Advances in Science and Technology

Research Journal
Volume 11, Issue 3, September 2017, pages 246–252 Research Article
DOI: 10.12913/22998624/76100

FEEDER TYPE OPTIMISATION FOR THE PLAIN FLOW DISCHARGE PROCESS

OF AN UNDERGROUND HOPPER BY DISCRETE ELEMENT MODELLING

Jan Necas1, Jakub Hlosta1, David Zurovec1, Martin Zidek1, Jiri Rozbroj1, Jiri Zegzulka1

1
VSB-Technical University of Ostrava, ENET Centre, 17. listopadu 15/2172, 708 33 Ostrava-Poruba, Czech
Republic, e-mail: jan.necas@vsb.cz

Received: 2017.05.15 ABSTRACT

Accepted: 2017.07.01 This paper describes optimisation of a conveyor from an underground hopper intended
Published: 2017.09.03 for a coal transfer station. The original solution was designed with a chain conveyor
encountered operational problems that have limited its continuous operation. The Dis-
crete Element Modeling (DEM) was chosen to optimise the transport. DEM simula-
tions allow device design modifications directly in the 3D CAD model, and then the
simulation makes it possible to evaluate whether the adjustment was successful. By
simulating the initial state of coal extraction using a chain conveyor, trouble spots
were identified that caused operational failures. The main problem has been the in-
creased resistance during removal of material from the underground hopper. Revealed
resistances against material movement were not considered in the original design at
all. In the next step, structural modifications of problematic nodes were made. For
example, the following changes have been made: reduction of storage space or instal-
lation of passive elements into the interior of the underground hopper. These modifica-
tions made were not effective enough, so the type of the conveyor was changed from
a drag chain conveyor to a belt conveyor. The simulation of the material extraction
using a belt conveyor showed a significant reduction in resistance parameters while
maintaining the required transport performance.

Keywords: Drag Chain Conveyor, Discrete Element Method, Simulation.

INTRODUCTION take into account the dynamic effects arising from

the storage and transport of bulk materials. When
Underground hoppers are mostly used for designing transport and storage facilities, it is ap-
short-term storage of bulk material. A great ad- propriate to use studies and knowledge from vari-
vantage of these storage facilities is the easy and ous areas of transport and handling equipment de-
quick unloading of the input material e.g. by signed not only for bulk materials [3, 4, 5, 6, 7, 8,
truck or rail. Almost every underground hopper is 9]. In this case, the problem of optimisation of the
equipped with a dispenser that conveys the bulk existing underground hopper and the drag chain
material to the related technology. The optimal conveyor is being solved. The existing drag chain
design of underground hoppers and dispensing conveyor system is designed for the processing
devices depends on many factors, operating con- of black coal. During operation, there are operat-
ditions and properties of the stored material [1, 2]. ing problems that limit continuous transport. The
The initial design of transport and storage devices Discrete Element Method (DEM) was used to op-
begins almost always with a theoretical calcula- timise the current unsatisfactory state. The advan-
tion of the dimensional and performance param- tage of optimising transport and storage processes
eters of individual parts. The calculations should using DEM simulations is the computer-based

