Research Article

Branna Journal of Engineering and Technology (BJET)

Vol. 2, No. 1: May, 2020, pp. 41 – 58

Optimizing Cotton Fiber Mixing Cost with Respect to Quality of Yarn: Using
Integer Programming
Abdelkader Kedir Biadglignea*, Getachew Basa Bonsaa, Hailekiros Sibhato Gebremichaela
a
School of Mechanical and Industrial Engineering, Ethiopian Institute of Technology-Mekelle, Mekelle University,
Mekelle 231, Ethiopia;

*Corresponding author; e-mail address: [email protected]

Abstract

In the globalized and dynamic environment, garment and textile industries are required to produce high-
quality yarn and fabrics at optimum cost in order to get profit and competitive advantages. This, in turn,
needs the proper integration of fiber quality in the process of yarn production which is heavily dependent
on cotton fiber selection and cotton fiber blending. The quality of the resulting yarn is very important in
determining the profitability of the company. The yarn should have optimal product characteristics with
minimum cost. This can be achieved through a rigorous program that determines the right fiber quality in
the process of yarn manufacturing. This involves the process of the right cotton fiber selection and cotton
fiber blending. In this study, an Integer Programming has been formulated for the cotton-blending
problem that determines the right mix required to manufacture a specific yarn count. The objective is to
minimize the cost of quality in the form of quality give away while satisfying demand of each yarn count
with its required quality characteristics. The model developed has been tested and verified using testbed
data collected from a case company, MAA Garment, and Textile Factory. Excel 2016 has been used as a
solution approach on a computer with 4GB of RAM and intel core i5 processor. The solution to the
problem has been found in less than a fraction of a second. The solution found has the potential of saving
762,955.37 ETB annually that would otherwise have been wasted as a quality giveaway.

Keywords: Fiber, Integer programming, Optimum mixing ratio, Yarn, Yarn quality.

Submitted on: November 11, 2019 Accepted on: April 29, 2020
 42
p- ISSN: 2664-0791, e-ISSN: 2664-0716

1. Introduction

In today’s competitive market, manufacturing and service organizations strive to reduce their
costs and maximize their profitability through the application of various cost reduction
techniques. In the case of the Garment and Textile factories, the proper integration of fiber
quality in the process of yarn production which is heavily dependent on cotton fiber selection
and cotton fiber blending is an important cost reduction technique[1]-[3]. The use of individual
fiber requires a high degree of certainty of individual characteristic effects on quality and
processing performance which extremely dependent on Manufacturers’ experiences in deciding
these levels. However, a deficit in one fiber characteristic could be compensated by an
improvement in another, which makes fiber blending an important and essential process to
minimize cost while meeting desirable quality characteristics of the yarn [4]. In addition,
desirable fiber quality is one that provides high-quality yarn, good processing performance and
maximum yarn yields. These parameters necessitate mixing between different fibers
characteristics during processing [5]-[6]. In this regard, researchers have investigated the various
types of fiber mixes and the characteristics of the mixed yarn made from the various types of
fibers. The findings show that different types of cotton fiber in the yarn have different features
depending on luster, weight, stiffness, comfort, and price of the fabric [5]-[10]. These
characteristics exert a greater or smaller impact on the yarn produced depending on the blending
ratio of the pieces of cotton [11]. Besides, yarn with the required characteristics can be produced
with the proper mixing of different types of cotton fibers [7], [10]-[11]. The blending ratio is one
of the most important factors that affect the performance of the blending fabric and the quality of
yarn produced [12]. The determination of the blending ratio mainly involves the development of
multi-component blending [11].

