International Journal of Clothing Science and Technology

A prediction method of material consumption for flat-knitted sweaters

Xuan Luo, Gaoming Jiang, Honglian Cong,
Article information:
To cite this document:
Xuan Luo, Gaoming Jiang, Honglian Cong, (2017) "A prediction method of material consumption for
flat-knitted sweaters", International Journal of Clothing Science and Technology, Vol. 29 Issue: 5,
pp.661-672, https://doi.org/10.1108/IJCST-06-2016-0065
Permanent link to this document:
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

https://doi.org/10.1108/IJCST-06-2016-0065
Downloaded on: 12 September 2017, At: 04:23 (PT)
References: this document contains references to 15 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 5 times since 2017*
Access to this document was granted through an Emerald subscription provided by emerald-
srm:126209 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.

 The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0955-6222.htm

Prediction
A prediction method of method of
material consumption for material
consumption
flat-knitted sweaters
Xuan Luo, Gaoming Jiang and Honglian Cong 661
Engineering Research Center for Knitting Technology, Jiangnan University,
Received 4 June 2016
Wuxi, China Revised 22 March 2017
Accepted 30 March 2017

Abstract
Purpose – A method for predicting the material consumption of a sweater is presented before it is knitted.
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

It can be achieved with the five basic models combined with the parameters related to the dimensions of the
knitting machine and needles. The paper aims to discuss these issues.
Design/methodology/approach – Based on the parameters of the needle bar flat knitting machine, the
sweater is modeled with five basic structures. The mathematical expression of each basic structure can
be derived with corresponding parameters under some consumptions. In following, the predictive weight of
the sweater can be formulated with the expression of the length of the basic structures and the linear density
of the yarn.
Findings – To evaluate the performance of the proposed scheme, experiments of three types of sweaters on
four different knitting machines are carried out. The results show that the proposed method can achieve the
performance with the bias values by percentage ranging from −1.54 to −2.84 percent.
Research limitations/implications – Due to the present limited research, more experiments could
not be carried out. To improve the performance and robustness of the proposed method, statistical
performance measures such as the statistical mean and variance in massive experiments will be studied in
the further research.
Practical implications – The evaluation of the material consumption can be obtained before it is knitted
with the known basic parameters related to the machine and yarn.
Originality/value – This paper derives the general expressions of five basic structures based on the
corresponding parameters of knitting machine. The predictive weight of the sweater is expressed according
to the above basic structures before the sweater is knitted.
Keywords Flat-knitted sweater, General expression, Material consumption, Weight prediction
Paper type Research paper

Introduction
The material consumption of flat-knitted sweaters is very important to the manufacturers,
related to the product quality and the economic benefit. From the view point of practice, a
simple and direct way to obtain the weight and the material consumption of the sweater is
by weighing the sweater. However, for this method, it is required that the desired sweater
should first be knitted, which means that if the material consumption of the sweater does
not satisfy the requirement, it is very difficult to modify the design.
To alleviate the above drawback, a method to predict the weight of fabrics, a software
system called “proKNIT” was developed to predict the weight of plain-knitted fabrics and rib
stitch based on the length of fabrics (Gravas et al., 2005, 2006). The software is programmed
referring to theories about the loop length by previous researchers. A simple loop model was
proposed to deal with the single knitted structure to approximate the yarns and knitted
structures using a geometrical method (Peirce, 1947; Munden, 1959). Subsequently, researchers
have developed complex models to improve the accuracy of length prediction of the knitted
structure. The yarn diameter as a variable was involved in the expression of the length
International Journal of Clothing
of the knitting loop (Postle, 1968). Mechanical relaxation or chemical treatments were used Science and Technology
to keep the plain jersey fabrics dimensionally stable (Knapton et al., 1975). A factorial Vol. 29 No. 5, 2017
pp. 661-672
analysis of many combinations of factors was applied in the length calculation of the plain © Emerald Publishing Limited
0955-6222
knit due to influence of the shrinkage rate (Baird and Foulds, 1968; Knapton, 1977). DOI 10.1108/IJCST-06-2016-0065
 IJCST Researchers expanded the above theories about the length calculation of the plain jersey
29,5 fabrics to the fields of double jersey fabrics and rib knitted structure (Knapton et al., 1968;
Knapton, 1969; Knapton and Fong, 1970, 1971; Smirfitt, 1965).
From the above, the corresponding techniques for prediction of the weight and the
material consumption can be summarized as follows. For the direct measurement of
the weight of the sweater, the measurement be carried out before the sweater is knitted.
662 For the “proKNIT” software system and corresponding theories, it is investigated on the
length calculation of the stitch of plain knitted and rib structures. For a sweater, it is
composed of many other structures besides the above two. Very little study has been carried
out for predicting the stitch length of other structures. In addition, the length of loop based
on geometrical method depends on many parameters in comparison to the knitting machine,
and the measurement of these parameters is susceptible to other factors such as force.
In contrast, the parameters of the needle bar flat knitting machine are relatively reliable.
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

