PREFACE

In pursuance of the press restructuring approved by the Prime Minister and the
proposal made by the Ministry of Education and Training, the Ministry of Culture and
Information issued the Decision No 519/GP-BVHTT dated November 25th, 2002,
licensing the publication of the University of Danang journal entitled Science and
Technology.
The Press Bureau under the Ministry of Culture and Information issued the official
dispatch No 816/BC dated August 10th, 2006, permitting the University of Danang to
publish the Science and Technology every two months instead of every three months.
The National Scientific and Technological Information Centre under the Ministry
of Science and Technology issued the official dispatch No 44/TTKHCN-ISSN dated
February 6th, 2007, agreeing to grant an international standardized code, namely ISSN
1859-1531 to the UD journal Science and Technology.
The Press Bureau under the Ministry of Information and Communications issued
the official dispatch No 210/CBC of March 5th, 2008, allowing the UD journal Science
and Technology to be published in English and French in addition to the Vietnamese
publication.
On September 15th, 2011, the Ministry of Information and Communications issued
the Decision No 1487 /GP-BVHTT, granting an additional licence which permits the UD
journal Science and Technology to increase the publication period frequency from
every 2 months to every month and to raise the number of pages from 80 to 150 pages.
The UD journal Science and Technology has been established in view of
announcing and introducing scientific researches in the fields of teaching and training,
offering information on the domestic and foreign scientific research results that serve the
education in colleges and universities and propagating the State and Partys guidelines
and policies in education and training as well as in scientific and technological researches.
Over the past 40 years, the UD journal Science and Technology has been
published as the continuity and development of the scientific information journals and
conference proceedings published at the University of Danang and its member
universities.
The journals editorial staff wishes to receive many-sided contributions of the
scientists and educators from within the university and from other colleges and
universities inside and outside the country so that the journals quality will be more and
more improved in the future.
Journals Editorial Staff

CONTENTS
ISSN 1859-1531 THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL 1

ENGINEERING AND TECHNOLOGY

Studying the method for sensitivity analysis of ozone formation in urban and rural areas using
CMAQ
Nguyen Phuoc Quy An

Study of designing and manufacturing simili film-coating machine

Tao Quang Bang, Hoang Van Thanh

Heavy metals accumulation of food-crops grown in agricultural areas affected by industrial

wastewater, Da Nang, Vietnam
Doan Chi Cuong, Masaki Takaoka, Vo Van Minh, Phan Nhat Truong

Research on magnetomechanical coupling relation in amorphous metal core transformers

Do Chi Phi, Le Van Doanh, Phung Anh Tuan

14

Improving stability for independent power control of DFIG with SFOC and DPC during GRID
unbalance
Nguyen Thanh Hai, Vo Viet Cuong

20

Effects of ethylene glycol, temperature and pH on the size and the dispersion of the platinum
nanoparticle catalyst on carbon support
Hoang Anh Huy, Ho Thi Thanh Van

25

A study of the curvature of a thick ALN film grown on a trench-patterned -Al2O 3 template using
X-ray diffraction
Dinh Thanh Khan, Nguyen Quy Tuan

30

The application of fuzzy entropy to selecting features of partial discharge in high voltage
cable joints
Nguyen Tung Lam

33

A study on improving properties of aligned multi-walled carbon nanotube/ epoxy composites

Tran Huu Nam, Vu Minh Hung, Vo Quoc Thang

38

Physical layer secrecy performance analysis of TAS/ MRC system over rayleigh/ nakagami
fading channels
Nguyen Van Tho, Van Phu Tuan, Vo Tan Loc, Ha Dac Binh

43

Study of student engagement levels in class activities

Hoang Van Thanh, Luu Duc Binh

48

Optimizing P.I.D parameters in control accelerometers and gyroscopes in self - balancing

quadrotors
Vu Van Thanh, Huynh Thanh Tung

53

A method for fruits recognition using image processing techniques

Pham Xuan Thuy

58

Research on reducing residue of fipronil pesticide by vetiver grass in constructed wetland a case study in Quang Nam province, Vietnam
Pham Phu Song Toan

63

Design of silicon wires based directional couplers for microring resonators

Le Trung Thanh, Nguyen Canh Minh, Nguyen Van Khoi, Bui Thi Thuy, Nguyen Thi Hong Loan

67

Human - computer interaction based on IMU sensors

Huynh Thanh Tung, Vu Van Thanh

71

NATURAL SCIENCES
Applying model-view-controller model (MVC model) in building a dormitory management
information system for The University of Danang, University of Science and Technology
Dang Hoai Phuong, Nguyen Van Dong, Dinh Thi Hong Huyen, Pham Thi Thu Hien, Lam Thi Hong Tam

