b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Available online at www.sciencedirect.com

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

Research Paper

Computer simulations to maximise fuel efficiency

and work performance of agricultural tractors in
rotovating and ploughing operations
Jin W. Lee, Jae S. Kim, Kyeong U. Kim*
Department of Biosystems & Biomaterials Science and Engineering, Seoul National University, Seoul 151-921, South
Korea

article info

This study was conducted to investigate the effects of five control variables of a tractor:

Article history:

ballast, tyre inflation pressure, transmission gear, engine speed, and work load on three

Received 5 June 2014

fuel efficiency parameters: fuel consumption per work hour (FC), fuel consumption per

Received in revised form

tilled area (FCA) and specific volumetric fuel consumption (SVFC). This was done for

27 August 2015

moldboard ploughing and rotovating operations by computer simulation. A tractor model

Accepted 26 November 2015

was constructed with four sub-models: engine and power train, fuel consumption, tractive

Published online xxx

performance, and draught and power requirement. The simulated fuel efficiency values
were in a range of 3.3e6.5% error in average when compared with those obtained from field

Keywords:

experiments carried out in a paddy field under the same operational conditions. Based on

Agricultural tractor

these results, the tractor model was considered acceptable for simulations to find a general

Fuel efficiency

relationship between the fuel efficiency parameters and the control variables.

Computer simulation

Using the tractor model, 162 simulations were performed under the various combina-

Mouldboard ploughing

tions of the control variables on the basis of a full factorial design. The simulation results

Rotovating

were used to develop linear regression models from which strategies can be established to
maximise fuel efficiency. The best strategy reduced FC, FCA, and SVFC by 81.3, 61.1, and
52% under ploughing, and by 58.9, 75.7 and 28.6% under rotovating operations, respectively,
when compared with those for the worst strategy.
2015 IAgrE. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

The tractor is one of the largest fuel consumers among

agricultural machines. In Korea, agricultural tractors
consumed 329,000 kL in 2010 which was about 17% of the
total tax-free fuel allocated for agricultural use. This consumption was a 4% increase compared to 2009 (NACF.,
2010). The tractor population is growing continuously and
* Corresponding author.
E-mail address: kukim@snu.ac.kr (K.U. Kim).
http://dx.doi.org/10.1016/j.biosystemseng.2015.11.012
1537-5110/ 2015 IAgrE. Published by Elsevier Ltd. All rights reserved.

the tractor power also appears to increase in recent years.

This trend is expected to continue and so is tractors' fuel
consumption in Korea. In addition, tractors should meet
emission standards both in domestic and overseas markets.
In response to such circumstances, tractor manufacturers
are required to develop technologies to increase fuel efficiency of tractors and at the same time to reduce their
emissions without any power loss.

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Nomenclature
a
A, B,C
Ar
b
Bn
CI
Cr
d
ei
Fd
Gi
GTR
h
H
hz
Ki
L1
L2
L3
L4
MRR
mt
Nt
PL
Pr
Pw

Acceleration of tractor, m s
2

Constants depending on implement types
Vertical projected area of implement, m2
Unloaded tyre section width, m
Mobility number
Cone index, kPa
Specific draught torque, Nm m
2
Unloaded tyre diameter, m
Longitudinal offset of soil reaction to wheel (i r
for rear, f for front), m
Draught of implement, N
Speed reduction ratio of ith gear, decimal
Gross traction ratio
Unloaded tyre section height, m
Gross traction force, N
Ground clearance of implement, m
Soil texture parameter (i 1 for fine, 2 for medium
and 3 for course soils)
Longitudinal distance between rear axle and mass
center of implement, m
Longitudinal distance between mass center of
tractor and rear axle, m;
Longitudinal distance between mass center of
tractor and front axle, m
Longitudinal distance between front axle and
mounting position of ballast, m
Motion resistance ratio
Tractor mass, kg
Rotational speed (i e at engine, R at rear tyre, F at
front tyre), RPM
Tractor power, kW
PTO power, kW
inflation pressure of tire, Pa

Various studies have been conducted to improve fuel efficiency of tractors. Gear up and throttle down (GUTD) is a wellknown operational technique to reduce fuel consumption of
tractors in field operations, particularly during ploughing and
transportation.
Grisso and Pitman (2001) analysed the effects on tractor's
fuel consumption of engine speed and gears at the same
drawbar power using performance test data published by the
Nebraska Tractor Test Laboratory. They realised that 5e30% of
fuel can be saved if the engine speed is reduced by 20e30% and
the drawbar is loaded within 75% of the maximum engine
power. Peca et al. (2010) analysed the effects of engine speed
and gear on the overall power efficiency in tractor operations.
They reported that the overall power efficiency increased by
10e20% when the engine speed was reduced from 2200 to
1750 rpm. Proper adjustments of tyre inflation pressure and
ballast also contributed to improving fuel efficiency of tractors
rquez, 2009). Gee(Serrano, Peca, Marques da Silva, & Ma
Clough, Pearson, and McAllister (1982) found that if a tyre is
loaded in a range of 70e140% of the optimum dynamic load,
the tractive force would not be significantly reduced relative to
the maximum traction. Lancas, Upadhyaya, Sime, and Shafii

ri
rs
s
T
Ti
V
va
vt
Wb
wi
Wi
Wj
wk
Wt
d
h
u

Rolling radius of tyre (i r for rear tyre, f for front

tyre), m
Unloaded radius of tyre, m
Slip of driving tyres, decimal
Ploughing depth, m
Torque (i e at engine, R at rear tyre, F at front
tyre), N m
Tractor velocity, km h
1
Actual velocity of tractor, m s
1
Theoretical velocity of tractor, m s
1
Weight of ballast, N
Width of tillage implement, m
Weight of implement, N
Vertical tyre load (j r for rear, f for front), N
Weight acting on the axle in % of total weight
(k R for rear axle, F for front axle), %
Total tractor weight, N
deflection, m
Transmission efficiency, decimal
Rotational speed of rotovator shaft, rpm

Abbreviations
Ba
Ballast
CG
Center of gravity
DF
Degree of freedom
Es
Engine speed
FC
Fuel consumption in, L h
1
FCA
Fuel consumption per tilled area, L ha
1
FE
Fuel efficiency parameters
SVFC
Specific volumetric fuel consumption, L kW
1 h
1
Ti
Tire-inflation pressure
Tr
Transmission gear ratio
Wl
Work load