246
 Advances in Science and Technology Research Journal Vol. 11 (3), 2017

verification of planned design changes [10]. This Black coal has a wide range of particle size dis-
replaces the step of experimental design chang- tributions, approximately from 0 mm to 100 mm.
es that do not guarantee removal of the original Other parameters are bulk density ρS = 1100 kg/m3
problems. These direct structural modifications and maximum material humidity θ = 15%. A 3D
are always costly both in time and money. The model of a cross-cut of the underground hopper
DEM simulation modelling significantly reduces and drag chain conveyor is in Fig. 1.
the development time of new devices as well as of In real-time operation, there were problems
optimising existing devices [11, 12]. At the begin- with the occurrence of bridging in the under-
ning, it is very important to define the mechani- ground hopper in the spaces above the drag chain
cal and physical properties of the bulk material to conveyor channel. The bridging was causing
be processed. These properties include grain size blockages of material above the conveyor, and the
and shape, material density, shear modulus and drag chain conveyor transported only a minimal
the Poisson constant. In our case, it is black coal. amount of material. The second major problem
The mechanical and physical properties of the occurred with the drag chain conveyor. When the
material together with the transport system model underground hopper was filled with more mate-
are the input parameters for the creation of the ac- rial (approximately 1/3 of the total volume of the
tual DEM simulation. During the simulation pro- underground hopper) and after a certain period of
cess of the material transport, it is possible to de- time had elapsed, the movement of the conveyor
tect different physical quantities at different time was stopped. The reason was overloading of the
points, such as the vector velocity field, the force drive, which could not overcome the emerging
and torque effects of the particles on the structure drag resistance. Other undesirable effects during
in any direction, etc. [13, 14, 15]. operation were manifested by deformations of the
drive elements and the shredding of the material
DESCRIPTION OF THE CURRENT at the point of material extraction from the hopper.
PROBLEM Based on the above facts, which stem from
real operation and available technical param-
The current problematic situation concerns the eters, a 3D CAD model of the current state was
underground hopper with a capacity of V = 30 m3 created, see Fig. 1. The 3D CAD model was im-
and the drag chain conveyor. The drag chain con- ported into EDEM Academic, which is used to
veyor is dimensioned for the amount of transported create DEM simulations. The model of the un-
material Q = 100 t·h-1 with the installed drive pow- derground hopper and drag chain conveyor has
er P = 11 kW. The length of the transport channel been assigned physical properties and kinematic
L = 16 m and speed of the drag chain conveyor parameters. In the second step, a bulk mate-
is v = 0.74 m·s-1. The combination of the under- rial resembling black coal was created (Contact
ground hopper and the drag chain conveyor is used Material + Material Definition). The individual
to process the black coal, which is delivered to the particles were assigned the measured mechani-
underground hopper by truck at regular intervals. cal and physical properties. Then the calibration

Fig. 1. Cross-cut of an underground hopper and a chain conveyor

247
Advances in Science and Technology Research Journal Vol. 11 (3), 2017

Fig. 2. Illustration of preparation of the DEM simulation

simulation was performed (Material Calibration). Furthermore, DEM simulation discovered in-
The basic material calibration was performed creased values of the total force which affects the
on a comparative experiment where the angle drive elements during the movement of the drag
of repose of real material with DEM simulation chain conveyor. Total force effect on the drive el-
was compared. After the necessary calibration ements expresses the total drag resistance. From
has been done, the material was generated in a the 36-second time period of the DEM simula-
prepared underground hopper model. The whole tion, the average value of FT = 40000 N has been
process of preparing the DEM simulation is visu- read. The course of FT in time and the distribution
ally represented in Figure 2. of forces on the drive elements of the drag chain
The DEM simulation of the current state has conveyor are illustrated on Fig. 4. Total drag re-
confirmed the large compressive forces occur- sistance is the main parameter when calculating
ring in the material while movement of the drag and dimensioning the performance of the drag
chain conveyor. Large compressive forces oc- chain conveyor. Following the established total
curred especially in the area where the material force from the DEM simulation, a control calcu-
from the underground hopper was extracted. The lation of the required power of the drag chain con-
size of the compressive force in the material was veyor was carried out according to equation (1).
approximately 1000 N. The critical areas of the The control calculation showed that the proposed
compressive forces in the underground hopper drive is seriously underpowered. The drive power
are shown in Fig. 3. for this application should be at least 45 kW.

