You are on page 1of 9

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Available online at www.sciencedirect.com

ScienceDirect
journal homepage: www.elsevier.com/locate/issn/15375110

Research Paper

Three-dimensional discrete element modelling


(DEM) of tillage: Accounting for soil cohesion and
adhesion
Mustafa Ucgul*, John M. Fielke, Chris Saunders
Barbara Hardy Institute, School of Engineering, University of South Australia, Mawson Lakes, SA 5095, Australia

article info

Recent studies have shown that the discrete element modelling (DEM) can effectively be

Article history:

used to model the soil-tool interaction if both an appropriate contact model and DEM

Received 21 June 2014

parameters are used. Ucgul, M., Fielke J.M and Saunders, C. (2014a) Biosystems Engineering.

Received in revised form

121: 105e117 showed that hysteretic spring contact model (HSCM) can give accurate pre-

2 November 2014

dictions of draft and vertical forces for a range of sweep tillage tool geometries operating in

Accepted 7 November 2014

a cohesionless soil. In this study a linear adhesion/cohesion model was integrated with the

Published online

HSCM to model the cohesive behaviour of soil and its interaction with a tillage tool. The
proposed contact model was validated with the direct shear tests performed and further

Keywords:

validation was achieved via DEM simulation of the interaction between soil and tillage tool

Draft

tests. The results showed that the suggested contact model can be used to predict both

Vertical force

tillage draft and vertical tillage forces for varying speed, operating depth, moisture content

Speed

and compaction levels. The results also showed that the DEM parameters determined in

Depth

our earlier paper can be directly used in the suggested model without further re-

Friction angle

calibration. It was observed that the soil bulk density has a significant effect on the predicted tillage forces and when the measured wet bulk density was used in the simulations
a good correlation with measured forces was able to be achieved. Thus, the method presented has good potential to accurately model tillage forces in a range of soil and operating
conditions.
2014 IAgrE. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

Using tillage tools to manipulate soil demands high energy


inputs due to; the large amount of soil which must be moved
and the inefficient methods of energy transfer to the soil.
Development of efficient tillage tools is essential to reduce
energy consumption and to increase agricultural production.

* Corresponding author.
E-mail address: Mustafa.Ucgul.@unisa.edu.au (M. Ucgul).
http://dx.doi.org/10.1016/j.biosystemseng.2014.11.006
1537-5110/ 2014 IAgrE. Published by Elsevier Ltd. All rights reserved.

Tillage forces (draft and vertical) are related to the tool geometry, soil and operating conditions. The quantification of
tillage forces with regard to different soil conditions is crucial
to design better tools (Zadeh, 2006). Internal friction, cohesion
and adhesion are vital soil parameters that affect tillage
forces. Internal friction is a characteristic of particles sliding
over each other. Cohesion is defined as the molecular force
between similar particles that acts to unite them. Adhesion is

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Nomenclature
Ac
Eeq
Fc/a
Fn
Fsn
Fdn
rc
req

Contact area (m2)


Equivalent Young's modulus (MPa)
Cohesion and/or adhesion force (N)
Normal total contact force (N)
Normal contact force (N)
Normal damping force (N)
Contact radius (m)
Equivalent radius (m)

Greek letters
x
Cohesion energy density (J m3)

