Chapter One: Introduction

1.1. Background

Teff [Eragosties tef(Zucc.) Trotter] is indigenous cereal crop in Ethiopia with largest
share of area (22.7%, 2.4 million hectors) under cereal cultivation and third (i.e. after
maize and wheat) in terms of grain production (16.3%, 24.4 million quintals) (Central
Statistical Agency, 2007) as cited in (Geremaw, 2007). Teff is used for both human
and animal feed in Africa. As a human food source, the seeds are normally ground
into flour. The flour is fermented and used to make injera, a sour-dough type of flat
bread. It is also used for making porridge and an alcoholic drink. Because the flour is
essentially gluten free, it is gaining popularity among those who suffer from gluten
allergies (Davison, 2003).

It is well known by Ethiopians and Eritreans for its superior nutritional quality. It
contains 11% protein, 73% complex carbohydrate, 2.8% ash, and 2.5% fat (Geremew,
2004). It is an excellent source of essential amino acids, especially lysine, the amino
acid that is most often deficient in grain foods. Teff contains more lysine than barley,
millet, and wheat and slightly less than rice and oats. Teff is also an excellent source
of fiber and iron, and has many times the amount of calcium, potassium and other
essential minerals found in an equal amount of other grains. When teff is used to
make injera, a short fermentation process allows the yeast to generate more vitamins
(Davison, 2003).

The color of the teff grains can be ivory, light tan to deep brown or dark reddish
brown purple, depending on the variety. There are three types of teff. White teff is the
preferred type but only grows in certain regions of Ethiopia. White teff grows only in
the Highlands of Ethiopia, requires the most rigorous growing conditions, and is the
most expensive type of teff. Red teff, the least expensive form and the least preferred
type, has the highest iron content. In persons living in areas of the country where
consumption of red teff is most prevalent, hemoglobin levels were found to be higher
with a decreased risk of anemia related to parasitic infection. The third main type of
teff, brown teff, has moderate iron content. Ethiopia is the considered the site of origin
of teff. Teff was domesticated in Ethiopia between 4000–1000 BC

1|Page
Physical and engineering properties are important in many problems associated with
the design of machines and the analysis of the behavior of the product during
agricultural process operations such as handling, planting, harvesting, threshing,
cleaning, sorting and drying. Solutions to problems in these processes involve
knowledge of their physical and engineering properties (Irtawange, 2000). Principal
axial dimensions of teff grains are useful in selecting sieve separators and in
calculating power during the milling process. They can also be used to calculate
surface area and volume of variety seeds which are important during modeling of
grain drying, aeration, heating and cooling.

Bulk density, true density, and porosity (the ratio of inter granular space to the total
space occupied by the grain) can be useful in sizing grain hoppers and storage
facilities; they can affect the rate of heat and mass transfer f moisture during aeration
and drying processes. Grain bed with low porosity will have greater resistance to
water vapor escape during the drying process, which may lead to higher power to
drive the aeration fans. Cereal grain kernel densities have been of interest in breakage
susceptibility and hardness studies (Ghasemi Varnamkhasti et al., 2007).

Differences in grain moisture content can result in a significant variation in the

processing characteristics of the grain. Hence, the objective of this study was to
determine physical properties of teff seeds and pressure drop during aeration system,
as a function of moisture content in the range of 12.01 to 25.01% (wet basis) which
can help out in the design of handling for teff production.

Undesirable materials such as light grains, weed seeds, chaff, plant leaves and stalks
can be removed with air flow, when grains, fruits and vegetables are mechanically
harvested. In addition, agricultural materials are routinely conveyed using air stream
in pneumatic conveyers. If these systems are not used properly, they could cause
problems. For example, in a combine harvester, if the air speed is low, the materials
would not be separated from each other and there will be extra foreign material with
the product. If air speed is high, the product will be exhausted along with extra
material and product loss will increase. For conveying agricultural material, the range
of proper air streams should be used. With low air speed, there is stagnation in the

2|Page
system, or with high air speed, there is not only energy lost, but also grains may be
broken (Khoshtaghaza et. al, 2006)

The proper air speed can be determined from aerodynamic properties of agricultural
materials. These properties are terminal velocity and drag coefficient. If an object is
dropped from a sufficient height, the force of gravity will accelerate it until the drag
force exerted by the air, balances the gravitational force. It will then fall at a constant
velocity called the terminal velocity (Mohsenin, 1970):

1
M .g   a *Vt 2 * C d * A
2

Where, M is mass of the object (kg), g is gravitational acceleration (m/s2), Cd is drag

coefficient, ρ is air density (kg/m3), A is projected area (m2), and Vt is terminal velocity
(m/s). From this equation, the drag coefficient of an object can be found from its
terminal velocity:

2mg
Cd = 2
vt ρ a A

Usually, a horizontal wind tunnel is used to measure drag coefficient of large objects.
In this method, external parameters such as size and velocity are varied and values of
drag coefficient are obtained over a wide range of Reynolds number. But for small
particles (like grain seeds), the drag force cannot be measured directly by this method.
So drag coefficient of agricultural materials are calculated from their terminal velocity
(Eq.2) which is experimentally measured.

Carman (1996) measured the terminal velocity of lentil seeds at different moisture
contents by free fall method. From the top of a dropping tube at various heights, a
seed was allowed to fall. The duration of the fall was plotted as a function of vertical
distance. The slope of the linear portion of the distance versus time curve indicated
the terminal velocity of the seed. He found that as the moisture content of the lentil
seed increased, its terminal velocity also increased linearly.

3|Page
In another experimental method, a vertical wind tunnel is used for finding the
suspension velocities of the particles in an air stream. Bilanski and Lal (1965)
measured terminal velocities of wheat kernel and straw by a vertical wind tunnel.

The drag coefficient of grains, which is a function of Reynolds number, lay within the
limits of a sphere (0.44) and of a cylinder (1.0) depending on the shape of the grain. In
this thesis the terminal velocity of teff was determined in order to find the effects of
mass and moisture content of teff seeds on terminal velocity (Khoshtaghaza et. al,
2006).

Ensuring high quality of cereal grain bulk in storage is assuming an increasingly great
importance. There are various methods of cereal grain preservation, but the most
frequently applied is the method of active aeration. Aeration requires a mechanical
ventilation system that can be used to manage grain temperatures by moving air with
the desired properties through the grain mass preventing moisture movement and
accumulation therefore maximizing grain storage life (Foster and McKenzie, 1979) as
cited at (Maier.D et al). In the course of that process, knowledge of air flow resistance
through grain deposit is highly significant for practical purposes [Kizun, Kusińska
2004]. Airflow resistance depends on the air velocity, grain deposit thickness, and on
the properties of the grain material (i.e. kind of grain, its bulk density, porosity,
content of contamination, and moisture content) [Siebenmorgen et al. 1987,
Sokhansanj et al. 1990, Jayas and Muir 1991, Dairo and Ajibola 1994, Giner and
Dienisienia 1996, Waszkiewicz 1999, Nimkar et al. 2002, Ray et al. 2002, Kusińska
2005].

The value of airfl ow resistance can be also strongly affected by the pouring density of
grain that may depend on its moisture content or on the method of silo charging
[Molenda et al. 2005a, Molenda et al. 2005b]. Airfl ow resistance is not a constant
feature and depends on the duration and conditions of grain storage [Szwed 2000].
Any particulate or granular material in storage undergoes compaction under the effect
of its own weight, which results in deformation of grains and has a detrimental effect

4|Page
on the grain quality features [Szwed, Kusińska 2005]. That process may cause an
increase in airfl ow resistance in grain bedding,

The quality of cereal grains in storage will deteriorate to an unacceptable level if

they are not kept dry and cool. To model the drying and cooling process, an accurate
knowledge of the airflow distribution is required. Design of efficient systems for
drying and aeration of grains requires proper design of electric motor and compressor
or fan selection, which can only be achieved with information on airflow resistance of
the grains. The pressure drop through a bed of grain depends on the airflow rate,
method of filling, the surface area and shape, configuration of voids, the variability of
particle size, grain bed depth and crop moisture content (Shedd , 1953).

Shedd (1953) reported that foreign materials mixed with grain increases the resistance
to airflow if the foreign material is finer than the grain. Similar results were reported
for resistance to airflow of grains, seeds, other agricultural products and perforated
sheets (ASAE, 1992a) as cited at (Jekayinfa, 2006) . Similar studies were also
reported by Calderwood (1973) for milled rice, Kumar and Muir (1986) for wheat and
barley, Dairo and Ajibola (1994) for sesame seed, Al-Yahya and Moghazi (1998) for
barley grain and Jekayinfa (2001) for cocoa beans. The most commonly used model is
the one proposed by Shedd (1953) where he presented curves relating airflow and
pressure drop per unit depth of grain. Because of their simplicity and ease of handling,
Shedd’s curves are widely used by many designers to estimate pressuredrops in
grains. The curves were estimated based on the formula:

Q  a  P b 1
Where,
Q = airflow rate (m s m )
3 -1 -2

Δ P= Pressure drop per unit depth (Pa/m) and

a, b = constants and are related to moisture content for some grains.

