Validation of the generalized coarse-graining model in multi-particle simulations

doi:10.11949/0438-1157.20250338

Abstract

Abstract:

Coarse-grained models are widely used to accelerate discrete element method simulations. For polydisperse particle systems, the number of particles increases with the cube of the particle size ratio. Traditional coarse-grained models cannot reduce the huge computational cost caused by the increase in particle size ratio because they use the same scale to scale particles of different sizes. To solve this problem, a variable-scale generalized coarse-grained model is proposed based on the consistency of the dimensionless contact equation, includes a new model named the constant surface energy (CSE) model, alongside the existing constant relative overlap (CRO) and constant absolute overlap (CAO) models. The applicability of the variable-scale generalized coarse-grained model in multi-particle simulations is verified through the angle of repose and uniaxial compression process. The CRO and combined CSE-CRO models show better consistency with the original bed using uniform and variable scaling ratios, respectively. The results show that the CSE-CRO model can reduce the calculation time by about 80% when the small particle scaling ratio is 2, and the mean relative error is less than 2%.

Key words: coarse-grained model, discrete element method, variable scale, angle of repose, uniaxial compression, particle, particle size distribution, simulation

摘要：

粗粒化模型被广泛用于加速离散元方法模拟。对于多分散颗粒系统，颗粒数量随颗粒粒径比的三次方增加。传统粗粒化模型由于使用相同的比例缩放各粒径颗粒，无法减少由于粒径比增大导致的巨大计算成本。为解决该问题，基于无量纲接触方程的一致性，提出了变比例广义粗粒化方法，通过引入“N对1接触”假设，构建了更符合物理实际的变比例接触模型，并通过调整等效杨氏模量减小几何误差。在变比例广义粗粒化方法的基础上，通过休止角和单轴压缩过程验证其在多颗粒场景下的适用性。结果显示，该方法在大颗粒粒径不变、小颗粒放大两倍时，可以减少计算时间约80%，相对误差均值低于2%。

关键词: 粗粒化模型, 离散元方法, 变比例, 休止角, 单轴压缩, 粒子, 粒度分布, 模拟

CLC Number:

TQ 015.9

Yanwei FANG, Guanqing LIU, Yiyang ZHANG, Zepeng ZHU, Zhu FANG, Shuiqing LI. Validation of the generalized coarse-graining model in multi-particle simulations[J]. CIESC Journal, 2025, 76(11): 5630-5644.

方延玮, 柳冠青, 张易阳, 朱泽鹏, 方筑, 李水清. 变比例广义粗粒化方法的多颗粒场景验证[J]. 化工学报, 2025, 76(11): 5630-5644.

Figures/Tables 18

Table 1 Scaling criteria for the generalized coarse-grained model

广义粗粒化模型特例	$χ$	$l v$	$l ω$	$l γ$	$l E$	$K M$	$l μ S$	$l θ R$	$l μ T$	$l t J K R$
CSE	2	$1$	$l R - 1$	$1$	$l R - 2.5$	$l R - 1$	$1$	$1$	$l R - 0.5$	$l R 2$
CRO	1	$1$	$l R - 1$	$l R$	$1$	$1$	$1$	$1$	$1$	$l R$
CAO	0	$1$	$l R - 1$	$l R 2$	$l R 2.5$	$l R$	$1$	$1$	$l R 0.5$	$1$

Table 1 Scaling criteria for the generalized coarse-grained model

广义粗粒化模型特例	$χ$	$l v$	$l ω$	$l γ$	$l E$	$K M$	$l μ S$	$l θ R$	$l μ T$	$l t J K R$
CSE	2	$1$	$l R - 1$	$1$	$l R - 2.5$	$l R - 1$	$1$	$1$	$l R - 0.5$	$l R 2$
CRO	1	$1$	$l R - 1$	$l R$	$1$	$1$	$1$	$1$	$1$	$l R$
CAO	0	$1$	$l R - 1$	$l R 2$	$l R 2.5$	$l R$	$1$	$1$	$l R 0.5$	$1$

Fig.1 Schematic diagram of repose angle simulation (solid circles represent original particles, particle color represents particle type, and solid arrows represent velocity vectors)

