DESIGN AND DEVELOPMENT OF DOUBLE CHAMBER

CENTRIFUGAL DE-HULLER FOR MILLETS

By
Er. S. GANESAN, M.E (Ag.)
(I.D.No: 10-641-002)

DEPARTMENT OF FOOD AND AGRICULTURAL PROCESS ENGINEERING

AGRICULTURAL ENGINEERING COLLEGE & RESEARCH INSTITUTE
TAMIL NADU AGRICULTURAL UNIVERSITY
COIMBATORE-641 003

2015
 DESIGN AND DEVELOPMENT OF DOUBLE CHAMBER
CENTRIFUGAL DE-HULLER FOR MILLETS

Thesis submitted in partial fulfillment of the requirements for the award of the
Doctor of Philosophy in Agricultural Processing and Food Engineering
to the Tamil Nadu Agricultural University, Coimbatore.

Er. S. GANESAN, M.E (Ag.)

(I.D.No: 10-641-002)

DEPARTMENT OF FOOD AND AGRICULTURAL PROCESS ENGINEERING

AGRICULTURAL ENGINEERING COLLEGE & RESEARCH INSTITUTE
TAMIL NADU AGRICULTURAL UNIVERSITY
COIMBATORE-641 003

2015
Acknowledgement
 ACKNOWLEDGEMENT

I kindly offer my humble salutations at the feet of Murugan, who kindly answered my
prayers and keep on pushing me towards good things and providing the necessary support from various
grounds of life.

I respectfully thank Dr.B.R.Ambedkar, Mr.Periyar and various other unknown faces who are
the reason for what I am now. I thank the millet farmers for providing us the right material for
research work.

I am deeply indebted to my Chairman of the advisory committee Dr.N.Varadharaju, Professor

and Head, Post Harvest Technology Center, for his love and affection, valuable guidance, Consistent
sharing of ideas and knowledge and I wish to specially thank him for providing the opportunity.

I would like to express my gratitude to Dr.R.Kailappan, Retd. Professor, as member of my

advisory committee for his esteemed help and encouragement offered during the entire course of study.

I would like to express my gratitude to Dr.D.Malathi, Professor, Post Harvest Technology

Center, as member of my advisory committee for heresteemed help and encouragement offered during
the entire course of study and I wish to specially thank her for selecting me and providing the
opportunity to conduct research in this area. I also wish to thank IDRC for the monetary support.

I would like to express my gratitude to Dr.A. Nirmalakumari, Professor, as member of my

advisory committee for her esteemed help and encouragement offered during the entire course of study.

I would like to express my gratitude to Dr.P.Rajkumar, Professor, as member of my advisory

committee for his esteemed help and encouragement offered during the entire course of study.

I would like to express my gratitude to Dr.R.Pandiarajan, Professor and Head, Department

of Food and Agricultural Process Engineering as for his support and encouragement during the period
of study.

I owe my indebtness to Dr.K.Thangavel, Dr.R.Visvanathan, Dr.R.Kasthuri, Dr. V. Thirupathi,

Dr.S.Ganapathy, Dr. Z.John Kennady, Dr. P.Vennila, Dr. Amirtham, Professors, Dr.M.Balakrishnan,
Associate professor and Dr. Arulmari, Dr.C.Indhu Rani Assistant professors, Department of Food
and Agricultural Process Engineering and Post-harvest Technology Centre for their periodic
monitoring and whole hearted help to undergo research work.
Abstract
 ABSTRACT

DESIGN AND DEVELOPMENT OF DOUBLE CHAMBER CENTRIFUGAL

DE-HULLER FOR MILLETS

By
Er. S. GANESAN

Degree : Doctor of Philosophy in Agricultural Processing and

Food Engineering

Chairman : Dr. N. VARADHARAJU, Ph.D.,

Professor and Head,
Post Harvest Technology Centre,
Agricultural Engineering College and Research Institute,
Tamilnadu Agricultural University,
Coimbatore – 641 003.

2015

“Millets are one of the oldest foods known to humans & possibly the first cereal
grains to be domesticated”. Millets are classified into major millets and minor millets
based on the usage and size. Pearl millet, Sorghum and Maize are regarded as major
millets while finger millet, foxtail millet, little millet, kodo millet, proso millet and
barnyard millet are categorized as minor millets.

One huge task is to improve the post harvest processing operations that comprises
of threshing, winnowing, destoning, grading, de-hulling, cleaning, grading and packaging.
Machines for processing millets are not readily available. As a reason, millet processing
is done manually with indigenous resources or else they are processed with paddy
processing machines.

Millets are encased by a tough outer aleurone layer. The aluerone layer alongwith
the husk should be removed for futher processing of the grains. The process of removing
the outer husk alongwith the aleurone layer is called de-hulling. De-hulling is done
conventionally by using pestle and mortar with minimal spray of water or by using stone
underrunners. Both these processes are laborious, time consuming and demands lot of
patience. To overcome this situation, millets were de-hulled using paddy de-hullers.
Utilization of paddy de-hullers to de-hull millets may provide a temporary solution but it
is not going to solve problem permanently due to the reason that these machines are
customized only to de-hull paddy and not millets. To sort out this crisis situation, it is
very important to come up with machines that can successfully process millets. Due to
the above mentioned facts a double chamber centrifugal de-huller was designed and
developed to overcome to de-hull millets.

The physical properties of the grains and kernels were studied. It was found that
physical properties like bulk density, true density and porosity decreased with an increase
in moisture content. The frictional properties like static coefficient of friction and
coefficient of internal friction increased with an increase with moisture content. The
aerodynamic property increased with increase in moisture content. Among the basic
machines tested, the centrifugal de-huller gave good results.

Preliminary studies were conducted for three millets at three different moisture
contents viz., 10, 12, 14 percent moisture content (w.b.) using three types of machines
available namely rubber roll de-huller, abrasive de-huller and centrifugal de-huller.
The maximum de-hulling efficiency of 86.75, 81.56 and 87.50 percent was recorded for
little millet, foxtail millet and proso millet with a centrifugal de-huller. The abrasive
machine and rubber roll de-huller gave poor results in terms of broken grain percentage
in the outlet. Three passes were required for all the millets. The capacity of the machine
was decided as 75 kg/h since there was no machine available to process materials in that
range. Based on the results obtained with a centrifugal de-huller and to reduce the
number of passes a double chamber centrifugal de-huller was developed.

A double chamber centrifugal de-huller was designed, developed and evaluated

for its performance with three millets namely little, foxtail and proso millet. The machine
was tested by varying the parameters like moisture content (10, 12 and 14 percent w.b.),
peripheral velocity (24.92, 26.39 and 27.86 m/s), number of vanes (3,4,5 and 6) and
de-hulling chamber surface (Grooved, Angular and Flat) and the de-hulling efficiency
and broken kernel was found out.

The individual components were designed and then they were assembled together.
The major component parts of the de-huller were the impeller, casing and separation
chamber. The impeller parts were fabricated with plates. The impeller consists of crown
plate, base plate and vanes. The casings were moulded with cast iron. A blower was
fabricated and fitted to the machine for separating the grains and husk.

The centrifugal dehuller developed in this study was evaluated for its performance
in terms of de-hulling efficiency and broken grain percentage. The machine was tested by
varying the moisture content of grains (10, 12, 14 percent, w.b.), peripheral speed of the
impeller (24.92, 26.39, 27.86 m/s), number of vanes (3, 4 and 5) and de-hulling chamber
surface (Grooved, Angular and Flat). The de-hulling efficiency and broken grain percentage
was found out.

Maximum hulling efficiency of 93.87, 92.89 and 94.95 percent was recorded for little,
foxtail and proso millet respectively. The statically reports confirmed that all the individual
parameters had a significant effect on the hulling efficiency of all the millets. The broken grain
percentage recorded for the given de-hulling efficiency was 6.04, 5.35 and 5.05 percent
respectively.

The parameters optimized based on the performance results were 10 percent moisture
content, 27.86 m/s peripheral speed, 5 number of vanes and grooved casing combination gave
best results for little, foxtail and proso millet.

The recovery of the kernels by manual and mechanical methods was compared. It was
found that there was maximum recovery of kernels for double chamber centrifugal de-huller
when compared to the conventional abrasive de-huller. A 3 hp three phase motor was used for
the studies. Nutritionally superior quality grains were obtained with a centrifugal de-huller.
The capacity of the machine was 75 kg/h and the cost of the machine is Rs. 85,000/-. The cost
of processing one kilogram of grains was Rs. 2.30/- .
 CONTENTS

Chapter No. Title Page No.

I INTRODUCTION 1-4

II REVIEW OF LITERATURE 5-28

III MATERIALS AND METHODS 29-71

IV RESULTS AND DISCUSSION 72-118

V SUMMARY AND CONCLUSION 119-123

REFERENCES

ANNEXURES
 LIST OF TABLES

Table Title Page

No. No.

1.1 Proximate Composition of Minor Millets in Comparison With 3

Rice And Wheat (Per 100 G)

2.1 Structural Features of Millet Grains 7

2.2 Mineral Composition of Minor Millets (Mg) 8

2.3 Essential Amino Acid Composition (Mg/g) And Chemical Score of 9

Millets

3.1 Specification of Abrasive Dehuller 36

3.1. Specification of Rubber Roll Sheller 37

3.2. Specification of Centrifugal Dehuller 38

3.3 Design Parameters 66

4.1. Effect of Moisture Content on Engineering Properties of Millet 75

Grains

4.2. Effect Of Moisture Content on Engineering Properties of Millet 76

Kernel

4.3. Composition For Little, Foxtail and Proso Millet at 12 Percent 77

Moisture Content (W.B.)

4.4. De-Hulling Efficiency (%) and Broken Kernel (%) of Available 78

Machines

4.5. ANOVA for the Effect of Moisture Content, Peripheral Velocity, 83

Number of Vanes and De-hulling Surface on De-hulling
Efficiency for Little Millet

4.6. ANOVA for the Effect of Moisture Content, Peripheral Velocity, 98

Number of Vanes and De-hulling Surface on Broken Kernel for
Little Millet

4.7. ANOVA for the Effect of Moisture Content, Peripheral Velocity, 102
Number of Vanes and De-hulling Surface on De-hulling
Efficiency of Foxtail Millet
Table Title Page
No. No.

4.8. ANOVA for the Effect of Moisture Content, Peripheral Velocity, 105
Number of Vanes and De-Hulling Surface on Broken Kernel for
Foxtail Millet

4.9. Anova for the Effect of Moisture Content, Peripheral Velocity, 109
Number of Vanes and De-hulling Surface on De-hulling
Efficiency for Proso Millet

4.10. ANOVA for the Effect of Moisture Content, Peripheral Velocity, 112
Number of Vanes and De-hulling Surface on Broken Kernel for
ProsoMillet

4.11. Recovery of Kernel In g Per Kg of Millet at 10% Moisture 115

Content

4.12. Nutrient Availability per 100 g of De-hulled Millets 116

 LIST OF FIGURES

Figure Page
Title
No. No.

3.1 Abrasive Type De-huller 36

3.2 Feed Hopper 40

3.3 Feed Housing 41

3.4. Feed Housing (Side View) 42

3.5. Feed Housing (Front View) 42

3.6. Schematic Diagram Of First De-hulling Chamber 44

3.7. Schematic Diagram Of Second De-hulling Chamber 45

3.8 Crown Plate 47

3.9. Base Plate 48

3.10. Three Vane Impeller 54

3.11. Four Vane Impeller 55

3.12. Five Vane Impeller 56

3.13. Impeller 57

3.14 Separation Chamber 58

3.15. Double Chamber Centrifugal De-huller for Millets 65

4.1 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

82
and Moisture Content on De-Hulling Efficiency for Little Millet

4.2 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

97
and Moisture Content on Broken Kernel for Little Millet

4.3 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

101
and Moisture Content on De-Hulling Efficiency for Foxtail Millet
Figure Page
Title
No. No.

4.4 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

104
and Moisture Content on Broken Kernel for Foxtail Millet

4.5 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

108
and Moisture Content on De-Hulling Efficiency for Proso Millet

4.6 Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface

111
and Moisture Content on Broken Kernel for Proso Millet
 LIST OF PLATES

Plate Page
Title
No. No.

3.1. Abrasive type dehuller 36

3.2. Rubber roll sheller 37

3.3. Single chamber centrifugal dehuller 38

3.4. Feed hopper 41

3.5 Feed housing 42

3.6. Dehulling surfaces 1- grooved casing, 2- angular casing, 3-Flat 43

Casing, 4-Outer casing

3.7. Three vane impeller 53

3.8. Four vane impeller 53

3.9. Five Vane impeller 53

3.10 Separation chamber 59

3.11 Double chamber centrifugal de-huller for millets 64

4.1 Little millet-I- grain, II- kernel and III-husk 99

4.2 Foxtail millet, I- Grain, II-Kernel 106

4.3 Proso millet-I- grain, II- kernel and III-husk 113

 LIST OF ANNEXURES

Annexure
Title
No.

Annexure 1. Effect Of Peripheral Speed on De-Hulling Efficiency, Broken Kernel

Percentage For Little Millet

Annexure 2. Effect Of Peripheral Speed on De-Hulling Efficiency, Broken Kernel

Percentage For Foxtail Millet

Annexure 3. Effect Of Peripheral Speed on De-Hulling Efficiency, Broken Kernel

Percentage For Proso Millet

Annnexure 4. Estimation of Cost of the Double Chamber De-Huller

 LIST OF PUBLICATIONS
Year and
NAAS
S.No Journal No. NAAS NO. TITLE AUTHORS Volume of
Rating
Publication
1 T1946 1975 2.74 Influence of S.Ganesan and Jan, 8-2
moisture content N.Varadharaju Issue, 2015
on physical
properties of
little millet
1 T1947 1975 2.74 Influence of S.Ganesan and Jan, 8-2
moisture content N.Varadharaju Issue, 2015
on physical
properties of
foxtail millet
 ABBREVIATIONS

AOAC – Association of Official Analytical Chemists

Kg – kilogram
h – hour
g – gram
% – percentage
o
C – degree Celsius
Φ – sphericity
mm – millimetre
3
m – cubic metre
2
m – square metre
m – metre
ε – porosity
ρb – bulk density
ρt – true density
θ – angle of repose
r – radius
µs – coefficient of static friction
Fe – frictional force
Ne – normal force
µi – coefficient of internal friction
et al., – and others
Fig. – figure
Figs. – figures
m/s – metre per second
η – Dehulling efficiency
Rs – Rupees
ANOVA – Analysis of variance
cm – centimetre
β – angle
N – Newton
D – diameter
hp – horsepower
w.b. – wet basis
M – moisture content
º – degree
rpm – revolution per minute
s – second(s)
viz. – namely
L – length
T – thickness
V – volume
min – minute(s)

Q – Feed rate

Vg – Volume

ρg – Density of grains

Dcpo – M

– m2

– m/s

Vir – Radial velocity at inlet

Die – Inlet vane edge diameter

Dcp – Crown plate diameter

Dbp – Base plate diameter

Tcp – Crown plate thickness

Tbp – Base plate thickness

Wi – Width of the impeller

Dive – Diameter of inner vane edge

De – Diameter of the eye

Rcv – Radius of crown plate

– Velocity at inner vane edge

Vrive – Radial velocity at inner vane edge

– Degree

– Tangential velocity at inlet

Introduction
 CHAPTER I

INTRODUCTION

“Millets are one of the oldest foods known to humans and possibly the first cereal
grains to be domesticated”. Millets are classified into major millets and minor millets
based on the usage and size. Pearl millet, sorghum and maize are regarded as major
millets while finger millet, foxtail millet, little millet, kodo millet, proso millet and
barnyard millet are categorized as minor millets.

Millets have been main staples of the people of semi-arid tropics of Asia and
Africa for centuries where other crops do not grow well. They have been cultivated since
time immemorial. Essential similarities of these millets are their resilience and ability to
thrive in harsh environments with very little inputs.

Millet grains are considered to be one of the most important sources of dietary
proteins, carbohydrates, vitamins, minerals and fibre for people all over the world.
The nutrient composition of millets compares well with other cereals (Malleshi, 1989).
The tiny "grain" is gluten-free and contains nutrients such as magnesium, calcium,
manganese, tryptophan, phosphorus, fibre and antioxidants. The carbohydrates of millets
consist mainly of starch (amylo pectin) and sustained amount of dietary fibre. Glutelins
form the major protein source indicating that millets are a good source of sulphur amino
acids. The fat content in millet contains good proportions of unsaturated fatty acids,
which are the essential fatty acids, required in our diet. Millets contain a higher
proportion of dietary fibre. There is also a good presence of vitamins such as thiamine,
riboflavin and niacin (Malleshi, 1989).

Finger millet has thirty times more calcium than rice while every other millet has
at least twice the amount of calcium compared to rice. In their iron content, foxtail and
little millet are so rich that rice is nowhere in the race. While most of us seek a
micronutrient such as B-Carotene in pharmaceutical pills and capsules, millets offer it in
abundant quantities. It also has high contents of polyphenols and phytochemicals that are
known to be having anti-fungal and anti-bacterial properties. Millet foods have high
sustaining power and low glycemic index. Diabetic subjects are advised to eat millets
instead of rice. The incidence of duodenal ulcer is practically nil among millet eaters.
Millet foods are considered cool foods because we do not feel thirsty after consumption
of millet diets (Malleshi and Hadimani, 1993). The proximate composition per 100 grams
of minor millets is furnished in the Table 1.1.

A wide range of products can be prepared from millets. Breads, porridges,

steamed and boiled products, fermented foods, drinks and beverages, snack foods and
baby foods, idly and dosa formulations, supplementary foods such as weaning foods,
extruded products, chapathis, rotis, mudde, popped and extruded products, and chapattis
etc., can be prepared from millets and millet based formulations.

Despite possessing all these extraordinary qualities, millets are still facing a
decline in cultivation and consumption of millets. The major reasons for this trend are
change in consumer’s food habits, rapid rate of urbanization, amount of time and energy
required to prepare millet based foods, lack of exclusive harvesting and processing
machines, lack of awareness about the nutritional and socio-economic value of millets,
poor support from policy makers and gaps in research and development in the area of
millet processing (Malleshi and Hadimani, 1993).

Improving the post harvest processing operations of the millet is one of the major
tasks. Machines for processing millets are not readily available. As a reason, millet
processing is done manually with indigenous resources or else they are processed with paddy
processing machines. Manual millet processing is time consuming and laborious, while the
mechanical processing utilizing rice processing machines resulted in poor quality products.
Hence it is important to concentrate on development of machines for millets.

Millets are encased by a tough outer aleurone layer. The aluerone layer along with
the husk should be removed for further processing of the grains. The process of removing
the outer husk along with the aleurone layer is called de-hulling. De-hulling is done
conventionally by using pestle and mortar with minimal spray of water or by using stone
under runners. Usage of paddy de-hullers for millets provide a temporary solution but it
is not solving the problem permanently since the efficiency of de-hulling is less with
more breakage. To sort out this crisis situation, it is very important to come up with
machines that can successfully process millets (Chakraverty et al., 2003).
Table 1.1. Proximate composition of minor millets in comparison with rice and wheat (per 100 g)

Minerals Fibre Carbo- Calcium Phosphorus Thiamine

Name Protein (g) Fat (g)
(g) (g) hydrates (g) (mg) (mg) (mg)

Finger millet 7.3 1.3 2.7 3.6 72.0 344 283 420

Poroso millet 12.5 3.1 1.9 7.2 70.4 14 206 400

Foxtail millet 12.3 4.3 3.3 8.0 60.9 31 290 590

Little millet 7.7 4.7 1.5 7.6 67.0 17 220 300

Kodo millet 8.3 1.4 2.6 9.0 65.9 27 188 330

Barnyard millet 6.2 2.2 4.4 9.8 65.5 11 280 300

Rice (milled) 6.8 0.5 0.6 0.2 78.2 45 160 -

Wheat 11.8 1.5 1.5 1.2 71.2 41 306 -

Source : Malleshi (1989)

 De-hulling of cereals is done by utilizing three principles namely abrasive,
attrition and shear. Abrasive de-hulling involves a compression and rubbing on the
material by passing it between a stone emery roller and a concave sieve. De-hulling by
using rubber roll de-huller is another method which is employed to de-hull the grains by
subjecting them to shear and compression. There are drawbacks in the existing machines
like less hulling efficiency, generation of heat in the product due to friction, thorough
polishing of grains and more breakage. The cost of operation is also on the higher side
for these equipments and it required frequent maintenance. In order to de-hull the millets
more efficiently with less breakage, suitable modifications have to be incorporated in the
existing machine by thorough designing of various components. Hence this study on
“Design and development of double chamber centrifugal de-huller for millets” was taken
up with the following objectives.

1. To study the engineering properties of minor millet grains.

2. To study the performance of the existing de-hullers and select the suitable one
for minor millets.

3. To design and incorporate modifications in the selected de-hulling machine to

improve the efficiency.

4. To evaluate the developed de-huller for its performance to improve the

efficiency.

5. To analyze the nutrient content of de-hulled millets and work out the cost
economics of the machine.
Review of Literature
 CHAPTER II

REVIEW OF LITERATURE

This chapter explicitly reviews the past research work carried out by various
researchers on physical, mechanical and aerodynamic properties of grains. It also reviews
work that covers important unit operations of grain processing like precleaning, grading
and de-hulling of minor millets.

Cereals, pulses, fruits, and vegetables are the important food crops in the world.
They are the bulk sources of calories, proteins and nutrients. These crops, along with
spices and plantation crops, play an important role in the economies of many countries.
To supply adequate quantity of grains and other crops to meet the demand created by the
expanding world population is a major challenge to mankind (Chakraverty et al., 2003).
Cereal grains have been considered as the source of carbohydrates to supply food energy
to the diet. Cereal grains, especially rice and wheat, provide the bulk of energy consumed
on earth (Seetharam et al., 1989).

The physical properties of neem nut were determined in the moisture range of 7.
6% to 21% (w.b). These include physical dimensions , crushing strength , 1000 nut mass ,
relative mass of kernel and shell , angle of repose , porosity , bulk density , particle
density and coefficient of static friction. The length and stem-end diameter of the nut
ranged from 12 to 87 mm to 16 to 20 mm and 6 mm to 86 mm to 8 to 52 mm respectively.
The crushing strength of the nut measured both along the longitudinal axis and a
diametral axis decreased with increase in moisture content . The mass of 1000 nuts, per
cent content of the mass of the kernel in the nut and angle of response of neem nut
increased with increase in moisture content. The porosity , bulk density and particle
density decreased linearly as the moisture content increased . The coefficient of static
friction on various surfaces increased with increase in moisture content. The plywood
surface of fered the maximum friction followed by mild steel, galvanized iron and glass .

Cereals are the most important staple food for most people living in the developed
and developing countries. In the developed countries, 70% of the cereal production is
used as animal feed whereas in developing countries, 68 to 98% of the cereal production
is used for human consumption. The principal cereal crops are wheat, oats, rice, sorghum,
millet and maize (FAO, 1995).

The cereal crops that are grown for their edible fruit are generally called grain, but
botanically referred to as caryopsis. The cereal seed consists of two major components, the
endosperm and embryo or germ. The endosperm encompasses the bulk of the seed and is the
energy source of stored food. An outer wall called the pericarp that develops from the ovary
wall encases the endosperm Rooney et al. (1986) and Sharma et al. (1996).

Millet is a collective term referring the grain of a large number of small-seeded

grasses that are grown as cereal crops for food or feed. Millet is the sixth most important
cereal crop in terms of global production. It is grown in over 40 countries, predominantly
in Africa and Asia. (FAO, 1995).

Millets are warm weather grasses which are often grown in diverse soils, climates
and arid and semi-arid environments. While developing countries in Asia still produce
the majority of the world's millets, Africa is becoming the hub of production. Millet
production in Africa has risen 25 per cent since the early 1970s, and its place in domestic
diets is growing steadily. All other regions of the world, however, have registered declines in
total output (largely due to changes in agricultural policy in China and the former USSR
which resulted in dramatic reductions in areas sown to foxtail and proso millet, respectively),
and even in Africa per capita production has dropped notably (FAO, 1995).

The most important cultivated millet species are: pearl millet also known as
bulrush millet, proso millet also known as common millet; foxtail millet, Japanese
barnyard millet, finger millet also known as birds food millet or African millet, and kodo
millet of India. Other millets include little millet, tef millet and fonio millet. (Sullins and
Rooney, 1971).

The millet crops are strongly associated with tribal agriculture. Despite the
importance of minor millets, the development of technologies for cultivation, utilisation
and processing lag behind that of other major cereals. Minor millets (also referred to as
small millets) have received far less attention than sorghum in terms of cultivation and
utilization . Millets are valuable cereals with an old tradition in food and feed. With new
food applications this coarse cereals can be re-discovered. Addition of food processing
techniques can change the texture and taste of a product. It has a high potential for
improvement of nutrient availability (Seetharam et al., 1989).

2.1. Structural Features of Millet Grains

The basic kernel structure and anatomical components are similar in sorghum and
millets. The principal anatomical components can be distinguished: the pericarp (outer
covering), endosperm (starchy part), and germ (oily part) (Sullins and Rooney, 1971;
Rooney et al., 1986; Sharma et al., 1996). In foxtail, little, finger and proso millets the
pericarp is like a sack, loosely attached to the endosperm at only one point. In these
utricle-type kernels the pericarp easily breaks away to expose the inner endosperm.
The kernels of sorghum and pearl millet are of the caryopsis type, in which the pericarp is
completely fused to the endosperm and requires slightly higher amount of energy to break
the pericarp. The relative distribution of these three main components of the kernel varies
according to the grain’s structure. In pearl millet, the distribution of pericarp, endosperm and
germ are 8.4, 75.0 and 16.5 per cent, respectively (Abdelrahman et al., 1984). The ratio of
endosperm to germ in pearl millet is 4.5:1, while in the sorghum kernel it is 8.4:1. In
common and finger millets, the germ is very small and therefore the endosperm to germ ratio,
11:1 to 12:1, is much higher than in sorghum. Protein in millet varies from 5.6 to
14.8 per cent (Egli et al., 2002). The structural feature of the millets is as follows (Table 2.1)

Table 2.1. Structural features of millet grains

1000 grain
Grain Type Shape Colour
mass* (g)

Foxtail Utricle Elliptical Yellow, pale 1.86

millet yellow

Proso millet Utricle Globose Yellow, green 4.7 - 7.2

Little millet Utricle Elliptical Green, yellow 2.6

Source : FAO (1995); * at 12 per cent moisture content (w.b.)

2.2. Nutritional Characteristics of Minor Millets

Although extensive information is available on proximate composition of millets

and processing (Usha et al., 1996; Sripriya et al., 1997), their phenolic content and
associated antioxidant properties have surprisingly not been investigated to the same
extent. Nutritive potential of millets in terms of protein, carbohydrate and energy values
are comparable with popular cereals like rice, wheat, barley or bajra (Malleshi and
Hadimani, 1993). Barnyard millet is a small grain which is rich in minerals when
compared to wheat and four times richer in fat, seven times richer in minerals and twice
richer in calcium compared to rice (Geervani et al., 1997).

The millets are low in phytic acid and rich in iron and calcium (Mbiti et al.,
2000). Kodo millet and little millet have 37 - 38 per cent of dietary fibre, which is the
highest among the cereals (Mamiro et al., 2001); the fat has higher Poly Unsaturated
Fatty Acids (PUFA) (Malleshi and Hadimani, 1993; Malleshi et al., 1986); the mineral
content is also higher than rice or wheat (Table 2.2), (Rooney et al., 1986). Finger millet
has the highest calcium content of 344 mg/100g as per Gopalan et al. (1989). Thompson
(1993) have found out that the millets also contain phytates, phenols, tannins, trypsin
inhibitory factors, and dietary fibre which act as ‘antinutrients’ by chelating metals or
inhibiting enzymes. It is now established that phytates, phenols and tannins can contribute to
antioxidant activity important in health, aging and metabolic diseases as per Camara et al.
(2003). Table 2.2. and Table 2.3 illustrates the mineral and essential amino acid composition of
minor millets .

Table 2.2. Mineral composition of minor millets (mg/g)

Grain P Mg Ca Fe Zn Cu Mn
Foxtail millet
Whole 422 81 38 5.3 2.9 1.60 0.85
Dehulled 360 68 21 2.8 2.4 1.40 0.60
Little millet
Whole 251 133 12 13.9 3.5 1.60 1.03
Dehulled 220 139 13 9.3 3.7 1.00 0.68
 Grain P Mg Ca Fe Zn Cu Mn
Kodo millet
Whole 215 166 31 3.6 1.5 5.80 2.90
Dehulled 161 82 20 0.5 0.7 1.60 1.10
Common millet
Whole 281 117 23 4.0 2.4 5.80 1.20
Dehulled 156 78 8 0.8 1.4 1.60 0.60
Barnyard millet
Whole 340 82 21 9.2 2.6 1.30 1.33
Dehulled 267 39 28 5.0 3.0 0.60 0.96
Finger 320 137 398 3.9 2.3 0.47 5.49
millet
a
Expressed on a dry-weight basis. Source : Sankara Rao and Deosthale (1980)

Table 2.3. Essential amino acid composition (mg/g) and chemical score of millets

Tryptophan
Methionine

Threonine
Isoleucine

Chemical
Tyrosine
lalanina
Leucine

Cystine

Pheny-
Lysine

Valine

score
Grain

Foxtail
475 1044 138 175 - 419 - 194 61 431 41
millet
Little
416 679 114 142 - 297 - 212 35 379 33
millet
Kodo
188 419 188 94 - 375 213 194 38 238 55
millet
Common
405 762 189 160 - 307 - 147 49 407 56
millet
Barnyard
288 725 106 133 175 362 150 231 63 388 31
millet
Finger
275 594 181 194 163 325 - 263 191 413 42
millet

Source : (Hulse et al.,1980) and (Burton et al., 1972).