Because of the limitation of Eqn. (1) for being able to predict airflow resistance over
only a narrow range of airflow rate (Q = 0.00056 to 0.203 m 3 s-1 m-2), Hukill and Ives

5|Page
(1955) also proposed an empirical equation which accounts for the non-linear nature
of resistance to airflow data. The equation is of the form;

CQ 2
P  2
ln  1  dQ 

Where,
C, d = constants for a particular grain.
Equation 2 is applicable over a wide airflow range of 0.01 to 2.0 m3 s-1 m-2.

It is important to determine the effects of some physical properties such as moisture

content, bulk density, bed depth and airflow rate on the resistance to airflow through
the bean seeds. The knowledge of these relationships would assist in the design of
dryers and aeration systems for teff seeds.

1.2. Statement of the Problem

Due to teff seed importance many research has be conducted by different researchers.
Physicochemical Characteristics of Grain and Flour in 13 teff variety (Geremew,
2007), teff demonstration planting results for 2003 (Davison, 2003), observations of
commercial teff production in Nevada during 2006 (Davison, 2003), aerodynamic
properties of teff grain and straw material (Zewdu, 2007), moisture dependent
physical properties of teff seed (Zewdu and Solomon, 2007).

Proper designing of teff storage system and handling system as well as determining
the factors affect these phenomena is crucial. Since these helps maintain grain quality
and reduce its loss. The unnecessary loss of grain leads the society food insecurity.
The majority of the population of Ethiopia is estimated to be food insecure and the per
capital food production has dropped by 13% per annual against a 3% population
growth (World Bank 1994). This production drop mainly due to post-harvest losses.
Post-harvest losses include the following:
 Harvesting loses
 Drying loses
 Transportation loses
 Storage loses (loses due to spoilage the result of absence aeration system)
 Handling loses (loses due to lack of proper conveying)
Therefore, in order to reduce the post harvest losses above, measures should be taken
to minimized by improving storage system, applying proper handling and using
appropriate mechanism for transportation. The solution to handling and storing
high moisture grain is to either remove the excess moisture, or cool the grain
mass. This can be achieved by passing air through the grain mass, the ability to dry
grain or simply control temperature. Hence, in order to reduce storage and handling
loses this thesis intended to develop and determine pressure drop in bulk aeration and

6|Page
pneumatic conveyor for Ethiopian indigenous grain teff, so that proper storage and
handling design can be done by taking the following in to consideration:
 the rate of airflow passing through the grain,
 the distribution effectiveness of air delivery and
 the thermodynamic properties of the air passing through the grain.

1.3. Thesis Objective

General objective

This study is to address the whole aeration systems and pneumatic conveying system,
i.e. from the feeding point to the receiving tank, including all typical components on
it. Consequently, the final aim of the study is to formulate a reliable design technique,
which can be used in design of aeration system of teff seeds.

Specific objective

The main specific objective of this study is:

a. to formulate for pressure drop determination of aeration system and pipe lines
of pneumatic conveying of teff seeds
b. determine the effect of moisture content, particle size and air flow rate on
resistance characteristics of teff grain
c. determine the pressure losses of blow tank feeding with top discharge facility
d. fit the data obtained to selected models to predict air flow resistance through a
bed clean bulk teff grain and pneumatic transport pressure drop across a short
straight section and a standard 90o bend.

7|Page
Chapter Two: Review of Literatures

In this literature review, gas-solid flow in pipes is described with help of some
suggested mechanisms that have been proposed in literature. The bulk aeration system
and drop will be reviewed. And also pneumatic conveying system application, type
and component parts will be presented.

2.1. Bulk Storage Aeration System and Pressure Drop

The distribution of air in a grain storage structure has a significant effect on the
ecosystem of the stored grain mass. Most grain storage structures have a non-uniform
airflow distribution due to the variations in material properties of the grain mass, the
geometry of the storage structure, and/or the design of the aeration system. Airflow is
generally assumed to be uniform in silos with fully perforated floors and non-uniform
in silos with aeration ducts, pads or partially perforated floors. The airflow
distribution can also be non- uniform in a silo with a hopper bottom, peaked grain
from overfilling, inverted grain from partial unloading, and high fine material
concentration in the core of the grain mass (Bartosik and Maier, 2006) as cited at
(Garg. D., Maier, D.E.).

Knowledge of the flow field of air in the grain mass and the pressure drop is essential
to designing grain aeration systems. Additionally, most of the ecosystem models
developed to predict heat and mass transfer in grain storage structures during aeration
(Maier, 1992; Chang et al., 1993; Montross, 1999) as cited at (Garg. D., Maier, D.E.)
assume airflow rate to be uniform through the grain mass. Solving and integrating the
non-uniform airflow distribution into existing ecosystem models will result in a more
accurate estimate of the heat and mass transfer during storage.

Stored grain moisture levels are also influenced by the temperature of the
external environment. A warm or cold condition outside the storage cause convection
currents inside the grain storage and this causes overall moisture migration within the
grain mass. In cool weather conditions the outer of the storage will have a downward
convection current. The middle of the storage will remain warm and when reaching

8|Page
the lower part of the storage the current will then rise with the warmth. The end result
is that there can be moisture migration to accumulate under the surface near the peak
of the storage. Alternatively if grain is much cooler while in storage than its external
environment the reverse moisture migration will occur. In this circumstance there
will be rising currents along the outside of the storage and falling currents in
the centre with moisture accumulation at the base of the storage. This is called
moisture migration (Figure 2.1) and prevented by aeration (CBH Group et al, 2006).

Figure 2.1. (a) When the external temperature is lower than grain temperature

Figure 2.1 (b) when the ambient temperature is greater than the grain mass
temperature.

9|Page
2.1.1. Aeration Systems

When looking at the management opportunities offered by aeration it is

important to understand the physical characteristics of grain and how it behaves
while in storage. Kernels of grain are living organisms that ‘respire’ using the same
biological process as all other living things. They take in oxygen and combined with
carbohydrates, undergo the respiration reaction to produce carbon dioxide, water and
heat.

O2  Carbohydratte  CO2  H 2 O  Heat

This living process is ongoing in grain and the surrounding environment including
insect and bacteria activity. At lower temperatures and moisture the biological
processes of the grain and the surrounding environment are slower and more
dormant.

Figure 2.2. Grain is a living mass

Limiting the level of biological activity in stored grain has two significant benefits:
• It helps maintain the grain kernels in a dormant but living state ready for
germination or milling.
• It minimizes bacterial and insect growth so the grain kernels are not attacked
from the outside.
The following conditions support low levels of biological activity in stored grain:
• Low moisture content

10 | P a g e
 • Lower temperature
Both low moisture and temperature combine for optimal long-term storage. A higher
temperature and moisture content in stored grain causes a higher level of biological
activity (grain, bacteria and insect). As shown in the respiration equation (Figure
2.2) this process of biological activity produces water and heat which then
further stimulates the biological activity of the grain storage. For this reason grain
can self-heat and rapidly deteriorate in quality if moisture and temperature are not
managed.

2.1.2. Pressure drop in bulk storage

Air flow through beds composed of seed or grain particles is frequently used in air
pollution control processes (Dairo & Ajibola, 1994) and therefore expression is
needed to predict pressure drop across beds due to the resistance caused by the
presence of particles. The drop in pressure for flow through a bed of particles provides
convenient method for obtaining a measure of the external area of particles (Coulson
et al., 1991). Air flow resistance data of cereal grains and oilseeds have been analyzed
using many different equations (Alagusundaram and Jayas, 1990) as cited at
(Nalladaru et al, 2002). The equation is given by Shedd (1953) and Hukill and Ives
(1955) as cited at (Agullo, 2005). Shedd suggested the use of his equation for narrow
ranges of air flow rates only (.005-0.3 m3/s-m2) (Nalladurai, et al, 2002). Shedd’s
equation is:

B
 P 
V  Ac  
 Lg 
 

2.1

Hukill and Inves (1955) proposed an empirical equation to represent the air flow
resistance data over an air flow range of 0.01-2.0 m 3/s-m2. The Hukill and Ives
equation is:

P aV 2
 2.2
Lg ln 1  bV 

11 | P a g e
Both Shedd and the Hukill and Ives equations were used to analyze experimental data
obtained in the current study. The empirical equations thus developed were used to
predict pressure drops at various air flow rates. Mean relative percent: error of
prediction e in % was calculating using

1 n  Pi  M i 
e 
n i 1  M i
x100
 2.3

Darcy showed that the average velocity, as measured over the whole area of the bed,
is directly proportional to the driving pressure and inversely proportional to the
thickness of the bed. This relation is known as Darcy’s law (Coulson et at., 1991) and
can be written as

 P
vK 2.4
l
Where: P is the pressure drop across the bed in N/m 2; l is the thickness of the bed in
m; v average velocity of the fluid, defined as (1/A)(dV/dt) in m/sec; A is total cross
sectional area of the bed in m 2; V is the volume in m 3 of the fluid flowing in time t in
sec; and K is constant depending on the physical property of the bed and fluid. The
linear relation between the rate of flow and the pressure difference indicates that the
flow is streamline. This is expected because the Reynolds number of the flow through
the pore spaces in the granular material is low, since both the velocity of the fluid and
the widths of the channels are normally small. The resistance to flow then arises
mainly from viscous drag. Therefore, equation (2.4) can be expressed as:

  P 
v  K
  P 
B 2.5
 l  l

Where:  is the viscosity of the fluid in kg/ms; and B is termed the permeability
coefficient for the bed, and depends only the properties of the bed (Coulson et at.,
1991) as cited at (Tabak, et al, 2004). An aeration system consists of an air fan, air
plenum, aeration ducts (or perforated floors), grain column and anemometer shown as

12 | P a g e
figure 2.3 below. The fan may be either of centrifugal or axial type, depending on the
static pressure and airflow rate required by the storage unit (Maier & Montross, 1997).

