Tractor design for small farms in resource limited markets

by
Guillermo Fabin Diaz Lankenau
B.S., Tecnol6gico de Monterrey (2012)
S.M., Massachusetts Institute of Technology (2017)
Submitted to the Department of Mechanical Engineering
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
February 2020

@ Massachusetts Institute of Technology 2020. All rights reserved.

Signature redacted
Author .............................
.

Department of Mechanical Engineering

December 9, 2019

Certified by......................Signatureredacted
Amos G. Winter,V
Associate Professor
Thesis Supervisor

Signature redacted
Accepted by..................
MASSACHUSETTS INSTITUTE NicolasG. Hadjiconstantinou
OfTECHNOLOGY -. hairman, Department Committee on Graduate Theses

FEB 0 5 2020
LIBRARIES
,
j

s
 Tractor design for small farms in resource limited markets

Guillermo Fabiin Diaz Lankenau

Submitted to the Department of Mechanical Engineering

on December 9, 2019, in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy

Abstract
This thesis describes the design of a tractor for small farms (<2 ha) in resource limited
markets, particularly India, and the analytical framework used to arrive at the design. Indian
smallholder farmers typically rely on draft animals, which compared to tractors are more
expensive to maintain, more exhausting to use, slower, and incompatible with many modern
farming tools and methods. These disadvantages are detrimental to the farmer's income
and crop yields. However, existing small tractors are too large and expensive to directly
replace draft animals. The presented tractor design is unique in its ability to compete with
draft animals' physical dimensions, pulling performance, and sale price, while retaining key
tractor advantages like compatibility with modern tools, low maintenance costs, and reduced
drudgery. This tractor features motorcycle-like controls and seating, inline drive wheels,
stabilization via an outrigger arm or a specially-developed balance board attachment, and
the ability to attach implements ahead or behind the rear axle. The design was created to
satisfy unmet farmer requirements identified during on-site interviews with Indian farming
stakeholders.
Before deviating from the conventional tractor design, a comprehensive description,
from a historical and physical perspective, of why the conventional tractor came to be
was elucidated. Then, the proposed tractor design was conceived by leveraging historical,
physics-based, and user-focused insights. Experimental results with an instrumented proof-
of-physics prototype validated the new tractor could produce traction forces as predicted by
the analytical framework used to create the design, as well as meet or exceed the maximum
pulling forces generated by draft animals.
A functional prototype of the tractor was built, and its ability to complete key farming
operations was demonstrated on a Massachusetts farm. The vehicle was able to complete
plowing, disc harrowing, rotary tilling, planting, cultivating, spraying, and towing of a trailer
per Indian industry specifications. A study was conducted to assess whether the vehicle
would meet the needs of small and marginal farmers in India through on-site, one-on-one
interviews with 24 farmers in Karnataka, Gujarat, and Tamil Nadu. Farmers generally
reported that the prototype tractor would meet their needs, with an average likelihood of
4.8/5 that they would use the vehicle for planting, inter-cultivation, and spraying, and an
average likelihood of 3.8/5 that they would use the tractor for primary or secondary tillage.

Thesis Supervisor: Amos G. Winter, V

Title: Associate Professor

3
4
Acknowledgments

Thank you to my advisor, Amos Winter, for your support and guidance through out this

journey. It has been fantastic to travel India with you and experience Global Engineering

through your eyes. Working with you has made me a better engineer.
Thank you to Maria Yang and Alexander Slocum for being part of my thesis committee.
I deeply appreciate your valuable advice and you sharing your expertise with me. Thank
you for all your help.
Thank you to Gwyndaf jones for your generous help, time, and valuable advice. Thank
you to the wonderful members of GEAR Lab for their support and friendship - it has meant a

lot me. Thank you to Susan Amrose, Julia Sokol, Brett Johnson, Victor Prost, Dan Dorsch,
Kameron, Sahil Shah for helping me prepare the writing in this thesis and my defense slides.

Thank you also to Dan Dorsch, along with Elliot Donlon, Jeffrey Costello, Shane Pratt, and
Rashed Al-Rashed for helping me execute field tests. Thank you also to Michael Buchman,
Murthy Arelekatti, and Natasha Wright for their support and friendly chats.
Thank you to the TATA Center at MIT and Mahindra Tractors for sponsoring this
research and assisting me in scheduling interviews in India as well as gaining knowledge
of the Indian Market. Thank you to Mahindra's knowledgeable team including Aravind
Bharadwaj, Vivek Gupta, Nagendra Pansare, Sameer Deo, Ravindra Shahane, Muthiah
Saravanan, Ashwinikumar Parhate, and Ganesh Sadkar.Thank you also to the Deshpande

Foundation and Vigyan Ashram for assisting me during my interviews in India and gaining
knowledge of the Indian Market.
Thank you to my parents, Magdalena and Guillermo, for believing in me always and,
with seemingly endless patience and love, enriching my life with support, advice, and a
caring ear. Thank you to my siblings Magdalena and Rodrigo for helping me embrace my
quirkiness and always being a part of my life and support system despite long distances.
Thank you to Jaya, my amazing wife, for supporting me through the challenging times

and making the good times incredibly joyous. Our love and mutual support gives me confi-
dence and anticipation for a wonderful, exciting life.

5
THIS PAGE INTENTIONALLY LEFT BLANK

6
Contents

1 Introduction 17

1.1 The need for mechanization in farms . . . . . . . . . . . . . . . . . . . . . . 17

.
1.2 The conventional tractor design is not meant for small farms worldwide . . . 18
1.3 Bullocks can be preferable over tractors in small farms . . 19

.
1.4 The Bullkey design . . . . . . . . . . . . . . . . . . . . . . 20

.
1.5 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . 21

.
2 An Engineering Review of the Farm Tractor's Evolution to a Dominant
Design 23
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 23
.

2.2 History of farm tractor evolution to a dominant design . . 24

.

2.2.1 Design features of the conventional farm tractor . . 24

.

2.2.2 1900 to 1920: early history of the modern farm tractor . . . . . . . . . 25

2.2.3 1920 to 1950: farm tractors converge on a common design to power

Am erican farm s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
.

2.2.4 Emergence of the dominant tractor design features between 1910 and
1940 ....................................... 30
2.3 Analytical modeling of the conventional tractor's design . . . . . . . . . . . 32
.

2.3.1 Conventional tractor dimensions and relevant forces . . . . . . . . . 32

.

2.3.2 Qualitative description of importance of soil-tire interaction in tractor

design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
.

2.3.3 Model for interaction of a single drive tire with soil . . . . . . . . . . 34

.

2.3.4 Validation of Tractor Model with Published Data . . . . . . . . . . . 38

.

2.4 Results and discussion: analytical model insights into tractor dominant design 38

7
 2.4.1 Advantages of conventional tractor weight distribution . . . . . . . . 38

.
2.4.2 Advantages of tillage tool rigid mounting behind the rear axle . . . . 40

.
2.4.3 Advantages of conventional rear wheel drive . . . . . . . . . . . . . . 42

.
2.4.4 Advantages of four wheel rectangular layout . . . . . . . . . . . . . . 43

.
2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

.
3 Design of a specialized tractor to replace draft animals in small farms 45

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

.
3.2 Physics behind maximizing traction performance . . . . . . . . . . . . . . . 48

.
3.3 Design exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

.
3.3.1 Physics-based design insights . . . . . . . . . . . . . . . . . . . . . . 55

.
3.3.2 Comparison of tractor layouts . . . . . . . . . . . . . . . . . . . . . . 58

.
3.3.3 Predicted performance . . . . . . . . . . . . . . . . . . . . . . . . . . 61

.
3.4 Proof-of-concept vehicle design . . . . . . . . . . . . . . . . . . . . . . . . . 62

.
3.5 Field testing and performance results . . . . . . . . . . . . . . . . . . . . . . 66

.
3.5.1 Field testing methods . . . . . . . . . . . . . . . . . . . . . . . . . . 66

.
3.5.2 Field performance results . . . . . . . . . . . . . . . . . . . . . . . . 68

.
3.5.3 User feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

.
3.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

.
3.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

.
4 Investigation of viability to replace draft animals with all-wheel-drive mo-
torcycles on small Indian farms 73
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
.

4.2 Description of Farmer Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

.

4.3 Prototype Vehicle Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

.

4.3.1 Prototype vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

.

4.3.2 Balance Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

.

4.3.3 Implement Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . 83

.

4.4 Methods - Field Validation of Operations and Assessment Interviews 85

4.4.1 Methods - Field Validation in Massachusetts . . . . . . . . . 85
.

4.4.2 Massachusetts farmer interviews . . . . . . . . . . . . . . . 87

.

4.4.3 Indian farmer interviews . . . . . . . . . . . . . . . . . . . . 87

.

8
 4.5 Results - Field Validation of Operations and Assessment Interviews 88

.
4.5.1 Field Validation in Massachusetts ..... 88
4.5.2 Results - Massachusetts farmer interviews 89
4.5.3 Feedback from Indian farmers . . . . . . 91

.
4.6 Discussion . . . . . . . . . . . . . . . . . . . . . 93

.
4.6.1 Satisfying user needs . . . . . . . . . . . 93

.
4.6.2 Limitations and Future Work . . . . . . 93

.
4.7 Conclusions . . . . . . . . . . . . . . . . . . . . 95

.
5 Conclusion 97
5.1 Motivation of Research ............... . . . . . . . . . . . . . . . . 97
5.2 Contributions from Research . . . . . . . . . .. . . . . . . . . . . . . . . . . 98
5.2.1 Evolution of conventional tractor . . . . . . . . . . . . . . . . . . . . 98
5.2.2 Design of Bullkey for small farms . . . . . . . . . . . . . . . . . . . . . . 99
5.2.3 Viability of Bullkey for replacing bullocks . . . . . . . . . . . . . . . . 99
5.3 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
.

5.3.1 Central tillage tool improved mechanism . . . . . . . . . . . . . . . . 101

.

5.3.2 Next steps for distribution of Bullkey . . . . . . . . . . . . . . . . . . 103

.

5.4 Conclusion. . . . . . . . . . . . . . . . . . . . . . . 106

.

6 Appendix 115

9
THIS PAGE INTENTIONALLY LEFT BLANK

10
List of Figures

1-1 Images of the proposed Bullkey tractor. . . . . . . . . . . . . . . . . . . . . . 21

2-1 Graphic chronology of tractor evolution as influenced by historical context

and stakeholder expectations. . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2-2 Sample of tractor design layouts from 1910 to 1920. . . . . . . . . . . . . . . . 27
2-3 Graphic chronology of tractor evolution into conventional small tractor design. 28
2-4 Free body diagram for farm tractor in 2D. . . . . . . . . . . . . . . . . . . . . 33
2-5 Stress under rigid driven wheels rolling on deformable soil. . . . . . . . . . . . 35
2-6 Parameters of tire perimeter for calculation of forces at soil interface. . . . . . 37
2-7 Comparison of tractor model as described in section 2.3 to published experi-
m ents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

2-8 Simulation data for tractor configurations with varying weight, weight distri-
bution, and draft load. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

2-9 Drawbar pull data compiled from Nebraska Tractor Test Archives 1691. ... 41

3-1 Cumulative costs of farming 1 ha with bullocks or small farm tractors. .... 46
3-2 Relationship of tractor sales price to tractor mass for common Indian tractors,
as compiled by the author. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3-3 Force free body diagram for a conventional small tractor . . . . . . . . . . . . 49
3-4 Parameters of tire perimeter for the calculation of forces at the tire soil interface. 51

3-5 Illustration of tire-soil interaction and multi-pass effect used in analysis. . . . 53

3-6 Sensitivity analysis of drawbar pull and tractive efficiency at 15% tire slip (a
typical heavy tillage operating point 1901) for a conventional small tractor. . . 54
3-7 (A) Isometric view with labeled soil engaging components and (B) force free
body diagram for Bullkey tractor. . . . . . . . . . . . . . . . . . . . . . . . . . 55

11
3-8 Tractor layouts considered for Bullkey. . . . . . . . . . . . . . . . . . . . . . . 59
3-9 A comparison of the drawbar pull versus slip performance in weak to strong
agricultural soil for a 500 kg hypothetical implementation of a conventional
tractor and Bulkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3-10 Bullkey prototype vehicle highlighting the implementation of desirable design

features for a small tractor intended to replace a pair of bullock . . . . . . . . 63

3-11 Close-up views of sensor installation examples. . . . . . . . . . . . . . . . . . 65
3-12 Mechanical diagram and CAD of the structure used to measure tillage forces
on the Bullkey proof-of-physics prototype. . . . . . . . . . . . . . . . . . . . . 66

3-13 Examples of measured forces for prototype configurations tested. . . . . . . . 69

4-1 Ownership costs over 15 years for a bullock pair, a small tractor, and the
Bullkey tractor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4-2 Locations visited during the development of the Bullkey tractor prototype. . . 75
4-3 Overview of the Bullkey design. . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4-4 Bullkey with balance board between crop rows. . . . . . . . . . . . . . . . . . 79
4-5 The balance is a rigid, wheeled platform connected to the motorcycle frame
by a ball hitch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4-6 FORCE FLOW PATHS WHEN USING BALANCE BOARD (TOP IMAGE)

AND IN CONVENTIONAL MOTORCYCLE (BOTTOM IMAGE). White

arrows represent the force applied by the user legs on and opposed by the
black arrow reaction forces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4-7 Close up views and key components of the farming implements used during
field testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4-8 Images of the researchers performing key operations of interest to Indian
farmers with Bullkey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4-9 Self-described likelihood of two interviewed Peruvian farmers for doing an
activity with Bullkey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4-10 Self-described likelihood of interviewed Indian farmers for doing an activity
with Bullkey tractor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

5-1 Images of people riding on tools to help them sink into the ground. . . . . . . 102
5-2 Tests to explore crank handle height. . . . . . . . . . . . . . . . . . . . . . . . 103

12
5-3 The conventional tool motion path (top) compared to the tool path proposed
by the authors (bottom). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5-4 AN ALTERNATIVE BALANCE BOARD DESIGN CONCEPT ALSO CRE-

ATED BY THE AUTHORS. The two arms could be rigidly coupled to rotate
together (four total pin joints) or independent (five total pin joints). The
arms are assumed to be rigidly coupled. . . . . . . . . . . . . . . . . . . . . . 1 05

6-1 Larger vehicle images. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

.
6-2 Pages 1 and 2 of booklet used to interview farmers in India. . . . . . . . . 131
6-3 Pages 3 and 4 of booklet used to interview farmers in India. . . . . . . . . 132

6-4 Pages 5 and 6 of booklet used to interview farmers in India. . . . . . . . . 133

6-5 Pages 7 and 8 of booklet used to interview farmers in India. . . . . . . . . 134

6-6 Pages 9 and 10 of booklet used to interview farmers in India. . . . . . . . . 135

6-7 Pages 11 and 12 of booklet used to interview farmers in India. . . . . . . . . 136

6-8 Four bar mechanism with variables used in equations labeled. . . . . . . . . 137

6-9 Examples of implementation of four bar linkage (top) and four bar linkage
with handle for manual control (bottom). . . . . . . . . . . . . . . . . . . . 137
.
6-10 Instants in the mechanism motion. Between instants 5 and 1, the tool is

coming down so the user does not need to apply force at the handle and can
let go of it. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6-11 Views of concept sliding rail mechanism on motorcycle. . . . . . . . . . . . . . 140
6-12 Instants in sliding rail mechanism motion. . . . . . . . . . . . . . . . . . . . . 141
6-13 Instants in the cable and slot mechanism motion. . . . . . . . . . . . . . . . . 142

13
THIS PAGE INTENTIONALLY LEFT BLANK

14
List of Tables

3.1 Occurrence of desirable design characteristics in evaluated tractor layouts of

Fig.3-8 .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Parameters for a conventional tractor and the Bullkey compared in Figure 3-9. 61

3.3 Basic parameters for Bullkey prototype. . . . . . . . . . . . . . . . . . . . . . 65

4.1 Major user needs identified via interviews with stakeholders in India and
background research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2 Key Bullkey dimensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

6.1 Production data for vehicles discussed in tractor evolution. . . . . . . . . . . . 115

6.3 Breakdown of costs to an Indian farmer for purchasing a tractor (financed
and upfront) or a pair of bullocks. [81, 80, 22, 31, 30, 47] . . . . . . . . . . . . 116
6.2 Layouts of Ch. 2 matched to production vehicles. . . . . . . . . . . . . . . . . 117
6.4 Summary of results from field test experiments compared to model predictions.119

6.5 Common tractors sold in India with their mass, engine power, and lowest
typical sale price. The tractors are sorted by mass. Data collected by authors
from online tractor sale websites [65, 96]. . . . . . . . . . . . . . . . . . . . . . 122
6.6 Soil properties used to generate plots. . . . . . . . . . . . . . . . . . . . . . . 123
6.7 Overview of electronics used for data collection. . . . . . . . . . . . . . . . . . 124
6.8 Basic dimensions for tools used. . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.9 Basic dimensions for attachment systems. . . . . . . . . . . . . . . . . . . . . 126
6.10 Breakdown of costs to an Indian farmer for bullock ownership. . . . . . . . . . 127
6.11 Breakdown of costs to an Indian farmer for purchasing (financed and upfront)

or renting a tractor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

15
THIS PAGE INTENTIONALLY LEFT BLANK

16
Chapter 1

Introduction

This thesis proposes a small tractor concept, called Bullkey, specifically designed for small
farms in low resource settings - particularly in rural India. The chapters move from a review
of the evolution of the conventional tractor (Ch. 2), to the implementation and testing of a
tractor layout with a high drawbar pull to mass ratio (Ch. 3), to the creation of a Bullkey
prototype and an evaluation of its ability to satisfy the needs of Indian farmers (Ch. 4). The
socioeconomic research and evaluation of conventional tractors that motivated this research
are summarized here in Chapter 1.

1.1 The need for mechanization in farms

Tractors are an icon of industrialized, modern farming and their presence has been noted
as a differentiator between farming in developed vs. developing countries139] [32]. In 1950,
the USA Census Bureau stated the benefits of mechanizing American agriculture [981:

The increased use of mechanical power on farms has influenced agriculture more
than any other factor during the present century. The changes from horses and
mules to tractorsfor farm work has made available an acreage of cropland greater
than the total increase in cropland during the half century for the production, di-
rectly or indirectly, of meat, milk, eggs, and other food. The use of the tractor
and related equipment for farm work, and the use of farm trucks for hauling and
automobiles for traveling have increasedthe rate at which farm work is done and
has increased the capacity of agriculturalworkers, enabling considerable numbers

17
 of farm workers to leave the farms or to engage in non-farm work, notwithstand-
ing considerableincreases in total farm production. Tractors and power-operated
equipment have made an increase in the size of farms possible. The substitution

of tractorsfor animal power has also made available additionalpower for farm
use. With a tractor, the farmer of 1950 probably turned out twice the amount of
farm products for market as his father did with a team of horses 50 years earlier.
Moreover, less of the farmer's time was required to care for the tractor than to
raisefeed for and to produce and care for horses that were replaced by the tractor.

There is high correlation worldwide between farm productivity and available tractor
power 1331[73]1[401[391. The Indian Agricultural Ministry [521estimates that farm tractors'
precision and timeliness increases farm yields by 5 to 20%, reduces waste seeds and fertilizer
by 15 to 20%, and reduces the labor time required on a farm by 20 to 30%. Bullock owners
are currently burdened with the high ownership costs of the animals, which outweigh initial
costs over time. The total cost of buying and maintaining a pair of bullocks for 15 years is
over two times higher than the equivalent costs for a tractor.

1.2 The conventional tractor design is not meant for small

farms worldwide

The majority of farms in the world (84%) are less than 2 ha in size [58]. Small farms
are particularly common in India, where the average farm size has steadily decreased from
2.28 ha in 1971 to 1.08 ha in 2016 [75]. Most small farmers use a pair of bullocks for
all or most of their farming operations, supplementing the rest with pure manual labor

or a hired tractor (particularly for primary and secondary tillage) [52][31]. Bullocks are
compact, highly maneuverable, and have a low capital cost, making them well-suited to the
technical and economic constraints of small farms. Conventional tractors, which are an icon
of modern farming and typically associated with high farm yields [391, have not yet been
able to reproduce these key bullock features [64][76]. As a result, farmers are constrained to
the slow speed of bullocks and a lack of access to suitable modern, more effective made-for-
tractor tools [3111301.
This misalignment between the tractor and needs of the small farmer is, in part, because
conventional tractors were designed for a larger field size. The conventional small tractor

18
was originally intended to operate in much larger farm holdings than what is typical around
the world [44]. The conventional small tractor largely found its form in the USA between
1910 and 1940 [44]1107117711611 [1041, where the annual average farm size since 1900 has
been at least 50 times greater than the present global average. Globally, at present, 72% of
farms are <1 ha, while less than 2% of US farms have been <1 ha since 1900 [581 [3311141

[991 [591. As a result, for many small farmers, replacing their bullocks with a conventional
tractor would require adjusting crop spacing (particularly for crops taller than a tractor's
ground clearance), or sacrificing farmland as headland area to give the tractor space to
maneuver at row ends. It might also require improving road access to their field due to the
tractor's larger size and poorer off-roadability.

1.3 Bullocks can be preferable over tractors in small farms

As part of this study, the authors spoke about local agricultural practices and the suitability
of existing tools with stakeholders of Indian small farming at 12 locations in India including
the states of Maharashtra, Tamil Nadu, Gujarat, Rajasthan, Madhya Pradesh, West Bengal
(Fig. 4-2).
Stakeholders included farmers, research organizations, governments, manufacturers, and
dealers. A key observation from these visits was that farmers would use bullocks not only
because of their low capital cost (about a third that of tractors) but because they have

functional advantages over conventional small tractors. Some of these advantages are:

" Bullocks have a smaller width than tractors. This allows the bullocks to walk between
rows of growing crops further into the season than a tractor could. In some cases even
allowing the farmer to use a tighter row spacing than would be possible with tractors.

" Bullocks require less headland area. Headland area is space left at field ends to allow
farm vehicle to maneuver between rows. Bullocks can make tight turns and are able
to step over crops or rows to avoid damaging them when turning.

" Bullocks have better off-roadability. They only require narrow trails to advance, similar
to human foot paths. They use legged locomotion which allows them to step over
obstacles, including foot-high bunds of soil between farm fields. And they have a self-
preservation instinct that can be reassuring to farmers when moving past cliffs or other

19
 dangerous features. Tractors, in an effort to achieve high crop (i.e. ground) clearance
have high centers of mass. This makes tractors more dangerous on sloped terrain than
bullocks.

•Bullocks are easier to control at slow speeds and in tight spaces. This was particularly

important to farmers operating in very wet fields where they are concerned with the

tractor's straight tracking, ability to access field edges, and difficulty of salvaging if it
gets stuck.

A tractor likely to be adopted by small Indian farmers should have the bullock's ability to
access narrow inter-row spaces and the bullock's low capital cost while retaining the tractor's

relatively low upkeep costs, efficient use of resources, better ability to interface with modern
tools, and low user fatigue.

1.4 The Builkey design

The research presented in this research manifested in the design of Bullkey, which has
features and performance to address Indian farmer needs. Bullkey is a portmanteau of
bullock and key - indicating its goal of being the key to unlocking the bullock market to

mechanization.
Bullkey features a three-wheeled layout that supports nearly all of the vehicle's mass on
inline drive wheels and shifts the tillage tool ahead of the rear axle (Fig. 1-1). This layout
increases the load on all drive tires during plowing and allows the trailing drive tire to roll
on already-compacted terrain - improving drawbar pull while also preventing the vehicle
from rolling over backward. To stabilize itself in roll, Bullkey can utilize an outrigger arm
or a custom-designed balance board system. The outrigger arm has high ground clearance
and can straddle growing crop. The balance board utilizes some user effort for stability but

keeps the Builkey overall width comparable to bullocks; this allows Bullkey to fully operate
between rows of growing crop without the need to straddle them.

Validation of the Bullkey design was two-fold: the design was field tested to validate the
physics model that drove its design, and the design was discussed with 24 small farmers in
India to receive their feedback on its viability as a replacement for bullocks. The field tests
demonstrated that the Bullkey tractor layout can generate more drawbar pull per unit mass

than a conventional tractor. The user interviews in India showed that farmers were able to

20
Figure 1-1: Images of the proposed Bullkey tractor. This design is discussed in detail in
Chapters 3 and 4. It features inline drive wheels, an outrigger wheel, and an option for a
centrally located tillage tool or conventional rear mounted tools.

identify Bullkey's ability to replace bullocks on farms and would be willing to purchase it
for a price equal to or higher than the predicted sales price for the Bullkey design.

1.5 Outline of Thesis

Chapter 2 discusses the origins of the conventional tractor. It elucidates the key features
of the conventional tractor design and what motivated their development. This is done via
a review of the tractor's history from an engineering perspective and the application of a
tractor physics model to study key design parameters. Understanding the reasoning behind
the conventional tractor design can empower the engineer to more confidently modify it to
meet the needs of new users. This review provided inspiration and motivation for the design
of Bullkey.
Chapter 3 explains the motivation and reasoning for the Bullkey design. It also de-
scribes a proof-of-physics prototype created to validate the performance advantages of Bul-
lkey. It is shown that the Bullkey design can generate more drawbar pull per unit mass than
conventional tractors and is a viable design for accessing narrow spaces on farm fields.
Chapter 4 describes the creation of a Bullkey prototype capable of performing farming
operations key to small Indian farmers, the field tests of that prototype, and the feedback
from one-on-one interviews with Indian farmers about their perception on the Bullkey's
viability for performing those operations. It is shown that Bullkey can succesfully perform

the farming operations it would be expected to and that farmers perceive as a vehicle that

21
integrates the key advantages of both tractors and bullocks.
Chapter 5 summarizes the results from this thesis and discusses its impact, along with
high value directions for future work. Bullkey has been shown to have high potential for

replacing bullocks in small Indian farms. Future work could interview Indian farmers who
have tested the tractor on their own farm fields and refine the design to get it ready for

production. Additionally, the physics model used to create Bullkey could be refined further
to increase its accuracy at low tire slips.

