DESIGN AND DEVELOPMENT OF A SMALL POTATO-GRADING

MACHINE USING CAPRON NET

D. GHANBARIAN1,4, N.N. KOLCHIN2, S.R. HASAN BEIGI3 and R. EBRAHIMI1

1
Department of Agricultural Machinery
Faculty of Agriculture, Shahrekord University,
PO Box 115 Iran
2
Russian Research Institute of Agricultural Machinery (VISKHOM)
127247, Russia
3
Department of Agricultural Technical Engineering
Abureyhan Campus, Tehran University
Iran

Accepted for Publication July 24, 2008

ABSTRACT

Grading potatoes by size is an important operation for preparing tubers

for seed and for commercial purposes. Farmers in Iran and in developing
countries grade potatoes by hand, which is very slow and costly. This work
concerns the development of a potato-grading machine for small-scale
farmers. The design is characterized by a spiral structure covered with a
capron net to form a cribriform rotating drum as the grading surface. Potato
tubers are fed into the drum, moved forward by the spiral structure, and
dropped through holes of the net according to their dimensions. The design
parameters included net stiffness (T), rotating drum speed (n) and feeding rate
(q). The grading accuracy of the machine was evaluated to determine optimum
condition of machine performance. An acceptable grading accuracy of about
70% with a low tuber damage of 5.5% were obtained at T = 0.14 N/mm,
n = 27 rpm and q = 2,500 kg/h as optimum condition.

PRACTICAL APPLICATIONS

Commercial potato-grading machines are designed for medium-

to large-scale farms, and are not suitable for the majority of farmers in

4
Corresponding author. TEL: +98-381-4424546; FAX: +98-381-4424428; EMAIL: Ghanbarian51@
yahoo.com

Journal of Food Process Engineering •• (2009) ••–••. All Rights Reserved.

© Copyright the Authors 1
Journal Compilation © 2009 Wiley Periodicals, Inc.
DOI: 10.1111/j.1745-4530.2008.00331.x
2 D. GHANBARIAN ET AL.

developing countries. In these countries, grading of potato tubers, especially

for seed purpose, is performed by hand, which is very time-consuming and
labor-intensive. A small grading machine has been developed that can sort
five times more tubers per person per unit time than the rate achieved by hand
grading. The machine is very simply constructed and costs approximately
US$800, which is considerably cheaper than current commercial grading
machinery.

INTRODUCTION

The potato (Solanum tuberosum L.) is the third most important food crop
in Iran, after rice and wheat. Potatoes are cultivated in an area of 1.9 ¥ 109 m2
with an annual production of 4.83 ¥ 109 kg (MJA 2007). In many developing
countries such as Iran, potatoes are produced by farmers with less than 3 to
4 ha under cultivation (Gregory 2002). Grading potato tubers by size in these
countries, especially for seed, is done by hand, which is very time-consuming
and is not cost-effective.
There are four basic types of potato-grading machines in current use:
(1) endless screen grader (conveyor); (2) roller grader; (3) cribriform drum
grader; and (4) reciprocating riddle grader (Kolchin 1982). These machines
are designed for medium- to large-scale farms and are not suitable for most
farmers in developing countries. A reciprocating riddle grader was developed
in India with a throughput of 2,500–3,000 kg/h (Shyam et al. 1990). Ten to
14 attendants were required for the operation of this machine, which worked
manually and electromechanically. A similar machine that takes three to five
people to operate was designed for small-scale farms in Bolivia (Butler et al.
2005). A reciprocating riddle grader was designed in Iran (Khojastapour
1996), but it was not based on research into the physical properties of Iranian
potatoes. Potato damage is a problem in reciprocating machines and they
work better with spherical tubers (Witney and McRae 1992). A potato-grading
machine with parallel conical rollers (Butler et al. 2005) was designed in
Peru, but its complex construction makes it inappropriate for small farms.
There is very little information in the literature concerning the design and
development of small potato graders. We studied the physical properties of
Iranian potatoes in order to design an appropriate small grading machine that
is simple in construction and less labor-intensive than grading by hand. In
view of the design demands, the use of a capron net on a drum-type size
grader was investigated.
 SMALL POTATO-GRADING MACHINE 3