Outlet of air

Anemometer

Pressure transducer
Grain column
P

Perforated floor

Centrifugal fan

Air plenum

Figure 2.3. Scheme of the apparatus for measuring airflow resistance in grain column

2.1.3. Factors affecting airflow resistance of grains in bulk

One of the primary causes of non-uniform airflow distribution is variation in the

material properties of the grain mass. Airflow resistance is a function of particle size
and porosity of the grain. Therefore, a number of material properties like distribution
of fine material, loading method, moisture content, and compaction cause non-
uniform airflow distribution. Airflow resistance increases when silos are filled using
spreaders as the amount of fines in the grain increases. When silos are filled using a
central fill conveyor or gravity spout, fines tend to concentrate towards the center of
the silo and chaff moves towards the silo walls. This creates a region of lower
resistance near the walls and higher resistance in the center of the silo. Orientation of
the grain kernels can also cause a non-uniform airflow distribution as the grain kernels
are not exactly spherical in shape. The non-spherical shape and random orientation of
the grain kernels result in airflow resistance that is different in every direction. Grain

13 | P a g e
also undergoes compaction during storage due to vertical pressure exerted by the
grain mass in the silo, which is influenced by the grain type, bulk density, and
coefficient of friction between the grain and the wall, moisture content, angle of
internal friction and filling method (Thompson et al., 1987) as cited at (Garg, et al).
Giner and Denisienia (1996) showed that the experimentally determined pressure drop
decreased by up to 30 % as moisture content increased from 12.8 % to 22.3 % in clean
wheat beds. Molenda et al. (2005) concluded that the effect of grain orientation on
airflow resistance was negligible, but the fill method significantly affected the airflow
resistance.

The prediction of airflow resistance is fundamental to the design of efficient drying

and aeration systems. For long term storage of grains, they must be kept cool and dry.
Drying or cooling is done by forcing air through the grains to remove high moisture
and temperature gradients within the bulk. An important step in designing drying and
aeration system is sizing the fans. Fans that are misplaced or sized incorrectly lead to
failure of the entire system (Sheley, 2000) as cited at Nalladurai et al 2001. Resistance
to airflow through a bed of grains and seeds usually expressed in terms of pressure
drop. The air flow –pressure drop relationship are useful in the mathematical
modeling of air-flow pressure patterns and air flow distribution in stored grain masses
(Brooker, 1961). The resistance to the air flow of cereal grains and oil seeds has been
studied for 70 years (Stiniman et al., 1931) as cited at (Nalladurai et al 2001). The
pressure drop depends on a number of product and environmental factors such as:
airflow rate, bed depth, bulk density, presence of amount, size and distribution of
foreign materials, grain moisture content and surface and shape characteristics of
grain (Dairo& Ajibola, 1994).

2.2. Pneumatic Conveying System and Pressure Drop

2.2.1. Pneumatic Conveying System and Its Application

Pneumatic conveying is a common in-plant transport system for bulk material which
has been used successfully in the chemical (soap powders, detergents), food (sugar,
flour), cosmetics (talc, face powder) or energy (coal and ash) industries. The major

14 | P a g e
advantages of pneumatic conveying systems are their enclosed nature, flexibility and
easy automation. A general pneumatic conveying system is shown in figure 2.4.

Figure 2.4: Pneumatic Conveying Plant

There are two basic types of pneumatic conveying; dilute phase (or suspension flow)
and dense phase (figure 2.5 and 2.6) where the predominant flow mechanism is a non
suspension mode of flow. While dilute phase systems are generally the most reliable
and offer the greatest flexibility in design, the relatively high conveying velocities
(generally in excess of 15 m/s) lead to significant operational problems including
particle attrition and erosive wear of pipelines.

The choice of whether to design for dilute phase conditions or for dense phase can be
a difficult choice for the designer. In general, dilute phase conveying has a greater
tolerance and can be safer with regard to reliability and the sensitivity of the system to
changes in material properties. However, for materials that is erosive or abrasive and
for materials that are fragile, dilute phase systems are generally not suitable. Lowering
conveying velocities can have a very significant effect in reducing the unwanted side
effects of product degradation (or attrition) and erosive wear of the system. In these
situations, there is a strong justification for using dense phase conveying.

15 | P a g e
Figure 2.5. Schematic diagram of typical dilute phase conveying system

Figure 2.6. Schematic diagram of typical dense phase conveying system

2.2.2. Advantages and Limitations of Pneumatic Conveying

In recent years, pneumatic transport systems are being used much more often,
acquiring market sectors, in which other types of transport were typically used,
especially in the fields of bulk solids handling and processing.
A well designed pneumatic conveying system is often more practical and economical
method of transporting materials from one point to another than alternative
mechanical system (belt conveyor, screw conveyors, vibrating conveyors, drag
conveyors and other methodologies) because of key reasons:

16 | P a g e
  By pneumatic conveying system, one can reduce the maintenance and
manpower cost.
 It is possible to move material vertically by simply installing a vertical section
of pipe section pipe with sufficiently high velocity of the gas to transport solid.
 Pneumatic systems are totally enclosed and if can operate entirely without
moving parts coming into contact with conveyed material. Being enclosed
these are relatively clean, more environmentally acceptable and simple to
maintain
 They are flexible in terms of routing and expansion. A pneumatic system can
convey a product any place a pipe line can run.

On the other hand, high power consumption, wear and abrasion of materials and
equipment and the limited conveying distance (1km maximum due to the economical
purpose) are the disadvantages of pneumatic conveying.

2.2.3. Major Components of Typical Pneumatic Conveying

There are a number of components in a pneumatic conveying plant, which are

required to achieve the particular duty condition. Usually, a typical conveying system
comprises different zones where distinct operations are carried out. In each of these
zones, some specialized equipments are required for successful operation of the plant.
Any pneumatic conveying system usually consists of four major components
(Chandana, 2005):
1. Conveying gas supply :– to provide the necessary energy to the conveying gas,
various types of compressors, fans, blowers and vacuum pumps are used as a
primer mover
2. Feeding mechanism: – to feed the solid to the conveying line, a feeding
mechanism such as rotary valve, screw feeder, etc, is used.
3. Conveying line: - this consists of all straight pipe lines of horizontal and/or
vertical sections, bends and other auxiliary components such as valve.
4. Separation equipment:- at the end of the conveying line, solid has to be
separated from the gas stream in which it has been transported. For this

17 | P a g e
 purpose, cyclones, bag, filters, electrostatic precipitators are usually are
usually used in the separation zone.

2.2.4. Pneumatic Conveying System Pressure Drop

Certain minimum conveying velocity must be maintained to keep the material in

suspension and flowing. A velocity that is too small will impede the material
conveying capability of the system and unnecessary high velocities will increase
pressure drop and therefore, additional energy will be required to overcome the
resistance. The conveying velocity and hence air flow rate is greatly influenced by
material characteristics. Particle sizes, size distribution, mean particle size and particle
density; all have an effect on minimum conveying velocity, pressure drop, air flow
etc. Properties such as moisture content, cohesiveness and adhesiveness may cause
flow problem through the vessel and valves. It is not just different materials! Different
grades of exactly the same material can exhibit totally different performances. Thus a
conveying system designed for one material may be totally unsuitable for another. For
practical purpose, a conservative design approach is to keep the ratio of material to air
below 1:2 proportions by mass (Bhetia. A).

2.2.5. Pressure Drop Determination

The basic step in design of pneumatic conveying system is the correct estimation of
total pressure drop in the conveying line and is estimated by either summing the
individual of contributions of air and solid by using empirical development equations.
The first method considers contribution associated with wall friction, particle friction,
particle acceleration and support of particles. Cramp and Priesty (1924) proposed this
approach and it has been followed by Vogt and White (1948), Hariu and Molstad
(1949), Pinkas and Troy (1952) and may others either in horizontal or vertical
pneumatic conveying as cited at (Raheman, et al, 2001). In general, the pressure drop
in horizontal pneumatic conveying line is represented by the following equations:

Δpt = Δp a a + Δp a s  Pf a  Pfs 2.3

18 | P a g e
The performance of a pneumatic conveying system, in terms of achieving a given
material flow rate, depends essentially on the system resistance. Higher the system
resistance, higher will be the pressure drop in the system or higher will be the static
pressure of fan (Bhetia. A). The usual assumption of pressure drop determination in
gas-solid two-phase flow is to consider the total pressure drop as being comprised of
two hypothetical pressure drop components, i.e., due to the flowing gas alone and the
additional pressure drop attributable to the solid particles. In this classic approach, the
pressure loss of air remains constant with respect to different loadings and qualities of
the conveyed materials. Under steady-state operation, the acceleration losses drop out
and equation (2.3) reduces to