 Physical and mechanical properties of mature okra (Hibiscus esculentus L.) seeds
from Ahvaz in Iran were evaluated by S.P. Singh et al. in 2012. The physical and
mechanical properties were evaluated at four moisture content levels of 7.1, 10, 15 and
20% dry basis (d.b). In this moisture range, seed length, width, thickness, geometric
diameter, mass of 1000 seeds increased from 5.096 to 5.677 mm, 4.476 to 4.878 mm,
4.239 to 4.608 mm, 4.585 to 5.035 mm and 56.615 to 65.779 g, respectively. The angle
of repose, volume, surface area and sphericity increased from 21.2 to 24.3°, 5.367 to
6.396 m3, 66.18 to 79.64 m2 and 90.07 to 0.92, respectively. The true density, bulk
density and porosity decreased from 1096.34 to 1002.16 kg/m3, 627.4 to 576.2 kg/m3 and
41.1 to 38.01%, respectively. The rupture force and coefficient of static friction on
aluminum, rubber, plywood, iron and galvanized iron sheets increased with increasing
moisture content.

2.3. Engineering Properties of Food Grains

The properties of millets are essential for the design of equipments used for
handling, storing and processing grains. They affect the conveying characteristics of solid
materials by air or water (Sahay and Singh, 2010).

2.4. Physical Properties of Food Grains

The properties of various grains and seeds have already been determined by the
earlier researchers for various agricultural materials like pigeon pea (Shepherd and
Bhardwaj,1986), green gram (Dutta et al., 1988), pumpkin seeds (Joshi et al., 1993),
cumin seed (Singh and Goswami, 1996), soybean (Deshpande et al., 1993), karingda
seeds (Suthar and Das, 1996), cashew nuts (Oloso and Clarke, 1993), lentil seeds (Carman,
1996), guna seeds (Aviara, et al., 1999), coffee (Chandrasekar and Viswanathan, 1999),
white lupin (Ogut, 1998), neem nuts (Viswanathan et al., 1996), pistachios (Hsu et al., 1991),
millets (Balasubramaniam and Viswanathan, 2010), pomegranate seeds (kingsly et al.,2006),
Chick Pea seeds (Konak et al., 2002), legumes (Laskowski et al., 1997), Melon seeds
(Makanjuola, 1972), Wheat and Corn (Nelson, 1980), green gram(Nimkar et al., 2001).
2.4.1. Geometry and dimensions

The geometry and design are very important parameters for desiging of
equipments for separation and handling. According to Mohsenin (1970), the sphericity is
given be equation 2.1.

(LWT )3
1

φ= …..(2.1)
L

where,

φ - Sphericity, dimensionless

L - Length of grain, mm

W - Width of grain, mm

T - Thickness of grain, mm

Geometric mean diameter is given by Sreenarayanan et al. (1988) and Sharma et al.
(1985) as

Dg = (LWT )3
1
…..(2.2)

Dg
φ= …..(2.3)
L

where,

Dg - Geometric mean diameter of grain, mm

2.4.2. Porosity, bulk density and true density

Bulk density, true density and porosity are the most important parameters for
designing agricultural equipments. The bulk density, true density and porosity are related
as in the equation 2.4. According to Mohsenin (1970) the porosity of grains and cereals
are given by

 (ρ g − ρ b )100 
ε =  …..(2.4)
 ρg 
where,

ε - Porosity, per cent

ρb - Bulk density of grain, kg/m3

ρg - True density of grain, kg/m3

2.4.3. Thousand grain mass

Arora et al. (1999) measured one thousand grain mass for wheat and corn using a
digital balance with a resolution of 1 mg and observed that the grain weight varies
linearly with moisture content for the study carried in the moisture content between
8.24 and 27.07 per cent (db).

2.4.4. Angle of repose

The angle of repose, was determined by Janusz Laskowski et al. (2005) using the
following equation 2.5.

 2h 
ϕ = tan −1   …..(2.5)
 d 

where,

ϕ - Angle of repose, degree

d - Diameter of flat steel plate, mm

h - Height of the food powder heap formed over the flat steel plate, mm

2.5. Frictional Properties of Food Grains

The following section enumerates the frictional properties of the grains viz.,
internal friction and static friction which are very important in designing equipments in
which the material is in direct or indirect contact with the equipment parts.

Many researchers have observed an increase in coefficient of static and dynamic

friction for various food grains with an increase in moisture content. The effect of
moisture content was found to be pronounced on the coefficients of static and dynamic
friction for sorghum grains (Stewart et al., 1969); barley, oats, shelled corn, soybean and
wheat (Brubaker and Pos, 1965); beans and peanut (Chung and Verma, 1989); pigeon pea
(Phirke et al., 1996); soybean, red kidney beans and unshelled peanuts (Tsang-Mui-
Chung et al., 1984); sunflower seed and kernel (Gupta and Das, 1998).

Thompson and Ross (1983) measured friction of wheat against galvanized steel
and found the coefficient of friction versus moisture content characteristic with a
maximum for 20 per cent of grain moisture content. These authors suggested that at moisture
content from 16 per cent to 20 per cent kernels became soft and deformation around the
asperities took place that generated stronger bonds than in the case of hard, dry grains.

Tsang-Mui-Chang et al. (1984) developed a friction device (later modified by

Chung and Verma, 1989) to determine the coefficient of dynamic friction for soybean
and corn. Phirke et al. (1996) reported that the dynamic coefficient of friction at higher
speed was further decreased with an increase in speed for pigeon pea. Soybeans had
significantly higher angles of internal friction than coated cottonseeds and shelled corn.
Murthy and Bhattacharya (1998) found that the angle of repose of black pepper increased
while flowability decreased at higher moisture contents, especially above 14 per cent. Oloso
et al. (1993) have reported the frictional and strength characteristics to weaken with increase in
moisture content.

2.6. Aerodynamic Properties of Food Grains

Shellard and MacMillon (1978) reported that the most fundamental forces
involved in the separation of wheat grains from chaff and straw are the terminal velocity
and the aerodynamic drag. Gorial and O’Callaghan (1990) examined the feasibility of
separation of good grains from other grain materials in horizontal air stream and
concluded that it is possible to separate 95 per cent of good grain from materials of other
grain at an air velocity of 11 m/s in a duct of length of 160 cm.

The terminal velocities of seeds at different moisture contents were measured

using a cylindrical air column (Joshi et al., 1993). For each experiment, a sample was
dropped into the air stream from the top of the air column, up which air was blown to
suspend the material in the air stream. The air velocity near the location of the seed
suspension was measured by a hot wire anemometer having a least count of 0.01 m/s.
It was found that the terminal velocity of pumpkin seeds increased from 4.7 to 6.5 m/s
due to an increase in moisture content from 4 to 40 per cent (d.b.).

2.7. Processing of Minor Millets

The key step between harvest of cereals and finished, processed cereal-based foods is
the milling operation. The history of milling operations is very rich and as old as human
civilization. Historians have considered the level of milling technology as an indicator of the
development of a given civilization at that point in time. (Chakraverty et al., 2003)

Although the milling tools and the processes have changed radically, the basic
purpose of milling remains essentially to improve palatability and digestibility of the
grain. Different cereal grains pose unique challenges to the miller and each possess
unique quality attributes in meeting requirements to produce widely differing end
products. Thus, the milling processes for each of these ranges are different from each
other. Nevertheless, there are certain processing elements that are common to all, and
these follow:

1. All the incoming grain is cleaned, inspected, and stored according to quality
before milling.

2. Preparation of grain by tempering or conditioning to facilitate removal of hulls,

bran, and germ from the endosperm.

3. If it is received unhulled, sometimes this preparation of the grain before further

milling or processing may require de-hulling.

4. Careful removal of the hulls, bran, or germ from the endosperm in the actual
milling process, either by de-hulling or pearling, and polishing the grain (oats,
barley, and rice), or by a series of gradual grinding and sifting to release the pure
endosperm in either granular form (corn and durum wheat) or fine powder form
as in meal and flour (rye and wheat).

5. Physical or chemical treatment to enhance product quality.

6. Packaging, storage, and handling of finished products. (Chakraverthy et al.,

2003).
2.7.1. De-hulling

De-hulling is the removal of the outer pericarp and testa (hull) during rocessing of
cereal grains, grain legumes, nuts and oilseeds (Omobuwajo et al., 1999). Usually
de-hulling of millet is done manually in several parts of the country in the absence of a
suitable device ( Singh et al., 2011).

It is unusual, in any human society, for cereals or coarse cereals to be eaten as

uncooked whole seeds (Hulse et al., 1980). Processing involves the partial separation and/or
modification of the three major constituents of the millets - the germ, the starch-containing
endosperm and the protective pericarp. In general, the first operation in processing cereal
or coarse cereal is usually the separation of offal (the portion not normally used for
human consumption) from the edible portion. The offal consists of the pericarp and
sometimes the germ and removal of the offal is referred as decortication or de-hulling
(Bassey and Schimdt, 1989).

2.7.1.1. Manual Pounding

Traditionally the husk portion is separated from the edible portion by moistening
the grain with 10% water to facilitate removal of the fibrous bran and separation of the
germ and the endosperm. (Perten, 1983). The traditional de-hulling is laborious and time
consuming. The drudgery involved in manual de-hulling can be overcome by developing
suitable machines.

2.7.1.2. Mechanical de-hulling

A machine comprising a roller that cracks open the hull, an oscillating cam
follower that removes the cracked hull through repeated shearing against a stationary
wall, and an aspiration unit which sifts the hull from the endosperm was developed for
de-hulling breadfruit seeds and tested for its performance. The optimised parameters were
a throughput of 64 kg/hr, yield of 75%, de-hulling efficiency of 85% and breakage of 1%
(Omobuwajo et al., 1999).

While there are many machines available for processing cereals, there is unfortunately
no well-proven industrial process available to satisfy entirely for making white products
from coloured minor millets. Grain, which should be fairly dry, is crushed and pulverized
by the backward and forward movement of the hand-held stone on the lower stone.
Generally, women do this unpleasant, laborious and inefficient hard work. Perten, (1983)
reported that a woman working hard with a pestle and mortar can decorticate 1.5 kg per
hour, providing a non-uniform poor keeping quality products. Dry, moistened or wet
grain is normally pounded with a wooden pestle in a wooden or stone mortar. Grain was
moistened by adding about 10 per cent water to loosen the fibrous bran, and also to
separate the germ and the endosperm, if desired. Although this practice produces slightly moist
flour, many people temper the grain in this way before they pound it. Decortication is
sometimes accomplished by using rice de-hullers or other abrasive de-hullers. Millets would
probably be more widely used if processing were improved and if sufficient good-quality flour
were made available to meet the demand (Eastman, 1980).

Technically, there are three types of de-hulling or decortication that can be

employed to minor millets:

Attrition usually means particles getting smaller due to their corners or surface
irregularities being knocked off. Attrition is a serious, yet little understood problem in
handling of food materials, which may be considered responsible for economical losses
in the food industry.

Bassey and Schmidt (1989) studied the roller mills in Africa and explained that
the mills consisted of rollers made of stone or stone coated with emery, where the grains
are rubbed by the rollers in the de-hulling chamber to remove the husk from the grains.

The collective experiences to date led to the conclusion that de-hullers can have a
significant role in the food systems of sorghum and millet in sub-saharan Africa. Many
factors define the extent of that role, and many are under the control of those who
determine national policies on food, agricultural production, and household food security.
Abrasive-disk de-hullers are strong contenders as solutions for problems where:

• Women and children daily devote much labour-intensive time to manually

processing their grain before it can be consumed;

• Food preferences favour those cereal grains that do not grow well in the dry areas,
and where the absence of appropriate processing machinery contributes to an
impoverishment of indigenous dryland agriculture; and
 • Antinutritional components, polyphenols, in the grain have not been fully
removable with traditional processing techniques; this has probably resulted in a
lowered nutritional status, especially for the segment of the population that is
most at risk.

• Evidence from applied research indicated that de-hullers are technically capable
of contributing to the solution of these three problems (Bautista et al., 2004).

Abrasiveness of bulk solids, i.e. their ability to abrade or wear surfaces with
which they come into contact is considered a property closely related to the hardness of
the material. The hardness of powders or granules is defined, in direct analogy with the
definition of hardness of solid materials, as the degree of resistance of the surface of a
particle to penetration by another body. It can be implied from the relative hardness of
the particles and the surface with which they are in contact, using Morhs’ hardness scale.

Munck et al. (1982) described a new industrial milling process developed in

Denmark, which does not involve abrasive milling. Decortication is achieved by a steel
rotor rotating the grain mass within a generally cylindrical chamber. When the grain is
properly tempered, the pericarp is rubbed off by the movement of one seed against
another. However, when the grain is too dry, abrasion of the internal components of the
mill becomes severe.

Lorenz (1979 and 1983) observed that the phytate content of common millet
varieties ranged from 170 to 470 mg per 100 g of whole grain, and de-hulling resulted in
a reduction of 27 to 53 per cent in phytate content. On de-hulling, phytin phosphorus
decreased by 12 per cent in common millet, 39 per cent in little millet, 25 per cent in
kodo millet and 23 per cent in barnyard millet (Sankara Rao and Deosthale, 1980).

De-hulling can remove 40 to 50 per cent of both phytate and total phosphorus.
Bioavailability of iron in sorghum for human subjects was found to be affected more by phytin
phosphorus than by tannin content of the grains (Radhakrishnan and Sivaprasad, 1980).
On pearling of sorghum grain, a significant increase in ionizable iron and soluble zinc
content indicated are improved bioavailability of these two micronutrients, which was
attributed partially to the removal of phytate, fibre and tannin along with the bran portion
during pearling (Sankara Rao and Deosthale, 1980).
 Ratio of entry to exit diameter, rotational speed, entry and exit axial breadths,
blade angles, number of blades and shape of the blades are the major areas of focus that
plays a vital role in the performance of the centrifugal pumps ( Bruno ECK, 1973).

Melon seeds were shelled in a rotating impeller by Okokon et al. (2010) The
seeds exit the impeller and impact a cylindrical ring shelling the seeds. Because of the
force of impact, some of the seeds are broken, which deteriorate in storage and make low
market value. An analytical method was used to determine the factors affecting the
impact force on the ring. Experimental compression tests were carried out to determine
the static forces to break melon seeds. Some seeds were also shelled with an experimental
shelling machine and the number of broken seeds was counted.

The analysis showed that the factors affecting the impact force were impeller
speed, seed cross- section area at impact and mass ratio. The mean forces to break melon
seeds were 13.14 × 10-3, 19.62 × 10-3 and 19.55 ×10-3 N for orientations breadth wise,
lengthwise with tip up and lengthwise with tip down, respectively.

De-hulling is the removal of the outer pericarp and testa (hull) during processing
of cereal grains, grain legumes, nuts and oilseeds (Omobuwajo et al., 1999). Usually
de-hulling of millet is done manually in several parts of the country in the absence of a
suitable device ( Singh et al., 2011).

It is unusual, in any human society, for cereals or coarse cereals to be eaten as

uncooked whole seeds (Hulse et al., 1980). Processing involves the partial separation
and/or modification of the three major constituents of the millets - the germ, the
starch-containing endosperm and the protective pericarp. In general, the first operation in
processing cereal or coarse cereal is usually the separation of offal (the portion not
normally used for human consumption) from the edible portion. The offal consists of the
pericarp and sometimes the germ and removal of the offal is referred as decortication or
de-hulling (Bassey and Schimdt, 1989).

Douglas C. Doehlert and Michael S. McMullen (2001) conducted studies on two

mechanical methods of oat dehulling i.e. compressed air dehulling and impact dehulling, and
found to produce reliable results. They reported that the results were strongly influenced by
dehulling conditions. Optimal dehulling conditions represented compromises between
unfavourable extremes. Correct aspiration strength was critical to accurate groat percentage
determination. The authors have also mentioned that a secondary aspiration was highly
desirable after compressed-air dehulling to remove hulls remaining with the groats after
dehulling. Also, increased mechanical stress on oats exerted either by the number of
passes through the impact dehuller, or by the air pressure in the compressed-air dehuller,
resulted in higher dehulling efficiency, but increased groat breakage as well. Dehulling
efficiency decreased as moisture increased from 7.5 to 15%, but increased as moisture
was further increased to 30%. In contrast, groat breakage with impact dehulling
decreased as moisture increased from 7.5 to 30%. A new equation for groat percentage
calculation was introduced where the mass of hulled oats remaining after dehulling was
subtracted from the mass of the original oat sample, so that poor dehulling efficiency
does not influence the groat percentage.

Amuthan et al. (2001) had reported that they had made modification to a
centrifugal dehuller that was already available for paddy. It was found that the paddy
desheller can be successfully used for deshelling sunflower with minor changes. The
performance of the machine in terms of percent shelling efficiency was found to be
maximum at 87.72% with four vanes, 8 cm radius of curvature of the vanes at 2600
m/min speed at 6.5% moisture content (w.b.) of the seed.

Singh et al. (2010) conducted studies on de-hulling of barnyard millet and found
that de-hulling process based on impact and shear principle was found to be successful
for de-hulling of barnyard millet grains and found that the grains obtained were of better
quality. They also had reported that “an impeller consisting of five impellers in series was
used commercially” for completely de-hulling.

Singh et al. (2011) developed a 40-50 kg/h capacity centrifugal dehuller driven by
a 3.7 PS electric motor and optimized for process and machine parameters. The machine
consisting of 9 canvas strips over periphery of the impeller and over hanging 3 mm width
of canvas strip) run at a speed of 8.6 m/s peripheral speed with 5 passes gave an
efficiency of 88.3±2.8% for grains at a moisture content of 8.4% d.b.

The de-hulling efficiency and broken kernel percentage was calculated using
equation 2.6 and 2.7.
 ….(2.6)

Where,

η = De-hulling efficiency, %

Mu= Mass of the unhulled grain, g

Mb= Mass of broken grain, g

Mt = Total mass of the grains taken for de-hulling, g

......(2.7)

Rebaccca Chin et al. (2012) created portable and easy to use mill for little millet
and found that the efficiency of the designed rubber roller and centrifugal mill
combination to be 96.52% with only 1.7% broken grains, with the RPM for the motor and
one rubber roller at 1100 and another rubber roller of RPM 1640 with a 0.55mm spacing
between the rubber rollers. The achieved efficiency stated above satisfies the criteria of
achieving 95% milling efficiency with less than 2% broken grains. The mill was portable
since it was built as compactly as possible around the parameters of the rotor and rubber
rollers, each measuring less than 8 inches in diameter. This mill was easy to use since it
employed an on/off switch to perform the milling and also used a simple plate that was
manually controlled to release the grain. This milling combination requires less human
labour, therefore it made it easier for the mill operator to use. The level of maintenance
required on the mill was minimal since the parts were properly enclosed. Even though
there were many moving parts, they were expected to have a long lifespan since the
chosen materials were strong metals and the rubber rollers used were designed to
withstand high forces on pavement which are greater than the force exerted by the millet.
The environmental conditions to the mill operator was favourable since the rubber rollers
and the centrifugal mill are each contained within a casing, therefore any dust that is
created from the milling process is contained. The milling efficiency was greater than
95% and 100 % separation was achieved. There was no need for pre-treatments since
these results achieve the stated goals. The milling capacity of 20 kg per hour is not as
high as the goal of 100 kg per hour, but since the machine was a prototype of the mill and
to be used for demonstration purposes and proof of concepts, this capacity was found to
be adequate. In order to increase milling capacity, longer rubber rollers could be used
which would allow for more contact area available for the grains. The estimated cost of
the materials of this mill is $581 CND if constructed in India. Energy inputs were kept to
a minimum since only one motor was used for both the rubber roller milling and the
centrifugal fan operations. The rubber roller and centrifugal mill combination was
designed, built, prototyped, tested and optimized to achieve the goals of the project. Since
it resulted in a high milling efficiency and separation of the grains and hulls, this mill can
be used to further the adoption of millet by rural farmers in India.

A refined design of centrifugal Sheller for muskmelon seed decortications was

developed and evaluated by Singh Ranjeet and Mangaraj Sukdev (2013). The developed
machine has a capacity of 5 kg/h which operates on 1400 r.p.m emery roll speed and seed
moisture content of 25 % (w.b), with this optimum condition, the shelling efficiency,
seed damage and capacity was 51.0 %, 32 % and 5.0 kg/h, respectively. The decortication

cost per kg of muskmelon seed using developed machine was Rs. 1.20 as compared

with manual methods of Rs.100.00.

The centrifugal force was found out using the equation 2.8

Cf = mv2/r ……………… . (2.8)

Cf - Centrifugal force, N. m - Unit mass of muskmelon, kg

V - Velocity, m/s. r -Radius, m

Estimation of crude fibre

The method of Maynard (1970) was followed for the estimation of crude fibre.
Two gram of ground material was extracted with petroleum ether to remove fat (initial
boiling temperature 35-38ºC and final temperature 52 º C). The extracted sample was
dried and boiled with 200 ml of sulphuric acid (Appendix I) for 30 min. with bumping
chips. Then it was filtered through muslin cloth and washed with boiling water until
washings are no longer acidic.
 Then the filtrate was boiled with 200 ml of sodium hydroxide solution (Appendix I)
for 30 min. and filtered through muslin cloth again and washed with 25 ml of boiling
1.25 per cent H2SO4, three 50 ml portions of water and 25 ml alcohol.

The residues were removed and transferred to ashing dish (preweighed, W1) and
dried for 2 h at 130±2ºC. Then the dish was cooled in a dessicator and weighed (W2).
Finally the dish was ignited for 30 min. at 600±15ºC and cooled in a dessicator and
reweighed (W3).

(W2-W1) – (W3-W1)
Per cent crude fibre in ground sample = -------------------------- x 100 …….(2.9)
Weight of the sample
Estimation of ash content

The ash content was determined by the method described by FAO (1986). Three
gram of defatted sample was weighed into a crucible of known weight. It was placed in
muffle furnace by setting the temperature to 600ºC and incinerated for 2 h. The crucible
along with ash was weighed, to prevent moisture absorption. Its weight is recorded and
the process of incineration was repeated until a constant weight was obtained.

(B-C)
Per cent crude fibre in ground sample = --------- x 100 …….(2.10)
A
Where,

A = Weight of the sample in grams

B = Weight of the crucible + Constant weight of ash in grams

C = Weight of empty crucible in grams

(B-C) = Weight of ash in grams

Estimation of Protein

The protein content in the sample was estimated using Micro-Kjeldahl method
(Sadasivam and Manickam, 1992). Protein content was estimated by Kjeladahl method
using a laboratory kjel plus equipment (Pelican equipments, model – REC 22238-A2,
Chennai) to estimate the protein content.

The kjeldahl method was carried out in three steps.

Digestion : The decomposition of nitrogen in organic samples takes place utilizing a
concentrated acid solution. This is accomplished by boiling a homogenous sample in
concentrated sulphuric acid and digestion catalyst. The end result is ammonium sulphate
solution.

Distillation: Distillation involves adding base to the acid digestion mixture to convert
ammonium to ammonia. This is followed by boiling. Finally ammonia gas is condensed
and trapped in a receiving solution (Boric acid).

Titration: The quantity of ammonia ions in the receiving solution was quantified by
titrating against concentrated hydrochloric acid(Hcl). Then the percentage of Nitrogen
can be calculated.

Preparation of Sample

About 2 mg of the ground sample was transferred into the digestion tube. To this
10 ml of concentrated sulphuric acid and 3 g of digestion activator (mixture of potassium
sulphate or sodium sulphate with copper sulphate in 5:1 ratio) were added.

The content was left undisturbed over night to enhance digestion and then the
material was digested in the digestion blocks at a temperature of 420º C. The sample was
distilled by passing steam; ammonia liberated due to addition of alkali which trapped in
4 per cent biric acid. The boric acid in ammonia was titrated against the standardized
Hydrochloric acid with mixed indicator (0.1 g bromocrescol and 0.1 g methyl red in
400 ml of 95 per cent ethanol). The titre value was noted down at the end point when
colour turns from green to light pink. The per cent nitrogen was calculated using the titre
value. The protein content was obtained by multiplying the per cent of nitrogen with the
factor 6.25.

(V1 – V2) x Normality x 14.01

Nitrogen (%) = --------------------------------------- x 100 …….(2.11)
W x 1000
Where,

V1 = Volume of sulphuric acid, ml

V2 = Volume of blank, ml
 W = Weight of the sample, g

N = Nitrogen, per cent

Protein content = N x 6.25 per cent.

Determination of reducing sugars

Reducing sugars was determined by the Nelson – Somogyi method (Somogyi,

1952). Hundred milligram of the sample was weighed and the sugars were extracted with
hot 80 per cent ethanol (5 ml each time). The supernatant was collected and evaporated
by keeping it on a water bath at 80 º C and 10 ml of water was added to dissolve the
sugar. Aliquots of 0.2 or 0.4 ml were pipetted out in separate test tubes and volume was
made upto 2 ml. Working standard of 0.2, 0.4, 0.6, 0.8 and 1 ml of distilled water. To these
standards, blanks and samples, 1 ml of alkaline copper tartarate reagent (Appendix I) was
added and placed in boiling water for 10 min. Then the tubes were cooled and 1 ml of
arsenomolybolic acid reagent (Appendix I) was added to all the tubes and the volume was
made up to 10 ml and the absorbance of blue colour was read at 620 nm after 10 min.

Absorbance corresponds to 0.2 ml of test = x mg of glucose

x
10 ml contains = ------ x 10 mg of glucose …… (2.12)
0.2
Estimation of iron

Iron content of grains were estimated by the method as described by Wong

(1928). The ash solution of the sample prepared by dry ashing was used for colour
development. Five millilitre of the sample was taken and 0.5 ml concentrated sulphuric
acid, 1.0 ml potassium persulphate and 2.0 ml thiocyanate were added and the colour
intensity was measured at 480 nm. Blank was prepared with 5.0 ml distilled water and
reagents were added as per the sample excluding ash solution. Standard iron solution was
prepared by taking 1.0 ml, 2.0 ml, 3.0 ml, 4.0 ml and 5.0 ml iron standard and made upto
5.0 ml and the reagents were added as aid for the sample.

OD of the sample x 0.1 x Total volume of ash solution

Iron (mg/100 g) = --------------------------------------------------------------------- -x 100.. (2.13)
OD of standard x 5 x Weight of the sample taken for ashing
Estimation of Calcium

The method of Ward and Johnston (1962) was followed for the estimation of
calcium. An aliquot (20 to 100 ml) of the ash solution obtained by drying by dry ashing
was pipetted to a 250 ml beaker and 20 to 25 ml of the distilled water was to it.
Ten millimetre of saturated ammonium oxalate solution and 2 drops of methyl red
indicator were added. The solution was made slightly alkaline by the addition of diluted
ammonia and then slightly acidic with a few drops of acetic acid until the colour faint
pink (pH 5.0). Then it was heated to the boiling point and allowed to stand at room
temperature for at least 4 h or preferably overnight. Then, the solution was filtered
through Whatman No. 42 filter paper and washed with water, till the filtrate was oxalate
free. (since Hcl was for preparing the original solution, it is convenient to test for the
absence of chloride using AgNO3). Then the point of the filter paper was broken with
platinum wire or pointed glass rod and the precipitate was first washed using hot dilute
H2SO4 (1+4) from wash bottle into the beaker in which the calcium was precipitated.
After washing it with hot water, titrated while still hot (temperature 70 to 80 º C) with
0.01 N KMno4 till the first permanent pink colour appeared.

Titre x 0.2 x Total volume of ash solution

Calcium (mg/100 g) = --------------------------------------------------------------- ----------x 100
Volume taken for estimation x Weight of the sample
taken for ashing
……… (2.14)
Determination of pH

The pH was determined by using a digital pH meter (Systronics µpH system 361,
Ahmedabad, India). The pH meter was standardized with double distilled water of pH 7.0
and buffers at pH 4.0 and 9.2. After standardization, pH of different treatments samples
was measured.

Estimation of titrability acidity

Titratable acidity was estimated based on the method described by Association of

Official Analytical Chemist (A.O.A.C, 1991) and expressed as g 1-1 tartaric acid.

One ml of phenolphthalein indicator was added to 200 ml of hot boiled deionized

water. Five ml of porridge sample was transferred volumetrically to this and titrated
against 0.05 N standard NaOH. The end point was the appearance of pink, which
coincided with pH 8.2. The titrable acidity was calculated as follows and expressed in g
1-1 of tartaric acid.

(Volume of base) x (Normality base) x 0.075 x 1000

Titrable acidity
(g 1-1 of tartaric acid) = --------------------------------------------------------------------- (2.15)
Volume of sample (ml)
Where,

Volume of base = Volume of NaOH used in ml.

Normality base = Normality of NaOH.