defined as the molecular force of attraction between unlike


bodies that acts to hold them together.
In past DEM models of tillage, soil cohesion and adhesion
have not been applied simultaneously. These models can be
classified into three categories as per Tsuji et al. (2011). In the
first category, cohesive and adhesive forces due to a liquidbridge between particles are simulated based on the physical interactions that occur at a microscopic particle level. This
approach requires detailed examination of the inter-particle
interactions. However, due to the complexity of the equations and the irregularities of the particle sizes, contact
models in this category have not yet been employed in any
studies regarding DEM simulation of soil-tool interaction.
In the second category of models, the cohesive and adhesive forces are modelled based on the physical interactions at
a macroscopic level. Instead of a detailed experimental examination of inter-particle interactions, a calibration process
is used to determine parameters to match measured results
(Tsuji et al., 2011). The parallel bond contact model (PBCM) is a
common contact model used in this category and has been
employed in the DEM modelling of cohesive soil-tool interaction studies by Zhang and Li (2006), Zhang, Chen, Li, and Xu
(2008), Tamas and Jori (2010), Sadek, Chen, and Liu (2011), Mak,
Chen, and Sadek (2012), Okayasu et al. (2012) and Chen,
Munkholm, and Nyord (2013). In the PBCM, contact is
modelled by a finite-sized glue bond. The parallel bond can
resist the normal and the tangential forces up to a predefined
maximum normal and shear strength (EDEM, 2011). If the
normal stress exceeds the maximum normal stress or the
shear stress exceeds the maximum shear stress the bond
breaks and is removed from the model. Thereafter interaction
between particles is calculated by using the frictional characteristics of the Hertz-Mindlin contact model (Potyondy &
Cundall, 2004). When the cohesive strength between particles is zero, the PBCM reduces to the Hertz-Mindlin contact
model. When evaluated in the study of Ucgul, Fielke, and
Saunders (2014a), the Hertz-Mindlin contact model in DEM
simulations of soil tillage yields produced unrealistic vertical
force predictions. In order to overcome the problem that PBCM
does not accurately predict vertical forces, the stiffness of the
bond can be calibrated using trial and error methods to reduce
the normal and tangential forces acting upon the bond (Chen
et al., 2013; Mak et al., 2012). Using this method, stiffness must
be recalibrated whenever the bulk density of the granular
media changes. Although the PBCM provides satisfactory

299

results for modelling of cohesion, it can only be used for


modelling particleeparticle interactions and not adhesive
particleetool interactions.
In the third category of models, additional cohesion and
adhesion forces are directly added to the normal and/or
tangential contact forces. This category of model was used in
the DEM simulations of Asaf, Rubinstein, and Shmulevich
(2006, 2007), Shmulevich, Asaf, and Rubinstein (2007) and
Shmulevich (2009) where they added cohesive forces to the
normal and tangential contact forces. Momozu, Oida,
Yamazaki, and Koolen (2003) also used this third category as
the method to model adhesion. In their study an adhesion
force was added in the normal direction while modelling the
interaction between particle and tool. To date, there has not
yet been a study using DEM modelling of soil-tool interactions
which employs the contact models of category three to model
simultaneously both adhesion and cohesion.
The objective of this study was to evaluate the use of the
hysteretic spring contact model (HSCM), which was suggested
to be used for DEM modelling of the soil-tool tillage interactions by Ucgul et al. (2014a) and Ucgul, Fielke, and
Saunders (2014b) and add to it both adhesion and cohesion
using the third category method described above, to simulate
the passage of a tillage tool through a sandy loam soil. This
work was based on the hypothesis that a cohesive sandy loam
soil can be modelled by adding cohesion and adhesion forces
to the parameters of a sandy soil (beach sand) that had similar
particle sizes. This was first evaluated using a series of direct
shear tests of a sandy loam soil at varying moisture contents
and compaction levels that provided three different levels of
cohesion, as performed by Fielke (1988). Secondly, tillage
using a sweep tillage tool was simulated at different levels of
adhesion, cohesion, speed, depth and compaction levels to
match experimental results of Fielke (1988). If the measured
and predicted results agreed for both types of tests using the
one set of parameters, this would show that the model is valid.

2.
The addition of cohesion and adhesion to
the HSCM
In this study EDEM 2.4 software which was developed by DEM
Solutions Ltd. in Edinburgh (United Kingdom) was used. In
order to simulate tillage of a granular material the following
steps were undertaken: 1) the material and interaction properties of the particles were entered in to the software, 2) the
geometries designed using 3D modelling software (i.e. shear
boxes, tillage tool and the soil-bin) were imported, 3) the
particle shape was modelled and a particle factory used to
generate the particles for the simulation, 4) the simulation
time was entered and the simulation undertaken and 5) the
results were analysed using videos, particle movement and
force data. The main advantages of using EDEM 2.4 are that
can provide accurate solutions and has extensive postprocessing tools, for graphing data obtained from simulations and to create videos and JPG images of the simulation.
These help users to analyse the simulation in detail. In addition the package has an application programming interface
(API) allows users to write their own code.