The spinning performance and quality of yarn are dependent on mixing [12]-[13]. Nevertheless,
traditionally the quality of cotton fiber was determined based on the fiber grade, length, and
fineness. The blending process was also dominantly based on individual expertise and
experience, making the mixing process highly subjective, time-consuming, and tedious [13]. In
this regard, massive blending with different qualities of bales based on grades or growth areas
were made to reduce variability [9]. However, the rising cost of labor, storage, equipment, and
raw material makes the old blending approach largely impractical. Moreover, the traditional fiber
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 43
p- ISSN: 2664-0791, e-ISSN: 2664-0716

testing was tedious and time-consuming resulting in a small proportion of the bales being tested
in a long time [4], [14].
Different properties of yarn are spun for different fabrication purposes [12]. While the
production of woven fabric needs high strength to sustain the weaving process, the premium
weaving products need a finer touch yarn. These requirements determine the use of a high
proportion of cottons with high strengths and micronairs [3]. Other requirements such as the
color of the fabric also determine the brightness and yellowness of cotton that need to be used.
Such a mixing recipe is a key element to associate the consumption of cotton types with different
yarn products [13]. Nevertheless, few papers investigated and formulated the cotton mixing
techniques for economic production of yarn. For example, [15] proposed a multi-objective
optimization model to find the optimal blending ratio of cotton in yarn production; [16]
developed a fuzzy optimization model to maximize the strength of yarn with a minimum quality
of cotton for economic considerations; [11] used a fuzzy linear program to optimize raw material
cost in the cotton spinning industry; [17] utilized genetic algorism to search for woven fabric
parameters for obtaining desired fabric weight at minimum cost; [7] applied the Linear
Programming Technique to determine the optimum mixing ratio to produce desired yarn quality
at minimum cost; [13] suggested an empirical formula to mix cotton for the optimal yarn quality,
[18] used a multivariate quality loss function approach to simultaneously optimize multiple yarn
characteristics to minimize the total loss in the quality of the considered product. Still, these
works merely focused on specific properties of yarn that could not cover a wide spectrum of
products.
The spinning mills, quality of cotton, and proper mixing are the critical parameters for producing
proper and high-quality yarn [13],[15], [19]. Moreover, the different sorts, crops, batches, and
bales make cotton to have an important impact on the structural, physical, and mechanical
features [14]. These properties have strong effects on the processing performance and quality of
the yarn [19]. Therefore, proper selection and mixing of cotton are important in the production of
yarns with specified quality requirements [10], [13]. Apart from the quality implication, raw
cotton is the largest component that contributes 60-70 percent of the total production cost of
cotton yarn [7], [8],[20]. A fraction of savings in cotton cost will substantially increase the gross
profit of the mill [4], [8], [20]. Hence, the economic impact of the blending of cotton results from
the ability to reduce the cost and improve the quality of yarn produced [4], [8]. This is achieved
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 44
p- ISSN: 2664-0791, e-ISSN: 2664-0716

by controlling the percentage of fiber components in the blend and the use of available quality
cotton efficiently.
Through time statistical methods were developed to determine the overall fiber properties and to
decide mixing parameters for achieving particular yarn quality fast and accurately[9]. These
techniques can reduce the man-hour and time required and have an economic impact as the
appropriate blending results in cost reduction with the required quality [4], [21]. The judicious
mixing of different types of cotton fibers is an optimization problem for obtaining raw material
with the required quality at a minimum cost in the cotton spinning industries [11]. In this study,
the Integer Programming technique has been used to determine the appropriate mixing ratio to
produce the desired yarn quality at optimum cost.
In Ethiopia, the cotton blending for yarn production is done in a traditional process. The mill
operators determine the mix ratio of bales based on intuition and personal experience, which
results in a high cost of blend and quality giveaways. This process consumes a long time, creates
delays and variations in the quality of yarns produced. Therefore, developing a systematic
mixing that maintains the lowest possible variation in cotton quality parameters and that improve
the yarn production and quality at optimal cost is imperative. Hence, the objective of this paper is
to formulate the Integer Programming (Mathematical model) to determine the mixing ratio of
various types of cotton fibers to produce desired yarn quality at optimum cost. The cotton
blending process involves complex quality control; it is particularly responsive to Integer
Programming (IP) techniques. By the use of IP, the mill operator can determine the specific
allocation of raw cottons required to produce a given blended yarn at minimum cost subjected to
stated restrictions on yarn quality and raw cotton availabilities. The model is solved and verified
by taking data from the spinning mill section of a case company. Besides, the model is
implemented in a spreadsheet to make it practically applicable and to make use of this model by
the operators with little mathematical knowledge or skill.
The remaining part of the paper is organized as follows: The research methodology which
describes the data source, data analysis procedures, bale management process, model
development, and mathematical formulation of the problem is presented in section two. The
results and discussion part which consists of optimization equation formulation, spreadsheet
representation of the mathematical model, computational results, and discussion of the numeric
result is organized under section three. Finally, the conclusion part is described in section four.
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 45
p- ISSN: 2664-0791, e-ISSN: 2664-0716