22
Chapter 2

An Engineering Review of the Farm

Tractor's Evolution to a Dominant
Design

2.1 Introduction

Tractors are an icon of industrialized, modern farming and their presence has been noted as
a differentiator between farming in developed versus developing countries [39] {32]. There is
high correlation worldwide between farm productivity and available tractor power {3911331
[73] [40]. In 1950, the USA Census Bureau summarized the benefits of mechanizing American
agriculture during the past 50 years198]: mechanical power on farms influenced agriculture
more than any other factor; mobile, powered farm equipment increased the rate at which
farm work was done and has increased the capacity of agricultural workers, enabling con-
siderable numbers of farm workers to leave the farms or to engage in non-farm work; and
tractors and power-operated equipment made an increase in the size of farms possible.
This chapter provides historical and physics context for why the modern tractor evolved
to its dominant design. This understanding and the analysis framework used to explain

it are valuable tools for evaluating new tractor designs. The work presented herein is of
contemporary relevance. Farm mechanization is growing in emerging markets and common
pitfalls of engineers designing for those markets include poor assumptions about the needs
of end users and a lack of knowledge about the specific problem space, cultural norms.

23
and technical area[109). This chapter can help these engineers by supplying them with the
context in which tractors were originally developed (and thus a reference point to compare
to their target market) and by accelerating their development of expertise on farm tractors

by understanding the underlying physics. A historical background on tractor design can also
help gain insights on new innovation opportunities; for example, relatively new autonomous
driving and precision farming technologies encourage the development of novel tractor-like
vehicle platforms [1101. Engineers may more confidently break away from the dominant
tractor design if they understand its origins.

2.2 History of farm tractor evolution to a dominant design

2.2.1 Design features of the conventional farm tractor

The conventional small tractor produced today found its form mostly in the USA between

1910 and 1940 11071 [77] [611[104]. The most salient features of the "conventional tractor"
or "dominant tractor design" are:

" Four wheels laid out in a rectangular pattern, attached to front and rear axles.

" Pneumatic tires on all wheels, with the rear tires having larger diameter and width
than the front tires.

" Front wheel steering.

• Rear-wheel-drive with the wheels joined by a differential axle (front wheel assist some-
times present).

• Independently controlled braking force at each rear wheel (two brake pedals are pro-
vided).

" Engine rests over the front axle.

" Operator sits between both axles, usually only slightly ahead of rear axle.

" Trailing implement behind the rear axle, option to attach implement rigidly (via
"Three-Point Hitch").

" Rear-pointing, engine-powered exposed shaft behind the rear axle for powering imple-
ments ("Power Take-Off").

24
 * Engine crankcase and transmission case used as structural components.

Similar to many products, these characteristics did not evolve solely to improve farm field
performance; they were also determined by pressure to lower manufacturing and distribution
costs, improve marketability, increase versatility and ease-of-use, and comply with govern-
ment regulations. Some of these pressures and their impacts are discussed in Sections 2.2.2

and 2.2.3. They are also summarized in Fig. 2-1.

To emphasize the other design directions the nascent tractor industry could have taken,
Fig. 2-2 shows 24 production tractor layouts from between 1910 and 1920. Layouts varied
widely in traction gear (mostly combinations of tracks, wheels, and drums), number of axles,
driver position, tool position, and overall dimensions.
Few of these designs would have a significant and lasting impact on the industry, however.
Figure 2-3 highlights a selection of production tractors whose most salient features would
influence future models and later become enduring characteristics of the dominant tractor
design.

2.2.2 1900 to 1920: early history of the modern farm tractor

In 1903 the term "tractor"was first coined in advertisements by the Hart Parr Gasoline En-
gine company of Charles City, Iowa. At the time, horses and mules were the primary source
of draft power in the burgeoning American farming industry. In the USA the Homestead Act
of 1862 was still ongoing with minor revisions and motivated farmers to extend westward
from the northeastern cities. Earnest farmers tilled the wild soil and rapidly expanded the
total amount of available arable land [107. The large tractors (often steam powered) of the
time were more capable than animals at tilling the expansive tracts of land in the Midwest
prairies but were also unwieldy and expensive. These tractors were specialized tools aimed

at heavy tillage of large areas and as mobile motors to power crop processing machines.
During the late 1910s the agricultural industry in the USA became highly profitable as

food exports increased dramatically to feed resource depleted Europe and Russia during
and after WWI. Between 1915 and 1920 the agricultural output of Eastern Europe and
Northwestern Europe dropped by half and a third, respectively 1361. During those five years
USA farms sold at almost continually rising prices that drove their net income to almost
triple, and farmland prices to more than double, as farming became a more attractive
investment [21][371. American farms grew in number and size, yet farm labor was more

25
 Homestead Wes t Expansion Automobile
of railroad
replaces horse

Migration West 3yrs after

iWWI Ends i Highway
i
System
New land for Economic Few Companies
| Struggle K Survive Limit use of steel
wheels on roads

Plow large tract sPush to "modernize" No federal Transport Produce

of vir land farmin farm subsidies on Roads

Reduce Desire for Many jobs

damagte to soil urban life e in cities
to a 'Operate tractor
in same physical Scarcity
of laborers
space as horses

Improve sa fety, Increase

Box] Nomencaturecomfort and speed productivity
.4oof operati0+ on I

cI¶U
of laborer

oeveFaturE
oveu MnentOV
User Needs Sodal eactor n

t High drawbar pull perunit weight

m# High drawbar pull 4. Three-point hitch
$ Reduced cost * Pneumatic Tires
& Better perfonance/access on roads ) Automatic draft control
Better reliability A Four wheel layout
Lower weight O Power Take Off
Smaller size
' Easier to use and/or more versatile

Figure 2-1: Graphic chronology of tractor evolution as influenced by historical context and
stakeholder expectations.

scarce as the rural youth went to fight in WWI and later returned preferring an urban
lifestyle. Farm tractors became an attractive way to multiply the capacity of each laborer

26
 Steered Wheel Driven Wheel Driven & Steered Wheel [Idle Wheel

Steered Track Driven Track Driven & Steered Track 0Driven Drum ATool Q Driver

A B C D E F G H

I K L M N 0 P

QR S T U W X

Figure 2-2: Components with a dotted outline represent multiple possible locations in other-
wise identical layouts. Designs A through J are rear axle driven, K through P are front axle
driven, and Q through X are driven by both front and rear axles. A list of some production
tractors using each layout can be found in Appendix Al.

[981.
The blooming tractor industry innovated quickly as it received feedback from a rapidly

expanding customer base and adopted engineering knowledge from its younger but more

refined cousin, the automobile 1201. It was often the case that the farmer who owned a

tractor still had to own horses, which were more maneuverable and smaller, for cultivation

operations [20][78]. Very large tractors that had been used to open large fields in the

expanding West were too specialized and would lay rusting with little or no use after that

initial heavy ploughing operation 1107] 1113]. The demand for a less expensive, smaller, and
1281
lighter tractor was growing, and manufacturers new and old rushed to fill the void [1121.

The first tractor to meet the demands of the common farmer in size and price was

offered by the Bull Tractor Company in 1913 (Fig. 2-3). This lightweight tractor had three
wheels with a single drive-wheel and had an initial price comparable to a team of horses.

27
 Ca. 1900, Migration West: Government allows homesteading of USA's West. Large tracts of virgin land must be plowed,
this is the tractor's first market and is dominated by large, steam powered models.
1902 Ivel 1908 Ilart-Parr 30-60 Frm Cit

During and soon after WWI: American farming is highly profitable as produce is exported to Europe and Russia. Scarcity
of laborers makes tractors an attractive option. Many tractor companies and designs emerge
to 1913 Bull 1914 Wallis Cub 914 Moline Universal

1916 Nilson eo 1916 Tur 1917 Fordson F

1920's crash: As foreign agriculture recovers, American produce exports plummet and national agriculture suffers Low
cash customers and a crashing national economy cause most tractor manufacturers to perish.
1921IHC 1530 1924 IHC FarmAll 1932 AllisChalmers U

1939 Ford 9N Dominant design emerges

Ford Motor Co.'s tractor arm had already pushed the industry with
its Model F and its unprecedentedly low price in the 1920's. In 1939
they returned with the novel "three-point hitch", as well as adopting
the major industry advancements from the previous two decades.

High drawbar pull per unit weight Structural crankcase and transmission
7 High drawbar pull .Z Three-point hitch
$ Reduced cost * Pneumatic Tires
d Betterperformance/access on roads ) Automatic draft control
Better reliability Four wheel layout
Lower weight * Power Take Off
Smaller size *Low sales ** Average sales Sales leader
Easier to use and/or more versatile Feature is: Q absent Q evolving present

Figure 2-3: Graphic chronology of tractor evolution into conventional small tractor design.
More vehicle data and larger images are found in Appendix A2. Photo Credits: 1902 Ivel
[70], 1908 Hart-Parr 15-30 [1141, 1909 Avery Farm City [291, 1913 Bull [1011, 1914 Wallis
Cub [94], 1914 Moline Universal 661, 1916 Nilson1911, 1916 Square Turn1921, 1917 Fordson
F 11051, 1921 IHC 15-30 1621, 1924 JHC FarmAll 1351, 1924 Allis Chalmers U 11001, 1939
Ford 9N 1181

28
By 1914 it was the best-selling tractor in the country [107]. The tractor industry still had
reliability issues and production volume challenges that it would learn to solve partially
from automobile experts becoming more involved. In 1917 Henry and Edsel Ford launched
Fordson tractors. Their Model F quickly became the best-selling tractor in the world and
would eventually be produced at a price and volume that would raise the entry barrier to
tractor manufacturing beyond what most smaller competitors couldmusterj107). The Model
F was already highly reminiscent of today's modern small tractor and also of the traditional
automobile layout. It had four wheels, front wheel steering, rear wheel drive, and a trailing
tool. Not yet incorporated were pneumatic tires, a rigid tool attachment system (Three-
Point Hitch), and an exposed engine-powered shaft for powering implements (Power Take
Off).

2.2.3 1920 to 1950: farm tractors converge on a common design to power

American farms

In 1920, 166 companies in the USA manufactured farm tractors and had a combined annual

production of 203,207 tractors. These were dramatic increases from 1910, when only 15
farm tractor companies were in business and had a combined production of 4,000 tractors

[43]. These 166 companies were competing to define the shape of the "farm tractor" and to
distinguish themselves through innovative designs (a sample of tractor layouts is shown in
Fig. 2-2) [20] [78].
During 1921, a dramatic shift occurred in USA production. Agricultural output in
Eastern and Northwestern Europe had quickly recovered to pre-WWI levels to suddenly
make them largely independent of imported food [36]. Farmers in the USA had misjudged
international demand and food overproduction caused the prices of agricultural produce to
plummet. Farmers abruptly found themselves unprofitable and with outstanding bank loans
used to purchase farmland that had since collapsed in value[37]. Farm tractor production
plunged by two-thirds from 203,277 units in 1920 to 68,029 units in 1921[43].

The Great Depression and Stock Market Crash of 1929 would keep American farmers in
a difficult position through the 1920s and 1930s. It forced tractor manufacturers to adapt to
a low cash flow style of farming. In February 1922, the "Tractor Price Wars" started when
Fordson (a Ford Motor Co. brand) slashed the price of its popular Model F from $625 to
$395 [107]. Over the next 20 years a fiercely price-competitive tractor market would see

29
manufacturers converge on similar designs. Many manufacturers would disappear in this
"war", from 166 manufacturers in 1920 to only 38 in 1930. However, the industry's annual

tractor production had rebounded to 196,297 units in 1930, very similar to the output of

1920 [43]. Yearly total production of American tractors would keep rising until reaching
a peak in 1951, when 564,000 tractors were manufactured. By 1950 there were over 3.6
million tractors operating in American farms (about 1 tractor for every 6 people living on
a farm) and the internal combustion engine had become the primary source of draft power
for farmers [98].
Farm tractor production between 1910 and 1950 was significantly higher in the USA
than elsewhere in the world. In any given year, production in the USA was at least ten
times higher than any other single country, and at least five times greater than the net
global production excluding the USA. American tractors were in high demand domestically
and also exported extensively[56][49.

2.2.4 Emergence of the dominant tractor design features between 1910

and 1940

Major innovations that shaped the dominant tractor design and products that exemplified
them are discussed below.
1914 - The Wallis Cub (Fig. 2-3) was the first tractor to use the engine crankcase and
transmission case as structural components. Launched in 1917, the Fordson F (Fig. 2-3)
would leverage this construction style to reduce production costs by using less materials
and more streamlined manufacturing than its competitors. The Fordson F's runaway suc-
cess in domestic and international markets would help cement the structural crankcase and
transmission as features of the dominant tractor design.
1916 - The Square Turn (Fig. 2-3) featured innovations that signaled trends to come. The
Square Turn's name boasts its ability to control both of its drive wheels independently for
tight turning (even reversing one wheel while the other drives forward). The dominant
tractor design that would emerge later features a differential axle and allows independent
braking of either rear drive wheel for tight turning (a skilled driver may also use independent
braking as a differential axle quasi lock). The Square turn also used engine power for lowering
and raising farming implements even while the tractor is stationary.
1921 - All tractors sold in Nebraska must henceforth go through the standarized Nebraska

30
Tractor Test, the results of which are public [691. This test would go on to become the
mandatory national, and later international, standard for tractor testing. For the manufac-
turers, outstanding performance in the test can provide a major marketing tool. Some of the
more marketable results from the test involve towing a braked vehicle behind the tractor,
an operation generally better suited to tractors designed to pull heavy tillage tools behind
them.
1921 - International Harvester introduces the Power Take Off (PTO) as an option in the 15-
30 tractor (Fig. 2-3), allowing the tractor's engine to power actuators in farming implements
through a rigid speed-controlled shaft instead of using a flat belt. Implement manufacturers
rush to take advantage of this innovation.
1925 - International Harvester introduces its Farmall "General Purpose" (GP) tractor (Fig. 2-
3). The Farmall series would become the best-selling tractor series ever in the USA. Com-
pared to most other tractors on the market it: was lighter, had higher ground clearance,
utilized smaller front wheels (enabling tighter turns), had adjustable track width, and it was
as advertised for cultivating, plowing, and cutting.
1932 - Firestone pneumatic tires are offered as standard equipment on the Allis Chalmers U
(Fig. 2-3). Pneumatic tires allowed tractors on the growing network of paved roads (where
steel, lugged wheel were not permitted) and enabled farmers to increase fuel efficiency and
operate at higher speeds.
Circa 1935 - Diesel engines are advanced enough to become standard in farm tractors
[104]. This improves reliability (especially after storage periods), and gives the tractor a
wider high-power RPM operating band.
1939 - Ford and Ferguson introduced the three-point hitch tool mounting system in the Ford
9N (Fig. 2-3). The system efficiently leverages draft forces from heavy tillage tools to improve
the tractor's drawbar pull performance; a better implementation of the idea pioneered by
the 1916 Nilson Tractor (Fig. 2-3). The 9N featured hydraulic-powered coarse and fine
control over implement vertical position, reducing the drudgery of tractor driving and tool
attachment while also increasing the tractor's field capacity (actual acres worked per hour).
The three point hitch is the standard today for mounting trailing farming implements.

31
2.3 Analytical modeling of the conventional tractor's design

Insights into success of the conventional tractor design are elucidated by modeling its per-

formance and then exploring the effects of altering the design. The modeling of a tractor

on soil can be separated into two inter-related parts: calculating the distribution of forces
among all tires (the tires hold the tractor afloat and propel it forward) and, given that
load distribution, calculating the power consumption and other performance metrics at each
individual tire. A detailed description of modelling off-road vehicles in uneven terrain can
be found in [97].

2.3.1 Conventional tractor dimensions and relevant forces

Calculation of the tractor-applied forces at the tire-soil interface requires a force balance

of the tractor and farm implement system. Under the assumption of the tractor being a
laterally symmetric rigid body, being in steady-state equilibrium, and all the wheels having
their rotation axes orthogonal to gravity and parallel to each other, the tractor free-body
diagram (Fig. 2-4) can be simplified to include only:

" overall center of mass location and magnitude

" draft tool force direction, magnitude, and origin (center of pressure)

" location of ground reaction force points and associated vectors

" tractor orientation with respect to gravity (uphill or downhill slope)

Overall, the sum of the vertical force Vf on the front wheels is

1
Vf = (WT(xrcos(6) - ygsin(6))
Xf + Xr
+ D(YD + cOs(a) - XDsin(a)) (2.1)

+ Wj(-xjcos() - yisin(6))),

and the sum of the vertical force V on the rear wheels is

1
V = +(W (xr cos() + ygsin(O))
of +zXr

+ D(-yD +cOs(a)+ XDSn(a)) (2.2)

+ W(( +xr + f)COS() + yrsin())),

32
 Y9 R f

Vf

fDr

Figure 2-4: Free body diagram for farm tractor in 2D.

where xf is the distance from the tractor center of gravity (CG) to the front axle, xr is the

longitudinal distance from the tractor CGto the rear axle, WT is the weight of the tractor,

9 is the ground slope angle, y. is the distance from the CG to the ground, D is the tillage

force, YD is the depth of the tillage tool center of pressure (COP), XD is the longitudinal

distance from the tillage tool COP to the rear axle, a is the angle of the drawbar force

vector relative to the ground slope, W1 is the weight of the implement, xz is the longitudinal

distance from the rear axle to the tillage tool CG, and yj is the distance from the ground

to the tillage tool CG.

It is assumed in the conventional tractor configuration that only the rear wheels are

driven. To move the tractor forward at a constant speed, the rear tires must provide the

net horizontal force

Hr = Bf + Br + Dcos(a) + (W, + W1 )sin(0), (2.3)

where Bf is the force from the soil on the front wheel opposing vehicle forward motion, Br

is the force from the soil on the rear wheel opposing vehicle forward motion, D is the tillage

force, WT is the weight of the tractor, and W1 is the weight of the implement.

The calculation of the actual wheel torque necessary to achieve Hr and the calculation

of resistance forces Bf and Br requires further analysis described in Section 2.3.3.

33
2.3.2 Qualitative description of importance of soil-tire interaction in trac-
tor design

A refined terramechanic design can reduce the power lost at the soil-tire interfaces, something

especially critical for farm tractors which seek to minimize fuel consumption and damage to
soil. While drivetrain mechanical losses in a small tractor can be under 5%, power conversion
at the tire-soil interface usually involves losses of 30 to 60% [161.
The two major causes of power loss are soil deformation and slippage at the tire-soil
interface [108]. The effects of soil deformation from wheeled vehicles are observed in the
ruts they leave behind. As the wheel rolls forward it deforms soil ahead of it (known
as "bulldozing"). This deformation requires energy but achieves no useful work. Slippage

occurs when the tangential speed of the tire contact is faster than the forward speed of the

vehicle. Presence of at least minimal slippage is unavoidable as for a thrust force to occur
the tire must exert a shear force on the soil (therefore causing soil deformation). When the

shear strength of the soil is low relative to the traction being generated, the shear stress
may result in large shear deformation and thus higher slippage.
An efficient terramechanic design must strike a balance between sinkage and slippage.
The amount of power lost to slippage and bulldozing are both correlated to ground pressure,
but with usually with opposite effects[108][90]. As ground pressure increases, the shear
strength of soils with a. frictional component (most natural soils) increases, and thus less
shear deformation is provoked by a given shear stress. This reduces slippage and energy
losses provoked by it. On the other hand, as the ground pressure increases so does the

sinkage of the tire into the soil, which results in more energy lost to bulldozing.

2.3.3 Model for interaction of a single drive tire with soil

The tire-soil model summarized here is an implementation of that described by Wong 1081,
which is commonly accepted in terramechanics.
For analytically studying the tire-soil interface, it is helpful to investigate it as a 2D
system and separate the net stress into normal stress (normal to the wheel perimeter) and
shear stress (tangent to wheel perimeter). All weight-bearing wheels generate a normal
stress on the soil. Only braked or powered wheels generate significant shear stress on the
soil. In Fig. 2-5 stress distributions at the soil-tire interface are shown for a rigid, smooth

34
wheel in a homogeneous soil. Note that the soil is deformed plastically, as the wheel moves
through it and thus the soil surface is lower behind the wheel than ahead of it.

a)

Vehicle Direction

Wheel Rotation .•

Forces on wheel
from soil normal
stress

b)

Vehicle Direction

Wheel Rotation

Forces on wheel
from soil shear
stress

Figure 2-5: Stress under rigid driven wheels rolling on deformable soil. Stress at the tire-soil
interface is separated for clarity into normal (a) and shear (b). Wheels are moving to the
right and rolling clockwise. Nomenclature according to reaction force direction (as felt by
wheel) are shown in soil.

In agricultural soils, the pressure required to penetrate into the ground increases with

depth. Soil pressure as a function of depth is commonly expressed in terramechanics using

Bekker's [101 or Reece's equations [54]. Reece's equation is used in this analysis because its
soil constants are not a function of tire contact patch size. The resulting equation for soil
pressure is

p=(ck + wyk )(z/w)", (2.4)

35
where p is soil normal stress, c is soil cohesion, k' is the cohesion constant, w is tire width, -s
is the soil bulk density, k is the friction constant, z is the depth below the soil surface, and
n is the depth exponent (an experimental value relating penetration depth to penetration
resistance).

The shear strength of frictional soil increases with pressure, and the pressure exerted
by soil increases with depth. This means that a wheel operating while sunk in soil may be
interacting with different soil shear strengths along its perimeter. The soil strength can be

reasonably predicted by the Mohr-Coulomb shear strength equation as a function of normal

pressure,

s = (c + ptan(0))(1 - e-j(i)/k), (2.5)

where s is soil shear stress, p is soil normal stress, c is soil cohesion, # is soil friction angle,
k is shear modulus, j is shear deformation, and i is slip at interface.
To calculate the total reaction forces experienced by the tire when contacting soil, the
shear and normal stresses can be integrated along the tire's casing. If the deformed tire is
assumed to take the shape in Fig. 2-6, it can be separated into three sections: front circular
arc of the tire, flat horizontal section at the bottom of the tire (the depth at which the tire
total pressure matches the soil pressure), and rear circular arc of the tire. Tire sinkage and
deformation are therefore defined by the angles 0, Of, and 0r.
The net vertical force may then be mathematically expressed as

I~f

V =wR [p()cos(O)+ s(,i)sin(O)]dO

+ w2RPtsin(9) (2.6)

wR] 10, [p(O)cos()- s(O, i)sin()]dO.

The horizontal force is expressed as

36
 Or Of '1\
L\
x

Figure 2-6: Parameters of tire perimeter for calculation of forces at soil interface.

Of
H =wR [-p(z)sin(6) + s(, i)cos()]dO

1-L(Oc,R)
±w] s(9)dx (2.7)
fr
±wR [p()sin(O)+s(O,i)cos(O)]d].

For both equations, H is drawbar pull, V is vertical ground reaction, w is tire width, R is
tire radius, p is soil normal stress, s is soil shear stress, z is the maximum depth of the tire
into soil, and L is the length of the tire's deformed flat section. The angles c,Of, and 0r
define the tire shape and sinkage into the soil (Fig. 2-6).
All tires are assumed to be rigid (0c is zero) if the maximum soil normal pressure p exerted
on them is less than the sum of the tire carcass stiffness and the tire inflation pressure. Once
the tire has sunk into the soil (depth z) to the point where the soil normal pressure p exceeds
the tire inflation pressure and carcass stiffness, it is assumed the tire starts deforming and
soil depth z remains constant (i.e. 0e grows to.support the extant vertical load, see Fig. 2-6).
It is often the case that nominal reaction forces at the tires are known but not the tire
sinkage, deformation, and slippage - these are needed to solve for power consumption and
maximum drawbar pull. These values must first be solved for idle wheels to find the their

37
horizontal force H, which will oppose vehicle motion and must be overcome by driven wheels.
Slippage i is assumed to be zero for tires that are idle (are not driven) and thus the soil
shear strength s(i) terms in Eq. 2.6 and Eq. 2.7 are zero. In this case only the angles 0,Of,

and 0r must be solved for. These are found by allowing the tire to sink into the soil until
the vertical load on that tire matches the soil reaction V from Eq. 2.6. All values are now

known to calculate from Eq. 2.7 the horizontal force H on the tire (which will be negative
and opposing vehicle motion in this case).

For the driven wheels, an optimization routine must be implemented to simultaneously

solve for slippage i as well as the tire deformation and sinkage parameters 0, Of, and 0,.
This is achieved by finding the solution that minimizes the power consumed at the wheel
while maintaining force equilibrium (vertical load at the tire is equal to V from Eq. 2.6 and

horizontal load is equal to H from Eq. 2.7).

2.3.4 Validation of Tractor Model with Published Data

To verify the model's accuracy as implemented, in this section its outputs are compared
to published data on production tractors. Experimental data were obtained from Battiato,
Diserens, and Sartori (7]{8], where four different sized production tractors were tested in
various soil conditions. To test a tractor's drawbar pull performance, it towed a braking
tractor behind it via an instrumented cable. The braking tractor was set to generate only
the desired horizontal drawbar pull force on the tractor being evaluated. The pulling cable
attachment height matched the CG height of the tractor being evaluated. The model shows
good agreement with the experimental data (Fig. 2-7).

2.4 Results and discussion: analytical model insights into trac-

tor dominant design

This section summarizes, based on governing physics, why the dominant tractor design is a
viable engineering product.