300

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Table 1 e DEM parameters used in the simulations.


Property
Density of sand particles (kg m3)
Shear modulus of sand (MPa)
Poisson's ratio of sand
Density of steel (kg m3)
Shear modulus of steel (MPa)
Poisson's ratio of steel
Yield strength of the sand (MPa)
Particle size distribution
Coefficient of restitution of sandesand
Coefficient of restitution of sandesteel
Coefficient of friction of sandesteel
Coefficient of rolling friction of sandesteel
Coefficient of rolling friction of sandesand for 5 mm radius particle
Coefficient of rolling friction of sandesand for 10 mm radius particle
Coefficient of friction of sandesand for 5 mm radius particles
Coefficient of friction of sandesand for 10 mm radius particle
Integration time step (s) for 5 mm radius particle
Integration time step (s) for 10 mm radius particle

To model the sandy loam soil used by Fielke (1988), in both


the direct shear and tillage tool tests, the method of determining the DEM parameters described in Ucgul et al. (2014a)
were used. The DEM parameters were considered in two
categories, namely material and interaction properties. The
material properties were obtained from a combination of
measurements and data from literature, as shown in Table 1.
The coefficient of restitution of sandesand, the coefficient of
restitution of sandesteel, the coefficient of friction of
sandesteel and the coefficient of rolling friction of sandesteel were obtained from a combination of measurements
and data from the literature, as shown in Table 1. The
interaction properties of the coefficient of friction of
sandesand and the coefficient of rolling friction of
sandesand, and the integration time step required for a
timely solution, were calibrated to adjust them for the particle sizes used in the simulations by Ucgul et al. (2014a). This
calibration process was based on matching simulation results to measured results for the angle of repose, disc penetration and cone penetration tests for a cohesionless beach
sand. The DEM parameters determined by Ucgul et al. (2014a)
were also validated in the study of Ucgul et al. (2014b).
To integrate cohesion and adhesion with the HSCM, their
respective forces were added to the normal contact forces. As
friction was assumed to restrict the tangential particle motion
in the governing equations (HSCM equations), the adhesion
and cohesion forces were not added to the tangential contact
forces. The magnitude of the cohesion and adhesion forces
were calculated as per EDEM (2011),
Fc=a x$Ac

(1)

where x is the cohesion energy density which is defined as the


energy needed to remove a particle from its near neighbours
divided by the total volume of the removed particle. In this
study cohesive energy density was assumed to be equal to the
cohesive strength with Ac as the contact area which was
calculated as,
Ac

p$r2c

(2)

Value

Source

2600
4.3  104
0.3
7865
7.9  104
0.3
0.588
0.95e1.05
0.6
0.6
0.5
0.05
0.16
0.407
0.53
0.57
0.000022
0.00008

Das (1997)
Asaf et al. (2007)
Asaf et al. (2007)
Budynas and Nisbett (2012)
Budynas and Nisbett (2012)
Budynas and Nisbett (2012)
Ucgul et al. (2014a)
Selected
Das (1997)
Das (1997)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)
Ucgul et al. (2014a)

where rc is the contact radius and was defined by Hertz (1882)


as;
rc


1=3
3$req $Fsn = 4$Eeq

(3)

Fsn

are the equivalent radius, equivalent


where req, Eeq, and
Young's modulus, and normal contact force, respectively as
explained in Ucgul et al. (2014a). After calculating the contact,
damping and cohesion/adhesion forces, the total normal force
was determined as,
Fn Fsn Fdn Fc=a

(4)

3.
Validation of the model using a direct
shear test
3.1.

Method for modelling a direct shear test

The first stage was to validate the proposed model. Comparisons were made between the experimental results of Fielke
(1988) and DEM predictions for a direct shear test. Fielke
(1988) conducted a series of direct shear tests to measure the
cohesion of a sandy loam soil taken from UniSA's Tillage Test
Track (85% sand, 3% silt and 12% clay) at a range of moisture
contents and soil bulk densities. From that data set three
conditions were chosen that gave a spread in values of cohesion, as shown in Table 2.
The simulations were made using nominal 10 mm radii
spherical particles (to reduce the computation time) with the

Table 2 e Measured soil values from Fielke (1988).