(1996) reduced fuel consumption by 6e20% and increased

productivity by 4.6e7.5% by maintaining proper tyre inflation
pressure during stubble-disking operations in tilled moist clay
soils. Casady (1997) developed guidelines for the management
of tyre inflation pressure and ballast to maximise tractive efficiency, to minimise soil compaction and to increase the life
of a tractor's drivetrain and profitability. A variety of educational materials have been developed to help tractor operators
increase fuel economy (Fulton, Raper, McDonald, & Tyson,
2006; Firestine, Furrey, Aslin, Crivella, & Sautter, 2007).
Tractor manufacturers also developed and commercialised
information and support systems such as ACET of Renault,
INFORMAT of Steyer, ECOplus of Kubota, ECOTRONIK of
Steyr, and Efficient Technology of Fendt to improve fuel
efficiency.
Park et al. (2010a) developed an eco-driving system to help
tractor operators select the proper gear and engine speed according to the work loads. They reduced fuel consumption by
an average of 69% in ploughing and 54% in rotovating operations with the aid of the system. Lee, Kim, Kim, and Choi (2011)
conducted field experiments under various soil conditions
and found that proper adjustment of tyre inflation pressure

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

and weight distribution on axles could reduce the fuel consumption by 10e65% in ploughing and 10e20% in rotovating in
flooded paddy fields.
Although it has been known that ballast, tyre inflation
pressure, engine speed and transmission gear affect fuel
consumption of tractors, few studies have dealt with how
such factors could be integrated into an automatic control
system of tractors to increase fuel efficiency. Zhang and
Chancellor (1989) used a control system that could adjust
ballast so that a desired vertical load could be maintained on
the front wheel of a 2WD tractor and increased fuel efficiency
by approximately 5e15% in tillage operations. Also, it has not
been known how such factors could be optimised according to
given field and load conditions. Developing an integrated
control system requires an understanding of interactions between factors affecting fuel consumption and how those
factors can be controlled most effectively. Computer simulation has been regarded as the most appropriate method of
doing this.
A number of researchers since the 1970s have proposed
theoretical and empirical models to predict the tractive
performance and fuel efficiency of tractors. Zoz (1970)
developed a graphical method of predicting drawbar pull,
speed and slip of driving tyres of 2WD tractors under the
predetermined soil conditions. Wismer and Luth (1973) proposed an empirical equation for tractive performance of
tyres on cohesive soils. Their equations were later revised by
Brixius (1987) to expand their applications to bias-ply tyre
and by Upadhyaya, Wulfsohn, and Jubbal (1989) to radial-ply
tyre. Evans, Clark, and Manor (1991) modified Brixius equation to improve its predictions for tractors on grass-covered
surfaces. They also developed a ballast selection map using
the modified equation. The Brixius equation was also modified by Kumar and Pandey (2009) to take soil characteristics
into consideration. Al-Hamed and Al-Janobi (2001), and
Pranav and Pandey (2008) have applied Visual C to coding
of computer simulation of the tractive performance and
ballast management. Simulation of tractor performance was
also conducted on the Matlab-Simulink environment by
Kolator and Biaobrzewski (2011) and on the Visual Basic by
 n, Linares, and Me
ndez (2008). Sahu and Raheman
Catala
(2008) also developed a Visual Basic-based decision support
system to help select a suitable tillage implement to 2WD
tractors and predict the field performance of the selected
tractoreimplement combinations. The other works relevant
to the field performance of tractors included those by
Serrano et al. (2003, 2007) and Lacour, Burgun, Perilhon,
Descombes, and Doyen (2014). However, no studies have
considered both of fuel consumption and work performance
at the same time to investigate the interrelations between
parameters representing their characteristics.
The objectives of this study were (a) to develop a tractor
model capable of simulating both fuel consumption and
work performance, (b) to investigate the effects of ballast,
tyre inflation pressure, transmission gear and engine speed
on fuel consumption and work performance using the tractor
model, and (c) to develop a strategy for controlling these
parameters most effectively to maximise fuel efficiency
within an acceptable work performance.

2.

Materials and methods

2.1.

Tractor simulation model

A tractor simulation model was built on the basis of a

48 kW TYM T623 MFWD diesel tractor (Tongyang Moolsan Co.,
Seoul, South Korea). Specifications of the T623 tractor are
given in Table 1.
The simulation model was built with four sub-models for
engine and power train, fuel consumption, tractive performance, and draught and power requirement. Each sub-model
was developed as follows:

2.1.1.

Engine and power train

The sub-model for the engine and power train was developed using the PTO performance data of the T623 tractor
tested by the OECD test code (OECD Code 2, 2010) considering a power transmission efficiency of 90%. The full-load
torque with full throttle was expressed as a second order
polynomial of engine speed, and the partial-load torque
both with full and reduced throttles by a linear function of
engine speed.
The torque available at the driving tyres was modelled
mathematically using the engine torque as follows:
TR

hTe wR Gi
2

(1)

TF

hTe wF Gi rf
2rr

(2)

NR

Ne
Gi

(3)

NF

NR rr
rf

(4)

where T is torque in N.m (eat engine, R at rear tyre, F at front

tyre); w is weight acting on the axle in % of total weight (R
for rear axle, F for front axle); h is transmission efficiency in
decimal; Gi is speed reduction ratio of ith gear; N is rotational speed in rpm (e at engine, R at rear tyre, F at front
tyre); r is rolling radius of tyre in m (r for rear tyre, f for front
tyre).

2.1.2.

Fuel consumption

A method proposed by Kim, Kim, and Kim (2010) was used

to construct a sub-model for fuel consumption under the
given load conditions. Contours of equal specific volumetric
fuel consumption (SVFC) of the model tractor were constructed as a function of engine speed and power as shown
in Fig. 1. SVFC was defined as the volume of fuel consumed
per unit of power produced in an hour. These contours
were used to predict the amount of fuel required for the
model tractor to perform the rotovating and ploughing
operations. The contour map was read by the tractor
simulation model through the lookup table module of
Matlab Simulink R2008a (The MathWork, Inc., Natick,
Mass., USA). Interpolation was also used to locate the SVFC
in between two contours.

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Table 1 e Specifications of MFWD tractor used for modelling in this study.

