MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.

1051/matecconf/202235707005
MMS 2020

The Use of Simulation Software using the

Discrete Element Method (DEM) for the Process
of Materials Comminution
Weronika Kruszelnicka1[0000-0002-9909-5112], Marek Macko2[0000-0002-8743-6602], Daniel
Łączny2[0000-0003-4228-4534], Patrycja Bałdowska-Witos1[0000-0003-0547-4400], and Jakub
Lewandowski2[0000-0003-2630-315X]
1
Bydgoszcz University of Science and Technology, Al. Prof. S. Kaliskiego 7, 85-796 Bydgoszcz,
Poland
2
Kazimierz Wielki University, Chodkiewicza 30 Street, 85-064Bydgoszcz, Poland

Abstract. Comminution processes are one of the most common processes

for processing energy materials, e.g. coal, biomass, and post-recycling
elements. The hitherto unsolved problem is the high energy consumption
of machines and the lack of precise descriptions of the phenomenon of
comminution in terms of the relationship between the design features of
mills and the properties of comminuted materials. The dynamic
development of simulation techniques based on advanced models and the
method of discrete elements allows for a certain mapping of occurring
phenomena. The purpose of the work is to illustrate the possibility of using
simulation software based on the discrete element method to model the
grinding processes in the shredders grinding assemblies. The paper
presents aspects of modeling the shape and size of particles, their
interactions and contacts with mills structural elements, as well as aspects
of crushing modeling in RockyDem software.

1 Introduction
The need for modeling of shredding and flow of bulk materials is mainly due to machine
problems occurring during their processing: loss of ability to move (transport) the bulk
material by blockage of dosing channels, transporting in feed supply systems, processing
processes, too high intermolecular friction, which causes the effect of material clogging,
which in consequence leads to material losses or deterioration of the quality of processed,
transported material, energy start, decrease in efficiency and effectiveness of processing
lines, and thus increase costs [1-5]. Therefore, it seems rational to model and search for
structural solutions of machines working with loose materials, taking into account the
modeling of the behaviour of these materials in contact with the working surface of these
devices.
The issues of modelling in mechanical engineering have been reflected in many works, e.g.
in the field of surface condition modelling, thermal stability and surface properties and
fatigue damage analysis [6-9].

© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons
Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/).
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

The main goal of optimization of machinery and equipment design for grinding of bulk
materials is to minimize costs, reduce energy consumption, while maintaining and even
improving the performance and functionality of these units [10-19]. Improvement of these
parameters is connected with appropriate selection of constructional features. For this
purpose computer support methods are used, e.g. [20-23].
The commonly used analytical methods make it possible to determine the geometric and
strength characteristics of machines, however, they do not take into account the behaviour
of loose materials during shredding [24-27].The appropriate characteristics of the materials,
especially in terms of grain size, have a great influence on the quality of the final products
in many branches, e.g. in the medical industry, in the manufacture of artificial organs or
devices supporting basic activities [28-32]. Computer applications based on solving
problems with the use of numerical methods become helpful in this respect. One of the
methods used to solve calculation problems of grinding and flow of bulk materials is the
Discrete Elements Method (DEM), which allows to simulate the behavior of materials
under machine conditions [1,2,33].
The Discrete Elements Method has recently been one of the most frequently used methods
in the analysis and simulation of the flow of bulk, granular materials, supporting the
optimization of machine and equipment design. Much attention is paid to issues related to
modeling of mixing processes [34,35], storage [1] and transport [36] of granular materials
and grinding, mainly of brittle materials, e.g. rocks and minerals [37-39]. The topic of
modelling granular materials, e.g. cereal grains, is much less frequently addressed [37].
Zeng at al. [37] performed a simulation of rice grain cracking using the method of discrete
elements, examining the influence of impact velocity and humidity on the crack course.
Their considerations concerned single grains, which in industrial applications does not take
into account the interactions between particles during crushing. Karwat et al. [2] indicated
that in modeling the flow of bulk materials it is important to correctly calibrate input
parameters on a macro scale, i.e. it is more important to map the behaviour of the entire
deposit than the dependence of individual particles. Boikov et al. [39] also pointed out the
importance of identifying the material behavior in the macro scale. They noted that in the
case of spherical particles it is necessary to determine 7 input parameters of the simulation,
which, if they change in a linear way, can be described by means of an approximation
function containing these parameters.
The development of a reliable simulation model, as close as possible to the actual behavior
of bulk materials, requires correct calibration of the model input parameters, e.g. particle
shape and size, coefficients of friction and restitution or bulk density, which are determined
experimentally in laboratory tests [2,35,38,39]. The purpose of the work is to illustrate the
possibility of using simulation soft-ware based on the discrete element method to model the
grinding processes in the shredders grinding assemblies. The paper presents aspects of
modeling the shape and size of particles, their interactions and contacts with mills structural
elements, as well as aspects of crushing modeling in RockyDem software.