Pt  Pfa  Pfs 2.4

2.2.5.1. ‘Air-Only’ Pressure Drop

The procedure involved in the determination of the air only pressure drop component,
is quite straightforward, since single-phase flow is well established with reliable
mathematical models such as Darcy-Weisbach’s. The system resistance (pipe wall
friction per unit area) can be estimated using following equation:

F f f  f v2
 2.5
A 2

If we carry energy balance over differential length, L 1 and L2, of a straight horizontal
pipe diameter D, the total force required to overcome friction drag must be supplied
by a pressure force giving rise to a pressure drop ∆P along the length L1 and L2. The
pressure drop force is
D2
∆P*Area of the pipe  P   
4
F
The friction force (force/area)*wall area of pipe     D  L
A
 v2 
So from equation (2.5)  f f     D  L
  f 
 2 

Therefore equating pressure equation and friction force

19 | P a g e
 D2  v2 
P      f f  f     D  L
4  2 

 v 2   L 
P  4   f f  f  
 2   D 

From above expression pressure drop per unit length due to air will be as follows:

P f f ρ f v2
=4 2.6
L 2D

Thus, we can see that the pressure drop is

 Directly proportional to velocity squared
 Directly proportional to the conveying distance i.e. length of the pipe and
 Inversely proportional to diameter of the pipe (Bhatia. A)

2.2.5.2. Pressure Drop Due to Solid Particles

Although the possibility of the existence of a unique mathematical model to determine

the pressure drop component due to the presence of dispersed solid particles is very
low, because of the complex nature of two-phase gas solid flow in pipes, many
correlating equations have been proposed by various authors in different publications.
Some of these methods, which show comparatively better agreements with
experimental consequences. Pressure drop per unit length due to solid friction is given
by:

2
P f f ρs vs
=4 2.7
L 2D

Chapter Three: Mathematical Modeling

20 | P a g e
3.1. Bulk Aeration Mathematical Modeling

The relationship between the pressure gradient and velocity of the air through the
grain mass must be known in order to estimate the airflow distribution. The air
velocity used in equations describing this relationship is the superficial velocity,
which is calculated as the volume flow rate divided by the cross-sectional area of the
flow. Shedd (1953) plotted data for numerous grains with a wide range of airflow
rates and proposed a relationship that has been widely used by engineers for the
design of aeration systems and the sizing and selection of fans. Unfortunately,
Shedd’s expression is empirical in nature and contains no information about the
properties of the product being aerated, or the fluid flowing through the product.
Several researchers developed expressions for pressure drop through packed beds that
have some physical basis. Darcy showed that the velocity of the fluid flowing through
a porous medium is directly proportional to the pressure drop (Darcy’s Law).
According to Reynolds (1900), the total energy loss for a fluid flow is the sum of the
viscous and kinetic energy losses. At low airflow rates (laminar flow), the resistance
offered by friction to the motion of the fluid is directly proportional to the viscosity
and velocity of the fluid. Darcy’s law holds for low flow rates where viscous forces
predominate and inertial forces can be neglected. Ergun’s, equation can be derived by
starting with the relationship between pressure drop due to friction, ∆P f and the
fanning friction, for the laminar and turbulent flow.

Pf 

4 f p L  v 2 
3.1
d e 2g c

By setting the values for the superficial velocity, v’ and hydraulic radius, rH equation
for pressure drop can be develop.

v'    v 3.2

  de
rH  3.3
61   

21 | P a g e
 3 f p   v' L  1   
2

Pf  3.4
de 3
Experimental data concluded that 3fp =1.75. The relationship between pressure drop
and fluid velocity was interest of many of experimenters who tried to find
relationships relating the factors. Blake-Kozeny, Scientists, developed an expression
that correlated the pressure drop to low (laminar) fluid flow rates (Ergun, 1952).

 1      u
2
P
 k1  3.5
L g c 3  de 2
Where de, L, , ∆P, k1, gc, u, are particle equivalent diameter, the height of the bed,
fluid viscosity, porosity, the pressure drop, the coefficient viscosity energy, the
acceleration of gravity at sea level, the fluid velocity.

At high airflow rates (turbulent flow), pressure loss is proportional to the product of
the air density and the square of the fluid velocity as viscous forces then become
relatively negligible. Ergun (1952) presented an equation for resistance to fluid flow
based on the Reynolds Theory. According to this equation, the total energy loss in a
packed bed should be treated as the sum of the viscous and kinetic energy losses. He
examined the equation from the point of view of its dependence upon flow rate,
properties of fluids (µ - viscosity and  – density), and porosity (), orientation,
particle diameter (de), shape, and surface of the granular solids:

P  1    u2
 k2   3.6
L gc 3  de

In the above equation, k2 is the kinetic energy and  is the density of the fluid in the
column. Prior to 1952, Sebri Ergun and A.A Orning tried to focus their research on
the idea that the total energy lost in a packed bed could be treated as the sum of
viscous and kinetic energy losses. It was believed that the transition for viscous to
kinetic dominance was smooth (Ergun, 1952, pg.89). Therefore, a relationship was
developed that related the pressure drop in packed bed over the entire range of
velocities when the flow rate the properties of the fluid, the fractional void volume,
shape, size and the surface of the granular solids were known.

22 | P a g e
  1     u  k   1      u2
2
P
 k1  3.7
g c 2   3 d e 2 g c 3   d e
2
L
Finally, Ergun completed the expression by finding values for k 1 and k2 and
combining all previous equations.

 1       u  1.75   1       u 2
2
P
 150  3.8
L g c 2   3 d e 2 3 g c  de

The sphericity of the packing material () can usually be ignored when assuming
perfect symmetry of packing. It is therefore presumed to be value of 1. When packing
in the bed are all spherical, the mixture can be modeled as crystalline, which involves
an ordinarily structured arrangement of the particles.
3.1.2. Pneumatic Conveying System Mathematical Modeling

In most pneumatic conveying, both solids flux and gas phase Reynolds number are
high and flow is usually turbulent. The effects of particle collisions and particle phase
turbulence must be considered in any mathematical model for simulating gas particle
flows. The shape and size of individual particles of material being transported may
differ and parameters characterizing shape and size must therefore be determined. The
characteristic relationships for pneumatic transport are generally measured in model
installations. The results are applicable to equipment of other dimensions of if
similarity criteria are known and the measurement results are plotted as the function
of a similarity characteristic. The equivalent diameter of a particle (grain ) may be
calculated from volume using the equation:

abc
V 
6
Where a, b, and c are major diameter, intermediate diameter and minor diameter
respectively.

d e   abc 
1/ 3

23 | P a g e
 abc   3 de

 3 de
V 
6

6V   3 d e

6V
de = 3 3.9
π

The shape coefficient expresses as the actual surface area as compared to the surface
area calculated using equivalent diameter

A
f= 2 3.10
πd e

In the case of plug transport, when the tube is filled completely by material, the
hydraulic diameter of bulk material, calculated as the ratio of the pore volume
between the grains to their surface area, is the characteristics parameter introducing
the concept of bulk porosity,

Vt   V
ε= 3.11
Vt

The hydraulic diameter may expressed as

V  V   ε   
t
dh = 6 =  1  ε d 3.12
A
e
f 

Geometrical similarity is ensured by the condition

D
= const 3.13
de

24 | P a g e
In addition, the shapes of the tube and material transported must also be similar.
Similarity of operation condition is ensured by adapting the specific load, i.e.,

Wm
= const 3.14
Wvm

Stationary flow for the transport medium is characterized by the Reynolds number
(ratio of the force of friction and inertia):
 f va D
Re = 3.15
f

The rato of inertia and gravitational forces gives the Froud number:

va
Fr = 3.16
gD

Chapter Four: Materials and methods

4.1. Sample Preparation

Samples that their moisture content should be raised were moistened with a calculated
quantity of distilled water by using the following Eq. (1) and conditioned to raise their
moisture content to the desired two different levels (Coskun et al., 2005):

25 | P a g e
 Wi  M f  M i 
Q 1
100  Mf

Where Q is the mass of added water (kg),

Wi is the initial mass of the sample (kg),
Mi is the initial moisture content of the sample (%, w.b.) and
Mf is the final moisture content of the sample (%, w.b.).

After making four levels of moisture contents, for selected teff seeds, the samples
were poured in polyethylene bags and the bags sealed tightly. The samples were kept
at 5 ºC in a refrigerator for a week to enable the moisture to distribute uniformly
throughout the sample. Before starting of each test, the required quantities of the
samples were taken out of the refrigerator and allowed to warm up to the room
temperature for about 2 h. The rewetting technique to attain the desired moisture
content in kernel and grain has frequently been used (Coskun et al., 2005; Garnayak
et al., 2008; Pradhan et al., 2008). All the physical properties and pressure drop of the
grains were determined at moisture levels of 12.01, 16.08, 20.71 and 25.01 % (w.b.).

4.2. Determination of engineering property tef

The seeds were cleaned manually to remove all foreign matter such as dust, dirt,
stones and chaff. To determine the size of the seeds, samples seeds were randomly
selected and their linear dimensions namely major dimension (length, L) and the
minor dimension (width, W) (Zewdu and Solomon, 2007) were measured using a
digital vernier caliper to an accuracy of 0.02 mm. The mass of seeds also measured by
an electronic digital balance to an accuracy of 0.0001 g. To evaluate 1000 grain seed
mass, randomly selected seeds from the bulk were weighed five times and averaged.
The moisture content of seed 1000 samples were determined using oven drying
method at temperature of 1052 for 24 hour (ASAE, 1994). Geometric mean diameter
(Dg), sphericity (), arithmetic mean diameter (Da) values were found using the
following formulae (Mohsenin, 1970; Sitkei, 1986).