2.4.1 Advantages of conventional tractor weight distribution

Modeling (Fig. 2-8) and historical data (Fig. 2-9) show that drawbar performance is maxi-
mized by placing 70 to 80% of the tractor's weight on the rear wheels. Shifting the center

38
 ofm arIsotnf aI*,n ramantalate
50p Model: 5700lbs Tractor
0 Experiment: 5700lbs Tractor
SAModel: 9000lbs Tractor
45 A Experiment: 9000lbs Tractor
' Model: 11300lbs Tractor
4-w -- -0-0- ' KExperiment: II300lbs Tractor
40
0- 0 4 Model: 15000lbs Tractor
0 Experiment:
- - - Tractor
15000lbs -

-
35

30' m -€ D *-

625 AA
-LA A& &g, A-A-J&-A
-9
or Ao-&

l5~
100

Recommended Real-World Slip

0
5 10 15 20 25 30
slip (%)

Figure 2-7: Comparison of tractor model as described in section 2.3 to published experi-
ments. Experimental data from work of Battiato, Diserens, and Sartori [71181. Model has its
best accuracy between 5% AND 20% slip, Which is the range recommended for farm tractor
operation [16][108][90][31].

of mass forward reduces puffing capacity while shifting it backward produces a negligible
increase in pulling capacity but dangerously increases the risk of upending the tractor.
Figure 2-8 shows modeling results for the effect on tractor pulling performance from
weight, weight distribution, and draft magnitude. In Fig. 2-8, weight distribution is defined
as the value when the vehicle is static at zero drawbar pull. The effective weight distribution
during operation is accounted for during simulation calculations. Note that when moving
along the 'Weight distribution on rear axle (%)" axis, power required to move (color bar
value) is reduced by shifting weight backwards until it asymptotes at around 70% of the
tractor weight on back wheels.
The historical data in Fig. 2-9 were obtained from the Nebraska Tractor Test archives

[69]. These tests are a standardized method to evaluate the performance of farm tractors.
Test results are public and often used by manufacturers when promoting their tractors. For
the manufacturers, one of the more marketable parts of the test involves the maximum
generated drawbar pull force when towing a braked vehicle behind the tractor. Before 1950
the tests were performed on soil instead of the concrete track now used. Farm tractors below

39
 25 hp tested between 1941 and 1950 (Fig. 2-9) were selected for comparison to the trends
established in Fig. 2-8.
For the Nebraska Tractors Tests, engineers employed by the manufacturer whose tractor
was being tested were allowed to ballast their vehicles as they preferred before testing began.
The preferred setups provide valuable insight into what adjustments the company's engineers
believed would maximize their tractor's performance. In Fig. 2-9, it can be observed that
the engineers would generally set up their tractors to maximize drawbar performance by

increasing vehicle mass and placing 70 to 80% of the tractor's total weight on the rear
wheels. These adjustments are supported by the findings in the Section 2.3 model.

20

18

16
x104
14
2 0

12
2-
10
1.5
8

0 6
3000 0-

4
2000
0 2

0
70 80 90 0 draft magnitud e (N)
Weight distribution on rear axle (%)

Figure 2-8: Simulation data for tractor configurations with varying weight, weight distribu-
tion, and draft load. Demonstrates that optimal weight distribution for drawbar pull is 70%
to 80% of vehicle mass on the rear wheels. The semi-transparent purple frontier on the left
represents where tractor wheels slip fully without generating progress or where any wheel
sinks past its radius. The semi-transparent brown frontier on the right represents when the
tractor upends and flips backwards.

2.4.2 Advantages of tillage tool rigid mounting behind the rear axle

Rigidly mounting heavy tillage tools behind the rear axle is a major characteristic of the

dominant tractor design. The utilization of draft forces to increase the vertical load on the

40
 <25hp Wheeled Tractors, 1941-1950 Nebraska Tests on Dirt Track
3000 0 Rear Bias Ballasted(%) Rear Bias Factory(%) A Mass Ballasted (kg) 0 Mass Factory (kg) 90

2500 8
70
2000 60.
504
21500
.2 40
~1000 3
20
500
100.

0 0

Figure 2-9: Drawbar pull data compiled from Nebraska Tractor Test Archives [691. Ahighly
publicized part of the Nebraska Test measures maximum drawbar pull, this table shows how
engineers would set up their vehicles for the test. Notice that, in aneffort tomaximize
performance, company engineers wouldballast theirtractors to have about 70to 80% of the
total vehicle mass on the rear wheels.

rear wheels (and thus the maximum drawbar pull) was acritical enabler to reducing the

size and cost of tractors to the point where they could be ageneral purpose tool for the
common farmer. From the free body diagram (Fig. 2-4) and the terramechanics theory in
Section 2.3, the following observations are made about rear tool mounting: it isefficient for
tillage, it isconvenient for the user, and it facilitates manufacturing. Abackground on the
introduction of this "Three Point Hitch" mounting style is given in Section 2.2.4.
The maximum pulling force the tractor can produce increases approximately constantly

with the tillage tool draft force D.As seen in Fig. 2-4, tillage tool draft force Dgenerates
a net moment on the tractor - the backward component of Dacts on the relatively short

lever arm YD,while the downward component acts on the longer lever arm XD. The net
moment wants to lift the front end wheels off the ground. This increases the vertical load
on the rear wheels Vr, which augments the soil's shear strength and thus the traction force

Hr generated at agiven soil-tire slip. It also reduces the vertical load on the front wheels

V, reducing soil bulldozing force Bfdue to tire sinkage.

41
 Mounting the farming implement on a hydraulically actuated rear hitch is practical and
safe for the operator. The driver need only reverse the implement-less tractor towards an
implement, lock the implement attachment points, and drive away. Placing the implement
behind the rear axle does not directly constrain the length or width of the implement since
the tractor tires will not physically interfere with the implement or immediately drive over
soil that the implement has already worked on. This in an important advantage for more
powerful tractors that can pull several ground engaging "bottoms" at once. Finally, the

driver is physically safe from the implement behind them while driving the tractor. This
can be especially important for implements that have moving parts powered by the engine
or that launch significant amounts of debris.
Placing the implement (tillage tool) behind the rear axle can facilitate manufacturing

via beneficial component packaging and a short load path between the traction gear and
tool. The drive axle, hydraulics, and power take-off (PTO) shaft are all around the same
location where engine power is being delivered (thus creating a short load path between
heavy drawbar pull tools and the tractor's driven tires). Additionally, the space behind
the rear axle can be fully dedicated to the implement, its attachment linkages, and its
power sources (hydraulics and PTO). This setup also allows placing the engine over the
front wheels and using the drivetrain's transmission case and crank case as the structural
"frame" of the tractor which minimizes the amount of components, facilitates fabrication,
and reduces mass. This method of construction was critical to helping Henry Ford create
farm tractor assembly lines11071.

2.4.3 Advantages of conventional rear wheel drive

Driven side-by-side rear wheels connected via a differential axle are well suited for farm
tractors - some of their advantages are tight turning, simple construction, and being able to
leverage trailing tool forces for improved performance.
Tight turning is a feature of the conventional layout for two main reasons: a large steering
angle enabled by small radii front wheels and the ability to independently brake each rear
wheel.
Tractor construction was simplified by side-by-side rear drive wheels since it allowed use
of a standard differential axle (as also engineered for emerging automobiles) and required
less structural reinforcements. If front (steered) wheels are driven, additional linkages must

42
be added to the driveline to enable steering. Since most heavy draft implements are attached
closely behind the rear axle, the tractor load path from the rear wheels' traction force to
the implement attachment points is short and most of the tractor's structure need not be
reinforced to support the tool's loads.
A tillage tool attached behind the tractor will cause the effective weight distribution of
the tractor to shift rearward. This added vertical load on the rear wheels can increase their
maximum tractive force as supported by production vehicle data141 and modeling.

2.4.4 Advantages of four wheel rectangular layout

A four wheel rectangular layout is sound from a manufacturing, stability, and terramechanics
perspective. By rectangular, it is meant that the front and rear axle dimensions allow the
rear wheels to run over the "ruts" or "tracks" formed by the front wheels. In other words,
both right wheels are longitudinally inline with each other, as are the left wheels. More than

four wheels would increase manufacturing and maintenance complexity (and cost). Fewer
than four wheels decreases the stability of the tractor [26][89].
There are two key advantages of a rectangular wheel layout from a terramechanics per-
spective: improved crop yields and better tractive efficiency. It is less detrimental to crop
yields to drive over the same patch of soil in the field multiple times (as is the case with
inline front and rear wheels) than it is to drive over more areas of soil on the field only once.
Applying this when planning routes for field operations is called "Controlled Traffic" and has

been proven beneficial in farm fields across the world [1021 [31 [241 [791 [1711251. Each tire
pass strengthens (compacts) the patch of soil it runs on, making it a better rolling surface
for trailing tires. This means the idle front wheels can partially "pre-compact" the soil for
the driven rear wheels, thus improving the vehicle's maximum drawbar pull and tractive

efficiency11081141][51].

2.5 Conclusions

This research chapter describes how the dominant farm tractor design evolved mostly in the

USA between 1910 and 1940. It provides a historical reflection on the priorities that drove
the creation of the tractor. It also uses engineering analysis to justify why the conventional
tractor design is well suited to farming and has endured the test of time.

43
 The conventional farm tractor evolved not only to maximize farming performance but
also to satisfy intense pressures from the social and political context it developed in. In par-
ticular, an unpredictable economic backdrop and sharing engineering with the new automo-

tive industry rapidly pushed the tractor industry towards standardization and competitively
priced mass manufactured vehicles by a few large corporations.
The conventional farm tractor has a sensible design that is easy to use, easy to manu-
facture, and offers efficient performance. In particular, at least within the constraints of its
standard overall layout, it has nearly optimal weight distribution and tillage tool attachment.

Contemporary engineers may wish to use this chapter to explore which features of the
existing dominant tractor design are applicable to their own novel tractor layout implemen-
tations. Consider that the dominant tractor design evolved mostly in the USA between

1910 and 1940 and is sold worldwide today; however, during that period, and since, Amer-
ican farms have been about 100 times larger than at least 80% of today's farms globally
[58][141199][591. As mechanization increases in developing and emerging markets, engineers
may question if the dominant tractor design is the ideal one to sell to these small farms.
At possibly a different end of the price spectrum, new technologies like autonomous driv-
ing may enable design freedoms for modern tractors that were not available in the 1940s.
This Chapter provides insights on tractor features that enhance user comfort and farming
productivity, as well as the physics behind tractor performance, which may be of value to
engineers designing new farming equipment for developing and developed markets around
the globe.

44
Chapter 3

Design of a specialized tractor to

replace draft animals in small farms

3.1 Introduction

A tractor designed specifically to meet the needs of small farmers in India, who would
otherwise use draft animals, has the potential to create significant impact by improving

farmers' economic health and India's overall food production capacity [50, 111, 63]. Small
farms (< 2 ha) are common; in India the average farm size has steadily decreased from 2.28
ha in 1971 to 1.08 ha in 2016 [75], and globally (84%) of farms are less than 2 ha in size

[58. Most small farmers use a pair of bovine draft animals known as bullocks (sometimes
called oxen in other countries) for all or most of their farming operations, supplemented by
manual labor or a hired tractor 152, 31]. Bullocks are compact, highly maneuverable, and
have a low capital cost, making them well-suited to the technical and economic constraints
of small farms.
Conventional tractors, which are an icon of modern farming, are able to produce much
higher farm yields than bullocks [39, 33, 73, 40]. Farm tractors increase the capacity of each
agricultural worker and enable larger, more profitable farms 198]. The Indian Agricultural
Ministry estimates that farm tractors increase farm yields by 5 to 20%, reduce wasted seeds
and fertilizer by 15 to 20%, and reduce farm labor by 20 to 30% [52]. However, tractors
have not yet been able to replace key bullock features of maneuverability and compactness
that are essential to work on a small Indian farm [57. Tractors also have a high upfront

45
cost that puts them out of reach of many small farmers in low income regions[87]. As a
result, small-scale farmers are constrained to the slow speed of bullocks and a lack of access
to suitable modern, more effective made-for-tractor tools [31, 30].
Although tractors are more expensive upfront, they are less expensive than bullocks in
the long term. Fig. 3-1 shows the initial cost and 15 year operating costs for a bullock pair,
a financed tractor, and a tractor bought upfront. A bullock pair is approximately twice as
expensive as using tractors over 15 years. There are alternatives to acquiring tractors at
full price up front, like financing and renting, but they are inaccessible to many farmers

[27, 38, 87] and have financial drawbacks. For example, those who rent tractors forgo using
the tractor for supplemental income work and risk not having access a tractor when they

need one if demand is high. An ideal vehicle would retain the low upfront cost of the bullocks
and the low overall cost of the tractor (e.g. the proposed vehicle in Fig. 3-1). Such a vehicle
would have an higher value proposition than both bullocks and tractors.

12
U Initial Purchase Cost
10
a Other Costs for first 15 years

2-
0
Bullock Tractor Tractor Proposed
Pair Financed Bought Vehicle
Upfront

Figure 3-1: Cumulative costs of farming 1 ha with bullocks or small farm tractors. The
ownership costs of bullocks outweigh their low initial purchase cost. Ownership costs are
recurring costs like maintenance, loan payments, animal feed, and fuel. The purchase cost
shown for Bullkey is a target as mentioned by farmers during interviews - the final price
is not yet determined. A breakdown of the costs shown here is provided in Appendix Bl.
Values are from [81, 80, 22, 31, 30, 47].
.

To elucidate both the financial and functional requirements of a tractor specialized for
small farmers, the author interviewed stakeholders of small farming in India regarding local

46
agricultural practices and the suitability of existing alternatives. The author spoke with
stakeholders at 12 locations in the Indian states of Maharashtra, Tamil Nadu, Gujarat,
Rajasthan, Madhya Pradesh, and West Bengal. Stakeholders included farmers, research
organizations, governments, and tractor manufacturers and dealers. A key observation from
these visits was that small farmers used bullocks both because of their low capital cost and
because of bullocks' suitability to the narrow inter-row spaces in a farm field. Bullocks
have a smaller width than tractors and are more maneuverable. These characteristics allow
bullocks to walk between rows of growing crops later into the season when crops are taller
and wider, leaving less space between crop rows. Compared to tractors, bullocks require
less space to turn at row ends, and can better traverse unfinished dirt paths leading to
farm fields. These critical features of low upfront cost and ability to access narrow spaces
are generally not present in commercially available small tractors. The few tractors that

approach the purchase price of bullocks cannot match the bullock's maximum pulling force,
a key requirement for seamlessly replacing them.

10
9 small local manufacturers
0.
8
7 *4

0
*

5
%

0 500 1000 1500 2000 2500 3000

tractor mass (kg)

Figure 3-2: Relationship of tractor sales price to tractor mass for common Indian tractors,
as compiled by the author. The full dataset is provided in Appendix B3. To be competitive
with bullocks' sale price, a tractor would have to cost ~100k INR.

In conventional tractors, lower cost often comes at the expense of pulling force. Pulling

force is related to a tractor's mass, which is correlated to purchase price. To be sold for

a price comparable to bullocks (~100k INR, as shown in Fig. 3-1), a tractor would likely

have a mass between 350 and 500 kg given the current trends of the Indian tractor market

47
(Fig. 3-2) [31, 4]. The layout of a conventional, rear-wheel-drive tractor with a mass of
350-500 kg would only produce a maximum pulling force of -60% of its weight (2060 N to
2940 N) in near ideal tilling conditions, and closer to ~35% of its weight in soft soils (1200 N
to 1720 N) [13, 115, 41. This could not, under most conditions, match the maximum pulling
force of a bullock pair (-2800 N [42, 1061).
A lightweight tractor capable of replacing bullocks in small farms, and thereby improving

farmers' livelihoods, must match bullocks' pulling force, their purchase price, and their
ability to enter narrow spaces. No major manufacturers currently make tractors in the

bullock price range of <100k INR or with dimensions comparable to bullocks. The lightest
tractors (Fig. 3-2) have limited pulling force (approximately 1720 N to 2940 N depending
on soil conditions and actual mass) and are unable to access narrow inter-row spaces. These

vehicles are made by small-volume local manufacturers near smallholder farms 1821 and are
not widely distributed. A manufacturer capable of making and widely distributing a low-
cost tractor with the characteristics to near seamlessly supplant bullocks could likely access
~80 million farmers currently underserved by mechanization [82, 76].
The goal of the research presented herein was to codify the parametric behavior of tractor

performance, and combine this understanding with market insights to generate a tractor
architecture well suited to the needs of Indian small farmers. We present our modeling
approach, proposed design, and evaluation of a novel tractor.

3.2 Physics behind maximizing traction performance

A terramechanics-based physics model of tractors' performance derived in our prior work

[44] (fully described in Ch. 2) was used to gain parametric insights on how the design of a
small tractor could be manipulated to maximize drawbar force per vehicle mass. Figure 3-
3 shows a free body diagram of the main forces acting on a farm tractor overlaid on a
conventional tractor layout. For a tractor to perform an operation successfully, two main
conditions based on the free body diagram must be met: (I) the vehicle must not tip over,
which necessitates positive vertical ground reaction forces Vf and V,; and (II) the tractor
must achieve forward motion, which occurs when the drawbar pull force F (the sum of
traction forces H and bulldozing forces B in Fig. 3-3) is greater than the tool draft force
parallel to vehicle motion, Dcos(a). Additionally, when building a low-cost tractor, it is of

48
interest to include one more condition: (III) maximize the drawbar pull to mass ratio of the
tractor.

Rr 0 X1 W,
Bf

Vf Br XD Y1
E) Vr Hr yD
Vr

the tractor
Figure 3-3: Force free body diagram for a conventional small tractor. Shown are
and
ground reaction forces (Vf,Vr,B,Br, and Hr), which support the tractor weight (WT)
including ground slope (0), tractor CG location
tool draft (D). Key dimensions are shown,
(xD andyD),
(Xj, Xr, and yg), tool CG location (xi and yi), tool draft center of pressure
and tool draft angle (a).

Condition (I) was evaluated by assuming the vehicle would rotate at points directly

below the wheel axles and solving for the reaction forces at V, andV, giving respectively:

1
Vf= f+J(WT(xrcos() - yqsin(O))

+ D(YD + cos(a) - XDsin(a)) (3.1)

+W(-xcos(6) - yisin(6))),

and

1
V= (WT(xrcos() + ygsin(6))
Xf±+Xr

+ D(-yD + cos(a)+XDs(a)) (3.2)

+W1((xI+ Xr+Xf)COS(O) + yjsin(6))).

axle, xr is the
Where of is the distance from the tractor center of gravity (CG) to the front

longitudinal distance from the CG to the rear axle, WT is the weight of the
tractor. 6 is the
force,
ground slope angle, yg is the distance from the CG to the ground, D is the tillage

49
YD is the depth of the tillage tool center of pressure (COP), XD is the longitudinal distance
from the tillage tool COP to the rear axle, a is the angle of the draft force vector relative
to the ground slope, W1 is the weight of the implement, x is the longitudinal distance from

the rear axle to the tillage tool CG, and yj is the distance from the ground to the tillage
tool CG.
Checking condition (II) and designing for condition (III) required an analysis that con-
sidered the physics of tire-soil interactions to calculate traction force H and tire bulldozing
force B. The soil exerts a pressure on the tire (normal to the wheel perimeter) and a shear
stress (tangent to wheel perimeter). All weight-bearing wheels generate a normal stress on

the soil (i.e. flotation). Only braked or powered wheels generate significant shear stress
on the soil (i.e. traction). The normal and shear stresses at the tire-soil interfaces were
calculated from the soil's mechanical behavior.
To calculate the soil pressure p along the tire's perimeter, a common equation used in
terramechanics was applied [54]:

p = (ck + wyk)(z/w)", (3.3)

where c is soil cohesion, k is the cohesion constant, w is tire width, y, is the soil bulk

density, k is the friction constant, z is the depth below the soil surface, and n is the depth
exponent (an experimental value relating penetration depth to penetration resistance).
The soil shear stress s is a function of tire-soil pressure and soil properties, and is scaled
by deformation at the tire soil interface represented by term 1 - e-j(i)/k [101:

s = (c + ptan(#))(1 - e-j(i)/k) (3.4)

where # is soil friction angle, k is shear modulus, and j(i) is the shear displacement at the
tire-soil interface, which is a function of tire slip i. Tire slip i is defined as 1 - , where S

is the forward speed of the vehicle, while R and w respectively are the effective radius and

the angular velocity of the wheel being evaluated for slip.

To calculate the total reaction forces on the tire when contacting soil, the shear and
normal stresses were integrated along the tire's casing. If the deformed tire is assumed to
take the shape in Fig. 3-4, it can be separated into three sections: a circular are at the front
of the tire, a flat horizontal section at the bottom of the tire (the depth at which the tire

50
total pressure matches the soil pressure), and a circular arc of the rear of the tire. Tire
sinkage and deformation can therefore defined by the angles 0, Of, and 0, in Fig. 3-4.

Figure 3-4: Parameters of tire perimeter for the calculation of forces at the tire soil interface.

OrrO
Each tire's vertical (flotation) force must satisfy Eq. 3.5. From this, the tire shape angles

Oc,Oy, anlGr can be solved for via acontrol strategy as shown in 85].

V =wR] [p()cos()+±s(O,i)si(O)]dO

+±w2RPtsin(O) (3.5)

+wRj [(co()- s(O.i)sin()|dO]

0 force Bcannowbecalculated usingEqs.3.6 and3.7,

Thetraction forceH and bulldozing

respectively.

The trctionforce H =wRj

and ulldozng[s(O,i)cos()
fce a o O)]dO
ecluae sn q.36ad37

wL(O 7 s()dx (3.6)

Gor
±wR [p(0)sin(0) + s(0,i)cos(0)]dO]

51
 rf
B =wR] [-p(z)sin(O)]dO (3.7)

In these expressions, w is tire width, R is tire radius, and L is the length of the tire's
deformed flat section.
The drawbar pull from a single tire is the difference between its traction force H and its
bulldozing force B (Fig. 3-3). The drawbar pull of the tractor is the sum of the drawbar
pull from all of its tires. For a tractor with n number of tires, this is

n
F = Z(Hv - Bv). (3.8)
v=1

The forces exerted on agricultural soil by tires affect the soil's mechanical properties
(apparent in the plastic deformation in the soil in Fig. 3-4). Each tire pass compacts and
strengthens the patch of soil it rolls over, improving the surface for trailing tires151, 41].
Compaction is accounted for as an increase in the soil's cohesion c and bulk density 711081.
Figure 3-5 is an idealized diagram demonstrating the interactions of inline drive tires on soil
during loading, unloading and reloading.
Figure 3-6 presents a sensitivity study of drawbar pull and tractive efficiency for a con-
ventional tractor in a common soil, loamy sand, to highlight the influence of key design

parameters on small tractor performance. Tractive efficiency is the efficiency in converting

power at the drive axle(s) into useful work. It is defined as 7 = (F*S)/P where, q is tractive
efficiency and P is power delivered to the wheel. The terramechanics model described here
is highly non-linear and depends on a large number of inputted tractor parameters. Fig-
ure 3-6 demonstrates that the maximum drawbar pull (net horizontal force) is approximately
linearly related to the tractor mass for a large range of values.
In conventional tractors this leads to two usage trends: tractors are ballasted to increase
their mass for work when a high drawbar pull is required, and high drawbar pull tools are
typically mounted to the tractor behind the rear drive axle. This results in the rear axle
supporting both the vertical draft forces and the tool weight, and in weight transfer from
the front axle to the rear axle due to the moment generated by these forces. However,

52
 Forward Travel
Tire Roll

Pass
x1
2 ndPass

F2 = H2 - B 2 F1 H1 - B,
F2 > F1

h2
Soil stronger above h2 Soilstrongerabovehi

Figure 3-5: Illustration of tire-soil interaction and multi-pass effect used in analysis. h is the
depth of the compaction effect on the soil. The 2nd pass tire, which is rolling on compacted
soil, can generate more drawbar pull F than it would on fresh soil since it sinks less into the
soil (reducing bulldozing force B) and the soil can provide a higher shear force (increasing
traction force H).

while maximum drawbar pull may increase, tractive efficiency may decrease - showing the
importance of correctly matching tractor mass to tire size (and thus ground contact shape
and pressure distribution). Too little weight on the tires or tires that are too wide may
apply insufficient pressure to the soil, resulting in a soil that requires excessive deformation
j to produce sufficient drawbar pull and therefore in power losses. Excessive weight on tires,
or tires that are too thin for the required weight, will increase pressure on the soil to a
detrimental degree, causing the tires to sink into the soil and exacerbating power losses to
bulldozing force B.

3.3 Design exploration

The tractor performance model from Sec. 3.2 was used to identify beneficial design features
to incorporate in tractors that are well suited to small Indian farmers. These features were
combined to create the Bullkey tractor layout (Fig. 3-7). The Bullkey name is a portmanteau
of bullock and key - indicating its goal of being the key to unlocking the bullock market to
mechanization.

53
 texture: loamy sand; cover: maize stubble
9000
-Tire Width
+Tire Radius
8000 +Vehicle Mass
7000 +Rear Bias
+Tool Location
$6000

5000,
4000
3000
-5 0 -25 0 25 50
70 7

65-

S60

Tire w idtm
>50 -- Tire Radius
Veh icle Mass:
45 Rea rBias
IToo I Location!
40
-50 -25 0 25 50
Variation(%)
Nominal Values
Tire Width Fr.: 132mm Mass Rear Bias: 0.55
Tire Width R.: 203.2mm Tool Loc.: 0.6m after rear axle
Tire Radius: 272mm Tool Mass: 100kg
Vehicle Mass: 850kg Tool soil depth: 200mm

Figure 3-6: Sensitivity analysis of drawbar pull and tractive efficiency at 15% tire slip
(a typical heavy tillage operating point 1901) for a conventional small tractor. Data were
generated using the terramechanics model described in Sec. 3.2. Soil conditions for typical
loamy sand are from 181. In the sensitivity analysis, variables were varied t50% from their
nominal value.

54
 A)

Front drive wheel

KM
B) AM-

B)Xj WT+I X,

R 09

Bf Hf XD 0
Br

\ _ _ D

Figure 3-7: (A) Isometric view with labeled soil engaging components and (B) force free
body diagram for Bullkey tractor.

3.3.1 Physics-based design insights

The physics-based theory of Sec. 3.2 led to insights about the behavior of lightweight tractors
that can improve their design and functionality for smallholder Indian farmers currently
relying on bullocks as a source of draft power. The following design strategies incorporated
into the Bullkey tractor (Fig. 3-7) maximize traction performance while incorporating or
improving on many of the bullock maneuverability advantages described in Section 3.1.
Support the tractor mass almost exclusively on driven wheels: Only driven wheels apply
a positive (drawbar pull generating) shear stress, s, on the soil. The maximum drawbar pull,
F, that a tire can generate (Eqs. 3.6-3.8) is limited by soil shear strength, which depends
on tire-soil pressure, p, and soil cohesion, c (term c + ptan(<) in Eq. 3.4). The soil's shear
strength can be improved by increasing pressure or by increasing soil cohesion. such as via

55
soil compaction induced by inline drive wheels (as in Fig. 3-5). Increasing tire-soil pressure
for drive wheels is best achieved by placing more vertical load on the tires, because, reducing
tire width, w, or radius, R, to lower their contact area would also scale down the magnitude
of traction force H in Eq. 3.6. It is also beneficial, to limit pressure on non-driven wheels
to only what is needed for stability. Idle wheels detract from the tractor's drawbar pull F
since they generate no measurable traction force H (given that s = 0) and can still generate
a significant bulldozing force B which increases with applied tire-soil pressure p (Eq. 3.7).