MATERIALS AND METHODS

Physical Characteristics of Potato Tubers

We examined the physical characteristics of the potato variety Morfana,
which is one of the most commonly grown varieties in Iran. A total of 500
samples were selected randomly from different farms in Hamedan province
during the potato harvest. A caliper was used to measure three perpendicular
axes for each tuber with 0.1 mm accuracy. The weight of each tuber was
measured on an electronic balance to 0.02 g. The mean, standard deviation,
maximum, minimum, coefficient of variation for each perpendicular axis, and
mass were calculated (Table 1).
Coefficients of sliding and rolling friction of the tubers on the capron net
were determined using a laboratory slope-meter apparatus consisting of a
topless box with its upper side covered by a capron net. The diameter of the net
yarns was 4 mm, and they were made of woven fibers. The stiffness of the net
(T) was defined as the force required for the unit vertical displacement of the
net at the middle of the mesh surface (Ghanbarian 2005):
F
T= (1)
δ
where F is vertical force (in N) and d is vertical displacement (in mm) at the
middle of the mesh.
The stiffness of the net was adjusted by two regulators placed on two
adjacent sides of the box. The tubers were put on the capron net parallel with
and vertical to the box length for determining the sliding and rolling coefficient
of friction, respectively (Kasparov et al. 1962). The slope-meter was pivoted
slowly until the tubers started to move, and the angle was read from the
slope-meter. The experiments were done with different sizes of tubers at two
levels of net stiffness, 0.1 and 0.8 N/mm.

TABLE 1.
THE PHYSICAL CHARACTERISTICS* OF THE POTATOES (MORFANA VARIETY)

Statistical parameters

Physical Maximum Minimum Mean Standard Coefficient

properties deviation of variation

a 170 30.1 73.3 22.7 30.9

b 109 26.3 60.5 16.9 27.9
c 112 26.3 51.4 13.1 26.4
Mass 833.41 11.12 155.9 128.82 82.6

* a, b and c are major, intermediate and minor diameters, respectively.

4 D. GHANBARIAN ET AL.

Modeling the Mass of Tubers by Dimensional Attributes

Microsoft Excel 2003 spreadsheet software (Microsoft Inc., California)
was used to analyze the data and determine the regression models between
the mass and the size parameters of tubers. The size parameters were (1)
major; (2) intermediate; and (3) minor diameters in addition to parameter
S = 0.708 b2 + c 2 (Goryachkin 1968). Various regression models were used
to estimate the mass of a tuber from the size parameters. The model
m = 0.0008 S3.0412 (R2 = 0.93) gave the best fit and was used to determine
the preliminary dimensions of the mesh holes. Therefore, the new grading
machine will sort potato tubers by parameter S.
Design Considerations and Machine Construction
It is important that the machine provides efficient and cost-effective
grading for a wide range of tuber sizes and shapes. Furthermore, it should be
of simple construction and be operable by an electric motor or by hand. The
machine must be able to accept manual feeding, which will usually be done by
emptying baskets, which is an intermittent process. Among the different types
of common potato-grading machines, the cribriform drum grader is considered
the simplest, and intermittent feeding has the least influence on the performance
of this type of grader (Kolchin 1982). Therefore, a small drum-type grading
machine was built and developed (Kolchin et al. 2006). To reduce the mechani-
cal damage to tubers, which is a recognized drawback of this type of machine,
a capron net was used as the grading surface. The machine was required to
sort potato tubers into three sizes: small (<50 g), medium (50–80 g) and large
(>80 g). The final dimensions of the mesh holes for each part of the grading
surface were calculated by the following equation (Kolchin 1982; Peleg 1985):