2. Research Methodology

2.1 Data source and data analysis procedures

This study was conducted in the Northern part of Ethiopia, Tigray Regional State. MAA
Garment and Textiles Company was taken as a case study. The data collected from the case
company was used to demonstrate and verify the optimization equation formulated using integer
programming.

2.2 Model development

One of the fundamental problems that a mill operator faces during the production of yarns is
determining the proper mix ratio of bales so that the cost of mixed fiber is minimal and the
desired quality of the mix achieves the required result. Currently, with the lack of a cotton blend
model, mill operators determine the mix ratio of bales based on intuition and personal
experience, which results in a high cost of a blend, which oftentimes occurs in the form of a
quality giveaway. In addition to the high cost of a blend, the current mixing ratio determination
by the operator is time-consuming creating unnecessary delays and is also a primary source of
variability in the quality of yarns produced. Hence, in order to produce a least-cost blend and a
uniform quality yarn, it is imperative to develop a mathematical model that would determine the
appropriate amount of cotton to blend from each bale at minimum cost.
The model proposed in this paper will enable the case company and more specifically the
spinning mill planner to make an informed decision regarding how much and which of the cotton
types be procured. This minimizes the cost of bale mixing by determining the right (optimal)
mix, the number of yarn defects which would arise from wrong blends, the number of quality
giveaways (super standard yarns) and enables to procure the optimal quantity with the right
quality of each type of bales that would minimize the cost of procurement and model costs.
Table1 below shows the cotton blending problem formulation, a tabular representation of the
problem makes easier to understand the problem.

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 46
p- ISSN: 2664-0791, e-ISSN: 2664-0716

Table 1. Tabular formulation of the problem

Yarn count (j)

10 20 30 40 60 Total amount of Supply
Cotton used
1 X11 X12 X13 X14 X15 X11 + X12 +…+ <= S1
X15
Cotton 2 X21 X22 X23 X24 X25 . <= S2
Type (i) 3 X31 X32 X33 X34 X35 . <= S3
. . <= .
. . <= .
. . <= .
N Xn1 Xn2 Xn3 Xn4 Xn5 Xn1 + Xn2 +…+ <= Sn
Xn5
Total amount X11 + X21 . . . X15 + X25
of Cotton +…+ Xn1 +…+ Xn5
used
= = = = =
Demand D1 D2 D3 D4 D5

The above table designates the essence of the cotton mixing problem, where the X’s represent
the amount of cotton form cotton type i that should be used to produce yarn type j. Hence the
problem of cotton mixing is the problem of finding the right amount of cotton with the right type
of fiber characteristics to meet the demand of the different types of yarn counts that should be
produced during a planning horizon. The demand and supply section, D’s and S’s respectively,
specify the amount of yarn required to be produced and the number of cotton bales available
respectively.

2.2.1 Existing bale management process

Currently, the case company makes raw cotton procurement decisions based on the assumption
that bales of high quality (specifically, fiber staple length) should be procured as much as
possible. If the total number of such bales is not enough for a specific production planning
horizon, then the next best bales will be acquired. This process continues until the total number
of bales procured matches the total demand. When this process is being carried no effort is made
to investigate alternative options (mixes of bales for example), which could minimize the total
cost of procurement and blending. That is, to produce a specified cotton yarn, with the minimum
quality give away cost, the operator needs to know the optimal blend ratio, which involves the

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 47
p- ISSN: 2664-0791, e-ISSN: 2664-0716

consideration of complex variety factors including, cotton prices, cotton quality, and supply of
each type of cotton.