Item

Specification

Unballasted weight, kN
Weight distribution (Front:Rear), %
Transmission (main/sub/creep)
Engine

24.4
47.6:52.4
Forward 24, Backward 24 (4/3/2)
Caterpillar C3.3T 2232/2200
48 kW @ 2200 rpm
47.61
540, 733 @ rated engine speed
11.2e20 8PR (drive tyres)
14.9e30 8PR (drive tyres)
1130
984
727

Manufacturer and Model

Rated power, kW
PTO
Rated power, kW
Rated speeda, rpm
Tyre
Front
Rear
Longitudinal distance from CG to front axle, mm
Longitudinal distance from CG to rear axle, mm
Longitudinal distance from ballast to front axle, mm
a

Rated PTO speed was determined for different PTO gear ratios at the rated engine speed.

tyres and the smaller of which was determined as tractive

force of the model tractor. Motion resistance was included to
take account of soil compaction. Tyre deflection was estimated using Eq. (8) which was developed in this study on the
assumption that the tyre inflation pressure equals its groundcontact pressure which is related to the tire and soil
characteristics.
v

u
2 !

u
W
d rs
t r2s

2pw b

(8)

where Pw is tyre inflation pressure in Pa; rs is unloaded radius

of tyre in m.
Vertical loads acting on the tractor tyres vary with ballast
and weight transfer due to attached implements. The vertical
tyre load was modelled mathematically by using Eqs. (9)e(11).

Fig. 1 e Contours of equal specific volumetric fuel

consumption of the model tractor.

Wr








Wt L3 ef
Wb L4
ef Wi L1 L2 L3 ef Fd hz



L2 L3
er ef
(9)

2.1.3.

Tractive performance

Tractive performance model was constructed by using the

empirical equations for gross traction and motion resistance
developed by Brixius (1987). The equations for bias-ply tyres
were expressed as a function of tyre slip and mobility number
as follows:





GTR 0:88 1
e
0:1Bn 1
e
7:5s 0:04
MRP

1 0:5s
p 0:04
Bn
Bn

!


1 5 hd
CI:b:d
Bn 1000
W
1 3 bd

(5)

(6)

(7)

where, GTR is gross traction ratio; MRR is motion resistance

ratio; Bn is mobility number; W is vertical wheel load in N; CI is
cone index in kPa; b is unloaded tyre section width in m; d is
unloaded tyre diameter in m; h is unloaded tyre section height
in m; d is tyre deflection in m; s is slip of driving tyres
(decimal).
Net traction force calculated from Eqs. (5) and (6) was
compared with that delivered by the engine to the driving

Wf

Wt L2
er Wb L2 L3 L4
er
Wi L1 er
Fd hz



L2 L3
er ef
(10)

ef MPRf rf ; er MPRr rr

(11)

where W is vertical tire load in N (r for rear, f for front); Wt is

total tractor weight in N; L1 is longitudinal distance between
rear axle and mass center of implement in m; L2 is longitudinal distance between mass centers of tractor and rear axle in
m; L3 is longitudinal distance between mass centers of tractor
and front axle in m; L4 is longitudinal distance between front
axle and mounting position of ballast in m; e is longitudinal
offset of soil reaction to wheel in m (r for rear, f for front); Wb is
weight of ballast in N; Wi is weight of implement in N; hz is
ground clearance of implement in m; Fd is draught of tillage
implement in N.
Equation of motion for the tractor model in the longitudinal direction can be written by Newton's second law of motion
as follows:
mt a H
Fd MRR  Wt

(12)

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

where mt is tractor mass in kg; H is gross traction force in N; a

is acceleration of tractor in m s
2.
Integrating the equation of motion with respect to time
yields velocity of the tractor as follows:
va

1
mt

Z
H
Fd
MRR  Wt dt

(13)

Slip of driving tyres can then be determined as follows:

s1

va
vt

(14)

where va is actual velocity of tractor in m s
1; vt is theoretical

velocity of tractor in m s
1 and is the product of angular velocity of the tyre and its rolling radius in a self-propelled
condition on a hard surface.

2.1.4.

Draught and power requirement

Draught of implement was estimated using the ASABE Standard D497.5 (ASABE, 2006) where it is expressed as



Fd 100  Ki A BV CV2 wi t

(15)

where Ki is soil texture parameter (i 1 for fine, 2 for medium,

and 3 for coarse soils in this study); A, B, C are constants
depending on implement types; wi is width of tillage implement in m; Vis tractor velocity in km h
1; t is ploughing depth
in m.
PTO power to drive the implements was estimated by
Pr

pAr Cr u
30

(16)

where Pr is PTO power in kW; Ar is vertical projected area of

implement in m2 and is the product of the tillage depth and
width; Cr is specific draught torque in Nm m
2; u is rotational
speed of rotovator shaft in rpm.
The specific draught torque for different soil types incorporated in the model is given in Table 2 (Kim, Kim, Koo, & Park,
2009).

2.2.

Development of simulation program

The tractor model was computer programmed with Matlab,

Simulink R2008a (The MathWork, Inc., Natick, Mass., USA) and
LabVIEW 2009 (National Instruments Corp., Austin, Texas,
USA). The Simulink solves the mathematical equations of the
sub-models and transmits the results to the other submodels. Simulation of tractor operations was performed in
LabVIEW and its results were monitored in the virtual
terminal.

2.3.

Field experiment for model validation

To verify the tractor model, ploughing and rotovating operations were conducted using the T623 tractor in a paddy field.
The performance data of the tractor including the engine
torque, engine speed, PTO torque, PTO speed, travel speed,
and fuel consumption were collected during the operations
and compared with those simulated by the tractor model
under the same field and operational conditions. Specifications of the implements used for the ploughing and rotovating
operations are given in Table 3. Average water content of 5
randomly collected soil samples from the surface of the paddy
field was determined to be 38.9% (d.b.) by the oven drying
method where the soil samples were subjected to drying at
105  C for 16 h and the percentage moisture was obtained as
the original mass minus the dry mass divided by the dry mass
of the samples. The soil texture was classified as loam according to the USDA method. Average soil cone index taken at
8 to 10 randomly selected locations in the field is given in
Table 4.
The ploughing and rotovating operations were conducted
under eight operational conditions in Table 5 because these
conditions could be easily achieved on the T623 tractor used in
this study. Vertical loads acting on the front and rear tyres
were measured from the tractor with no implement on the
horizontal flat surface and adjusted by adding the ballast
weight of 313.9 N to the front-end. The load distribution between the front and rear axles was 45:55 with no ballast
weight. The distribution was changed to two levels, 49:51 and
53:47, by adding four and eight ballast weights respectively.
The tyre inflation pressure was varied at two levels; 117 and
176 kPa for the front tyre and 83 and 124 kPa for the rear tyre
which were respectively 100 and 150% of the pressure recommended for those tyres by the Korean Industrial Standard
(KIS, 2009). The engine speed was set to two levels; the rated
speed (2200 rpm) and the speed at the maximum torque under
full load (1400 rpm). Each operational condition was applied
on the T623 tractor with no load before starting the
operations.
Fuel consumption was determined by measuring the difference in the flow rates of fuel entering and leaving the fuel
injection nozzle of the fuel supply system using a pulseoutput flow meter. Engine torque and speed were measured
at the input shaft of the transmission using a full-bridge strain