Shifting weight towards the driven tires is fundamental to achieving a high drawbar pull
to mass ratio. A pneumatic agricultural tire can generally generate as drawbar pull no more
than 80% of the vertical load it supports[115, 13}, and a conventional tractor design has 50

to 80% of its total mass on its driven rear wheels [441. Shifting more weight to the rear

wheels in this layout would increase the risk of upending the tractor and reduce vehicle
safety. A conventional tractor is therefore nominally able to pull up to 64% of its operating
weight (even less if considering the detrimental bulldozing forces from idle wheels) in near
ideal conditions, and much less in non-ideal conditions. Changing the layout to support all
of the mass on the drive wheels should increase the maximum pull capacity to 80% of the
vehicle's operating weight. If a tractor layout must use additional idle wheels for stability,
they should be designed so that stability can be achieved while only lightly loading the idle
wheels - therefore limiting the detracting opposing force they can generate and maximizing
the mass supported by drive wheels.
Match tire ground pressure to required soil shear stress by operating between 10% and
25% tire slip: A tire slip of 10 to 25% has been found to be an efficient compromise between
energy losses to soil shear deformation j (which is a function of tire slip i) and to soil
bulldozing B [10, 108, 115]. A well designed tractor should have its mass and tires sized
appropriately to reach its desired drawbar pull F in that tire slip range. To increase the
drawbar pull generating traction force H, one must increase the applied soil shear stress
s, which increases with i (Eq. 3.4), or increase the tire contact area (term wR in Eq. 3.6).
Some soil shear deformation must always exist at the tire-soil interface to generate a traction
force H. Reducing tire-soil slip while maintaining constant applied shear stress requires
increasing the soil pressure, p, which is typically done by adding ballast to the tractor.
However, increasing pressure also results in a larger tire bulldozing force B (Eq. 3.7), which

is detrimental to drawbar pull F (Eq. 3.8). If instead, ground pressure is adjusted by

56
changing tire size (i.e. contact area), the wR term will be affected in both H (Eq. 3.6) and
B (Eq. 3.7), causing them both to either increase or decrease simultaneously. Therefore, an
all encompassing design rule cannot be given but it is recommended to use the model from

Sec. 3.2 to select tire sizes and a weight distribution that generate sufficient drawbar pull
while staying in the desirable tire slip range.
Use inline drive wheels with similar vertical loads: Compared to side-by-side wheels,
inline drive wheels increase tractor drawbar pull and efficiency because the rear drive wheel

operates on soil that has become stronger (higher cohesion, c, and bulk density, -ys) after
being compacted by the front drive wheel [108, 9]. In conventional tractors the front drive
wheels are much smaller and lightly loaded compared to their rear side-by-side drive wheels
- so the front wheels do not strengthen the soil significantly for the rear drive wheels. In
agriculture, soil compaction is often considered undesirable because it hinders crop growth.
However, inline drive wheels leverage a technique known as "controlled traffic", in which
one patch of soil is driven over multiple times rather than driving over more areas of soil
only once. This method takes advantage of the fact that if all tire passes are equivalent,
compaction will be highest after the first pass and much lower for subsequent passes [51, 861.
This method is less detrimental to crop yields and has been proven in farm fields across the
world [102, 3, 24, 79, 17, 25].
Add a mount for high drawbar tools between both driven axles: Adding a mount for high
drawbar tillage tools between the front and rear axles uses the downward forces from tillage
(D * sin(a)) to increase the vertical loading on both the front and rear wheels, respectively

Vf and VR (Fig. 3-7). This results in higher soil-tire pressure, p, and thus higher soil shear
strength (represented by c + ptan(<) in Eq. 3.4). If both axles are driven, this produces a
higher maximum traction force H at both drive tires and increases the tractor's maximum
drawbar pull F (Eqs. 3.6) and 3.8.
Additionally, the central mount improves steering authority and stability by firmly plant-
ing both wheels on the ground, which allows the operator to safely operate the proposed
tractor design near its performance limits. In contrast, the draft force, D, in the conven-
tional tractor design (Fig. 3-3) causes the front wheels to become unweighted; even though
the horizontal draft component, Dcos(a), is typically larger than the vertical component,
Dsin(a), it exerts a torque over a much shorter moment arm (YDvS. XD). This unweighting

of the front wheels can cause the vehicle to upend (i.e. tip over backwards) and severely

57
 Layout
Beneficial Design Features User Need Met A B C D
Weight transfer during tillage improves drawbar pull drawbar pull V I/
Weight transfer during tillage improves steering authority safety, comfort
Safe to operate near tillage force limits (will not upend) safety, drawbar pull
Tillage tool is near farmer's driving line of sight comfort, ease-of-use
All drive tires are in a single lane with farmer and tool narrow, low compaction

Table 3.1: Occurrence of desirable design characteristics in evaluated tractor layouts of

Fig. 3-8

injure the operator [44, 74, 34], and limits the operator's confidence when operating the

tractor near its performance limits. In India, tractors account for over 25% of farming ac-

cidents and the upending of tractors is a common cause of serious injury 168]. This risk is

mitigated by the added stability of mounting the drawbar tool between the front and rear

axles.

3.3.2 Comparison of tractor layouts

Bulkey was designed by combining the strategies discussed in Sec. 3.3.1 resulting from

physics modeling with insights gathered from farmer interviews, while utilizing advantageous

characteristics of existing small tractor designs. Major needs of Indian small farmers are

unmet by existing designs, including the ability to enter narrow (<70 cm) inter-row spaces

like bullocks can, and achieving a purchase price comparable to bullocks (-100k INR) while

generating sufficient drawbar pull. A successful design should meet these needs and also

account for other important considerations farmers use when evaluating tractors, like soil

compaction and ease of operation. Additionally, the design must maintain desirable features

of existing tractors relative to bullocks, such as reduced ownership costs, reduced drudgery,

and improved farming productivity 152, 31, 76]. The analysis in this section shows that

Indian small farmer needs could be better met by a novel tractor layout - particularly with

respect to the location of drive wheels and the location of tillage tools.

Possible tractor layouts (Fig. 3-8) were selected for evaluation with respect to user needs

because they are either currently popular in India (layouts A and B), have been well adopted

in other countries by farms smaller than their national average (layout C) [44], or include
the features identified as desirable for the Bullkey design (layout D). These layouts have

distinct configurations: (A) is a conventional small farm tractor with side-by-side steering

idle wheels on the front axle, side-by-side drive wheels on the rear axle, and tools behind

58
the rear axle, (B) is a tricycle tractor similar to the conventional tractor layout but with a
single front idle wheel, and (C) has a design similar to a conventional tractor but with tools
ahead of the rear axle. The proposed Bullkey layout, (D), has inline drive wheels and tillage
tools between the front and rear drive wheels.

Steer Wheel Drive Wheel Idle Wheel

Drive & Steer Wheel Tool Operator

C D
iA B

Figure 3-8: Tractor layouts considered for Bullkey. A and B are typical small tractor layouts
in India. C is an alternative vintage design that was considered. D is the chosen Bulkey
layout.

Tool location impacts user comfort and safety, along with the tractor's drawbar pull
capability. Placing the tool behind the rear axle, as in layouts (A) and (B), improves comfort
by keeping soil detritus away from the driver during tillage and, more importantly, improves
drawbar pull by transferring weight to the driven rear axle during tillage. However, this
weight transfer is also detrimental to comfort and safety, as it unweights the front wheels,
resulting in loss of steering authority or, ultimately, in upending the tractor. Placing the
tool between the front and rear axles, as in layouts (C) and (D), improves comfort and safety
by placing the tool's action near the farmer's driving line of sight and eliminating the risk
of upending the tractor.
Layout (D) is singular in its ability to enter narrow spaces. Layouts (A), (B), and (C)
are limited by their side-by-side drive wheels, which prevent them from straddling crop
rows taller than their low ground clearance. In the case of (B), the situation is worsened
by the front wheel requiring a third travel lane - meaning the rows must be widened to
accommodate the full vehicle in a single inter-row space or the vehicle must straddle two
rows of crop. In these layouts, the major mass components - engine, transmission, and
operator - are between, not in line, with the drive wheels. As such, for a vehicle of this

59

m
configuration to straddle crops, a large amount of mass would have to be elevated above
the crop height. In the case of tall crops, this is deleterious to the vehicle's stability and
would limit its ability to use ground engaging tools. The inline drive wheels configuration

of Bullkey, layout (D), places all the major mass components in line with the drive wheels.
This naturally places the components in a narrow package, allowing access to inter-row lanes
and maintaining a low center of mass. Since the outrigger wheel does not need to generate
traction or provide steering, it does not need to bear much weight and can be attached

via a simple high ground clearance extension arm from the main tractor frame (Fig.3-7A).

This allows the outrigger arm to straddle tall crops and the tractor to generate a single
compaction lane (that of its drive wheels).

The side-by-side drive wheel configuration in layouts (A), (B) and (C) allows for differ-
ential steering, which can be an advantage in some situations. Differential steering is the
simultaneous application of different torques on each of two side-by-side drive wheels, which
generates a moment on the tractor body and causes it to rotate in yaw. Differential steering
can reduce the tractor's turning radius and enables the driver to maintain some control

even when the steering authority of the front wheel is low (e.g., when the front wheels are
unweighted). This could be replicated in (D) by a differential drive-line and steering system
that allows the rear tire to be completely braked (i.e. stopped) while the front wheel is
turned 900 and driven, therefore pivoting the whole vehicle around the rear tire's contact
patch.
The novel layout, (D), was selected for Bullkey because it combines the drawbar pull
advantages of weight transfer of (A) and (B) with the improved safety and comfort of (C).
Additionally, Bullkey has a unique ability to operate in narrow spaces. The advantages of
Bullkey, both in terms of drawbar pull and usability, are significant (Table 3.1) and allow it to
meet the needs of small farmers in India elucidated in Sec. 3.1. The inline drive wheels allow

Bullkey to enter narrow spaces currently only accessible to bullocks. The combination of
the wheel placement and a central tool location improves the vehicle's drawbar pull per unit
mass. Thus, the Bullkey design meets the required drawbar pull with a lower overall mass,
lowering the purchase price for the user relative to a conventional tractor that can produce
equivalent drawbar pull. Bullkey also meets the farmers' needs for improved comfort and
safety by providing improved visibility of the tillage tool and eliminating the risk of upending
the tractor during tillage. Additionally, soil compaction, which is detrimental to crop growth,

60
is reduced by limiting the vehicle to a single compaction lane. Bullkey (Fig. 3-7) is thus
uniquely capable of providing the benefits of both a pair of bullocks and a tractor.

3.3.3 Predicted performance

To demonstrate the relative performance advantages of Bullkey, its predicted drawbar force

(using the Sec. 3.2 model) was compared to that of an equal mass tractor and a pair of
bullocks. The mass of the modeled Bullkey and conventional tractor was set to 500 kg
because market trends of cost-to-mass ratio suggest that a tractor with a cost comparable to
a pair of bullocks would have a mass 500 kg or less (Fig. 3-2). Dimensions of the conventional
tractor, except for mass, are the same as on the Mahindra Yuvraj (Yuvraj NXT 215 by

Mahindra Tractors, India (601), a popular small tractor in India. The Bullkey dimensions
are those of the prototype vehicle described in detail in Sec. 3.4. Dimensions for both vehicles
are shown in Table 3.2.

Tractor Layout Conventional Bullkey

Vehicle Mass (kg) 500 500
Rider Mass (kg) 60 60
Weight Front/Rear(%) 45/55 50/50
Wheelbase (m) 1.5 1.3
Tool Horz. from CG (m) -1.5 0
Plow depth (m) 0.13 0.13
Tire sidewall height (m) 0.085 0.165
Tire diameter (m) 0.72 0.64
Tire width (m) 0.2 0.2
Tire Pressure (psi) 8.7 8.7

Table 3.2: Parameters for a conventional tractor and the Bulkey compared in Figure 3-9.

The pulling force of a pair of bullocks was calculated for comparison with these tractors.

The bullocks' pulling force has two values, a steady pull and a maximum pull. The steady,
sustained pull is about 15% of the animals' combined weight (each bullock has a mass of
-300 kg[5]), while the maximum pull can be as much as 50% of the animals' combined
weight [83, 106. The maximum pull plays a critical role - it allows the animals to briefly
pull a tillage tool through a harder patch of soil. A tractor that cannot reach an equivalent
maximum pull would become stuck in similar situations and require a decrease in drawbar

pull (by reducing tool depth) to proceed. Minimizing the unplanned depth adjustments
during operations improves the quality of the work and reduces drudgery. Therefore, it is

61
 valuable to have Bullkey match the maximum pulling force of bullocks to negate the need
for depth adjustments in any tillage situations where the bullocks could pull through.

Bullkey Conventional tractor

4500--
4000

3500 Bullocks maximum

Z3000k
2500

12000
/

~1500
1000
500
Bullocks steady state
$

2 5 10 15 20 25 30
slip (%)

Figure 3-9: A comparison of the drawbar pull versus slip performance in weak to strong
agricultural soil for a 500 kg hypothetical implementation of a conventional tractor and
Bullkey (more details Table 3.2). The drawbar pull of a bullock pair has been added for
reference. Soil data in Appendix B1.

The model predicted that Bullkey can exceed the maximum pull of bullocks over a
significantly wider range of soil conditions than conventional tractors can (Fig. 3-9). This
translates to improved usability of Bullkey over other light tractors by reducing the likelihood
of the vehicle being bogged down during tillage.

3.4 Proof-of-concept vehicle design

A prototype vehicle was built to validate the Bullkey concept and evaluate the model of
tractor traction performance that was used to design it (Fig. 3-10). The prototype incorpo-

rates key Bullkey design features, including: supporting nearly all the vehicle's mass on its
two inline drive wheels, incorporating a centrally mounted tillage tool, and incorporating a
lightly loaded outrigger wheel that can straddle rows of growing crop. The prototype was
built on a Rokon Scout motorcycle (Scout by Rokon International Inc., New Hampshire

[841), which is an all-wheel-drive, two-wheeled motorcycle meant for heavy off-road duty. A

62
Figure 3-10: Bullkey prototype vehicle highlighting the implementation of desirable design
features for a small tractor intended to replace a pair of bullock. These features include
two inline drive wheels supporting almost the full vehicle weight, a manually controlled and
centrally mounted heavy tillage tool, and motorcycle-type controls. Gym weights were used
for ballast at the front, rear, and over the outrigger wheel.

63
removable frame (next to the driver in Fig. 3-10) was attached to the left side of the motor-
cycle to control the tillage tool position and record the forces it experienced. An outrigger
arm extended parallel to the rear axle of the motorcycle, also on the left side (behind the
driver's left in Fig. 3-10). The outrigger wheel's axle was in the same vertical plane as the

rear drive wheel axle, making side-slip during slow speed turning negligible for the outrigger
wheel. The parameters of the test vehicle are given in Table 3.3.
The prototype mass could be varied between 192 kg and 305 kg during testing. This

mass range allowed testing of the tractor physics model at drawbar pull loads comparable
to bullocks but without overloading the stock frame and transmission of the Rokon. The
transmission began slipping at drawbar forces produced by the 305 kg tested configuration,
and so the prototype could not be tested at the maximum expected production mass of

500 kg. The Rokon, which weighs only 98 kg, had crucial benefits not present in other
heavier vehicles, including a unique inline drive wheel system and a frame designed for 20
cm wide tires. Building the Bullkey prototype with a commercially-available base vehicle
allowed its most critical features to be evaluated without the time and financial burden of
manufacturing an entirely new operator-safe vehicle. The prototype design incorporated the
full proposed novel layout and was thus suitable for evaluating the drawbar pull force and
overall functionality of the Bullkey concept. In combination with our prior work validating
the traction model with published data for heavier commercial tractors [44], the prototype
can be used to validate the physics model presented here, and so its predictions for other
mass configurations should be accurate.
The prototype was designed to evaluate if the lightweight Bullkey tractor could achieve
the predicted high drawbar pull force at tire slips recommended for plowing (15 to 25% [901)
and also have the ability to enter narrow inter-row spaces. Sensors were mounted to record

tillage tool forces (equal and opposite to the tractor's generated drawbar pull when parallel
to the tractor's pull), tillage tool force location, acceleration, and tire slip (Fig. 3-11). Tillage
tool forces and their location were isolated for measurement by the attachment structure
described in Fig. 3-12 and shown in Fig. 3-11A. The desired operating tillage tool depth
was controlled by a Haacon 1524 rack and pinion jack (1524 SS by Haacon, Germany [481)
and recorded by a string potentiometer (CWP-S by CALT, China 1151) that attached to
the pinion housing and the rack. The horizontal and vertical components of the tool force
could be resolved from the three axial load cells (104-500 by DYLY, China [191) that exactly

64
 Bullkey proof-of-physics prototype
Base Vehicle ROKON Scout 184]
Mass unballasted 192 kg
Mass supported by front wheel 82.5 kg
Mass supported by rear wheel 94 kg
Mass supported by outrigger 15 kg
Wheelbase 1.3 m
Rear ballast to rear axle 0.56 m
Front ballast to front axle 0.48 m
Turn radius (no lean) 1.4 m
Tire pressure 7 psi
Tire model TITAN 489XT [951
Tire size 12" rim, 8" x 25"
Tool used 0.3 m wide furrower

Table 3.3: Basic parameters for Bullkey prototype.

Figure 3-11: Close-up views of sensor installation examples. Views of prototype shown are
(A) left-side, (B) outrigger wheel right side, and (C) rear axle left side

65
 Gnd. 1 Bearing I (Lin.) Bearing 2 (Lin.) - Parallel rail behind it

Gnd. 0 Load Cell 1 Rack & Pinion

Bearing 2 (Rot.)

o Load Cell 1
0 0 Bearing 3 (Lin.)
Load Cell 3
d Gnd. 0
Bearing 2l Id
Gnd. 2 Rack & Pinion
(Rot.)
Bearing I
Load Cell 2 Slider (Lin.)

.
Load Cell 3 Slider
Soil F rce X
Plow
Load Cell 2
Figure 3-12: Mechanical diagram and CAD of the structure used to measure tillage forces
on the Bullkey proof-of-physics prototype. Gnd. stands for ground (i.e. fixed to Bullkey's
frame), Lin. for linear, and Rot. for rotary.

constrained the motion of the tillage tool (Fig. 3-12). To exactly constrain each load cell,
they were mounted in conjunction with a single degree-of-freedom linear (HSR15-600-A

by Joomen, China [53]) or rotary (513267 Wheel Bearing and Hub by MOOG, USA1671)
bearing and placed to be the only load bearing elements in the force path of the loads they
measured. These three load cells also allowed spatial resolution of the center of pressure for
forces exerted on the tillage tool (along the x dimension in Fig. 3-12).
All three wheels were fitted with magnetic proximity sensors, with 10 evenly spaced
magnets placed on each wheel (Fig. 3-11B). Tire slip can be calculated using the rotation of
these senors, assuming that the idle outrigger wheel has near zero slip and can be used as the
reference point for distance travelled. An accelerometer (ADXL335 by Adafruit, USA121)
was placed at the rear axle to provide higher time resolution on vehicle speed and to assist
in confirming short-term measurements from the outrigger wheel rotations (Fig. 3-11C).
Appendix B4 lists the sensors used for measurement and their characteristics.

3.5 Field testing and performance results

3.5.1 Field testing methods

Field tests were performed on a working farm in Massachusetts to validate the terramechan-
ics model, investigate the traction performance of the Bullkey prototype, and obtain user

66
feedback after operating the vehicle among actual growing crops. Traction performance tests
were conducted at different tillage depths and different ballasting levels to alter the vehicle

mass distribution from the base distribution specified in Sec. 3.4. Ballast on the front and
rear ballast trays varied between 0 and 56 kg ±0.5 kg, operator mass (of the author) during
recorded tests was 79 kg 1 kg, and tire pressure was set to 41 kPa (6 PSI) ±4 kPa. For
each configuration, the tractor was driven in a straight line at about 1I m/s for 30 to 50 in
with a 30 cm wide furrowing tool at a constant depth between 12 and 19 cm. Tool depth was
set for each configuration to force tire slip to be near 20% - approximately the upper limit
of what would be useful on a farm field and thus close to the vehicle's maximum practical
drawbar pull {115][13].
In addition to sensor data, field tests with the prototype vehicle provided an opportunity
to gain valuable feedback on the usability of the proposed tractor design. Six local Mas-
sachusetts farmers observed the field tests and provided their feedback in a spoken survey.
The survey was approved by MIT's Committe on Use of Humans as Experimental Subjects
(COUHES). In addition to the drawbar pull tests on open farm fields, qualitative tests were
performed by driving Bullkey between growing crop using a 15 cm wide sweep tillage point
at 3 to 6 cm depth. A sweep is a thin "V" shaped tillage tool used cut weeds at their root

between rows of growing crop during intercultivation.

The sensor data collected was processed to have a similar format as the initial simulations
(Fig. 3-9). First, the collected time-force signals were passed through a 1 Hz low pass filter.
This filtering frequency was selected because the 30 cm long tool travels at least three

characteristic lengths every second. Then, the distance traveled by all wheels was calculated
by summing the new distance traveled each time a wheel magnet (Fig. 3-11B) was detected,
using linear interpolation to fill in the distance travelled for intervals between detections.
The distance travelled between magnet detections is 27r/10 * R, where 27r/10 is the angular
spacing between neighboring magnets in radians (there are 10 magnets per wheel) and R is
the effective radius of the wheel (estimated by counting the number of wheel rotations to

travel 30 m under the test conditions). The three (one per wheel) distance-travelled vectors
were then processed through a 1 Hz low pass filter as well.

The drawbar pull versus tire slip binned data shown in Fig. 3-13 were generated by
the following procedure: (1) The highest drive tire slip was selected at each timestamp
and stored along with the drawbar pull measured at that timestamp to generate a slip vs.

67
drawbar matrix. (2) This matrix was then rearranged so that all drawbar pull instances were
assigned to the closest integer slip (i.e. all slip instances >13.5% and <14.5% were assigned
to the 14% slip bin). (3) Finally, in each slip bin the average, minimum, and maximum

drawbar pull were obtained and stored. As presented in Fig. 3-13, squares represent the
average drawbar pull at that tire slip bin while the error bars represent that maximum
and minimum drawbar pull recorded at that tire slip bin. Further details are presented in
Appendix B5.

3.5.2 Field performance results

Figure 3-13 compares the drawbar pull performance for each of the Bullkey mass config-
urations tested against the steady state and the maximum pulling force of a bullock pair

[83, 106, 42], as well as to the model-predicted performance for the soil conditions during
the test and for the range of common farm soil conditions. The results validated that the

physics model from Sec. 3.2 made predictions for the maximum drawbar pull that are suf-
ficiently accurate to inform tractor design. The model average absolute error compared to

experimental data was 7% at 15% slip, 9% at 20% slip, and 12% at 25% slip. The standard
deviation for the absolute error was 4% at 15% slip, 5% at 20% slip, and 8% at 25% slip.
The drawbar versus slip results are presented in more detail in Appendix B2. All tested
configurations comfortably surpassed the steady-state pulling of bullocks. The maximum
drawbar pull for the 305 kg Bullkey configuration, despite being limited by the test soil not
being at the upper limit of strength for agricultural soils, was near to the maximum pulling
force for a pair bullocks. More importantly, given the demonstrated accuracy of the model,
it is expected that a heavier Bullkey (up to 500 kg) would be able to match or exceed the
maximum pulling force of a bullock pair for any common agricultural soil condition, as was

predicted in Fig. 3-9. This cannot be matched by a conventional tractor layout of the same
mass. The properties for the soils used for the model are provided in Appendix B4.

3.5.3 User feedback

On-site farmers who observed the Bullkey prototype during field tests said that the vehicle
had valuable and unique benefits for small farmers. Farmers appreciated the ease with
which the tool could be observed during tillage and the tall height of the outrigger arm,
which allowed the vehicle to easily straddle crop rows. They also commented positively on

68
 Bullock pulling force instantaneous maximum & steady-state avg.
[i] Model predicted range of pulling forces for agricultural soils
- Model predicted pulling force for soil conditions of experiments
-+-Average and limits of pulling force from experiments
3000
2500
2000
1500
1000!,
5005
Front: 95 kg, Rear: 96 kg Front: 136 kg, Rear: 137 kg

3000
X_2500
1 2000-
1500

500
0 50 Front: 95 kg, Rear: 137 kg Front: 136 kg, Rear: 96 kg
0
3000
t2500
12000
1500
.01000 Frnt- - - - - --
500 Front: 136 kg, Rear: 153 kg Front: 152 kg, Rear: 153 kg
0
5 10 15 20 25 305 10 15 20 25 30
Tire Slip (%) Tire Slip (%)

Figure 3-13: Examples of measured forces for prototype configurations tested. Indicated in
each plot are the masses supported by the Bullkey prototype's front and rear wheels when
static and with no driver on board. The masses were adjusted by adding and removing
ballast.

69
the Bullkey prototype's ability to plough deeper than they would have expected from such
a small vehicle. Farmers were initially concerned that the tillage tool's lateral offset from
the drive tires might cause Bullkey to veer off-track - this concern was allayed when they

watched Bullkey maneuver and saw that it was easy to drive the vehicle in a straight line
under all conditions. The farmers also had some suggestions for improving the vehicle. They
suggested having a mount for low drawbar force tools, like those used during intercultivation,

set up behind the rear axle and in line with the drive wheels in order to provide better access

to narrow rows while crops are growing.