According to the above discussion, this paper focuses on the material consumption
prediction of the sweater expressed by its weight. To do this, the sweater is divided into five
basic structures, namely plain stitch, tuck stitch, loop transfer, float stitch and rib stitch.
The basic structures are modeled with the parameters of the needle bar flat knitting
machine combined with their geometrical structures, considering that the parameters of the
knitting machine are more reliable than those of the basic structures. With the expression of
the length of basic structures and the linear density of the yarn, the predictive weight of the
sweater is derived. Three types of sweaters on four different knitting machine are knitted as
examples to verify the performance of the proposed methods in this paper.
The experimental results show that the proposed method can achieve good performance
with the error by percent ranging from −1.54 to 2.84 percent.

Structure of flat-knitted sweater

In general, knitted fabrics are classified into two different types: single-side and double-side
fabrics. Single-side fabrics are knitted by single needle bar flat knitting machines and are
used for the body and sleeve. Double-side fabrics are made on double needle bar flat knitting
machines and are used for lower hem, cuff and neckline. The knitting process of a single and
double knitting machine is illustrated in Figure 1.
Figure 1(a) demonstrates the front view of the single knitting machine, (b) shows the
left-hand side view of yarn on the latch, and (c) depicts the structure of the double knitting
machine. For simplifying the following prediction method, some assumptions and symbols
are given as follows. The section of the hook of latch needle is assumed to be cylindrical with
the diameter denoted as “b.” The symbol “r” represents the radius of the bearded needle.
The section of the yarn is treated as a circular one with diameter “d.” The outer of sinker
loop is regarded as a semi-circle due to the effect of the sinker and the thickness of the sinker
is written as “a.” The distance of knitting bed is written with “h.” The symbols “t” and “x”
represent the needle pitch and sinking depth.

Models of the basic structures

Although there are various types of fabrics, they can be regarded as the combinations of
five basic structures, namely plain stitch, tuck stitch, loop transfer, float stitch and rib stitch.
Predicting the material consumption of a sweater is equivalent to predicting the material
consumption of the basic structures within the sweater. To do this, mathematical models of
the basic loops length are built related to the parameters of the needle bar on the knitting
machine. For convenience, the symbol l1 represents the length of the plain stitch.
The lengths of tuck stitch, loop transfer, float stitch and rib stitch are represented with the
symbols l2, l3, l4 and l5.
 (a) (b) Prediction
C
method of
C x material
a d
x consumption
b B
d b
r
B
r 663

N1 t N2

(c)
N2
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

x
b

N1 t Figure 1.
Structure of the
Notes: (a) Front view of single knitting machine; (b) left side view of yarn on the latch; knitting machine
(c) double knitting machine