REAGENTS USED

a. Chromic acid solution

Potassium dichromate - 100 g
Conc. Sulphuric acid - 500 ml
Water - 1000 ml

b. Gram Stain Solutions: -

1. Crystal violet solution -
Crystal violet - 10 ml
Ammonium oxalate - 4g
Ethanol - 100 ml
Distilled water - 400 ml

2. Iodine solution -
Iodine - 1g
Potassium iodide - 2g
Ethanol - 25 ml
Distilled water - 100 ml
 3. Alcohol
Distilled water - 5 ml
Ethanol - 95 ml

4. Counter stain
2.5% safranin in ethanol - 10 ml
- 100 ml
c) Nitrate reduction test
1. 5 N Acetic acid - 294 ml of glacial acetic acid + 706
ml. of distilled water
2. Alpha naphthylamine reagent - 5.0 ml 2-naphthalyamine + 1000
ml of 5 N acetic acid.
3. Sulfanalic acid - 8.0 g sulfanalic acid + 1000 ml. of
5 N acetic acid

d) Reducing sugars estimation:

1. Alkaline Copper Tartrate

(1). Dissolve 2.5 g anhydrous sodium carbonate, 2.0 g sodium bicarbonate, 2.5 g
potassium sodium tartrate and 20.0 anhydrous sodium sulphate in 80 ml water and
make up to 100 ml.

(2). Dissolve 15.0 g copper sulphate in a small volume of distilled water. Add one
drop of sulphuric acid and make up to 100 ml.

(3). Mix 4.0 ml of B and 96 ml of solution A before use.

2. Arsenomolybdate reagent:

Dissolve 2.5 g ammonium molybdate in 45 ml water. Add 2.5 ml sulphuric acid

and mix well. Then add 0.3 g disodium hydrogen arsenate dissolved in 25 ml
water. Mix well and incubate at 37ºC for 24 – 48 hours.

3. Standard glucose solution : Stock: 100 mg in 100 ml distilled water.

4. Working standard : 10 ml of stock diluted to 100 ml with distilled water

[100 µg/ml].
e) Crude fibre estimation :

1. Sulphuric acid solution (0.255 ± 0.005 N) : 1.25 g concentrated sulphuric acid

diluted to 100 ml (concentration must be checked by titration).

2. Sodium hydroxide solution (0.313 ± 0.005 N) : 1.25 g sodium hydroxide in

100 ml distilled water (concentration must be checked by titration with standard
acid).
Materials and Methods
 CHAPTER III

MATERIALS AND METHODS

This chapter deals with the materials and methods employed to determine the
engineering properties of selected raw minor millet grains and kernels, and design and
development of a double chamber centrifugal dehuller.

3.1.Selection of Raw Materials

The project is sponsored by International Research Development Centre, Canada.

International Research Development Centre had conducted survey’s in the millet growing areas
and decided to lay emphasis on development of processing machines ranging from 50 to
100 kg for three millets namely little millet, foxtail millet and proso millet was considered for
the studies since the area under this millet has come down drastically. Little millet, foxtail
millet and proso millet had been selected for the studies. Bulk grains of little millet (var. CO4),
foxtail millet (var. CO7) and proso millet (var. CO5) were purchased from the Department of
Millets, Tamilnadu Agricultural University, Coimbatore, India. The grains were stored in
closed containers for further processing.

3.2. Pre-cleaning of the Millets

Pre-cleaning was done to remove foreign matter, immatured grains and impurities.
Cleaning and grading was done by a lab model specific gravity separator of 50 kg/hr
capacity (make : Westrup, Denmark) and was used for de-hulling .

3.3. Moisture Content of Grains

The moisture content of the grains was determined by using hot air oven method
as per the procedures recommended by Association of Official Analytical Chemists
(AOAC, 1995). About 2-3 g sample was ground and kept in the oven for about one hour
at 130 ± 2 0C. After one hour, the grains were removed from the oven and were placed in
a desiccator to cool down and then weighed. The drop in weight of grain was measured
based on the differences between initial weight and final weight. The moisture content in
wet basis (w.b.) was calculated following the recommendations of AOAC (1995) as in
equation 3.1.
 The moisture content in wet basis = (Wm / Wt) x 100 --------- (3.1)

Where,

Wm – weight of the moisture, g

Wt – total weight of the grain, g

3.4. Moisture Simulation

The engineering properties of raw minor millets were investigated within the
simulated moisture content range of 10 to 14 per cent wet basis, since the post harvest
operations like storage, handling and processing are usually performed within this range of
moisture content. The grain samples of the desired moisture levels were prepared by adding
calculated amount of distilled water in accordance with equation 3.2 (Sacilink et al., 2002)
and mixing thoroughly to attain the required moisture content. The grain samples were
then sealed in 300 micron thick polyethylene bags and kept in a refrigerator and
maintained at a temperature of 4±1oC for a period of 10 days. The samples were then
taken from the refrigerator and temperature was brought down to ambient room
temperature. The moisture contents of the samples were checked for 10, 12, 14 per cent
(w.b.) as per the procedures recommended by AOAC (1995). Grains with higher
moisture contents were shade dried to attain the required moisture content.

 m f − mi 
Q = Wi  
 100 − m f
 
…..(3.2)

where,

Q - mass of water to be added, g

Wi - initial mass of grain sample, g

mi - initial moisture content of grain sample, per cent (w.b.)

mf - final moisture content of grain sample, per cent (w.b.)

All experiments were conducted in the laboratory at an ambient temperature of about

30±2oC and relative humidity of 55–65 per cent.
3.5. Engineering Properties

Engineering properties of food materials have a great impact on design and

utilization of processing and control systems. The methods used to determine the
engineering properties such as geometry and dimensions, thousand kernel mass, bulk
density, true density, porosity, angle of repose, static and internal coefficient of friction
and hardness for minor millet grains are given in this section.

3.6. Physical Properties

The physical properties such as size, shape, volume, bulk density, porosity, true
density, coefficient of internal friction, coefficient of external friction and terminal
velocity are very important for designing agricultural machinery and hence the physical
properties for little millet, foxtail millet and proso millet were determined by following
standard procedures.

3.6.1. Size of the material

The dimension of the grains were determined by picking hundred number of

grains and measuring their three principal dimensions (major, medium and minor
diameters) using a digital vernier caliper with a least count of 0.01 mm. The sphericity was
calculated as per the equation 3.3 (Mohsenin, 1986).

(LWT )3
1

φ= …..(3.3)
L

where,

φ - sphericity, decimal

L - length of grain, mm

W - width of grain, mm

T - thickness of grain, mm

3.6.3.Thousand grain mass

The grain mass of 1 kg was roughly divided into 10 equal portions. 100 grains
were randomly picked from each portion and mixed thoroughly. 1000 grains picked were
weighed using a digital electronic balance having a least count of 0.01 g. The measurement
was repeated five times and the mean was reported (Singh and Goswami, 1996).

3.6.4. Bulk density, porosity and true density

The minor millet grains were filled with a gentle tap in a circular container of
known volume (0.149x10-3 m3) and the mass of the content was weighed in an electronic
balance (0.01 g accuracy). The bulk density was then calculated as the ratio of mass (kg)
to volume (m3). The measurement was repeated five times and the mean was reported.
Toluene displacement method was used to determining the true density. It is evident from
the literature (Dursun and Dursun, 2005) that the absorption of toluene (C7H8) by the
grain and kernel is lesser as compared to water. The actual volume of grain will be equal
to toluene displaced (Singh and Goswami, 1996; Ogut, 1998; Konak et al., 2002). The
porosity of minor millet grains was determined using the Day’s apparatus (Mohsenin,
1986). The same apparatus was used by Balasubramaniam et al. (2010), Singh et al.
(2010) for millets respectively. The true density was calculated from the measured values
of bulk density and porosity using the following relationship (Mohsenin, 1970).

 (ρ g − ρb ) 
ε =  x 100 …..(3.4)
 ρ 
 g 

where

ε - porosity, per cent

ρb - bulk density of grain, kg/m3

ρg - true density of grain, kg/m 3

3.7. Frictional Properties

The frictional properties such as coefficient of friction and angle of repose are
important in designing machinery where there is a direct or indirect contact between the
material and the machine components.The grains will be subjected to friction while they
move inside the machine. The friction may act as a barrier for the movement of the
material and so it is necessary to study the frictional properties of the grains (Sahay and
Singh, 2010).
3.7.1. Angle of repose

The angle of repose is a major parameter for designing the feed hopper and feed
housing of the dehuller. The grains slide down from the hopper to feed housing due to
gravity and the slope provided in the feed hopper. The angle of repose for minor millet
grains was determined by following the procedure and calculated using equation 3.5
(Sreenarayanan et al., 1988).

θ = tan-1 (h / r) ….(3.5)

where

θ - angle of repose, degree.

h - height of the heap, cm.

r - radius of the platform, cm.

3.7.2. Coefficient of static friction

The materials interact with the surface of the metal bodies while they move within
the machine. It is essential to study the coefficient of static friction to know its influence
in designing the machine accordingly (Stewart et al., 1969). For the determination of
coefficient of static friction between the minor millet grains and the surface, an
experimental set up consisting of an open-ended sample holding container, pulley, thread
and loading pan were used. The minor millets were filled in the sample holder (Ns),
incremental weight was added to the loading pan (Fs) until the sample holder started
moving overcoming the static friction between minor millet and the sheet surface (mild
steel sheet) and the same was recorded as static frictional force. The experiment was
performed with grains of different moisture contents with test surfaces of mild steel and
galvanized iron. The coefficient of static friction (μs) was calculated with the help of
equation 3.6 (Sreenarayanan et al., 1988).

µs = Fs / Ns …..(3.6)
Where,

μs - Coefficient of static friction, dimensionless

Fs - Frictional force for static friction, g

Ns - Normal force for static friction, g

3.6.3. Coefficient of internal friction

For the measurement of coefficient of internal friction, the apparatus consisting

of a horizontally split cylinder was used (Thajudhin Sheriff, 1996). The lower portion was
fixed and the upper portion was removable, attached with a loading pan. Minor millet grains
were poured in the cylinder without any compaction and the weights (Fi) were added in the
loading platform incrementally to determine the force (Ni) required for sliding the minor
millet grains contained in the upper portion of the cylinder against the minor millet grains in
the lower cylinder. The internal frictional coefficient (µi) was determined from the equation
3.7 given by Sreenarayanan et al. (1988) and Thompson et al. (1983).

Fi
µi = ….. (3.7)
Ni

where,

µi - coefficient of internal friction, dimensionless

Fi - frictional force for internal friction, g

Ni - normal force for internal friction, g

3.7. Aero-dynamic Properties

The aerodynamic properties especially the terminal velocity of agricultural products

has great importance in considering the design of air and water conveying systems. It also
influences the designing of separation equipments (Arora and Kumar, 1999).

3.7.1. Terminal velocity

The terminal velocity of the particle is defined as velocity equal to the air velocity
at which the particle remains suspended in a vertical pipe. The terminal velocities of each
component of mixture decide the range of air velocity to be used for a definite extent of
separation (Arora and Kumar, 1999). An apparatus consisting of a vertical acrylic pipe of
50 mm diameter, a centrifugal blower driven by a 0.25 hp motor was used to determine
the terminal velocity of the grains. Layers of wire mesh were placed at the bottom of the
vertical pipe to get uniform velocity distribution within the chamber. The seeds were fed
into the vertical tube and the air velocity was adjusted at which 90 per cent of the
material floated against the gravity (Singh et al., 2009). The air velocity was measured
using an anemometer.

3.8. Composition of Grains

Determination of the composition is a deciding attribute to quantify the amount of

kernel, bran and husk present in the grains. The composition of the grains was determined by
removing the husk from randomly selected grain sample consisting of 100 grains (Malleshi,
1993). The selected grains were placed between two rough emery sheets and smoothly
rubbed to just remove the husk portion alone. The weight of the kernel, bran and husk was
determined.

3.9. Performance Evaluation of Conventional Dehullers Used for Millets

De-hulling is the process which removes the husk from the grain and makes it
available for consumption. The husk present in millets was removed using available
machines like abrasive dehuller, rubber roll sheller and centrifugal dehuller. The machines
were evaluated for their performance with 100 kg of the grains based on the standards
suggested by manufacturers. De-hulling efficiency and broken grain percentage were
evaluated for little, foxtail and proso millet at 10,12 and 14 per cent moisture contents (w.b.).

The Abrasive roller dehuller (Krisha Shellers, Madurai) as presented in Plate 3.1
and Fig. 3.1 consists of tapered roller of 250 mm diameter with 230 mm length fitted with
adjustable grooves between the concave and roller. Two sieves of various sizes are
attached to separate the grains into unhulled, dehulled and broken materials. A blower
with shutter arrangements to separate the husk. A variable speed motor 5 hp capacity was
fixed to operate the dehuller at different rpm to get maximum de-hulling efficiency.
 Fig. 3.1 and Plate 3.1. Abrasive type dehuller

Table 3.1. Specification of Abrasive type dehuller

Length of Power
Name of the Size of Peripheral
Principle concave, required,
machine roller, mm speed, m/s
mm hp

Abrasive type Abrasion 250 mm 230 mm 5 hp 15.75

dehuller
 Feed hopper

Pulley

Clearance
adjustment

Belt

Motor

Plate 3.2. Rubber Roll Sheller (Satake Model)

Table 3.2. Specifications of the rubber roll sheller

Size of Capacity,
Overall dimensions, mm
rubber roller t/h
Type Power, hp
Length Diameter, Length, Breadth, Width,
mm mm mm mm mm

PNU 254 254 1 800 1000 2650 5

 Feed hopper

De-hulling Chamber

Separation box

Husk outlet

Plate 3.3. Single chamber centrifugal de-huller

Table 3.2 Specifications of Single Chamber Centrifugal De-huller

1. Name of the : Millet dehuller

machine

2. Capacity : 50 kg/hr

3. Grains : Little millet, Foxtail millet

3. Size of : 12 cm x 5.3 cm
impeller(diameter) 12 cm dia x 3.5 cm
width(ID)

4. Size of casing : 22 cm x 5 cm (2 No’s)

5. RPM of motor : 1420 rpm

6. RPM of impeller : 5000 rpm

7. RPM of blower : 1420 rpm

8. Motor : 2 HP single phase motor

9. Blower type : Suitable blower

 A single chamber de-huller as given in Plate 3.3 and specifications as given in
Table 3.2 were designed, developed and tested for its perforformance. A maximum
de-hulling efficiency and less brokens were obtained for little millet, foxtail millet and
proso millet with centrifugal de-huller while the rubber roll sheller and abrasive de-huller
gave poor results. Hence it was decided to incorporate modifications in the centrifugal
de-huller to improve the de-hulling efficiency for minor millets. The centrifugal de-huller
works on the principle of impact which removes the husk from the grains. The grains fed
through the feed hopper are subjected to centrifugal force by an impeller rotating at high
speed. At this high speed the husk is split from the grains and the kernel is brought out.
The machine has unique feature of procesing even minimum quantity of grains with less
power. Nowadays the government is interested in promoting small de-hulling machines
for the development of livelihood of small and marginal farmers. Keeping this in view
the capacity of the machine to be developed was fixed as 75 kg/h. In the conventional
de-huller, atleast two to three passes are required to dehull the grains depending on the
nature of the millet. In order to reduce the drudgery involved in the single chamber
de-huller it was decided to develop a centrifugal de-huller with two de-hulling chambers.

3.10. Design and Development of a Double Chamber Centrifugal De-huller for

Selected Millets

A double chamber centrifugal dehuller of 75 kg/h was developed for de-hulling

little, foxtail and proso millet. The machine consists of components like feed hopper and
housing, impeller, de-hulling chamber, separation chamber, grain outlet, husk outlet,
motor and transmission components. The machine components were designed and then
fabricated.

3.10.1. Design of components

3. 10.1.1. Feed hopper and housing

A feed hopper of the size 400 mm x 400 mm x (167 mm,125 mm, 92.45 mm) on
one side and mm was fabricated using 2.65 mm mild steel sheet. Feed rate of the grains
depends on two major factors namely slope of the feed hopper and size of the shutter
opening. A lever arrangement was provided at the base of the feed hopper to adjust the
 All dimensions are in mm
Fig. 3.2. Feed Hopper
slope by 5 to 8º as recommended by Chakraverthy et al.(2003) for easy flow of the
grains. The shutter was provided with a screw adjustment to vary the clearance in the
feed hopper. The detailed view of the feed hopper is shown in Fig.3.2 and Plate 3.4.

Feed regulator

Feeding Chamber

Plate 3.4. Feed Hopper

The feed housing is the component connecting the feed hopper and the hulling
chamber. Grains from the feed hopper enter the impeller through the housing. A feed
housing was fabricated with an opening at the top. The size of the opening was 7.5 cm x
7.5 cm. The outlet of the housing is a circular ring of 6.8 cm diameter and 3 cm length.
The housing was moulded with cast iron and was machined to ensure smooth flow of
grains. The housing enclosed a bearing at one side and supported the impeller on the
other side. The shaft passes through the housing and was connected to the impeller.
The schematic diagram of the feed housing is presented in Fig. 3.3 and 3.4.a,b. Also the
feed housing is presented in 3.5.

Fig. 3.3. Feed housing

 3.4. Side view
3.5. Frontview

Fig 3.4. Feed Housing-Side view and Fig. 3.5 Front view

Grain entry

Grain outlet
Base

Plate 3.5. Feed Housing (Front View and Isometric view)

3.10.2. Development of De-hulling Chamber and Impellers

De-hulling chamber is another component of more importance in a centrifugal

dehuller on which the grains hit the impeller and the kernels come out of the husk. The
chamber consists of two parts namely grain hitting surface and a cover. The diameter of
the chamber has to be around 2 to 4 cm more than that of the impeller in order to have
maximum de-hulling . Hence the diameter for the top chamber was taken as 16 cm and
the bottom chamber was taken as 18 cm. The width of the chamber with the hitting
surface and cover was decided to be 5 cm, so that the entire impeller assembly can be
fixed inside the chamber with adequate clearance. Different de-hulling surfaces are
presented in plate 3.6.

1 2 3 4

Plate. 3.6. De-hulling surfaces

Plate 3.6.1- grooved casing, 2- angular casing, 3- Flat Casing, Outer casing

It was decided to have three different hitting surface configurations namely the
flat surface, angular with grooved surface and flat surface with grooves. Necessary
provisions were made in the moulds for fitting the shaft and grain outlet as well. Exactly
2 mm groves were provided to the flat and angular de-hulling chambers on their inner
surface to facilitate tearing of the husk. The schematic diagram of the de-hulling
chambers is represented in Fig.3.6 and Fig.3.7 and Plate 3.6.

3.10.2.1. Design of flow rate

The capacity of the machine to be manufactured was chosen as 75 kg/h. 75 kg/h

was selected to meet the demand of making machines in the capacity range of 50 to
100 kg/h. To maintain this capacity the volume and flow rate were calculated from the
values of bulk density and volume. this capacity. The flow rate was calculated by using
equation 3.7 (Sahay and Singh, 2010).
 Hitting surface
Impeller eye
Impeller eye
Hitting surface

Flat de-hulling chamber

Grooved de-hulling chamber

Impeller eye
Hitting surface

Angular de-hulling chamber

Fig 3.6. Schematic diagram of first de-hulling chamber

 Impeller eye
Hitting surface

Flat de-hulling chamber

Hitting surface
Impeller eye

Grooved de-hulling chamber

All dimensions are in mm
Impeller eye
Hitting surface

Angular de-hullling chamber

Fig 3.7. Schematic diagram of second de-hulling chamber

 We know that density is ratio of mass/volume

So, M= ….(3.7)

A value slightly higher than the bulk density of the grains was taken for the
design in order to have a flexibility in volume of grains to be handled. So the bulk
density of 750 kg/ m3 was considered for the design.

From equation 3.8 (Sahay and Singh, 2010), the volume of the material was
calculated as

V = M/ ….(3.8)

= 75/750

= 0.1 m3 for one hour

Therefore the flow rate/s = 0.1 m/3600 = 2.78 x 10-5 m3 /s

To ensure the flow of grains into the impeller a minimum pressure difference should
be maintained between the hopper and impeller entry point. Grains continuously move
towards the inlet of the rotating impeller due to a pressure difference of 5-10 m between the
point of entry and point of exit of housing. The grains in the feed hopper are at atmospheric
pressure (1.03 kg/cm2), so a low pressure was created between the impeller centre and
housing inlet to create smooth flow of grains (William C. Osborne, 1977).

3.10.2.2. Design of impeller components

The crown plate, base plate and vanes are the main components of an impeller.
The individual components were designed and then fabricated. The components design is
important to have a smooth entry and exit of the grains into and out of the impeller.

3.10.2.3. Crown plate

Crown plate is the front part of the impeller. The grains enter the impeller through
the opening given in the crown plate. The crown plate as given in Fig. 3.8 was designed
first. The diameter of the crown plate inlet opening (Dci) depends upon the feed rate.
The velocity of flow (Vci ) at the crown plate inlet opening should be kept minimum to
avoid prerotation at the entry point of the impeller. The value was assumed as 3 m/s for
 All dimensions are in mm

Fig. 3.8. Crown plate

 All dimensions are in mm

Fig. 3.9. Base plate

determining the crown plate inlet opening diameter (Dci) as per recommendations by
pump designers. Bruno Eck (1973). Based on the flow rate the diameter of crown plate
inlet opening was calculated using equation 3.9 (Bruno Eck , 1973).

Diameter of crown plate,

….(3.9)

Where, Vci is the velocity of flow at the suction flange, m/s.

The effective diameter was calculated by taking into account the shaft diameter
and the leakage losses. The shaft passes through the housing into the impeller. To allow the
shaft to pass through and then assure smooth flow of grains into the impeller, a factor of
safety of 25 was taken as per pump designer’s recommendation (William C. Osborne, 1977).
The diameter of the crown plate inlet with a factor of safety works out to 0.0680 m.

3.10.2.4. Diameter of eye of impeller

The grains enter the impeller through the eye of the crown plate at a minimum
velocity. The inlet velocity C0 should be more than the crown plate inlet velocity , so the
value of the inlet velocity was taken as 3.2 m/s. Bruno Eck (1973) have suggested that the
diameter of the impeller should be slightly higher than the eye of the impeller. We know
that the diameter at the inlet of crown plate is 0.068 m. So, the diameter of the impeller
eye (De) was fixed as 7.0 m, which is 0.012 m greater than the crown plate inlet diameter.

3.10.2.5. Inlet Outlet vane edge diameter

The inlet vane edge is the point of starting of the vanes. Normally the vanes start
immediately from the outer periphery of the eye. As per recommendation, the diameter at
inlet vane edge should be equal to impeller eye diameter and outlet edge diameter (Do
should be twice the diameter of the eye of the impeller (Church, 1972).

Therefore, when Do = 2 Di = 2 x Do = 0.14 m ----- (3.10)

3.10.2.6. Crown plate and base plate diameter

The crown plate and base plate should be of the same size as in fig 3.8 and 3.9.
So the base plate diameter was fixed as 0.14 m. The thickness of the crown plate was
chosen as 3 mm and the thickness of the base plate was fixed as 5 mm for safety purpose.
The crown plate of 3 mm thickness and base plate of 5 mm thickness. The diameter of the
plates was base plate and crown plate of diameter 0.14 m was fabricated.

3.10.2.7. Width at inlet

The amount of grains passing through the impeller is decided based on the size of
the impeller. The width of the impeller is very important parameter to be designed for
smooth flow as well. Total width of the impeller is the width of the clearance between
crown plate and base plate. The vanes are fixed in between the crown plate and base
plate. The width is decided based on the d/b ratio. The d/b ratio should be atleast 5 for
smooth flow as per pump manufacturer’s recommendation (William C. Osborne, 1977).
We know that the diameter of the impeller is 0.14 m. So the breadth was calculated using
equation 3.11 (William C. Osborne, 1977).

------ (3.11)

d/5 = b, 0.14/5 = 0.028

Deducting the thickness of the crown plate plus the base plate, the width of the
impeller was 0.020 mm which is equal to 2 cm.

3.10.2.8. Design of vanes and their angles

The vanes guide the grains to the point of impact. The inlet vane angles, exit vane
angles guide the grains to the point of impact and they were designed for further
fabrication of the impeller.

3.10.2.8.1. Inlet vane angle

The grains enter the impeller radially at an angle of 900. Grains travelling straight
from the housing, takes a 900 turn at the point of entry in the impeller. The entering
grains make an angle at the point of entry and it can be calculated using equations

3.12 and 3.13 given by William C. Osborne (1977). The velocity at the eye of the
impeller was 3.20 m/s. Usually the grains enter the vane area at a velocity ( which is
slightly higher than the velocity (Co) at the entry point., so the velocity was fixed as 3.40 m/s
The tangential velocity ( and entry velocity ( are important for calculation of
angle of entry of the grains. The value of u1 can be found out using equation 3.12 and
substituted in equation 3.13 to know the angle at which the grains enters (Church,1972).

….(3.12)

3.14 x 0.07 x 3800

u1 = = 13.92 m/s
60

By substituting the values of and

……..(3.13)

0.25

= 14.11°

The recommendation says that the inlet vane angles between 10° and 25° will
give high efficiencies. So as per the recommendation the inlet vane angle was fixed as
20° (Church, 1972).

3.10.2.8.2. Outlet vane angle and dimension

The grains are guided towards the spot of impact by the vanes. The inlet vane
angle allows the material to enter correctly and travel in the designed pathway while the
outlet vane angle directs it to the spot of impact. Pump manufacturers normally adopt a
vane angle of 15° to 40° for outlet vane angles (Bruno Eck, 1973). The outlet vane angle
(β2) of 30° was selected for the study. The grains travel out of the impeller at very high
velocities. The velocity (Cu2) of the material when there is no barrier is the theoretical
velocity of the grains which is calculated from equation 3.14. For calculation of absolute
velocity, the tangential and radial velocities should be known. Cm2 is the radial velocity
of grains at the outlet. The value of the outlet radial velocity should be less than the inlet
radial velocity (Church, 1972). Since the inlet radial velocity was 3.50 m/s. The outlet
radial velocity Cm2 was fixed as 3.25 m/s for good performance.

….(3.14)

= 21.97 m/s

The grains do not actually come out with the theoretical velocity. So we have to apply a
correction factor of η∞ = 0.7 for the calculation of velocity of exiting grains Church
(1972). The velocity ( at which the grains come of the impeller was calculated using
the equation 3.15 (Church, 1972).

η∞ ….(3.15)

= 15.38 m/s

The actual absolute velocity can be calculated from the equation 3.16, (Bruno Eck, 1973).
2
2) ….(3.16)

= 15.90 m/s

The absolute velocity is the vector sum of the tangential and radial
velocities. The actual absolute velocity is the final velocity at which the grains exit the
impeller. They should be considered for further calculations. Similarly the velocities were
found out for the peripheral speeds of 26.39 and 24.92 m/s respectively and reported.
The exit velocities calculated were 14.89 m/s and 13.89 m/s.

3.10.2.8.3. Number of vanes

The recommendations for number of vanes for handling bulk materials like small
size solids, sewage, sludge and slurry is 3 to 6 vanes and radius of curvature is 6.0 cm
(Bruno Eck, 1973). It was proposed to find out the best impeller among 3 vane, 4 vane
and 5 vane impellers respectively. So, three impellers were constructed with the designed
dimensions. All the calculations are made for the maximum peripheral speed, so the
machine will work safely for the other peripheral speeds as well. For the second chamber,
the design is the same since we are using the same dimensions of the impeller for both
the chambers. The radius of curvature of the vanes was 6.0 cm for all the impellers. The
base plate alongwith the three vanes, four vanes and five vanes are given in Plates 3.7,
3.8, and 3.9 and fig 3.10, 11, 12. An impeller of the size 140 mm x 33 mm was fabricated
as given in the Fig. 3.13. The crown plate, vanes and base plate were arranged in
sequence and they were riveted to each other.

Plate 3.7. Three Vane impeller

Vanes

Base plate

Plate 3.8. Four vane impeller Plate 3.9. Five vane impeller
 All dimensions are in mm

1. Vane 2. Base plate

Fig. 3.10. Three vane impeller

 All dimensions are in mm

1. Vane 2. Base plate

Fig.3.11. Four vane impeller

Fig.3.12. Five vane impeller
All dimensions are in mm

Fig. 3.13. Impeller

 Three impellers of the size diameter 14 cm with eye diameter 7 cm and hub
diameter 2.5 cm were selected for the studies. Two separate outlets were provided for the
collection of grains and husk separately. A provision was made with a circular net
having a diameter of 20 cm at the husk collection end so that the husk is collected
without causing any pollution.

3.12. Separation Chamber

The grains after processing, enter the separation chamber whose function is to
separate the husk from the dehulled grains. The separation chamber is made of two
zones. The top zone reduces the speed of the kernels and the husk by means of baffles.
Four baffles made of mild sheet of 30 cm length and 8 cm width was fixed at an
inclination of 60º to reduce the speed of the grains. The bottom zone was fitted with a
standard centrifugal blower of capacity 0.0216 m3/s and it is run by a 0.5 hp motor.
The diagram of the separation chamber is presented in Fig 3.14 and Plate 3.10.
 Blower
Husk Outlet

Husk Outlet

Plate 3.10. Separation Chamber (Front view and side view)

3.10.2.9. Force acting on the grains

The grains come out of the impeller at high velocities. The velocity required to
split open the husk and bring out the grains had already been calculated and the force at
which the grains come out is calculated from Newton’s second law of motion. The grains
coming out of the impeller is thrown at the casing with the given absolute velocity. The
force required for de-hulling can be calculated using Newton’s second law of motion.
The centrifugal acceleration is calculated using the equation (3.17) given by (Khurmi,
2012).