Table 3 e Specification of implements used for ploughing

and rotovating operations.
Item

Table 2 e Specific draught of soils in the simulation

model.
Soil type or
tractive surface
Sandy loam
Loam
Clay loam
Grass

Specific draught
(N m
2)

Specific draught
torque (Nm m
2)

25,000e42,000
35,000e49,000
49,000e70,000

600e1500
1500e2900
2500e3400
2900e3900

Length  Width
 Height (mm)
Type
Ploughing width (mm)
Center of resistance
rearward from
rear axle (mm)

Specification
Rotovator

Moldboard
plough

970  2035  1130

2140  1805  1165

Trailed
1800
1318

Mounted
1810
1583

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Table 4 e Soil properties of paddy field.

Plot

Soil texture

Paddy field

Average water
content, (% d.b)

Loam

38.9

Average cone index by depth (kPa)

0 mm

50 mm

100 mm

150 mm

200 mm

250 mm

201

615

732

735

1485

2425

Table 5 e Operational conditions for model verification.

Condition

Vertical axle load

(Front:Rear in %)

1
2
3
4
5
6
7
8

45:55
45:55
49:51
49:51
53:47
53:47
53:47
53:47

Transmission gear
(gear ratio)

M-2
M-3
L-3
M-3
L-3
M-3
M-2
M-3

(122.2)
(79.4)
(253.7)
(79.4)
(253.7)
(79.4)
(122.2)
(79.4)

gauge system and a magnetic pick-up sensor. A flange-type

transducer was mounted on the counter shaft between the
PTO and implement and used to measure the PTO torque. The
PTO speed was calculated by multiplying the engine speed
and the PTO gear ratio. The tillage depth, a vertical distance
from the field surface to the furrow bottom, was measured at
five to eight randomly selected locations in the field using a
wooden ruler after each test operation was completed. All
performance data were collected when the tractor operated at
a constant speed set by the levels of the engine speed and
transmission gear.

2.4.

Engine speed (rpm)

Computer simulation

Computer simulations of the tillage and rotovating operations

were conducted to estimate the fuel consumption per work
hour (FC), fuel consumption per tilled area (FCA) and specific
volumetric fuel consumption (SVFC) with the engine speed
(Es), transmission gear ratio (Tr), tyre-inflation pressure (Ti),
ballast (Ba) and work load (Wl) varied at three levels as shown
in Table 6. Each level of the Es, Tr, Ti, and Ba was determined as
the same as in the field test for the model validation. In
addition, an engine speed of 1400 rpm and tyre inflation
pressures of 59 kPa for the front and 41 kPa for the rear were
added. They are respectively the intermediate speed between
the rated engine speed and the speed at the maximum torque,
and the tyre inflation pressure equal to the 50% of the KIS

2200
1400
1400
2200
1400
2200
2200
1400

Tyre inflation
pressure (kPa)
Front

Rear

117
117
176
176
117
117
176
176

83
83
124
124
83
83
124
124

(Korean Industrial Standard) recommendation. The transmission gear was set to the gears that were commonly used
for field operations at rated engine speed. The work load was
varied by changing the tillage depth at two levels of 100 and
150 mm for the ploughing, and the PTO speed at two levels of
540 and 733 rpms for the rotovating operations. The simulation was performed with four control variables each consisting of three levels under two workloads in the ploughing and
rotovating operations, resulting in a total of 162 simulations.

2.4.1.

Statistical analysis of simulation results

The factorial analysis of variance was performed on the

simulation results using the SAS 9.3 statistical program (SAS
Institute Inc., Cary, NC, USA). The fuel efficiency parameters
FCA and SVFC were calculated from FC respectively by using
Eqs. (17) and (18), and FC, FCA and SVFC were considered as
the dependent variables for the analysis.
FCA

10  FC
3:6  va  wi

SVFC

(17)

FC
pL

(18)

where FC is fuel consumption per work hour in L h
1; FCA is

fuel consumption per tilled area in L ha
1; SVFC is specific
volumetric fuel consumption in L (kWh)
1;PLis tractor power
in kW.

Table 6 e Levels of control variables for computer simulations.

Level

1
2
3

Weight distribution
(Front:Rear in %)

45:55
49:51
53:47

Transmission gear (gear ratio)

L-3 (253.7)
M-2 (122.2)
M-3 (79.4)

Engine speed (rpm)

1400
1800
2200

Tyre inflation
pressure (kPa)
Front

Rear

59
117
176

41
83
124

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

The control variables Es, Tr, Ti, Ba, and Wl were considered
as the independent variables. The factorial analysis was then
to test the null hypothesis that the control variables have the
same effect on the fuel efficiency parameters at a confidence
level of 95%. This analysis considered only the main effects of
the control variables because the effects of their interactions
were insignificant.

2.4.2. Multiple regression analysis for integrated control

strategy
To establish a control strategy for maximising fuel efficiency,
a multiple regression analysis was performed between the
fuel efficiency parameters and the control variables identified by the analysis of variance as significantly affecting
them in the ploughing and rotovating operations. Based on
the regression analysis, fuel efficiency parameters were
expressed as a linear function of the control variables Es, Tr,
Ti, Ba, and Wl as follows:
FE a0 a1 Ba a2 Ti a3 Tr a4 Es a5 Wl

(19)

ploughing operations were simulated under eight operational

conditions in Table 5. The simulated values were compared
with those obtained from field experiment under the same
operational conditions as shown in Tables 7 and 8. The percentage errors of the all simulated values ranged from 0.1 to
17.2% in the rotovating and from 0.4 to 10.9% in the ploughing
operations. Although the error range was greater in the
simulation of the rotovating operation, an average simulation
error was greater in the ploughing operation. An average error
of each simulation was respectively 4.5% and 4.1%. Compared
with errors of 1.5e17.3% obtained in model simulations conducted by Kumar and Pandey (2009), the simulated values
agreed relatively well with the experimentally determined
ones Thus, the tractor model was considered as an acceptable
one for the simulation purpose to investigate the interrelationship between the fuel efficiency parameters and the
control variables.