2 Fundamentals for DEM simulation of comminution

Over the years, the method of discrete elements originally developed by Cundall and Strack
[40,41] has developed dynamically and is used not only for modeling bulk material flows,
but also for shredding modeling [35,37,42]. The first DEM models for rock crushing were
developed in the early 21st century. Potyondy and Cundall [43] developed and introduced a
particle-bonded model for crushing processes, which assumed that single particles consist
of spheres glued together. This model was developed and modified by Cho et al. [44], Tan
et al. [45], Ivars et al. [46]. A different strategy was introduced by Cleary et al. [47],

2
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

according to which spherical particles are replaced by densely packed smaller spheres
(constituting offspring) after exceeding the value of energy corresponding to the fracture,
which made it possible to map crushing processes [47,48]. This method has been extended
so that it is possible to model non-spherical particles [49].
DEM is a numerical method that treats particles individually and motion equations are
solved at the level of single particles, which has made DEM applicable to modeling
discontinuous centers [48]. In order to predict the behavior of a particle, Newtonian
equations of motion and inter-particle contact models are solved [48]. Some of them will be
presented in the next section.

2.1 Possibilities of simulation software

Many commercial applications using DEM are currently available, e.g. RockyDEM,
EDEM, Newton, ThreeParticle/CAE, PASSAGE®/DEM Software, Bulk Flow Analyst or
open-open source codes such as LIGGGHTS (LAMMPS improved for general granular and
granular heat transfer simulations).
The possibilities of computer modelling of granular materials using the Discrete Elements
Method (DEM) will be presented on the example of the commercial RockyDEM software,
which enables to model the movement of bulk materials in machines and devices such as:
belt conveyors, shredders, silos. The software also allows to model particles of any shape and
model their motion on the basis of determining the type of contacts, coefficients of friction and
restitution and to shred non-spherical particles.
In RockyDEM, the calculations are mainly based on Newton's laws and gravity (Fig. 1)
including intermolecular contacts (normal forces, tangential forces and adhesive forces),
with information at the particle scale (Fig. 2) [35,50]. Motion equations are solved using the
finite difference method.

Fig. 1. DEM calculation loop.

The selection of the contact model is one of the first steps in shredding modeling.
The available contact models (Fig. 2) can be used in certain specific cases. The Hysteretic
Linear Spring (HLS) force model is used to model the behavior of compressible and non-
compressible materials, while the Linear Spring Dashpot (LSD) model is only used for
modeling non-compressible materials. Hertzian Spring Dashpot (HSD) model is used for
modeling non-linear materials. Coulomb Limit (CL) can be used with all normal force
models and Mindlin-Deresiewicz (MD) only with the HSD model. Additionally, adhesion
forces can be included in the contact models. Constant Adhesive (CA) and Linear Adhesive
(LA) can be used with all normal force models, while Leeds Adhesive (LD) only with HLS
and JKR only with HSD. The HSD models in combination with MD are most commonly
used in bulk material flow modelling [2].

3
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Fig. 2. Applicable contact models.

To perform the simulation it is possible to import your own geometry or to select conveyor
and feeder models from the application library. For moving parts, a suitable type of motion
is assigned: translational, rotational, vibrating, force-fed, periodic. It is then necessary to
prepare your own particle shape or to choose the shape and size of the particle from the
available shapes (fig. 3.)
a) b) c) d)

e) f) g) h)

Fig. 3. Particle shapes: a) sphere, b) straight fiber, c) polyhedron, d) sphero-cylinder, e) sphero-

polygon, f) sphero-polyhedron, g) briquette, h) faceted cylinder.

Simultaneous simulation of particles of different shapes and dimensions is possible. The

dimensions can be mapped on the scale of the actual grain dimensions, based on screen
dimensions or calculated equivalent sphere diameter. The next step in preparing the
simulation is to determine material properties such as density or bulk density, Yung's
modulus or Poisson's coefficient, and then contact parameters (e.g. static and dynamic
friction coefficient, tangential stiffness ratio, restitution coefficient) resulting from selected
contact models (Fig. 2).
For non-spherical particles two shredding models are available: Ab-T10 model and Tavares
breakage model. Particle decay is performed using the Voronoi fracture algorithm [51].
Ab-T10 is a probability model of particle decay calculated from the value of accumulated
contact energy [52]:
 L 
P( ecum ) 
1  exp  Secum  (1)
 Lref 

where:ecum – cumulative specific energy, S - selection function coefficient, L - particle
size,Lref – reference size.
Tavares breakage model is based on the probability of upper-truncated log-normal
distribution of unit crack energy [53,54]:

4
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

1  ln e *  ln e50  
P0
( e) 1  erf    (2)
2   2 2  
where:e* - relative particle specific fracture energy,e50 – median particle specific fracture
energy,2 – variance of the log-normal distribution of fracture energy.
For mapping the creation of particles of a given size after grinding, an appropriate particle
size distribution based on Gaudin-Schumann or Incomplete Beta can be adjusted. In order
to determine the grinding parameters it is necessary to perform one of the following tests:
drop weight testing, impact load target testing or repeated impact testing to obtain the minimum
particle size distribution energy for a specific reference dimension.
The analysis module allows to determine particle velocity, collision energy, particle size
and mass after grinding as well as to determine the loads exerted by the material on
machine components, including wear modelling.