26 | P a g e
 
D g = LW 2  1
3 4.1

=
 LW  2
1
3
4.2
L

L + 2W
Da = 4.3
3

The bulk density (ρb) was determined with a weight per hectoliter tester which would
calibrate in kg per hectoliter. The grains would not compacted during the test (Jain &
Bal, 1997 et al). The true density (ρt) and volume (V) were determined by using the
toluene displacement method (Mohsenin, 1970; Sitkei, 1986). The porosity was
calculated from the following equation (Mohsenin, 1970):
 ρ 
ε = 1001  b  4.4
 ρt 

The terminal velocities of the seeds are measured using a wind column. A roots
blower (rotary positive displacement blower) was used to develop air velocities. Air
flow is regulated by adjusting the blower speed by a motor with frequency converter.
For each test, a sample was dropped into the air stream from the top of the air column.
Air was blown upward to suspend the seed in the air stream. The air velocity near the
location of the seed suspension was measured by digital anemometer. The terminal
velocity for each of the seed was measured ten times and the average terminal
velocity for each seed was determined. The dimensionless drag coefficient
characterizes the interaction between seed and air flow and is expressed by the
formula (Mohsenin, 1970; Sitkei, 1986).

2mg
Cd = 2 4.5
vt ρ a A

4.3. Determination of pressure drop

27 | P a g e
4.3.1. Experiment Apparatus

Figure 3.1 shows a diagram of equipment. It consists of fan, control valve, air duct, air
plenum chamber, a test column, a pressure drop measuring system and air flow
measurement system. The centrifugal fan was used to create air pressure and is
preceded by manually operated control valve (not seen on the picture). The
manometer was installed next to control valve to measure pressure created by
centrifugal fan. To avoid propagating vibrations, a flexible rubber tube was used to
connect the air plenum. The flow was conducted to the inlet of air plenum to create a
uniform velocity profile. To assist this effect, the air was passed through a perforated
sheet metal or mesh with diameter 0.40mm of holes between the air plenum and test
column.

The test column consists of a cylinder of 1.0m long and 0.1m of internal diameter.
The pressure measuring taps was installed at four points with interval of 0.15m above
the air plenum to compare pressure drops. After the column, the air was flow towards
an outlet with diameter of 0.05m. This outlet air tube was used to measure air
velocities. To measure air velocity digital-hot anemometer was used in order to
minimize errors.

Outlet of air

Anemometer

Pressure transducer
Grain column
P

Perforated floor

Centrifugal fan

Air plenum

28 | P a g e
Fig.3.1. Aeration system experimental apparatus

4.3.2. Determination of Pressure drops

To determine pressure drops the test column will be filled with teff seeds to the depth
of 0.60m at a given level of moisture content. Two method of filling was used: dense
and loose. Pressure drops of dense and loose fill method was compared. The first
0.15m of test column above air plenum chamber was used for straightening the air
(Agulo et al, 2005). Pressure drop measurement was started from the air tap (T 1) at a
height of 0.15m above air plenum. This pressure was taken as a reference tap and
pressure drops for subsequent taps was the differences in static pressure between this
tap and the subsequent above the air plenum.

Pressure in a Column of Liquid or Gas

Invention of the U-tube manometer allowed the early investigators into fluid mechanics to
confirm that pressure was directly related to the sum total of the forces acting on a
surface. If you were to stand on the seashore, the pressure on you would be the weight of
the air column directly above you. That pressure has been given the name of ‘one
atmosphere’. If you were to dive to a depth of 10 meters (about 32 feet) there would now
be an added pressure on you of the weight of water above plus the weight of the air
column. By international agreement (a convention) ‘absolute’ pressure includes the
pressure of the column of air whereas ‘gauge’ pressure does not. Gauge pressure is the
pressure showing on a pressure indicator dial and is one atmosphere less than absolute
pressure. The pressure of 1-m depth of water is found from the formula –

Pressure = Density x Gravity x Height of liquid column P = ρ gh

29 | P a g e
The difference of the level of the liquid on both sides of the U tube, the unknown
pressure P for gas fluid C can be determine with fluid static formula as below,

P  Pr ef   B g h (1)

The gauge pressure of P can be determinate with

Pgauge  P  Pr ef   B g h (2)

The unit is a Pascal. Gravity has the value of 9.8 m/sec 2 at sea level. For simplicity of
multiplication the value 10 m/sec 2 will be used. The density for water is 1,000 kg per
cubic meter at 20 oC. Putting all the know values back into the pressure equation gives –

P = ρ g h = 1000 kg/m3 x 10 m/sec2 x 1 m = 10,000 Pa = 10 kPa

30 | P a g e
The calculation shows that 1-meter of water is equal to about 10 kPa, which means 30
meters of water produces a pressure of nearly 300 kPa. One atmosphere of air pressure at
sea level is 101 kPa. This means the pressure at 30 meters depth below sea level is 300
kPa gauge pressure or about 400 kPa absolute pressures. The same formula can be used to
calculate negative, or vacuum, pressures.

The pressure difference in a inclined u-tube can be expressed as

P 
ΔP = ghsin( ) B ρ θ Eq (3)
where
θ = angle of column relative the horizontal plane

How to Use a U-Tube Manometer

Figure No. 1 shows three manometers open to atmosphere. The left one has the same
pressure in both legs and the liquid levels are the same on both sides. The U-tube in the
center shows a pressure applied to the left leg of 100 kPa. The water level in the left leg

31 | P a g e
has gone down and the level in the right leg has gone up. The difference in the height of
water between the two legs is 10 meters. Since the liquid is water, each meter height
represents 10 kPa and a 10 meter high water column represents 100 kPa gauge pressure.
The remaining U-tube shows a pressure of 100 kPa as well but this time mercury is used
in the tube. The height of mercury is now 750 mm. The density of mercury is 13.6 times
that of water. Because mercury is so much heavier than water the same pressure raises a
correspondingly lower column of liquid.

Figure No. 1 U-tube Manometers with Water and Mercury

Figure No. 1 U-tube Manometers with Water and Mercury If a manometer were used to
measure a vacuum the column of liquid would be drawn up toward the vacuum and the
difference in the height of liquid between the two legs would be a measure of the vacuum
pressure below atmospheric pressure.
Making a U-Tube Manometer

To make a U-tube manometer requires a clear plastic tube mounted in the shape of a ‘U’
onto a board marked with a graduated scale. The pressure to be measured determines the
selection of the liquid used in the tube. The U-tube liquid’s density and the pressure being
measured determine the height of the liquid column and the corresponding height of the
backing-board.

32 | P a g e
4.4. Determination of power requirement and capacity

To determine the conveying power requirement, the voltage, the current, and the
power factor drawn by the electric motor for the roots blower was measured by using
voltmeter, ampere meter, and power factor meter, respectively. Conveying capacity
depends on mass of seed, air moved, and the physical characteristics of the seeds
(Hellevang, 1985).

Chapter Five: Results and discussion

5.1. Determination of engineering properties of teff seeds

The teff seeds were cleaned and all foreign materials such as dust, dirt, stones and
chaff were removed. Engineering properties of the teff seeds were determined using
different level of moisture contents. Different moisture contents of samples were

33 | P a g e
determined using oven drying method at temperature of 105  2 for 24 hours (ASE,
1994).
The size of seeds at different moisture contents were determined by randomly selected
samples and their linear dimensions major dimension (length) and minor dimension
(width) were measured using a digital vernier caliper to an accuracy of 0.02 mm. The
thousand grain of teff seeds were measure by electronic digital balance to accuracy of
0.001 gm the teff seeds selected from the bulk were weighed five times.

The level zero and the moisture content average result was found on wet basis was
12.01% and the engineering properties at this moisture content level were found as
follows.

Major dimension (length) 1.05 mm and minor dimension (width) 0.62 mm, thousand
grains mass was 0.322 gm.

Geometric mean diameter (Dg), sphericity () and mean diametric (Da) values were
found using the following formula (Mohsenin, 1970, Sitikei, 1986).

   
1/ 3 1/ 3
Dg  LW 2  1.05  0.622  0.74mm


 LW  2 1/ 3


 1.1 0.62  2 1/ 3

 70.38%
L 1.05

Da 
 L  2W    1.05  0.62  2   0.76mm
3 3

The equivalent mean diameter (de) at the level zero moisture content was determined
using the procedure adapted by the Kaleemullah and Gunnasker (2002) and Socilik et
al. (2003), by considering its effective diameter of the thousand grain mass and true
density as:

34 | P a g e
 1/ 3 1/ 3
 6  M 1000   6  0.322 
de       0.775mm
   0 1000    1320.5 1000 

The bulk density (b) was determined with weight per hectoliter tester which would
calibrate in kg per hectoliter. The grain was not compacted during the test (Jain & Bal,
et al. 1997). The result at the zero level moisture content of bulk density was 795.53
Kg/m3.
The true density (t) and volume were determined by using the toluene displacement
method (Mohsenin, 1970, Sitikie, 1986). The result at the zero level moisture content
was 1320.5 Kg/m3. The porosity was calculated from the following equation
(Mohsenin, 1970):

    795.53 
  100 1  b   100 1    39.92%
 t   1324.03 

The same procedure was followed for the other remains moisture content and the
following results were found.
The level one moisture content average result found on wet basis was 16.08 and the
engineering properties at this moisture content level were found as follows. The two
dimension major dimension (length) and minor dimension (width) were 1.07 mm and
0.63 mm respectively and average thousand grains mass was 0.328 gm.