75

Reactive multi-agent model for collaborative filtering - based recommender systems

Tran Thi Ngoc Trang, Le Viet Man, Nguyen Minh Duc

80

SOCIAL SCIENCES
FDI attraction for industry development experiences from developed countries and
applications for Laos
Nguyen Ngoc Anh, Lamngeun Xayasene

85

A review of societal marketing

Nguyen Thi Xuan Trang

90

HUMAN SCIENCES
Applying group work activities to improve English speaking skills for nursing students at Danang
University of medical technology and pharmacy
Pham Dang Tram Anh, Luu Quy Khuong

94

Positions and strategies of self-repair in English and Vietnamese conversations

Nguyen Th Minh Hanh, Nguyen Thi Quynh Hoa

100

Applying grammatical metaphor to academic writing

Phan Van Hoa, Nguyen Van Vui

105

Linguistic features of assertives in presidents welcome messages on official websites of

colleges and universities in the USA
Nguyen Thi Quynh Hoa, Doan Thi Hong Nguyen

109

An investigation into semantic features of hyperbole used in football commentaries in English

and Vietnamese newspapers
Luu Quy Khuong, Cao Nguyen Minh Thong

115

An investigation into semantic feature of English slang containing words denoting animals
Dinh Quang Trung

121

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

STUDYING THE METHOD FOR SENSITIVITY ANALYSIS

OF OZONE FORMATION IN URBAN AND RURAL AREAS USING CMAQ
Nguyen Phuoc Quy An
The University of Danang, University of Science and Technology; ngphquyan@gmail.com
Abstract - The m ajority of ozone form ation occurs when NOx, CO
and VOC react in the atmosphere in the presence of sunlight.
However, the ratio of VOC and NOx largely influences the
formation of ozone. Therefore, the Comm unity Multiscale Air
Quality (CMAQ) modeling system is used in a sensitivity analysis
of ozone with nine different em ission scenarios by reducing VOC
and NOx emissions. The capital metropolis of Seoul and the island
of Gang-hwa, considered as typical urban and rural sites
respectively, are chosen for the scope of this study. From the
results of the sensitivity analysis of ozone formation in urban and
rural areas, it is considered that ozone concentration in urban and
rural appear in VOC limited area of EKMA (Empirical Kinetic
Modeling Approach).
Key words - ozone; analysis; CMAQ; sensitivity; urban; rural

1. Introduction

variation in the output of a statistical model can be

attributed to different variations in the inputs of the model.
In this study, sensitivity of ozone formation in the urban
and the rural of Korea will be analyzed. The capital
metropolis of Seoul and the island of Gang-hwa,
considered as the typical urban and rural, respectively, are
chosen for the scope of this study. In the sensitivity study,
the peak O3 concentration for each scenario will be
compared with the base-case. Special emphasis will be
focused on the impact of VOC and NOx emission sources.
Besides, sensitivity analysis of ozone formation indicates
that reducing VOC or NOx emission affect the greatest
reduction or increase in peak of ozone concentration in the
urban and in the rural.

The levels of air pollutants are increasing rapidly in

many mega cities of the developing world. Urban air
pollution has increased rapidly with urban populations,
numbers of motor vehicles, and fuel with poor
environmental performance, badly maintained roads and
ineffective environmental regulations. Ozone is one of the
air pollutant emissions which are the predominant factors
affecting air quality. Ozone is the most severe air pollution
problems in the world. It has serious impacts on human
health and ecosystems, and is very difficult to control. In
particular, the ground level ozone is responsible for a
variety of adverse effects on both human being and plant
life. To protect the humankind from such adverse health
effects, early information and precautions of high ozone
level need to be supplied in times.

The Community Multiscale Air Quality (CMAQ)

modeling system is used in a sensitivity analysis of ozone
with 9 different emission scenarios by reducing VOC
andNOx emissions using the same meteorological input
and chemical transport schemes. The meteorological field
is the Mesoscale Model, Version 5 (MM5) and the
emission inventory model is Sparse Matrix Operater
Kernel Emissions (SMOKE) Modeling System. In
addition, this study will analyse the effect of VOC and
NOx on the sensitivity of ozone formation in the urban and
in the rural to have projects which can control strategies for
VOC and NOx emissions to reduce ozone concentration in
Seoul and Gang-hwa.