2.2 Calibration and simulation

The calibration and validation of the model is a key element in creating a simulation using
the discrete elements method, which determines the reliability and credibility of the results
obtained [55]. When preparing a simulation and deciding on the choice of the calibration
method, it is important to be aware of the difference between the concepts of material
properties and model parameters. Material properties are essentially physical characteristics
of the material, which can be determined on a micro scale (relations between single
particles) or on a macro scale (properties determined for a set of particles) [55]. On the
other hand, the model parameters refer to selected physical properties of the material
defined in the micro scale and may vary depending on the selected contact model or used in
DEM code modeling [55,56].
Depending on the research possibilities and approaches to determine material properties (at
the particle or particle group level) two approaches to the calibration process are
distinguished: the Bulk Calibration Approach (BCA) and the Direct Measuring
Approach.(DMA) [55,57] (Fig. 4).

5
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Fig. 4. Calibration approaches of DEM models.

The Bulk Calibration Approach. BCA is based on the calibration and mapping of the bulk
material behavior in the DEM model based on results of bulk material properties tests. In
this case, the bulk density, bulk stiffness, angle of repose and others are determined in the
laboratory tests, after which the input parameters are changed interactively in the DEM
model until the material response is consistent with the experimental tests [55,57].
Unfortunately, the disadvantage of this approach is that once determined parameters for one
sy-mulation do not have to be suitable for the same material in another type of simulation
[57].
Particle size and shape determination. In preparing the DEM model it is important to
determine the parameters of particle shape and size. The shape and size of particles for
different moisture levels of biomass grains is determined by means of granulometric
analysis, thus it is possible to obtain the distribution of particle size distribution,
characteristic dimensions and grain shape parameters [58-60].
Determination of bulk density. The bulk density of a material is a parameter that determines
the volume of the material in the loose state, taking into account the pores and space
between the grains. The value of the bulk density is influenced by, among others, humidity,
specific density as well as the degree of filling of the preset volume. The bulk density is
determined from the relation [2,36,58]:
m  m1
B  2 (3)
V
where: ρB – bulk density in loose state, g∙cm-3, m2 – weight of measuring cylinder and
sample in loose state, g, m1 – weight of measuring cylinder, g, V – capacity of measuring
cylinder, cm3.

6
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Determination of the angle of repose. The angle of natural embankment is called the angle that
is formed between the plane and the cone formed during free pouring of material on the plane.
This angle may be measured directly or by reference to a relationship [2,35].

2h
AoR  arctan (4)
D
where: AoR – natural angle of the repose, h – height of the cone, mm, D – length of the
cone base, mm.

a) b)

h
AoR
Pouring Emptying

Rotating Piling
Fig. 5. Methods of determining the angle of repose: a) geometric method of determining the angle of
repose, b) methods of cone repose.

In the secondary case, image analysis and a quick-frame camera are used to record the
arrangement of material. The rotating /drum method is used to determine the dynamic AoR,
while others in Fig. 5 allows to determine the static AoR.
The Direct Measuring Approach.(DMA). In DMA, the material parameters for individual
particles are determined, which then serve as the corresponding input data in the DEM
model. The problem is the determination of material properties of very small sizes [55,57].
There is also no certainty that after entering the obtained parameters into the DEM model
the material behavior will correspond to the actual one [55,61,62]. The appropriate
precision of the model in this case depends on the accuracy of the particle shape mapping
and selection of the appropriate contact model [55,63].
Determination of particle-surface coefficient of friction. Determination of the coefficient of
friction can be carried out using the test of the slope equilibrium with a variable slope angle
and interchangeable substrates. The material is placed on a movable plate and starts to lift
the plate until the material starts to slip evenly. In such an arrangement the tilt angle of the
equatorial or the length of the equatorial throw together with its height can be directly
measured (Fig. 2a) [35]. Based on the results obtained, the coefficient of friction can be
calculated from the formula:
FTS h
S 
 
tg (5)
FN l

7
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

a) b)
h h
α α
l l

Fig. 6. Example of determining the coefficient of friction by means of the inclined test a) for the
particle-plate pair, b) for the particle-pair - particle.

When considering the movement of bulk material similar to a spherical or cylindrical

shape, enabling it to roll over an inclined surface, the rolling friction moments and the
values of the rolling friction coefficient (RFC) should be determined. For spheres for
viscoelastic materials, the RFC without slip is [64]

krol 
 2

1  32  1   1    1  2  
2

(6)
3  32  21   Y 2 
 

where: ƞ1, ƞ2 – dissipative constants. In work[64] it was shown that knowing the RC for a
given material, the value of RFC can be calculated.
Determination of the internal friction coefficient (particle-particle). The coefficient of
internal friction between particles can also be carried out using the experiment on a sloping
equilibrium. In this case, particles of the material are additionally permanently placed on
the equilibrium, after which the loose particles will slide (Fig. 2b) [12].
Determination of the restitution coefficient. The restitution coefficient is a measure
of particle inflexibility during a collision. It is extremely important for determining and
predicting the trajectory of materials after the impact. In simple terms, it describes the
inflexible collision of a particle with a plane, and can be defined as [65,66]:

v'
 (7)
v
where: v – particle velocity before collision, m∙s-2, v’ – velocity of the particle after
collision, m∙s-2. For an ideal fully elastic collision, the coefficient of restitution is 1.
In practice, this coefficient is within the range from 0-1. Simplifying with the omission of
friction forces, this coefficient can be determined on the basis of the knowledge of changes
of potential energy of the body Ep1 falling from the height h1 and potential energy after the
reflection of Ep2 from the surface to the height h2 according to formula [35]:

Ep 2 mgh2 h2
   . (8)
Ep1 mgh1 h1

The RC is a measure of the inelasticity of particles during a collision. In simplified terms, it

describes the collision of an inflexible particle with a plane, and can be defined as [65,66]

v' vi  1 ti 1  ti
i
   (9)
v vi ti  ti 1

8
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

g g
vi   ti  ti 1   ti , i 2, 3,.... (10)
2 2
where: v – particle velocity before collision, m∙s-2, v’ – particle velocity after collision in
time ti, m∙s-2. It can be measured as the time delay between successive impacts of particles
on a hard surface by piezoelectric force sensors or accelerometers mounted in the plate.
Delay, and thus RC, can also be measured by recording the sound of the particles hitting the
disc[67].
RC is a function of impact velocity, particles masses, radii and materials parameters[67].
Depending on interaction force law, the RC can be determined by solving the equations of
motion. For linear-dashpot force, the RC is constant and independent of collision speed[68]

  
 d0 exp  
  (11)
 2 km /  2
 1 
 

In fact, RC is dependent on speed and its fluctuations are observed, which result from the
microscopic roughness of the surface of the spheres hitting the plate, causing the transfer of
energy between the translational and rotational degree of freedom. RC increases as the
speed of impact decreases. At constant speed, the fluctuation probability distribution of RC
is non-Gaussian[58].
Calibration of shredding parameters. Shredding modelling requires, in addition to
calibrating the contact parameters, the determination of the input parameters for particle
division. The literature describes a number of tests that allow the determination of the
grinding energy and particle size after grinding, including Bond Ball Mill Grindability test,
Bond Rod Mill Grindability test, Bond Low-Energy Impact test, JKTech drop-weight test,
SAG mill comminution test [69]. The brief characteristics of these tests are presented in
Table 1.

Table 1. Characteristics of selected grindability tests.

Mill Top Closing Sample Sample

Test diameter, size, size, requested, consumed, Type
m mm mm kg kg
Locked-
Bond Ball Mill 0.305 3.3 0.149 10 Y
cycle
Bond Rod Mill 0.305 13 1.2 15 8 Locked-cycle
Bond Low-Energy
- 75 - 25 10 Single particle
Impact
JKTech drop-
- 63 - 75 25 Single particle
weight
Sag Mill
- 32 - 20 5 Single particle
Comminution

The Ab-T10 is one of the shredding models in the RockyDEM software and its parameters
can be determined with the JKTech drop-weight test, so it will be discussed in more detail.
The JKTech drop-weight test is divided into three parts: the first determination of resistance
to impact cracking (material size 63-13.2 mm), the second determination of resistance to

9
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

abrasion, breaking (particles size 53-37.5 mm), the third determination of mean density and
dispersion for 30 particles [69].
In the Impact test, five dimensional fractions of the material are tested in three series
differing in the shredding energy. In each series, 10 to 30 particles are shredded. In the
tests, the level of shredding energy is known, which is determined by the weight and height
from which it is lowered onto the material. The samples are then subjected to a sieve
analysis after shredding. As a result of the test, a crack curve t 10 characteristic of the
material is obtained, as determined by equation [69]:

t10 A 1  e

 bEcs
 (12)
where: t10 –the percent weight of fragments thatpasses 0,1 of its original size A, b – model
parameters, Ecs – unit shredding energy(kWh/t).

3 Example of simulating particle crushing using DEM

The shredding simulation model was developed in the Rocky DEM environment.
The crushing of grain was modelled, based on the results of grain size, force, stress and
compression work of rice in a static compression test presented in [70]. First, the spatial
form of rice grain was modeled and the working arrangement of the piston and plate for
particle compression was mapped on the Instron 5966 machine (Fig. 7a). The simulation
assumed a mass of the piston equal to 100 kg, and took into account the interaction of
gravity force and earth acceleration (g=9, 81 m/s2). The piston descending speed was set at
0.005 m/s. Table 2 shows the dimensional parameters of the modelled rice grain and Figure
7b shows the modelled grain shape.

Table 2. Input parameters for rice size and shape modeling.

Input parameter Value

Particle shape Polyhedron
Number of angles 110
Superquadric degree 3.0
Vertical aspect ratio 0.8
Horizontal aspect ratio 3.50
Sieve size 1.91 mm
Bulk density of rice 750 kg/m3
Young's modulus of rice 143.66 MPa
Poisson's Ratio 0.4

10
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Fig. 7. a) Modelled piston-plate compression system, b) Modelled shape of rice grain.

Then contacts and friction coefficients were determined and verified. Mindlin-Deresiewicz
contact model was used, adhesion forces were omitted. Other parameters used in the
contact model are shown in Table 3.

Table 3. Parameters used in the simulation.

Value
Parameter
Particle - wall Particle-particle
Static friction 0.2123 [62] 0.8 [63]
Dynamic friction 0.1610 [62] 0.1610 [62]
Tangential stiffness ratio 0.5 [24] 0.2 [63]
Restitution coefficient 0.5 [24] 0.1 [63]
Rolling resistance - - 0.3 [64]

The Ab-T10 and Gaudin-Shumann models were chosen as the grinding model for both
grains for the reproduction of particle size distribution after grinding. The grinding
modeling parameters for rice are shown in Table 4.

Table 4. Shredding modelling parameters.