2.2.2 Mathematical formulation of the problem

Consider a customer order for multiple types of yarn counts with their desired amount and
unique quality characteristics. Each type of yarn count has the desired level of quality
characteristics which are described by the values of its Fineness, Strength, Length, and yarn
count [10]. A yarn count is a number giving a measure of the yarn linear density. The linear
density is defined as the mass per unit length. It measures yarn coarseness or fineness. A brief
description of the important quality parameters is given below, which is essential in
understanding the mathematical model and the relations defined in those models. In order to
construct the model we need to define the design variables and parameters of the model, which
are defined below: Important quality parameters of the various yarn counts which have been
utilized in developing and solving the cotton blending model include as shown in Table 2 below:

Table 2. Quality parameter yarn counts used

No. Count 2.5% span Micronaire value Bundle strength Trash Content
Length (µg/inch) (gms/tex) (%)

1 20s 26-28mm 3.8-4 20.5-21.5 4.0

2 30s 27-29mm 3.6-3.8 21-22 4.0
3 40s 29-31mm 3.5-3.7 22-24 3.5
4 60s 31-33mm 3.3-3.6 25-27 3.0
5 80s 32-34mm 3.3-3.6 26-27 2.5
6 100s 33-35mm 3.3-3.6 26-28 2.0

Notation:
Design Variables
( )

Parameters
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 48
p- ISSN: 2664-0791, e-ISSN: 2664-0716

The Objective Function

The objective is to minimize the total cost of the blend while meeting the desired level of quality
for each type of yarn counts to be produced. Hence, the Objective Function is:

∑∑

Constraints

Fineness- Fiber fineness is a relative measure of the diameter or linear density of an individual
fiber. Cotton fineness is typically measured by an instrument, called a Micronaire that
determines the rate of airflow through a sample of fixed weight. The finer the fiber, the larger the
total surface area is in a standard sample which implies more resistance to airflow. Micronaire
fineness values are read directly from an air gage. Fineness is linear, directly proportional to the
number of neps (tangles of fiber) per square inch and also to spinnable limit (the spinnable limit
is the yarn count at which a particular blend meets the ends down specification). Since a high nep
count must be avoided, the fineness must ordinarily fall within the specified maximum and
minimum limits depending on the yarn count desired.
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 49
p- ISSN: 2664-0791, e-ISSN: 2664-0716

Staple length- staple length also called fiber length is defined usually as the upper-half mean
length (UHML) or 2.5% span length (2.5%SL) from a Fibrogram beard. The case company uses
Uster HVI Fibrograhp tester instrument to measure the fiber length and other important fiber
parameters. Two different kinds of fiber length measurements can be generated from a
Fibrogram1; mean lengths and span lengths. Mean lengths, for example, UHML, which is the
mean length by numbers of the longer half (50%) of the fiber by weight, and the mean length
(ML) are more commonly used since they relate the mean of percentages of fibers represented in
the Fibrogram. The following mathematical relationship describes the staple length of the final
mixed fiber as a weighted mean of the individual fiber length of each bale.

1. Fiber strength- is the measure of the force required to break a sample of 15-20 grams of
fiber, and is measured in the Presley index (P.I).

2. Demand- this equation related the amount of yarn type required to be produced with the
required amount of bales necessary to produce the required yarn.

3. Supply- the following constraint enforces the constraint that the demand for all types of
fiber bales cannot exceed the available amount of each type of bale.

1
ASTM D4604, Standard Test Method for Measurement of Cotton Fibers by High
Volume Instruments (HVI).
BJET Vol. 2, No. 1, May, 2020: 41 – 58
©EiT-M, Mekelle University
 50
p- ISSN: 2664-0791, e-ISSN: 2664-0716

3. Results and Discussions

Based on the model developed and mathematical equation formulated for the problem identified
and data collected the results and discussions are described below.

3.1 Formulation of the optimization equation

Table 3 below shows part of the data used to test and measure the performance and validate the
model proposed in this paper.