Sample
1
2
3

Moisture
content (%)

Bulk density
(kg m3)

Cohesion
(kPa)

Friction
angle (o)

1
15
13

1320
1780
1880

3
15
22

36.8
32.2
34.9

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

actual radii varying from 0.95 to 1.05 times the nominal size. In
order to evaluate the effect of particle size on simulation results, several simulations were also repeated using 5 mm radii
particles. The particles were placed in a shear box of
200  100  100 mm (see Fig. 1a). To apply the desired normal
stress, the top wall of the shear box was set as a servo wall (as
per Sadek et al., 2011). The vertical velocity of the top wall was
changed automatically to maintain the desired normal stress
to match the normal stress settings of Fielke (1988) of 15.3, 51.6
and 103 kPa. The shearing process (Fig. 1b) was simulated by
keeping the upper shear box stationary and moving the lower
shear box horizontally at 0.020 mm s1 as used by Fielke
(1988). The shear stress was determined as the total force on
the opposing wall of the lower shear box divided by the area of
the sheared section.
For the DEM simulation, the bulk density of the particles in
the model was matched to that of the soil used in each
respective test. For the soil Sample 1 (Table 2) which had the
lowest density, the particles were able to be placed randomly
and achieve the required wet bulk density. For the Samples 2
and 3 (Table 2) which had a much higher values of density, the
particles needed to be arranged in a face-centred cubic packing arrangement (a particle is in the centre of the six sides of a

301

cube and one particle in each corner of the cube) with the
particle spacing chosen so as to achieve the required wet bulk
density. The particle packing condition was not arranged to
model the high moisture content of soil but only to achieve the
soil density measured by Fielke (1988). As the cohesive soil
consists of soil particles and water, the weight of the wet soil
was used to accurately represent the mass of soil being moved
by the tillage tool.

3.2.

Results of modelling the direct shear test

The measured and predicted normal stress versus shear stress


graphs are presented in Fig. 2 from which cohesive strength
and friction angle were able to be determined. As shown in
Fig. 2 and Table 3, there was only a small variation in values
between the experimental and simulation results. Thus, the
results show that to model a direct shear test using a sandy
loam soil a good prediction can be achieved using beach sand
DEM parameters, measured bulk density and cohesive energy
density equal to the measured cohesive strength. The results
showed that more accurate results could be obtained using
the smaller 5 mm radius particle size. However, the current
computational capacities make the use of these smaller particles not feasible for tillage studies because of the long
simulation times.

4.
Validation of the model using sweep
tillage
4.1.

Fig. 1 e DEM simulation of direct shear test; (a) start and (b)
end of simulation.

Method for modelling sweep tillage

The second stage, to validate the proposed model, examined


the interaction between soil and a sweep tillage tool that was
examined in the experiments of Fielke (1988). This study
evaluated a sweep tool, as shown in Fig. 3, with a 400 mm
width, 32 mm lift height, 10 rake angle, 70 sweep angle and
3 mm cutting edge height. The tests were carried out in the
University of South Australia Tillage Test Track with a
2500 mm wide and 300 mm deep sandy loam soil at 4, 8 and
12 km h1 speeds (Fielke, 1988). The tests were conducted at 3
moisture content levels (4, 8 and 12%) and at two levels of
compaction created using a 1 m diameter  2 m width roller
that was empty (1000 kg) and filled with water (3000 kg) with
the values for the properties of the soil shown in Tables 4 and
5, respectively. An extended octagonal ring transducer
(Godwin & Spoor, 1977) was used to measure the draft and
vertical forces.
For the simulations, the HSCM and the DEM parameters
shown in Table 1 were used. The cohesion and adhesion
values used in the simulations were those shown in Tables 4
and 5. A nominal 10 mm radius particle size was used with
the particles randomly generated in the range of 0.95e1.05
times the nominal size. The particles were placed in a virtual
bin of 2500 mm long  1500 mm wide  300 mm deep (Fig. 4).
The particles were placed randomly and were able to match
the required density for the 4% moisture content soil. The
particles needed to be placed in a face centred cubic packing
arrangement to achieve the higher bulk density of the 8% and
12% moisture content soils. For the 3000 kg compaction soil a