F = ma c , where m is the mass of the particle, kg

ac is the acceleration of the particle, m/s2

v2
ac = ---- ----- (3.17)
r

(15.90) 2
ac = ----------- = 3611 m/s2
0.07
where v = velocity of material coming out in m/s and r is the radius of the impeller. The
acceleration value for a peripheral speed of 24.92 was calculated as 2756 m/s2 and at
26.39 it was calculated as 3167 m/s2. The absolute velocity of the grains at 24.92, 26.39
and 27.82 m/s was substituted for v and the force was calculated for a single grain as
given in equation 3.18. We know that the mass of the 1000 grain mass is 5 g, From this
value, the average mass of a single grain mass was calculated as 5 x 10-3.

The grains come out of the periphery of the impeller with high centrifugal force.
The centrifugal force for a rotating material is mass multiplied by the rotational
acceleration and is given in equation 3.18 (Khurmi, 2012).

F = m. ac --------(3.18)

Force, kg. m/ s2 = 5 x 10-3 x 3611 = 18.5 kg. m/s2

= 18.5 N

For peripheral speed 26.39 m/s , force = 5 x 10-3 x 3167.32

= 15.83 N

For peripheral speed 26.39 m/s , force = 5 x 10-3 x 2756.00

= 13.78 N

Torque ( Ts) was calculated experimentally. The impeller was mounted on a shaft
and the torque required to rotate the impeller was found out by rolling a rope on the
peripheral of the impeller. A provision for adding weights at the end of the rope was
provided. The weight was added till the impeller started rotating and the weight at which
the impeller started rotating was noted down. Weight required for rotation was around
2.86 kg to get the rotational motion for the 14 cm diameter impeller. The maximum
torque required was using equation, (Khurmi, 2012).

T=Wxr -------- (3.19)

T = 2.86 x 7 = 20 kg.cm

Where, T is the torque in kg.cm , W is the weight added to rotate the impeller in kg and r
is the radius of the impeller in cm. The shaft is made from mild steel rod, so for
designing the diameter the safe stress, Ss was assumed to be 450 kg/cm2, (Khurmi, 2012).
The shaft diameter was calculated using the equation proposed by Khurmi (2012).

-------- (3.20)

=0.60 cm 6.0 mm

Shaft diameter ( ) of 6.0 mm satisfies the requirements of torque only. The

effect of bending moment is also to be considered for designing the shaft (Khurmi, 2012).
A factor of safety of ranging between 4 to 4.5 should be provided to the shaft to
overcome the stresses due to bending moment and twisting bending moment (Khurmi,
2012). So the shaft diameter was taken as 25 mm based on the suggestions. The hub
diameter should to be 10 to 15 mm more than more than the diameter of shaft (Bruno
Eck, 1973). Thus, the diameter of hub was fixed as 30 mm.

3.10.2.10. Calculation of power requirement

Power required to operate a single impeller was calculated from the equation 3.21

hp required = ….(3.21)

Power require for two

= 1.02 hp
impellers = 2.12 hp

3.10.2.11. Design of blower

A centrifugal blower was designed and fabricated to separate the grain and the
husk based on their difference in densities. The grain husk mixture coming out of
de-hulling chamber falls into the slow down box and pass through a set of baffles due to
which they are slowed down and are sprinkled into the blower chamber. The husk will be
blown by the blower while the grains will drop into the grain outlet. The flow of the
grains can be regulated by means of shutters provided below the second hulling chamber.
The air can be regulated by shutters provided at the sides of the blower.

Torque was calculated by suddenly stopping the impeller at full load condition
and force required to move the impeller was determined using a pulley and loading a pan
and it was determined as,

Weight required to move impeller, kg = 3.00 kg

Diameter of pulley = 9 cm

= 27 kg-cm

= 0.27 kg-m

hp required (Khurmi, 2012) = ….(3.21)

A shaft size of 25 mm was selected for safe design so that the shaft will overcome
the bending and twisting moments.

3.10.2.12. Total power required by the machine

The total power required for the operating the machine is the sum of power
required to operate the impellers + Power required to operate the blower + Transmission
losses.

Power required by the both the chambers = (1.06 x 2) = 2.12 hp

Power required by the blower = 0.50 hp

Considering all the transmission losses to be 10%, the total power for the impeller
is 2.66 hp. By adding the power required for the impellers and the blower the power
required was calculated as 2.92 hp. The total power required to operate the machine.
A standard 3 hp, three phase motor was selected for the study.
3.13. Transmission Components

Shafts, step up and step down pulleys, bearings and v-belts are the components
which were used for transmission of energy from the motor to the machine parts.

3.14. Frame and Assembly

The machine components were fabricated as per the design and fitted in a frame.
The frame was made of standard M.S. channels, L-angles, and flats of size 540 x 640 x
840 mm. The machine components were assembled in the frame and is shown in Figure.
3.13 and Plate. 3.11.

3.15. Variable Speed Drive

The required peripheral velocities of 24.92 (3400 rpm), 26.89 (3600 rpm) and
27.86 (3800 rpm) m/s were attained by using a variable speed drive.

3.16. Total Power Required by the Machine

The total power required for the operating the machine is the sum of power
required to operate the impellers + Power required to operate the blower + Transmission
losses.

Power required for both the impellers = (1.06 x 2) = 2.12 hp

Power required by the blower = 0.50 hp

Total power required = 2.62 hp

Transmission loss = 10 %

Nett power required = 2.88 hp ~ 3.0 hp

3.17. Design of Experiments

As per the recommendations of Chakraverty et al.(2003), the moisture content of

the grains was selected within the grain storage and handling range. The number of vanes
recommended for handling small solids are 3 to 6 vanes as per Church (1972) and hence the
impellers with 3,4,5 and 6 vanes were selected. So the number of vanes were selected as 3,4,5
and 6 for the studies. The feed rate of 75 kg/h was maintained for all the experiments.
 Side view Front View

1. De-hulling 2. De-hulling 3. Grain outlet 4. Feed hopper

chamber I chamber II

Plate 3.11. Double chamber centrifugal de-huller for millets

 All dimensions are in mm

1. Feed 2. Dehulling 3. Dehulling 4. Separation

hopper chamber-1 chamber-2 chamber

5. Blower 6. Motor 7. Feed housing

Fig. 3.15. Double Chamber Centrifugal De-huller for millets

Three De-hulling chamber surfaces namely grooved, angular and flat surfaces were
chosen for the studies. The design of experiments and their levels used in the present
study are as follows (Table 3.3).

Table 3.3. Independent and Dependent variables

S.No Independent Variables Levels Used

1. Moisture content of grains, % (w.b.) 10, 12, 14

2. Peripheral velocity of impeller, m/s 24.92 (3400 rpm)

26.89 (3600 rpm)
27.86 (3800 rpm)

3. Number of vanes in impeller, Nos 3

4
5
6

4. De-hulling chamber configuration Grooved

Angular
Flat surface

Dependent Variables

1. De-hulling Efficiency (η), (%)

2. Broken grain (Bg,) (%)

The machine was tested and with no load and it was found that a peripheral
velocity beyond 27.86 m/s gave more vibration and makes unusual sounds. So the
peripheral velocity was restricted to 27.86 m/s.

3.18. Performance Evaluation of the Double Chamber Centrifugal Dehuller

During the preliminary trials, poor de-hulling efficiency and more breakage of
kernels was observed with the impeller having six vanes. This may be due to the fact that
the clearance between the vanes is only 5 mm and that has hindered the entry of grains
into the impeller and caused rubbing which resulted in more breakage. Hence the
experiments were confined with the impellers having 3, 4 and 5 vanes only.
3.19. De-hulling Efficiency and Broken Grain Percentage

Five random samples of 25 g were collected from the kernel outlet and the
samples were separated into three major divisions namely whole kernels, broken grains
and unhulled grains. The De-hulling efficiency and broken grain percentage were
calculated using the equations 3.26 and 3.27 proposed by Singh et al.(2009).

….(3.26)

Where,

η = De-hulling efficiency, %

Mu= Mass of the unhulled grain, g

Mb= Mass of broken grain, g

Mt = Total mass of the grains taken for De-hulling, g

......(3.27)

3.20. Chemical Analysis

The three millets namely little millet (CO4), foxtail millet (CO7) and proso millet
(CO5) were dehulled in the developed unit. The samples of the de-hulled kernels and
husk were collected separately. 5 kg of the millet was taken for each trial and it was de-
hulled at a rate of 75 kg/h.

25 kg of the same varieties of little, foxtail and proso millet were dehulled by
maintaining the operational parameters in the conventional abrasive type dehuller
available at Peraiyur in Madurai district. The kernel and the husk were weighed after the
De-hulling process and the samples were compared with those obtained from double
chambered dehuller for the nutrients.
3.20.1. Estimation of moisture

The moisture content of the sample was estimated by the hot air oven method
suggested by AOAC (1995). About 5.0 g of sample was weighed accurately and dried in
an air oven at 1100C. The drying was continued till a constant weight was obtained. The
moisture content was expressed as percentage.

3.20.2. Estimation of protein

Protein was analysed by the amount of nitrogen available in the sample by micro
kjeldhal method (AOAC, 1980). One gram of sample was transferred into 250 ml
digestion flask along with one to two gram of catalyst mixture and 25 ml of concentrated
sulphuric acid. The catalyst mixture consists of 2.5 g of powdered selenium-dioxide,
10 g of potassium sulphate and 20 g of copper sulphate. The sample was digested until
the solution became colourless.

The digested sample was made upto 100 ml with distilled water in a volumetric
flask. A known amount of aliquot was transferred into the distillation flask. To this 10 ml
of saturated sodium hydroxide solution was added. The solution was distilled and the
ammonia evolved was trapped in boric acid placed in a beaker at the tip of the condenser.
The solution was titrated against the N/70 hydrochloric acid for the end point, until the
colour changes. The same procedure was repeated to get the blank titre value and the
nitrogen content of the sample can be calculated. The nitrogen value multiplied by factor
6.25, gives the crude protein content of the sample in per cent.

3.20.3. Estimation of fat

The fat content of the sample was estimated by the method described by AOAC,
(1995). The lipid in the sample was extracted with petroleum ether (60-80oC) in a
soxhlet-apparatus for 16 hours. The solvent was evaporated and the remaining residue
was weighed. The fat content was expressed in percentage.

3.20.4. Estimation of total carbohydrate

An aliquot of 0.2 ml of the supernatant was taken in a test tube and the volume
was made upto 1.0 ml using distilled water. To this 3.0 ml phenol solution was added and
mixed well. Then 5ml of concentrated sulphuric acid was then added from a fast flowing
pipette and agitated. After 10 min, the absorbance of the solution was measured at
490 nm. Standard curve was prepared by taking glucose in the range of 20 - 100 µg and
the amount of total carbohydrate was calculated using the standard curve.

3.20.5. Estimation of crude fibre

The crude fibre content was determined by the method described by Sadasivam
and Manickam (1996). The dried sample was taken in a beaker and 200 ml of 1.25 per cent
H2SO4 was added and boiled for 30 min. The contents were filtered through muslin cloth
and washed with distilled water until the washings were no longer acidic. The residue
was transferred into the same beaker and boiled with 1.25 per cent NaOH for 30 min and
filtered through a muslin cloth, washed with 50ml of distilled water and 25ml of alcohol.
The residue was transferred into a pre weighed silica crucible, dried for 2-4 hrs at 1300C,
cooled and weighed. It was ignited and ashed for 30 min at 6000C, cooled and weighed.
The loss in weight due to the fibre content was expressed in percentage.

3.20.6. Estimation of ash

About 5 to 10 g of the sample was weighed accurately into a crucible (which has
previously been heated to about 6000C and cooled (AOAC, 1995). The crucible was
placed on a clay pipe triangle and was first heated over a low flame till all the material
was completely charred followed by cooling in desiccators and weighed. To ensure
completion of ashing the crucible was again heated in the muffle furnace for ½ hour,
cooled and weighed. This was repeated till two consecutive weights obtained were same
and the ash was almost white or greyish white in color.

3.20.7. Preparation of ash/mineral solution for estimation of calcium and iron

The mineral solution of all samples were prepared by dissolving the ash obtained after
ashing the samples in a muffle furnace in dilute hydrochloric acid (1:1). This mixture was then
heated over a water bath to dryness before another 5 ml of the solution was added. It was
heated further over the water bath until the sample started fuming and at this point, the crucible
was retrieved and its contents filtered into a100 ml volumetric flask using Whatman No.40
filter paper. After thorough rinsing of the crucible and the filter paper, the volume was made up
to the mark with distilled water in the flask. Aliquots of this mineral solution were taken for the
estimation of all the minerals in this study.
3.20.8. Estimation of calcium

Two ml of prepared ash solution and calcium standard was taken in duplicate.
One ml of ammonium oxalate was added to each test tube. The solution was allowed to
stand for 30 min with shaking at intervals and centrifuged. The supernatant was discarded
and to the residue 3 ml of ammonia was added and again centrifuged. The supernatant
was discarded and 2 ml of 1 N sulphuric acid was added and kept in the boiling water
bath for 8 minutes and later the hot solution was titrated against 0.02 N potassium
permanganate till the appearance of end point pale pink colour. Blank was prepared with
2 ml of sulphuric acid and titrated against 0.02 N potassium permanganate. The
difference between the two titration indicated the volume of potassium permanganate
required to titrate the oxalic acid formed from calcium oxalate. The result was expressed
as mg of calcium per 100 g of sample (AOAC, 1995).

3.20.9. Estimation of iron

Two ml of ash solution was taken in a test tube to which 1.0 ml of saturated
potassium sulphate and 1.0 ml of 30 per cent sulphuric acid were added and made upto to
8.5 ml with double distilled water. About 1.5 ml of 3 N potassium thiocyanate was added
to the tube for colour development. The intensity of colour was read at 530 nm in a
colorimeter. A standard graph was drawn using standard iron solution (ferrous
ammonium sulphate). The mg per cent of iron was calculated by the values on the
standard graph (AOAC, 1980).

3.21. Cost Economics

The cost of the double chamber centrifugal dehuller was calculated based on the
material used for fabrication and the labour required to develop and fabricate the
machine. The cost economics for the newly fabricated double chamber centrifugal
dehuller was calculated following the methods recommended by Palnabendu et al.(2007).
The fixed cost can be calculated using equation 3.28.

Fixed cost of the unit/year, (Rs) = i(i+1)N ….(3.28)

(i+1)N-1
where,

I - interest, decimal

N - life span of the unit, years

C - Cost of the unit, Rs

The variable cost of the newly fabricated dehuller was calculated by considering
the repairs and maintenance, electricity charges and cost of labour involved for
De-hulling . The cost of operation to dehull the millets was calculated from the values of
fixed cost and variable cost worked out for optimum operating conditions.

3.22. Statistical Analysis

The various data observed were analyzed using a JMPR (SAS software). The
Analysis of variance (ANOVA) and mean table for different process parameters was
tabulated and the level of significance was reported.
Results and Discussion
 CHAPTER 4

RESULTS AND DISCUSSION

In this chapter, the results of various investigations carried out on engineering

properties of minor millet grains and the performance evaluation results of the double
chamber centrifugal de-huller are presented and discussed.

4.1. Physical properties of Grains and Kernel

Influence of moisture content on physical properties viz., sphericity, thousand

kernel mass, bulk density, true density, porosity, angle of repose, coefficient of static
friction, coefficient of internal friction and terminal velocity were determined for little
millet, foxtail millet and proso millet. The values recorded during the experimental study
for little millet, foxtail millet and proso millet grains are presented in Table 4.1 and for
kernel (de-hulled grain) in Table 4.2.

4.1.1. Sphericity

The value of sphericity of grain and kernel increased with increase in moisture
content. The higher sphericity was exhibited by little millet, foxtail millet and proso
millet, at higher moisture content. The increase in sphericity upon addition of moisture
have been reported for barnyard millet by Singh et al. (2010).

4.1.2. Surface area

The values of grain and kernel surface area increases linearly with increase of
moisture content. This increase in surface area was happened due to increase of
dimensions with increase of moisture content of grains and kernels. Similar trend of
increase have been reported by Altuntas and Yildiz (2007) for faba bean and Singh et al.
(2010) for barnyard millet.

4.1.3. Thousand kernal mass (g)

The 1000 kernel weight increased linearly with increase in moisture content.
Singh and Goswami (1996) found the same linear relationship between thousand weight
and moisture content for cumin seed. The 1000 seeds weight and kernels at wet basis
ranged from 3.25 to 4.16 g for little millet, from 3.45 to 4.21 g for foxtail millet, from
4.52 to 5.65 for proso millet on 10 and 14 per cent moisture content, respectively. Similar
results were reported by Joshi et al. (1993) for pumpkin seeds and kernels.

4.1.4. Bulk density

A linear decreasing trend with in bulk density, moisture content range of 10 to

14 per cent on wet basis were observed. Higher bulk density was exhibited by little
millet, foxtail millet and proso millet, at lowest moisture levels. Suthar and Das (1996)
found that the bulk density increased linearly with grain moisture increase for karingada
seeds. These discrepancies could be due to cell structure, volume and weight increase
characteristics of grains and seeds.

4.1.5. True density

The true density decreased with increase of moisture content. This decrease indicates
that there is a lesser increase in grain mass compared to increase in its volume with its
moisture content increase. These results agree with the fi ndings of Suthar and Das
(1996) with pumpkin seeds. Higher mass range of seeds, which is due to larger size of
seed, also increases the volume resulting in decreased true density. Many researchers
have reported the decrease in true density with moisture content increase (4 - 32.6% db)
for karingda seed (Suthar and Das 1996), chick pea seeds (Konak et al. 2002) and
rapeseed (Calisir et al. 2005).

4.1.6. Porosity

Porosity values decreased with increase in moisture content. Porosity values was higher
for little millet, foxtail millet and proso millet, at lowest moisture levels. Joshi et al. (1993) also
reported a linear decrease in porosity with increase in pumpkin seed moisture content.

4.1.7. Angle of repose

Angle of repose showed an increasing trend with moisture content. At lowest

moisture content, angle of repose was recorded lower for little millet, foxtail millet and
proso millet. At higher moisture content (14%, wb), proso millet exhibited a higher angle
of repose (30.16o), followed by foxtail millet and little millet. These results are
accordance with findings of Joshi et al. (1993) for pumpkin seeds and kernels.
4.1.8. Coefficient of static friction

The coefficient of static friction for minor millet against mild steel surface increased
with moisture content. Little millet had exhibited least coefficient of static friction at low
moisture content (10%, wb) and all three millets exhibited a higher value of coefficient of
static friction at higher moisture content (12 and 145% wb, rspectively). Similar results were
reported by Balasubramanian and Viswanathan (2010) for minor millets. The reason for
increased friction coefficient at higher moisture content may be due to water present in grains
offering increased adhesive force on contact surface. Among various contact surfaces, mild
steel offered higher coefficient of static friction as compared to stainless steel and aluminium
were reported by Singh and Goswami 1996) and rapeseed (Calisir et al., 2005).

4.1.9. Coefficient of internal friction

The coefficient of internal friction increased with moisture content. This may be
due to higher cohesion exhibited by minor millets at higher moisture content. A higher
coefficient of internal friction was offered by little millet, foxtail millet and proso millet at
high moisture content (12 and 14% wb). Similar results were reported by Balasubramanian
and Viswanathan (2010) for minor millets.

4.1.10. Grain hardness

Hardness showed a decreasing trend with increase in moisture content. At all

moisture contents, all three millets exhibited almost higher hardness. The small rupturing
forces at higher moisture content might have resulted from the fact that the seed became
soft and more sensitive to cracking at high moisture. It indicates that greater force was
necessary to rupture the seed with lower moisture. Similar trends were also reported by
Balasubramanian and Viswanathan (2010) for minor millets.

4.1.11. Terminal velocity

The values of terminal velocity increases linearly with increase of moisture

content. Similar trend was observed by Subramanium and Viswanathan (2007) for millets
grain and flours.
Table 4.1. Effect of Moisture Content on Engineering Properties of Millet Grains

Little Millet Foxtail Millet Proso Millet

Properties Moisture Content, %(w.b) Moisture Content, %(w.b) Moisture Content, %(w.b)

10 12 14 10 12 14 10 12 14

Length, mm 2.41 2.55 2.59 2.47 2.56 2.71 3.43 3.52 3.61

Breadth, mm 1.60 1.72 1.85 1.65 1.76 1.85 2.38 2.49 2.54

Thickness, mm 1.22 1.35 1.39 1.31 1.41 1.45 1.70 1.82 1.92

Equivalent diameter, Dge 1.68 1.80 1.88 1.74 1.85 1.94 2.40 2.51 2.62

Sphericity, ϕ 0.70 0.72 0.74 0.73 0.75 0.78 0.75 0.77 0.81

Surface area, mm2 8.86 10.18 11.10 9.51 10.75 11.82 18.09 19.79 21.57

Thousand kernel mass, g 3.25 3.75 4.16 3.45 3.86 4.21 4.52 5.01 5.65

Bulk density, kg/m3 711.88 691.95 649.24 705.34 658.50 612.22 775.73 724.80 710.76

True density, kg/m3 1533.14 1485.25 1342.39 1572.45 1400.00 1250.00 1791.63 1652.40 1575.60
Porosity, % 53.57 53.41 51.64 55.14 52.96 51.02 56.70 56.14 54.89
Angle of repose, º 31.79 35.37 39.00 30.54 33.82 36.12 31.38 36.86 38.30
Coefficient of static friction, μs Mild steel 0.27 0.33 0.36 0.32 0.35 0.39 0.35 0.37 0.41
Cast Iron 0.31 0.35 0.38 0.35 0.38 0.42 0.36 0.39 0.42
Coefficient of internal friction, μi Mild steel 0.49 0.52 0.54 0.52 0.56 0.59 0.38 0.41 0.46
Cast Iron 0.53 0.56 0.61 0.55 0.61 0.64 0.54 0.57 0.61
Hardness, N 24.90 22.89 19.89 23.50 22.25 18.75 19.89 18.25 17.75
Terminal Velocity, m/s 2.45 2.63 2.97 2.60 3.05 3.36 2.95 3.35 3.75
Table 4.2. Effect of Moisture Content on Engineering Properties of Millet Kernel

Little Millet Foxtail Millet Proso Millet

Properties Moisture content, % (w.b) Moisture content, % (w.b) Moisture content, % (w.b)
10 12 14 10 12 14 10 12 14
Length, mm 2.12 2.32 2.38 2.18 2.25 2.31 2.57 2.68 2.75
Width, mm 1.52 1.57 1.62 1.57 1.62 1.68 2.22 2.27 2.35
Thickness, mm 1.20 1.27 1.35 1.08 1.16 1.24 1.53 1.59 1.66
Equivalent diameter, Dge, mm 1.57 1.71 1.74 1.51 1.59 1.70 2.05 2.15 2.23
Surface area, mm2 7.64 9.20 9.52 7.16 7.94 9.07 13.20 14.52 15.62
Sphericity, ϕ 0.74 0.76 0.77 0.69 0.72 0.75 0.71 0.74 0.77
3
Bulk density, kg/m 770.03 765.32 743.65 784.18 762.63 756.58 779.13 768.67 753.09
True density, kg/m3 1678.45 1610.53 1530.32 1498.74 1444.07 1407.32 1750.32 1678.45 1610.53
Porosity, % 54.12 52.48 51.41 47.68 47.19 46.24 55.49 54.20 53.24
1000 kernel mass, g 1.51 2.58 3.12 2.71 3.45 4.01 3.21 4.15 4.76
Angle of repose, º 28.80 33.42 36.86 33.02 36.50 38.66 30.11 34.21 39.35
Coefficient of static friction, μs Mild steel 0.32 0.34 0.38 0.35 0.38 0.40 0.39 0.42 0.45
Cast iron 0.34 0.37 0.42 0.37 0.40 0.44 0.40 0.43 0.48
Coefficient of internal friction, μi Mild steel 0.52 0.55 0.58 0.55 0.58 0.61 0.58 0.61 0.64
Cast iron 0.56 0.61 0.64 0.59 0.64 0.66 0.62 0.66 0.70
Hardness, N 26.28 25.32 24.48 20.15 19.32 18.45 25.45 24.51 22.50
Terminal Velocity, m/s 2.20 2.42 2.65 2.35 2.51 2.90 2.60 3.03 3.35
4.1.1. Composition of selected millets

The thoeritical compostion was found out for 100 g of the grains to know the
amount of kernel, bran and husk present in the grain. The composition of the grains were
found out at 12 per cent moisture content (w.b.) and presented in Table 4.3.

Table. 4.3. Composition for Little, Foxtail and Proso Millet at 12 Per cent Moisture
Content (w.b.)

Millet Kernel, g Bran, g Husk, g

Little millet 71.20 5.11 23.69

Foxtail millet 72.13 5.43 22.44

Proso millet 63.62 6.99 29.39

From the Table 4.3. it is clear that proso millet has 63.32 g per 100 g per cent of
kernel whereas foxtail millet was having maximum kernel of 72.13 g per 100 g. The bran
content was ranging between 5.11 to 6.99 g per 100 g. The little millet had less amount of
bran followed by foxtail and proso millet. In the case of husk content proso millet was
possessing maximum amount of 29.39 g and little and foxtail millet were containing
22.44 and 23.69 g of husk respectively. Similar results have been reported by Seetharam
et al.(1993) for little millet and foxtail millet. However Riley et al.(1994) have reported
65 to 75 g , 5 to 10 g bran and 20 to 35 g husk for all the millets. The variation in results
may be due to the varietal differences taken for the study.

4.2. Performance Evaluation of Abrasive De-huller, Rubber Roll De-huller and

Centrifugal De-huller

About ten kilogram of little, foxtail and proso millet at three different moisture
contents (10, 12, 14 per cent, w.b.) were fed into the abrasive de-huller, rubber roll
sheller and centrifugal de-huller and the three machines were operated as per the
specifications of the manufacturers and tested for their performance The grains were
subjected to three passes in order to get maximum de-hulling efficiency. The process
variables were chosen as per the recommendations suggested by Amuthan et al. (2001)
and Singh et al. (2011). The performance results of the de-hullers are presented in
Table.4.4.
Table 4.4. De-hulling Efficiency (%) and Broken Kernel (%) of Available Machines

Moisture
Little millet Foxtail millet Proso millet
content

%, w.b. De-hulling efficiency, % De-hulling efficiency, % De-hulling efficiency, %

Rubber roll Abrasive Centrifugal Rubber roll Abrasive Centrifugal Rubber roll Abrasive Centrifugal
de-huller de-huller de-huller de-huller de-huller de-huller de-huller de-huller de-huller

10 75.25 77.85 81.50 74.33 78.75 82.55 76.75 78.32 80.67

12 72.50 75.75 77.85 72.57 73.68 78.98 75.48 76.74 82.54

14 65.75 72.50 75.00 70.58 72.86 75.76 72.75 75.78 78.84

Broken kernel, % Broken kernel, % Broken kernel, %

10 10.55 9.76 6.27 10.75 10.11 6.50 12.50 11.65 7.72

12 9.24 8.23 5.38 9.12 8.63 5.75 8.50 9.72 6.17

14 8.65 7.58 4.10 8.28 7.12 4.77 7.01 5.44 4.28

 The de-hulling efficiency recorded by rubber roll dehuller showed 75.25, 72.50
and 65.75 per cent for little millet followed by 74.33, 72.57 and 70.58 for foxtail millet,
and 76.75, 75.48 and 72.75 for proso millet at moisture content of 10, 12 and 14 per cent,
respectively. On the other hand, broken kernel per cent recorded 10.55, 9.24 and 8.65 per
cent for little millet, followed by 10.75, 9.12 and 8.28 for foxtail millet, whereas 12.50,
8.50 and 7.01 for proso millet at 10, 12 and 14 per cent, respectively.

The de-hulling efficiency assessed by abrasive dehuller displayed 77.85, 75.50

and 72.50 per cent for little millet followed by 78.75, 73.68 and 72.86 for foxtail millet,
and 78.32, 76.74 and 75.78 for proso millet at moisture content of 10, 12 and 14 per cent,
respectively. On the other hand, broken kernel per cent noted 9.76, 8.23 and 7.58 per cent
for little millet, followed by 10.11, 8.63 and 7.12 for foxtail millet, while 11.65, 9.72 and
5.44 for proso millet at 10, 12 and 14 per cent, respectively.

The de-hulling efficiency evaluated by centrifugal dehuller revealed 81.50, 77.85

and 75.00 per cent for little millet followed by 82.55, 78.98 and 75.76 for foxtail millet,
and 80.67, 82.54 and 78.84 for proso millet at moisture content of 10, 12 and 14 per cent,
respectively. On the other hand, broken kernel per cent showed 6.27, 5.38 and 4.10 per cent for
little millet, followed by 6.50, 5.75 and 4.77 for foxtail millet, while 7.72, 6.17 and 4.28
for proso millet at 10, 12 and 14 per cent, respectively.

In the present investigation, it is evident that three machines viz., rubber roller
dehuller, abrasive dehuller and centrifugal dehuller tested, of these, centrifugal dehuller
performs better pertaining to maximum de-hulling efficiency and minimum broken per
cent. Hence, it shows that impact principle has more advantage over abrasion, shear and
friction. It is thus justified that even less quantity of seeds can be dehulled using
centrifugal dehuller. In abrasive dehuller at least 10 kg grains should be fed for
continuous movement to achieving the effective processing of grains.

In abrasive dehuller and rubber roll dehuller the material subjected to heat which
may lead to degradation of the material resulting in loss of nutrients. On the other hand,
materials are not exposed to heat in impact dehuller, the centrifugal dehuller.
 Wear and tear of the rollers and abrasive stone significantly affect the results
adversely in terms of de-hulling efficiency. There are no wearing of materials in the
centrifugal dehuller, which gives long life to the machine.