Table 2 Simulation parameters of the repose angle process

模拟参数	模拟参数取值
模拟参数	相同比例缩放	变比例缩放
原始颗粒直径 $d$ /mm	$d 2 = 5$
原始颗粒粒径比 $β p$	$β p = 1,2$	$β p = 5$
粗粒比 $l r$	$l r = 1,1.5,2$	$l r, 1 = 1,1.5,2, 2.5; l r, 2 = 1$
密度 $ρ$ /(kg/m³)	$ρ p = ρ w = 3000$
杨氏模量 $E$ /GPa	$E p = E w = 0.03$
泊松比 $σ$	$σ p = σ w = 0.3$
表面能 $γ$ /(mJ/m²)	$γ p - p = γ p - w = 50$
碰撞恢复系数 $e$	$e p - p = e p - w = 0.5$
滑动摩擦因数 $μ S$	$μ S, p - p = μ S, p - w = 0.4$
临界滚动角 $θ R$ /rad	$θ R, p - p = θ R, p - w = 0.1$	$θ R, p - p = θ R, p - w = 0.05,0.1,0.2$
扭转摩擦因数 $μ T$	$μ T, p - p = μ T, p - w = 0.4$
小颗粒的体积分数/%	$Φ 1 = 25$	$Φ 1 = 2,4, 8$

Table 2 Simulation parameters of the repose angle process

模拟参数	模拟参数取值
模拟参数	相同比例缩放	变比例缩放
原始颗粒直径 $d$ /mm	$d 2 = 5$
原始颗粒粒径比 $β p$	$β p = 1,2$	$β p = 5$
粗粒比 $l r$	$l r = 1,1.5,2$	$l r, 1 = 1,1.5,2, 2.5; l r, 2 = 1$
密度 $ρ$ /(kg/m³)	$ρ p = ρ w = 3000$
杨氏模量 $E$ /GPa	$E p = E w = 0.03$
泊松比 $σ$	$σ p = σ w = 0.3$
表面能 $γ$ /(mJ/m²)	$γ p - p = γ p - w = 50$
碰撞恢复系数 $e$	$e p - p = e p - w = 0.5$
滑动摩擦因数 $μ S$	$μ S, p - p = μ S, p - w = 0.4$
临界滚动角 $θ R$ /rad	$θ R, p - p = θ R, p - w = 0.1$	$θ R, p - p = θ R, p - w = 0.05,0.1,0.2$
扭转摩擦因数 $μ T$	$μ T, p - p = μ T, p - w = 0.4$
小颗粒的体积分数/%	$Φ 1 = 25$	$Φ 1 = 2,4, 8$

Fig.2 Particle piles of the original particles (the yellow dashed line is the position of the cylinder wall, and the color indicates the particle coordination number): (a) βp=1 (monodisperse); (b) βp=2

Fig.3 Monodisperse particle piles with the same coarse-grained ratio (βp=1，lr=2, color represents particle type, red is original particles, blue is coarse-grained particles; the yellow dotted line is the position of the cylinder wall): (a) original particles; (b) CSE model; (c) CRO model; (d) CAO model

Fig.4 Bidisperse particle piles with the same coarse-grained ratio (βp=2,lr=2, red is original particles, blue is coarse-grained particles, the yellow dotted line is the position of the cylinder wall): (a) original particles; (b) CSE model; (c) CRO model; (d) CAO model

Fig.5 Schematic diagram of relative deflection error

Fig.6 The angle of repose and its relative error for the same coarse-grained ratio

Fig.7 Bidisperse particle piles with different coarse-grained ratios (βp=5,Φ1=8%,θR=0.1,lr,1=2.5,lr,2=1, red is original particles, blue is coarse-grained particles; the yellow dotted line is the position of the cylinder wall): (a) original particles; (b) CSE-CRO model; (c) CRO-CRO model; (d) CAO-CRO model

Fig.8 The repose angle and relative error of different coarse-grain ratios

Fig.9 Performance of the variable-scale generalized coarse-grained model in simulating the angle of repose process

Fig.10 Schematic diagram of uniaxial compression simulation (solid circles represent original particles, particle color represents particle type, and solid arrows represent velocity vectors)