Geometric mean diameter (Dg), sphericity () and mean diametric (Da) values were
found using the following formula (Mohsenin, 1970, Sitikei, 1986).

   
1/ 3 1/ 3
Dg  LW 2  1.07  0.632  0.75mm


 LW  2 1/ 3


 1.07  0.63  2 1/ 3

 70.24%
L 1.07

Da 
 L  2W    1.05  0.63  2   0.77mm
3 3

35 | P a g e
The equivalent mean diameter (de) at the level one moisture content was determined
using the procedure adapted by the Kaleemullah and Gunnasker (2002) and Socilik et
al. (2003), by considering its effective diameter of the thousand grain mass and true
density as:

1/ 3 1/ 3
 6  M 1000   6  0.328 
de       0.78mm
   0 1000    1319.23 1000 

The bulk density (b) was determined with weight per hectoliter tester which would
calibrate in kg per hectoliter. The grain was not compacted during the test (Jain & Bal,
et al. 1997). The result at the level one moisture content of bulk density was 791.41
Kg/m3.

The true density (t) and volume were determined by using the toluene displacement
method (Mohsenin, 1970, Sitikie, 1986). The result at the level one moisture content
was 1319.23 Kg/m3. The porosity was calculated from the following equation
(Mohsenin, 1970):

    791.41 
  100 1  b   100 1    40.01%
 t   1319.23 

The level one moisture content average result found on wet basis was 20.71% and
the engineering properties at this moisture content level were found as follows. The
two dimension major dimension (length) and minor dimension (width) were 1.09 mm
and 0.65 mm respectively and average thousand grains mass was 0.330 gm.

Geometric mean diameter (Dg), sphericity () and mean diametric (Da) values were
found using the following formula (Mohsenin, 1970, Sitikei, 1986).

   
1/ 3 1/ 3
Dg  LW 2  1.09  0.652  0.77mm


 LW  2 1/ 3


 1.09  0.65  2 1/ 3

 70.84%
L 1.09

36 | P a g e
Da 
 L  2W    1.09  0.65  2   0.80mm
3 3

The equivalent mean diameter (de) at the level two moisture content was determined
using the procedure adapted by the Kaleemullah and Gunnasker (2002) and Socilik et
al. (2003), by considering its effective diameter of the thousand grain mass and true
density as:

1/ 3 1/ 3
 6  M 1000   6  0.33 
de       0.79mm
   0 1000    1272.54 1000 

The bulk density (b) was determined with weight per hectoliter tester which would
calibrate in kg per hectoliter. The grain was not compacted during the test (Jain & Bal,
et al. 1997). The result at the level two moisture content of bulk density was 740.88
Kg/m3.

The true density (t) and volume were determined by using the toluene displacement
method (Mohsenin, 1970, Sitikie, 1986). The result at the level two moisture content
was 1272.54 Kg/m3. The porosity was calculated from the following equation
(Mohsenin, 1970):

    740.88 
  100 1  b   100 1    41.78%
 t   1272.54 

The level one moisture content average result found on wet basis was 25.01% and
the engineering properties at this moisture content level were found as follows. The
two dimension major dimension (length) and minor dimension (width) were 1.13 mm
and 0.68 mm respectively and average thousand grains mass was 0.351 gm.

Geometric mean diameter (Dg), sphericity () and mean diametric (Da) values were
found using the following formula (Mohsenin, 1970, Sitikei, 1986).

37 | P a g e
    
1/ 3 1/ 3
Dg  LW 2  1.13  0.682  0.81mm


 LW  2 1/ 3


 1.13  0.68  2 1/ 3

 71.27%
L 1.13

Da 
 L  2W    1.13  0.68  2   0.83mm
3 3

The equivalent mean diameter (de) at the level three moisture content was determined
using the procedure adapted by the Kaleemullah and Gunnasker (2002) and Socilik et
al. (2003), by considering its effective diameter of the thousand grain mass and true
density as:

1/ 3 1/ 3
 6  M 1000   6  0.351 
de       0.85mm
   0 1000    1094.2 1000 

The bulk density (b) was determined with weight per hectoliter tester which would
calibrate in kg per hectoliter. The grain was not compacted during the test (Jain & Bal,
et al. 1997). The result at the zero level moisture content of bulk density was 740.88
Kg/m3.

The true density (t) and volume were determined by using the toluene displacement
method (Mohsenin, 1970, Sitikie, 1986). The result at the level three moisture content
was 1272.54 Kg/m3. The porosity was calculated from the following equation
(Mohsenin, 1970):

    630.63 
  100 1  b   100 1    42.38%
 t   1094.54 
Table 1. Effect of moisture content on engineering properties of teff seeds

Moisture Engineering property of the teff seeds

38 | P a g e
 Geo Equi True Bulk
Content Length Width dia dia. M1000 Sphericity density density Porosity
3 3
(%) (mm) (mm) (mm) (mm) (gm) (%) (Kg/m ) (Kg/m ) (%)
12.01 1.05 0.62 0.74  0.775 0.322 70.38 1324.03 795.53 39.92
16.08 1.07 0.63  0.75 0.78 0.328 70.24 1319.23 791.41 39.47
20.71 1.09 0.65  0.77 0.79 0.33 70.84 1272.54 741.88 40.06
25.01 1.13 0.68  0.78 0.85 0.351 71.27 1094.2 630.63 41.13

5.1.1. Effect of moisture content on teff seed linear dimensions

Mean values of the size dimensions of teff seeds at different moisture contents are
presented in Table 1. As also seen in Table 1, all the dimensions increased with
increase of moisture content within the moisture range of 12.01-25.01% (w.b.). The
relationships between the axial dimensions (L and W) and moisture content of grain
(M) can be represented by the regression equations:

L  0.006 M  0.975
With values for R2= 0.9661

w  0.0046M  0.5602
With values for R2= 0.9569
By means of which the regression relationship was determined. This results show that
there is an important and positive relationship between moisture content of grain and
axial dimensions of grain.

39 | P a g e
 1.2
Linear dimensions, mm
1.1

0.9 Length

0.8 Width

0.7 Linear
(Length)
0.6 Linear
(Width)
0.5
10 12 14 16 18 20 22 24 26
Moisture content, %, w.b.

Fig.5.1. Effect of moisture content on teff seed linear dimensions

5.1.2. Effect of moisture content on teff seeds diameters

The variation of geometric mean diameter of teff seeds, with moisture content is
presented in (fig ). This show that the geometric mean diameter of teff seeds increased
from 0.74 to 0.78 mm in the moisture content range of 12.01 to 25.01% (w.b). A
linear relationship between geometric mean diameter and moisture content was
obtained and can be expressed using equation:

Dg  0.0032M  0.701

With the coefficient of determination R2 of 0.985, Where Dg was geometer diameter

of teff seeds (mm) and M was moisture content at wet basis (w.b).

The relationship between equivalent diameter and moisture content of grain is shown
in Fig. The equivalent diameter of the teff seeds increased from 0.775 to 0.85 mm
depending on the increase of moisture content. The relationship between diameter and
moisture content can be represented by the following equation (R 2=0.767):

d e  0.0054M  0.699

40 | P a g e
 0.86

0.84 Equi dia

Geometric M dia
Diam eters, m m

0.82

0.8

0.78

0.76

0.74

0.72
10 15 20 25
Moisture content, %, w.b

Fig.5.2. Effect of moisture content on teff seed diameters

5.1.3. Effect of moisture content on teff seed thousand grains mass

The variation of one thousand teff seeds grains mass weight, M 1000, with moisture
content is presented in (fig ). This show that the one thousand teff seeds weight
increased from 0.322 to 0.351 gm in the moisture content range of 12.01 to 25.01%
(w.b). A linear relationship between M1000 and moisture content was obtained and can
be expressed using equation:

M 1000  0.002 M  0.2952

With the coefficient of determination R2 of 0.8254, Where M1000 was one thousand
teff seeds weight (gm) and M was moisture content at wet basis (w.b).

41 | P a g e
 0.355

0.35

0.345

0.34
M1000, gm

0.335

0.33

0.325

0.32

0.315
10 15 20 25
Moisture content , % w.b.