⎡ μ σ 2 − μ1σ 22 ⎤
X12 = ⎢ 2 12 ⎥
⎣ σ1 − σ 2 ⎦
2

12
⎡ ⎛ σ1 ⎞ ⎤ (2)
⎢⎛ μ σ 2 − μ σ 2 ⎞ 2 μ2 σ 1 − μ1 σ 2 − 2σ 1 σ 2 ln ⎜⎝ σ ⎟⎠ ⎥
2 2 2 2 2 2

± ⎢⎜ 2 12 1 2
⎟⎠ −
2
⎥
⎢ ⎝ σ 1 − σ 2
2 σ 2
1 − σ 2
2 ⎥
⎢⎣ ⎥⎦

where X12 is the size of the mesh holes (S) (mm), m1 and m2 are the means of
S for small and medium sizes (mm), respectively, and s1 and s2 refer to the
standard deviation of S for small and medium sizes, respectively.
The dimensions of the mesh holes for the first and the second part of the
grading surface were calculated as 35 ¥ 35 mm and 45 ¥ 45 mm, respectively.
 SMALL POTATO-GRADING MACHINE 5

FIG. 1. THE TECHNICAL DRAWING OF POTATO-GRADING MACHINE

All measurements in mm.

The machine is composed of a spiral, a net, a power transmission system, an

electric motor, receiving trays and a feeding tray as shown in Fig. 1. The inside
diameter of the rotating drum containing the spiral is recommended to be
500–1,000 mm (Kolchin 1982). For a feeding rate of 1,500–3,000 kg/h, we
selected a drum length of 1,300 mm and a spiral pitch of 380 mm. The frame
is 2,200 mm long, 920 mm wide and 1,000 mm high, and is made of
40 ¥ 40-mm steel L-beams. The spiral is made of 10-mm diameter steel rod
and is connected to a 45-mm diameter steel shaft that can be driven by hand or
by an electric motor (a 368 W, 220 V motor was used in the model described
here). The feeding and receiving trays are made of 2.0 mm thick steel sheet.
Potato tubers are placed onto the feeding tray and then rolled down into the
rotating drum, where the spiral structure moves them forward. The tubers
move on the net and drop down to the receiving tray whenever the S parameter
of a tuber is less than the hole size of the mesh, so that small tubers are sized
before big tubers. The specifications of the machine are given in Table 2.

Machine Optimization
Generally, the optimization parameters to be addressed for any potato-
grading machine are feeding rate, accuracy of grading and lack of mechanical
damage to tubers (Ghanbarian 2005). Because of the use of a capron net,
6 D. GHANBARIAN ET AL.

TABLE 2.
SPECIFICATIONS OF GRADING MACHINE

Description Value

Throughput, kg/h 1,500–3,000

Rotating drum speed, rpm 27
Sizes of drum (length ¥ diameter), mm 1,300 ¥ 620
Overall dimensions 2,200 ¥ 920 ¥ 1,000
(length ¥ width ¥ height), mm
Power requirement, w 368
Mass (Manual drive/Electric drive), kg 45/90
Attendants, man 3

mechanical damage to tubers is decreased, and the feeding rate is less impor-
tant in a small-scale grading machine. The size of tubers is a key characteristic
of selection for seed potatoes; thus, accuracy of grading, Y, was considered
as the most important optimization parameter, and was determined by the
following equation (Kolchin 1982):

∑ mi (3)
Y= 1
mt

where mi is the mass of tubers of size i dropping correctly into receiving tray
size i (kg), mt is the total mass of graded tubers of all sizes (kg), and N is the
number of sizes.
The accuracy of grading (Y) is influenced mainly by the stiffness of the
net (T), the speed of the rotating drum (n), and the feeding rate (q). So, it can
be defined as a function of these three factors, i.e.:

Y = f (T , n, q ) (4)