302

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Fig. 2 e Normal stress vs. shear stress graphs.

density variation with depth was measured (Table 5) and the


modelled soil was created with the same layered density
profile. Each simulation was repeated three times as there was
a variation in results and the averages of the simulation results were taken as the final result.

4.2.

The effect of adding cohesion and adhesion

In order to show the change in tillage forces by including


cohesion and adhesion to the HSCM, the measured and predicted results for 75 mm depth at 12 km h1 in a 12% moisture
content soil and 3000 kg compaction were compared. As
shown in Table 6 the inclusion of both the cohesion and
adhesion forces to the HSCM improved the prediction of the
draft and vertical forces.

4.3.

The effect of compaction level

The simulation results for 1000 and 3000 kg compactions were


compared at 75 mm operating depth and a range of speeds (4,

8 and 12 km h1) and moisture contents (4, 8 and 12%). As


shown in Fig. 5 the simulation results were in good agreement
with the experimental results measured by Fielke (1988).
As shown in Fig. 5, the draft and vertical forces were able to
be accurately predicted by using the HSCM with measured
values of cohesion, adhesion and bulk density as input parameters to the model. The results were able to predict the
trends with regards to varying levels of soil moisture content
and level of compaction which influence the soil's values of
cohesion and adhesion.

4.4.
The effect of moisture content, speed and depth on
tillage forces
The effect of varying the moisture content on tillage forces at
four different tillage depths were simulated for three different
tool speeds for the condition of 3000 kg compaction. As shown
in Fig. 6, the simulation results were very close to the experimental results of Fielke (1988). Figure 7 confirms the good
correlation between the measured and predicted draft and

Table 3 e Measured and predicted soil parameters of cohesion and friction angles for 5 mm and 10 mm nominal radii
particles.
Sample Moisture Density
Measured
Predicted
Predicted
Measured
Predicted
Predicted friction
content (kg m3) cohesion (kPa)
cohesion
cohesion (kPa) friction angle angle (o) for 5 mm
friction
(%)
(kPa) for 5 mm
for 10 mm
(o)
radii particles
angle (o) for
radii particles radii particles
10 mm radii
particles
1
2
3

1
15
13

1316
1780
1880

3
15
22

2.8
12.3
18.4

2.2
11.7
17.2

36.8
32.2
34.9

33.4
31.4
34.1

31.3
30.5
31.7

303

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Table 5 e Soil properties for 3000 kg compaction used by


Fielke (1988).
Moisture
content (%)
4

12

Fig. 3 e Definition of the share wing geometry (Fielke,


1988).

vertical forces. This good correlation shows that the HSCM


model and the parameters used can provide a good prediction
of tillage forces for a range of soil conditions. It was also
observed that the bulk density had a large influence on the
tillage forces (especially on draft forces) and when the
measured density was used there was a good correlation with
measured forces.
As shown in Fig. 6, at each moisture content, the draft force
increased linearly with depth. However, for the vertical force
the response with increasing depth varied depending upon
the moisture content. At the 4% moisture content condition
(low density and no adhesion) the increase in vertical force
with depth was not linear and virtually independent of speed.
As the moisture content increased (increased density and
adhesion) the change in vertical force with depth become
more linear.
For the 75 mm depth of tillage the tillage force response to
soil moisture content is shown in Fig. 8. The increase in draft
with an increase in moisture content from 4 to 8 % can be
explained by the increase in cohesive strength. The increase
in draft from 8 to 12 % moisture content can be explained by
the lower draft expected from a reduction in soil strength
being counteracted by a greater increase in draft from the
increase in adhesion. The increase in vertical down force with
an increase in moisture content from 4 to 8 % can be explained
by the higher strength and increased bulk density of soil
increasing the downward force on the tool. The small increase
in vertical down force from 8 to 12 % moisture content can be
explained by the increase in soil density increasing the