Parameter value Value

Shredding model Ab-T10
Reference size 1.91 mm
Reference minimum specific energy 5410 J/kg
Selection function Coefficient 0.000184573 kg/J
Maximum t10 value 4%
Particle distribution Gaudin-Schumann
Minimum size 0.212 mm

Figure 8 shows the squeezed grain of rice and corn, generated. It should be noted that
particles with different geometries were obtained.

11
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Fig. 8. View of rice grain after squeezing.

A dependency was used to assess the compatibility of the simulation model with the real
one:
x 1sx
 1e
p  100% (13)
x1e
where: ∆p – relative error of the designated simulation parameter of the batch material
motion, x1s – the value of the designated parameter based on the simulation, x1e – the value
of the parameter designated based on the experiment.

The particle size distributions obtained as a result of simulation in a computer program and
as a result of an experiment on a testing machine are similar to each other, although in the
case of simulation finer particles were obtained (Fig. 9 and Fig. 10). It should be pointed
out that the simulation concerned one grain, while the results from the strength tests
concern 100 grains, so discrepancies may result.
The largest relative error of the share of particles after grinding for the simulation results
was recorded for particles in size classes below 0.68 mm, for these particles the error was
over 50%. As the particle size increased, the error value decreased logarithmically. The
constructed simulation model best represented the percentage share of particles with
dimensions above 1.46 mm, where the relative error was below 11%. The average relative
error of the cumulative percentage of particles after grinding was 28.43%.
The developed simulation model will be well suited to predict the size distribution of rice
grinding particles to larger fractions, as indicated by the obtained errors of the cumulative
percentage share of particles depending on the fraction size. It should be noted that in the
real experiment the biological material - rice grains, which are characterized by great
dimensional and structural diversity, was crushed, which causes a significant scattering of
the results of physical experiments and determination of their properties (see: [70]). The
method of discrete elements in case of prediction of granulometric composition takes the
best results of compliance for the grinding of brittle materials, rocks for which structural
diversity is smaller within one type of material, and calibration methods give more reliable
results. Given the large variety of rice grains and the fact that the models available in the
simulation software are rather adapted to the shredding of brittle materials, it is necessary to

12
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

conduct research
G raph o and
f ea sdevelop
m u rem en t re scalibration
u lts :C om b n i a tmethods
o
in, a l that
t sk f ie ryż śccan
i kan ybe
s a f successfully used in
. g
modelling the shredding
Re ef ren ce f ie
R aw da a t : of biological
l :C um u al tvie s z
materials.
ie d s i trb
i u to
i n R o cky D EM .re f

Q3%
[ ]

80
70
60
50
40
30
20
C um u a l tv
ie s z
ie d s i trbi u to
i n R o cky D EM .rd f
10 R ci e a fe
t r com p re s oi n .rd f

0 05
. 10
. 15
. 20
. xc_m n
i m
[m ]

Fig. 9. Particle size distribution distributions after grinding obtained as a result of physical expetion
and simulation.

100
Reliative error Average relative error
90
80
70 y = -31,19ln(x) + 105,06
Reliative error, %

R² = 0,9532
60
50
40
30
20
10
0
0,35 0,48 0,62 0,75 0,88 1,01 1,14 1,27 1,4 1,53 1,66 1,8 1,92 2,05 2,19
Particle size, mm

Fig. 10. Relative error of the cumulative share of particle size obtained as a result of simulation for
individual particle size classes.

4 Summary
The method of discrete elements has recently been one of the most frequently used methods
in the analysis and simulation of the flow of bulk and granular materials, supporting the
optimization of machine and equipment design. Much attention is paid to issues related to
the modelling of mixing processes, storage and transport of granular materials and
crushing, mainly of brittle materials such as rocks and minerals.
Calibrating and developing an adequate model using the discrete elements method is a
time-consuming process that requires a number of experiments to determine the input
parameters that determine the behavior of the bulk material. Unfortunately, sometimes

13
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

correct determination of input materials will not guarantee that the simulation model will
reflect reality.
Simulation programs for flow modeling and grinding of materials by means of discrete
elements facilitate the process of modeling, as they do not require entering own code
sequences and the user can use ready-made solutions which shorten the time of simulation
and calculations. Sometimes this can be a limitation, although in most programs it is also
possible to enter your own macros and numerical codes.
The process of shredding biological materials takes place in an uncontrolled and difficult to
predict manner. In the case of these materials, the development of simulation models
reflecting the actual behavior is problematic, especially as the properties of the materials
change within one species, variety and also with the change in humidity.
This paper presents the possibilities of simulation software and draws attention to the most
important elements when determining the input parameters of the simulation model and the
calibration process.