3.6 Discussion

The Bullkey prototype's measured maximum drawbar pull matched well to model predic-
tions in both trends and absolute values. This showed that the model is a useful tool to

evaluate potential tractor designs for traction performance and identify promising design di-
rections. The average absolute error of the model at high slips (when maximum drawbar pull
occurs) was generally less than 10%. The model's performance and its parametric, physics-

based foundation make it useful for exploring a large design space of previously unrealized
tractor designs. These capabilities of the model made it a powerful tool for identifying and
establishing the Bullkey design.
The prototype's field performance showed that the Bullkey design satisfied the outlined
user needs for an easy to use, highly maneuverable lightweight vehicle with high drawbar
pull capability for a low mass device. Bullkey was able to straddle rows of growing crops
on the field because of its configuration of inline drive wheels with an outrigger arm. This
enabled Bullkey to operate in narrow inter-row spaces like bullocks do - something that

is not possible with conventional tractors. Bullkey generated more drawbar pull per unit
mass than conventional tractors with rear drive wheels and rear mounted tools - this is
significant because mass is correlated approximately linearly with cost (Fig. 3-2). Bullkey's
performance on the field therefore suggests that a production-version of the tractor could be
sold at a lower cost for a given drawbar pull capacity than available tractors, enabling the
distribution of a tractor that can compete with the maximum pulling force and the purchase
price of bullocks.
Bullkey was comfortable to operate after some adjustments were made. A 20 kg ballast

70
was added to the outrigger wheel after early field tests and mitigated the risk of the tractor
rolling over sideways. During instrumented testing the operator would sit side saddle

-
a remnant habit from an earlier, taller version of the tillage tool attachment mechanism

shown in Fig 3-12 - which shifted the overall center of mass away from the outrigger wheel
and decreased stability. During later driving the operator sat as is conventional, straddling
the motorcycle frame and the tillage tool mounting frame, which was an improvement in
comfort. The front drive wheel never became unweighted during tests with heavy drawbar
loads (a common occurrence with conventional tractors [1, 26, 891), which enables Bullkey
to operate near its traction limits more safely than conventional tractors.
The proof-of-physics Bullkey prototype allowed testing to find its drawbar pull at slips
relevant to tillage and near its traction limits (15 to 25% tire slip). A limitation of the
presented work is that the accuracy of the modeled drawbar pull drops for slips under
10% (Fig. 3-13). It is possible that at lower slip the assumed soil deformation mechanics
are less applicable, or that the vehicle was at least partially relying on other methods of
forward propulsion during low slips (like its inertia when slowing down). These errors could
have been accentuated by the experimental methods, which focused on finding the tractor's
maximum drawbar pull at high tire slips and not on generating steady drawbar pull at
low tire slips. Future work could include experiments at constant low tire slips, to better
capture the performance of the model in those conditions and identify strategies for model
improvements. The model is usable for its design purposes in this Chapter, which is to
estimate the maximum drawbar pull of multiple designs, which occurs at high tire slips.
The Bullkey prototype was usable for farming operations that could be performed with
tillage tools mounted centrally on the vehicle (like plowing and furrowing). To add flexibility,
a future prototype could allow low drawbar force farming tools via conventional mounting
points behind the tractor, like a three point hitch and a pin or ball hitch. Future work could

also allow for ballast to be added without extending the overall length of the vehicle. A key
next step is to discuss the Bullkey vehicle with small Indian farmers - the target users
-

to solicit feedback on the vehicle design and usability. In these discussions, farmers could
also be asked if they might use Bullkey (with some attachments removed) as a conventional
two-wheel motorcycle for personal transportation. If Bullkey is viable as a two-wheeler, it
could replace both a pair of bullocks and a motorcycle for farmers, further increasing its
value proposition.

71
3.7 Conclusions

The presented tractor design, Bullkey, is novel in its high potential to concurrently match

bullocks' sales price, pulling strength, and unique ability to access a field with growing crop,

while also offering farmers major conventional tractor benefits like increased productivity,
lower maintenance costs, and improved comfort. This allows Bullkey to fulfill the unique
needs of small farmers in India, which are not currently being met by commercially available
tractors.
Bullkey has inline drive-wheels that support the majority of the its mass, a crop clearance

similar to a bullock team yoke, and a centrally located tillage tool attachment. Inline drive
enabled improved traction, reduced soil compaction, and operating in narrow inter-row

spaces between growing crop. Central tool attachment increased traction and improved
safety while also facilitating the operator maintaining control over the direction they are

driving as well as the tillage operation being performed. These beneficial design features were
identified by combining insights from a physics-based traction model and farmer interviews.
The traction performance predicted by the model was validated by building an instrumented
prototype of Bullkey and field testing it.
Replacing bullocks with a suitable farm tractor, such as the Bulkey design proposed
here, could increase farmer income by 20% and reduce their recurring expenses by 60%.
Farmer income could increase because of higher crop yields from more precise and timely
farm operations. Recurring expenses would be reduced because tractor maintenance is much
lower over the course of a year than the daily feed and care costs of bullocks. The findings

presented in this Chapter will be useful to engineers developing lightweight, high drawbar
pull vehicles and/or vehicles that are well suited to in-field use by small farmers in emerging
markets.

72
Chapter 4

Investigation of viability to replace

draft animals with all-wheel-drive
motorcycles on small Indian farms

4.1 Introduction

This Chapter describes the motivation, design, and validation for a farm tractor prototype
specialized to small farmers in low resource settings, particularly in India. In these settings,
farmers' prosperity is currently stymied by the limitations of draft animals. Draft animals
are inefficient and expensive to maintain compared to tractors (Fig. 4-1) [52, 39], but con-
ventional tractors cannot replace animal's small dimensions and low capital cost [64, 761.
This misalignment between conventional tractors and the needs of small farmers is, in part,
because conventional tractors were designed for larger fields than what is typical around the
world [44]. The majority of farms in the world (84%), and particularly in India (86%), are

less than 2 ha in size158, 75], whereas the conventional tractor largely evolved for farms in
the US that are at least 30 times larger [14, 99, 59].
A tractor layout specialized to the contemporary needs of small farms, called Bullkey,
has been previously introduced by the author [45] (fully described in Ch. 3). Bullkey is
a portmanteau of bullock and key - indicating its goal of being the key to unlocking the
bullock market to mechanization. In prior work work 145], Bullkey was shown to generate

more drawbar pull per unit mass than conventional tractors. This is important because mass

73
 12
MInitial Purchase Cost
10
a Other Costs for first 15 years
8

.$6

0
Bullock Tractor Tractor Proposed
Pair Financed Bought Vehicle
Upfront

Figure 4-1: Ownership costs over 15 years for a bullock pair, a small tractor, and the Bullkey
tractor. In addition to purchase price, other costs include fuel/food, and maintenance. An
ideal solution would have the low purchase price of bullocks and the low upkeep cost of
tractors - shown as Bullkey tractor. Financing or renting may be inaccessible to many
farmers. Renters forgo using the tractor for supplemental income work and potentially
timeliness of completing operations. 181, 80, 22, 31, 30, 47]. A breakdown of the costs
shown here is provided in Appendix C5 Tables 6.10 and 6.11.

is correlated to both tractor price [45] and drawbar pull [115, 13, 4]. A conventional tractor
that is light enough to be sold at a price competitive to bullocks would be too light to match

the maximum pulling force of the bullocks. Builkey overcomes that with a three-wheeled

layout that supports nearly all of the vehicle's mass on inline drive wheels and shifts the

tillage tool ahead of the rear axle. This layout increases the load on all drive tires during

plowing and allows the trailing drive tire to roll on already-compacted terrain - improving

drawbar pull while also preventing the vehicle from rolling over backward.

In Chapter 2, Bullkey was field tested to validate its traction performance and to pre-

liminary assess its usability with farmers' feedback. This Chapter expands on that work by

demonstrating the design's viability for performing specific and comprehensive agricultural

operations key to Indian farmers. Section 4.2 presents the needs of small Indian farmers

elucidated during field interviews and from background research that drove the design of

Bullkey. Section 4.3 describes Bullkey's overall design and how it was engineered to achieve

the operations required by Indian small farmers. Section 4.4 describes field tests in Mas-

74
sachusetts conducted to validate Bullkey's performance on these operations. This section
also presents feedback collected with small and marginal farmers in India on their views on
the tested tractor regarding its viability, the likelihood they would use Bullkey for various
farming operations, the overall design, and the price point at which they would purchase the
vehicle. The collected feedback suggests that Bullkey is useful and attractive to the targeted
user population.

4.2 Description of Farmer Needs

A) Mahindra ResearchValley.
B) Small and large farmers
C) JaitsarTractor Development Center
D) Farmers, partners, manufacturers
E) Vigyan Ashram, small farmers
F) Mahindra test site, farmers
G) Central Inst of Ag. Eng.
H) Central Mech. Eng.. Research Inst.
1) Agricultural Engineering, lIT
J) Small Tractor Manufacturers
K) Mahindra Farm Marketing
L) Desphande Foundation, farmers
M) Small farmers, tractor dealers

Figure 4-2: Locations visited during the development of the Bullkey tractor prototype. The
final interviews assessing the developed concept were performed at locations with white
background, and were not conducted with the same farmers as the initial user needs assess-
ment.

BuIlkey was conceptualized based on interviews on local agricultural practices and the
suitability of existing tools with stakeholders of Indian small farming at 12 locations in India,
including the states of Maharashtra, Tamil Nadu, Gujarat, Rajasthan, Madhya Pradesh,
West Bengal (Fig. 4-2). Farmer interviews were approved by MIT's Committee on the

Use of Humans as Experimental Subjects (COUHES). Stakeholders interviewed included

farmers, research organizations, governments, manufacturers, and farm tractor dealers. A
key insight from these visits was that farmers used bullocks not only because of their low
capital cost (about a third that of tractors) but because they have functional advantages
over conventional small tractors. Bullocks have a smaller width than tractors and are more
maneuverable. This allows the bullocks to walk between rows of growing crops further into
the season, require less space to turn at row ends, and better traverse unfinished dirt paths

75
 Major design requirements from user needs Reference Alternative (in gray)
Aspect Bullkey Bullock Pair Small Tractor
Purchase cost (INR) ~100000 80000 265000
Ownership cost (INR/year) <12500/Ha 93000[22] 12500/Ha[47]
Overall Width (m) 1.7 2.1 1.7
Required Path Width (m) <0.7 0.6 to 0.9 1.7
Headland needed (m) <1.5 1.5 2.6
Max. Drawbar Pull (N) >2800 2800 [421 4600
Daily work hours unlimited 5 unlimited
Top road speed (km/h) 26 4 26

Table 4.1: Major user needs identified via interviews with stakeholders in India and back-
ground research. Unless otherwise noted, bullock values are from the author's farmer in-
terviews and tractor values are from the Mahindra Yuvraj NXT 215 (Mahindra Tractors,
India 160]) - a market leader in the small tractor segment. Actual Bullkey price will be
affected by distribution possibilities.

leading to farm fields.

Also elucidated during the visits were key farmer-required field operations that are typi-
cally completed by a small tractor or bullocks. Early in the season and prior to planting, seed
bed preparation is completed via plowing, disc harrows, and/or use of a rotavator. Next,
planting is executed by precision seed drills that can position seeds at consistent depths
and spacings both laterally and longitudinally. While the crop is growing, intercultivation
is performend by mechanically removing weeds between rows of crops, often done with "S"
or "C" shaped cultivator tines. This tillage-based deweeding is typically supplemented by
manual laborers who pick weeds between crops within a row. Concurrently, spraying of
fertilizer, herbicide, and other liquid-based inputs may be performed. When crops are tall
(above 30 cm), this is often done by manual laborers carrying backpack-based sprayers since
bullocks are not compatible with most sprayers and tractors are much wider than crop row
spacing but do not have sufficient ground clearance to straddle the crop. Throughout the
season, moving inputs and outputs between farms and towns by haulage trailers is an im-
portant usage case for both bullocks and tractors. The identified operations of interest are
in agreement with prior work [761 [38][311.
From the aforementioned key targeted agricultural operations, and using information
from our project partner's field research, design specifications for Bullkey were established
and are provided in Table 4.1. Bullkey should be comparable to bullocks in purchase cost
(Fig. 4-1), width, drawbar pull, headland required, and have ability to traverse unfinished

76
dirt paths. Bullkey should be comparable to conventional small tractors in ownership cost
(Fig. 4-1), user comfort, daily work hours, and road speed. Finally, it is desirable that
Bullkey have a familiar interface to users, since a current barrier to tractor adoption is
the training required to operate them [7611381. Critically, farmers must also be willing to
purchase Bullkey at a price that allows those in its supply chain to earn a profit.
A design that can succeed in the Indian small farmer markets should be able to demon-
strate in field testing - and intuitively convince holder farmers of - its ability to perform
plowing, disc harrowing, rotavator tillage, seed drill planting, deweeding with a cultivator,
and pulling a trailer. In the next sections, we present the design of a Bullkey prototype and
demonstrate its ability to complete these operations.

4.3 Prototype Vehicle Design

Three-point hitch All drive controls on handlebar 7

& ball tow hitch

V Outrigger wheel
tool
lN~t rn~k Central tillage

Figure 4-3: Overview of the Bullkey design. The CAD drawing on top left highlights key
features of proposed tractor design. This is supplemented by three pictures of the physical
prototype where those features are clearly visible.

77
 Bullkey dimensions with outrigger arm
Base Vehicle ROKON Scout
Mass 125 kg
Mass supported by front wheel 60 kg
Mass supported by rear wheel 65 kg
Wheelbase 1.3 m
Turn radius (no lean) 1.4 m
Overall width 1.2 m
Path Width 0.6 m
Outrigger arm height 1 m
Drive Tire pressure 6 psi
Drive tire size 12" rim, 8" x 25"
Outrigger Tire pressure 20 psi
Outrigger tire size 3" wide, 26" diam.

Table 4.2: Key Bullkey dimensions.

To validate the performance of the Bullkey concept tractor on field operations that are
important to Indian farmers, a prototype vehicle capable of the aforementioned functions
was built.

4.3.1 Prototype vehicle

The BuIlkey prototype was built by modifying a ROKON Scout utility motorcycle (ROKON,
New Hampshire [841) with factory installed all-wheel-drive. Major modifications included
the addition of a centrally-located tillage tool attachment, an outrigger arm, a rear mounted
three-point hitch compatible with conventional small tractor tools, a ball hitch for towing,
and interfaces for ballasting. The prototype's key features are highlighted in Fig. 4-3 and
its main dimensions are listed in Table 4.2.
This design can generate a high tillage drawbar pull for two main reasons: a high pro-
portion of its total mass is supported by the front and rear drive tires, and tillage forces
from the centrally mounted tool increase the vertical load on both drive tires [451. The latter
not only further improves traction but also allows safe operation near the vehicle's traction
limits. By contrast, conventional tractors with a rear mounted tool can rollover backward
during heavy tillage due to the tillage forces unweighting the front tires - a dangerous and
common situation 111126][89].
Two tested options for stabilizing the motorcycle are presented: Bullkey's default rigid
outrigger arm with a third wheel to the operator's left and aligned with the rear drive wheel

78
(Fig. 4-3), and a specially-created, human-powered stabilization design. The outrigger arm
has high ground clearance and adjustable width, allowing it to straddle rows of crop. Its
alignment with the rear wheel keeps it from side-slipping during turns. The human powered
stabilization design allows Bullkey to retain an overall width comparable to a single bullock,
therefore avoiding the need to straddle crop. It is described in Section 4.3.2.
The Bullkey prototype described here meets the farmer needs outlined in the previous
section. Its overall dimensions allow it to operate in a spaces similar to bullocks - something
not possible with conventional tractors due to their large width and low axle height that
prohibits straddling crops over 0.3 m tall. Bullkey's inline drive wheels and central tillage
tool location allow it to generate more drawbar pull per unit mass - enabling it to theo-
retically be sold at a lower price than a conventional tractor of comparable traction since
mass and sales price are correlated [451. Finally, its tool attachment points make it safe
to operate and compatible with the tools needed to complete key tasks mandated by small
Indian farmers.

4.3.2 Balance Board

crop grows to
0.4 to 0.6 m widthcrophei
reaches 1.2 to 2 m

row spacing when planted,

tpically ~0.9 m

Figure 4-4: Bullkey with balance board between crop rows. The balance board allows
Bullkey to operate between rows of tall growing crop by keeping all of its ground contact
points in a single lane narrower than the distance between adjacent crops.

79
 To enable Bullkey to operate in the same overall space as a single bullock, a human-
powered roll stabilization attachment for utilitarian two-wheeled vehicles was created [57].
This device is called a balance board and is narrower than the motorcycle's handlebars

-
keeping all ground contact points within a single lane under 0.52 m wide and maintaining
the stock motorcycle overall width (Fig. 4-4)). The balance board allows the motorcycle to
operate late into the season between tall rows of growing crop. It also allows the motorcycle
to lean relative to the ground, which is beneficial for comfortable operation in side slopes or
when turning at higher speeds (Fig. 4-5).

Ball hitch joint from Bullkey to rigid balance board

Adjusting to side slope

Figure 4-5: The balance is a rigid, wheeled platform connected to the motorcycle frame
by a ball hitch. The user steps on the balance board to stabilize themselves but can still
comfortably stay upright on side slopes or lean in turns.

The balance board provides a rolling platform under the motorcycle for the operator
to place their feet on. It is attached via a ball hitch underneath the motorcycle frame and

80
directly behind the front wheel. Nominally, the rotation of the balance board is independent
of the motorcycle rotation for a large range of motion. When driving at slow speeds the
motorcycle is unstable in the roll direction and may start to tip sideways. When side roll
initiates, the user can press down on the balance board (which remains parallel to the

ground) with the leg on the side towards which the motorcycle is tipping. This, in practice,
has a very similar stabilization effect as pressing against the ground (as one would do without
the balance board) but has two major advantages: (1) the reach to the balance board is
much shorter than to the ground, allowing the driver to maintain a natural riding position;

and (2) since the balance board is moving forward with the motorcycle, the rider is pressing
down on a surface that is largely static relative to them (as opposed to dragging a foot on
the ground or tip-toeing on the ground).
The basic operation principle of the balance board is converting an internal force (the
user's feet on the motorcycle foot pegs) to an external force (user's feet on the not fully
constrained balance board). This stabilizes the motorcycle in roll. By contrast, with nor-
mal foot pegs, the leg forces would be redistributed internally between the foot pegs and
motorcycle frame. The balance board is also enabled by having all of the Bullkey controls
be hand actuated, thereby fully liberating the legs and feet for other tasks.
Figure 4-6 demonstrates the basic operating principle of the balance board. In the top
row the user leg force flow paths are shown going through the balance board and to the
ground. This stabilizes the motorcycle in roll since the balance board can rotate relative to
the motorcycle body and its contact points are laterally offset from the vehicle centerline. By
contrast, with normal foot pegs (bottom image) the force goes through the rigid motorcycle
frame and to the vehicle centerline - thereby applying no roll stabilization.
Compared to the outrigger arm, the balance board has the advantage of being agnostic
to crop height since it keeps the overall vehicle identical to the stock motorcycle and all
ground contact points in a single lane narrower than the motorcycle handlebar - allowing it
to roll freely between rows of growing crop. However, the balance board also increases the
width of the contact footprint of the vehicle within a row, since its wheels are not inline with
the drive wheels. By contrast, the outrigger wheel rolls in the center of the row neighboring
the motorcycle's path, but its outrigger arm must be tall and wide enough to straddle the
growing crop between the drive wheels and the outrigger wheel.

81
 roi

roll pin joint ground reaction

internal force

Figure 4-6: FORCE FLOW PATHS WHEN USING BALANCE BOARD (TOP IMAGE)
AND IN CONVENTIONAL MOTORCYCLE (BOTTOM IMAGE). White arrows
repre-
sent the force applied by the user legs on and opposed by the black arrow reaction forces.

82
4.3.3 Implement Utilization

ounts to side of motorcycle 3 point hitch mounts

Control levers
Gang adjustment

Rack and pinion Hitch tongue

Spinning blades
Notched disc gangs
Tillage tool Rotary bearings Support wheels

Seed hoppers 3 point hitch moun ts 3 point hitch mounts

Nozzles
Adjustable height
Chain drive
Seed tu be Precis( application Pump

Tank
Hitch tongue Support wheels
Support wheels
Distance tracking wheels Compliant tines

Figure 4-7: Close up views and key components of the farming implements used during field
testing.

The following is a list of implements that were tested with the Bullkey prototype. These

implements were selected because they perform the operations needed by Indian small farm-
ers from, as outlined in Section 4.2. These tools perform seed bed preparation, planting,
intercultivation, and spraying to a degree equal or better than bullocks. Close-ups of the
implements are shown in Fig. 4-7 and tool dimensions are provided in Appendix C1.
Plow: unearths soil from 10 to 20 cm depth to loosen and dry it. In the prototype, this

tool is mounted between the front and rear axles. The plow used was 20 cm wide with depth
controlled manually via a Haacon 1540 jack (Haacon, Germany [481) located adjacent to the
driver. This tool is expected to be used exclusively with the prototype in the outrigger arm
configuration.
Rotavator: breaks up large soil clumps near the surface. In the prototype, the rotary
tiller is attached behind the rear axle with a single vertical pin hitch connection. Via control
levers extending forward from the tiller, the transmission engagement and tillage depth can
be adjusted. The rotary tiller is a model Field Tuff (Field Tuff, Illinois [721), with 0.92 m
tillage width, 0.3 m diameter blades, and powered by a 208cc Briggs & Stratton (Briggs
&

8:3
Stratton, Wisconsin {121) 9.5 hp engine. This tool can be used with the prototype stabilized
by either the outrigger arm or the balance board.
Disc harrows: improve top soil texture for planting. In the prototype, the disc harrow
(Kolpin Outdoors, Minnesota [551) consists of eight 0.3 m diameter discs with a total en-
gagement width of 1.37 m. The disc harrow is mounted at the back of the vehicle to the
three-point hitch. The hitch can fully raise the tool at row ends or for transportation, and
then lower it for engagement during tillage. This tool is expected to be used exclusively
with the prototype in the outrigger arm configuration.

Planter: lays down seed at controlled depth and spacing. In the prototype, the planter
(Field Tuff, Illinois [711) is mounted behind the rear axle to the three-point hitch. The
planter was set up with two planting units spaced 0.9 m apart and planted black beans and

soy beans at a 3cm depth. At row ends the planter can be fully raised (without its wheel
touching the ground). During planting, engagement is controlled via the three-point hitch

and the tool is partially carried by its wheels. When transporting the planter, it can be
fully carried by Bullkey or partially supported by the tractor's wheels. When planting, it is
recommended the planter be stabilized using the outrigger arm Bullkey configuration.
Cultivator: mechanically removes weeds between rows of growing crops. In the proto-
type, the cultivator (Black Boar, North Carolina 1111) is attached to the three-point hitch
behind the rear axle. The cultivator has multiple S-shaped spring tines that engage the soil
between rows of growing crop at a 2 to 5 cm depth - tearing up weeds. When cultivating,
the motorcycle can be stabilized by the outrigger arm or the balance board. When crops
are large, it is recommended to use the balance board with two S-tines to allow access to
narrow spaces that would typically only be accessible by bullocks.
Sprayer: pulling or carrying a tank to supply liquids (often fertilizers or pesticides) to
plants at their leaves or roots. The sprayer (VEVOR Machinery, China [1031) used has a 60
L tank, a 9 bar pump, and six nozzles evenly spaced on a boom spanning a 1.5 m width.
The boom can be set to heights between 0.4 in and 2 in inincrements of 0.1 in. When the
sprayer tank is carried by the motorcycle, it is recommended the vehicle be stabilized by
the outrigger arm. If the sprayer tank is towed on a trailer, the balance board may be used
for stabilization - allowing access to narrow spaces between rows of tall crops and spraying
them with a tall boom height.
Trailer: used to transport farm inputs and outputs as well as for supplemental income.

84
The trailer (locally made in workshop) used in testing had a mass of 225 kg and a track
width of 1.4 m. The overall length of the tractor and trailer is 3.1 m. The trailer is mounted
to the prototype at a ball hitch located 0.3 m above the ground behind the rear axle.

4.4 Methods - Field Validation of Operations and Assessment

Interviews

Elii
--*W~o~i ~ H

F4 **

Figure 4-8: Images of the researchers performing key operations of interest to Indian farmers
with Bullkey. A) Plow B) Disc harrows C) Cultivator D) Rotavator E) Planter F) Trailer
G) Sprayer on its trailer H) Sprayer on motorcycle.

4.4.1 Methods - Field Validation in Massachusetts

Field tests of the Bullkey prototype were performed at a small farm in Carlisle, MA with
silt loam and sand loam soil. The farm sells locally grown organic produce in the area,
including berries, eggplant, leafy greens, vegetables, peppers, and cantaloupe. The goals

of the field tests in Massachusetts were threefold: to validate the feasibility of the concept
vehicle for farming, to capture media to show farmers in India during interviews, and to

receive hands-on feedback from two Peruvian small farmers temporarily working at the
farm. The prototype's tillage drawbar ability was earlier tested in detail as seen in [45].

85
 Bullkey's ability to perform the previously described farming operations needed by small
Indian farmers was assessed by using the vehicle for plowing, disc harrowing, rotavating,
planting, intercultivation, spraying, and trailer pulling operations per the specifications of

{88], a popular Indian handbook on farm machinery. The operations tested represent the key
needs of small Indian farmers with best practice settings. The specifications were selected
to ensure that the operations were representative of how Indian farmers would use Bullkey,
so its success in performing the operations as specified would demonstrate its suitability for
meeting the needs outlined by farmers. The test dimensions are summarized in Appendix
C1 and are also described here.
Primary tillage consisted of plowing with 20 and 30 cm wide furrowing bottoms (one

at a time) at depths between 10 and 20 cm. These operating dimensions were enough to
generate drawbar forces in excess of 70% of the vehicle's weight (beyond what would be
expected of a conventional tractor 111511131) and generally operate the tires at 10 to 20%
slip. These tests are discussed in detail in [45].
Secondary tillage included disc harrowing and rotavating to break up soil clumps and
prepare seed beds. Disc harrowing was performed by two sets of four 30 cm diameter notched
discs. The two sets were angled at 120 degrees relative to each other and covering a 1.37 cm
width, forming a forward pointing "V" centered along the motorcycle's drive line. The discs
were operated at 10 cm depth to till rows of approximately 50m length before being lifted
by the three-point hitch at row ends to allow for no engagement during headland turning.
The rotavator consisted of 24 blades 30 cm in length, organized in sets of four to form six
rotating crosses (see Fig. 4-7). The machine was set to cut at 12 cm depth and was towed
through 50 m long field rows. Motor-to-blade and blade-to-soil engagement were controlled
by separate, manually actuated levers.
Post seed-bed preparation operations included planting, and simulated crop care (inter-
cultivation) by an s-tine cultivator and a spraying unit. The planter utilized two seed drills
spaced to plant rows 90 ci apart. One was filled with soy bean seeds and one with black
beans. Seeds were in inserted into soil every 10 cm at 3 cm depth. Insertion depth was
controlled by Bullkey's three-point hitch mechanism. The cultivator utilized two S-shaped
tines at 2 to 6 cm depth for 50 m rows. The tines were 2.5 cm wide and set at 25 cm apart.
Towing the 60 L sprayer tank was done with the outrigger arm and balance board. A 1.68
m wide sprayer boom with four nozzles was set at 40, 160, and 210 cm heights - representing

86
heights valuable for vegetables crops as well as tall crops in the late stages of their growth.
Finally, a trailer was towed with the Bullkey prototype as a test, and to haul equipment
to facilitate other tests. The trailer measures 1.4 m long and gave the vehicle an overall
length of 5.4 m. Its base mass was 225 kg and it was loaded with at least 200 kg of equipment,
then driven at up to 20 km/h. It was attached to the ball hitch at the rear of the motorcycle.