Soil Layer
(mm)

Wet bulk
density
(kg m3)

Cohesive
strength
(kPa)

Adhesive
strength
(kPa)

0e25
25e50
50e75
75e100
0e25
25e50
50e75
75e100
0e25
25e50
50e75
75e100

1269
1391
1430
1430
1540
1644
1689
1689
1742
1837
1864
1864

7.3

7.9

5.6

downward force on the tool but this effect was reduced by the
increase in adhesion which acts on the tool's wing in an upward direction. Hence, while the force response to changes in
soil moisture content is complex, it can be modelled using the
appropriate parameters.

5.

Conclusion

In this study the HSCM was used to model soil. By using the
particle contact parameters obtained for a cohesionless beach
sand, and adding normal contact forces for cohesion and
adhesion of the soil, the model was able to accurately simulate
direct shear tests and sweep tillage measured in a sandy loam

Table 4 e Soil properties of top 75 mm soil layer for


1000 kg compaction level used by Fielke (1988).
Moisture
content (%)
4
8
12

Wet bulk
density
(kg m3)

Cohesive
strength
(kPa)

Adhesive
strength (kPa)

1338
1605
1781

4.9
5.3
4.6

0
0
3

Fig. 4 e Screen captures of a soil-tool simulation taken


from the EDEM software.

304

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Table 6 e Comparison of the results with and without


cohesion/adhesion for 75 mm depth, 12 km h1 speed
and 3000 kg compaction.
Draft
Downward vertical
force (N)
force (N)
HSCM
HSCM Cohesion
HSCM Cohesion Adhesion
Fielke (1988)

768
782
820
840

276
301
287
241

soil. The direct shear test simulations, using 5 and 10 mm


nominal radius particles, showed that when measured values
of the cohesive strength and bulk density are used the model
could predict both cohesive strength and friction angle. The
most accurate results were obtained when using smaller

particles but this greatly increased computation time. The


sweep tillage simulations showed that even when using
10 mm nominal radius particles the model could still accurately predict draft and vertical forces with regards to both
changes in depth and speed for a range of soil conditions that
result from various levels of soil moisture content and
compaction. In comparing the measured and predicted tillage
forces for these various conditions for a 1:1 slope, the draft
force line had R2 0.93 and simultaneously the vertical force
line had R2 0.89. This indicates a very accurate prediction of
tillage forces. Thus, the HSCM with the addition of cohesion
and adhesion contact forces, and using actual soil parameters
such as wet bulk density, is a valid method for predicting both
draft and vertical tillage forces in a sandy loam soil. Further
work is now needed to look at other soil types which contain a
greater percentage of clay.

Fig. 5 e Effect of compaction on draft and vertical forces for 75 mm depth of tillage at varying moisture content levels (a) 4%,
(b) 8%, and (c) 12%.

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

305

Fig. 6 e Effect of tool speed and tillage depth on draft and vertical forces at 3000 kg compaction and moisture contents (a) 4%,
(b) 8% and (c) 12%.

Fig. 7 e Correlation at 3000 kg compaction between the measured and predicted (a) draft and (b) vertical forces.

306

b i o s y s t e m s e n g i n e e r i n g 1 2 9 ( 2 0 1 5 ) 2 9 8 e3 0 6

Fig. 8 e Effect of the moisture content at varying tool speeds for 75 mm depth and 3000 kg compactions on (a) draft and (b)
vertical forces.