References

1. J.P. Lopes Neto, J.W.B. Nascimento, J.I. Marques, C.A. da Costa nMechanical
properties of grain in silos for simulation designs. Eng. Agríc. 2016, 36, 573–580
(2016),
2. B. Karwat, R. Machnik, J. Niedźwiedzki, M. Nogaj, P. Rubacha, E. Stańczyk
Calibration of bulk material model in Discrete Element Method on example of perlite
D18-DN. Eksploatacja i Niezawodność 2019, 21, 351–357 (2019).
3. M. Bembenek Research and prospects for new areas of using roller presses. Przem.
Chem. 2017, 96, 1845–1847.
4. V. Mannheim Empirical and scale-up modeling in stirred ball mills. Chem. Eng. Res.
Des.2011, 89, 405–409 (2011).
5. M. Jewiarz, K. Mudryk, K., M. Wrobel, J. Fraczek, K. Dziedzic Parameters Affecting
RDF-Based Pellet Quality. Energies 2020, 13, 910, (2020).
6. T. Domański, A. Sapietová, M. Sága Application of Abaqus Software for the
Modeling of Surface Progressive Hardening. Procedia Engineering 2017, 177, 64–69.
7. P. Krawiec, L. Różański, D. Czarnecka-Komorowska, Ł. Warguła Evaluation of the
Thermal Stability and Surface Characteristics of Thermoplastic Polyurethane V-Belt.
Materials, 13, doi:10.3390/ma13071502 (2020).
8. M. Sága, P. Kopas, M. Uhríčík Modeling and Experimental Analysis of the
Aluminium Alloy Fatigue Damage in the case of Bending – Torsion Loading.
Procedia Engineering 2012, 48, 599–606 (2012).
9. M. Hryniewicz, M. Bembenek, A. Janewicz, B. Kosturkiewicz Agglomeration of fine-
grained materials in roll presses with asymmetrical compaction unit. Przem. Chem.
2015, 94, 2223–2226 (2015).
10. A. Tomporowski, J. Flizikowski, J. Wełnowski, Z. Najzarek, T. Topoliński, W.
Kruszelnicka, I. Piasecka, S. Śmigiel Regeneration of rubber waste using an
intelligent grinding system. Przemysł Chemiczny 2018, 97, 1659–1665 (2018).
11. W. Kruszelnicka, P. Bałdowska-Witos, R. Kasner, J. Flizikowski, A. Tomporowski, J.
Rudnicki Evaluation of emissivity and environmental safety of biomass grinders
drive. Przemysł Chemiczny 2019, 98, 1494–1498 (2019).
12. J. Flizikowski, W. Kruszelnicka, M. Macko The Development of Efficient
Contaminated Polymer Materials Shredding in Recycling Processes. Polymers 13, 713
(2021).

14
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

13. M. Macko, A. Mrozinski Work parameters research of wood pellet machine. AIP.
Conference Proceedings, 2077. DOI: 10.1063/1.5091899 (2019).
14. M. Macko, A. Mroziński, A. Prentki A. Simulations CAE of wood pellet machine.
MATEC Web of Conferences 254, 02028 (2019).
15. A. Tomporowski Stream of efficiency of rice grains multi-disc grinding. Eksploatacja
i Niezawodność - Maintenance and Reliability 2012, 14 (2012).
16. W. Kruszelnicka, R. Kasner, P. Bałdowska-Witos, J. Flizikowski, A. Tomporowski
The Integrated Energy Consumption Index for Energy Biomass Grinding Technology
Assessment. Energies 2020, 13, 1417 (2020).
17. M. Macko, A. Mroziński Computer Aided Design of Wood Pellet Machines. In:
Rusiński E., Pietrusiak D. (eds) Proceedings of the 14th International Scientific
Conference: Computer Aided Engineering. CAE 2018, p. 454-46. Lecture Notes in
Mechanical Engineering. Springer, Cham (2019).
18. A. Marczuk, W. Misztal, P. Savinykh, N. Turubanov, A. Isupov, D. Zyryanov
Improving Efficiency of Horizontal Ribbon Mixer by Optimizing Its Constructional
and Operational Parameters. Eksploat. Niezawodn. 2019, 21, 220–225 (2019).
19. G. Zając, A. Węgrzyn Analysis of work parameters changes of diesel engine powered
with diesel fuel and faee blends. Eksploatacja i Niezawodnosc-Maintenance and
Reliability 2008, 17–24 (2008).
20. M. Macko, K. Tyszczuk, G. Śmigielski, J. Flizikowski, A. Mroziński The use of CAD
applications in the design of shredders for polymers. MATEC Web Conf. 157, 02027
(2018)..
21. J. Flizikowski, M. Macko Competitive design of shredder for plastic in recycling. In
Proceedings of the Tools and methods of competitive engineering, vols 1 and 2;
Lausanne, Switzerland, pp. 1147–1148 (2004).
22. M. Macko, A. Mroziński, J. Flizikowski Design and Utility of Specialist
Comminution Set-Up for Plastics and Organic Materials. American Society of
Mechanical Engineers Digital Collection, pp. 397–402 (2012).
23. J. Flizikowski, M. Macko The influence of design of shredder on quality of product
(Conference paper) REWAS'04 - Global Symposium on Recycling, Waste Treatment
and Clean Technology, 2841-2842, code 66515 (2005).
24. J.B. Flizikowski, W. Kruszelnicka, A. Tomporowski, A. Mrozinski A study of
operating parameters of a roller mill with a new design. AIP Conference Proceedings
2019, 2077, 020018 (2019).
25. K. Piotrowska, W. Kruszelnicka, P. Bałdowska-Witos, R. Kasner, J. Rudnicki, A.
Tomporowski, J. Flizikowski, M. Opielak Assessment of the Environmental Impact of
a Car Tire throughout Its Lifecycle Using the LCA Method. Materials 2019, 12, 4177,
(2019)
26. J.B. Flizikowski, A. Mrozinski, A. Tomporowski Active monitoring as cognitive
control of grinders design.; AIP Conference proceedings: Melville, New York, 1822,
p. 020006 (2017).
27. M. Tymoszuk, K. Mroczek, S. Kalisz, H. Kubiczek An investigation of biomass
grindability. Energy, 183, 116–126 (2019).
28. M. Macko, Z. Szczepański, E. Mikolajewska, J. Nowak, D. Mikolajewski Repository
of 3D images for education and everyday clinical practice purposes. Bio-Algorithms
and Med-Systems, 13, 2, 111-116 (2017).
29. M. Macko, Z. Szczepański, D. Mikołajewski Design and manufacture of artificial
organs made of polymers. 23rd Polish-Slovak Scientific Conference on Machine
Modelling and Simulations (MMS), 2018. Book Series: MATEC Web of
Conferences, 254, (2019).