4.4.2 Massachusetts farmer interviews

Two Peruvian farmers who typically farm with animals in their home country attended the
field tests in Massachusetts and provided feedback (Farmer interviews were approved by
MIT's Committee on the Use of Humans as Experimental Subjects). One was a terrace

farmer from the Andes who grows potatoes, corn, pumpkin, wheat, barley and quinoa. His
crop rows are usually 10 to 15 m long. The second farmer was a jungle farmer who grows
fruit trees, especially papaya. His crop rows are usually 15 to 20 m long. Both had been

farming for about 25 years with heavy dependence on animal and manual labor.
After two field days of testing, the farmers were interviewed for 90 minutes at the end
of the second day. They were asked the same questions as farmers in India (Appendix C3)
with one exception - due to market differences, instead of being asked to estimate a price
they would purchase Bullkey for, they were asked about the likelihood that they and their
neighbors might purchase Bullkey if it were available. This questionnaire is discussed more
in the next subsection.

4.4.3 Indian farmer interviews

Twenty-four Indian small farmers who use bullocks and/or small tractors were interviewed
one-on-one in the states of Gujarat, Karnataka, and Tamil Nadu (Farmer interviews were
approved by MIT's Committee on the Use of Humans as Experimental Subjects). Their
median farm size was 2.4 ha and 63% of the farmers currently utilize bullocks. Twelve

different crops are grown by the farmers interviewed, including: cotton (16 farmers), maize
(10), wheat (9), peanut (8), onion (7), chilli (7), watermelon (5), and rice (3). These crops
represent low vegetable crops, tall crops, and wide creeper crops - which account for the
types of the nine major crops of India1271.

Indian farmers were interviewed individually for 45 to 90 minutes of conversation guided

by prepared questions that can be separated into five categories: farm dimensions, demo-

87
graphic, farm tool perception, likelihood of adoption, and comments on prototype shown.
Farm dimension and demographic questions were closed form but sometimes followed up by
impromptu questions to inquire more about an unexpected response. Farm tool perception
and comment on prototype questions were open ended with the intention of gaining insights
into the user needs and if they were met by the proposed tractor. Finally, numerical ques-
tions were asked to the farmers to suggest an accessible purchase price as well, and to, using
a Likert scale, rate their likelihood of using the tractor for different operations. Questions
asked to farmers are provided in Appendix C3.
To better explain to the farmers the functionality of Bullkey, they were provided with
a graphic booklet during the interview. The booklet contains four sections: a cover page
showing the Bullkey prototype performing all operations, a graphics-based overview page

summarizing the tractor's capabilities, more detailed pictures of the prototype performing
each operation, and an overview of the balance board design with its usage scenarios. The
booklet was described to farmers via a translator expert in farming. The description typically
took about 20 minutes and included answering questions from the farmer. The booklet

supplied to the farmers is provided in the Appendix C4 Figs. 6-2 to 6-7.

4.5 Results - Field Validation of Operations and Assessment

Interviews

4.5.1 Field Validation in Massachusetts

Operations performed in field tests with the Bullkey prototype demonstrated its ability
to satisfy the needs of small Indian farmers. Bullkey performed the required operations

per the specifications of the guide, Prasad Singh's Indian Farm Machinery handbook [881,
demonstrating that it can do the operations identified in Section 4.2.
The field tests suggested that Bullkey was comfortable and easy to set up to perform
various farm operations. When plowing, the tool could be comfortably inserted into soil and
extracted from the soil by the driving operator. The rotavator's 38 cm diameter pneumatic
wheels always remained in contact with the ground and were used to adjust cut depth when
the tool was engaged. The rotavator could also easily be towed by the Bullkey prototype
at 15 to 20 km/h between fields. Rows of 50 m length were planted and the planter could
be fully lifted (including wheels) by the Bullkey prototype at row ends to facilitate turning

88
(with appropriate front wheel ballasting). The trailer was used to easily carry equipment
from the farm parking area to field-side. The trailer could be comfortably towed using the
outrigger arm or balance board, even when loaded.

Testing the Bullkey prototype highlighted some potential areas for improvement in future
designs. With the prototype tractor, at least 55 kg of ballast on the front wheel was required
to prevent front end lift when the planter was fully raised. In a more refined version of the
prototype this could be reduced by placing the three-point hitch closer to the rear axle.
During testing, drivers would naturally shift their weight forward on the motorcycle to
prevent front end lift when possible. While the balance board worked comfortably for
towing the trailer or sprayer, using it with the cultivator tines was viable but challenging
due to the raised vehicle center of mass and the soil reaction forces generated by the engaged
tines.

4.5.2 Results - Massachusetts farmer interviews

After observing the Bullkey field tests, two Peruvian farmers who have typically farmed
with draft animals in their home country were interviewed. Overall, both farmers reported
that they thought Bullkey was very usable for its intended operations and expressed their
interest in looking to purchase it for their home farms if it were available. The farmers
thought Bullkey would be easy to integrate into their work flow in Peru and thought there
was a high likelihood (5/5) that the vehicle could be locally adopted in Peru (Fig. 4-9).
The mountain farmer thought that Bullkey was particularly valuable because of its pre-
dicted low price and its improved mobility compared to conventional tractors. He also ap-
preciated Bullkey's light weight, which lie said would make it easier to transport up rugged
mountain trails, and that adequate temporary bridges could be constructed to move it across
rivers, which is not possible with conventional tractors. The jungle farmer thought general
driving training would be easier than on a typical tractor since most people he knows in Peru
drive motorcycles, and thus would likely find the idea of using one on the farm attractive. He
also valued that Bulkey could access spaces that a conventional tractor would not be able
to and needed less space to turn. The farmers independently identified many of the original
design requirements for Bullkey - road utility, turning radius, purchase cost, and vehicle size
- as attractive features. They also suggested benefits of the design not previously identified,
like that its light weight could enable it to reach geographically inaccessible regions. The

89
 Very Likely
5
4
'3

Unlikely

Mountain Farmer Jungle Farmer

Figure 4-9: Self-described likelihood of two interviewed Peruvian farmers for doing an ac-
tivity with Bullkey. These farmers were present to observe while Bullkey performed every
operation described in this Chapter. "Local Adoption" refers to the likelihood that they
and their neighbors would use the tractor for farming in Peru.

farmers' interest in these unique design aspects suggests that Bulkey has use for resource

limited small farm settings globally, even though the initial interviews that drove the design

requirements were conducted in India.

The two farmers were generally confident in Bullkey's ability to perform the outlined

farm operations (Fig. 4-9) and rated the likelihood they would use Bulkey very highly
for plowing (average 5/5), cultivating (5/5), spraying (5/5), and secondary tillage (4.5/5).
Both farmers thought BuIlkey was excellent for planting operations, and one noted that he

was particularly happy with the consistent seed depth and spacing the tractor was able to

achieve. Both farmers thought the tractor very convenient for de-weeding with a cultivator

and would enable them to spray more conveniently than they have been able to before. The

mountain-based farmer said that where it is hard to access crops over 1 m with a conventional

tractor, he thinks Bullkey would be "perfect" in those conditions. He said that lie valued

that Bullkey could spray both tall crops and close to the ground. The jungle-based farmer

liked the sprayer setup with its tallest arms and thought it was well suited to the trees he

grows.

Informal discussions with American farmers attending the field tests suggested that

Bullkey could also be valuable in certain situations in the United States. One American

farmer said that since they grow many crops in a relatively small farm, conventional tractors

90
available to them in the USA are often too large for their row lengths and spacing. This
forces them to use more manual labor than the farm manager would like and they said that
a design like Bullkey could be a valuable product for them for farming in Massachusetts.

4.5.3 Feedback from Indian farmers

Overall, Indian farmers were attracted to the possibility of a low cost alternative to trac-
tors that could access narrow spaces and spaces between tall crops like bullocks but was
less expensive to maintain and could work longer hours than animals. Sixty-seven per-

cent of farmers said that a fault of conventional tractors designs is that they are too big
for intercultivation. Ninety-two percent of farmers were satisfied with the Bullkey's width

and 88% were satisfied with Bullkey's weight, two of the major design differentiators with
conventional tractors.
Indian farmers saw a financial value in Bullkey. On average, they said they would be

willing to pay 123,000 INR for Bullkey (standard deviation: 27,500 INR, min: 85,000 INR,
max: 200,000 INR). One of the 24 interviewed farmers chose not to answer the question
oi price point. Based on the initial user needs interviews prior to its inception, Bullkey
was targeted to have a price point of 100,000 INR. The assessment interviews suggested
that the final design would be valuable to farmers at a higher price point, which makes a
product-version of Bullkey more financially viable. The final price point of Bullkey can only
be determined once more is known about its manufacturing and distribution costs.
All interviewed farmers saw Bullkey as a viable road vehicle as a two-wheeled motorcycle,
which was an important design requirement. The average required minimum top speed

reported was 33 km/hr (min: 17 km/lir, max.: 50 km/hr). The farmer who was willing
to accept the slowest top speed said that he preferred a slow vehicle so he would not have
to register it with the government. The preference for a petrol or a diesel engine was

evenly split. A larger concern for farmers than fuel type was maintenance cost and fuel
consumption.
The interviewed farmers thought Bullkey would be able to easily perform all tasks be-
tween planting and harvesting, including intercultivation and spraying (Fig. 4-10). Farmers
rated tasks that would normally be done by bullocks as those they would be most likely to
use Bullkey for - planting (4.7/5), inter-cultivation (4.8/5), and spraying (4.9/5). Farmers
rated Bullkey well but not as highly for tasks that they would use rented medium size trac-

91
 tors for - primary tillage (plowing) (3.75/5), secondary tillage (3.8/5), and trailer operations
(4.1/5). Farmers thought Builkey would perform better at some
tasks than any alternative,
most commonly: intercultivation (80%), spraying (75%), ownership costs (66%), and
small
field seed drill (33%).

iCiiii
Very Likely

0 5

Unlikely
b

Figure 4-10: Self-described likelihood of interviewed Indian farmers for doing an activity
with Bullkey tractor.

The farmer's feedback on the two stabilization configurations was solicited.

Seventy
percent of farmers preferred the outrigger arm and 30% preferred the balance board.
With
the balance board configuration, farmers were concerned about the total width
of ground
contact points for the vehicle. The balance board track width is larger than the width
of
one small tractor tire (which typically operates in single adjacent rows), forcing it
to roll
closer to the seed beds than the farmers are used to and would prefer. The balance
board
is too narrow and short to straddle a row of growing crop like the outrigger arm
or even a
tractor (if the crop is short) would. Ground contact width in a single row is important
to
farmers because it determines how close to the origin of the plant's roots soil compaction
is
occurring.

Ninety-six percent of farmers (all but one) felt confident in the viability of the outrigger
arm for their operations. The unconvinced farmer was hesitant about the "too narrow"
outrigger tire being able to roll on his soft soil - a concern that was also mentioned
by
another farmer. Forty-five percent of farmers said they would find the outrigger arm
even
more usable if it could have a higher ground clearance, allowing them to use it with
even full
grown crops. The desired height varied between 1.4 and 2.1 m. Two farmers also requested
a greater range of lateral adjustment for the outrigger arm, from 0.7 to 1.4m instead
of the

92
current range of 0.9 to 1.2 m.

4.6 Discussion

4.6.1 Satisfying user needs

Originally, the major design requirements identified (Table 4.1) were related to drawbar
pull, purchase and ownership cost, width, path width, turning radius, user comfort, daily
work hours, and road speed. The conducted interviews showed that many of the design
requirements - width, turning radius, user comfort, and road utility - have been well met.
The drawbar pull requirement was validated during earlier technical assessments [45]. The
interviewed farmers were delighted by Bullkey's small size, its ability to maneuver in tight
spaces, and its clear road utility as a two-wheel drive motorcycle. The Bullkey prototype
can travel at 45 km/hr, well above the average minimum top speed requested at 33 km/hr.
While this is not the final vehicular configuration, this functionality shows that a vehicle can
meet the road utility and other farming design requirements simultaneously. Farmers were
generally willing to purchase Bullkey at a higher price point than anticipated, which provides

added flexibility for future vehicle manufacturing considerations. Given this information and
the low cost of the design, a product version of Bullkey should meet the identified purchase
and ownership cost.
The three assessments presented here - field tests, and interviews with farmers in Mas-
sachusetts and India - show that Bullkey meets the intended design requirements. Feedback
from interviewed farmers on how and why they would like to use Bullkey also confirmed the
insights from the original user needs assessment. Some of the collected feedback suggested
that Bullkey may have more general usability (e.g. in small farms in developed countries
and mountainous regions) than originally expected.

4.6.2 Limitations and Future Work

From the functional tests themselves, the major identified areas for future improvement are
moving the rear-mounted three-point hitch mechanism closer to the rear axle and lowering

its center of mass. Placing the mechanism longitudinally closer to the rear axle will lower
the upending moment exerted on the front wheels when lifting heavy implements like the
planter. Lowering the mechanism's center of mass will facilitate using it with the balance

93
board. These changes should be straightforward to implement in a motorcycle frame that
is custom-built for Bullkey to feature three-point hitch arm pivots near the rear axle.
The farmers identified other areas for future improvement. Some of the feedback may
be addressed technically and others may be addressed via clear operational instructions for
the vehicle. In India, some interviewed farmers felt that Bulkey could be even narrower

(for example, by allowing seating with knees together like in a scooter and/or by reducing
handlebar width) or that Bullkey was too lightweight to adequately perform tillage opera-
tions. The latter group's concerns have been addressed in [45], where it is shown Bulkey
can pull more per unit mass than conventional tractors. While 70% of farmers were content
with BuIlkey's towing capability of 600 kg, 30% wanted the ability to tow at least 1000 kg.

The need to transport to centralized storage a day's worth of harvest in a single run was
given as a major reason for large towing capacity. Current towing capacity is limited by the
ability to quickly stop at road speeds and safely handle road speed turns with with a towed
load. Towing capacity could be improved by having connectors to interface with trailers
that feature an independent braking system and/or by limiting Bullkey's top speed when
towing.
One farmer was concerned that the vehicle may not be stiff enough in roll to use a planter
correctly, and that if the planter tracking wheels lost traction with the ground it might not
seed at appropriate intervals. This did not show itself as an issue during field testing. One
farmer expressed concern regarding misalignment between the implement wheels (like those
of the planter) and Bullkey's wheels. They wanted alignment between the wheels to reduce
the number of lanes of compacted soil on the field. One farmer mentioned that since the dirt
paths around him have pronounced wheel ruts, he was unsure how the motorcycle would
perform while driving on the hump between wheel ruts (potentially with the outrigger in
one rut) and pulling a trailer which has its wheels in the ruts. Aligning the wheels of trailers

differently relative to Bulkey could be readily achieved by laterally shifting the tow hitch
mount towards the outrigger wheel or with custom implements that have a wheel layout to
address this.
The Peruvian farmers suggested that users who have previously only driven draft animals
may need training on which farming tools they should use, particularly for choosing a tillage
tool appropriate for soil conditions. The jungle-based Peruvian farmer mentioned that he
would like the option to add more than 55 kg of ballast over the front wheel in order to

94
utilize an even larger planter with the tractor (raising the planter off the ground transfers
weight away from the front wheel). Reinforcing the frame to allow for additional ballast
should be straight forward in a final design.

4.7 Conclusions

This Chapter described why conventional tractors are not well suited to small farmers (<2
Ha) in India, and similar small farm markets globally. This is reflected in the majority of
Indian small farmers relying on other sources of draft power, mainly bullocks and manual
labor. A more suitable tractor design proposed by the author, called Bullkey, has dimen-
sions akin to bullocks. This facilitates efficient use of scarce farmland, while retaining many
advantages of conventional tractors including their lower ownership cost and longer operat-
ing hours. A Bullkey prototype was farm tested and demonstrated the ability to perform
functions important to Indian farmers. At the test site, Peruvian farmers who have mostly
farmed with draft animals observed field tests with Bullkey and were interviewed for feed-
back. They felt the design could satisfy their needs and those of their neighbors at their
home farms.

Bullkey's design and its field performance, including images, were then discussed one-
on-one with small farmers in three Indian states. These farmers felt Bullkey met their needs
and suggested a purchase price above that which was identified as a target. The Bulkey
design is well positioned to transition into a production vehicle. The design is based on
well understood technologies, and the field tests and user studies presented here show that
combining these technologies yields a functional farm vehicle able to operate in spaces not

accessible with other options in the market.

The distribution of Bullkey could increase small farm mechanization. This would make
farms more efficient in their use of resources (including labor time), increase crop production,
and improve the quality of life for the farmers through reduced drudgery and increased
income.

95
THIS PAGE INTENTIONALLY LEFT BLANK

96
Chapter 5

Conclusion

This thesis proposes a small tractor concept, called Bullkey, designed for small farms in low
resource settings - particularly in rural India. The chapters provided a review of the evolution
of the conventional tractor and why it is inadequate for small Indian farms (Chapter 2), a
model of tractor physics and the user needs considered to create Bullkey (Chapters 3 and
4), the implementation and testing of a Bullkey proof-of-physics prototype demonstrating
its high drawbar pull per unit mass capabilities (Chapter 3), and the creation of a Bullkey
functional prototype that was tested performing key operations to Indian farmers and then
evaluated by the farmers (Chapter 4).

5.1 Motivation of Research

The majority of farms in the world (84%) are less than 2 Ha in size {58]. Small farms are
particularly common in India, where the average farm size has steadily decreased from 2.28

Ha in 1971 to 1.08 Ha in 20161751. Most small farmers use a pair of bullocks (bovine draft
animals) for all or most of their farming operations, supplementing the rest with manual
labor or a hired tractor (particularly for primary and secondary tillage) [5211311. Bullocks
are compact, highly maneuverable, and have a low capital cost, making them well-suited

to the technical and economic constraints of small farms. Conventional tractors, which are
an icon of modern farming and typically associated with high farm yields 1391, have not
yet been able to reproduce these key bullock features164][761. As a result, farmers are
constrained to the slow speed of bullocks and a. lack of access to suitable modern, more
effective made-for-tractor tools[31][30].

97
 This misalignment between the tractor and needs of the small farmer is, in part, because
conventional tractors were designed for a larger field size. The conventional small tractor
was originally intended to operate in much larger farm holdings than what is typical around

the world [44]. The conventional small tractor largely found its form in the US between 1910
and 1940 [44] [10711771 [61] 11041, where the annual average farm size since 1900 has been
at least 50 times greater than the present global average. Globally, at present, 72% of farms
are <1 Ha, while less than 2% of US farms have been <1 Ha since 1900 [58] 1331 [14 [99]
[59]. As a result, for many small farmers, replacing their bullocks with a conventional tractor
would require adjusting crop spacing (particularly for crops taller than a tractor's ground
clearance), or sacrificing farmland as headland area to give the tractor space to maneuver

at row ends. It might also require improving road access to their field due to the tractor's
larger size and poorer off-roadability.

5.2 Contributions from Research

5.2.1 Evolution of conventional tractor

Chapter 2 explains the origin and merits of the dominant farm tractor design, which has
endured since the 1940s. Understanding the origins and rationale for this dominant design
enables engineers to appreciate its merits, and understand its drawbacks, which could be
addressed in future designs. Additionally, the methods used in this research to study tractor
evolution are applicable to the study of other products with a longstanding dominant design.
Two themes are covered in Chapter 2: first, the historical context that directed the farm
tractor's design evolution is presented; and second, a terramechanics-based tractor model is
used to analyze why the dominant design is conducive to good performance. The prominent
characteristics of the dominant tractor design are its weight distribution, wheel layout, tool
location, and construction. It's weight distribution maximizes drawbar pull by placing 70

to 80% of the total vehicle weight on the rear wheels. Shifting the weight forward reduces
pulling force while shifting it backward produces a negligible increase in pulling capacity
while dangerously increasing the risk of upending the tractor. The tractor has four wheels
arranged in a rectangular pattern - the rear wheels are driven while the front ones are usually
idle. Rear wheels are of large diameter to increase ground clearance and tractive efficiency.
Front wheels are of small diameter to allow for a large steering angle despite a narrow track

98
width. A narrow track width reduces the space required for making a U-turn at field ends
and improves access to farm spaces. Inline front and rear wheels are desirable for ease of
driving between rows and to best harness soil compaction. Attaching implements behind
the rear axle leverages tillage forces to increase maximum drawbar pull and enables using
large tools. The tractor's crankcase and transmission housing are structural components

-
this reduces mass and manufacturing complexity.

5.2.2 Design of Bullkey for small farms

Chapter 3 describes the motivation, design, and testing of a specialized farm tractor designed
to replace draft animals in small farms, particularly in rural India. The vehicle is specially
suitable for this use environment because it matches the low capital cost of animals and
has their unique ability to operate in narrow inter-crop-row lanes while retaining the major
advantages of tractors, such as low maintenance cost and reduced operator physical effort.
The presented vehicle was designed based on user needs and our implementation of a detailed
terramechanics model. The proposed tractor design improves drawbar pull per unit mass
compared to conventional tractors by applying nearly the full vehicle's weight on the drive
wheels, placing drive wheels inline, and locating the tillage tool between both axles. A

proof-of-physics prototype of this design was built based on an all-wheel-drive motorcycle

retrofitted with an outrigger wheel for stabilization, and a tillage tool attachment point
between the front an rear axles. The prototype was instrumented to measure drawbar pull
and tire slip to validate the terramechanics model and quantify traction performance. During
field tests on a working farm, the vehicle successfully operated in the narrow spaces between
growing crops that would typically not be accessible to a low cost (< 15lip) conventional
small tractor. Traction experiment results showed that the proposed design is capable of
pulling with more drawbar force per unit mass than conventional tractors and that this

performance can be accurately predicted by the provided physics model. Initial farmer
feedback on the design confirmed its high potential for performing tillage in small farms and

other farming operations with further improvements.

5.2.3 Viability of Bullkey for replacing bullocks

Chapter 4 describes the design, functional testing, and user feedback for a tractor special-
ized for small farms in low resource settings, particularly India. The presented tractor is

99
unique in its ability to compete with bullocks' physical dimensions, pulling performance,
and sale price, while retaining key tractor advantages like compatibility with modern tools,
low maintenance costs, and reduced drudgery. This tractor features motorcycle-like controls
and seating, inline drive wheels, stabilization via an outrigger arm or a specially-developed
balance board attachment, and the ability to attach implements ahead or behind the rear
axle. A prototype of the tractor was built and its ability to complete key farming operations
was demonstrated in a Massachusetts farm. The vehicle was able to complete plowing, disc
harrowing, rotary tilling, planting, cultivating, spraying, and towing of a trailer per Indian
industry specifications. Two Peruvian small farmers drove the vehicle and thought it could
be a valuable replacement for draft animals in their domestic farms. A study was conducted

to assess whether the vehicle would meet the needs of small and marginal farmers in India
with 24 farmers in Karnataka, Gujarat, and Tamil Nadu. Farmers were shown photos and
videos of the operating vehicle and performance data, and were asked numerical and open
ended questions about the vehicle and their general farming experiences and needs during
45-90 minute one-on-one interviews. Indian farmers interviewed generally reported that the

prototype tractor to be would meet their needs and suggested they would be willing to
purchase the vehicle for around 123,000 Indian Rupees, about 22% higher than the price
target with which the tractor was designed. The interviewed farmers' reported an aver-
age likelihood of 4.8/5 that they would use the vehicle for planting, inter-cultivation, and
spraying and al average likelihood of 3.8/5 that they would use the tractor for primary or
secondary tillage. Overall, in this work I found that the proposed vehicle is better suited
than conventional small tractors to replacing draft animals in small Indian farms.

5.3 Future Work

Future work should demonstrate the Bullkey tractor on-site in small Indian farmers for
the end users and other key stake holders to evaluate. This will give the stakeholders
a clearer understanding of how Bullkey performs, allowing the researchers to collect more
accurate data on Bullkey's likelihood of market success. Before demonstrating the tractor in
India, the major opportunities for improvement are in the attachment interfaces for farming
implements.

Improvements can be made to both rear and central tool attachment points. The rear

100
tool attachment three point hitch (an industry standard) can be kept mostly as is but should
be mounted closer to the rear axle - this will reduce the amount of front wheel ballast needed
when lifting heavy implements. The central tillage tool attachment could be improved in
its motion mechanics and the control input ergonomics.
It's motion mechanics should enable two things: the weight of the tractor to preload
the tool during initial insertion into the soil and, once soil draft is pulling the tool down,
for the tool to sink into the soil to user-set depth without constant user input. The former
is already present in the current design and is important because of the light weight tools
Bullkey is expected to use; the latter is not present in the current design and is important
for presenting the user with a familiar and user-friendly work process. In a conventional
tractor, tillage tools can be lifted out of the soil but not pushed into the soil - tools sinks
into the soil to a user-set depth initially under their own weight and, once deep enough into
the soil, due to soil draft forces. However, conventional tractors utilize hydraulic power that
allows for heavier tools than a human operator can lift in manually powered mini tractor
systems. Mini tractor operators in India hire an additional laborer to stand on the tool
during insertion into the soil, effectively increasing the tool's weight, and then step off the
relatively lightweight tool when it must be raised (Fig. 5-1). Hiring this extra laborer adds
to operating costs.