Table 3. Fiber Quality characteristics of different bales

Quality Characteristics Price Quantity
Micronaire value (µg/inch) Length Strength SCI Grade Per of bale
(mm) (g/tex) Kg
1 5.1 28.51 27.8 111 35.3 93
2 4.9 28.43 28.9 109 35.5 110
Cotton Type

3 5.02 28.12 28.9 116 36 280

. . . . . . .
. . . . . . .
. . . . . . .
30 3.72 26.33 25.3 80 30.1 133

As can be noticed from Table 3, the price of cotton per kg is different for different lots, which is
because the modeling of the quality giveaway is considered in this case to quantify a cost. Even
though the cost of cotton for 1 kg ranges usually from 28 to 36 Birr depending on the type of
cotton, the real cost (that is the internal cost of cotton to the company) of cotton could be
quantified using the critical quality characteristics of the cotton fiber [18]. According to the
literature and industrial practice, the following data shown in Table 4 is thought to represent the
contribution of each quality characteristic towards the price.

Table 4. Cost contribution of the four critical fiber characteristics

Quality Characteristics
Staple Length Fiber SCI Grade Fineness
Strength (micronaire)
Cost 45 40 10 5

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 51
p- ISSN: 2664-0791, e-ISSN: 2664-0716

Contribution in
%

Once the data regarding quality, price, and maximum available quantity of bales for the different
types of raw cotton have been collected and organized, these values have been substituted into
the generic model developed. Then the performance of the model was tested and validated
accordingly. The following equation has been obtained by taking part in the data and substituting
those values into the general model.

Objective Function:

∑∑

Upon substitution of all the values, 210 terms for the objective function are generated.

Similarly, the constraints could be formulated, and the expanded form of all the constraints is
shown below:

1. Fineness

2. Staple length

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 52
p- ISSN: 2664-0791, e-ISSN: 2664-0716

3. Fiber Strength

4. Non-negativity Constraint

3.2 Spreadsheet representation of the mathematical model

This spreadsheet illustrated in Table 5 below represents the cost matrix, which is the cost of
manufacturing a certain type of yarn using a given type of cotton. It also has the supply
constraint and the supply available of each type of cotton in bales.

Table 5. Cost Matrix

Yarn Type
10 20 30 40 60 Supply Supply Available(in
Constraint Bales)
1 35.30 35.30 35.30 35.30 35.30 0.00 120
2 35.50 35.50 35.50 35.50 35.50 0.00 110
3 36.07 36.07 36.07 36.07 36.07 0.00 280
4 34.19 34.19 34.19 34.19 34.19 0.00 380
5 36.31 36.31 36.31 36.31 36.31 0.00 157
Cotton Type

6 36.25 36.25 36.25 36.25 36.25 0.00 122

7 33.44 33.44 33.44 33.44 33.44 0.00 146
8 32.95 32.95 32.95 32.95 32.95 0.00 128
9 32.95 32.95 32.95 32.95 32.95 0.00 183
10 34.96 34.96 34.96 34.96 34.96 0.00 102
11 34.70 34.70 34.70 34.70 34.70 0.00 148
12 32.33 32.33 32.33 32.33 32.33 0.00 96
13 30.01 30.01 30.01 30.01 30.01 0.00 89

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 53
p- ISSN: 2664-0791, e-ISSN: 2664-0716

14 31.08 31.08 31.08 31.08 31.08 0.00 50

15 31.59 31.59 31.59 31.59 31.59 0.00 87
16 33.86 33.86 33.86 33.86 33.86 0.00 310
17 30.15 30.15 30.15 30.15 30.15 0.00 133
18 36.01 36.01 36.01 36.01 36.01 0.00 119
19 36.63 36.63 36.63 36.63 36.63 0.00 104
20 38.19 38.19 38.19 38.19 38.19 0.00 110
21 36.04 36.04 36.04 36.04 36.04 0.00 117
22 34.27 34.27 34.27 34.27 34.27 0.00 102
23 36.16 36.16 36.16 36.16 36.16 0.00 71
24 37.92 37.92 37.92 37.92 37.92 0.00 84
25 32.89 32.89 32.89 32.89 32.89 0.00 590
26 32.89 32.89 32.89 32.89 32.89 0.00 400
27 32.07 32.07 32.07 32.07 32.07 0.00 977
28 33.61 33.61 33.61 33.61 33.61 0.00 491
29 32.51 32.51 32.51 32.51 32.51 0.00 128
30 32.73 32.73 32.73 32.73 32.73 0.00 106

Table 6 below represents the variable values, which are the amount of cotton that should be used
from a given type of cotton (1, 2…) in producing a required type of yarn (10’s, 20’s,…).