3.2.

Analysis of factors affecting fuel efficiency

3.

Results and discussion

Table 9 shows the results of the analysis of variance conducted on the simulation results to estimate the effects of the
control variables on FC, FCA and SVFC in the ploughing and
rotovating operations. In the rotovating operation, the Fvalues for the transmission gear ratio, engine speed and work
load on SVFC were respectively 168.27, 476.54 and 618.42, and
the p-values associated with them were all less than 0.0001.
This indicated we can accept at a confidence level of 95% that
the transmission gear ratio, engine speed and work load
affected significantly the SVFC. However, the F-value for the
ballast was 0.02, and the p-value was 0.9848 which was greater
than the F-values so that we cannot accept that the ballast
affected the SVFC. Similarly, it was accepted that the effects of
the transmission gear ratio, engine speed, and work load on
FC and FCA were all significant both in the rotovating and
ploughing operations. However, the tyre inflation pressure
affected FCA and SVFC and ballast affected FCA significantly
only in the ploughing operation. This was attributable to the
slip of the driving tyres and soil compaction which changed
with the ballast and tyre inflation pressure as indicated by
Serrano et al. (2009). It was also noted that the effects of the
transmission gear, engine speed, and work load differed
depending on the efficiency parameters. FC and SVFC were
more affected by the engine speed and work load than any
other control variables in the rotovating and by the transmission gear and engine speed in the ploughing operations.
The transmission gear and work load had the largest effect on
FCA both in the rotovating and plowing operations. These
agreed with results from the previous works (Grisso and
Pitman, 2001; Serrano et al., 2003; Peca et al., 2010; Park
et al., 2010b).
Consequently, the adjustment of engine speed, transmission gear and work load was considered as the most
effective way to increase the fuel efficiency of tractors in the
tillage operations.

3.1.

Validation of tractor model

3.3.

where FE is the fuel efficiency parameter FC, FCA or SVFC; a0 ,

a1 , a2 , a3 , a4 , a5 are constants to be determined depending on
the fuel efficiency parameters and the types of tillage
operations.
Each of the fuel efficiency parameters may have a set of
different coefficients ai even under the same tillage operation.
The coefficient value indicates the degree of the control variable's effect and its sign does the increase or decrease of the
effect. In other words, the fuel efficiency increases with
negative coefficients and decreases with positive coefficients.

2.4.3.

Development of control strategy

The linear regression model indicated that the fuel efficiency

can be optimised under the given load conditions by adjusting
the control variables according to the signs of their coefficients
in the model. To increase the fuel efficiency, the control variables with negative coefficient must be increased and those
with positive coefficient decreased. However, there are operational ranges where the control variables can be actually
varied. The engine speed may be adjusted between two speeds
which deliver the maximum torque and the maximum power
respectively. The ballast and tyre inflation pressure can be
changed with tractor weight, its distribution on the axles and
the types of implement attachment. Transmission gear ratio
can be varied from its highest to lowest values. These operational ranges are determined by engine, transmission and
structural characteristics of the tractor to be used.
The best strategy can be obtained by maximising the control variables with negative coefficient and minimising those
with positive coefficient within their operational ranges. The
worst strategy can be obtained vice versa.

Using the tractor model, PTO power, draught force, engine

power, tractor speed, FC, FCA and SVFC for the rotovating and

Control strategy to maximise fuel efficiency

Table 10 shows the coefficients of the FC, FCA and SVFC

models determined by a multiple regression analysis with a

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Table 7 e eExperimental and predicted tractor performance parameters for rotovating operation.
Condition

PTO power (kW)

1
2
3
4
5
6
7
8

Experimental

Simulated

25.7
15.4
14.7
24.8
17.7
27.5
25.1
14.9

25.6
15.7
14.8
27.1
17.1
28.4
25.6
15.6

FC (L h
1)

Condition

Experimental
1
2
3
4
5
6
7
8
a

% Errora

Simulated

38.1
24.1
18.1
42.8
21.7
44.1
38.3
24.3

35.3
23.8
18.9
42.2
21.5
44.2
37.2
25.5

SVFC (L (kW
1 h
1)

Simulated
12.2
7.0
5.8
13.7
6.5
14.1
12.6
7.4

Experimental
0.1
2.1
0.4
9.1
3.3
3.4
1.8
4.5
% Error

13.3
6.9
5.4
14.3
6.4
14.2
13.4
6.7

Engine power (kW)

Experimental
8.2
1.2
8.1
3.9
0.7
0.6
5.9
9.8

Tractor speed (km h
1)

% Error

Simulated

4.1
3.8
1.1
5.8
1.1
5.4
3.8
3.5

4.0
3.8
1.1
5.9
1.1
5.8
3.8
3.6

7.3
1.2
4.1
1.4
0.6
0.1
2.8
4.9

FCA (L ha
1)

% Error

Simulated

0.47
0.46
0.39
0.61
0.40
0.54
0.59
0.47

Experimental

Experimental

0.47
0.44
0.39
0.51
0.38
0.50
0.49
0.47

0.2
3.3
0.6
17.2
5.3
8.9
15.7
0.1

18.1
10.1
26.6
13.6
32.9
14.5
19.7
10.7

% Error

0.8
0.2
0.6
2.0
1.4
7.8
2.4
3.9
% Error

Simulated
16.7
10.2
28.3
12.7
32.3
13.2
17.9
11.2

8.2
1.2
8.1
3.9
0.7
0.6
5.9
9.8

% Error 100  (Simulated
Experimental)/Experimental.