5.3.1 Central tillage tool improved mechanism

The current Bulkey tillage tool control mechanism is non-backdriveable in both directions

(upward and downward). This is behavior stems from it being powered by a hand cranked
worm gear that drives the pinion gear in a rack and pinion system. It allows applying the
tractor's weight to the soil-working tool in much the same way as a typical vehicle lifting
jack pushes on the ground. However, it also prevents soil draft forces from moving the tool
downward to a pre-set depth without constant user input. Instead, the user must manually
crank the mechanism until it reaches the desired depth.
This could be improved by a ratchet-wrench-like direction switch on the crank handle,
making the system non-backdriveable in only one user-selected direction. During insertion
the user should set the switch to be able to push the tool down but not pull it up. In other
words, during insertion turning the crank handle increases the minimum tool soil depth but
does not interfere with soil draft forces pulling the tool further down. Then during lifting of

101
 -1

Figure 5-1: Images of people riding on tools to help them sink into the ground. These are
tools mounted on conventional three point hitch linkages but that are manually operated.

the tool, the direction switch should be flipped to now allow the user to pull on the tool and
raise it out of the soil. During active tillage, the user can flip the switch as needed to adjust
tool depth. The tool should have an adjustable stop point to set maximum tool depth. This

stop point could be achieved by a sliding collar on the tool rack or by a rack with holes for
pin that limits motion.
If the current tool control mechanism is to be implemented, another improvement to
explore is raising the crank handle by about 30 cm. This would improve the mechanism's
user ergonomics by not requiring the operator to bend down when turning the crank handle.
To explore this adjustment a simple fit test device able represent raising the crank handle

by 3 to 60 cm was installed on the Bullkey prototype (Fig. 5-2A). Two researchers between
167 cm and 180 cm then tested raising and lowering the device's crank handle, agreeing on
a most comfortable height 31.5 cmi above the current height. This height allowed them to
turn the crank handle in a comfortable riding position (Fig. 5-2B) instead of having to bend
down as must be done on the current prototype (Fig. 5-2C).

Alternatively to incorporating a ratcheting feature on the current mechanism, a new

mechanism with a motion path which intrinsically leverages the tractor weight for tool
insertion could be investigated. In this design the tool would move backward and downward

102
 "A'At

Figure 5-2: Tests to explore crank handle height. (A) a simple device to test crank handle
heights up 60 cm higher than the current design. (B) The user's motion with a handle height
~31.5 cm higher than the current design. (C) The user's motion with the current handle
height.

during insertion. Therefore the soil drag forces pushing the tool back are now also pushing it

down (a similar working principle to a typical door stop). By contrast, a conventional three
point hitch inserts the tool down and forward (see Fig. 5-3). An improved tool motion

path for small tractors with low lifting strength should do two things: 1) Insert the tool into
the soil downward and backward, thus utilizing the torque created by the soil drag force. 2)
Lift the tool upward and backward - thus allowing the tool to reverse into the furrow (i.e.

the rut it created) during lifting and reducing the amount of soil that must be disturbed to
lift. Some strategies for achieving this motion path are discussed in Appendix D.

5.3.2 Next steps for distribution of Bulkey

The current Bullkey prototype has important areas to be improved for mass manufactured
and distribution. This areas for improvement mainly revolve around improvements the

limitations of the prototype caused by its requirements to be instrumented and viable for
being manufactured by the author in a small machine shop. The central tool attachment
should be made more compact and the rear tool attachment should be placed closed to the
rear axle. As suggested by users, the outrigger arm should have greater adjustability laterally

103
Figure 5-3: The conventional tool motion path (top) compared to the tool path proposed
by the authors (bottom). The latter is well suited to very small tractors since it requires
less force to operate.

and vertically to suit all crops. The balance board could be improved with an alternative

mechanism that still exploits the same principles. Finally, a new vehicle platform should be

built from the ground up to consider safety and have a reverse gear.
The current tool attachment mechanisms are unnecessarily large. The central tool at-
tachment currently occupies about half the length of the vehicle, this was to facilitate instru-

mentation and is unnecessary for production. As discussed in Section 5.3.1, the mechanism
could be made very compact by consisting only of a narrow jack. This would liberate more
space for the user to comfortably straddle the motorcycle frame with their legs. The rear
tool attachment method on the prototype was limited by the attachment points on the base

ROKON vehicle. A production attachment system should be moved closer to the rear axle
of the motorcycle, this would improve the ease-of-use for heavy rear-mounted tools and make

the tractor more compact.

The proof-of-concept balance board shown in this thesis is functional and has high po-
tential to be a viable implementation. However, it is not a formally optimized design in
dimensions or mechanical layout - that optimization is left for future work. For discussion
of design challenges and freedoms in a balance board, I present an alternative design that

104
improves clearance to the motorcycle frame, eliminates side-to-side motion during turning,
and allows reversing without the balance board position being unstable. This alternative
design replaces the ball hitch of the implemented design with four to five pin joints. This
alternative design has advantages and disadvantages compared to the implemented design.
In brief terms, the current design is simpler but its motion is less ideal.

longitudinal pivot between

frame and balance board yoke

[foot pedal

casterwheed

arms

arms pivot
along yoke axis

Figure 5-4: AN ALTERNATIVE BALANCE BOARD DESIGN CONCEPT ALSO CRE-

ATED BY THE AUTHORS. The two arms could be rigidly coupled to rotate together (four
total pin joints) or independent (five total pin joints). The arms are assumed to be rigidly
coupled.

Looking to the future, it would be interesting to consider making this tractor electric if

solar panel and battery technologies become capable of sustaining its performance at a low
cost. Bullkey's two-wheel drive transmission could be simplified by having two independent

electric motors, one for the front wheel and one for the rear wheel. Additionally, this would
enable the implementation of a computer-controlled traction control system that further
improves Bullkey's drawbar pull capabilities. Batteries could be placed low on the vehicle

105
to achieve a low center of mass and their weight would contribute to improving traction.
The tool control systems could also be electrically powered.

5.4 Conclusion

In this thesis a tractor design specialized to small Indian farms was created and prototyped.

This tractor is called Bullkey and utilizes inline drive wheels, a centrally located tool, and
stabilization fitments that do not interfere with its ability to access narrow spaces between
crop rows. The drive wheel and tool locations augment its traction performance, allowing it

to maintain a pulling performance comparable to bullocks while being sold at a lower price
than conventional tractors. Accessing narrow inter-row spaces is a key advantage of bullocks
over conventional tractors; Bullkey's ability to replicate that accessibility increases its value
to farmers and enables it to fully replace bullocks.
To study existing tractor designs and explore ways to augment Bullkey's traction perfor-
mance in a narrow package of dimensions comparable to bullocks, a tractor physics model
was implemented. Leveraging the model, the Bullkey design was identified and a proof-
of-physics prototype created to test its traction performance and to confirm the modelling
behind it. This prototype validated the Bullkey's high drawbar pull per unit mass perfor-
mance and ability to access narrow inter-row spaces between growing crops.
After validating the performance of the Bullkey, the prototype was modified to improve
its overall usability and allow it to perform farming operations key to Indian farmers. After
these modification, the Bullkey demonstrated in field tests its ability to perform the required
operations. Afterwards, the Bullkey design was discussed one-on-one with 24 Indian farmers
in their home villages. The results from these interviews showed that farmers believed
Bullkey could replace bullocks in their farming operations and that their suggested purchase

price was higher than the sale price Bullkey was designed for.
The Bullkey design has high potential for productization. It is designed to fulfill a well
documented user need, it is built on solid engineering principles, and the target users have
expressed an understanding of its advantages as well as an interest in purchasing it for a
price higher than its initial target sale price.

106
Bibliography

[1] Mohammed Shu'aibu Abubakar, Desa Ahmad, and Fatai Butola Akande. A review
of farm tractor overturning accidents and safety. Petranika Journal of Science and
Technology, 2010.

[2] Adafruit. Adxl335 - 5v ready triple-axis accelerometer (+-3g analog out), October
2019. https://www.adafruit.coi/product/163.

[31 Diogenes L. Antille, John McL. Bennett, and Troy A. Jensen. Soil compaction and con-
trolled traffic considerations in australian cotton-farming systems. Crop and Pasture
Science, 2016.

[4] V. N. Murthy Arelekatti, Douglas H. Bj6rkdal, Carmen W. Graves, Anthony Wong,

Armen Mkrtchyan, and Amos G. Winter V. Proof-of-concept evaluation of a low-cost
and low-weight tractor for small-scale farms. In ASME IDETC 2014, Buffalo, New
York, 2014.

[5] Aruna Pal and P.N. Chatterjee. Field level study on the buffalo bullock, an excellent
draught animal. Buffalo Bulletin, 32, 2013.

[6] Automation Direct. Mafm1-a0-lh, October 2019. https://www.automationdirect.com.

[71 Andrea Battiato and Ettiene Diserens. Influence of tyre inflation pressure and wheel
load on the traction performance of a 65 kwmnfwd tractor on a cohesive soil. Journal
of AgriculturalScience, 2013.

[8] Andrea Battiato, Ettiene Diserens, and L. Sartori. Traction performance simulation
for mechanical front wheel drive tractors: towards a practical computer tool. Journal
of AgriculturalEngineering, 2013.

[9] Andrea Battiato, Ettiene Diserens, and L. Sartori. Tractor traction performance sim-
ulation on differently textured soils and validation: a basic study to make traction and
energy requirements accessible to the practice. Soil and Tillage Research, 2017.

[10] Mieczyslaw Gregory Bekker. Theory of Land Locomotion. University of Michigan

Press, 1956.

[111 Black Boar. S-tine cultivator, 2019. https://blackboaratv.com/products/s-tine-

cultivator.

[121 Briggs& Stratton. Cr 950 engine, 2019. https://www.briggsandstratton.com/na/en

us product-catalog/engines/utility-engines/r950-series.html.

107
1131 W.W. Brixius. Traction prediction equations for bias ply tires. ASAE Paper, 1987.

[141 USA Farm Census Bureau. 1969 usa census of agriculture. USA Farm Census Bureau,
1973.

[151 CALT 500mm Measure Range 0-5Kohm Draw Wire Potentiometer. Web-
site with details on sensor, 2019. https://www.amazon.com/dp/BO1FHDL81Y/
ref=cm sw em_r_mtdp_U_RZfQDbFM2TDZ6.

1161 William W. Cassidy. Tractor tire and ballast management. University Extension,
University of Missouri Columbia, 2017.

[17] Tim Chamen. Controlled traffic farming - from worldwide research to adoption in eu-
rope and its future prospects. In InternationalConference of AgriculturalEngineering,
Zurich, 2014.

1181 Charles0. Creative commons license, August 2011. Wikimedia.

[191 China Enconder. Pull pressure force sensor s-type load cell 300kg 500kg 1.5t
for concrete mixing station batching scale hopper scale pressure testing ma-
chine (500kg), October 2019. https://www.amazon.com/Pressure-Concrete-Station-
Batching-Testing/dp/B01HHO5RNQ.

1201 CM Eason. Tendency of farm tractor design. Transactions of the American Society of
Agricultural Engineers, pages 59-67, 1916.

[21] W.W. Cochrane. The Development of American Agriculture - A HistoricalAnalysis.

University of Minnesota Press, 1993.

[22] Damodaran, Harish. What it might cost to save gauvansh countrywide. The
Indian Erpress, April 2017. indianexpress.com/article/explained/gau-rakshaks-
cattle-protection-project-cow-hansraj-ahirwhat-it-might-cost-to-save-gauvansh-
countrywide-4625488/.

[23] DataQ Instruments. Di-2108 8-channel high speed usb data acquisition system, Octo-
ber 2019. https://www.dataq.com/products/di-2108/.

[24] Gustavo Soares de Souza, Zigomar Menezes de Souza, Reginaldo Barboza da Silva,
Ronny Sobreira Barbosa, and Fernando Silva Ara jo. Effects of traffic control on the
soil physical quality and the cultivation of sugarcane. Revista Brasileira de Cidncia
do Solo, 2014.

[25] Jodi Dejong-Hughes. Tires, traction and compaction. University of Minnesota Exten-
sion, 2017. Extension UMN website.

1261 Ivan Demsar, Rajko BERNIK, and Joze Duhovnik. A mathematical model and nu-
merical simulation of the static stability of a tractor. Agric. conspec. sci., 77, 2012.

[27] Department of Agriculture, Cooperation and Farmer Welfare. Annual report 2017 to
2018. Government of India Ministry of Agriculture Farmers Welfare, 2019.

[281 Lynn W. Ellis. The problem of the small farm tractor. Scientific American, pages
518-519,525-528,1913.

108
[291 Engineering and Contracting. Image in public domain, May 1909.

[301 Engineers at Mahindra and Mahindra Co. Conversations with farmers and researchers
in india on tractor adoption. met multiple times from 2014 to sept. 2019.

1311 Engineers at Mahindra and Mahindra Co. Conversations with mahindra and mahindra
tractor company on tractor adoption in india, 2016.

[321 FAO (Food and Agricultural Organization). Fao, the first 40 years: 1945-1985, 1985.

133] FAO (Food and Agricultural Organization). Mechanization for rural development: A
review of patterns and progress from around the world. Integrated Crop Management,
2013.

[34] Farm, Ranch Safety, and Health Association of British Columbia. Agricultural equip-
ment safety. SA FE Work Manitoba Communicatios, 2016.

[351 FarmAll Tractor. Public domain, 1925. Tractors Wikia.

[36] Giovanni Federico. The growth of world agricultural production 1800-1938. Research
in Economic History, Vol. 22, 2004.

[37] Giovanni Federico. Not guilty? agriculture in the 1920s and the great depression. The
Journal of Economic HistoryVol.14,2005.

138] Food, Agricultural Strategic Advisory, YES BANK. Research (FASAR), and German
Agribusiness Alliance at OAV German Asia-Pacific Business Association (GAA). Farm
mechanization in india. the custom hiring perspective. Indian Ministry of Agriculture
and Farmers Welfare, 2016.

[39] Andrew Foster and Mark R. Rosenzweig. Barriers to farm profitability in india: Mech-
anization, scale and credit markets. World Bank Resources, 2010.

[401 G Spoor and R Carillon and L Bournas and EH Brown. The Impact of Mechanization.
John Wiley and Sons Ltd., Chichester, New York, "1987.

[41] Bahareh Ghotbi, Francisco Gonzalez, Jbzsef Kdvecses, and Jorge Angeles. Effect of
multi-pass on the mobility of wheeled robots on soft terrain. In ECCOMAS Thematic
Conference on Multibody Dynamics, Barcelona, 2015.

[421 Michael R. Goe and Robert E. McDowell. Animal traction guidelines for utilization.
Cornell InternationalAgriculture Mimeograph, 1980.

[43] R.B. Gray. The Agricultural Tractor 1855-1950. American Society of Agricultural
Engineers, 1954.

[44] Guillermo F. Diaz Lankenau and Amos G. Winter V. An engineering review of the
farm tractor's evolution to a dominant design. ASME Journal of Mechanical Design,
2018. doi:10.1115/DETC2018-86285.

[45] Guillermo F. Diaz Lankenau and Amos G. Winter V. Design of a specialized tractor
to replace draft animals in small farms. In Review., 2019.

109
[46] Guillermo F. Diaz Lankenau and Amos G. Winter V. Investigation of viability to re-
place draft animals with all-wheel-drive motorcycles on small indian farms. In Review,
2019.

[471 H. Goel, V. Kumar. Automobiles, sixth gear. Kotak Institutional Equities, January
2013.

[48] Haacon Lifting Technology. Company website with details on products,

2019. https://www.haacon.com/en/commercial-vehicle-equipment/rack-and-pinion-
jacks/rack-and-pinion-jack-1524.php.

[49] HansBinswagner. Na. NA, 2005.

[50] Peter B. R. Hazell. Is small farm-led development still a relevant strategy for africa
and asia? Oxford Scholarship, 2015.

[51] I.C. Holm. Multi-pass behaviour of pneumatic tires. Journalof Terramechanics,pages

47-71,1969.

[52] Indian Agrigultural Ministry. Presentation on farm mechanization before parliamen-

tary consultative committee, January 2013.

[53] Joomen. Joomen enc set 15-600mm 2x linear guideway rail 4x square type car-
riage bearing block, October 2019. https://www.amazon.com/Joomen-15-600mm-
Guideway-carriage-bearing/dp/B1CMGOOMS.

[54] J.Y. Wong and A. Reece. Prediction of rigid wheel performance based on the analysis
of soil-wheel stresses part i. performance of driven rigid wheels. Journal of Terrame-
chanics, pages 81-98, 1967.

1551 Kolpin Outdoors. Atv utv dirtworks tool attachment - disc plow kit, 2019.
https://www.kolpin.com/disc-plow.

[561 R.T. Kudrle. Agricultural Tractors: A World Industry Study. Ballinger Publishing
Company, 1975.

157] Guillermo F. Diaz Lankenau, Lea Daigle, Samuel H. Ihns, Eric Koch, Jana Saadi,
Patrick Tornes, Jessica M. Wu, and V Amos G. Winter. Design of a human-powered
roll stabilization attachment for utilitarian two-wheeled vehicles. ASME IDETC, 2019.
Anaheim, CA.

1581 S.K. Lowder, J. Skoet, and S. Singh. What do we really know about the number and
distribution of family farms worldwide? background paper for the state of food and
agriculture. ESA Working Paper No. 14-02. Rome FAO, 2014.

[591 James M. MacDonald, Penni Korb, and Robert A. Hoppe. Farm size and the organi-
zation of u.s. crop farming. USDA Economic Research Report - 152, 2013.

[60] Mahindra Tractors. Company website with details on vehicles, 2019.

https://www.mahindratractor.com/tractor-mechanisation-solutions/tractor/yuvraj-
215-nxt.

110
[61] Scott Marshaus. Dominant design: How the ferguson system revolutionized mecha-
nized agriculture. Master's thesis, University of Winsonsin-Eau Claire, History De-
partment, 2015.

[62] McCormick Deering Instruction Manual. 15-30 hp GEAR DRIVE TRACTOR manu-
factured by International Harvester Company. Public domain, 1920.

[63] C.R. Mehta, N.S. Chandel, P.C. Jena, and Anamika Jha. Indian agriculture count-
ing on farm mechanization. Agricultural Mechanization in Asia, Africa AND Latin
America, 50, 2019.

1641 C.R. Mehta, N.S. Chandel, T. Senthikumar, and Kanchan K Singh. Trends of agri-
cultural mechanization in india. CSAM Policy Brief, 2014.

[65] Manik Mital and Rahul Gawade. Tractor price, new tractors, buy and sell used trac-
tors, October 2019. https://tractorguru.in/.

[661 Moline Plow Company Advertisement. Public domain, March 1918. Gas Power, vol-
umen 15, issue 9.

[67] MOOG Suspension Parts. moog-513267 I front wheel bearing and hub assembly I acura
mdx 4wd, October 2019. https://www.moog-suspension-parts.com/moog-513267.

[681 Pranab Kumar NAG and Anjali NAG. Drudgery, accidents and injuries in indian
agriculture. Journalof IndustrialHealth, 42, 2004.

[691 Nebraska Tractor Test Lab. Test procedure and historical archives are well documented
in official website., June 2018. https://tractortestlab.unl.edu/.

[701 North Berdfordshire Gazette. Image in public domain, January 1903.

[71] Northern Tool. Field tuff 3-pt. hobby seed planter - 0.22 bushel capacity, model
num. ftf-cbp3pt, 2019. https://www.northerntool.com/shop/tools/product 200622245
200622245.

[72] Northern Tool. Field tuff tow-behind tiller - 36in. width, model num. atv-3665, 2019.
https://www.northerntool.com/shop/tools/product 200622251 200622251.

[73] NSL Srivastava. Farm power sources, their availability and future requirements to
sustain agricultural production. Indian Council of Agricultural Research, 2006.

[74] Derek S. Oden. Harvest of hazards the farm safety movement. PhD thesis at Iowa
State University, 2006.

[75] Minisitry of Agriculture for Government of India. Agriculture census 2015-2016. Agri-
culture Census Division, 2019.

[76] Indian Council of Food and Agriculture. Farm mechanization national round table
conference, 2017.

[77] Alan L Olmstead and Paul W. Rhode. Reshaping the landscape: The impact and
diffusion of the tractor in american agriculture, 1910 - 1960. The Journalof Economic
History, pages 663-698, 2001.

111
[78] Raymond Olney. Signs of progress in the farm tractor field. Power Farming, pages
9-11, 1918.

179] Ismet Onal. Controlled traffic and widespan tractors. Tarim MakinalariBilimi Dergisi
(Journal of Agricultural Machinery Science), 2012.

[801 Philip, A J. Cow protection. Indian Currents, April 2017.

www.indiancurrents.org/cow-protection-1558.php.

1811 Prabu M J. A tilting cart offers relief to animals and workers in the field.
The Hindu, April 2010. www.thehindu.com/sci-tech/agriculture/A-tilting-cart-offers-
relief-to-animals-and-workers-in-the-field/articlel6371609.ece.

[821 B. Sanjeeva Reddy, Indravarapu Srinivas, Ravikanth V. Adake, C.R. Thyagraj,

K. Sammi Reddy, and C. Srinivas Rao. Small farm mechanization technologies and
transfer strategies. Indian Farming, 65, 2015.

[831 Research and Experimental Centre for Tropical Mechanical Agricultural Equipment.
The employment of draught animals in agriculture. Food and Agriculture Organization
of the United Nations, 1972.

[84] Rokon International Inc. Company website with details on vehicles, 2019.
https://www.rokon.com/bikes/scout.

[851 Carmine Senatore. Prediction of mobility, handling, and tractive efficiency of wheeled
off-road vehicles. PhD dissertation, Virginia Polytechnic Institute and State University,
Department of Mechanical Engineering, 2010.

[861 Hassan Shibly. Analysis of the effect of soft soil's parameters change on planetary
vehicles' dynamic response. Journal of Automation, Mobile Robotics and Intelligent
Systems, 11, 2017.

[87] Sukhpal Singh. Agricultural machinery industry in india a study of grwoth, mar-
ket strategy, and business strategies. Centre for Management in Agriculture. Indian
Institute of Management Ahmedabad, 2009.

[881 Triveni Prasad Singh. Farm machinery. PHI Learning Pvt. Ltd., 2016.

[891 David W. Smith. Safe tractor operation: Rollover prevention. Texas Agricultural and
Mechanical University. AgriLife Extension, 2005.

[90] Smitth and Grisso. Using tillage horsepower more efficiently: Selecting speed, slip and
ballast. In Conservation Tillage Proceeding 9:79-81, 1990.

[91] Spirex Radiator ad in Country Gentleman Magazine. Public domain, May 1918.

[921 Square Turn Company Advertisement. Public domain, 1916.

[93] Tacuna Systems. Load cell amplifier strain gauge amplifier, October
2019. https://tacnasystems.com/products/amplifiers-conditioners/strain-gauge-or-
load-cell-amplifier-conditioner-interface-2/.

[94] The Tractor Field Book. Public domain, 1916.

112
 [95] Titan Tire. 489 xt, October 2019. http://www.titantirestore.com/titan-489-xt-atv-
tires.html.

[96] Tractor Junction. New tractors price in india, top tractor brands, specs, photos,
reviews and videos, October 2019. https://www.tractorjunction.com/tractors.

1971 Martin Udengaard and Karl Iagnemma. Analysis, design, and control of an omni-
directional mobile robot in rough terrain. Journal of Mechanical Design, Vol. 131,
2009.

[98] USA Census Bureau. Agriculture 1950: Changes in agriculture, 1900 to 1950, 1950.
Web PDF.

[99] Economic Research Service USDA. U.s. farms: Numbers, size, and ownership. Struc-
ture and Finances of U.S. Farms: 2005 Family Farm Report EIB-12, 2005.

[100] User: BulldozerD11. Creative commons license, 2009. Tractors Wikia.

[1011 User: BulldozerD11. Shared to crowd sourced farm tractor website, 2016. Tractors
Wikia.

[1021 Diego L. Valera, Jess Gil, and Juan Agiiera. Design of a new sensor for determination
of the effects of tractor field usage in southern spain: Soil sinkage and alterations in
the cone index and dry bulk density. Sensors, 2012.

[103] Vevor Machinery Equipment. Company homepage, 2019.

http://www.globalsources.com/vevor.co.

[104] W. J. White. Economic history of tractors in the united states - economic history
association, June 2018. https://eh.net/encyclopedia/economic-history-of-tractors-in-
the-united-states//.

[105] Wade and Dunton Motors Inc. Vintage advertisement image in public domain., 1917.

[106] Peter R. Watson. Animal traction. Peace Corps. by TransCentury Corporation, 1981.
[107] Robert C. Williams. Fordson, Farmall, and Poppin' Johnny: A History of the Farm
Tractor and Its Impact on America. Champaign: University of Illinois Press, Urbana,
Illinois, 1987.

1108] J.Y. Wong. Terramechanicsand Off-Road Vehicle Engineering. Elsevier, Oxford, UK,
2010.

[109] Amy E. Wood and Christopher A. Mattson. Designing for the developing world:
Common pitfalls and how to avoid them. Journal of Mechanical Design, Vol. 138,
2016.

[110] Yongua Xiong, Lei Tian, Tofael Ahamed, and Bin Zhao. Development of the reconfig-
urable data acquisition vehicle for bio-energy crop sensing and management. Journal
of Mechanical Design, Vol. 134. 2012.

[1111 Sandeep Yadav and S. Kumar Lohan. Tractor and implement ownership and utilization
of haryana. AgriculturalMechanization in Asia, Africa AND Latin America, 37, 2006.

113
[1121 Arnold P. Yerkes and LM Church. The gas tractor in eastern farming, 1918.

[113] Arnold P. Yerkes and H.H. Mowry. Farm experiences with the tractor, April 1915.

[114] H. Zell. Creative commons license and gnu free documentation license, July 2016.
Wikimedia.

[1151 Frank M Zoz and Robert D Grisso. Traction and tractor performance. In Agricultural
Equipment Technology Conference, pages 1-47, 2003.

114
Chapter 6

Appendix

Appendix A: Chapter 2 notes

It is the authors' intent that these tables will provide engineers with context to better
understand the tractors studied in Chapter 2. Table 6.1 are tractors that were evaluated
for their key innovations or historical significance. Figure 6-1 shows images larger images
for these tractors. Table 6.2 shows production tractors with the vehicle layouts discussed in
Chapter 2.