The conditions for optimum machine performance were determined

according to Response Surface Methodology (Montgomery 1991) and by the
use of STATGRAPHICS Plus software (Manugistic, Inc., Rockville, MD). For
this purpose, we selected the multiple linear regression shown as Eq. (5):

Y = a1 X1 + a2 X 2 + a3 X3 + a0 (5)

where a1, a2, a3, a0 are the regression coefficients, and X1, X2, X3 are the
independent variables (T, n and q, respectively).
 SMALL POTATO-GRADING MACHINE 7

TABLE 3.
LEVELS AND VARIATION INTERVALS OF
THE FACTORS AT SECOND STAGE OF EXPERIMENTS
(BOX–BEHNKEN DESIGN)

Levels

Factors Low Basic High Variation interval

X1 0.1 0.2 0.3 0.1

X2 26 31 36 5
X3 1,800 2,500 3,200 700

The feeding rate (X3) was adjusted by a conveyer and an inverter. The
rotating drum speed (X2) was controlled by the inverter. A special regulator was
used to adjust the stiffness of the net (X1) by pulling it in two directions. Based
on the size of the tubers, they were marked with three different colors. The
experiments followed a factorial experimental design of 23 with two additional
central points. Each treatment was done in triplicate. After each repeat, the
accuracy of grading was determined by Eq. (3). If the model is not suitable in
the first stage, the second stage of optimization should proceed. Determination
of the regression coefficients and analysis of variance (ANOVA) showed that
Eq. (5) is not suitable to describe the grading process. Thus, there is a curva-
ture at the response surface of the model and preliminary experiments were
done in the correct zone. Therefore, a multiple nonlinear regression model was
selected at the second step of the experiments and is shown as Eq. (6):

Y = a1 X1 + a2 X 2 + a3 X3 + a4 X12 + a5 X 22 + a6 X32 + a7 X1 X 2
(6)
+ a8 X 2 X3 + a9 X1 X3 + a0

In this stage, 15 experiments were done according to the Box–Behnken

design (Montgomery 1991). The levels and variation intervals of the factors
are shown in Table 3.

RESULTS AND DISCUSSION

The coefficients of rolling and sliding friction of potato tubers on the

capron net with different degrees of net stiffness were determined as 0.46 and
1.07, respectively. The results showed that the coefficients of friction are not
correlated with the stiffness of the net and they are almost 1.5-fold greater than
similar amounts on other surfaces, such as galvanized steel or fiberglass.
Therefore, compared with similar machines that use only 12–16% of the
8 D. GHANBARIAN ET AL.

FIG. 2. EFFECT OF NET STIFFNESS AND ROTATING DRUM SPEED ON ACCURACY

OF GRADING

grading surface (Kolchin 1982), the tubers move up further on the internal
surface of the drum, increasing the effective surface in the new grading
machine.
The results of the experiments were approximated by a regression model
in the form of a second-degree polynomial:

Y = −5.2188T − 1.23n + 2.0038q + 2.8937T 2 − 1.7388n2

(7)
−3.15Tn + 62.59

The ANOVA showed that Eq. (7) is significant at the 95% confidence
level (R2 = 0.88). Eq. (7) allows us to estimate the effect of the major param-
eters (T, n and q) on the accuracy of grading (Y) and to determine the optimum
condition of machine performance. The effect of parameters T, n and q on the
accuracy of grading are shown in Figs. 2–4. The upper limits of net stiffness
and rotating drum speed are not suitable for grading potato tubers. Under these
conditions, while potato tubers continuously leap on the net, they pass all the
way through the rotating drum and do not have time to drop through the mesh
holes. The effect of parameter q on the accuracy of grading is shown in Figs. 3
and 4. It is clear that to obtain maximum accuracy of grading, the tubers must
be in a single layer on the grading surface; otherwise, the accuracy of grading
is rapidly decreased. In this study, the increase of feeding rate led to increased
grading accuracy, which may be because more collisions between tubers
result in a better position and orientation to drop through the mesh holes. The
optimum condition of machine performance was determined by taking
partial derivatives from Eq. (7) as follows: T = 0.14 N/mm, n = 27 rpm and
q = 2,500 kg/h.
 SMALL POTATO-GRADING MACHINE 9