Fig.5.3. Effect of moisture content on teff seed thousand grain mass

5.1.4. Effect of moisture content on teff seed Sphericity

The relationship between sphericity and moisture content of grain is shown in Fig.
The sphericity of the teff seeds first decreased from 70.38 to 70.24% then increased
from 70.24 to 71.27% depending on the increase of moisture content. The relationship
between sphericity and moisture content can be represented by the following equation
(R2=0.8298):

  0.757 M  69.286

71.4

71.2

71
Sphericity, %

70.8

70.6

70.4

70.2

70
10 15 20 25
Moisture content, % w.b

Fig.5.4. Effect of moisture content on teff seed sphericity

42 | P a g e
5.1.5. Effect of moisture content on teff seed Densities

The true density of teff seeds at 12.01, 16.08, 20.71 and 25.01% (w.b) moisture
contents the values varied from 1324.23 to 1094.20 Kg/m3 (fig ). The relationship
between true density of grain of with moisture content was as follows:

t  1564.80  16.927 M

With values for R2 of 0.779

Where M is moisture content at wet basis (w.b).

The values of bulk density of teff seeds for different moisture contents level varied
from 795.53 to 630.63 Kg/m3 (fig ). The bulk density of grains was found to be linear
the following relationship with moisture content:

b  970.96  12.52 M
With values for R2 of 0.8443

The regression equations indicate that the increase of moisture content caused a
decline both in bulk density and in true density. It was also observed that the increase
of moisture content of grain depending on structure of fiber in grainy products
affected bulk density and true density in studies made by Gupta and Das (1997),
Baryeh (2001), Sahoo and Srivastava (2002), Aviara et al. (2005), Altuntaþ et al.
(2005), Mwithiga and Sifuna (2005) and Yalçýn (2006) as cited at Kiber et al. (2008).

43 | P a g e
 1400 900

1320

Bulk Density, Kg/m3

True Density, Kg/m3

800
1240

1160 True density

Bulk density 700

1080

1000 600
10 15 20 25
Moisture Content, % w.b.

Fig.5.5. Effect of moisture content on teff seed densities

5.1.6. Effect of moisture content on porosity of teff seed

The change of porosity with moisture content is shown in Fig. 2d. The porosity of teff
seeds increased from 39.92 to 42.38% depending on the increase of moisture content.
The porosity () and the moisture content of teff seed can be correlated as:

  0.021M  37.13
Coefficient of determination R2= 0.912

43

42.5

42
Porosity,%

41.5

41

40.5

40

39.5
10 15 20 25
Moisture content, %, w.b

Fig.5.6. Effect of moisture content on teff seed porosity

44 | P a g e
5.2. Aerodynamic Property of teff

The proper air speed can be determined from aerodynamic properties of agricultural
materials. These properties are terminal velocity and drag coefficient. If an object is
dropped from a sufficient height, the force of gravity will accelerate it until the drag
force exerted by the air, balances the gravitational force. It will then fall at a constant
velocity called the terminal velocity (Mohsenin, 1970):

1
M .g   a *Vt 2 * C d * A
2

Where, M is mass of the object (kg), g is gravitational acceleration (m/s2), Cd is drag

coefficient, ρ is air density (kg/m3), A is projected area (m2), and Vt is terminal velocity
(m/s). From this equation, the drag coefficient of an object can be found from its
terminal velocity:

2mg
Cd = 2
vt ρ a A

Usually, a horizontal wind tunnel is used to measure drag coefficient of large objects.
In this method, external parameters such as size and velocity are varied and values of
drag coefficient are obtained over a wide range of Reynolds number. But for small
particles (like grain seeds), the drag force cannot be measured directly by this method.
So drag coefficient of agricultural materials are calculated from their terminal velocity
(Eq.2) which is experimentally measured.

Carman (1996) measured the terminal velocity of lentil seeds at different moisture
contents by free fall method. From the top of a dropping tube at various heights, a
seed was allowed to fall. The duration of the fall was plotted as a function of vertical
distance. The slope of the linear portion of the distance versus time curve indicated
the terminal velocity of the seed. He found that as the moisture content of the lentil
seed increased, its terminal velocity also increased linearly.

45 | P a g e
In another experimental method, a vertical wind tunnel is used for finding the
suspension velocities of the particles in an air stream. Bilanski and Lal (1965)
measured terminal velocities of wheat kernel and straw by a vertical wind tunnel.

The drag coefficient of grains, which is a function of Reynolds number, lay within the
limits of a sphere (0.44) and of a cylinder (1.0) depending on the shape of the grain. In
this thesis the terminal velocity of teff was determined in order to find the effects of
mass and moisture content of teff seeds on terminal velocity.

5.3. Determination of Pressure drops

To determine pressure drops the test column was filled with teff seeds to the depth of
0.60m at a given level of moisture content. Two method of filling was used: dense and
loose. Pressure drops of dense and loose fill method was compared. The first 0.15m of
test column above air plenum chamber was used for straightening the air (Agulo et al,
2005). Pressure drop measurement was started from the air tap (T 1) at a height of
0.15m above air plenum. This pressure was taken as a reference tap and pressure
drops for subsequent taps was the differences in static pressure between this tap and
the subsequent above the air plenum.

Four moisture levels 12.01, 16.08, 20.71 and 25.01% were used to determined
pressure drop through the grain column, the result indicate that when the moisture
content increases the pressure drop or static pressure decreases. With the above
moisture contents in the loose method the following pressure drops were observed
518.52, 503.70, 488.889 and 481.482 Pa per meter respectively. For the loose fill the
calculated pressure drops were 315.501, 309.809, 245.889 and 197.414 Pascal per
meter respectively.

Incase dense fill method with moisture contents 12.01, 16.08, 20.71 and 25.01% the
following pressure were observed 525.93, 522.963, 500.00 and 496.296 Pascal per
meter and calculated pressure drops for the dense fill were 333.454, 328.01, 295.16
and 225.57 Pascal per meter respectively.
5.4. Determination of power requirement and capacity of Blower Motor

46 | P a g e
To determine the conveying power requirement, the voltage, the current, and the
power factor drawn by the electric motor for the roots blower was measured by using
voltmeter, ampere meter, and power factor meter, respectively. Conveying capacity
depends on mass of seed, air moved, and the physical characteristics of the seeds
(Hellevang, 1985).

5.5. Mathematical Modeling

5.5.1. Bulk Aeration Mathematical Modeling

The relationship between the pressure gradient and velocity of the air through the
grain mass must be known in order to estimate the airflow distribution. The air
velocity used in equations describing this relationship is the superficial velocity,
which is calculated as the volume flow rate divided by the cross-sectional area of the
flow. Shedd (1953) plotted data for numerous grains with a wide range of airflow
rates and proposed a relationship that has been widely used by engineers for the
design of aeration systems and the sizing and selection of fans. Unfortunately,
Shedd’s expression is empirical in nature and contains no information about the
properties of the product being aerated, or the fluid flowing through the product.
Several researchers developed expressions for pressure drop through packed beds that
have some physical basis. Darcy showed that the velocity of the fluid flowing through
a porous medium is directly proportional to the pressure drop (Darcy’s Law).
According to Reynolds (1900), the total energy loss for a fluid flow is the sum of the
viscous and kinetic energy losses. At low airflow rates (laminar flow), the resistance
offered by friction to the motion of the fluid is directly proportional to the viscosity
and velocity of the fluid. Darcy’s law holds for low flow rates where viscous forces
predominate and inertial forces can be neglected. Ergun’s, equation can be derived by
starting with the relationship between pressure drop due to friction, ∆P f and the
fanning friction, for the laminar and turbulent flow.

Pf 

4 f p L  v 2 
3.1
d e 2g c

47 | P a g e
By setting the values for the superficial velocity, v’ and hydraulic radius, rH equation
for pressure drop can be develop.

v0 '   0  v  0.3992  0.08  0.032m / sec

 0  de0
rH   0.0858mm
6  1 0 

3 f p   v0 ' L   1   0 
2

Pf   39.31
d e 0 03
Experimental data concluded that 3fp =1.75. The relationship between pressure drop
and fluid velocity was interest of many of experimenters who tried to find
relationships relating the factors. Blake-Kozeny, Scientists, developed an expression
that correlated the pressure drop to low (laminar) fluid flow rates (Ergun, 1952).

 1    u
2
P  g c
 k1    285.39
L 3  de2
Where de, L, , ∆P, k1, gc, u, are particle equivalent diameter, the height of the bed,
fluid viscosity, porosity, the pressure drop, the coefficient viscosity energy, the
acceleration of gravity at sea level, the fluid velocity.

At high airflow rates (turbulent flow), pressure loss is proportional to the product of
the air density and the square of the fluid velocity as viscous forces then become
relatively negligible. Ergun (1952) presented an equation for resistance to fluid flow
based on the Reynolds Theory. According to this equation, the total energy loss in a
packed bed should be treated as the sum of the viscous and kinetic energy losses. He
examined the equation from the point of view of its dependence upon flow rate,
properties of fluids (µ - viscosity and  – density), and porosity (), orientation,
particle diameter (de), shape, and surface of the granular solids:

 1    u  1       u 2  315.504
2 2
P  g c
 k1  2 3  2  k2 
L   de 3   de

48 | P a g e
In the above equation, k2 is the kinetic energy and  is the density of the fluid in the
column. Prior to 1952, Sebri Ergun and A.A Orning tried to focus their research on
the idea that the total energy lost in a packed bed could be treated as the sum of
viscous and kinetic energy losses. It was believed that the transition for viscous to
kinetic dominance was smooth (Ergun, 1952, pg.89). Therefore, a relationship was
developed that related the pressure drop in packed bed over the entire range of
velocities when the flow rate the properties of the fluid, the fractional void volume,
shape, size and the surface of the granular solids were known.