Centrifugal dehuller consumes less power/energy under the study, since it works
on a single phase line than other two machines, which consumes huge power/energy and
requires three phase line, which may not be available at all place.

All the above machines need at least two passes to dehull the garins, handling of
the grains may be tiresome and tedious for labours. Hence, it is mandatory to develop
machines that will completely dehull the grains in a single pass. The double chamber
dehuller was developed to overcome the limitations of the above dehullers under the
investigation.

4.3. Design and Development of Double Chamber Centrifugal De-huller

The double chamber centrifugal de-huller consists of various components namely

feed hopper, housing, impeller, de-hulling chamber and separation chamber. All these
components were designed as per sections 3.10, 3.11, 3.12, 3.13 and 3.14. The designed
components were fabricated as per the specific dimensions and assembled.

4.4. Performance Evaluation of the Double Chamber de-huller with Millets

The developed double chamber de-huller was tested for its performance with
three millets viz., little, foxtail and proso millet. The peripheral velocity of the grain has
got the direct influence on de-hulling. The impellers were run at different speeds. During
the ideal run it was observed that the unit was having more vibration and was making
unusual sound above 3800 rpm (peripheral velocity of 27.86 m/s) and so the experiment
was restricted only to a maximum of 3800 rpm. Experiments were conducted with 3, 4, 5
and 6 vanes. It was observed that during t he trials with 6 vanes, the material did not pass
through the impeller smoothly and got stuck between the impeller and its housing as the
clearance between the vanes was only 5 mm and that hindered the entry of grains which
resulted in poor performance. Hence the experiments were restricted to 3, 4 and 5 vane
impellers only.
4.5. Performance Evaluation of Double Chamber Centrifugal De-huller for Little
Millet

Experiments were conducted with the developed double chamber de-huller by

changing the moisture content, number of vanes, peripheral velocity of the impeller and
de-hulling surfaces (Flat, Grooved and Angular), and the results were reported. The feed
rate for the experiment was maintained as 75 kg/hr. Grains entering the impeller gains
peripheral velocity while moving from the centre towards the periphery of the impeller.
The grains coming out of the impeller hits the de-hulling surface at high velocities and
due to impact the splitting of husk takes place and the kernel comes out of the husk.
The grain and husk after first impact, enters second chamber where the grains undergo
same process of de-hulling as in first chamber and then moves towards the separation
unit. The grain and husk gets separated and are collected at their respective outlets.

4.5.1. Effect of moisture content, peripheral velocity, number of vanes and de-
hulling surface on de-hulling efficiency for little millet

The De-hulling efficiency of the machine with different experimental parameters

like moisture content, peripheral velocity, number of vanes were tested and reported
graphically in Fig.4.1 and Annexure Table 1. The maximum de-hulling efficiency of
93.87 per cent was found at 10 per cent moisture content (w.b.) for a five vane impeller
with a peripheral velocity of 27.86 m/s and Grooved de-hulling surface while the lowest
de-hulling efficiency of 83.14 per cent was found at 14 per cent moisture content (w.b.)
for three vane impeller having a peripheral velocity of 24.92 m/s and flat de-hulling
surface.
Fig. 4.1. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface and
Moisture Content on De-hulling Efficiency for Little Millet
 From the Figure 4.1 and Annexure Table 1, it was clear that the de-hulling
efficiency increases with an increase in peripheral velocity and the reason for a decrease
in de-hulling efficiency with increase in moisture content may be due to the fact that
additional moisture content softens the grains and provide a cushioning effect to it by
filling the gaps within the grain with moisture. At lower moisture levels the grains are
hard in nature when subjected to impact, the husk tears apart from the kernels and so it
takes up the impact, tears the husk and brings out the kernels.

Good de-hulling efficiency was obtained when the number of vanes were
increased and due to proper guidance of the grains to the impact zone. The de-hulling
efficiency was maximum for a peripheral velocity of 27.86 m/s. The reason being the
grains when accelerated at high velocities are and possess sufficient energy to tear away
the husk from the kernels. Among all the de-hulling chambers, the Grooved surface gave
better de-hulling efficiency for little millet because the grains hit a sharp edge instead of
flat surface. The Grooveds facilitate in splitting out the husk and bringing out the kernel.
Plain surfaces of flat and angular de-hulling chamber recorded less de-hulling efficiency
since the grains were subjected to direct impact.

Table 4.5. ANOVA for the Effect of Moisture Content, Peripheral Velocity, Number
of Vanes and De-hulling Surface on De-hulling Efficiency for Little Millet

Degrees of Sum of
Source F Ratio Significance
Freedom Squares

Moisture content (M) 1 172.54627 194.2931 **

Peripheral Velocity (P) 1 184.59744 207.8631 **

Vanes (V) 1 416.32154 468.7925 **

De-hulling surface (C) 2 161.91197 91.1592 **

MxP 1 10.07111 11.3404 **

MxV 1 1.16771 1.3149 NS

PxV 1 1.13058 1.2731 NS

MxC 2 2.22547 1.2530 NS

 Degrees of Sum of
Source F Ratio Significance
Freedom Squares
PxC 2 8.03926 4.5262 *
VxC 2 37.96039 21.3724 **
MxPxV 1 0.20587 0.2318 NS
MxPxC 2 1.08968 0.6135 NS
MxPxVxC 2 2.54975 1.4356 NS

** Significant at 1% level; * Significant at 5% level

It is also clear from the ANOVA that the individual effect of moisture content,
peripheral velocity, number of vanes and de-hulling surface on de-hulling efficiency was
found to be significant at 1 per cent level. While considering the combinations moisture
content and peripheral velocity (M x P) and number of vanes and de-hulling surface
combination (V x C) gave significant results to the level of 1 per cent while peripheral
velocity and de-hulling surface (P x C) gave 5 per cent significant effect.

Singh et al. (2011) had conducted studies on de-hulling of barnyard millet using a
centrifugal de-huller and obtained a de-hulling efficiency of around 85 per cent in
7 passes with 8 m/s peripheral velocity at a moisture content of 7.5 per cent and reported
that an increase in peripheral velocity and decrease in moisture content influences the
de-hulling efficiency positively.

4.5.2. Effect of Moisture Content, Peripheral Velocity, Number of Vanes and

De-Hulling Surface on Broken Kernel for Little Millet

Experiments were conducted to study the effect of moisture content, peripheral

velocity, number of vanes and de-hulling surface on the broken kernel and are interpreted
in Fig.4.2 and Annexure Table 1. The highest broken percentage of 8.31 was found at
10 per cent moisture content (w.b.) for five vane impeller at 27.86 m/s peripheral velocity
and fitted with a flat de-hulling surface while the lowest broken percentage of 2.94 was
recorded at 14 per cent moisture content (w.b.) for three vane impeller at 24.92 m/s
peripheral velocity and fitted with a Grooved de-hulling surface. Hence it was concluded
that as the velocity increases the amount of broken kernels also increase.
 With the increase in the level of moisture percentage it was recorded that there
was decrease in percentage of broken kernels. This was due to the reason that increase the
moisture content leads to softening the grains and also produce more unhulled grains as a
resultant to produce less beroken kernels.

4.5.2.1. Effect of peripheral velocity, number of vanes, de-hulling surface and

moisture content on de-hulling efficiency for little millet

In 3 vane, at low moisture content of 10 per cent showed de-hulling per cent of
86.69 at Grooved surface, 85.84 at Angular surface and 85.19 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While de-hulling per cent per cent of 88.08, 87.03 and
86.52 at Grooved, Angular and Flat surface, respectively, were exhibited at peripheral
velocity of 26.39 m/s whereas 89.21, 88.16 and 87.88 per cent de-hulling at Grooved,
Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed de-hulling per cent of
87.93 at Grooved surface, 87.56 at Angular surface and 87.24 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 89.48, 88.76
and 87.99 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 91.15, 89.75 and 88.70 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed de-hulling per cent of
90.76 at Grooved surface, 88.94 at Angular surface and 87.79 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 91.72, 90.18
and 89.06 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 93.87, 91.69 and 89.63 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed de-hulling per cent of 85.96
at Grooved surface, 85.05 at Angular surface and 84.50 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 87.37, 86.36
and 85.72 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 88.18, 87.38 and 86.90 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.
 In 4 vane, at moisture content of 12 per cent showed de-hulling per cent of 87.23
at Grooved surface, 86.62 at Angular surface and 86.05 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 88.66, 88.21
and 87.09 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 89.41, 88.77 and 87.60 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed de-hulling per cent of 89.82
at Grooved surface, 88.23 at Angular surface and 86.78 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 91.04, 90.08
and 88.21 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 92.53, 91.29 and 88.99 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed de-hulling per cent of 85.64
at Grooved surface, 84.90 at Angular surface and 83.14 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 85.32, 85.10
and 83.75 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 85.34, 84.90 and 84.30 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent showed de-hulling per cent of 86.64
at Grooved surface, 86.17 at Angular surface and 85.51 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 86.95, 86.50
and 86.16 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 88.58, 88.02 and 87.26 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed de-hulling per cent of 88.47
at Grooved surface, 87.12 at Angular surface and 84.30 at Flat surface at peripheral
velocity of 24.92 m/s for little millet. While dehhulling per cent per cent of 89.60, 88.17
and 85.90 at Grooved, Angular and Flat surface, respectively, were exhibited at
peripheral velocity of 26.39 m/s whereas 90.90, 89.04 and 86.70 per cent de-hulling at
Grooved, Angular and Flat surface, respectively at peripheral velocity of 27.86 m/s.
4.5.2.2. Effect of peripheral velocity, number of vanes, de-hulling surface and
moisture content on broken kernel percentage for little millet

In 3 vane, at low moisture content of 10 per cent showed broken per cent of 3.52
at Grooved surface, 6.17 at Angular surface and 6.95 at Flat surface at peripheral velocity
of 24.92 m/s for little millet. While broken per cent of 3.80, 6.59 and 7.44 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.46, 6.97 and 7.94 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed broken per cent of 3.98
at Grooved surface, 6.61 at Angular surface and 7.21 at Flat surface at peripheral velocity
of 24.92 m/s for little millet. While broken per cent of 4.27, 6.85 and 7.68 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 5.08, 7.53 and 8.10 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed broken per cent of 5.28
at Grooved surface, 6.70 at Angular surface and 7.35 at Flat surface at peripheral velocity
of 24.92 m/s for little millet. While broken per cent of 5.43, 6.97 and 7.97 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 6.04, 7.46 and 8.31 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed broken per cent of 3.20 at
Grooved surface, 5.90 at Angular surface and 6.23 at Flat surface at peripheral velocity of
24.92 m/s for little millet. While broken per cent of 3.74, 6.45 and 6.80 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.09, 6.84 and 6.99 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 12 per cent showed broken per cent of 3.45 at
Grooved surface, 6.12 at Angular surface and 6.33 at Flat surface at peripheral velocity of
24.92 m/s for little millet. While broken per cent of 3.74, 6.51 and 7.07 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.02, 7.03 and 7.35 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed broken per cent of 3.94 at
Grooved surface, 6.18 at Angular surface and 6.60 at Flat surface at peripheral velocity of
24.92 m/s for little millet. While broken per cent of 4.36, 6.32 and 7.24 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.85, 6.79 and 7.53 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed broken per cent of 2.94 at
Grooved surface, 5.21 at Angular surface and 5.14 at Flat surface at peripheral velocity of
24.92 m/s for little millet. While broken per cent of 3.12, 5.76 and 5.93 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.50, 6.16 and 6.56 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent broken per cent of 3.07 at Grooved
surface, 5.30 at Angular surface and 5.38 at Flat surface at peripheral velocity of 24.92 m/s
for little millet. While broken per cent of 3.27, 5.75 and 6.27 at Grooved, Angular and
Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s whereas
3.57, 6.29 and 6.84 per cent broken at Grooved, Angular and Flat surface, respectively at
peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed broken per cent of 3.34 at
Grooved surface, 5.66 at Angular surface and 6.02 at Flat surface at peripheral velocity of
24.92 m/s for little millet. While broken per cent of 3.98, 5.98 and 6.79 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.42, 6.46 and 7.28 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.
4.5.2.3. Effect of peripheral velocity, number of vanes, de-hulling surface and
moisture content on de-hulling efficiency for foxtail millet

In 3 vane, at low moisture content of 10 per cent showed broken per cent of 85.24
at Grooved surface, 81.22 at Angular surface and 80.90 at Flat surface at peripheral
velocity of 24.92 m/s for foxtail millet. While broken per cent of 86.72, 82.57 and 79.18
at Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity
of 26.39 m/s whereas 88.17, 84.40 and 82.54 per cent broken at Grooved, Angular and
Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed broken per cent of 87.27
at Grooved surface, 83.17 at Angular surface and 80.50 at Flat surface at peripheral
velocity of 24.92 m/s for foxtail millet. While broken per cent of 88.62, 84.87 and 82.27
at Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity
of 26.39 m/s whereas 90.10, 86.01 and 83.50 per cent broken at Grooved, Angular and
Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed broken per cent of 89.32
at Grooved surface, 85.13 at Angular surface and 82.45 at Flat surface at peripheral
velocity of 24.92 m/s for foxtail millet. While broken per cent of 91.04, 87.33 and 84.85
at Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity
of 26.39 m/s whereas 92.87, 91.01 and 88.60 per cent broken at Grooved, Angular and
Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed broken per cent of 86.19 at
Grooved surface, 85.10 at Angular surface and 83.94 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 87.47, 86.17 and 85.19 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 88.58, 87.05 and 86.80 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 12 per cent showed broken per cent of 86.49 at
Grooved surface, 85.68 at Angular surface and 84.90 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 87.34, 87.00 and 86.31 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 89.50, 88.08 and 87.65 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed broken per cent of 90.88 at
Grooved surface, 89.14 at Angular surface and 87.60 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 91.01, 90.71 and 89.29 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 91.15, 90.85 and 90.50 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed broken per cent of 85.43 at
Grooved surface, 84.21 at Angular surface and 83.53 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 86.00, 85.47 and 84.33 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 87.67, 86.72 and 85.44 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent broken per cent of 86.00 at Grooved
surface, 85.15 at Angular surface and 84.57 at Flat surface at peripheral velocity of 24.92
m/s for foxtail millet. While broken per cent of 86.76, 86.03 and 85.59 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 88.00, 87.04 and 86.54 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed broken per cent of 88.20 at
Grooved surface, 87.60 at Angular surface and 86.67 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 89.64, 88.66 and 87.69 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 90.71, 89.95 and 88.61 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.
4.5.2.4. Effect of peripheral velocity, number of vanes, de-hulling surface and
moisture content on broken kernel percentage for foxtail millet

In 3 vane, at low moisture content of 10 per cent showed broken per cent of 2.78
at Grooved surface, 4.99 at Angular surface and 6.35 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 3.40, 5.75 and 6.81 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.76, 6.41 and 7.31 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed broken per cent of 3.32
at Grooved surface, 6.20 at Angular surface and 6.83 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 3.95, 6.50 and 7.22 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.57, 7.13 and 7.66 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed broken per cent of 4.27
at Grooved surface, 5.73 at Angular surface and 6.79 at Flat surface at peripheral velocity
of 24.92 m/s for foxtail millet. While broken per cent of 4.75, 6.24 and 7.29 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 5.35, 6.94 and 7.89 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed broken per cent of 2.41 at
Grooved surface, 4.81 at Angular surface and 6.08 at Flat surface at peripheral velocity of
24.92 m/s for foxtail millet. While broken per cent of 3.06, 5.64 and 6.95 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.45, 6.25 and 6.93 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 12 per cent showed broken per cent of 3.04 at
Grooved surface, 6.00 at Angular surface and 6.29 at Flat surface at peripheral velocity of
24.92 m/s for foxtail millet. While broken per cent of 3.34, 6.25 and 6.70 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.74, 6.74 and 7.24 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed broken per cent of 4.11 at
Grooved surface, 5.57 at Angular surface and 6.35 at Flat surface at peripheral velocity of
24.92 m/s for foxtail millet. While broken per cent of 4.35, 5.97 and 6.95 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.68, 6.61 and 7.61 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed broken per cent of 2.03 at
Grooved surface, 4.81 at Angular surface and 5.45 at Flat surface at peripheral velocity of
24.92 m/s for foxtail millet. While broken per cent of 2.66, 5.01 and 6.08 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 2.95, 5.69 and 6.68 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent broken per cent of 2.62 at Grooved
surface, 5.38 at Angular surface and 6.04 at Flat surface at peripheral velocity of 24.92
m/s for foxtail millet. While broken per cent of 2.93, 6.14 and 6.52 at Grooved, Angular
and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s whereas
3.16, 6.52 and 6.88 per cent broken at Grooved, Angular and Flat surface, respectively at
peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed broken per cent of 3.80 at
Grooved surface, 5.17 at Angular surface and 6.10 at Flat surface at peripheral velocity of
24.92 m/s for foxtail millet. While broken per cent of 4.24, 6.33 and 6.73 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.45, 7.21 and 8.33 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.
4.5.2.5. Effect of peripheral velocity, number of vanes, de-hulling surface and
moisture content on de-hulling efficiency for proso millet

In 3 vane, at low moisture content of 10 per cent showed broken per cent of 88.80
at Grooved surface, 84.32 at Angular surface and 83.12 at Flat surface at peripheral
velocity of 24.92 m/s for proso millet. While broken per cent of 90.19, 85.19 and 83.99
at Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity
of 26.39 m/s whereas 91.32, 86.32 and 85.12 per cent broken at Grooved, Angular and
Flat surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed broken per cent of 90.04
at Grooved surface, 85.04 at Angular surface and 83.84 at Flat surface at peripheral
velocity of 24.92 m/s for proso millet. While broken per cent of 91.59, 86.59 and 85.39 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 93.26, 88.26 and 87.06 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed broken per cent of 92.87
at Grooved surface, 87.87 at Angular surface and 86.67 at Flat surface at peripheral
velocity of 24.92 m/s for proso millet. While broken per cent of 93.83, 88.83 and 87.63 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 94.45, 90.00 and 88.80 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed broken per cent of 86.67 at
Grooved surface, 2.21 at Angular surface and 80.37 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 88.06, 83.08 and 81.24 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 89.19, 84.21 and 82.37 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 12 per cent showed broken per cent of 87.91 at
Grooved surface, 82.93 at Angular surface and 81.09 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 89.46, 84.48 and 82.64 at
Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 91.13, 86.15 and 84.31 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed broken per cent of 90.74 at
Grooved surface, 85.76 at Angular surface and 83.92 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 91.70, 86.72 and 84.88 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 92.87, 87.89 and 86.05 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed broken per cent of 84.67 at
Grooved surface, 79.65 at Angular surface and 78.08 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 86.06, 80.52 and 78.95 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 87.19, 81.65 and 80.08 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent broken per cent of 85.91 at Grooved
surface, 80.37 at Angular surface and 78.80 at Flat surface at peripheral velocity of 24.92
m/s for proso millet. While broken per cent of 87.46, 81.92 and 80.35 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 89.13, 83.59 and 82.02 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed broken per cent of 88.74 at
Grooved surface, 83.20 at Angular surface and 81.63 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 89.70, 84.16 and 82.59 at
Grooved, Angular and Flat surface, respectively, were exhibited at peripheral velocity of
26.39 m/s whereas 90.87, 85.33 and 83.76 per cent broken at Grooved, Angular and Flat
surface, respectively at peripheral velocity of 27.86 m/s.
4.5.2.6. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface and
Moisture Content on Broken Kernel Percentage for Proso Millet

In 3 vane, at low moisture content of 10 per cent showed broken per cent of 2.97
at Grooved surface, 5.62 at Angular surface and 6.40 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 3.25, 6.04 and 6.89 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.91, 6.42 and 7.39 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at low moisture content of 10 per cent showed broken per cent of 3.43
at Grooved surface, 6.06 at Angular surface and 6.66 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 3.72, 6.30 and 7.13 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.53, 6.98 and 7.55 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at low moisture content of 10 per cent showed broken per cent of 4.73
at Grooved surface, 6.15 at Angular surface and 6.80 at Flat surface at peripheral velocity
of 24.92 m/s for proso millet. While broken per cent of 4.88, 6.42 and 7.42 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 5.05, 6.91 and 7.76 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 12 per cent showed broken per cent of 2.65 at
Grooved surface, 5.35 at Angular surface and 5.68 at Flat surface at peripheral velocity of
24.92 m/s for proso millet. While broken per cent of 3.19, 5.90 and 6.25 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.54, 6.29 and 6.44 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 12 per cent showed broken per cent of 2.90 at
Grooved surface, 5.57 at Angular surface and 5.78 at Flat surface at peripheral velocity of
24.92 m/s for proso millet. While broken per cent of 3.19, 5.96 and 6.52 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.47, 6.48 and 6.80 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 12 per cent showed broken per cent of 3.39 at
Grooved surface, 5.63 at Angular surface and 6.05 at Flat surface at peripheral velocity of
24.92 m/s for proso millet. While broken per cent of 3.81, 5.77 and 6.69 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 4.30, 6.24 and 6.98 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 3 vane, at moisture content of 14 per cent showed broken per cent of 2.39 at
Grooved surface, 4.66 at Angular surface and 4.59 at Flat surface at peripheral velocity of
24.92 m/s for proso millet. While broken per cent of 2.57, 5.21 and 5.38 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 2.95, 5.61 and 6.01 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.

In 4 vane, at moisture content of 14 per cent broken per cent of 2.52 at Grooved
surface, 4.75 at Angular surface and 4.83 at Flat surface at peripheral velocity of 24.92
m/s for proso millet. While broken per cent of 2.72, 5.20 and 5.72 at Grooved, Angular
and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s whereas
3.02, 5.74 and 6.29 per cent broken at Grooved, Angular and Flat surface, respectively at
peripheral velocity of 27.86 m/s.

In 5 vane, at moisture content of 14 per cent showed broken per cent of 2.79 at
Grooved surface, 5.11 at Angular surface and 5.47 at Flat surface at peripheral velocity of
24.92 m/s for proso millet. While broken per cent of 3.43, 5.43 and 6.24 at Grooved,
Angular and Flat surface, respectively, were exhibited at peripheral velocity of 26.39 m/s
whereas 3.87, 5.91 and 6.73 per cent broken at Grooved, Angular and Flat surface,
respectively at peripheral velocity of 27.86 m/s.
Fig.4.2. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface
and Moisture Content on Broken Kernel for Little Millet
 The high impact force provided to the grains at higher peripheral velocity had
resulted in more breakage of kernels and a lower peripheral velocities the impact force
provided by the impeller was found to be inadequate to split open the husk and also to
break away the kernels.

A five vane impeller performed well by distributing the grains evenly within the
impeller which guided and accelerated all the entering grains towards the exit. Hence it
gave good impact to the grains due to which more brokens were obtained. In a 3 vane
impeller the grains were diverted from its pathway due to more space available between
the vanes and resulted in low broken kernels.

Experiments with Grooved de-hulling surface gave less broken kernels which was
due to the fact that in the Grooved surface chamber Grooveds facilitate the splitting of the
husk and reduce the broken kernels whereas as a direct high impact force to the grains by
the flat and angular surface lead to more damage resulting in more broken kernels than
the Grooved de-hulling chamber.

Table 4.6. ANOVA for the Effect of Moisture Content, Peripheral Velocity, Number
of Vanes and De-hulling Surface on Broken Kernel for Little Millet
Degrees of Sum of
Source F Ratio Significance
Freedom Squares
Moisture content (M) 1 58.4761 443.8476 **
Peripheral Velocity (P) 1 34.4357 261.3758 **
Vanes (V) 1 15.3150 116.2448 **
De-hulling surface (C) 2 394.6022 1497.563 **
MxP 1 0.04201 0.3189 NS
MxV 1 0.28418 2.1570 **
PxV 1 0.00006 0.0004 NS
MxC 2 1.14268 4.3366 NS
PxC 2 0.40411 1.5337 NS
VxC 2 3.48854 13.2394 **
MxPxV 1 0.00781 0.0593 NS
MxPxC 2 0.71452 2.7117 NS
MxPxVxC 2 0.04763 0.1808 NS
** Significant at 1% level; * Significant at 5% level
 The ANOVA data projected in Table 4.6 proves that the individual parameters
influence the results significantly at one per cent level. The combinations, number of
vanes with moisture content ( M x V) and number of vanes with de-hulling surface (V x
C) had a significant effect on broken kernels of the millets.

Omobuwajo et al.(1999) has reported that an increase in peripheral velocity and

moisture content increases the broken percentage for bread fruit.

4.5.3. Optimization of independent variables for little millet

Optimization of the parameters such as moisture content, peripheral velocity,

de-hulling surface and number of vanes was done based on the results of the de-hulling
efficiency and broken kernel obtained for little millet. It is clear from the above
discussion that changes in moisture content, peripheral velocity and number of vanes
significantly affect the de-hulling efficiency and broken percentage as per ANOVA
(Table 4.5 and 4.6). For Grooved de-hulling chamber, the de-hulling efficiency of 93.87
per cent at moisture content of 10 per cent (w.b.), while the broken kernel per cent was
6.04 at a peripheral velocity of 27.86 m/s, 5 number of vanes and Grooved de-hulling
chamber. The grains and the kernel are depicted in Plate 4.1.

I II III

Plate 4.1. Little millet-I- grain, II- kernel and III-husk

4.6. Performance Evaluation of Double Chamber Centrifugal De-huller for Foxtail
Millet

To evaluate the double chamber centrifugal de-huller at a feed rate of 75 kg/h.

Different parameters like moisture content of the grains, number of vanes, peripheral
velocity and de-hulling surface were experimented.

4.6.1. Effect of moisture content, peripheral velocity, number of vanes and de-hulling
surface on de-hulling efficiency for foxtail millet

The investigations made for finding the de-hulling efficiency with respect to
moisture content, peripheral velocity, number of vanes and de-hulling surface are
presented in Fig.4.3 and Annexure Table 2. It is evident from the figure that de-hulling
efficiency increased with decrease in moisture content. The minimum and maximum
values obtained for de-hulling efficiency was 83.53 and 92.48 per cent respectively.
The grains at 10 per cent moisture content (w.b.) with the operation parameters of
peripheral velocity of 27.86 m/s and Grooved de-hulling surface recorded maximum de-
hulling efficiency (92.48%), while grains at 14 per cent moisture content recorded the
minimum de-hulling efficiency (83.53%). The low moisture content grains are harder and
porous, so the grains will have enough potential to bear the impact force while they are
hit upon the surface of the de-hulling chamber and produce more de-hulling efficiency.
Similarly parameters like peripheral velocity, number of vanes and de-hulling surface has
a positive effect on the de-hulling efficiency of foxtail millet.
Fig 4.3. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface and
Moisture Content on De-hulling Efficiency for Foxtail Millet
Table 4.7. ANOVA for the Effect of Moisture Content, Peripheral Velocity, Number of
Vanes and De-hulling Surface on De-hulling Efficiency of Foxtail Millet

Degrees of Sum of
Source F Ratio Significance
Freedom Squares
Moisture content (M) 1 134.58936 141.8296 **
Peripheral Velocity (P) 1 197.58511 208.2142 **
Vanes (V) 1 702.12536 739.8961 **
De-hulling surface (C) 2 155.80083 82.0911 **
MxP 1 0.02739 0.0289 NS
MxV 1 22.53193 23.7440 **
PxV 1 1.27618 1.3448 NS
MxC 2 1.50618 0.7936 NS
PxC 2 1.34900 0.7108 NS
VxC 2 6.39118 3.3675 **
MxPxV 1 0.05120 0.0540 NS
MxPxC 2 0.97106 0.5116 NS
MxPxVxC 2 1.87236 0.9865 NS

** Significant at 1% level; * Significant at 5% level

The ANOVA report shown in Table 4.7 proves that the individual parameters
influence the results significantly at one 1% per cent level. While the combination effect
of moisture content with number of vanes (M×V) and number of vanes with de-hulling
surface (V×C) had a significant effect (P<0.01) on de-hulling efficiency of grains.

From Fig.4.3, it is evident, that an increase in peripheral velocity provides more

impact to de-hull the grains and thus increases the efficiency. Moist grains in presence of
more moisture content showed lower de-hulling efficiency. Due to the lack of proper
guidance of grains in three vane impeller, the de-hulling efficiency is less when compared
to five vane impeller. In Grooved surface, grains were subjected to point impact and
scratching effect to the husk has helped in easy removal of husk.
 Amuthan et al.(2004) have conducted studies on de-hulling of sunflower using a
single chamber centrifugal de-huller and obtained a de-hulling efficiency of 85 per cent at
2880 rpm and at a moisture content of 8.0 per cent.

