainly focused on a particular region, e.g.…”

“…Some studies [17][18][19][20][21][22] use a combined approach of DEM and CFD (computational fluid dynamics) to investigate the influence of gas phase on solid flow in a BF, mainly focused on the raceway where gas-solid interaction is strong. To date, most of DEM applications to BF are mainly on the basis of two-dimensional (2D) slot models [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27]. On the other hand, experimental studies in three-dimensional (3D) models have been carried out (e.g.…”

Scaling of discrete element model parameters for cohesionless and cohesive solid

“…There are several possible solutions (Mio et al, 2009) for the speed-up of DEM simulation, such as optimization of the hardware and the software, including improved DEM algorithm, parallel computing, and simplifying the calculation process. Common ways to simplify the calculation process are done, for example, using a lower spring stiffness, using mono-sized particles, using a cut-off distance for long range forces (Mio et al, 2009) etc.…”

“…There are several possible solutions (Mio et al, 2009) for the speed-up of DEM simulation, such as optimization of the hardware and the software, including improved DEM algorithm, parallel computing, and simplifying the calculation process. Common ways to simplify the calculation process are done, for example, using a lower spring stiffness, using mono-sized particles, using a cut-off distance for long range forces (Mio et al, 2009) etc. Other possibilities are the use of higher particle density in quasi-static simulation (Sheng et al, 2004b) known as density scaling, reduction of number of particles by scaling the system size down or scaling up the size of parti-Paper published in Powder Technology (2015)-use for reference/citation cle.…”

Numerical Investigation of the Transient Multiphase Flow in an Ironmaking Blast Furnace

“…The application of DPS to BF has been made by some researchers on various aspects, including burden distribution at the top, [18][19][20][21] gas-solid flow in BF shaft [22][23][24][25][26] and raceway, [27][28][29][30][31][32][33] and solid behaviour in the hearth. those studies are mainly focused on a particular region, e.g.…”