Plain stitch
Plain stitch is a typical structure of a sweater composed of needle loop, sinker loop and loop
arm. Figure 2 illustrates the mathematical model of the plain stitch derived from the real
loop on the knitting machine. Some assumptions are given for the components of the plain
stitch as follows: the needle loop hooked by the latch needles, denoted as “N1” and “N2” in
Figure 2(a), is assumed to be a semi-circle and the section of the latch needle is assumed to be
cylindrical; loop arm, denoted as “BC,” can be treated as the small slope in Figure 2(b), which
can be computed with Pythagorean theorem; and the sinker loop can be seen as a circular
segment with the diameter referring to Figure 2(b) and the symbol “r” represents the radius
of the bearded needle. The expression of the diameter can be determined with t-d-b, where
the symbols “t,” “d” and “b” represent the needle pitch, the diameter of the section of the yarn
and the diameter of the section of the latch needle, respectively, as mentioned above.
Based on the above assumptions, one can obtain the length of each segment of the plain
stitch according to the following equations:
p
l AB ¼ ðbþd Þ (1)
4
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi b
l BC ¼ ðx þrbÞ þr 2 r þ (2)
2
p
l CD ¼ ðtdbÞ (3)
4
 IJCST (a) (b)
29,5

D
C

664 t-d-b x
d
b
B r

A
N1 N2
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

N1 N2
t/2
Figure 2.
Plain stitch
Notes: (a) Structure of plain stitch; (b) model of plain stitch

Subsequently, the length of the plain stitch, l1, can be determined with the following equation:

l 1 ¼ 2  ðl AB þl BC þl CD Þ (4)

Tuck stitch
Tuck stitch is comprised of three segments, namely needle loop, sinker loop and loop arm.
As the knitting needles ascend to the height of tuck stitch, the new yarn does not form the
new loop while it enters into the hook with the old loop and presents the arc shape.
The structure of tuck stitch is shown in Figure 3(a) and the model is depicted in
Figure 3(b). Figure 3(a) displays the on-line procedure of tuck stitch. During the knitting
process, the sinker acts on the sinker loop immediately. Through the effect between the
loop stitch formed by latch needle “N1” and the old needle formed by latch needle “N2,” the
tuck stitch at “N2” is drawn out and separate itself from the sinker after the whole knitting
procedure is finished.

(a) (b)

D
C
a

x
d
b
N1 N2 r B

N1 N2
Figure 3.
Tuck stitch t/2
Notes: (a) Sructure of tuck stitch; (b) model of tuck stitch
 From the figure, it can be observed that the sink loop of the tuck stitch is attached with the Prediction
sinker. The model can be analyzed under some assumptions as follows: similar to the method of
structure of the plain stitch, the needle loop is assumed to be cylindrical; the length of the leg material
can be calculated using the Pythagorean theorem; and the section of the sink loop is
assumed to be cylindrical with the diameter “a,” which is equal to the thickness of the sinker. consumption
Therefore, the segments of the tuck stitch can be determined with the following equations:
p 665
l AB ¼ ðbþd Þ (5)
4
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
l BC ¼ x2 þ ðtabdÞ2 (6)
4
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

p
l CD ¼ ða þdÞ (7)
4
In the following, the length of the tuck stitch, l2, can be obtained with the segments
according to the following equation:
l 2 ¼ 2  ðl AB þl BC þl CD Þ (8)

Loop transfer
The loop transfer is derived from the plain stitch by transferring the needle loop and the
sinker loop from one needle to another, which means that the length expression of the loop
transfer is identical to that of the plain stitch, although some loop deformation is made from
the plain stitch. For example, the structure of the loop transfer is illustrated in Figure 4,
where it is obvious that the loop is transferred from the latch needle “N1” to the latch needle
“N2” directly. Therefore, the expression of the loop transfer, l3, has the similar form to that of
the plain stitch in the following equation:
l 3 ¼ l 1 ¼ 2  ðl AB þl BC þl CD Þ (9)