4.7.2. Effect of peripheral velocity on broken kernel of foxtail millet at different moisture
contents

In the studies made, it was noted that the broken kernel percentage of grains was
affected by moisture, peripheral velocity, number of vanes and de-hulling surface and the
result obtained are graphically represented in Fig 4.4 and Annexure Table 2, it is clear
that broken percentage has increased in relation to moisture content. The range of values
obtained for broken percentage was between 7.89 per cent and 2.03 per cent. A maximum
broken percentage of 7.89 per cent was recorded for foxtail millet at 10 per cent moisture
content, 27.86 m/s peripheral velocity, 5 number of vanes and flat de-hulling chamber.
A minimum broken percentage of 2.03 per cent was recorded for foxtail millet at
14 per cent moisture content, 24.92 m/s peripheral velocity, 3 number of vanes and
Grooved de-hulling chamber. The cushioning effect provided due to high moisture at
high moisture contents minimized the de-hulling efficiency of the grains. However the
de-hulling efficiency was good for low moisture grains due to their hard texture.
The broken kernel increases with increase in number of vanes and peripheral velocity of
the impeller.
Fig 4.4. Effect of Peripheral velocity, Number of Vanes, De-hulling Surface and
Moisture Content on Broken Kernel for Foxtail Millet
Table 4.8. ANOVA for the Effect of Moisture Content, Peripheral Velocity, Number
of Vanes and De-hulling Surface on Broken Kernel for Foxtail Millet

Degrees of Sum of
Source F Ratio Significance
Freedom Squares

Moisture content (M) 1 23.13511 63.2347 **

Peripheral Velocity (P) 1 34.96338 95.5647 **

Vanes (V) 1 31.50416 86.1097 **

De-hulling surface (C) 2 433.67366 592.6757 **

MxP 1 0.15038 0.4110 NS

MxV 1 0.07156 0.1956 **

PxV 1 0.07468 0.2041 NS

MxC 2 0.99780 1.3636 NS

PxC 2 0.61478 0.8402 NS

VxC 2 12.01056 16.4141 **

MxPxV 1 0.00080 0.0022 NS

MxPxC 2 0.12387 0.1693 NS

MxPxVxC 2 0.15378 0.2102 NS

** Significant at 1% level; * Significant at 5% level

From the ANOVA Table 4.8 it is clearly indicates that there was significant effect
of individual parameters on the broken kernel percentage at 1 per cent level of
significance while the combinations moisture content and number of vanes (M x V), de-
hulling surface and number of vanes (V x C) gave significant effect at 1 per cent level.

The increase in number of vanes results in more broken kernels due to proper
guidance to the zone of impact. In a three vane impeller, the force required on grains was
less breakage occured. High peripheral velocities produced high broken kernels due to
the fact that grains get accelerated to high velocities at high speed impeller. The grains
were thrown at the de-hulling chamber with high peripheral velocity and so good de-
hulling results occured. Similarly more moisture in the grains provided a cushioning
effect to the grains which reduced the broken kernel percentage which in turn reduced the
de-hulling efficiency. The necessary splitting force and scratching of husk are provided
by the Grooveds present in the inner surface of the de-hulling chamber, which resulted in
less kernels compared to other two types of surfaces.

Similar results were reported by Singh et al.(2011) for barnyard millet. They have
reported that an increase in peripheral velocity increases the broken kernel and also a
decrease in moisture content increases the de-hulling efficiency of barnyard millet.

4.7.3. Optimization of independent variables for foxtail millet

Optimization of the moisture content, peripheral velocity, de-hulling surface and

number of vanes was done based on the results of the de-hulling efficiency, broken kernel
percentage obtained for foxtail millet. It is obvious from the above discussion that any
changes in moisture content, peripheral velocity and number of vanes significantly would
affect the de-hulling efficiency and broken kernel percentage as per ANOVA (Table 4.7 and
4.8). For Grooved de-hulling chamber, the de-hulling efficiency of 92.48 per cent was
recorded for foxtail millet, while the broken percentage was 5.35 for a peripheral velocity of
27.86 m/s, 5 number of vanes and Grooved de-hulling chamber. The machine performance
was best at 10 per cent moisture content, peripheral velocity of 27.86 m/s, 5 vanes and
Grooved de-hulling chamber. The de-hulled grains and kernels are depicted in Plate 4.2.

I II

Plate 4.2. Foxtail millet- I-Grain and II-Kernel

4.8. Performance Evaluation of Double Chamber Centrifugal De-huller for Proso
Millet

The developed double chamber centrifugal de-huller was operated at a feed rate of
75 kg/h and tested for proso millet by altering the parameters namely moisture content of
grains, peripheral velocity of the impeller, number of vanes and de-hulling chamber
surface. The effect of these parameters on the de-hulling efficiency and broken kernel
was studied by conducting experiments and the results are reported.

4.8.1. Effect of moisture content, peripheral velocity, number of vanes and de-
hulling chamber surface on de-hulling efficiency for proso millet

Experiments were conducted to study the effect of moisture content, peripheral

velocity, number of vanes and de-hulling surface on the de-hulling efficiency of the
grains and the results are presented in Fig.4.5 and Annexure Table 3. It is evident from
the Fig 4.5 that de-hulling efficiency had decreased with an increase in moisture content.
The de-hulling efficiency was ranging between 78.01 to 94.95 per cent. The maximum
de-hulling efficiency of 94.95 per cent was recorded at 10 per cent moisture content
(w.b.) for a peripheral velocity of 27.86 m/s and Grooved surface while a minimum value
of 78.01 per cent was recorded at 14 per cent moisture content with flat de-hulling
surface and a peripheral velocity of 24.92 m/s. Unlike in the other two millets, two main
physical properties that influence the de-hulling efficiency are hardness and porous
nature of the grains. Hardness and pores in the grains will be maximum at low moisture
contents, so the grains would have had strength to bear the impact and so good de-hulling
was obtained. Other parameters like peripheral velocity, number of vanes and de-hulling
surface had showed a positive effect on the de-hulling efficiency of proso millet.

As the number of vanes increased, the grains were properly routed towards the
hitting surface and gave good results of de-hulling efficiency. The path of the trajectory
also influenced the grains from hitting the de-hulling chamber. Likewise, increase in
peripheral velocity also had an increasing effect on the de-hulling efficiency. The
combination effect of all the above mentioned parameters along with low moisture
content gave maximum de-hulling efficiency.
Fig 4.5. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface and
Moisture Content on De-hulling Efficiency for Proso Millet
Table. 4.9. Anova for the Effect of Moisture Content, Peripheral Velocity, Number
of Vanes and De-hulling Surface on De-hulling Efficiency for Proso
Millet

Degrees of Sum of
Source F Ratio Significance
Freedom Squares

Moisture content (M) 1 1 267.52415 **

Peripheral Velocity (P) 1 1 124.43408 **

Vanes (V) 1 1 632.84961 **

De-hulling surface (C) 2 2 101.00259 **

MxP 1 1 2.23603 NS

MxV 1 1 0.27100 **

PxV 1 1 1.25238 NS

MxC 2 2 0.82908 NS

PxC 2 2 1.01809 NS

VxC 2 2 14.03796 **

MxPxV 1 1 2.39805 NS

MxPxC 2 2 0.66667 NS

MxPxVxC 2 2 0.38057 NS

** Significant at 1% level; * Significant at 5% level

The effect of each parameter on de-hulling efficiency was statistically proved by

ANOVA (Table 4.9) and Annexure Table 3. The ANOVA clearly indicates that all the
individual parameters had significant influence on the de-hulling efficiency at 1 per cent
while the combinations moisture content and number of vanes (M x V), number of vanes
and de-hulling surface (V x C) affected the de-hulling efficiency at 1 per cent level of
significance.

The above results clearly confirm that an increase in peripheral velocity, number
of vanes and decrease in moisture content of grains increased the de-hulling efficiency.
The Grooved surface gave good results for all the moisture levels.
 Similar results were reported by Singh et al.(2011) for barnyard millet. They have
reported that an increase in peripheral velocity increases the broken kernel and also a
decrease in moisture content increases the de-hulling efficiency of barnyard millet.

4.8.2. Effect of moisture content, peripheral velocity, number of vanes and de-hulling
chamber configuration on broken kernel for proso millet

A number of experiments were conducted to study the relation between moisture

content, peripheral velocity, number of vanes and de-hulling surface on the broken kernel
of the grains and the results are presented in Fig.4.6 and Annexure Table 3. The
maximum and minimum broken kernels were recorded as 7.76 per cent (at 10 per cent
moisture content, 27.86 m/s peripheral velocity, 5 number of vanes and flat de-hulling
chamber) and 3.87 per cent (at 14 per cent moisture content, 24.92 m/s peripheral
velocity, 3 number of vanes and Grooved de-hulling chamber). Increase of moisture
content reduced the broken kernel. The moisture content of 12 and 14 per cent (w.b.)
gave 30 to 55 per cent less brokens compared to that of 10 per cent (w.b.) when the
de-huller was operated at 27.86 peripheral velocity with Grooved surface for five vane
impeller. The reduction in breakage may be due to the softening of the grains at elevated
moisture contents.
Fig 4.6. Effect of Peripheral Velocity, Number of Vanes, De-hulling Surface and
Moisture Content on Broken Kernel for Proso Millet
 From the Figure 4.6 and Annexure Table 3, it is clear that a decrease in moisture
content increases broken kernels. High peripheral velocity trials conducted had resulted
in more breakage. At high peripheral velocities the grains were subjected to more
momentum than at lesser peripheral velocities. The high momentum made the grains hit
the surface of the de-hulling chamber with high impact force causing more broken
kernels. Among all the de-hulling chambers tested the Grooved one resulted in less
breakage of the kernels. This may be due to the fact that the grains were hitting only on
the wedge of the Grooveds and so they were subjected to less impact. The main reason
for this result is that the grains are rightly targeted at high speeds towards the impact zone
resulting in good splitting of the grains.

Table 4.10. ANOVA for the Effect of Moisture Content, Peripheral Velocity,
Number of Vanes and De-hulling Surface on Broken Kernel for
ProsoMillet
Degrees of Sum of
Source F Ratio Significance
Freedom Squares
Moisture content (M) 1 60.24361 210.7743 **
Peripheral Velocity (P) 1 18.74080 65.5684 **
Vanes (V) 1 5.52227 19.3208 **
De-hulling surface (C) 2 118.13841 206.6654 **
MxP 1 0.25521 0.8929 NS
MxV 1 0.15263 0.5340 **
PxV 1 0.19678 0.6885 NS
MxC 2 3.25815 5.6996 NS
PxC 2 0.60008 1.0498 NS
VxC 2 5.65334 9.8897 **
MxPxV 1 0.00245 0.0086 NS
MxPxC 2 0.13069 0.2286 NS
MxPxVxC 2 0.18311 0.3203 NS
** Significant at 1% level; * Significant at 5% level

The ANOVA Table 4.10 and Annexure Table 3 clearly indicates that all the
individual parameters had significant influence on the broken kernel at 1 per cent while
the combinations moisture content and number of vanes (M x V), number of vanes and
de-hulling surface (V x C) influenced the broken kernel at 1 per cent level of significance.

Studies conducted by Amuthan et al.(2004) on de-hulling of sunflower seeds

using a centrifugal de-huller conclude that an increase in peripheral velocity, decrease in
moisture content, increase in number of vanes influences the broken kernel percentage.

4.8.3. Optimization of independent variables for proso millet

Optimization of the moisture content, peripheral velocity, de-hulling surface and

number of vanes was done based on the results of the de-hulling efficiency and broken kernel
percentage obtained for proso millet. It is obvious from the above discussion that changes in
moisture content, peripheral velocity and number of vanes significantly affect the de-hulling
efficiency, broken kernel percentage as per ANOVA (Table 4.9 and 4.10). For Grooved
de-hulling chamber, the de-hulling efficiency of 94.95 per cent was recorded for proso millet,
while the broken kernel percentage was 5.05 per cent for a peripheral velocity of 27.86 m/s, 5
number of vanes and Grooved de-hulling chamber. The performance of the machine was the
best at 10 per cent moisture content with a peripheral velocity of 27.86 m/s, 5 vanes and
Grooved de-hulling chamber. The de-hulled grains and kernels are depicted in Plate 4.3.

I II III

Plate 4.3. Proso millet-I- grain, II- kernel and III-husk

4.9. Optimization of Process Parameters for De-hulling the Millets

The results of the double chamber centrifugal de-huller have shown that the little
millet, foxtail millet and proso millet could be de-hulled successfully. The performance
will be adjudged the best when the machine gave maximum de-hulling efficiency with
less broken kernels. From the analysis of the results it was observed that the maximum
de-hulling efficiency was 93.87, 92.48 and 94.95 per cent for little, foxtail and proso
millet at 10 per cent (w.b.) moisture content for a five vane impeller, Grooved
de-hulling chamber and a peripheral velocity of 27.86 m/s. The same set of parameters
i.e. 10 per cent moisture content (w.b.), 5 vane impeller, Grooved de-hulling chamber
surface and peripheral velocity of 27.86 m/s gave broken kernels of 6.04, 5.35 and
5.05 per cent for little, foxtail and proso millet respectively. Hence it was concluded that
the optimum parameters for the double chamber centrifugal de-huller for the selected
millets are 10 per cent moisture content (w.b.), 27.86 m/s peripheral speed and Grooved
de-hulling chamber surface.

4.10. Recovery of Kernels

The de-hulled samples obtained from the Conventional de-huller, Abrasive

de-huller, Single Chamber Centrifugal De-huller and was analyzed for the recovery of
kernels compared with that of the Double Chamber Centrifugal De-huller and the results
are presented in Table 4.11.
Table 4.11. Recovery of kernel in g per kg of millet at 10% moisture content

centrifugal de-huller, g
Rubber roller dehuller

abrasive de-huller, g

Double chamber
Conventional
Manual
Millets

Kernel with bran, g

Kernel with bran, g

Kernel with bran, g

Kernel with bran, g

Kernel, g

Kernel, g

Kernel, g

Kernel, g
Bran, g

Bran, g

Bran, g

Bran, g
Little 579.50
787.72 61.31 726.41 675.00 52.64 622.36 610.00 30.5 785.82 40.00 745.82
millet

Foxtail
749.33 59.69 689.64 666.33 53.68 612.65 635.00 45.0 595.00 787.87 50.33 737.54
millet

Proso
686.21 70.41 615.80 625.00 57.89 567.11 650.00 43.00 607.00 725.00 70.32 654.68
millet

The recovery of the kernel in the millets were analyzed by manual method using
and conventional de-huller, abrasive de-huller, rubber roller sheller and centrifugal de-
huller and the results are furnished in Table 4.10. The millets used were little, foxtail and
proso millet. It could be seen from the table that in the manual method the kernel
recovery in little, foxtail and proso millet were 726.41, 689.64 and 615 g/kg respectively.
This recovery was devoid of the husk and the bran. The recovery of kernel from the
centrifugal de-huller was divergent from the values of the manual method. The kernel
recovery was documented to be 762.43, 744.85 and 693.11 g/kg in little, foxtail and
proso millet respectively. These values were higher when compared to the manual and
conventional abrasive de-hullers, proving that the centrifugal de-huller had removed the
husk alone and not the bran and thus the recovery of the kernel was higher.

The lowest kernel recovery was found in conventional abrasive de-huller,

irrespective of the millets used. The recovery of kernel in this machine was 655.50 for
little millet, 643.50 for foxtail millet and 615.80 g/kg in proso millet.

It is clear from the table that the centrifugal de-huller had removed the husk alone
and the recovery of the kernel with bran was almost equal to the amount of kernel
available in the three millets taken for study. In the little and foxtail millet the recovery
was less by 2.5 and 0.5 per cent respectively compared to that of the available kernel
whereas it was more by 0.7 percentage in the case of proso millet. The variation in the
recovery may be due to non-uniformity of the pressure applied for removal of the husk.
Compared to the conventional de-huller the recovery in the double chamber centrifugal
de-huller was 16.33 per cent, 15.70 per cent and 22.00 per cent for little millet, foxtail
millet and proso millet respectively. The increased recovery obtained in the double
chamber de-huller was due to the retention of the bran layer in the kernel. In the
conventional de-huller not only the bran is removed but also some portion of the kernel.
That is why the kernel obtained is lesser by 9.8 per cent, 6.6 per cent and 7.6 per cent in
little, foxtail and proso millet respectively.

4.12. Nutrient availability per 100 g of de-hulled millets

Little millet Foxtail millet Proso millet

Nutrient
De-hulled Polished De-hulled Polished De-hulled Polished

Moisture(g) 6.8 7.6 6.59 7.48 6.75 7.32

Protein (g) 11.4 9.5 14.68 12.3 15.87 12.5

CHO (g) 74.8 68 65.18 60.9 76.84 70.4

Fat (g) 5.4 3.1 7 4.3 2 1.1

Ash (g) 1.6 0.73 1.9 0.72 1.52 0.7

Calcium (mg/100 g) 17.2 13.1 38.25 31 14 8.65

Phosphorous (mg/100 g) 412.4 262.7 453 290 398.2 206

Iron (mg/100 g) 13.45 9.3 4.02 2.8 2.02 0.8

Crude fibre (g/100 g) 9.24 6 9.83 6.52 5.38 2.1

Table 4.12 gives the nutrient content of de-hulled millets (de-hulled using double
chamber centrifugal de-huller) and polished millets (polished using abrasive de-huller). The
millets included little millet, foxtail millet and proso millet. The contents analyzed were
moisture, protein, carbohydrate, fat, ash, calcium, phosphorous, iron and crude fibre.
 From the table, it could be inferred that the moisture content of the polished
millets was slightly higher than the de-hulled millets. The moisture content was 6.8 and
7.6 (little millet), 6.59 and 7.48 (foxtail millet) and 6.75 and 7.32 g/100 g (proso millet)
for the de-hulled and polished millets respectively.

There was a notable difference in the protein content between the de-hulled and
polished millets. De-hulled millets had comparatively higher protein content of 11.40,
14.67 and 15.87 g/100 g for little millet, foxtail millet and proso millet respectively.
Hence it is evident that polishing of the grains had led to a loss of protein. On the
contrary, the carbohydrate cotent of the de-hulled millets was lesser when compared to
the polished millets. The highest carbohydrate content was recorded as 76.84 g/100 g in
polished proso millet whereas the lowest carbohydrate content was recorded as
60.90 g/100 g in de-hulled foxtail millet.

With reference to the fat content of the millets, the results indicated that the
process of de-hulling and polishing had a remarkable influence. The de-hulled millets had
the higher fat content, while the fat content of the polished millets were low. This might
be due to the removal of bran in the polished millets. The little millet had a fat content of
5.40 g/100 g(de-hulled), 3.10 g/100 g (polished), foxtail had a fat content of 7.0 g/100 g
and 4.3 g/100 g and the proso millet recorded a fat content of 2.0 (de-hulled) and 1.1
g/100 g (polished) respectively.

The ash content of the de-hulled little, foxtail and proso millet were 1.60, 1.90 and
1.52 g/100 g while that of polished little, foxtail and proso millet were 0.73, 0.72 and
0.70 g/100 g respectively. The above data specified that removal of bran content reduced
the ash content of the grains.

The results of the mineral composition analysis revealed that the calcium,
phosphorous and iron of de-hulled millets were higher irrespective of the type of millets
used in the study. Accordingly the calcium, phosphorous and iron content of 17.20,
412.40 and 13.45 mg/100 g for de-hulled little millet; 38.25, 453.00 and 4.02 mg/100 g
for de-hulled foxtail millet; 14.0, 398.20 and 2.02 mg/100 g for de-hulled proso millet
respectively.
 Millets owe their nutritional significance due to the presence of fibre content in it.
This study had shown that, processing methods such as de-hulling or de-hulling and
polishing created a huge impact in the crude fibre content of the millets as it is obvious
from the Table.4.12. The data pointed out that the crude fibre content of the de-hulled
little millet was 9.24 and polished little millet was 6.00; de-hulled foxtail millet was 9.83
and polished foxtail millet was 6.52; de-hulled proso millet was 5.38 and polished proso
millet was 2.10 g/100 g. These results clearly indicated that there is a significant loss of
crude fibre in the polished millets because of the removal of the bran content. Therefore
the nutritional significance will be vanished in the polished millets.

On the whole, for the significant difference in the nutrient content that existed
between the de-hulled nutrient content was due to the fact that in conventional de-hullers
the grains are subjected to rubbing which removed the bran content and the grain gets
polished. On the contrary, the centrifugal de-huller subjected the millets only to impact
and retained the bran yielding de-hulled millets alone. Therefore the centrifugal de-huller
had resulted in millets which were nutritionally superior.

4.11. Cost Economics

The cost of the newly fabricated double chamber centrifugal de-huller for millets
was estimated and found out to be Rs. 85,000/-. The capacity of the machine was 75 kg/h
for all the millets. The cost of processing one kilogram of the material was Rs.2.30 for
double chamber de-hulling while it was Rs. 37.50/- for manual de-hulling, Rs. 20/- using
abrasive dehuller, Rs 25/- using rubber roll sheller and Rs.10/- for single chamber
dehuller. The cost economics is worked out as in Annexure 4 and Annexure 5.
Summary and Conclusion
 CHAPTER V

SUMMARY AND CONCLUSION

Millet grains are considered to be one of the most important sources of dietary
proteins, carbohydrates, vitamins, minerals and fibre for people all over the world. The
nutrient composition of millets compares well with other cereals (Malleshi, 1989). The tiny
"grain" is gluten-free and contains nutrients such as magnesium, calcium, manganese,
tryptophan, phosphorus, fibre and antioxidants. Despite possessing all these extraordinary
qualities, millets are still facing a decline in cultivation and consumption due to the non-
availability of proper processing machines.

Millets are encased by a tough outer aleurone layer. The aluerone layer along with
the husk should be removed for further processing of the grains. The process of removing
the outer husk is called de-hulling. De-hulling is done conventionally by using pestle and
mortar with minimal spray of water or by using stone under runners. Both these processes
are laborious, time consuming and demands lot of patience. To overcome this situation,
millets were de-hulled using paddy de-hullers. Utilization of paddy de-hullers to dehull
millets may provide a temporary solution but it is not going to solve problem
permanently due to the reason that these machines are customized only to de-hull paddy
and not millets.

Preliminary studies were conducted for three millets at three different moisture
contents viz., 10, 12, 14 per cent moisture content (w.b.) using three types of machines
available namely rubber roll sheller, abrasive machine and centrifugal de-huller.
The maximum hulling efficiency of 86.75, 81.56 and 87.50 per cent was recorded for
little millet, foxtail millet and proso millet with a centrifugal de-huller. The abrasive
machine and rubber roll sheller gave poor results in terms of broken grain percentage in
the outlet. Three passes were required for all the millets. The capacity of the machine was
decided as 75 kg/h since there was no machine available to process materials in that
range. Based on the encouraging results obtained with a centrifugal de-huller and to
reduce the number of passes a double chamber centrifugal de-huller was developed.
 In the present investigation a double chamber centrifugal de-huller was designed,
developed and evaluated for its performance with three millets namely little, foxtail and
proso millet. The machine was tested by varying the parameters like moisture content
(10, 12 and 14 per cent w.b.), peripheral velocity (24.92, 26.39 and 27.86 m/s), number of
vanes (3,4,5 and 6) and de-hulling chamber surface (Grooved, Angular and Flat) and the
de-hulling efficiency and broken kernels were found out. The cost of the machine and
also the cost of operation of the machine was found out.

Conclusions drawn from the study

The experiments were conducted in the moisture range of 10, 12 and 14 per cent
since the moisture content of the grains during processing and handling was found to be
in that range.

The knowledge of engineering properties of minor millet grains is very important

and essential in the design of machines for primary and secondary processing products.
Hence, the engineering properties of minor millet viz., geometry and dimension, thousand
kernel mass, bulk density, true density, porosity, coefficient of static and internal friction
and terminal velocity at 10, 12 and 14 per cent moisture content (w.b.) were determined
following the standard procedures.

The sphericity of minor millet grains ranged between 0.68 to 0.74 for little millet grain;
0.69 to 0.76 for foxtail millet grains; 0.72 to 0.78 for proso millet grains respectively.
The thickness of minor millet kernels ranged between 0.65 to 0.71 for little millet grain; 0.67 to
0.73 for foxtail millet grains; 0.69 to 0.75 for proso millet grains respectively
The sphericity increased with an increase in moisture content.

Thousand grain mass for little millet, foxtail millet and proso millet at 14 per cent
moisture content (w.b.) were found to be maximum of 2, 2.91 and 4.05 g respectively for
the grains. The value of thousand kernel mass at 14 per cent for these millet grains were
2.28, 2.47, 3.48 g. It was found that the thousand kernel mass increased with moisture
content for both the grains and kernels.

The bulk density for the little millet grain was 668.57, 673.33 and 725.37 kg/m3 at
10 per cent moisture content and the minimum bulk density was 643.82, 651.46 and
698.42 kg/m3 at 14 per cent moisture content (w.b.). The bulk density for the kernels
were 685.35, 695.34 and 751.48 kg/m3 at 10 per cent moisture content (w.b.). The bulk
density was found to decrease with an increase in moisture content for both the grains
and the kernels. Similar trends in bulk density were noted down for foxtail millet and
proso millet.

The true density for little millet, foxtail millet and proso millet grains were
1345.25, 1356.28 and 1401.27 kg/m3 at 10 per cent (w.b.). Maximum true density for
little, foxtail and proso millet kernels were 1405.04, 1426.64 and 1437.50 kg/m3
respectively at 10 per cent moisture content. Similar trends were found out for 12 per cent
and 14 per cent moisture contents for the kernels. The true density was found to decrease
with increase in moisture content.

The porosity values at 10 per cent (w.b.) moisture content for little millet, foxtail
millet and proso millet grains were 48.21, 52.36 and 57.75 per cent, respectively. At 14
per cent (w.b.) moisture content, the porosity values for these minor millet grains were
45.36, 48.12 and 53.26 per cent respectively. The porosity values decreased with increase
in moisture content. It was found that the porosity value decreased with an increase in
moisture content for the grains and the kernels.

The porosity values at 10 per cent (w.b.) moisture content for little millet, foxtail
millet and proso millet kernels were 39.54, 43.01 and 47.25 per cent respectively. At 14
per cent (w.b.) moisture content, the porosity values for these minor millet grains were
35.25, 40.18 and 43.26 per cent respectively. The porosity values decreased with increase
in moisture content. It was found that the porosity value decreased with an increase in
moisture content for the grains and kernels also.

The coefficient of static friction for mild steel surface increased with moisture
content. The coefficient of static friction for little millet, foxtail millet and proso millet at
10 per cent moisture content against mild steel surface was found to be 0.27, 0.32 and
0.35. At 14 per cent (w.b.) moisture content, the static friction values for these minor
millets were 0.36, 0.39 and 0.41 respectively.

The coefficient of static friction for mild steel surface increased with moisture
content for the kernels also. The coefficient of static friction for little millet, foxtail millet
and proso millet at 10 per cent moisture content against mild steel surface was found to
be 0.32, 0.35 and 0.39. At 14 per cent (w.b.) moisture content, the static friction values
for these minor millets were 0.38, 0.40 and 0.45 respectively. Similar increasing trends in
static friction were observed for little millet, foxtail millet and proso millet grains and
kernels against a cast iron surface also.

The coefficient of internal friction for mild steel surface increased with moisture
content. The coefficient of static friction for little millet, foxtail millet and proso millet at
10 per cent moisture content against mild steel surface was found to be 0.53, 0.55 and
0.56. At 14 per cent (w.b.) moisture content, the coefficient of internal friction values for
these minor millets was 0.61, 0.64 and 0.67 respectively.

The coefficient of static friction for mild steel surface increased with moisture
content for the kernels also. The coefficient of static friction for little millet, foxtail millet
and proso millet at 10 per cent moisture content against mild steel surface was found to
be 0.32, 0.35 and 0.39. At 14 per cent (w.b.) moisture content, the static friction values
for these minor millets were 0.38, 0.40 and 0.45. Similar increasing trends in static
friction were observed for little millet, foxtail millet and proso millet grains and kernels
against a cast iron surface also.

The terminal velocity at 10 per cent (w.b.) moisture content for little millet, foxtail
millet and proso millet grains were 2.45, 2.60 and 2.95 m/s. At 14 per cent (w.b.) moisture
content, the terminal velocity values for these minor millet grains were 2.90, 3.30 and 3.75
m/s. The terminal velocity values increased with an increase in moisture content.

The terminal velocity values at 10 per cent (w.b.) moisture content for little millet,
foxtail millet and proso millet kernels were 2.15, 2.50 and 3.25 m/s. At 14 per cent (w.b.)
moisture content, the porosity values for these minor millet grains were 3.25, 3.65 and
4.10 per cent respectively. The terminal velocity values increased with increase in moisture
content for the kernels also.

The theoretical composition of little millet was 71.20 g kernel, 5.11 g bran and
23.49 g husk; for foxtail millet it was 72.13 kernel, 5.43 g bran and 29.39 g husk; for
proso millet 63.62 g kernel, 6.99 g bran and 29.39 g husk. The kernel content was more
for foxtail millet.
 The individual components were designed and then they were assembled together.
The major component parts of the de-huller are impeller, casing and separation chamber.
The impeller parts were fabricated with plates. The impeller consists of crown plate, base
plate and vanes. All the parts were fabricated with mild steel plate. The casings were
moulded with cast iron and then machined. The blower was fabricated and fitted to the
machine for separating the grains and husk.

The centrifugal de-huller developed in this study was evaluated for its
performance in terms of hulling efficiency and broken grain percentage. The de-hulling
efficiency and broken grain percentage was found out and reported.

Maximum hulling efficiency of 93.87, 92.89 and 94.95 per cent was recorded for little,
foxtail and proso millet respectively. The broken grain percentage recorded for the given
hulling efficiency was 6.04, 5.35 and 5.05 per cent respectively. The parameters that influenced
the machine performance were optimized based on the performance results. 10 per cent
moisture content, 27.86 m/s peripheral speed, 5 number of vanes and grooved casing
combination gave best results for little, foxtail and proso millet.