15
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

30. J. Kopowski, D. Mikołajewski, M. Macko, I. Rojek Bydgostian hand exoskeleton -

own concept and the biomedical factors. Bio-Algorithms and Med-Systems. DOI:
10.1515/BAMS-2019-0003 (2019).
31. J. Kopowski, I. Rojek, D. Mikołajewski, M. Macko 3D printed hand exoskeleton -
own concept. Lecture Notes in Mechanical Engineering, pp. 298-306 (2019).
32. M. Macko, Z. Szczepański, D. Mikołajewski, E. Mikołajewska, S. Listopadzki The
Method of Artificial Organs Fabrication Based on Reverse Engineering in Medicine.
In: Rusiński E., Pietrusiak D. (eds) Proceedings of the 13th International Scientific
Conference. RESRB 2016. Lecture Notes in Mechanical Engineering. Springer,
Cham. (2017).
33. Z. Liang, Y. Li, L. Xu, Z. Zhao Sensor for monitoring rice grain sieve losses in
combine harvesters. Biosystems Engineering 147, 51–66 (2016),
34. J. Hlosta, L. Jezerská, J. Rozbroj, D. Žurovec, J. Nečas, J. Zegzulka DEM
Investigation of the Influence of Particulate Properties and Operating Conditions on
the Mixing Process in Rotary Drums: Part 2—Process Validation and Experimental
Study. Processes 2020, 8, 184 (2020).
35. J. Hlosta, L. Jezerská, J. Rozbroj, D. Žurovec, J. Nečas, J. Zegzulka DEM
Investigation of the Influence of Particulate Properties and Operating Conditions on
the Mixing Process in Rotary Drums: Part 1—Determination of the DEM Parameters
and Calibration Process. Processes 2020, 8, 222 (2020).
36. J. Pachón-Morales, H. Do, J. Colin, F. Puel, P. Perré, D. Schott DEM modelling for
flow of cohesive lignocellulosic biomass powders: Model calibration using bulk tests.
Advanced Powder Technology, 30, 732–750 (2019).
37. Y. Zeng, F. Jia, Y. Xiao, Y. Han, X. Meng Discrete element method modelling of
impact breakage of ellipsoidal agglomerate. Powder Technology. 346, 57–69 (2019).
38. X. Wang, J. Yi, Z. Zhou, C. Yang Optimal Speed Control for a Semi-Autogenous Mill
Based on Discrete Element Method. Processes 2020, 8, 233 (2020).
39. A.V. Boikov, R.V. Savelev, V.A. Payor DEM Calibration Approach: design of
experiment. J. Phys.: Conf. Ser. 1015, 032017 (2018),
40. P.A. Cundall, O.D.L. Strack A discrete numerical model for granular assemblies.
Géotechnique 29, 47–65 (1979).
41. Z. Yan, S.K. Wilkinson, E.H. Stitt, M. Marigo Discrete element modelling (DEM)
input parameters: understanding their impact on model predictions using statistical
analysis. Comp. Part. Mech. 2, 283–299 (2015).
42. L.F. Orozco, D.-H. Nguyen, J.-Y. Delenne, P. Sornay, F. Radjai Discrete-element
simulations of comminution in rotating drums: Effects of grinding media. Powder
Technology, 362, 157–167 (2020).
43. D.O. Potyondy, P.A. Cundall A bonded-particle model for rock. 41, 1329–1364
(2004).
44. N. Cho, C.D. Martin, D.C. Sego A clumped particle model for rock. International
Journal of Rock Mechanics and Mining Sciences, 44, 997–1010 (2007).
45. Y. Tan, D. Yang, Y. Sheng Discrete element method (DEM) modeling of fracture and
damage in the machining process of polycrystalline SiC. Journal of the European
Ceramic Society, 29, 1029–1037 (2009).
46. D. Mas Ivars, M.E. Pierce, C. Darcel, J. Reyes-Montes, D.O. Potyondy, R. Paul
Young, P.A. Cundall The synthetic rock mass approach for jointed rock mass
modelling. International Journal of Rock Mechanics and Mining Sciences, 48, 219–
244 (2011).
47. P.W. Cleary, M. Prakash, M.D. Sinnott, M. Rudman, R. Das Large Scale Simulation
of Industrial, Engineering and Geophysical Flows Using Particle Methods. In Particle-
Based Methods: Fundamentals and Applications; Oñate, E., Owen, R. (Eds.);