Fig. 3. Displaying areas of compression forces in a underground hopper: a) real situation, b) DEM simulation

248
 Advances in Science and Technology Research Journal Vol. 11 (3), 2017

Fig. 4. a) Dependence of total force FT on time; b) Loadings on the drag elements of the chain conveyor

posed, which should reduce the resistance param-

(1) eters during the movement. To reduce the loading
area of the material to the bottom of the channel, a
P … required power of the drive unit [kW] passive element was applied which was placed in
FT … total drag resistance [N] the underground hopper space. A further modifica-
v … speed of the drag chain conveyor [m.s-1] tion consisted in limiting the length of the drive
ƞ … overall efficiency of mechanical drive trans- portion of the conveyor located in the underground
mission [-] hopper space. Structural modifications were made
only in a virtual 3D model. Subsequently, a new
OPTIMISATION OF THE FEEDER PROCESS DEM simulation was performed, see Fig. 5.
Evaluated results from the DEM simulation
The reasons for these consequences were iden- have shown that the proposed solution reduces
tified based on the compression forces found in the the performance of the drag chain conveyor.
material and the total drag resistance. High drag However, the effectiveness of reducing the drag
resistances of the conveyor are generated from resistance parameters was not sufficient. The pas-
a large amount of material lying directly on the sive element was placed very close to the trans-
bottom of the conveyor channel. When moving port channel. This positioning of the passive ele-
the drive elements, the drag chain conveyor must ment has resulted in a limitation of the continuous
overcome big dynamic shocks when picking up flow of material. During the DEM simulation, air
the material from the underground hopper. As a bridges were formed between the passive element
result of this, design modifications have been pro- and the underground hopper walls.

Fig. 5. DEM simulation of material discharge process with a passive element

249
Advances in Science and Technology Research Journal Vol. 11 (3), 2017

Fig. 6. Diagram of the underground hopper with belt conveyor

Since the structural modifications were not ef- conveyors with supporting rollers and calculation
fective enough, the type of conveyor was changed of the drive performance parameters.
from a drag chain conveyor to a belt conveyor. A DEM simulation of transport of material from
condition of the change from the drag chain con- an underground hopper using a belt conveyor has
veyor to the belt conveyor was to maintain the shown a significant reduction in resistance param-
conveyed quantity of material Q = 100 t·h-1, belt eters in the transport process. The average total
conveyor speed v = 0.85 m·s-1 and to reduce the force applied to the conveyor belt during the sim-
required power of the belt conveyor drive to max. ulation was only FT = 12000 N. After deducting
Pmax = 15 kW. The capacity of the underground the total force values from the DEM simulation,
hopper was reduced to 20 m3. Structural changes a control calculation of the required output of the
were made only in the virtual model. The sche- belt conveyor drive unit according to equation (1)
matic diagram of the underground hopper with a was again performed. The required drive power
belt conveyor is shown in Fig. 6. Subsequently, a for this type of delivery device was calculated at
new DEM simulation was performed. P = 13 kW. The calculated power meets the condi-
The theoretical design of the belt conveyor tional maximum drive power Pmax = 15kW. From
was carried out according to the ISO 5048 stan- the course of the DEM simulation, it was also ob-
dard. The standard regulates the design of the vious that with the new passive element the mate-
equipment for the smooth transport of loads, belt rial flow disturbances did no longer occur. Fig. 7

Fig. 7. DEM simulation of transport of material from an underground hopper using a belt conveyor