The recovery of the kernels by manual and mechanical methods was compared. It was
found that there was maximum recovery of kernels for double chamber centrifugal de-huller
when compared to the conventional abrasive de-huller. A 3 hp three phase motor was used for
the studies. The capacity of the machine was 75 kg/h and the cost of the machine is
Rs. 85,000/-. The cost of processing one kilogram of grains was Rs. 2.30/- .

Future line of work

The machine can be semi-automated. Regulators may be provided for changing the
speeds. The material consumption and the speed of operation may be cut down. The machine
can be slightly modified for de-hulling barnyard and kodo millet. Attachment of a grader to
the machine. Attachment of destoner and grader to the machine.
Reference
 REFERENCES

Abdelrahman,A., R.C.Hoseney and E.Varriano-Marston. 1984. The proportions and

chemical compositions of hand-dissected anatomical parts of pearl millet. Journal
of Cereal Science. 2: 127-133.

Altuntas, E., Yildiz, M., 2007. Effect of moisture content on some physical and
mechanical properties of faba bean (Vicia faba L.) grains. Journal of Food
Engineering. 78: 174–183.

Amuthan, G., Subramanian, P., and P.T. Palaniswamy. 2001. Modifiction made on
centrifugal paddy sheller for sunflower seed shelling. Agricultural mechanization
in Asia, Africa and Latin America. 32(3): 51-53.

AOAC. 1980. Official Methods of Analysis. 14th ed. Association of Official Analytical
Chemists, Arlington, Virginia, USA.

AOAC. 1995. Official Method of Analysis. 16th ed. Association of Official Analytical
Chemists, Arlington, Virginia, USA.

Arora, S. and S.S.Kumar. 1999. Separation behaviour and aerodynamic characteristics of

different varieties of wheat. Agricultural Engineering Today 23 (1&2).

Aviara, N.A., M.I.Gwandzang and M.A.Hque. 1999. Physical Properties of guna Seeds.
Journal of Agricultural Engineering Research 73(2): 105-111.

Balasubramaniam, S. and R. Viswanathan. 2010. Influence of moisture content on

physical properties of minor millets. Journal of Food Science and Technology.
47 (3): 279-284.

Bassey, M.W. and O.G.Schmidt. 1989. Abrasive-disk dehullers in Africa -from research
to dissemination Ottawa, Canada, Centre de recherche pour le dévelopement
International. pp.97.

Bautista,R.C., T.J.Siebenmorgen., A.Mauromustakos. and R.M.Burgos. 2004. Small

Sample Mill Protocol Development: Evaluation of a Genogrinder 2000 Wells
Rice Research Studies AAES Research Series. pp.529.
Brubaker, J.E. and J.Pos. 1965. Determining static coefficients of friction of grains on
structural surfaces. Transactions of the American Society of Agricultural Engineers
8(1):53-55.

Bruno Eck (1973). Fans. Design and Operation of Centrifugal Axial Flow and Cross
Flow Fans. Pergamon Press, Newyork. (pp.89–106)

Burton,G.W., A.T.Wallace and K.O.Rachie. 1972. Chemical Composition and Nutritive

Value of pearl millet (P. typhoides (Burm) Stapf and EC Hubbard) grain. Crop
Science. 12:187.

Calisir, S., T. Marakoglu, H. Ogut and O. Ozturk. 2005. Physical properties of rapeseed
(Brassica napus oleifera L.). Journal of Food Engineering. 69: 61–66.

Camara,E. and M.A.Amaro. 2003. Nutritional aspect of zinc availability. International

Journal of Food Sciences and Nutrition. 54: 143–151.

Carman, K. 1996. Some physical properties of lentil seeds. Journal of Agricultural

Engineering Research 63:87-92.

Chakraverty Amalendu, Arun S. Mujumdar, G.S. Vijaya Raghavan and Hosahalli

S. Ramaswamy (2003), pp: 253-455, Marcel Dekker Inc. Newyork

Chandrasekar, V. and R.Viswanathan. 1999. Physical and thermal properties of coffee.

Journal of Agricultural Engineering Research. 73: 227-234.

Chung, J.H and L.R.Verma. 1989. Determination of friction coefficients of beans and
peanuts. Transactions of the American Society of Agricultural Engineers 32(2):
745-750.

Church H. Austin. 1972. Centrifugal Pumps and Blowers, Kreiger Publishing Company.

Deshpande,S.D., S.Bal. and T.P.Ojha. 1993. Physical properties of soybean. Journal of

Agricultural Engineering Research. 56: 89-98.

Dutta, S.K., V.K.Nema and R.K.Bhardwaj. 1988. Physical properties of gram. Journal of
Agricultural Engineering Research 30(4):259-268.

Douglas C. Doehlert and Michael S. McMullen (2001) . Optimizing Conditions for

Experimental Oat Dehulling. Cereal Chem. 78(6):675–67.
Eastman, P. 1980. An end to pounding a new metrical flour milling system in use in Africa.
Ottawa, Canada, Centre de recherche pour le développement international : 63.

Egli,I., L.Davidsson, M.A.Juillerat, D.Barclay, and R.F.Hurrell. 2002. The influence of

soaking and germination on the phytase activity and phytic acid content of grains
and seeds potentially useful for complementary feeding. Journal of Food Science.
67(9): 3484–3488.

FAO. 1995. ‘Sorghum and millets in human nutrition’. (FAO Food and Nutrition Series,
No. 27) Rome.

Geervani, P. and B. O. Eggum. 1989. Nutrient composition and protein quality of minor
millets. Plant Foods Hum Nutr. 39: 201-208.

Gopalan,C., B.V.Ramashastri, and B.V.Balasubramanian. 1989. Nutritive value of

Indian foods. New Delhi: ICMR Offset Press: 47-91.

Gorial,B.Y. and J.R. Callaghan. 1990. Aerodynamic properties of Grain/Straw Materials.

Journal of Agricultural Engineering Research. 46: 275-290.

Gupta,R.K. and S.K.Das. 1998. Friction Coefficients of Sunflower Seed and Kernel on
Various Structural Surfaces. Journal of Agricultural Engineering Research.
71: 175-180.

Hazbavi (2013), Moisture dependent physico-mechanical properties of Iranian Okra

(ablemoschus esculentus L.) seed. Vol: 12(42). 6098-6106.

Hsu, M.H., J.D.Mannapperuma and R.P.Singh. 1991. Physical and thermal properties of
pistachios. Journal of Agricultural Engineering Research. 49:311-321.

Hulse,J.H., E.M.Laing, and O.E.Pearson. 1980. Sorghum and the millets: their composition
and nutritive value. New York, Academic Press: 997.

Janusz Laskowski, Grzegorg Lysiak, Stanislaw Skonecki. 2005. ‘Mechanical properties

of granular agro-materials and food powders for industrial practice - Part II-
Material properties for grinding and agglomeration.

Joshi, D.C., S.K.Dass and R.K.Mukherjee. 1993. Physical properties of pumpkin seeds.
Journal of Agricultural Engineering Research. 54:219-229.
Khurmi, R.S. and J.K. Gupta 2012. A Textbook of Machine Design. Eurasia Publishing
House Pvt, Ltd. Ram Nagar. New Delhi.

Kingsly,A.R.P., D.B.Singh, M.R.Manikantan and R.K.Jain. 2006. Moisture dependent

physical properties of dried pomogranate seeds (Anardana). Journal of Food
Engineering. 75: 492-496.

Konak, M., K.Carman and C.Aydin. 2002. Physical properties of chickpea seeds.
Biosystem Engineering. 82(1): 73-78.

Laskowski,J. and S.Skonecki. 1997. Influence of moisture on the physical properties and
parameters of the compression process of legumes’ seeds. Institute of Agrophysics.
11: 245-256.

Lorenz,K. 1983. Tannins and phytate content in proso millets (Panicum miliuceum).
Cereal Chemistry. 60: 424-426.

Lorenz,K., W.Dilsaver, and L.Bater. 1979. Proso millets. Milling characteristics, proximate
composition nutritive value of ours. Cereal Chemistry. 57(1): 14-16.

Makanjuola, G. A. 1972. A study of some of the physical properties of melon seeds.

Journal of Agricultural Engineering Research, 17 (1), 128 – 137.

Malleshi, N.G. 1989. Processing of small millets for food and industrial uses. In: Small
millets in global agriculture. Seetharam, A., K.W. Rieley and G. Harianarayana
(Eds.). Oxford and IBH Publishing Co. Pvt. Ltd., New Delhi.

Malleshi,N.G. and N.A.Hadimani. 1993. Nutritional and technological characteristics of

small millets and preparation of value-added products from them. In: K.W. Riley,
S.C. Gupta,A. Seetharam, and J.N. Mushonga (Eds.), Advances in small millets
(pp.271–287). New Delhi: Oxford and IBH Publishing Co Pvt. Ltd.

Malleshi,N.G., H.S.R.Desikachar, and S.Venkat Rao. 1986. Protein quality evaluation

of a weaning food based on malted ragi and green gram. Qual. Plant. Plant Foods
for Human Nutrition. 36: 223-230.

Mamiro,P.R.S., J.Van Camp, S.Mbithi-Mwikya, and A.Huyghebaert. 2001. In vitro

extractability of calcium, iron and zinc in finger millet and kidney beans during
processing. Journal of Food Science. 66(9): 1271–1275.
Mbithi-Mwikya,S., J.Van Camp, Y.Yiru, and A.Huyghebaert. 2000. Nutrient and
antinutrient changes in finger millet (Eleusine coracana) during sprouting.
Lebensmittel Wissenschaft und Technologie. 33(1): 9–14.

Mohsenin, N. N. 1986. Physical properties of plant and animal materials, Gordon and
Breach Science Publisher. New York.

Mohsenin, N.N. 1970. Physical properties of plant and animal Materials. Gordon and
Breach Science Publishers. New York.

Munck,L., K.E.Bach Knudsen, and J.D.Axtell. 1982. Milling processes and products as
related to kernel morphology. In: L.W. Rooney and D.S. Murty, éds. Proceedings
of the International Symposium on SorghumGrain quality, Hyderabad, India,
28-31 October 1981:200-210.

Murthy, C.T., and S.Bhattacharya. 1998. Moisture dependant physical and uniaxial compression
properties of black pepper. Journal of Food Engineering. 37(2): 193-205.

Nelson, S.O.1980. Moisture-dependent kernel and bulk density relationship for wheat and
corn. Transactions of the ASAE. 23(1): 139-143.

Nimkar, P.M. and Chattopadhyay, P.K. (2001). Some physical properties of green gram.
Journal of Agricultural Engineering Research 80 (2): 183-189.

Ogut, H.1998. Some physical properties of white lupin. Journal of Agricultural

Engineering Research. 56: 273–277.

Oloso, A.O. and B. Clarke. 1993. Some aspects of strength properties of cashew nuts.
Journal of Agricultural Engineering Research.55:27-43.

Okokon. F.B, E.Ekpenyong, C.Nwaukwa, N. Akpan and F.I.Abam (2010), Analysis of

the impact forces on Melon seeds during shelling”. Agricultural Engineering
International: CIGR Ejournal. Manuscript. FP 1276.Vol. XII. March

Omobuwajo, T.O., H.C. Ikegwuoha, O.A. Koya and M.T. Ige (1999). Design,
construction and testing of a dehuller for African breadfruit (Treculia Africana)
seeds. Journal of Food Engineering. 42: 173-176

Palnabendu and Sarkar Sahadev. 2007. Statistical Concept and Application.

Perten, H. 1983. Practical experience in processing and use of millet and sorghum in
Senegal and Sudan. Cereal Foods World. 28: 680-683.

Phirke, P.S., N.G.Bhole and S.H. Adhaoo.1996. Friction Coefficient of Pigeon Pea Grain
with Abrasive Surfaces Used for Dehulling. Agricultural Mechanization in Asia,
Africa and Latin America. 27(4):27-31.

Radhakrishnan, M.R and J.Sivaprasad. 1980. Tannin content of sorghum varieties and their
role in iron bioavailability. Journal of Agriculture and Food Chemistry. 28: 55-57.

Rebacca Chin, Nicholas Matlashewski and Kristian Swan (2012). Bree 495 Design 3, Final
Design Report Presented to Dr.Grant Clark.

Rooney, L.W. and C.M. McDonough.1986. Food quality and consumer acceptance of
pearl millet. Proceedings of the International Pearl Millet Workshop Hyderabad,
Inde, 7-11 april 1986, p. 43-61. Patancheru, Inde, ICRISAT.

Sacilinkt, K., R.Ozturk and R.Kesikin. 2002. Some physical properties of hemp seeds.
Biosystems Engineering. 86(2):191-198.

Sadasivam, D. and A. Manikam. 1996. Biochemical Methods. New International

Pvt. Ltd.

Sahay, K.M. and K.K.Singh. 2010. Unit operations of Agricultural processing. Vikas
Publishing Hose Private Limited. Delhi-110 014.

Sankara Rao, D.S. and Y.G.Deosthale. 1980. Effect of pearling on mineral and trace
element composition and ionizable iron content of sorghum. Nutrition Reports
International.22: 723-728

Seetharam, A., K.W.Riley and G.Harinarayana.1989. Small millets in global agriculture.

New Delhi, Inde, Oxford and IBH.

Sharma, A. and A.C.Kapoor.1996. Levels of antinutritional factors in pearl millet as

affected by processing treatments and various types of fermentation. Plant Foods
for Human Nutrition. 49:241-252.

Sharma,S.K., R.K.Dubey and C.K.Teckchandain.1985. Engineering properties of black

gram, soybean and green gram. Proc. ISAE (India). 3:181-185
Shellard, J.E. and R.M.Mcmillan. 1978. Aerodynamic properties of threshed wheat
material. Journal of Agricultural Engineering Research. 23:273-281.

Shepherd, H. and R.K.Bhardwaj .1986. Moisture-dependent physical properties of pigeon

pea. Journal of Agricultural Engineering Research.35:227–234.

Singh K P, Mishra H N, Saha S. 2009. Moisture–dependent properties of barnyard millet

grain and kernel. J. Food Engineering. 96, 598–606.

Singh, K, P., H,N.Mishra and S.Saha. (2011). Design, development and evaluation of
barnyard millet dehuller. Journal of Agricultural Engineering. 48(3): 17-25.

Singh, K.K. and T.K.Goswami. 1996. Physical properties of cumin seed. Journal of
Agricultural Engineering Research. 64 (2): 93-98.

Singh, K.P., H.N. Mishra and Supradip Saha. 2010. Moisture-dependent properties of
barnyard millet grain and kernel. Journal of Food Engineering. 96: 598–606.

Singh Ranjeet and Mangaraj Sukhdev. 2013. Development and Evaluation of Centrifugal
Sheller for Muskmelon Seed. International Research Journal of Biological
Sciences. Vol 2(3): 7 – 10.

Sreenarayanan, V. V., R .Viswanathan, and V. Subramaniyan. 1988. Physical and thermal

properties of soybean. Journal of Agricultural Engineering, ISAE, 25(4):76-82.

Sripriya, G., A.Usha and T.S. Chandra.1997. Changes in carbohydrate, free amino acids,
organic acids, phytate and HCl extractability of minerals during germination and
fermentation of finger millet (Eleusine coracana). Food Chemistry. 58:345-350.

Stewart, B.R., Q.A.Hossin and O.R.Kunze.1969. Friction Coefficients of Sorghum Grain on

Steel, Teflon and Concrete Surfaces. Transactions of the ASAE. 12(4):415-418.

Subramanian, S., R. Viswanathan. .2003. Thermal properties of minor millet grains and
flours. Biosyst Eng. 84:289–296.

Sullins, R.D. and L.W.Rooney. 1971. Physical changes in the kernel during reconstitution
of sorghum grain. Cereal Chemistry. 48:567.

Suthar, S.H and S.K.Das. 1996. Some Physical Properties of karingda (Citrullus lanatus
Mansf) Seeds. Journal of Agricultural Engineering Research 65(1):15-22.
Thajudhin Sherif, J.1996. Studies on the influence of extruder and processing parameters
on the characteristics of cassava flour extrudates. Unpublished Ph.D., Thesis.
Tamilnadu Agricultural University. India.

Thompson, L.U. 1993. Potential health benefits and problems associated with antinutrients in
foods. Food Research International. 26:131-149.

Thompson, R.A. and G.W. Isaac .1967. Porosity determination of grains and seeds with
air compression pyconometer. Transactions of the ASAE. 10(5): 693-696

Thompson, S.A. and I.J.Ross.1983. Compressibility and Frictional Coefficient of Wheat.

Transactions of the ASAE. 26(4):1171-1180.

Tsang-Mui-Chung, L.R.Verma and M.E.Wright. 1984. A Device for Friction Measurement of

Grains. Transactions of the ASAE. 27(6):1938-1941.

Usha,A., G.Sripriya and T. S.Chandra. 1996. The effect of fermentation on the primary
nutrients in foxtail millet (Setaria italica). Food Chemistry. 56:381-384.

Viswanathan, R., P.T. Palanisamy, L.Gothandapani and V.V.Sreenarayanan. 1996.

Physical properties of neem nut. Journal of Agricultural Engineering Research.
63:19-26.

William C.Osborne. 1977. Fans. Pergamon press. Newyork.

Annexures
 Annexure 1
Effect of peripheral speed on de-hulling efficiency, broken kernel percentage and for little millet

Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat

3 VANE 24.92 86.69 85.84 85.19 3 VANE 24.92 85.96 85.05 84.50 3 VANE 24.92 85.64 84.90 83.14
26.39 88.08 87.03 86.52 26.39 87.37 86.36 85.72 26.39 85.32 85.10 83.75
27.86 89.21 88.16 87.88 27.86 88.18 87.38 86.90 27.86 85.34 84.90 84.30
4 VANE 24.92 87.93 87.56 87.24 4 VANE 24.92 87.23 86.62 86.05 4 VANE 24.92 86.64 86.17 85.51
26.39 89.48 88.76 87.99 26.39 88.66 88.21 87.09 26.39 86.95 86.50 86.16
27.86 91.15 89.75 88.70 27.86 89.41 88.77 87.60 27.86 88.58 88.02 87.26
5 VANE 24.92 90.76 88.94 87.79 5 VANE 24.92 89.82 88.23 86.78 5 VANE 24.92 88.47 87.12 84.30
26.39 91.72 90.18 89.06 26.39 91.04 90.08 88.21 26.39 89.60 88.47 85.90
27.86 93.87 91.69 89.63 27.86 92.53 91.29 88.99 27.86 90.90 89.24 86.70
Broken kernel percentage
Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat
3 VANE 24.92 3.52 6.17 6.95 3 VANE 24.92 3.20 5.90 6.23 3 VANE 24.92 2.94 5.21 5.14
26.39 3.80 6.59 7.44 26.39 3.74 6.45 6.80 26.39 3.12 5.76 5.93
27.86 4.46 6.97 7.94 27.86 4.09 6.84 6.99 27.86 3.50 6.16 6.56
4 VANE 24.92 3.98 6.61 7.21 4 VANE 24.92 3.45 6.12 6.33 4 VANE 24.92 3.07 5.30 5.38
26.39 4.27 6.85 7.68 26.39 3.74 6.51 7.07 26.39 3.27 5.75 6.27
27.86 5.08 7.53 8.10 27.86 4.02 7.03 7.35 27.86 3.57 6.29 6.84
5 VANE 24.92 5.28 6.70 7.35 5 VANE 24.92 3.94 6.18 6.60 5 VANE 24.92 3.34 5.66 6.02
26.39 5.43 6.97 7.97 26.39 4.36 6.32 7.24 26.39 3.98 5.98 6.79
27.86 6.04 7.46 8.31 27.86 4.85 6.79 7.53 27.86 4.42 6.46 7.28
 Annexure 2
Effect of peripheral speed on dehulling efficiency, broken kernel percentage for foxtail millet
Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat
3 VANE 24.92 85.24 81.22 80.90 3 VANE 24.92 86.19 85.10 83.94 3 VANE 24.92 85.43 84.21 83.53
26.39 86.72 82.57 79.18 26.39 87.47 86.17 85.19 26.39 86.00 85.47 84.33
27.86 88.17 84.40 82.54 27.86 88.58 87.05 86.80 27.86 87.67 86.72 85.44
4 VANE 24.92 87.27 83.17 80.50 4 VANE 24.92 86.49 85.68 84.90 4 VANE 24.92 86.00 85.15 84.57
26.39 88.62 84.87 82.27 26.39 87.34 87.00 86.31 26.39 86.76 86.03 85.59
27.86 90.10 86.01 83.50 27.86 89.50 88.08 87.65 27.86 88.00 87.04 86.54
5 VANE 24.92 89.32 85.13 82.45 5 VANE 24.92 90.88 89.14 87.60 5 VANE 24.92 88.20 87.60 86.67
26.39 91.04 87.33 84.85 26.39 91.01 90.71 89.29 26.39 89.64 88.66 87.69
27.86 92.87 91.01 88.60 27.86 91.15 90.85 90.50 27.86 90.71 89.95 88.61
Broken kernel percentage
Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat
3 VANE 24.92 2.78 4.99 6.35 3 VANE 24.92 2.41 4.81 6.08 3 VANE 24.92 2.03 4.81 5.45
26.39 3.40 5.75 6.81 26.39 3.06 5.64 6.95 26.39 2.66 5.01 6.08
27.86 3.76 6.41 7.31 27.86 3.45 6.25 6.93 27.86 2.95 5.69 6.68

4 VANE 24.92 3.32 6.20 6.83 4 VANE 24.92 3.04 6.00 6.29 4 VANE 24.92 2.62 5.38 6.04
26.39 3.95 6.50 7.22 26.39 3.34 6.25 6.70 26.39 2.93 6.14 6.52
27.86 4.57 7.13 7.66 27.86 3.74 6.74 7.24 27.86 3.16 6.52 6.88
5 VANE 24.92 4.27 5.73 6.79 5 VANE 24.92 4.11 5.57 6.35 5 VANE 24.92 3.80 5.17 6.10
26.39 4.75 6.24 7.29 26.39 4.35 5.97 6.95 26.39 4.24 6.33 6.73
27.86 5.35 6.94 7.89 27.86 4.68 6.61 7.61 27.86 4.45 7.21 8.33
 Annexure 3

Effect of peripheral speed on de-hulling efficiency, broken kernel percentage for proso millet

Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat

3 VANE 24.92 88.80 84.32 83.12 3 VANE 24.92 86.67 82.21 80.37 3 VANE 24.92 84.67 79.65 78.08
26.39 90.19 85.19 83.99 26.39 88.06 83.08 81.24 26.39 86.06 80.52 78.95
27.86 91.32 86.32 85.12 27.86 89.19 84.21 82.37 27.86 87.19 81.65 80.08
4 VANE 24.92 90.04 85.04 83.84 4 VANE 24.92 87.91 82.93 81.09 4 VANE 24.92 85.91 80.37 78.80
26.39 91.59 86.59 85.39 26.39 89.46 84.48 82.64 26.39 87.46 81.92 80.35
27.86 93.26 88.26 87.06 27.86 91.13 86.15 84.31 27.86 89.13 83.59 82.02
5 VANE 24.92 92.87 87.87 86.67 5 VANE 24.92 90.74 85.76 83.92 5 VANE 24.92 88.74 83.20 81.63
26.39 93.83 88.83 87.63 26.39 91.70 86.72 84.88 26.39 89.70 84.16 82.59
27.86 94.95 90.00 88.80 27.86 92.87 87.89 86.05 27.86 90.87 85.33 83.76
Broken kernel percentage
Grooved Angular Flat Grooved Angular Flat Grooved Angular Flat
3 VANE 24.92 2.97 5.62 6.40 3 VANE 24.92 2.65 5.35 5.68 3 VANE 24.92 2.39 4.66 4.59
26.39 3.25 6.04 6.89 26.39 3.19 5.90 6.25 26.39 2.57 5.21 5.38
27.86 3.91 6.42 7.39 27.86 3.54 6.29 6.44 27.86 2.95 5.61 6.01
4 VANE 24.92 3.43 6.06 6.66 4 VANE 24.92 2.90 5.57 5.78 4 VANE 24.92 2.52 4.75 4.83
26.39 3.72 6.30 7.13 26.39 3.19 5.96 6.52 26.39 2.72 5.20 5.72
27.86 4.53 6.98 7.55 27.86 3.47 6.48 6.80 27.86 3.02 5.74 6.29
5 VANE 24.92 4.73 6.15 6.80 5 VANE 24.92 3.39 5.63 6.05 5 VANE 24.92 2.79 5.11 5.47
26.39 4.88 6.42 7.42 26.39 3.81 5.77 6.69 26.39 3.43 5.43 6.24
27.86 5.05 6.91 7.76 27.86 4.30 6.24 6.98 27.86 3.87 5.91 6.73
 Annexure 4

Estimation of Cost of The Double Chamber Dehuller

A. Material cost
S.No. Materials Quantity Amount, Rs.
1. 18 gauge M.S. sheet , 8’ x 4’ 3 No’s 8000
2. 16 gauge M.S. sheet , 8’ x 4’ 1 No 4000
3. ‘L’ Angles (32x6x6 mm) 3 No’s 3600
4. Plate (25.4x 6.25 mm) 2 No’s 1000
5. Pulleys 8 No’s 5000
6. 1.5” O.D MS shaft 4m 1000
7. 3 H.P three phase motor (1440 rpm) 1 no 10000
8. Bearings with Pillow blocks 8 No’s 5000
9. Bolts & Nuts 4 kg 320
10. Gas cutting and machine charges - 8000
11. Wire 5m 440
12. Main switch box 1 no 1000
13. Primer and Paint 1.5 liter 6000
14. Labour cost 1 unit 12000
15 Belt 8 No’s 4000
16 Casting 1 unit 8000
17 Blower box 1 No 3500
Unforeseen expenditure 4000
Total cost including labour Rs. 84860/-
Final cost rounded off value Rs. 85,000/-
 Annexure 5

Estimation of Cost of Operation of The Double Chamber Centrifugal Dehuller

Assumptions:

Capacity of the centrifugal dehuller = 75 kg/h

Life span of the dehuller (N) = 10 years
Initial cost of the thresher (C) = Rs.85,000/-
Annual usage (B) = 300 days,
1 shift per day i.e. 8 hours / day or
2400 hours/ year
Interest (i) = 12.5 per cent per annum
Electrical charges = Rs.7/unit

Fixed cost per year

i (i+1)n
Fixed cost of the unit (C) = ---------------- x C
(i+1)n + 1

0.125 (0.125+1)10
= -- ----------------------- x 85000
(0.125+1)10 +1

= Rs. 8123/ year

Fixed cost = Rs. 3.38/h

Variable cost

• Repairs and maintenance = 5 per cent of the initial cost

= Rs.85,000 × 0 .05

= Rs.1.46/h
Cost of labour = (Number of labours × Cost of labour ×
N umber of shifts )= (6×200×1)

= Rs. 1200/day

= Rs. 150/h

3. Cost of electricity = Rs. 7/h

Horse power of the dehuller = 3 hp

= 2.23 kW

Cost of electrical power = Rs. 7/ unit

Cost of electricity = (2.23×7) = 15.61/h

Total variable cost = (Repair and maintenance

+ Cost of labour

+ Cost of electricity)

= 1.46 +150 +15.61= Rs. 167.06/h

Total operating cost = fixed cost + variable cost

= Rs. 3.38 + Rs. 167.06

= Rs. 170.44

~ Rs. 170/h

Cost of operation = Rs. 170/h

Cost for 1 kg = Rs. 170/75

= Rs. 2.27/kg
Conventional method

Two labours can de-hull 2 kg/h, so per day two labours can de-hull 8 kg/day

Cost involved for manual hulling with two labours = Rs. (2 x 200)/8

= Rs. 50/kg

Quantity of grains dehulled = 10 kg

Time required to dehull 1 kg of grains manually = 0.50 h

Time required to dehull 75 kg of grains = 37.50 h

Time required to dehull 75 kg of grains by machine = 1h

Savings in time = (37.50 – 1)

= 36.50 h

= (36.50/1) x 100

= 3650 %

Savings in labour cost = Rs. 50 – 2.27

= Rs. 47.73

= 2102.64 %

From the cost economics calculation, it was observed that the cost involved for
processing one kilogram of millet is Rs.2.30/- and cost involved for processing 1 kg of
millet by conventional means is Rs.50/-. It was realized that there is a saving of
2102 percent in cost of processing and 3650 percent in time involved.