Table 8 e Experimental and predicted tractor performance parameters for ploughing operation.
Condition

Draught force (kN)

1
2
3
4
5
6
7
8

Experimental

Simulated

10.8
10.5
9.4
12.1
10.5
12.2
11.4
10.6

11.0
11.0
10.0
12.5
10.0
12.6
11.1
11.0

FC (L h
1)

Condition

Experimental
1
2
3
4
5
6
7
8
a

% Errora

Simulated
8.2
5.6
2.6
11.3
2.6
11.3
8.2
5.5

Experimental

Simulated

18.0
18.7
5.2
32.9
5.32
31.1
17.9
18.7

18.9
17.8
5.2
31.4
5.2
31.6
18.6
17.8

2.3
5.1
6.6
3.9
4.3
3.5
2.2
3.7
% Error

8.3
5.4
2.4
11.8
2.3
11.2
8.5
5.3

Engine power (kW)

SVFC (L (kW
1 h
1)

Experimental

1.0
4.3
9.6
4.5
10.9
1.3
4.2
4.5

0.71
0.54
0.87
0.57
0.66
0.52
0.63
0.49

Tractor speed (km h
1)

% Error

Experimental

Simulated

3.9
3.4
1.0
6.2
1.2
6.4
4.3
3.7

3.8
3.6
1.1
5.8
1.3
5.9
3.9
3.8

4.6
4.8
0.6
4.4
2.5
1.8
3.8
5.0

FCA (L ha
1)

% Error

Simulated
0.70
0.50
0.82
0.56
0.73
0.55
0.67
0.47

Experimental
1.2
6.8
6.1
1.7
9.8
6.1
5.6
2.6

11.5
8.9
12.7
10.7
10.7
9.7
11.1
7.9

% Error

2.1
6.5
9.5
6.5
5.6
7.8
7.3
3.4
% Error

Simulated
11.9
8.5
12.5
10.7
11.2
10.6
11.4
8.0

3.4
4.9
1.1
0.4
3.9
9.1
2.9
0.5

% Error 100  (Simulated
Experimental)/Experimental.

Table 9 e F-values (p-values) of ANOVA test for ploughing and rotovating operations.
Dependent
variables

Ballast (Ba)

Tyre inflation
pressure (Ti)

FC

0.07
0.19
0.00
5.71
0.02
2.38

0.08
0.24
0.32
10.82
0.66
4.57

FCA
SVFC

Rotovating
Plough
Rotovating
Plough
Rotovating
Plough

(0.9340)
(0.8231)
(0.9996)
(0.0041)
(0.9848)
(0.0962)

(0.9202)
(0.7876)
(0.7272)
(<0.0001)
(0.5192)
(0.0119)

Transmission gear
ratio (Tr)
10.44 (<0.0001)
995.21 (<0.0001)
7358.30 (<0.0001)
694.88 (<0.0001)
168.27 (<0.0001)
648.37 (<0.0001)

Engine speed (Es)

4013.30
838.65
68.13
472.68
476.54
148.40

(<0.0001)
(<0.0001)
(<0.0001)
(<0.0001)
(<0.0001)
(<0.0001)

Work load (Wl)

369.16 (<0.0001)
245.62 (<0.0001)
396.40 (<0.0001)
2538.93 (<0.0001)
618.42 (<0.0001)
5.53 (0.0200)

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Table 10 e Coefficients (p-values) of linear regression model of fuel efficiency parameters applicable within operational
ranges of the control variables.
Dependent
Variable

Operation

FC

Ploughing
Rotovating
Ploughing
Rotovating
Ploughing
Rotovating

FCA
SVFC

Intercept

Ballast
adjustment

Tyre
inflation
pressure

a0

a1

a2

14.05

6.29
1.04
37.89
0.66
0.33

(<0.001)
(0.003)
(0.294)
(<0.001)
(<0.001)
(<0.001)

0.05 (0.005)

Engine
speed/10

a3

0.01 (<0.001)
0.0002 (0.007)

coefficient of determination greater than 0.9. To decrease FC,

FCA and SVFC, the positive-coefficient variables must be
decreased while the negative ones need to be increased. In
other words, the positive-coefficient variables have negative
effect on the fuel efficiency and vise verse. It is noted from
Table 10 that engine speed must be reduced as low as
possible both for the ploughing and rotovating operations to
increase the fuel efficiency. However, the effects of the
transmission gear and work load differed depending on the
types of tillage operations and the fuel efficiency parameters.
A negative effect of the higher transmission gear was seen on
only FC and a positive effect on FCA and SVFC both in the
ploughing and rotovating operations. The controlling of the
transmission gear and engine speed reduced FCA and SVFC
as the same way as in the GUTD operations. Reducing the
work load decreased FC and FCA, and increased SVFC both in
the ploughing and the rotovating operations. The effect of
tyre inflation pressure was seen only in the ploughing operation where FCA and SVFC increased with tyre inflation
pressure. This was similar to the result of the previous
research done by Serrano et al. (2009). The positive effect of
the ballast was seen only for the FCA in the ploughing
operation. In general, FCA can be decreased by increasing the
tractor speed and SVFC by operating the tractor in full-load
condition. The best strategy to decrease FC, FCA and SVFC
can be obtained by Eq. (19) with its coefficients determined

Transmission
gear

Work load

a4

1.22 (<0.001)
0.07 (<0.001)

1.08 (<0.001)

6.04 (<0.001)

0.064 (<0.001)

0.005 (<0.001)

0.06
0.07
0.05
0.03
0.002
0.001

R2

a5

(<0.001)
(<0.001)
(<0.001)
(<0.001)
(<0.001)
(<0.001)

0.30 (<0.001)
0.01 (<0.001)
1.01 (<0.001)
0.02 (<0.001)

0.003 (0.024)

0.0001 (<0.001)

0.91
0.98
0.96
0.99
0.90
0.90

according to the tillage operations and fuel efficiency parameters as shown in Table 10. For example, the best strategy
for SVFC in the plowing operation can be obtained by a liner
regression model
SVFC 0:6645 0:0002Ti
0:0639 Tr 0:0016 Es
0:0028 Wl
from which, the least SVFC can be achieved by increasing the
transmission gear ratio and work load while reducing the tyre
inflation pressure and engine speed as much as possible
within their operational range. Applying this strategy to the
tractor used in this study, the engine speed and tyre inflation
pressure must be adjusted to the level 1 and the transmission
gear ratio to the level 3 to obtain the best fuel efficiency within
the operational ranges given in Table 6.

3.4.