Table 6. Variable matrix Table

Variable Matrix
Cotton Type 10 20 30 40 60
1 0.00 0.00 0.00 0.00 0.00
2 0.00 0.00 0.00 0.00 0.00
3 0.00 0.00 0.00 0.00 0.00
4 0.00 0.00 0.00 0.00 0.00
5 0.00 0.00 0.00 0.00 0.00
6 0.00 0.00 0.00 0.00 0.00
7 0.00 0.00 0.00 0.00 0.00
8 0.00 0.00 0.00 0.00 0.00
9 0.00 0.00 0.00 0.00 0.00
10 0.00 0.00 0.00 0.00 0.00
11 0.00 0.00 0.00 0.00 0.00
12 0.00 0.00 0.00 0.00 0.00
13 0.00 0.00 0.00 0.00 0.00
14 0.00 0.00 0.00 0.00 0.00
15 0.00 0.00 0.00 0.00 0.00
16 0.00 0.00 0.00 0.00 0.00
17 0.00 0.00 0.00 0.00 0.00

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 54
p- ISSN: 2664-0791, e-ISSN: 2664-0716

18 0.00 0.00 0.00 0.00 0.00

19 0.00 0.00 0.00 0.00 0.00
20 0.00 0.00 0.00 0.00 0.00
21 0.00 0.00 0.00 0.00 0.00
22 0.00 0.00 0.00 0.00 0.00
23 0.00 0.00 0.00 0.00 0.00
24 0.00 0.00 0.00 0.00 0.00
25 0.00 0.00 0.00 0.00 0.00
26 0.00 0.00 0.00 0.00 0.00
27 0.00 0.00 0.00 0.00 0.00
28 0.00 0.00 0.00 0.00 0.00
29 0.00 0.00 0.00 0.00 0.00
30 0.00 0.00 0.00 0.00 0.00

3.3 Discussions

Figure1 below represents how the spreadsheet model is designed to solve the mathematical
optimization equation. Following Figure 1, the result shown in Table 7 is obtained upon solving
the mathematical model presented in the previous section.

Figure 1 Spreadsheet model

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 55
p- ISSN: 2664-0791, e-ISSN: 2664-0716

Table 7 below indicates how much kg of cotton from each type of cotton should be used to
produce a required amount of yarn type. According to the result in Table 7, yarn count 10 could
be produced using cotton type 7,8,12, and 16 using of 146,1,43, and 310 kg of cotton
respectively. However, even though using one kg of cotton type 8 would result in the minimum
cost and the desired quality parameters, in practice the operator would shift the one kg of cotton
type 8 to cotton type 7.