N1 N2
Figure 4.
Structure of loop
transfer
 IJCST Float stitch
29,5 As the knitting needle which forms the float stitch does not hook the new yarn, the new yarn just
passes over a knitting needle and forms a piece of thread which connects with the adjacent loops.
Hence, float stitch can be treated as a line segment with the distance same as of the gauge
distance. Figure 5(a) demonstrates the physical structure of the float stitch under the
microscope and Figure 5(b) depicts the mathematical model for the float stitch.
666 The length of the float stitch, l4, can be determined with the following equation, where the
symbol “t” represents the distance between two knitting needles as mentioned above:
l4 ¼ t (10)

Rib stitch
1 + 1 rib stitch is a typical double-side rib stitch, consisting of a front-side rib stitch and a
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

back-side rib stitch. In general, the front-side rib stitch and the back-side rib stitch do not lie on
the same plane. As a result, the sinker loop is distorted connecting the front-side rib stitch to
the back-side rib stitch. Figure 6(a) shows the physical structure of the double-side rib stitch.
For convenience, the mathematical model in Figure 6(b) is obtained with the projection of rib
stitch on the same plane under the following assumptions: the line segments “AB” and “CD”
are attached on the inner side of the needle and can be considered as arc segments, in specific,
quarter circle; the yarn “BC” can be divided into three different segments due to the effect of
the front needle bar and the back needle bar. When the three segments are projected on the
same plane, they are collinear and can be obtained with using a Pythagorean theorem.
The segments of the rib stitch and the length of rib stitch, l5, can be determined with the
following equations:
p
l AB ¼ l CD ¼ ðb þdÞ (11)
4
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2
t
l BC ¼  bd þ ð2x þhÞ2 (12)
2

l 5 ¼ l ABCD ¼ l AB þl BC þl CD (13)

(a) (b)

A B

N1 N2 N3
N1 N2 N3
t
Figure 5.
Float stitch
Notes: (a) Structure of float stitch; (b) model of float stitch
 (a) (b) Prediction
N2 method of
A
material
N2 b
B consumption
x
667

h
d

N1 x
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

b
C

D
t/2
N1 Figure 6.
Rib stitch
Notes: (a) Structure of rib stitch; (b) model of rib stitch

Prediction of the material consumption

To predict the material consumption of the sweater, a matrix P is defined with a form in the
following equation:
2 3
p11 p12    p1n    p1N 0
6p 7
6 21 p22    p2n    p2N 0 7
6 7
6 ^ ^ ^ ^ 7
6 7
P M N ¼ 6 7 (14)
6 pm1 pm2    p1n    pmN 0 7
6 7
6 ^ ^ ^ ^ 7
4 5
pM 1 pM 2    pM n    pM N 0

where pmn is the number of stitches of structure n in the mth course, n ¼ 1, …, N0 ( ¼ no. of
basic structures ¼ 5 in the text), m ¼ 1, …. M ( ¼ no. of courses):
 T
R ¼ r1 ; r2 ; . . .rm ; . . .rM
where ρm is the linear density of the yarn in the mth course and [·]T represents the transpose
operation
 of a vector: 
L ¼ l 1 ; l 2 ; . . .; l n ; . . .l N 0 where ln is the length of nth basic structure.
Then, the weight is:
M X
X N0
G¼ pmn rm l t (15)
m¼1 n¼1
and in a matrix form is:
G ¼ RT PLT (16)
The proposed method for predicting the material consumption can be used for estimating
the material consumption of the sweater before it is knitted. Therefore, the manufacturers
 IJCST can control the cost of the sweater and improve the efficiency. In this condition, the
29,5 designer can integrate the proposed method into design system of a sweater. The process
can be summarized as follows: one can design the sweater according to the garment style
and pattern with the help of computer-aided design (CAD) system; with the module of
CAD system, a number of courses can be obtained; the lengths of the five basic structures
and yarn consumption can be determined according to machine parameters and yarns
668 parameters; and the gross cost is calculated with typing in unit price of yarns. The flow
chart of weight calculation of the sweater is illustrated in Figure 7, where m and M
represent the index of course and maximum course of the sweater, respectively.
Unlike the conventional methods, this method can derive the basic structures with the
parameters of the needle bar flat knitting machine and predict the weight and the price of
the sweater before it is knitted.
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