Effectiveness of control strategy

To evaluate the effectiveness of the control strategy given by

the linear regression model, FC, FCA and SVFC were simulated
under the two different operational conditions as shown in
Table 11. Of the two conditions, one was considered as the
best strategy and the other as the worst one. Each condition
was set by the combinations of the control variables in Table 6
and work load with coefficients determined by the linear
regression models in Table 10.

Table 11 e eOperational conditions to evaluate the effectiveness of the control strategy.

Fuel efficiency
model
FC

Operational
conditions

Ba (F:R in %)

Ti (kPa, F:R)

Ploughing
Rotovating

FCA

Ploughing

53:47
45:55

59:41
176:124

Rotovating
SVFC

Ploughing
Rotovating

a
b

59:41
176:124

Tr (ratio)

Es (rpm)

Wl (mm
or rpm)a

Designationb

L-3 (253.7)
M-3 (79.3)
L-3 (253.7)
M-3 (79.4)
M-3 (79.4)
L-3 (253.7)
M-3 (79.4)
L-3 (253.7)
M-3 (79.4)
L-3 (253.7)
M-3 (79.4)
L-3 (253.7)

1400
2200
1400
2200
1400
2200
1400
2200
1400
2200
1400
2200

100
150
540
733
100
150
540
733
150
100
733
540

FC1P
FC2P
FC1R
FC2R
FCA1P
FCA2P
FCA1R
FCA2R
SVFC1P
SVFC2P
SVFC1R
SVFC2R

mm is for the tillage depth of moldboard ploughing operation and rpm is for PTO speed for the rotovating operation.
1 and 2 denote the best and worst strategies respectively. P and R denote the ploughing and rotovating operations. For example FCA1P
represents the fuel consumption per tilled area simulated with the best strategy for FCA in the ploughing operation.

10

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

Figures 2e4 show FC, FCA and SVFC estimated under the
two operational strategies. As expected, FC, FCA and SVFC
were reduced significantly by the best control strategy. FC
was 2.7 L h
1 (FC1P) by the best strategy and 14.5 L h
1 (FC2P)
by the worst in the plowing operation, an 81.3% improvement for the best strategy compared to the worst one.
Similarly, the best strategy for FC reduced FC by 58.9% (between FC1R and FC2R) in the rotovating operations. The best
strategy for FCA and SVFC also reduced FCA by 61.1% (between FCA1P and FCA2P) and SVFC by 51.9% (between
SVFC1P and SVFC2P) in the ploughing, and by 75.7%

(between FCA1R and FCA2R) and 28.6% (between SVFC1R

and SVFC2R) in the rotovating operations respectively.
However, the best strategy for FC reduced FCA by 48.6% in
the rotovating and 13.1% in the plowing operations which
are much less than the reductions obtainable with the best
strategy for FCA. When the best FC strategy for the
ploughing operation was applied to SVFC, SVFC was reduced
by only 31.5% while FC was reduced by 81.3% as mentioned
previously. It was noticed that the best strategy worked well
only for those cases that it was designed for. In other words,
the best strategy must be established differently depending
on the types of tillage operation and fuel efficiency parameters. It must be noted that higher fuel efficiency comes
with larger values of FC and FCA but with smaller values of
SVFC.

4.

Fig. 2 e Fuel consumption per work hour simulated with

the best and worst strategies for FC (1 best strategy,
2 worst strategy, R rotovating, P ploughing).

Fig. 3 e Fuel consumption per tilled area simulated with

the best and worst strategies for FCA (1 best strategy,
2 worst strategy, R rotovating, P ploughing).

Fig. 4 e Specific volumetric fuel consumption simulated

with the best and worst strategies for SVFC (1 best
strategy, 2 worst strategy, R rotovating,
P ploughing).

Conclusions

This study was conducted to investigate the effects of ballast,

tyre inflation pressure, transmission gear, engine speed, and
work load on the fuel efficiency of a tractor in ploughing and
rotovating operations using a computer simulation method.
The fuel efficiency was considered in terms of three parameters: fuel consumption per work hour (FC), fuel consumption
per tilled area (FCA) and specific volumetric fuel consumption
(SVFC).
FC, FCA and SVFC were simulated for 162 operational
conditions made by a combination of the weight distribution
on front and rear axles, transmission gear ratio, engine speed,
and tyre inflation pressure each consisting of three levels in
the ploughing and rotovating operations using a tractor model
developed in this study. The mean errors of the simulated FC,
FCA, and SVFC were all ranged within 3.3e6.5% in both the
rotovating and ploughing operations when compared with
those obtained from the field experiment under eight operational conditions. Statistical analyses of the simulation results
showed that:
(1) Effects of the engine speed, transmission gear, ballast,
tire inflation pressure and work load on FC, FCA and
SVFC were different depending on the types of tillage
operations considered, namely, -ploughing and
rotovating.
(2) The engine speed, transmission gear, and work load
affected FC, FCA and SVFC significantly both in the
ploughing and rotovating operations.
(3) The tyre inflation pressure had significant effect on FCA
and SVFC and ballast had significant effect on FCA only
in the ploughing operation.
(4) To decrease FC, FCA and to increase SVFC, the engine
speed must be reduced near to the speed for the
maximum torque both in the rotovating and ploughing
operations. The engine speed had the greatest influence
on SVFC in ploughing and on FCA in rotovating
operations.
(5) The best strategies to maximise the fuel efficiency were
developed as linear combinations of ballast, tyre inflation pressure, transmission gear ratio, engine speed,
and work load in the form

b i o s y s t e m s e n g i n e e r i n g 1 4 2 ( 2 0 1 6 ) 1 e1 1

FE a0 a1 Ba a2 Ti a3 Tr a4 Es a5 Wl
with regression coefficients determined for different fuel efficiency parameters and type of tillage operations.
(6) FC, FCA and SVFC obtained by the best strategy were
2.7 L h
1, 8.8 L ha
1, and 0.4 L/(kWh)
1 in the ploughing and
5.4 L h
1, 8.0 L ha
1, and 0.3 L (kWh)
1 in the rotovating operations which were 81.3, 61.1, and 52.0%, and 58.0, 75.7 and
28.6% improvement over the same operations by the worst
strategy.