FIG. 3. EFFECT OF NET STIFFNESS AND FEEDING RATE ON ACCURACY OF GRADING

FIG. 4. EFFECT OF ROTATING DRUM SPEED AND FEEDING RATE ON ACCURACY

OF GRADING

On the basis of a feeding rate of 2,500 kg/h and the national production
average of 22,000 kg/ha, the machine can grade potato tubers from 1 ha in
8.8 h. An experienced human grader can sort 125 kg/h, and it would therefore
need a minimum of 20 people to sort the same mass of potatoes in the same
time. In contrast, three people are needed to operate the motor-driven machine,
or four people if the machine is hand-operated. So, the machine can sort five
times more tubers per person per unit time than hand grading. The machine
costs approximately US$800, which is considerably cheaper than existing
grading machines. The new grading machine performed with an acceptable
accuracy of grading of about 70% under optimum condition. The mechanical
damage to tubers in the form of external damage such as cracks and scuffing
was only 5.5%, which is less than the extent of damage caused by existing
grading machines.
10 D. GHANBARIAN ET AL.

ACKNOWLEDGMENTS

The authors would like to thank the Ministry of Science, Research and
Technology of Iran for the financial support. They would also like to thank
Dr. V.M. Khalancki of the Moscow Timiriazev Agricultural Academy for his
valuable guidance.

REFERENCES

BUTLER, G.P., BERNET, T. and MANRIQUE, K. 2005. Mechanization of

potato grading on small-scale farms: A case study from Peru. Exp. Agric.
41, 81–92.
GHANBARIAN, D. 2005. Substantiation of basic parameters of drum
potato grader using capron net. Tractors and Agric. Mach. 7, 31–33
(in Russian).
GORYACHKIN, V.P. 1968. About Potato Grading. Collected Works,
pp. 190–202, Kolos, Moscow (in Russian).
GREGORY, J.S. 2002. Maps, models, and muddles: World trends and patterns
in potatoes revisited. Potato Res. 45(1), 45–77.
KASPAROV, C.A., NIKOOLIN, L.B. and MILTCEVA, L.V. 1962. Physical
and mechanical properties of potato tubers. The Works of Viskhom. 32,
13–34 (in Russian).
KHOJASTAPOUR, M. 1996. Design and fabrication method of potato sorting
machine according to Iran conditions. MSc Thesis, Tehran University,
Iran (in Persian).
KOLCHIN, N.N. 1982. Complex Machine and Implements for Postharvest
Operations of Potatoes and Vegetables, Machinestroienie, Moscow
(in Russian).
KOLCHIN, N.N., LIAMENKOF, U.A., GHANBARIAN, D., KHALANCKI,
V.M., ROUBTCOF, C.V. and KRAKHOVETCKI, N.N. 2006. Drum-
type grading machine (Sortirovka barabannovo tipa). Russian patent
No. 2271092.
MINISTRY OF JIHAD -E- AGRICULTURE (MJA). 2007. Statistical
Yearbook. Iran (in Persian).
MONTGOMERY, D.C. 1991. Design and Analysis of Experiments, John
Wiley & Sons, New York.
PELEG, K. 1985. Produce Handling, Packaging and Distribution. AVI Pub.
Co. Inc., Connecticut, CT.
SHYAM, M., SINGH, R. and SINGH, V. 1990. Design and development of
potato grader. Agric. Mechan. Asia Africa Latin America, 21(1), 40–42.
 SMALL POTATO-GRADING MACHINE 11

WITNEY, B.D. and MCRAE, D.C. 1992. Mechanization of crop production

and handling operations. In The Potato Crop, the Scientific Basis for
Improvement, 2nd Ed. (P.M. Harris, ed.) pp. 570–607, Chapman and Hall,
London.