Table 7. Numeric Result

Yarn Type Variable Matrix
10 20 30 40 60 Supp. Supp. 10 20 30 40 60
Const. Ava.
1 35.30 35.30 35.30 35.30 35.30 120.00 120 0 0 0 120 0
2 35.50 35.50 35.50 35.50 35.50 110.00 110 0 0 0 110 0
3 36.07 36.07 36.07 36.07 36.07 0.00 280 0 0 0 0 0
4 34.19 34.19 34.19 34.19 34.19 380.00 380 0 0 0 380 0
5 36.31 36.31 36.31 36.31 36.31 157.00 157 0 0 0 157 0
6 36.25 36.25 36.25 36.25 36.25 122.00 122 0 0 0 122 0
7 33.44 33.44 33.44 33.44 33.44 146.00 146 146 0 0 0 0
8 32.95 32.95 32.95 32.95 32.95 128.00 128 1 0 36 0 91
9 32.95 32.95 32.95 32.95 32.95 183.00 183 0 0 0 0 183
10 34.96 34.96 34.96 34.96 34.96 102.00 102 0 0 0 102 0
11 34.70 34.70 34.70 34.70 34.70 148.00 148 0 0 0 96 52
12 32.33 32.33 32.33 32.33 32.33 96.00 96 43 53 0 0 0
13 30.01 30.01 30.01 30.01 30.01 89.00 89 0 89 0 0 0
14 31.08 31.08 31.08 31.08 31.08 50.00 50 0 0 50 0 0
15 31.59 31.59 31.59 31.59 31.59 87.00 87 0 0 0 87 0
16 33.86 33.86 33.86 33.86 33.86 310.00 310 310 0 0 0 0
17 30.15 30.15 30.15 30.15 30.15 133.00 133 0 0 132 1 0
18 36.01 36.01 36.01 36.01 36.01 119.00 119 0 0 0 119 0
19 36.63 36.63 36.63 36.63 36.63 103.00 104 0 0 0 103 0
20 38.19 38.19 38.19 38.19 38.19 2.00 110 0 0 2 0 0
21 36.04 36.04 36.04 36.04 36.04 95.00 117 0 0 95 0 0
22 34.27 34.27 34.27 34.27 34.27 102.00 102 0 0 101 0 1
23 36.16 36.16 36.16 36.16 36.16 71.00 71 0 0 0 71 0
24 37.92 37.92 37.92 37.92 37.92 0.00 84 0 0 0 0 0
25 32.89 32.89 32.89 32.89 32.89 590.00 590 0 590 0 0 0
26 32.89 32.89 32.89 32.89 32.89 400.00 400 0 400 0 0 0
27 32.07 32.07 32.07 32.07 32.07 977.00 977 0 0 977 0 0
28 33.61 33.61 33.61 33.61 33.61 491.00 491 0 134 357 0 0
29 32.51 32.51 32.51 32.51 32.51 128.00 128 0 128 0 0 0

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 56
p- ISSN: 2664-0791, e-ISSN: 2664-0716

30 32.73 32.73 32.73 32.73 32.73 106.00 106 0 106 0 0 0

Following the same fashion yarn count, 20 should be produced mixing cotton types 12, 13, 25,
26, 28, 29, and 30 with 53, 89,590, 400, 134, 128, and 106 kg of cotton respectively. To produce
yarn count 30, (36, 50, 132, 2, 95, 101, 977, 357) kg of cotton from cotton types of 8, 14, 17, 20,
21, 22, 27, and 28 should be used respectively. The results for the rest of the yarn types can be
seen from the table which are highlighted in green and bold. The cost of this daily production
schedule is 185,315.87 ETB. However, if the company were to use its production mix plan the
cost would have been 188,367.69 ETB per day. That is, with its production mix plan the
company would lose a total of 3,051.82 ETB per day and 762,955.37 ETB annually.

4. Conclusion

This paper has defined and formulated the cotton mix problem, commonly known in the textile
industry as, bale management using integer programming. The paper formulated a generic
mathematical model of the problem, data used from a case company to test the validity and
applicability of the model to a real-world problem. Then the model developed was executed
using EXCEL 2016 on a 4GB RAM Intel core i5 processor. The solution time required to solve
the problem took less than a fraction of a second, which makes the model applicable and easy to
use with an existing facility of the case company and similar companies. Besides, the model
developed and verified saves time for the operator, brings consistency in the bale mixing
requirement which ultimately boosts the company quality of yarn production. The major
advantage of this model is that it enables the company to save 762,955.37 ETB annually which
otherwise would have been wasted as quality giveaways.

The model developed has important managerial and practical implications in producing
consistent and better quality yarn products at minimum cost and minimal technical and
mathematical knowledge/skill requirements. Though it is not common in the case company and
other companies in Ethiopia, the use of mathematical modeling in the textile industry is common
and versatile. The use of such models make the production process consistent and productive, as
the subjective decision is minimized. Hence, modeling the manufacturing process using a

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 57
p- ISSN: 2664-0791, e-ISSN: 2664-0716

mathematical model like integer programming to aid decision making is critical and
recommended.

Acknowledgment

The researchers are indebted to Mekelle University and the Ethiopian Institute of Technology-
Mekelle for the financial support given.