Acknowledgements
The authors acknowledge the support of the University of
South Australia for granting of a post graduate scholarship to
Mustafa Ucgul and Australia's Grains Research and Development Corporation (GRDC) project USA00005 for funding the
computer and software.

references

Asaf, Z., Rubinstein, D., & Shmulevich, I. (2006). Evaluation of


link-track performances using DEM. Journal of Terramechanics,
43(2), 141e161.
Asaf, Z., Rubinstein, D., & Shmulevich, I. (2007). Determination of
discrete element model parameters required for soil tillage.
Soil and Tillage Research, 92(1e2), 227e242.
Budynas, R. G., & Nisbett, K. J. (2012). Shigley's mechanical
engineering design. New York: McGraw-Hill Education.
Chen, Y., Munkholm, L. J., & Nyord, T. (2013). A discrete element
model for soil-sweep interaction in three different soils. Soil
and Tillage Research, 126, 34e41.
Das, B. M. (1997). Advanced soil mechanics. Taylor&Francis.
EDEM. (2011). EDEM theory reference guide. Edinburgh, UK: DEM
Solutions.
Fielke, J. M. (1988). The influence of the geometry of chisel plough share
wings on tillage forces in sandy loam soil. The Department of Civil
and Agricultural Engineering University of Melbourne. Master
of Engineering.
Godwin, R. J., & Spoor, G. (1977). Soil failure with narrow tines.
Journal of Agricultural Engineering Research, 22(4), 213e228.
Hertz, H. (1882). Ueber die beruhrung fester elatischer korper.
Journal fur die Reine und Angewandte Mathematik, 92, 156e171.
Mak, J., Chen, Y., & Sadek, M. A. (2012). Determining parameters
of a discrete element model for soil-tool interaction. Soil and
Tillage Research, 118, 117e122.
Momozu, M., Oida, A., Yamazaki, M., & Koolen, A. J. (2003).
Simulation of a soil loosening process by means of the
modified distinct element method. Journal of Terramechanics,
39(4), 207e220.

Okayasu, T., Morishita, K., Terao, H., Mitsuoka, M., Inoue, E., &
Fukami, K. (2012). Modeling and prediction of soil cutting
behaviour by a plow. In International Conference of Agricultural
Engineering. Valencia, Spain.
Potyondy, D. O., & Cundall, P. A. (2004). A bonded-particle model
for rock. Journal of Rock Mechanics and Mining Sciences, 41(8),
1329e1364.
Sadek, M. A., Chen, Y., & Liu, J. (2011). Simulating shear behavior
of a sandy soil under different soil conditions. Journal of
Terramechanics, 48(6), 451e458.
Shmulevich, I. (2009). State of art modelling of soil tillage
interaction using discrete element method. In Presentation at
the 18th Triennial Conferance. ISTRO June 15e19, Izmir, TurkeyOral presentation unpublished.
Shmulevich, I., Asaf, Z., & Rubinstein, D. (2007). Interaction
between soil and a wide cutting blade using the discrete
element method. Soil and Tillage Research, 97(1), 37e50.
Tamas, T., & Jori, I. J. (2010). DEM analysis of the soil-tool (sweep)
interaction. In 17th World Congress of the International
Commission of Agricultural and Biosystems Engineering (CIGR),
Quebec City, Canada.
Tsuji, T., Nakagawa, Matsumoto, N., Kadono, Y., Takayama, T., &
Tanaka, T. (2011). 3-D DEM simulation of cohesive soilpushing behavior by bulldozer blade. Journal of Terramechanics,
49(1), 37e47.
Ucgul, M., Fielke, J. M., & Saunders, C. (2014a). Three-dimensional
discrete element modelling of tillage: determination of a
suitable contact model and parameters for a cohesionless soil.
Biosystems Engineering, 121, 105e117.
Ucgul, M., Fielke, J. M., & Saunders, C. (2014b). 3D DEM Tillage
simulation: validation for a sweep tool for a cohesionless soil.
Soil and Tillage Research, 144, 220e227.
Zadeh, S. R. A. (2006). Modelling of energy requirements by a narrow
tillage tool. Saskatoon, Canada: Department of Agricultural and
Bioresource Engineering, University of Saskatchewan. PhD:
190.
Zhang, R., Chen, B., Li, J., & Xu, S. (2008). DEM simulation of clod
crushing by bionic bulldozing plate. Journal of Bionic
Engineering, 5(1), 72e78.
Zhang, R., & Li, J. (2006). Simulation on mechanical behavior of
cohesive soil by distinct element method. Journal of
Terramechanics, 43(3), 303e316.

You might also like