Experimental results and analysis

In this section, three types of sweaters shown in Figure 8 are chosen to verify the
performance of the proposed prediction models. The sweaters are composed of the five basic
structures as mentioned above. In this experiment, the 24S × 3 acrylic is used to knit three
types of sweaters on the SHIMA SEIKI computer flat knitting machine SSG 122 E7, E12 and
28S × 3 wool/acrylic yarn is used to knit three types of sweaters on the general flat knitting
E6, E9 machine as shown in Figure 8.

Start

Initialization
m=1

Properties of
the mth row

Weight of Weight of Weight of Weight of Weight of

plain tuck stitch loop transfer float stitch rib stitch

Weight of
the mth row

Accumulative
weight of the
sweater

m=m+1

Figure 7.
Flow chart of m > M? End
weight prediction
 Prediction
method of
material
consumption

669
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

Figure 8.
Three types of
sweater for experience
 IJCST The corresponding parameters of the knitting machine are listed in Table I, and Table II
29,5 shows the number of the basic structures of the desired sweaters.
Substituting the parameters listed in Tables I and II into Equations (1)-(16), one can
obtain the predictive weight of the sweater in Table III, where the actual value and the
relative errors expressed in percentage are also given as a comparison.
From Table III, it can be observed that the proposed model can achieve the relative error
670 in the range from −1.54 to 2.84 percent and has a better performance. The errors occur due
to the following aspects: the mathematical models of the basic structures are slightly
different from the actual structures; the error lies in the measurement of the related
parameters of the knitting machine.
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

Thickness Diameter Radius of Sinking Distance of

Machine of yarn of needle needle Distance of Thickness of depth knitting bed
gauge (d/mm) (b/mm) (r/mm) needle (t/mm) sinker (a/mm) (x/mm) (h/mm)

E6 0.25 0.65 1.65 4.23 0.95 3.47 4.50

Table I. E7 0.29 0.60 1.44 3.629 0.77 3.47 4.50
Parameters of the E9 0.25 0.50 1.15 2.822 0.65 2.87 4.20
knitting machine E12 0.29 0.45 1.04 2.117 0.35 3.47 4.15

No. Plain stitch Tuck stitch Loop transfer Float stitch Rib stitch

Table II. 1 100,120 407 295 408 10,080

Number of loops of 2 99,691 314 170 284 8,934
the sweaters 3 105,204 412 590 399 30,360

Machine gauge Actual value Predictive value Error (%)

E6
Sweater 1 358.92 364.44 1.54
Sweater 2 357.42 355.58 −0.51
Sweater 3 400.89 397.20 −0.92
E7
Sweater 1 341.86 348.37 1.90
Sweater 2 340.38 342.54 0.63
Sweater 3 394.89 395.69 0.20
E9
Sweater 1 326.37 329.99 1.11
Sweater 2 325.87 321.96 −1.20
Sweater 3 340.88 342.48 0.47
E12
Sweater 1 278.35 286.24 2.84
Table III. Sweater 2 285.32 280.92 −1.54
Experimental results Sweater 3 300.30 302.16 0.62
 Conclusion Prediction
This paper developed a prediction model for material consumption prediction and method of
was proven by comparing the predictable result and the final weight of the sweater. material
Aiming to alleviate the drawback that very little study has been carried out in the
predicting length of the other structures except for plain stitch and rib stitch, this consumption
paper divided the basic structures of the sweater into five types, namely plain stitch,
tuck stitch, loop transfer, float stitch and rib stitch. The lengths of the basic structures are 671
expressed with the parameters related to the knitting machine. In the following, the
weight of the sweater is expressible with the number of the basic structures of the sweater
and the line density in a matrix form. To verify the proposed model, three types of
sweaters knitted on the different knitting machines were used as examples for evaluation
of the prediction model. In addition, the proposed method can predict the weight and the
material consumption of the sweater without the requirement that the sweater is knitted.
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

The experimental results show that the proposed model can achieve good performance
with the error by percentage ranging from −1.54 to 2.84 percent. A flow chart is also
provided. Moreover, the proposed method for predicting the material consumption of the
sweater can be integrated into a CAD system.