Table 3 Simulation parameters of the uniaxial compression process

模拟参数	模拟参数取值
模拟参数	相同比例缩放	变比例缩放
原始颗粒直径 $d$ /mm	$d 2 = 2$
原始颗粒粒径比 $β p$	$β p = 1,2, 3$	$β p = 10$
粗粒比 $l r$	$l r = 1,1.5,2$	$l r, 1 = 1,2, 4,6; l r, 2 = 1$
密度 $ρ$ /(kg/m³)	$ρ p = ρ w = 3000$
杨氏模量 $E$ /GPa	$E p = E w = 0.1$
泊松比 $σ$	$σ p = σ w = 0.3$
表面能 $γ$ /(mJ/m²)	$γ p - p = γ p - w = 50$	$γ p - p = γ p - w = 5,50,500$
碰撞恢复系数 $e$	$e p - p = e p - w = 0.5$
滑动摩擦系数 $μ S$	$μ S, p - p = μ S, p - w = 0.3$
临界滚动角 $θ R$ /rad	$θ R, p - p = θ R, p - w = 0.01$
扭转摩擦系数 $μ T$	$μ T, p - p = μ T, p - w = 0.3$
小颗粒的体积分数/%	$Φ 1 = 50$	$Φ 1 = 2,4, 8$

Table 3 Simulation parameters of the uniaxial compression process

模拟参数	模拟参数取值
模拟参数	相同比例缩放	变比例缩放
原始颗粒直径 $d$ /mm	$d 2 = 2$
原始颗粒粒径比 $β p$	$β p = 1,2, 3$	$β p = 10$
粗粒比 $l r$	$l r = 1,1.5,2$	$l r, 1 = 1,2, 4,6; l r, 2 = 1$
密度 $ρ$ /(kg/m³)	$ρ p = ρ w = 3000$
杨氏模量 $E$ /GPa	$E p = E w = 0.1$
泊松比 $σ$	$σ p = σ w = 0.3$
表面能 $γ$ /(mJ/m²)	$γ p - p = γ p - w = 50$	$γ p - p = γ p - w = 5,50,500$
碰撞恢复系数 $e$	$e p - p = e p - w = 0.5$
滑动摩擦系数 $μ S$	$μ S, p - p = μ S, p - w = 0.3$
临界滚动角 $θ R$ /rad	$θ R, p - p = θ R, p - w = 0.01$
扭转摩擦系数 $μ T$	$μ T, p - p = μ T, p - w = 0.3$
小颗粒的体积分数/%	$Φ 1 = 50$	$Φ 1 = 2,4, 8$

Fig.11 Uniaxial compression process of the original particle bed: (a) stress-strain curve; (b) volume fraction-stress curve

Fig.12 Uniaxial compression process with the same coarse-grain ratio

Fig.13 Uniaxial compression characteristic parameters and their relative errors for the same coarse-grained ratio

Fig.14 Compression process of coarse-grained particle beds with different coarse-grain ratios

Fig.15 Performance of the variable-scale generalized coarse-grained model in simulating uniaxial compression