250
 Advances in Science and Technology Research Journal Vol. 11 (3), 2017

Fig. 8. The course of mass flow rate based on DEM simulation

shows a DEM simulation of the transport of ma- conveyor was replaced by a classic belt conveyor.
terial from an underground hopper using a belt This radical change has brought very positive re-
conveyor where the particle speed can be read at sults. The average total force in the conveyor belt
each instant of the transport process. during the simulation was only FT = 12000 N. It
During the DEM simulation, other parameters should be noted that all design modifications were
can be monitored to allow us to better understand made on a virtual 3D CAD model. DEM simula-
specific transport or storage processes. In our case, tion allows us to apply different design solutions
the mass flow rate that the belt conveyor is able to and verify process functionality. This verification
transport is monitored. Fig. 8 shows the mass flow method prevents the production of non-functional
rate dependency on time. The figure shows the av- prototype devices. Based on the results obtained
erage value of the mass flow rate, which ranges in this paper, it is possible to assume that DEM
from 95 t·h-1 to 100 t·h-1. This range fulfils the set simulations represent a very effective tool in the
mass flow rate condition of Q = 100 t·h-1. design of conveying and handling equipment in
the field of bulk materials.
CONCLUSIONS
Acknowledgements
This paper describes verifying the possible
use of computer DEM simulations for designing This paper was conducted within the frame-
completely new or optimising existing conveyor work of the project LO1404: Sustainable devel-
and storage devices for bulk materials. Using dis- opment of ENET Centre.
crete modelling, it has been possible to identify
problematic areas in transporting coal from an REFERENCES
underground hopper using the drag chain con-
1. Ekmann J. M. and Le P. H. Coal Storage and Trans-
veyor. It was found that the original drag chain
portation. Reference Module in Earth Systems and
conveyor was heavily underpowered. The DEM Environmental Sciences, from  Encyclopedia of
simulation showed very high drag resistances, on Energy, 2004, 551-58.
average around FT = 40000 N. In the first phase of 2. Cloete F.L.D. A drag chain feeder/conveyor based
optimisation, the effort to reduce the magnitude on standard engineering components. Powder
of the resistive resistances was only a minor de- Technology, 20(1), 1978, 21-27.
sign change. A passive element was placed in the 3. Molnar V., Fedorko G., Stehlikova B., Michalik P.
underground hopper above the conveyor transport and Weiszer M. A regression model for prediction
channel, which only slightly reduced the drag of pipe conveyor belt contact forces on idler rolls.
resistances. In the second phase, the drag chain Measurement, 46(10), 2013, 3910–3917.

251
Advances in Science and Technology Research Journal Vol. 11 (3), 2017

4. Molnar V., Fedorko G., Stehlikova B., Tomaskova 9. Bigos P., Kulka J., Kubin K., Mantic M. and Kopas
M. and Hulinova Z. Analysis of asymmetrical ef- M. Necessity of experiment for technical state ver-
fect of tension forces in conveyor belt on the idler ification of supporting structure of foundry crane
roll contact forces in the idler housing. Measure- before its reconstruction. Experimental Stress
ment, 52, 2014, 22-32. Analysis-TU Liberec, 2009, 1-7.
5. Debski H., Teter A., Kubiak T. and Samborski 10. Cleary P.W. Large scale industrial DEM modelling.
S. Local buckling, post-buckling and collapse of Engineering Computations, 21(2/3/4), 2004, 169-204.
thin-walled channel section composite columns 11. Guo Y.C., Wang S., Hu K. and Li D. Optimization
subjected to quasi-static compression. Composite and experimental study of transport section lateral
Structures, 136, 2016, 593-601. pressure of pipe belt conveyor. Advanced Powder
6. Jachowicz T., Garbacz T., Tor-Świątek A., Gajdoš Technology, 27(4), 2016, 1318-1324.
I., Czulak A. Investigation of selected properties of 12. Grima A.P., Fraser T, Hastie D.B and Wypych P.W.
injection-molded parts subjected to natural aging. Discrete element modelling: trouble-shooting and
International Journal Of Polymer Analysis And optimisation tool for chute design, 2011, 1-26.
Characterization, 20 (4), 2015, 07-315.. 13. Coetzee C. J. and Els D. N. J. Calibration of dis-
7. Molnar V., Fedorko G., Stehlikova B., Kude- crete element parameters and the modelling of silo
las L. and Husakova N. Statistical approach for discharge and bucket filling. Computers and Elec-
evaluation of pipe conveyor’s belt contact forc- tronics in Agriculture, 65(2), 2009, 198-212.
es on guide idlers. Measurement, 46(9), 2013, 14. Coetzee C. J. Review: Calibration of the Discrete Ele-
3127-3135. ment Method. Powder Technology, 310, 2017, 104-142.
8. Kral J. and Kral J. Verification of a three axis mill- 15. Grima A. and Wypych P. Discrete element simula-
ing machine accuracy in the process of complex tion of a conveyor impact-plate transfer: calibra-
shaped part production. Applied Mechanics and tion, validation and scale-up. Australian Bulk Han-
Materials, 474, 2014, 261-266. dling Review, 3, 2010, 64-72.

252