References
Baird, K. and Foulds, R.A. (1968), “Felting shrinkage of plain-knitted wool fabrics its dependence
upon fabric structure and shrinkproofing level”, Textile Research Journal, Vol. 38 No. 7,
pp. 743-753.
Gravas, E., Kiekens, P. and Langenthove, L. (2005), “An approach to the ‘proKNIT’ system and its
value in the production of weft-knitted fabrics”, AUTEX Research Journal, Vol. 5 No. 4,
pp. 219-226.
Gravas, E., Kiekens, P. and Langenhove, L. (2006), “Predicting fabric weight per unit area of single
and double-knitted structures using appropriate software”, AUTEX Research Journal, Vol. 6
No. 4, pp. 223-237.
Knapton, J.J.F. (1969), “Geometry of complex knitted structures”, Textile Research Journal, Vol. 39 No. 9,
pp. 889-892.
Knapton, J.J.F. (1977), “How to knit spun yarns efficiently”, Knitting Times Year Book, Vol. 46 No. 18,
pp. 111-114.
Knapton, J.J.F. and Fong, W. (1970), “The dimensional properties of knitted wool fabrics part iv: 1×1 rib
and half-cardigan structures in machine-washing and tumble-drying”, Textile Research Journal,
Vol. 40 No. 12, pp. 1095-1106.
Knapton, J.J.F. and Fong, W. (1971), “The dimensional properties of knitted wool fabrics part
v-:interlock and Swiss double-piqué structures fully-relaxed and in machine-washing and
tumble-drying”, Textile Research Journal, Vol. 41 No. 2, pp. 158-166.
Knapton, J.J.F., Truter, E.V. and Aziz, A. (1975), “The geometry, dimensional properties, and
stabilization of the cotton plain-jersey structure”, Journal of the Textile Institute, Vol. 66 No. 12,
pp. 413-419.
Knapton, J.J.F., Ahrens, F.J., Ingenthron, W.W. and Fong, W. (1968), “The dimensional properties of
knitted wool fabrics part i: the plain-knitted structure”, Textile Research Journal, Vol. 38 No. 10,
pp. 999-1012.
Munden, D.L. (1959), “The geometry and dimensional properties of plain-knit fabrics”, Journal of the
Textile Institute Transactions, Vol. 50 No. 7, pp. 448-471.
Peirce, F.T. (1947), “Geometrical principles applicable to the design of functional fabrics”, Textile
Research Journal, Vol. 17 No. 3, pp. 123-147.
 IJCST Postle, R. (1968), “Dimensional stability of plain-knitted fabrics”, Journal of the Textile Institute, Vol. 59
29,5 No. 2, pp. 65-77.
Smirfitt, J.A. (1965), “Worsted 1×1 rib fabrics part II: some physical properties”, Journal of the Textile
Institute Transactions, Vol. 56 No. 6, pp. T298-T313.

Further reading
672 Kurbak, A. and Alpyildiz, T. (2008), “A geometrical model for the double Lacoste knits”, Textile
Research Journal, Vol. 78 No. 3, pp. 232-247.
Wilfried, R. and Yordan, Y. (2010), “Geometry modelling of warp knitted fabrics with 3D form”, Textile
Research Journal, Vol. 81 No. 4, pp. 437-443.

Corresponding author
Downloaded by Florida Atlantic University At 04:23 12 September 2017 (PT)

Xuan Luo can be contacted at: 653011306@qq.com

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com