References 46

[1]	Gong M, Azadi S, Gans A, et al. Erosion of a cohesive granular material by an impinging turbulent jet[J]. EPJ Web of Conferences, 2021, 249: 08011.
[2]	Metzger P T. Erosion rate of lunar soil under a landing rocket(part 2): Benchmarking and predictions[J]. Icarus, 2024, 417: 116135.
[3]	LaMarche C Q, Curtis J S. Cratering of a particle bed by a subsonic turbulent jet: effect of particle shape, size and density[J]. Chemical Engineering Science, 2015, 138: 432-445.
[4]	Dotson B, Valencia D S, Millwater C, et al. Cohesion and shear strength of compacted lunar and Martian regolith simulants[J]. Icarus, 2024, 411: 115943.
[5]	LaMarche C Q, Curtis J S, Metzger P T. Cratering of a lunar soil simulant, JSC-1A, by a turbulent subsonic jet[C]//Earth and Space. 2012: 36-44.
[6]	Zhang S, Ge W. Accelerating discrete particle simulation of particle-fluid systems[J]. Current Opinion in Chemical Engineering, 2024, 43: 100989.
[7]	Brandt V, Grabowski J, Jurtz N, et al. DEM and DEM-CFD modeling of systems with geometric constrictions using a new particle location based multi-level coarse graining approach[J]. Powder Technology, 2024, 436: 119447.
[8]	Kishida N, Nakamura H, Ohsaki S, et al. Surrogate model of DEM simulation for binary-sized particle mixing and segregation[J]. Powder Technology, 2025, 455: 120811.
[9]	Kishida N, Nakamura H, Ohsaki S, et al. Optimizing data-sampling period in a machine learning-based surrogate model for powder mixing simulations[J]. Powder Technology, 2025, 452: 120584.
[10]	Chu K W, Chen J, Yu A B. Applicability of a coarse-grained CFD-DEM model on dense medium cyclone[J]. Minerals Engineering, 2016, 90: 43-54.
[11]	Di Renzo A, Napolitano E, Di Maio F. Coarse-grain DEM modelling in fluidized bed simulation: a review[J]. Processes, 2021, 9(2): 279.
[12]	Sakai M, Koshizuka S. Large-scale discrete element modeling in pneumatic conveying[J]. Chemical Engineering Science, 2009, 64(3): 533-539.
[13]	Hilton J E, Cleary P W. Comparison of non-cohesive resolved and coarse grain DEM models for gas flow through particle beds[J]. Applied Mathematical Modelling, 2014, 38(17/18): 4197-4214.
[14]	Bierwisch C, Kraft T, Riedel H, et al. Three-dimensional discrete element models for the granular statics and dynamics of powders in cavity filling[J]. Journal of the Mechanics and Physics of Solids, 2009, 57(1): 10-31.
[15]	Radl S, Radeke C, Khinast J G, et al. Parcel-based approach for the simulation of gas-particle flows[C]//8th International Conference on CFD in Oil & Gas, Metallurgical and Process Industries, Trondheim. 2011, 23: 1084-1098.
[16]	Fang Y, Liu G, Zhang Y, et al. A generalized coarse-graining discrete-element method with variable scaling ratios based on non-dimensional contact equation[J]. Powder Technology, 2025, 452: 120569.
[17]	Qin Z Y, Zhou Q, Wang J W. An EMMS drag model for coarse grid simulation of polydisperse gas-solid flow in circulating fluidized bed risers[J]. Chemical Engineering Science, 2019, 207: 358-378.
[18]	Lu L, Xu J, Ge W, et al. EMMS-based discrete particle method (EMMS-DPM) for simulation of gas-solid flows[J]. Chemical Engineering Science, 2014, 120: 67-87.
[19]	Hu Y Z, Chan E L, Watanabe J I, et al. Inter-particle torque scaling in coarse grained DEM with rolling resistance and particle size distributions[J]. Powder Technology, 2024, 438: 119612.
[20]	Washino K, Chan E L, Nishida Y, et al. Coarse grained DEM simulation of non-spherical and poly-dispersed particles using scaled-up particle (SUP) model[J]. Powder Technology, 2023, 426: 118676.
[21]	Jin G, Zhou Z, Liu Y, et al. A coarse-grained discrete element method based on the principle of energy density mapping conservation: efficient simulation of particle dynamic mixing and interaction using larger particles[J]. Physics of Fluids, 2025, 37(1): 013377.
[22]	李博, 刘备, 张鹏, 等. 双尺度粗粒化离散元方法及煤散料试验验证[J]. 煤炭科学技术, 2024, 52(3): 225-235.
	Li B, Liu B, Zhang P, et al. A two-scale coarse-grained discrete element method and experimental verification of bulk coal[J]. Coal Science and Technology, 2024, 52(3): 225-235.
[23]	赵婷婷, 冯云田. 大规模颗粒系统的精确缩尺和粗粒化离散元方法[J]. 计算力学学报, 2022, 39(3): 365-372.
	Zhao T T, Feng Y T. Exact scaling laws and coarse-grained discrete element modelling of large scale granular systems[J]. Chinese Journal of Computational Mechanics, 2022, 39(3): 365-372.