Rate of De-hulling using abrasive, rubber roll shellers were Rs. 15/kg and
Rs. 20/kg.
Publications
384 in Biosciences 8(2), Print : ISSN 0974-8, 384-389,
Trends Trends in2015
Biosciences 8 (2), 2015

Influence of Moisture Content on Physical Properties of Little Millet

GANESAN, S. AND VARADHARAJU. N
Dept. of Food and Agricultural Process Engineering,
Agricultural Engineering College and Research Institute,
Tamilnadu Agricultural University,
Coimbatore, Tamil Nadu
email : [email protected]

ABSTRACT Hardness, coefficients of internal friction, static

coefficient of friction and terminal velocity can be
The physical properties of the little millet grain were
used in design and development of dehuller for
found out using various methods. The geometric mean
diameter, sphericity, grain surface area, 1000 grain millets. Bulk density, true density and porosity, are
mass, true density (toluene displacement method), important factors in designing of storage structures.
terminal velocity, dynamic angle of repose, coefficient The angle of repose of the grains and kernels can
of internal friction, coefficient of static friction at be used for designing the bins, silos, hoppers and
different surfaces mild steel and cast iron surfaces), storage structures (Mohsenin, 1993). Many
hardness of the grain increased with an increase of researchers have been determined the properties
moisture content from 11.11% to 16.28% moisture of different agricultural produces like millet
content dry basis. However bulk density decreased (Balasubramaniam and Viswanathan, 2010). Singh,
with an increase in moisture content. True density et al., 2009 determined the physical properties like
(toluene displacement method) was found to reduce bulk density and friction coefficient of millets.
for all experimental moisture range indicating the
presence of void space inside the grain and kernel. L Length, mm LMWG Little millet whole
grain
W Width, mm LMK Little millet kernel
Key words Moisture content, little millet.
T Thickness, mm LMWMS Little millet whole
grain on mild steel
Panicum sumatrense commonly known as little Ø Sphericity LMWCI Little millet whole
millet is commonly grown millet crop in the arid grain on cast iron
and semiarid region of the world. It was ρt True density V Terminal velocity,
(kg/ m3) m/s
domesticated in India and can grow in altitudes upto
ρb Bulk density, H Dynamic angle of
2100 m. It is mostly grown in south and central kg/ m3 repose,
India (Hulse, et al., 1980). This crop is harvested ε Porosity, % µi Coefficient of
at 25% (db) moisture content and reduced to internal friction
11.11% (db) for safe storage of the grains. It w.grain Whole grain µs Coefficient of static
contains protein 7.7%, fibre 7.7% and minerals friction
1.5%. It also contains iron, B vitamins, H Hardness
phosphorous, carbohydrates and fat in good The study focuses on some properties of little
proportions at a moisture content of 13.64% millet grain and kernel namely the mean diameter,
(db)(Malleshi, 1993). It is gluten free is sphericity, bulk density, true density (toluene
recommended for diabetic patients. The presence displacement and proximate composition methods),
of all the required nutrients in millet make them coefficient of internal friction, coefficient static
suitable for large scale utilization in the manufacture friction at different surfaces, dynamic angle of
of baby food, snack foods, dietary food etc. from repose, 1000 grain weight, surface area, porosity,
grain, kernel and flour (Balasubramanium and terminal velocity and hardness of little millet grain
Viswanathan, 2010). The study of physical, and kernel at moisture range 11.11% to 16.28%(db).
aerodynamic and mechanical properties of food
grain is important and essential in the design of MATERIALS AND METHODS
processing machines, storage structures and
Little millet (Panicum sumatrense) was
processes. The shape and size are important in the
obtained from the local market and cleaned by
separation from foreign material and in the design
manual and mechanical means to remove all foreign
and development of grading and sorting machineries.
 GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Little Millet 385

matter, broken and immature grains. Physical Porosity and true density: Porosity was
properties for raw little millet was investigated in determined by using the porosity apparatus
the simulated moisture content range of 11.1– (Sreenarayanan, et al., 1988). True density was
16.28% (db), since harvesting, transportation, calculated from the measured values of bulk density
storage and dehulling operations of minor millets and porosity using the following equation
are performed in this range. Grain samples of desired (Mohsenin, 1986).
moisture content levels were prepared by adding Bulk density(kg/m3)
calculated amount of distilled water in accordance
t = ———————— 3.3
with the following equation (Sacilink, et al., 2002)
and mixed thoroughly. (1-Porosity)

——————— 3.4
—————— 3.1

where, Q is the weight of water to be added where,  is the porosity (%),

(g); Wi is the initial weight of grain sample (g); mi b is the bulk density of grain, kg/m3, ñt is the
is the initial moisture content of grain sample (% true density (kg/m3)
db) and mf is the final moisture content of grain
sample (% db). Grain samples were sealed in Angle of repose: The filling angle of repose
polyethylene bags of 300  thickness. The samples was determined using the apparatus described by
were kept in a refrigerator maintained at 4±1°C for Sreenarayanan, et al., 1988. Angle of repose was
a minimum period of 10 days to reach uniform determined from the height and diameter of the
moisture content. The moisture contents of the naturally formed heap of grains on a circular plate.
samples were equilibrated to 11.1, 13.6, Coefficient of static and internal friction: The
16.28% db as per the procedures of AACC, coefficient of static friction apparatus consisted of
2000. Before each experiment, samples were a frictionless pulley fitted on a frame, a cylindrical
equilibrated at room temperature (30±2°C) for 2 h container of negligible weight with both ends
and the moisture was checked using the standard opened, loading pan and test surfaces
oven-dry method. All tests were conducted in the (Sreenarayanan et al. (1988). The container, placed
laboratory at an ambient temperature of about on the mild steel surface was filled with a known
(30±2°C) and relative humidity of 55–65%. quantity of material and weights were added to the
loading pan until the container began to slide. The
Thousand kernel weight: A grain weight of experiment was performed with millets of different
approximately 1 kg was roughly divided in to 10 moisture contents. The coefficient of static friction
equal portions and then 1000 numbers of grains was calculated as the ratio of weights added
were randomly picked from each portion, and (frictional force) and material mass (normal force)
weighed on a digital electronic balance. The as given below.
measurement was repeated for 5 times and the mean
value was taken.
Bulk density: A cubical container of volume ——————-—— 3.5
1.52x10-3 was filled with the grain and gently
tapped. Bulk density was calculated as the ratio of Coefficient of static and internal friction: The
weight of millets to volume of container. Average coefficient of static friction apparatus consisted of
of 5 replications was taken. Care was taken to avoid a frictionless pulley fitted on a frame, a cylindrical
compaction of grains in the container and filled to container of negligible weight with both ends
full volume. opened, loading pan and test surfaces (Singh, et
M al., 2009). The container, placed on the mild steel
b = ——— ———— 3.2 surface was filled with a known quantity of material
V and weights were added to the loading pan until
the container began to slide. The experiment was
where  is the bulk density in kg/m3, V is the performed with little millets of different moisture
volume in m3, M is the mass in kg. contents. The coefficient of static friction was
386 Trends in Biosciences 8 (2), 2015

calculated as the ratio of weights added (frictional throughout the entire cross sectional area of the
force) and material mass (normal force) as given vertical tube, wire meshes were placed at the bottom
below. For the measurement of coefficient of of the vertical tube. The air velocity was measured
internal friction, apparatus consisting of 2 cylinders, using a hot wire anemometer. The grains were fed
with one being a stationary and other one to slide into the vertical tube and the air velocity was
on the stationary one (diameter:height, 50:55 mm) adjusted such that the air stream carried no material
was used. Through a pulley-rope arrangement with away and the velocity at which 90 per cent of the
a loading pan, the top cylinder was made to slide grains which floated in air was taken as the terminal
on the stationary cylinder. Both the cylinders were velocity of the grains. The experiment was repeated
placed in position and sample was filled without for five times and the mean was recorded.
any compaction. Grain mass sample contained in
top cylinder (normal force) acts on top layer of RESULTS AND DISCUSSION
same material in the bottom stationary cylinder and Thousand kernel weight: The 1000 kernel weight,
incremental load applied in loading pan to slide the increased linearly with an increase in moisture
top pan (frictional force). The force required to content (Fig. 1). Singh and Goswami, 1996 for
slide the empty top cylinder was subtracted from cumin seeds found such linear relationship between
frictional force to get actual frictional force to thousand seed/kernel weight and moisture content.
overcome the friction due to material. Mean of 5
Bulk density: The bulk density, true density
replications was taken. The coefficient of internal
and porosity and grain hardness decreased
friction was calculated as the ratio of weights added
linearly with an increase of moisture content from
(frictional force) and grain mass (normal force) as
11.11 per cent to 16.28 per cent. A similar
given below.
relationship was reported by Balasubramaniam and
Viswanathan, 2010 for millets. The increase in
weight is due to change in cell structure of grains
———————— 3.6
and absorption of water between them. These
results agree with the findings of Suthar and Das,
where, e, i are the coefficient of static and 1996 with pumpkin seeds. Joshi, et al., 1993 also
internal friction, Fi , Fe are the frictional force for reported a linear decrease in porosity with increase
static and internal friction (g) and Ni, Ne are the in pumpkin seed moisture content. The small
normal force for static and internal friction(g) and rupturing forces at higher moisture content might
Ni , Ne are the normal force for static and internal have resulted from the fact that the seed became
friction (g) soft and more sensitive to cracking at high moisture.
Grain hardness: A single grain hardness of little Angle of repose, static friction and internal
millet at different moisture contents was measured friction: Angle of repose, static friction and
using texture analyser (Model TA-HD, Surrey, UK). coefficient of internal friction increased with an
Little millet was placed individually in its natural increase in moisture content (Fig. 1). The reason
resting position on the platform and load was applied for increased friction coefficient at higher moisture
until the grain crushed. The measurement was content may be due to water present in grains
repeated for 5 different samples and the mean was offering increased adhesive force on contact
taken. Experimental conditions followed were load surface. At lower moisture content, angle of repose
cell: 5 kg, test mode: measure force in compression, was low for little millet. Little millet at moisture
test option: return to start, pre test speed: 2 mm/ content (11.1%, db) showed lesser value and also
sec, test speed: 0.1 mm/s, post test speed: 2 mm/ exhibited a higher value of coefficient of static
sec, strain: 70% and test probe: P 4. friction at higher moisture content (16.28% db).
Similar results were reported by Kukelko, et al.,
Terminal Velocity
1988 for rapeseeds and by Chung and Verma, 1989
Terminal velocity of little millet was for beans. This may be due to higher cohesion
determined at different moisture levels. The terminal exhibited by little millets at higher moisture content.
velocity apparatus consisted of a vertical acrylic A higher coefficient of internal friction was offered
tube of 50 mm diameter and 400 mm length by little millet. Similar trends were also reported by
attached to a centrifugal blower powered by a 0.33 Kingsly, et al., 2006 that the hardness of dried
hp, 2850 rpm motor. To get uniform air velocity pomegranate seeds decreased linearly from 87 N
GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Little Millet 387

Mild steel Galvanized iron

388 Trends in Biosciences 8 (2), 2015

Fig. 1. Influence of moisture content on physical properties of foxtail millets

Sphericity (w.grain)
= 0.0289 x + 0.653 (R2 = 0.99) (4.1)
Sphericity (kernel)
= 0.03x + 0.620 (R2 = 1.00) (4.2)
2
M1000(w.grain) = 0.125 M + 2.1233 (R = 0.99) ….(4.3)
M1000(kernel) = 0.215 M + 1.6233 (R2 = 0.99) ….(4.4)
(b) (w.grain) = -12.035 M + 688.88 (R2 = 0.95) …..(4.5)
2
(b) (kernel) = -11.99 M + 698 (R = 0.99) …..(4.6)
(w.grain) = -1.415 M + 44.49 (R2 = 0.99) ......(4.7)
 (kernel)
= -1.415 M + 44.49 (R2 = 0.99) .....(4.8)
2
(t) (w.grain) = -29.07 M + 1259.40 (R = 0.99) …..(4.9)
(t) (kernel) = -24.645 M + 1308.90 (R2 = 0.99) ....(4.10)
(s) (w.grain) = 2.01 M + 26.537 (R2 = 0.96) …..(4.11)
(s) (kernel) = 2.79 M + 31.163 (R2 = 0.96) …..(4.12)
(i) (w.grain) = 0.0250 M + 0.4667 (R2 = 0.98) …..(4.13)
(i) (kernel) = 0.0300 M + 0.550 (R2 = 1.00) …..(4.14)
Hardness(w.grain) = -0.900 M + 27.160 (R2 = 0.99) …..(4.15)
2
Hardness(kernel) = -1.505 M + 25.723 (R = 0.96) …..(4.16)
Terminal Velocity (w.grain) = -0.900 M + 27.160 (R2 = 0.99) ….(4.17)
Terminal Velocity (Little) = -1.505 M + 25.723 (R2 = 0.96) ….(4.18)

to 50 N for the corresponding increase in moisture moisture content. Bulk density, true density,
content from 6.0 to 18.13% (db). porosity and grain hardness were found to be
inversely proportional to moisture content of little
Thousand kernel weight, angle of repose,
millet at the moisture range (11.1 to 25% db)
coefficient of static and internal friction for little
studied.
millet were found to be directly proportional to
 GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Little Millet 389

LITERATURE CITED technological characteristics of small millets and

preparation of value added products from them. In:
AACC 2000. Approved methods, American Association of Advances in small millets.
Cereal Chemists. 10th edn, Moisture determination, 44–
15A, Minnasota, St Polo. Mohsenin, N.N. 1986. Physical properties of plant and
animal ma- terials-II. Gorden and Beach Sci Publ,
Balasubramaniam, S. and Viswanathan, R. 2010. Influence London.
of moisture content on physical properties of minor
millets, J. Food science and technology 47(3): 279-284. Nimkar, P.M. and Chattopadhyaya, P.K. 2001. Some
physical properties of green gram. J. Agric. Eng. Res.
Chung, J.H. and Verma, L.R. 1989. Determination of friction 80:183–189.
coefû cients of beans and peanuts. Trans ASAE 32:745–
750. Sacilink, K., Ozturk, R. and Kesikin, R. 2002. Some physical
properties of hemp seeds. Biosyst Eng 86:191–198.
Joshi, D.C., Das, S.K. and Mukherjee, R.K. 1993. Physical
properties of pumpkin seeds. J Agric Eng Res 54:219– Shepherd, H. and Bhardwaj, R.K. 1986. Moisture-
229. dependent physical properties of pigeon pea. J. Agric.
Eng. Res. 35:227–234.
Kingsly, A.R.P., Singh, D.B., Manikantan, M.R. and Jain,
R.K. 2006. Mois- ture dependent physical properties Singh, K.K. and Goswami, T.K. 1996. Physical properties
of dried pomogranate seeds (anardana). J Food Eng of cumin seed. J. Agric. Eng. Res. 64:93–98.
75:492–496. Singh, K.P., Mishra, H.N. and Supradip Saha 2009. Moisture
Kukelko, D.A., Jayas, D.S., White, G. and Britton, M.G. dependent properties of barnyard millet grain and kernel.
1988. Physical properties of Canola (rape seed) meal. Journal of Food Engineering 96: 598-606.
Canadian Agric Eng 29(2):32–35. Sreenarayanan, V.V., Viswanathan, R. and Subramanian, V.
Malleshi, N.G. and Hadimani, N.A. 1993. Nutritional and 1988. Physical and thermal properties of soybean. J.
Agric. Eng. 25:76–82.

Received on 16-12-2014 Accepted on 23-12-2014

390 in Biosciences 8(2), Print : ISSN 0974-8, 390-395,
Trends Trends in2015
Biosciences 8 (2), 2015

Influence of Moisture Content on Physical Properties of Foxtail Millet

GANESAN, S. AND VARADHARAJU. N
Dept. of Food and Agricultural Process Engineering, Agricultural Engineering College and
Research Institute, Tamilnadu Agricultural University, Coimbatore, Tamil nadu
email : [email protected]

ABSTRACT storage structures and processes. The shape and

size are important in the separation from foreign
The physical properties of the foxtail millet grain were
material and in the design and development of
found out using various methods. The geometric mean
diameter, sphericity, grain surface area, 1000 grain
grading and sorting machineries. Hardness,
mass, true density (toluene displacement method), coefficients of internal friction, static coefficient
terminal velocity, dynamic angle of repose, coefficient of friction and terminal velocity can be used in
of internal friction, coefficient of static friction at design and development of dehuller of millets. Bulk
different surfaces mild steel and cast iron surfaces), density, true density and porosity, are important
hardness of the grain increased with an increase of factors in designing of storage structures
moisture content from 11.11% to 16.28% moisture (Mohsenin, 1986). The angle of repose of the
content dry basis. However bulk density decreased grains and kernels can be used for designing the
with an increase in moisture content. True density bins, silos, hoppers and storage structures. Many
(toluene displacement method) was found to reduce researchers have been determined the properties
for all experimental moisture range indicating the of different agricultural produces like millet
presence of void space inside the grain and kernel. (Balasubramaniam and Viswanathan, 2010). Singh,
et al., 2009 determined the physical properties like
Key words Moisture content, foxtail millet bulk density and friction coefficient of millets.
The study focuses on some properties of
Setaria italica commonly known as foxtail foxtail millet grain and kernel namely the mean
millet is commonly grown millet crop in the arid diameter, sphericity, bulk density, true density
and semiarid region of the world. It is the second (toluene displacement and proximate composition
most widely planted species of millet, and most methods), coefficient of internal friction, coefficient
important in East Asia. It is an annual grass with static friction at different surfaces, dynamic angle
slim, vertical, leafy stems which can reach a height of repose, 1000 grain weight, surface area,
of 120 to 130 cm. The seed head is dense, hairy porosity, terminal velocity and hardness of foxtail
panicle. The seeds are small and around 2 mm in millet grain and kernel at moisture range 11.11% to
diameter, and are encased in a thin papery hull which 16.28%(db).
is easily removed in threshing. This crop is
harvested at 25 per cent (db) moisture content and MATERIALS AND METHODS
reduced to 11.11 per cent (db) for safe storage of
Foxtail millet was obtained from the local
the grains. It contains protein 11.0 per cent, fibre 6
market and cleaned by manual and mechanical
percent and fat 4 per cent. It also contains iron, B
means to remove all foreign matter, broken and
vitamins, phosphorous and carbohydrates and fat
immature grains. Physical properties for raw minor
in good proportions at a moisture content of
millets were investigated in the simulated moisture
13.64% (db). It is gluten free is recommended for
content range of 11.1–16.28% (db), since
diabetic patients.
harvesting, transportation, storage and dehulling
The presence of all the required nutrients in operations of the millet are performed in this range.
millets make them suitable for large scale utilization Grain samples of desired moisture content levels
in the manufacture of baby food, snack foods, were prepared by adding calculated amount of
dietary food etc. from grain, kernel and flour distilled water in accordance with the following
(Malleshi, 1993). equation (Sacilink et al. 2002) and mixed thoroughly.
The study of physical, aerodynamic and
mechanical properties of food grain is important
and essential in the design of processing machines, ————— 3.1
 GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Foxtail Millet 391

where, Q is the weight of water to be added where,  is the porosity (%),

(g); Wi is the initial weight of grain sample (g); mi b is the bulk density of grain, kg/m3, t is the
is the initial moisture content of grain sample (% true density (kg/m3)
db) and mf is the final moisture content of grain
sample (% db). Grain samples were sealed in Angle of repose: The filling angle of repose
polyethylene bags of 300  thickness. The samples was determined using the apparatus described by
were kept in a refrigerator maintained at 4±1°C for Sreenarayanan et al. (1988). Angle of repose was
a minimum period of 10 days to reach uniform determined from the height and diameter of the
moisture content. The moisture contents of the naturally formed heap of grains on a circular plate.
samples were equilibrated to 11.1, 13.6, Coefficient of static and internal friction: The
16.28% db as per the procedures of AACC coefficient of static friction apparatus consisted of
(2000). Before each experiment, samples were a frictionless pulley fitted on a frame, a cylindrical
equilibrated at room temperature (30±2°C) for 2 h container of negligible weight with both ends
and the moisture was checked using the standard opened, loading pan and test surfaces
oven-dry method. All tests were conducted in the (Sreenarayanan et al. (1988). The container, placed
laboratory at an ambient temperature of about on the mild steel surface was filled with a known
(30±2°C) and relative humidity of 55–65%. quantity of material and weights were added to the
loading pan until the container began to slide. The
Thousand kernel weight: A grain weight of experiment was performed with millets of different
approximately 1 kg was roughly divided in to 10 moisture contents. The coefficient of static friction
equal portions and then 1000 numbers of grains
was calculated as the ratio of weights added
were randomly picked from each portion, and (frictional force) and material mass (normal force)
weighed on a digital electronic balance. The as given below.
measurement was repeated for 5 times and the mean
value was taken.
Bulk density: A cubical container of volume ——————-—— 3.5
1.52x10-3 was filled with the grain and gently
tapped. Bulk density was calculated as the ratio of
Coefficient of static and internal friction: The
weight of millets to volume of container. Average
coefficient of static friction apparatus consisted of
of 5 replications was taken. Care was taken to avoid
a frictionless pulley fitted on a frame, a cylindrical
compaction of grains in the container and filled to
container of negligible weight with both ends
full volume.
opened, loading pan and test surfaces (Singh, et
M al., 2009). The container, placed on the mild steel
b = ————— —————— 3.2 surface was filled with a known quantity of material
V and weights were added to the loading pan until
the container began to slide. The experiment was
where  is the bulk density in kg/m3, V is the performed with foxtail millets of different moisture
volume in m3, M is the mass in kg. contents. The coefficient of static friction was
calculated as the ratio of weights added (frictional
Porosity and true density: Porosity was force) and material mass (normal force) as given
determined by using the porosity apparatus below. For the measurement of coefficient of
(Sreenarayanan et al. 1988). True density was internal friction, apparatus consisting of 2 cylinders,
calculated from the measured values of bulk density with one being a stationary and other one to slide
and porosity using the following equation on the stationary one (diameter:height, 50:55 mm)
(Mohsenin 1986). was used. Through a pulley-rope arrangement with
Bulk density(kg/m3) a loading pan, the top cylinder was made to slide
ñt = ——————— —————3.3 on the stationary cylinder. Both the cylinders were
(1-Porosity) placed in position and sample was filled without
any compaction. Grain mass sample contained in
top cylinder (normal force) acts on top layer of
———— 3.4 same material in the bottom stationary cylinder and
incremental load applied in loading pan to slide the
392 Trends in Biosciences 8 (2), 2015

top pan (frictional force). The force required to increased linearly with an increase in moisture
slide the empty top cylinder was subtracted from content (Fig. 1). Singh and Goswami, 1996 for
frictional force to get actual frictional force to cumin seeds found such linear relationship between
overcome the friction due to material. Mean of 5 thousand seed/kernel weight and moisture content.
replications was taken. The coefficient of internal
Bulk density: The bulk density, true densiy and
friction was calculated as the ratio of weights added
porosity and grain hardness decreased linearly
(frictional force) and grain mass (normal force) as
with an increase of moisture content from 11.11
given below.
per cent to 16.28 per cent. A similar relationship
was reported by Balasubramaniam and
Viswanathan, 2010 for millets. The increase in
———————— 3.6 weight is due to change in cell structure of grains
and absorption of water between them. These
where, e, i are the coefficient of static and results agree with the findings of Suthar and Das,
internal friction, Fi , Fe are the frictional force for 1996 with pumpkin seeds. Joshi, et al., 1993 also
static and internal friction (g) and Ni , Ne are the reported a linear decrease in porosity with increase
normal force for static and internal friction(g) and in pumpkin seed moisture content. The small
Ni , Ne are the normal force for static and internal rupturing forces at higher moisture content might
friction (g) have resulted from the fact that the seed became
soft and more sensitive to cracking at high moisture.
Grain hardness: A single grain hardness of foxtail
millet at different moisture contents was measured Angle of repose, static friction and internal
using texture analyser (Model TA-HD, Surrey, UK). friction: Angle of repose, static friction and
Minor millet was placed individually in its natural coefficient of internal friction increased with an
resting position on the platform and load was applied increase in moisture content (Fig. 1). The reason
until the grain crushed. The measurement was for increased friction coefficient at higher moisture
repeated for 5 different samples and the mean was content may be due to water present in grains
taken. Experimental conditions followed were load offering increased adhesive force on contact
cell: 5 kg, test mode: measure force in compression, surface. At lower moisture content, angle of repose
test option: return to start, pre test speed: 2 mm/ was low for foxtail millet. Foxtail millet at content
sec, test speed: 0.1 mm/s, post test speed: 2 mm/ (11.1%, db) showed lesser value and also exhibited
sec, strain: 70% and test probe: P 4. a higher value of coefficient of static friction at
higher moisture content (16.28% db). Similar
Terminal Velocity
results were reported by Kukelko, et al., 1988 for
Terminal velocity of foxtail millet was rapeseeds and by Chung and Verma, 1989 for beans.
determined at different moisture levels. The terminal This may be due to higher cohesion exhibited by
velocity apparatus consisted of a vertical acrylic minor millets at higher moisture content. A higher
tube of 50 mm diameter and 400 mm length coefficient of internal friction was offered by foxtail
attached to a centrifugal blower powered by a 0.33 millet. Similar trends were also reported by Kingsly,
hp, 2850 rpm motor. To get uniform air velocity et al., 2006 that the hardness of dried pomegranate
throughout the entire cross sectional area of the seeds decreased linearly from 87 N to 50 N for the
vertical tube, wire meshes were placed at the bottom corresponding increase in moisture content from
of the vertical tube. The air velocity was measured 6.0 to 18.13% (db).
using a hot wire anemometer. The grains were fed
Thousand kernel weight, angle of repose,
into the vertical tube and the air velocity was coefficient of static and internal friction for foxtail
adjusted such that the air stream carried no material
millet was found to be directly proportional to
away and the velocity at which 90 per cent of the
moisture content. Bulk density, true density,
grains which floated in air was taken as the terminal
porosity and grain hardness were found to be
velocity of the grains. The experiment was repeated
inversely proportional to moisture content of foxtail
for five times and the mean was recorded.
millet at the moisture range (11.1 to 25% db)
RESULTS AND DISCUSSION studied.

Thousand kernel weight: The 1000 kernel weight,

GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Foxtail Millet 393

Mild steel Galvanized iron

394 Trends in Biosciences 8 (2), 2015

Fig. 1. Influence of moisture content on physical properties of foxtail millets

Sphericity (w.grain)
= 0.0289 x + 0.653 (R2 = 0.99) (4.1)
Sphericity (kernel)
= 0.03x + 0.620 (R2 = 1.00) (4.2)
2
M1000(w.grain) = 0.125 M + 2.1233 (R = 0.99) ….(4.3)
M1000(kernel) = 0.215 M + 1.6233 (R2 = 0.99) ….(4.4)
(b) (w.grain) = -12.035 M + 688.88 (R2 = 0.95) …..(4.5)
2
(b) (kernel) = -11.99 M + 698 (R = 0.99) …..(4.6)
(w.grain) = -1.415 M + 44.49 (R2 = 0.99) ......(4.7)
 (kernel)
= -1.415 M + 44.49 (R2 = 0.99) .....(4.8)
2
(t) (w.grain) = -29.07 M + 1259.40 (R = 0.99) …..(4.9)
(t) (kernel) = -24.645 M + 1308.90 (R2 = 0.99) ....(4.10)
(s) (w.grain) = 2.01 M + 26.537 (R2 = 0.96) …..(4.11)
(s) (kernel) = 2.79 M + 31.163 (R2 = 0.96) …..(4.12)
(i) (w.grain) = 0.0250 M + 0.4667 (R2 = 0.98) …..(4.13)
(i) (kernel) = 0.0300 M + 0.550 (R2 = 1.00) …..(4.14)
Hardness(w.grain) = -0.900 M + 27.160 (R2 = 0.99) …..(4.15)
2
Hardness(kernel) = -1.505 M + 25.723 (R = 0.96) …..(4.16)
Terminal Velocity (w.grain) = -0.900 M + 27.160 (R2 = 0.99) ….(4.17)
Terminal Velocity (Little) = -1.505 M + 25.723 (R2 = 0.96) ….(4.18)
 GANESAN and VARADHARAJU, Influence of Moisture Content on Physical Properties of Foxtail Millet 395

REFERENCES preparation of value added products from them. In:

Advances in small millets.
AACC 2000. Approved methods, American Association of
Cereal Chemists. 10th edn, Moisture determination, 44– Mohsenin, N.N. 1986. Physical properties of plant and
15A, Minnasota, St Polo. animal ma- terials-II. Gorden and Beach Sci Publ,
London.
Balasubramaniam, S. and Viswanathan, R. 2010. Influence
of moisture content on physical properties of minor Nimkar, P.M. and Chattopadhyaya, P.K. 2001. Some
millets, J. Food science and technology 47(3): 279-284. physical properties of green gram. J. Agric. Eng. Res.
80:183–189.
Chung, J.H. and Verma, L.R. 1989. Determination of friction
coefû cients of beans and peanuts. Trans ASAE 32:745– Sacilink, K., Ozturk, R. and Kesikin, R. 2002. Some physical
750. properties of hemp seeds. Biosyst Eng 86:191–198.

Joshi, D.C., Das, S.K. and Mukherjee, R.K. 1993. Physical Shepherd, H. and Bhardwaj, R.K. 1986. Moisture-dependent
properties of pumpkin seeds. J Agric Eng Res 54:219– physical properties of pigeon pea. J. Agric. Eng. Res.
229. 35:227–234.

Kingsly, A.R.P., Singh, D.B., Manikantan, M.R. and Jain, Singh, K.K. and Goswami, T.K. 1996. Physical properties
R.K. 2006. Mois- ture dependent physical properties of cumin seed. J. Agric. Eng. Res. 64:93–98.
of dried pomogranate seeds (anardana). J Food Eng Singh, K.P., Mishra, H.N. and Supradip Saha 2009. Moisture
75:492–496. dependent properties of barnyard millet grain and kernel.
Kukelko, D.A., Jayas, D.S., White, G. and Britton, M.G. Journal of Food Engineering 96: 598-606.
1988. Physical properties of Canola (rape seed) meal. Sreenarayanan, V.V., Viswanathan, R. and Subramanian, V.
Canadian Agric Eng 29(2):32–35. 1988. Physical and thermal properties of soybean. J.
Malleshi, N.G. and Hadimani, N.A. 1993. Nutritional and Agric. Eng. 25:76–82.
technological characteristics of small millets and
Received on 17-12-2014 Accepted on 22-12-2014