Manufacturer Model Years Engine Mass Price (2017) Units Made

Ivel "tractor" 1902-1920 18hp 1814kg £300 ($43,630) 900
Hart-Parr 30-60 1907-1918 60hp 9120kg $2,600 ($64,030) 3,798
Ford T 1908-1927 22hp 660kg $360 ($5,067) 14,689,525
Avery Farm & City 1909-1915 36hp 2100kg $2,500 ($67,750) N/A
Bull Little Bull 1913-1915 12hp 1315kg $335 ($8,450) 3,800
Wallis Cub 1914-1918 44hp 3855kg $2,480 ($55,495) 660
Moline Universal 1915-1918 27hp 1630kg $1,325 ($18,105) 20,000
Nilson 20-40 1916-1929 40hp 2380kg $925 ($33,230) N/A
Square Turn 18-35 1917-1925 35hp 3538kg $1,875 ($22,900) approx.700
Fordson (Ford) F 1917-1928 20hp 1215kg $395 ($5,760) 755,278
Int. Harvester 15-30 1921-1928 30hp 2653kg $1250 ($17,905) 157,366
Int. Harvester Farmall Reg. 1924-1932 20hp 1655kg $925 ($13,530) 134,647
Allis-Chalmers U 1929-1952 20hp 2086kg $125 ($21,240) 19,009
Ford 9N 1939-1942 20hp 970kg $585 ($10,291) 99,002

Table 6.1: Production data for vehicles in Figure 3 of Chapter 2.

115
 1902 Ivel

-HartN ri3-60
193 B 4 Wallis Cub 1914 Moline Universal

9 oro

Y 19171Ford,!orL-F

1921IRC 15-30 1924IIICVarmit

Figure 6-1: Larger images for some tractors that introduced innovations or have historical
significance as discussed in Chapter 2.

Appendix B: Chapter 3 notes

Appendix B1: Breakdown of costs for farming with bullocks or conven-

tional tractors in Indian small farms

Aspect Variable Tractor Bullock Pair Units

Capital Cost C 200000 80000 INR
Principal on Loan P 20 N/A % of C
Interest on Loan I 16 N/A % of C
Annual operating cost 1 bectare 0 12500 60000 INR./year
Time before replacement T 10 13 years
Resale value when replaced R, 50000 0 INR
Layout Manufacturer Model Years Motor Mass
Twin City 12/20 1919-1926 27hp 2268kg
A Russell Model C 20/40 1919-1924 40hp 3450kg
Huber 30/60 1912-1916 60hp 5000kg
Hart Parr 20-40 1912-1914 40hp 6000kg
B Samson Sieve Grip 1914-1918 25hp 2630kg
Wallis Cub 1913-1917 44hp 3855kg
Bull Little Bull 1913-1915 12hp 1800kg
Case 10/20 1914-1918 20hp 2304kg
Hart-Parr Little Devil 1914-1916 22hp 3015kg
D Common Sense 15/25 1914-1918 25hp 2700kg
Emerson Brantingham Model L 1916-1918 20hp 2500kg
E Gray Tractor Model B 1914-1918 25hp 2500kg
F Joilet Bates Steel Mule F 1915-1937 30hp 2200kg

Model B 12-20 1917-1920 20hp 1500kg

G Beltrail
Tom Thumb 12-20 1917-1920 20hp 1900kg
Joilet Bates Steel Mule C 1911-1919 30hp 2200kg
Yuba 20-35 1911-1916 35hp 3500kg
H Blewett Webfoot 53 1920-1922 53hp 4500kg
Holt 75 1913-1924 75hp 10432kg
I Killen Strait 30hp 1917-1919 30hp 2600kg
J Killen Strait 15-30 1913-1917 30hp 4300kg
Lawter 18/38 1914-1918 38hp 2950kg
K Boring 12/25 1916-1922 25hp 2050kg
Hackney Auto-Plow 1916-1922 36hp 3630kg
S.L. Allen Planet Jr. 1920-1935 2.31hp 250kg
L Moline Universal 1914-1918 27hp 1630kg
Allis-Chalmers 6-12 1919-1926 12hp 1134kg
M Acme 12-24 1918-1919 24hp 1450kg
N Rumely 8-16 1917-1919 16hp 2600kg
0 Victor Victor 1919 34hp 1950kg
P Bean Track-Pull 6/10 1918-1920 10hp 1400kg
Samson Iron Horse D 1918-1923 26hp 850kg
Olmstead Four Wheel Pull 1914-1920 50hp 3000kg
R Fitch Four Drive 20/30 1915-1918 30hp 1360kg
S Buckeye Junior 1912-1915 25hp 2500kg

20-28 1912-1916 30hp 2000kg

T Heer
Nelson 20-28 1917-1924 30hp 2000kg
U Rumely Ideal Pull 1916-1917 16hp 1500kg
V Post 12-20 1918-1920 20hp 1500kg
W John Deere Dain 1918-1919 24hp 2086kg
X Bullock Creeping Grip 1916-1919 20hp 3270kg

Table 6.2: Layouts of Chapter 2 matched to production vehicles.

117
To calculate the costs of ownership for 15 years, the equations below were used with values
from Table 6.3. Yearly maintenance costs were assumed to remain constant through time for

both tractor and animals. At the end of their useful life as draft animals, bullocks cannot

be sold in India [22] but reasonably maintained tractors can be sold for at least 25% of
their original value after 10 years [47]. The Bullkey tractor is assumed to have equivalent
ownership costs to a conventional tractor but with an estimated capital cost of 100,000 INR
and a corresponding resale price of 25,000 INR after 10 years. The capital cost for Bullkey
is a target as mentioned by farmers1461, not a final price.

Total ownership costs for tractor or bullock pair bought up front:

ownership cost total for 15 years = tO + 2C - R

Total ownership costs for tractor financed:

ownership cost total for 15 years= tO + 2C - R + 2(C - P)I

118
Appendix B2: Average drawbar pull results compared to model predictions

Avg. Pull (N) [actual(model)] and Model Error (%) at Tire Slip
Mass (kg) 10% slip 15% slip 20% slip 25% slip

Fr./R pull error pull error pull error pull error

95/96 1162(849) -26.9 1300(1161) -10.7 1472(1391) -5.5 1513(1564) 3.3
95/96 1297(849) -34.5 1333(1161) -12.9 1350(1391) -3 1479(1564) 5.7
95/137 1281(1086) -15.2 1313(1441) 9.7 1478(1697) 14.8 1425(1882) 32
136/96 1360(1096) -19.4 1461(1448) -0.9 1495(1698) 13.6 1612(1878) 16.5
136/96 1324(1096) -17.2 1516(1448) -4.4 1509(1698) 12.5 1616(1878) 16.2

136/137 1436(1326) -7.7 1624(1735) -6.8 1688(2015) 19.4 N/A(2211) N/A

136/137 1526(1326) -13.26 1623(1735) -6.8 1955(2015) 3.1 N/A(2211) N/A

136/153 N/A(1431) N/A 1989(1923) -3.3 2026(2148) 6 2158(2347) 8.8

136/153 1825(1431) -21.5 2009(1923) -4.3 2056(2148) 4.5 2141(2347) 9.6

152/153 1999(1653) -17.3 2056(1968) -4.3 2065(2262) 9.5 2317(2463) 6.3

152/153 1568(1653) 5.4 1694(1968) 16.2 2092(2262) 8.1 2110(2463) 16.7

Table 6.4: Summary of results from field test experiments compared to model predictions.

For reference, 64% drawbar pull to mass ratio for all configuration masses is 191 kg : 1199

N, 232 kg : 1456 N, 273 kg : 1714 N, 289 kg : 1814 N, 305 kg : 1914 N

119
Appendix B3: Common Indian Tractor Sale Prices and Size

Make Model Mass Motor Min. Price

Jyoti Sanedo 450 kg 12 hp 120k INR

Madhav Agro DI 510 515 kg 10 hp 215k INR
Blue Chemp Agr o0 MS-120 520 kg 12 hp 170k INR
VST Shakti MT 180 D JAI 2W 645 kg 19 hp 295k INR
VSt Shakti MT 180D 645 kg 18.5 hp 295k INR
Kubota NeoStar B2741 4WD 650 kg 27 hp 545k INR
VST Shakti MT 224 1D AJAI 4WD 740 kg 22 hp, 356k INR
Captain 120 DI 780 kg 15 hp 250k INR
Mahindra Yuvraj 215 NXT 780 kg 15 hp 200k INR
VST Shakti MT 171 DI Samraat 800 kg 13 hp, 275k INR
Sonalika GT 20 RX 820 kg 20 hp 300k INR
Sonalika GT 22 850 kg 22 hp 343k INR
Swaraj 717 850 kg 15 hp 260k INR

Captain 200 DI 885 kg 17 lip 265k INR

Captain 200 DI 4WD 940 kg 17 hp 310k INR

Farmtrac Atom 26 990 kg 26 hp, 480k INR

John Deere 3028 E 1070 kg 28 hp 565k INR

John Deere 3036 E 1295 kg 36 hp 740k INR

Farmtrac 60 1400 kg 50 hp 630k INR

TAFE 30 DI Orchard Plus 1400 kg 30 hp 420k INR

Swaraj 724 XM Orchard 1430 kg 25 hp, 395k INR

Massey Ferguson 1035 dl Maha Shakti 1700 kg 39 hp 495k INR
Eicher 364 Super DI 1710 kg 32 hp 471k INR

Massey Ferguson 1035 DI 1713 kg 35 hp 490k INR

Massey Ferguson 1134 MAHA SHAKTI 1720 kg 34 hp 470k INR
Eicher 242 1735 kg 25 hp 355k INR

John Deere 5005 1750 kg 33 hp 470k INR

John Deere 5039 D PowerPro 1760 kg 41 hp 570k INR

120
New Holland 3032 1760 35 hp 520k INR

Mahindra 275 ECO 1760 kg 35 lip 455k INR

John Deere 5042 D PowerPro 1810 kg 44 lip 625k INR

Powertrac 439 Plus 1850 kg 41 hp 530k INR
Kubota MU4501 1850 kg 45 hp 715k INR

John Deere 5205 1870 kg 48 hp 690k INR

Swaraj 735 FE 1895 kg 39 lip 550k INR

Powertrac ALT 4000 1900 kg 41 lip 530k INR

Farmtrac Champion 39 1940 39 lip 490k INR

Massey Ferguson 7250 Power 1950 47 lip 620k INR

Massey Ferguson 241 4WD 1950 42 lip 620k INR

John Deere 5050 D - 4WD 1975 50 lip 800k INR

Eicher 371 Super Power 1995 37 lip 475k INR

John Deere 5045 D - 4WD 2010 45 lip 770k INR

Mahindra 475 DI 2019 42 hp 545k INR

Swaraj 855 FE 2020 52 lip 710k INR

Mahindra Yuvo 575 DI 2020 45 lip 628k INR
Mahindra Yuvo 415 DI 2020 40 lip 570k INR

Eicher 380 Super DI 2045 40 lip 530k INR

New Holland 3600 Tx Heritage Edition 2055 47 lip 650k INR

New Holland 3630 TX Plus 2080 55 lip 725k INR

Mahindra 585 DI Power Plus BP 2100 50 lip 600k INR

John Deere 5310 2110 55 lip 789k INR

Digitrac PP 43i 2140 kg 47 lip 585k INR

Powertrac Euro 50 2170 kg 50 lip 610k INR

Swaraj 963 FE 2200 kg 60 lip 790k INR

Kubota MU5501 2200 kg 55 lip 870k INR

Farmtrac 45 EPI Classic Pro 2245 kg 48 lip 590k INR

New Holland Excel 4710 2255 kg 47 lip 660k INR
John Deere 5405 GearPro 2280 63 lip 850k INR

121
 Digitrac PP 46i 2470 kg 50 hp 630k INR
Sonalika WT 60 SIKANDER 2520 kg 60 hp 790k INR
Table 6.5: Common tractors sold in India with their mass, engine power, and lowest typical
sale price. The tractors are sorted by mass. Data collected by authors from online tractor
sale websites [65, 96].

122
Appendix B4: Soil properties and sensors used for tests

Soil Type

Property Weak Strong Actual

n 1.1 0.79 1
Cohesion (kPa) 0.6 20 3.3
Friction angle (deg) 28 18 33.7

k' (kN/m 2 ) 0.990 2354 74.6

k (kN/m 3 ) 1528 -4130 2080
Bulk density (kg/n 3 ) 1310 1580 1557

Table 6.6: Soil properties used to generate plots. Soils properties for limit conditions are
the upper and lower strength limits of soils published soil traction parameter tables [108][9].
Actual soil strength for field tests is from matching soil type, cone penetromenter data, and
soil moisture data from field test to the most appropriate soil parameters in [108].

123
 String Potentiometer (Tool Position)
Sensor[15] CALT CWP-S
String Length 500 mm
Resistance Range 0-500 kOhm

Force (Tool Loads)

Sensor[19] DYLY 104-500
Load Range +500 kg
Sensitivity 2.0 ± 0.05 mV/V
Non-Linearity ±0.03<%F-S
Hysteresis ±0.03<%F.S
Amplifier[93] Tacuna EMBSGB200
Amplification 1&2 0550x, 3 0220x
Magnetic Proximity (Wheel Rotations)

Sensor[6] MAFM1-A0-1H
Switching Freq. 5 kHz
Magnets Neodymium 10x3 mm
Magnets per rotation 10
sensor to magnet 5 mm
Acceleration (Vehicle Motion)

Sensor[2] Adafruit ADXL335

Range ±3 g
Bandwidth 50 Hz
Data Acquisition (All Sensors)

Device[23] DATAQ D12108

Resolution 16 bit

Capture Rate 10 kHz

Table 6.7: Overview of electronics used for data collection.

124
Appendix B5: Data processing from experiments

Processing of the collected sensor data was performed in MATLAB. First, the drawbar force
components were calculated at every instant. To define the drawbar components and their

point of application, soil force D and center-of-pressure position x are calculated from Eqns

6.1, 6.2 and 6.3.

D= (RA - RB) 2 + R2 (6.1)

a = tan-1 ( R R (6.2)
RA - RB

-RA*(d+1+q)+RB* (d+21+q) (6.3)

(RA - RB * sin(@) - Rc * cos(b)

Where, in reference to Figs. 3-12 and 3-3, RA is the tension force on Load Cell 1, RB
is the tension force on Load Cell 2, Rc is the tension force on Load Cell 3, 1 is the vertical
distance from Load Cell 1 to Load Cell 2, q is the vertical vertical distance from Load Cell
2 to the origin of distance x when d = 0, d is the distance the tool jack has been lowered
from its storage position, and y is the angle of attack for the plow.
After the data was loaded into MATLAB, the resulting time-force signals were processed
through a 1 Hz low pass filter, selected at this frequency because 1 second is the time it
takes the tool to travel three characteristic lengths. Then, the distance traveled by all wheels
was calculated by summing the new distance traveled each time a magnet was detected
(27r/10*Reffective) and using linear interpolation to fill in gaps when no magnet was detected.

The three (one per wheel) distance-travelled vectors were then processed through another
1 Hz low pass filter. The highest drive tire slip was selected at each instant, along with
its corresponding drawbar pull. This matrix was then rearranged so that all drawbar pull
instances were assigned to the closest integer slip (i.e. all slip instances between 13.5% and
14.5% were assigned to the 14% slip bin). Finally, the drawbar pull values in each slip bin
were averaged.

125
Appendix C: Chapter 4 notes

Appendix Cl: Tool dimensions table

Plow Rotavator Discs Planter Sprayer Cultivator Trailer

Balancing Outrig. Outrig. Outrig. Outrig. Outrig. Board Outrig.
or Board or Board
Attachment Central Pin 3-Point 3-Point pin or 3-Point Ball
moto frame
9kg (sprayer)
Mass 1kg 120kg 69kg 137kg 11kg (trailer) 21kg 225-425kg
<60kg (water)
0.68im (trailer)
Tool Width 0.20in 1.02n 1.37 1.5n 0.59cm (tank) 0.51m 1.4mn
1.68m (boom)
Net Length 2.17m 3.78mn 2.69m 3.60in 3.61m (trailer) 2.64i 5.4m
__________2.17m (frame)
Speed 1-3km/h 1-4km/h 3-4km/h 3km/h 3-6km/h 2-3km/h 5-20km/h
Depth 5-18cm 10cm 8-14cm 3-6cm N/A 3-10cm N/A

Table 6.8: Basic dimensions for tools used.

Central Attachment
Mass 32 kg
Jack axis to rear axle 0.6 m
Jack axis to centerline 0.32 m
Vertical travel range 0.3 m
Three-Point Hitch
Mass 36 kg
Overall width 0.51 m
Vertical travel range 0.38 m
Lower pins to rear axle 0.45 m
Ball/Pin Hitch
Height 0.3 m
Dist to rear axle 0.35 m

Table 6.9: Basic dimensions for attachment systems.

126
Appendix C2: Breakdown of costs for farming with bullocks or conven-
tional tractors in Indian small farms

Mass of one Bullock 500 kg

Purchase cost of one Bullock 40000 Rs.

Usable animal life 10 years

Lifetime of animal 13 years

Number of Bullocks 2

Daily feed 0.04 kg of food per kg of animal

Cost of feed 6 Rs./kg

Other costs (medical etc.) 3000 Rs. per bull per year
Purchase cost of ride-on bullock cart 30000 Rs.

Bullock cart maintenance 6000 Rs./year

Cart life 10 years

Table 6.10: Breakdown of costs to an Indian farmer for bullock ownership.

Capital Cost 200000Rs./Tractor

Principal on Loan 20%

Interest on Loan 16%

Tenure of Loan 5 years

Resale value after 10 years 400001Rs.

Annual operating cost per acre 5000Rs./year/acre

Tractor rent per hour 800Rs./hr

Hours per acre 20hrs/acre/year%

Table 6.11: Breakdown of costs to an Indian farmer for purchasing (financed and upfront)
or renting a tractor. A farmer may rent a tractor if they do not own one or own one that
is too small for the task. A farmer who owns a tractor can rent it and themselves out to
others for profit.

127
Appendix C3: Questions and booklet for farmer interviews

A) Farm
1) What is a typical row length for you?
2) What is typical row spacing for you?

3) What are your main crops?

4) How would you describe your soil?

B) Demographic
1) What is your experience in agriculture (approximate years, locations)?

2) How would you describe your role in a farm?

C) Farm Tools
1) How do you work the land? (bullocks, tractor number of people, who are the people?)
2) As applicable: How often do you use a farm tractor or bullock? What do you typically
use each for?
3) Do you purchase farm mechanization tools? What do you look for?
4) What difficulties do you have with your current bullocks, tractors or related tools?

Note: Vehicle is referred to as Bullkey in the following questions.

D) On a scale of 1 (least) to 5 (most):

How likely would you be to use Bulkey for the following:

1) for plowing?
2) for secondary tillage?
3) for planting?
4) for cultivating?
5) for spraying?
6) for trailer?

E) Open Questions

128
1) Looking at pics what do you think of the vehicle width? Follow ups: Does a narrower
vehicle facilitate farming operations? If so, for which operations? Which advantages would
you expect?
2) Does the vehicle look light enough for your needs? Follow ups: Would a lighter vehicle
facilitate farming operations? If so, for which operations? Which advantages would you
expect?
3) Are there task you do on your farm that you are not sure if Bullkey could manage?
Follow ups: Are there tasks you are concerned Bullkey would not be able to do? If so, which
operations? Why?
4) Are there farming operationsyou feel Bullkey is better suited to than any existing alter-
natives? Why?
5) What would you like to change about Bullkey? Why?
6) If you had a Bullkey, what operations would you use it for?
7) If Bullkey was available to buy, what, if anything, would you pay for it?

F) Additional conversation points

1) Would you prefer the outrigger arm or balance board? Why?

2) Would you prefer a diesel engine or a petrol engine? Why?
3) Would you use Bullkey as a motorcycle? If so, what is the lowest acceptable top speed.

129
Appendix D: Chapter 5 notes

Introduction

The author's suggested mechanism for attaching a tool near the tractor's center is a modified

linear motion jack, as discussed in Chapter 5. However, if one were interested in achieving
the elliptical tool path motion shown in Fig. 5-3, a different mechanism would be required.

In this appendix, three mechanisms for achieving this motion will be briefly discussed. It
is the author's intention to provide the reader with a starting point for mechanism design
ideation. It is not the author's intention to suggest a specific mechanism as the best solution,
that is left for future work.

Pin joints mechanism for elliptical path

A four bar mechanism is a well understood device for generating elliptical motion. It requires

only pin joints and rigid links, which enables it to be made robust for outdoor applications
without many complications. Using the nomenclature in Fig. 6-8 for a four bar linkage, the
tillage tool's location and orientation can be calculated from Eqn. 6.4.

[cos(Y) [sin(p) l cos(W) [cos(-6) 0l

qI- I-s --t =- (6.4)
Ssin([)cos()J sin(W) sin(--)J [0]
Where q crank length, r is rocker length , s is follower length, t is frame link length, 0
is frame link angle to horizontal, p is crank link angle to horizontal, and p is follower angle
to horizontal

An example implementation of this concept is shown in Fig. 6-9. In that implementation

the drive link CD is located near the output shaft of the engine and low in the motorcycle
frame. This would facilitate using either the engine power or an electric motor (where being
low is beneficial to CG location) to drive the mechanism. The mechanism can also be driven
manually. This can be achieved by extending the rear guiding link BA past point A, turning
it into a lever, or by adding a second mechanism that applies a force directly at one of the
mechanism pivots. The latter would turn the mechanism into a six bar linkage and is shown
in Fig. 6-9.
The mechanism shown in Fig. 6-9 was designed in MATLAB. The design space where
the following equation is true was explored for desirable motion paths using Eqn. 6.4.

130
 1

007m

Im
0.64m

0. 9-1.2m

Can match pulling force of two bulls. Powered inline

wheels deliver more pulling force per mass than tractors.

Same width and shorter length than bull. Comparable

crop clearance.

Half mass of bull. 200kg (base)+ 200kg (optional ballast).

One third the ownership costs (e.g. food, petrol) of

two bulls, and 15 to 25% more income than two bulls.

Figure 6-2: Pages 1 and 2 of booklet used to interview farmers in India.

131
 J

Figure 6-3: Pages 3 and 4 of booklet used to interview farmers in India.

132
 41

0I

Figure 6-4: Pages 5 and 6 of booklet used to interview farmers in India.

133
 8

Figure 6-5: Pages 7 and 8 of booklet used to interview farmers in India.

134
 Eu-

10

Figure 6-6: Pages 9 and 10 of booklet used to interview farmers in India.

135
Balance Board allows to remain upright and narrow at walking speed

11
Baac Board alost remai upih an naro on unvnteri

-Side
Slope Oacle Leaning in Turn
12

Figure 6-7: Pages 11 and 12 of booklet used to interview farmers in India.

136
 'A

q s

C V r8

•.... - . . .. .. .-... Soil Level

. . ..-..
Soil Force u
d

E - - - - - -

-
Figure 6-8: Four bar mechanism with variables used in equations labeled.

Figure 6-9: Examples of implementation of four bar linkage (top) and four bar linkage with
handle for manual control (bottom). More instants in themechanism motion are shown in
Fig. 6-10.

137
From this equation, after setting the crank arm angle y, the angles of the rocker arm (p)
and follower (cp) can be solved for.
The magnitude of t and the angle 0 were set to constants that matched well to viable
attachment points on the current prototype's frame. The lengths of the links q, r, and s
were then explored for arrangements where Eqn. 6.5 is met. Desirable designs will keep
rocker angle y approximately constant while the tool is being lowered into the soil. An
example of a viable motion path is shown in Fig. 6-10.

s+ r - q < t (6.5)

Concepts with other joint and link types

A mechanism that achieves the motion Fig. 5-3 could also use sliding contacts or cable links.
These components provide advantages like facilitating keep a constant tool angle through
the whole motion cycle without being limited to a circular motion path (as would be the
case with a four bar linkage) and decoupling tool depth from the orientation of the rigid
links in the mechanism.
The mechanism shown in Figs. 6-11 and 6-12 has parallel sliding circular rails rigidly
mounted to a structure that connects the outrigger arm to the motorcycle frame. On these
rails a submechanism slides that is effectively a four bar linkage which would otherwise
simply achieve a circular motion while keeping the tool at a constant angle of attack. The
tool still maintains a constant angle but is motion path shape has been favorably altered by
the rails.
The mechanism itself is driven by the light green'slotted link shown in Figs. 6-11 and 6-
12. This drive link does not fully constrain the location of the tool, instead the tool's location
is also determined by the forces applied to it. This degree of freedom allows the drive link to
have a longer lever arm while the tool is on the ground than when the tool is raised. It also

changes the ratio of vertical to downward motion of the tool throughout its stroke, allowing
a shallower insertion angle into the soil.
Another design alternative is shown in Fig. 6-13. This design uses a pin-slot joint and
a cable link. The location of the pin in the slot limits the maximum depth the tool can

reach. In an implementation of this design, the pin should lock at the top of the slot (for

138
 CO

07 P

Figure 6-10: Instants in the mechanism motion. Between instants 5 and 1, the tool is coming
down so the user does not need to apply force at the handle and can let go of it. A foot
support can be placed on the rocker link (shown) to allow the user to press down on the
tool during initial insertion into unusually hard soil.

139
 /"00

Figure 6-11: Views of concept sliding rail mechanism on motorcycle.

140
 four bar linkage sub-mec anism
keeps tool angle constant
( drive link
(rotates C CW)

rails

tool attachment railcarrage

(6) 7

Figure 6-12: Instants in sliding rail concept mechanism motion.

141
tool transportation) until released by the user. The pin should then fall in the slot to a
maximum depth set by the user (not necessarily the bottom of the slot). This maximum
depth could be set obstructing the pin's travel along the slot. The cable link allows the pin
the tool to be laid on the soil before insertion and then the soil to naturally pull the tool
down via its drag forces during forward motion. In other words, the cable limits the tool
maximum depth but not its minimum - similar to a conventional tractor hydraulic system.

pin/~ hancean lever attachment

slot for pin

pm push ing cup

cable
tool attachment
(

Figure 6-13: Instants in the mechanism motion.

142