Acknowledgement
This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Korean Government
(MOE) (2010-0023058).

references

Al-Hamed, S. A., & Al-Janobi, A. A. (2001). A program for

predicting tractor performance in Visual C. Computers and
Electronics in Agriculture, 31(2), 137e149.
ASABE Standard. (2006). ASABE D497.5 FEB2006: Agricultural
machinery management data. St. Joseph, MI, USA: American
Society of Agricultural and Biological Engineers.
Brixius, W. W. (1987). Traction prediction equations for bias ply tyres.
ASAE Paper No. 87-1622. St. Joseph, MI, USA: American Society
of Agricultural and Biological Engineers.
Casady, W. W. (1997). Tractor tyre and ballast management.
Agricultural Publication G1235. Columbia, MO, USA: University
of Missouri-Columbia.
 n, H., Linares, P., & Me
ndez, V. (2008). Tractor_PT: a traction
Catala
prediction software for agricultural tractors. Computers and
Electronics in Agriculture, 60(2), 289e295.
Evans, M. D., Clark, R. L., & Manor, G. (1991). An improved traction
model for ballast selection. Transactions of the ASAE, 34(3),
773e780.
Firestine, N., Furrey, L., Aslin, C., Crivella, F., & Sautter, J. A. (2007).
The farmer's handbook for energy self-reliance (pp. 20e26). South
Royalton, USA: The Institute for Energy and the Environment
Vermont Law School.
Fulton, J., Raper, R., McDonald, T., & Tyson, T. (2006). Fuel
conservation strategies for the farm. Alabama Cooperative Extension
System. ANR-1303. Auburn, Alabama, USA: Alabama A&M and
Auburn University.
Gee-Clough, D., Pearson, G., & McAllister, M. (1982). Ballasting
wheels tractors to achieve maximum power output in
frictional-cohesive soils. Journal of Agricultural Engineering
Research, 27(1), 1e19.
Grisso, R. D., & Pitman. (2001). Gear up and throttle down e Saving
fuel. Virginia Cooperative Extension Publication (pp. 442e450).
Blacksburg, VA, USA: Virginia Tech.
Kim, S. C., Kim, K. U., & Kim, D. C. (2010). Modeling of fuel
consumption rate for agricultural tractors. Journal of Biosystems
Engineering, 35(1), 1e9 (in Korean, with English abstract).
Kim, K. U., Kim, K. D., Koo, Y. M., & Park, K. J. (2009). Agricultural
machinery design 1. Seoul, Korea: Moon Woon dang (in Korean).

11

Kolator, B., & Biaobrzewski, I. (2011). A simulation model of 2WD

tractor performance. Computers and Electronics in Agriculture,
76(2), 231e239.
Korean Industrial Standard M 6752. (2009). Rubber tyres for
agricultural implements and machineries. Seoul Korea: Korean
Standard Association.
Kumar, R., & Pandey, K. P. (2009). A program in visual basic for
predicting haulage and field performance of 2WD tractors.
Computers and Electronics in Agriculture, 67(1e2), 18e26.
Lacour, S., Burgun, C., Perilhon, C., Descombes, G., & Doyen, V.
(2014). A model to assess tractor operational efficiency from
bench test data. Journal of Terramechanics, 54, 1e18.
Lancas, K. P., Upadhyaya, S. K., Sime, M., & Shafii, S. (1996).
Overinflated tractor tyres waste fuel, reduce productivity.
California Agriculture, 51(2), 28e32 (UC Davis, California,
USA).
Lee, J. W., Kim, K. U., Kim, D. H., & Choi, K. J. (2011). Effects of
distribution of axle load and tyre inflation pressures on fuel
efficiency of tractor operations. Journal of Biosystems
Engineering, 36(5), 303e313 (in Korean, with English abstract).
NACF.. (2010). Work guidelines for tax-free fuels. Seoul, Korea:
National Agricultural Cooperative Federation (in Korean).
OECD Code 2. (2010). Standard code for the official testing of
agricultural and forestry tractor performance. Paris, France:
Organization for Economic Co-operation and Development.
Park, S. H., Kim, Y. J., Im, D. H., Kim, C. K., Jang, Y., & Kim, S. S.
(2010b). Analysis of factors affecting fuel consumption of
agricultural tractor. Journal of Biosystems Engineering, 35(3),
151e157 (in Korean, with English abstract).
Park, S. H., Kim, Y. J., Im, D. H., Kim, C. K., Jung, S. C., Kim, H. J.,
et al. (2010a). Development of eco driving for agricultural
tractor. Journal of Biosystems Engineering, 35(2), 77e84 (in
Korean, with English abstract).
Peca, J., Serrano, J., Pinheiro, A., Carvalho, M., Nunes, M.,
Ribeiro, L., et al. (2010). Speed advice for power efficient
drawbar work. Journal of Terramechanics, 47, 55e61.
Pranav, P. K., & Pandey, K. P. (2008). Computer simulation of
ballast management for agricultural tractors. Journal of
Terramechanics, 45(6), 185e192.
Sahu, R., & Raheman, H. (2008). A decision support system on
matching and field performance prediction of tractorimplement system. Computers and Electronics in Agriculture,
60(1), 76e86.
 rquez, L. (2009). The
Serrano, J., Peca, J., Marques da Silva, J., & Ma
effect of liquid ballast and tyre inflation pressure on tractor
performance. Biosystems Engineering, 102, 51e62.
Serrano, J., Peca, J., Marques da Silva, J., Pinheiro, A., &
Carvalho, M. (2007). Tractor energy requirements in disc
harrow systems. Biosystems Engineering, 98, 286e296.
Serrano, J., Peca, J., Pinheiro, A., Carvalho, M., Nunes, M.,
Ribeiro, L., et al. (2003). The effect of gang angle of offset disc
harrows on soil tilth, work rate and fuel consumption.
Biosystems Engineering, 84, 171e176.
Upadhyaya, S. K., Wulfsohn, D., & Jubbal, G. (1989). Traction
prediction equations for radial ply tyres. Journal of
Terramechanics, 26(2), 149e175.
Wismer, R. D., & Luth, H. J. (1973). Off-road traction prediction for
wheeled vehicles. Journal of Terramechanics, 10(2), 49e61.
Zhang, N., & Chancellor, W. (1989). Automatic ballast position
control for tractors. Transactions of the ASAE, 32(4), 1159e1164.
Zoz, F. M. (1970). Predicting tractor field performance. ASAE Paper No.
70-118. St. Joseph, MI, USA: American Society of Agricultural
and Biological Engineers.