[24]	Lu L Q, Xu Y P, Li T W, et al. Assessment of different coarse graining strategies to simulate polydisperse gas-solids flow[J]. Chemical Engineering Science, 2018, 179: 53-63.
[25]	Kazidenov D, Khamitov F, Amanbek Y. Coarse-graining methods for the modified jkr contact model on a triaxial compression test[C]//Rock Mechanics/Geomechanics Symposium. 2022: ARMA-2022.
[26]	Kazidenov D, Khamitov F, Amanbek Y. Coarse-graining of CFD-DEM for simulation of sand production in the modified cohesive contact model[J]. Gas Science and Engineering, 2023, 113: 204976.
[27]	Grabowski J, Jurtz N, Brandt V, et al. Numerical investigation of segregation and mixing in bidisperse systems using the coarse-grained CFD-DEM approach[J]. Powder Technology, 2025, 458: 120922.
[28]	Kanjilal S, Schneiderbauer S. A revised coarse-graining approach for simulation of highly poly-disperse granular flows[J]. Powder Technology, 2021, 385: 517-527.
[29]	Yu Y, Duan F, Wang L, et al. An overlappable coarsening strategy for discrete element method simulations of bi-disperse granular flows[J]. Powder Technology, 2025, 455: 120765.
[30]	Marshall J S, Li S. Adhesive Particle Flow[M]. Cambridge: Cambridge University Press, 2014.
[31]	Chen S, Li S Q, Yang M M. Sticking/rebound criterion for collisions of small adhesive particles: effects of impact parameter and particle size[J]. Powder Technology, 2015, 274: 431-440.
[32]	Pachón-Morales J, Do H, Colin J, et al. DEM modelling for flow of cohesive lignocellulosic biomass powders: model calibration using bulk tests[J]. Advanced Powder Technology, 2019, 30(4): 732-750.
[33]	Roessler T, Katterfeld A. Scaling of the angle of repose test and its influence on the calibration of DEM parameters using upscaled particles[J]. Powder Technology, 2018, 330: 58-66.
[34]	Doan T, Indraratna B, Nguyen T T, et al. Interactive role of rolling friction and cohesion on the angle of repose through a microscale assessment[J]. International Journal of Geomechanics, 2023, 23(1): 04022250.
[35]	Hassanzadeh V, Wensrich C M, Moreno-Atanasio R. Elucidation of the role of cohesion in the macroscopic behaviour of coarse particulate systems using DEM[J]. Powder Technology, 2020, 361: 374-388.
[36]	Roessler T, Katterfeld A. Scalability of angle of repose tests for the calibration of DEM parameters[C]//The 12th International Conference on Bulk Materials Storage, Handling and Transportation (ICBMH 2016). 2016: 201.
[37]	Kosaku Y, Tsunazawa Y, Tokoro C. A coarse grain model with parameter scaling of adhesion forces from liquid bridge forces and JKR theory in the discrete element method[J]. Chemical Engineering Science, 2023, 268: 118428.
[38]	Mohajeri M J, Helmons R L J, van Rhee C, et al. A hybrid particle-geometric scaling approach for elasto-plastic adhesive DEM contact models[J]. Powder Technology, 2020, 369: 72-87.
[39]	McGlinchey D. Simulations in Bulk Solids Handling: Applications of DEM and Other Methods[M]. NYSE: John Wiley & Sons, 2023.
[40]	Müller D, Fimbinger E, Brand C. Algorithm for the determination of the angle of repose in bulk material analysis[J]. Powder Technology, 2021, 383: 598-605.
[41]	Wu Z Y, Li H W, Lu C Y, et al. Development and evaluations of an approach with full utilization of point cloud for measuring the angle of repose[J]. Computers and Electronics in Agriculture, 2023, 209: 107799.
[42]	Beakawi Al-Hashemi H M, Baghabra Al-Amoudi O S. A review on the angle of repose of granular materials[J]. Powder Technology, 2018, 330: 397-417.
[43]	Janda A, Ooi J Y. DEM modeling of cone penetration and unconfined compression in cohesive solids[J]. Powder Technology, 2016, 293: 60-68.
[44]	Thakur S C, Ooi J Y, Ahmadian H. Scaling of discrete element model parameters for cohesionless and cohesive solid[J]. Powder Technology, 2016, 293: 130-137.
[45]	Wiącek J, Molenda M. Representative elementary volume analysis of polydisperse granular packings using discrete element method[J]. Particuology, 2016, 27: 88-94.
[46]	De Pue J, Di Emidio G, Flores R D V, et al. Calibration of DEM material parameters to simulate stress-strain behaviour of unsaturated soils during uniaxial compression[J]. Soil and Tillage Research, 2019, 194: 104303.