 1    u  1      u2
2 2
P  g c
 k1  2 3  2  k2 
L   de 3   de
Finally, Ergun completed the expression by finding values for k 1 and k2 and
combining all previous equations.

 1    u  1 0    u2
2 2
P  gc
 150  2 0 3   1.75    315.504
L  0 de02  03   de0

The sphericity of the packing material () can usually be ignored when assuming
perfect symmetry of packing. It is therefore presumed to be value of 1. When packing
in the bed are all spherical, the mixture can be modeled as crystalline, which involves
an ordinarily structured arrangement of the particles.

49 | P a g e
8. References

1. ASAE. ASAE standard 352.2 –moisture measurement underground grain and

seeds. (1992) standard 352.2, St. Joseph
2. Aullo. J.o; Marenya. M.O. Air flow Resistance of Parchment Arabica Coffee.
Biosystem Engineering (2005) 91 (2), 149-156.
3. Bhetia . A: Pneumatic Conveying systems. PDH Engineer.com
4. Brooker, D. B.; Bakker-Arkema, F. W.; Hall, C. W. (1992). Drying and Storage
of Grains and Oilseeds. New York, NY, Van Nostrand Reinhold.
5. Brooker, D. B. Pressure patterns in grain drying systems established by numerical
methods(1961). Transactions of ASAE, 4, 72-74, 77.
6. CBH Group and the South East Premium Wheat Growers Association
(SEPWA). (2006). The WA Guide to High Moisture Harvest Management, Grain
Storage and Handling.
7. Chandana Rantnayak. A Compressive Scale Up Techniques for Pneumatic
Transport System. The Norwegian University of Science and Technology (2005)
8. Coskun, M.B., I. Yalçin and C. Özarslan, 2005. Physical properties of sweet corn
seed (Zea mays saccharata Sturt.). J. Food Eng., 74(4): 523-528.
9. Crozz. D.E and Pagano. A.M. Modeling Resistance to air flow through beds of
agropyron and corn estimation of power ventilation. Lat .Am.Appl. Res. v36.n1 Bahia
Balanca eng/marc2006

50 | P a g e
10. Dairo O U, Ajibola O O (1994). Resistance to air flow of bulk sesame seed.
Journal of Agricultural Engineering Research, 58(2), 99-105
11. Datta . B. K. and Ratnayake. C., A Possible Scaling Up Technique for
Dense Phase Pneumatic Conveying, Proceedings of the 4th International Conference
for Conveying and Handling of Particulate Solids, Budapest, Hungary, May 2003,
Vol. 2, pp. 11.74-11.79
12. Datta . B. K. and Ratnayake. C. An Experimental Study on Degradation of
Maize Starch during Pneumatic Transportation, Communicated to publish in
Particulate Science and Technology, Taylor & Francis Inc.
13. Ergun, S. Fluid flow through packed columns. Chemical Engineering Progress,
Vol. 48, 1952, No. 2, 89-94
14. Esraf Isik and Hulya Isik. The effect of moisture content of organic chick pea
(Cicer arietinumI) grain on the physical and mechanical properties. International
Journal of Agricultural Research 3(1) 40-51, 2008.
15. Fadel F. Erian, Leonard F, Pease III. Three Phase Upward flow in vertical pipe.
International Journal of Multiphase flow 33 (2007)498-509.
16. Garg. D, Maier. D.E. Modeling of non-uniform airflow distribution in large grain
silos using Fluent. 9th International Working Conference on stored product protection.
17. Geremew. Bultesa. Physiochemical characteristics of grain and flour in 13 teff
verities. Journal of Applied Sciences Research, 3 (12): 2042-2051, 2007
18. Geremew. Bultesa. TEFF. Bio system Engineering. (2004)
19. Giner. S.A. and Denisenia. E. Pressure Drop through Wheat as affected by Air
Velocity, Moisture Content and Fines. J. Agric. Engng. Res. (1996) 63, 73-86
Irtawange, S.V., 2000. The effect of accession on some physical and engineering
properties of African yam bean. Unpublished Ph.D. Thesis, Department of
Agricultural Engineering, University of Ibadan., Nigeria.

Jay Davison. 2003. Tef Demonstration Planting Results for 2003. Area Specialist,
University of Nevada Cooperative Extension Craig McKnight, Agricultural
Cooperator, Fallon Nevada

51 | P a g e
20. Jekanifa. S.O. “Effect of Airflow Rate, Moisture Content and Pressure Drop on
the Airflow Resistance of Locust Bean Seed”. Agricultural Engineering International:
the CIGR Ejournal. Manuscript FP 06 010. Vol. VIII. May, 2006.

21. José Helvecio Martins; Alexandre Manuel Mota; José Alberto Fonseca.
Automatic Control and Monitoring of Stored Grain Aeration systems by real time
data acquisition. University of Aveiro, Department of Electronic and
Telecommunications, 3810-193 Aveiro,

Kiber, Physical and mechanical properties of soybean

H. Kibar* and T. Öztürk
Department of Agricultural Structures and Irrigation, Faculty of Agriculture, University of Ondokuz Mayis, 55139,
Kurupelit-Samsun, TurkeyReceived March 7, 2008; accepted July 3, 2008

22. Kilickan. A; Guner. M. Pneumatic Characteristics of Cotton Seed. Bio system

engineering (2006) 95(4), 73-86
23. Khoshtaghaza. M and Mehdizadeh. R. “Aerodynamic Properties of Wheat
Kernel and Straw Materials”. Agricultural Engineering International: the CIGR
Ejournal. Manuscript FP 05 007. Vol. VIII. March, 2006.
24. Lacerda Filho, A. F.; Silva, J. S. (1995). Aeração de Grãos [Grains Aeration].
In: Pré-Processamento de Produtos Agrícolas [Pre-processing of Agricultural
Products]. Juiz de Fora, MG, Brazil, Silva, J.S. (editor), Instituto Maria [Maria
Institute], p.231-252.
25. LAI . F.S; FAN. L.T. Excessive Pressure drop through a Bed of Stored Grain
during Aeration. J. Agric. Engng. Res. (1976) 21, 207-209
26. Lasseran, J. C. (1981). Aeração de Grãos [Grains Aeration]. Série
CENTREINAR [CENTREINAR series], No. 2, Artes Gráficas Formato S.A., Belo
Horizonte, MG, Brazil, 131pp.
27. Maier, D.; Montross, M. (1997). Aeration Technology for Moisture
Management. In: Proceedings of the University of Illinois Grain Quality Conference –
Managing Moisture in Grains and Oilseeds. March 26-27.
28. Mills, D. Pneumatic Conveying: Dilute vs. Dense phase. Chemical Engineering
(2005)112(1):46-51

52 | P a g e
29. Mohsenin N N (1986) Physical properties of plant and animal materials, 2nd Edn.
Gordon and Breach science Publishers, New York.
30. Molenda, M., Montross, M. D., McNeill, S. G., and Horabik, J., 2005. Airflow
resistance of seeds at different bulk densities using Ergun’s equation. Transactions of
the ASAE 48, 1137-1145
31. Montross, M. D., 1999. Finite element modeling of stored grain ecosystems and
alternative pest control techniques. Ph.D. Dissertation, Purdue University, West
Lafayette.
32. Nalladurai. K, Alagusundaram. K, Gayathri. P; Air Resistance of Paddy and
its byproducts. Biosystem Engineering (2002) 83(1), 67-75.
33. Nehal Desai. Investigation of in gas-solid multiphase flows. North Carolina State
University (2003).
34. Raoufat. M.H, Clarke. B. Design and Development of a Packed-bed Continuous
Pneumatic Conveyor. J. Agric. Engng. Res. (1998) 71, 363-371
35. Rehman. H, Jindal. V.K. Pressure Drop Gradient and Solid Friction Factor in
Horizontal Pneumatic Conveying of Agricultural Grains. Applied Engineering
Agriculture vol-17(5):649-656
36. Reynolds, O., 1900. Papers on mechanical and physical subjects. Cambridge
University Press, Cambridge, England
37. Roby L, Mc Guckin; Mark A. Eiteman; Keshaw Das. Pressure Drop through
Food Waste Composite Containing Synthetic Bulking agents. J. Agric. Engng. Res.
(1999) 72, 375-384
38. Sarker, N.N. and M. Farouk, 1989. Some factors causing rice milling loss in
Bangladesh. Agriculture and Mechanization in Asia, Africa and Latin A merica. 202:
49-52.
39. Shedd C K. Resistance of grains and seeds to airflow. Agricultural Engineering,
(1953) 4(9), 61
40. Sitkei, G. Mechanics of agricultural materials. (1986). Budapest: Academia
Kiado.
41. Tabak. S; Askrov. B; V.; Rashidov. I. Tabak; Manor. G; Shmulevich. I. Air
Flow through Granular Beds Packed with Cottonseeds. Bioengineering (2004)
88(2),163-173

53 | P a g e
42. Zewdu.A.D. Aerodynamic Properties of Tef grain and straw material. Bio system
engineering 98(2007)304-309
43. Zewdu. A.D; Solomon. W.K. Moisture Dependent Physical properties of Tef
seed. Bio system engineering (2007)(1)57-63

54 | P a g e