References
[1] S. Raian, T. Siddiqua, M. L. R. Shanzid, and T. R. Tumpa, “Optimization of raw material mixing
in rotor spun yarn using analytical hierarchy process,” Int. J. Sci. Technol. Res., 2019.
[2] C. F. J. Kuo, C. Y. Kao, and H. J. Wei, “Optimization of open-end rotor spinning frame parameter
and estimation of relevant quality characteristics,” Polym. - Plast. Technol. Eng., 2011.
[3] J. Foulk, W. Meredith, D. McAlister, and D. Luke, “Fiber and yarn properties improve with new
cotton cultivar,” J. Cotton Sci., 2009.
[4] R. A. Sheikh and U. A. Lanjewar, “Decision Support System for Cotton Bales Blending Using
Genetic Algorithm,” Int. J. Comput. Appl., 2010.
[5] Y. E. e. Mogahzy, “Optimizing Cotton Blend Costs with Respect to Quality Using HVI Fiber
Properties and Linear Programming:Part I: Fundamentals and Advanced Techniques of Linear
Programming,” Text. Res. J., 1992.
[6] R. L. Long, M. P. Bange, C. D. Delhom, J. S. Church, and G. A. Constable, “An assessment of
alternative cotton fibre quality attributes and their relationship with yarn strength,” Crop Pasture
Sci., 2013.
[7] S. Bhattacharyya, D. Sur, and S. M. Chatterjee, “Optimization of Mixing Ratio and Mixing Cost
for the Production of Ring-Spun Cotton Yarn,” J. Nat. Fibers, vol. 5, no. 3, 2008.
[8] L. S. Admuthe and S. D. Apte, “Optimization of spinning process using hybrid approach involving
ANN, GA and linear Programming,” in Proceedings of the 2nd Bangalore Annual Compute
Conference, COMPUTE’09, 2009.
[9] Rajiwada, “Practices of cotton fiber selection and procurement for optimum mixing.,” Text. Eng.
Inst., 2009.
[10] M. Hassen and S. Faouzi, “Quality prediction & optimizing cotton blend using ANN.,” Indian
Text. J., 2007.
[11] S. Das, A. Ghosh, and D. (2013). Banerjee, “Optimization of raw material cost in cotton spinning
industry using fuzzy linear program.,” Ind. Eng. J., vol. 6, no. 6.
[12] Y. Yao and J. Duan, “Technical problems and research thoughts on development of multi-
component blending yarn.,” Adv. Mater. Res., vol. 317–319, pp. 817–822, 2011.
[13] S. K. Sharma, “Cotton Yarn : Quality Depends on Mixing Strategy,” Indian Text. J., vol. 124, no.
6, pp. 35–42, 2014.
[14] A. Majumdar, A. Das, P. Hatua, and A. Ghosh, “Optimization of woven fabric parameters for
ultraviolet radiation protection and comfort using artificial neural network and genetic algorithm,”
Neural Comput. Appl., 2016.
[15] Y. Zhang, B. Sun, Y. Yang, C. Han, W. Chen, and M. Li, “Multi-objective optimization of

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University
 58
p- ISSN: 2664-0791, e-ISSN: 2664-0716

blending ratio of Outlast fiber/cotton blending yarn,” in Advanced Materials Research, 2011.
[16] S. Das, A. Ghosh, and D. Banerjee, “Optimization of Raw Material Cost in Cotton Spinning,” Ind.
Eng. J., vol. 6, no. 6, 2013.
[17] J. J. Lin, “A Genetic Algorithm for Searching Weaving Parameters for Woven Fabrics,” Text. Res.
J., 2003.
[18] S. Chakraborty and A. Mitra, “A Multivariate Quality Loss Function Approach for Optimization
of Spinning Processes,” J. Inst. Eng. Ser. E, 2018.
[19] R. Harpa, A. Curteza, I. Cristian, and C. Piroi, “Optimizing the yarn quality by means of cotton
mixture design,” in ITC and DC: Book of Proceedings of the 4th International Textile, Clothing
and Design Conference - Magic World of Textiles, 2008.
[20] S. Yang and S. Gordon, “Accurate prediction of cotton ring-spun yarn quality from high-volume
instrument and mill processing data,” Text. Res. J., vol. 87, no. 9, pp. 1025–1039, 2017.
[21] S. Dame, The American Cotton handbook. 1965.

BJET Vol. 2, No. 1, May, 2020: 41 – 58

©EiT-M, Mekelle University