16
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

Computational Methods in Applied Sciences; Springer Netherlands: Dordrecht

(2011).
48. N.S. Weerasekara, M.S. Powell, P.W. Cleary, L.M. Tavares, M. Evertsson, R.D.
Morrison, J. Quist, R.M. Carvalho The contribution of DEM to the science of
comminution. Powder Technology, 248, 3–24 (2013).
49. G.W. Delaney, P.W. Cleary, M.D. Sinnott, R.D. Morrison, R.D. Novel application of
DEM to modelling comminution processes. IOP Conference Series: Materials Science
and Engineering, 10, 012099 (2010).
50. H. Pourtavakoli, E.J.R. Parteli, T. Poeschel, T. Granular dampers: does particle shape
matter? New J. Phys. 2016, 18, 073049 (2016).
51. Y. Lv, H. Li, X. Zhu, X. W Liu Discrete element method simulation of random
Voronoi grain-based models. Cluster Comput, 20, 335–345 (2017).
52. S. Fengnian, T. Kojovic Validation of a model for impact breakage incorporating particle
size effect. Internationa Journal of Mineral Processing, 82, 156–163 (2007).
53. L.M. Tavares Optimum routes for particle breakage by impact. Powder Technology,
142, 81–91 (2004).
54. L.M. Tavares Analysis of particle fracture by repeated stressing as damage
accumulation. Powder Technology, 190, 327–339 (2009).
55. C.J. Coetzee Review: Calibration of the discrete element method. Powder
Technology, 310, 104–142 (2017).
56. C. González-Montellano, J.M. Fuentes, E. Ayuga-Téllez, F. Ayuga, F. Determination
of the mechanical properties of maize grains and olives required for use in DEM
simulations. Journal of Food Engineering, 111, 553–562 (2012).
57. M. Marigo, E.H. Stitt Discrete Element Method (DEM) for Industrial Applications:
Comments on Calibration and Validation for the Modelling of Cylindrical Pellets.
KONA Powder and Particle Journal, 32, 236–252 (2015).
58. Food powders: physical properties, processing, and functionality; Barbosa-Cánovas,
G.V., Ed.; Food engineering series; Kluwer Acad./Plenum: New York (2005).
59. A. Tomporowski, J. Flizikowski, W. Kruszelnicka A new concept of roller-plate mills.
Przemysł Chemiczny 2017, 96, 1750–1755 (2017).
60. W. Kruszelnicka, A. Idzikowski, K. Markowska, R. Kasner Quality Index of Multi-
Disc Grinding Process of Grainy Biomass. QPI 2019 1, 503–511 (2019).
61. J. Quist, C. Evertsson Framework for DEM Model Calibration and Validation. 6.
62. T.A.H. Simons, R. Weiler, S. Strege, S. Bensmann, M. Schilling, A. Kwade A Ring
Shear Tester as Calibration Experiment for DEM Simulations in Agitated Mixers – A
Sensitivity Study. Procedia Engineering, 102, 741–748 (2015).
63. G.K.P. Barrios, R.M. de Carvalho, A. Kwade, L.M. Tavares Contact parameter
estimation for DEM simulation of iron ore pellet handling. Powder Technology, 248,
84–93 (2013).
64. N.V. Brilliantov, T. Poeschel Rolling friction of a viscous sphere on a hard plane.
Europhys. Lett., 42, 511–516 (1998).
65. M. Heckel, A. Glielmo, N. Gunkelmann, T. Pöschel Can we obtain the coefficient of
restitution from the sound of a bouncing ball? Phys. Rev. E, 93, 032901 (2016).
66. N. Gunkelmann, M. Montaine, T. Pöschel Stochastic behavior of the coefficient of
normal restitution. Phys. Rev. E, 89, 022205 (2014).
67. M. Montaine, M. Heckel, C. Kruelle, T. Schwager, T. Pöschel Coefficient of
restitution as a fluctuating quantity. Phys. Rev. E, 84, 041306 (2011).
68. T. Schwager, T. Pöschel Coefficient of restitution and linear–dashpot model revisited.
Granular Matter, 9, 465–469 (2007).
69. A. McKen, S. Williams An overview of the small-scale tests available to characterise
ore grindability. SGS Minerals Services (2006)..

17
MATEC Web of Conferences 357, 07005 (2022) https://doi.org/10.1051/matecconf/202235707005
MMS 2020

70. W. Kruszelnicka, A. Marczuk, R. Kasner, P. Bałdowska-Witos, K. Piotrowska, J.

Flizikowski, A. Tomporowski Mechanical and Processing Properties of Rice Grains.
Sustainability, 12, 552 (2020).
71. P.C. Corrêa, F.S. da Silva, C. Jaren, P.C. Afonso, I. Arana Physical and mechanical
properties in rice processing. Journal of Food Engineering, 79, 137–142 (2007).
72. J. Guowei, Q. BaijinG Discrete Element Method Simulation of Impact-Based
Measurement of Grain Mass Flow. In Proceedings of the 2011 International
Conference on Computer Distributed Control and Intelligent Environmental
Monitoring; IEEE: Changsha, Hunan, China, pp. 419–422 (2011).
73. D. Markauskas, R. Kačianauskas Investigation of rice grain flow by multi-sphere
particle model with rolling resistance. Granular Matter, 13, 143–148 (2011).

18