Tropospheric ozone is a trace gas which plays a key role

in the oxidizing capacity of the atmosphere. Ozone also
exerts a significant influence on the radiation budget of the
atmosphere owing to its properties as a greenhouse gas.
Major ozone sources and sinks in the troposphere are the air
mass exchange between the stratosphere and troposphere, in
photochemical production or destruction and surface dry
deposition. Taking into account that ozone precursors are
also anthropogenically emitted, tropospheric background
ozone levels have been modified during the last century [3].
Moreover, it can be swept away by prevailing winds, thus
leading to higher ozone concentrations in places far from the
sources of emission of the ozone precursors. Thus, the
concentration of ozone in different areas is not similar,
especially in the urban and in the rural.

2.1. Ozone

Ozone is a secondary pollutant formed through the

oxidation of volatile organic compounds (VOC) in the
presence of nitrogen oxides (NOx) and sunlight followed
by the combination of molecular oxygen (O2) and triplet
oxygen radical (O3P) [2]. Thus, the sensitivity analysis of
ozone will be performed by reducing VOC and NOx
emissions. Sensitivity analysis is the study of how the

2. Research Overview
2.1.1. Sources of ozone precursors
A problem of increasing concern is the presence of
photochemical smog in some urban and industrial regions.
The photochemical reaction of NOx (NO + NO2) and VOC
in the presence of sunlight originate in photochemical
smog. It is chemically characterized by a high level of
oxidant compounds, mainly O3.
NOx and hydrocarbon emissions from traffic are high
in urban areas so ozone tends to accumulate rapidly. A
considerable effect on the oxidizing capacity of the
troposphere which affects human health by causing
symptoms such as irritated eyes, cough, headache, chest
pains and, in extreme cases, lung inflammation coming
from the concentration of ozone. The ozone is also
associated with the corrosion of urban structures, the toxic
plants and leading to a decrease in vegetation. Moreover,
ozone can be swept away by winds so the higher ozone
concentration appears in places far from the sources of the
emission of the ozone precursors. Many regions worldwide
have been plagued by the air pollution of high surface

ozone arising from photochemical formation and

accumulation. The ozone is photochemically produced and
can accumulate to hazardous level in favorable weather
conditions, in the presence of volatile organic compounds
(VOC) and nitrogen dioxides (NOx).
2.1.2. Ozone control strategies
It is difficult to apply an optimized control strategy for
ozone, since complex chemical mechanisms are involved
in ozone generation. The ozone isopleth plot has been used
as a basis for applying control strategies historically. The
relationships between maximum ozone concentrations and
mixtures of NOx and VOC are shown in the isopleths plot.
The result of VOC and NOx mixtures being irradiated in
photochemical chambers is the isopleths plot. The O3 NOx VOC can be illustrated by isopleths plots generated
from applying a basic ozone model called the Empirical
Kinetic Modeling Approach (EKMA) to VOC and
NOxconcentrations [4]. The peak ozone as a function of
the ratio of VOC to NOx concentrations is shown in this
graph. There are two regimes with different O3 - NOx
VOC sensitivities, they are referred to as limited in the
graph, they are VOC limited and NOx limited.

Nguyen Phuoc Quy An

on the ozone air quality for low VOC/NOx ratio condition.

It is not realistic however, to use the ozone isopleths as
a basis for control strategies without detailed investigations
of VOC and NOx levels within a region. In the real
atmosphere, deposition process, existence of particulate
matter, turbulence and variations in radiation are believed
to be the primary causes of deviations from chamber
studies. Another difficulty in applying the ozone isopleths
method, is that the VOC/NOx ratio at a monitoring site
may not represent ratio in a region. The other approach in
determining an optimal control strategy for ozone is to use
air quality models. Air quality models have the capability
to include the emission and meteorological characteristics
of a region, therefore, they could be better tools to provide
bases for optimal ozone control strategies.
2.2. Model descriptions
CMAQ modeling system is the air quality modeling
system used in this study. The primary modeling
components in the CMAQ modeling system include:
Mesoscale Model Version 5 (MM5) is a meteorological
modeling system for the description of atmospheric states
and motions, Sparse Matrix Operating Kernel for
Emissions (SMOKE) models for processing man-made
and natural emissions that are injected into the atmosphere,
and the chemical transport model used in this study is the
Community Multiscale Air Quality Model (CMAQ).
2.2.1. Mesoscale Model Version 5 (MM5)
The Mesoscale Prediction Group in the Mesoscale and
Microscale Meteorology Division, National Center for
Atmospheric Research (NCAR) provide and support for
MM5 (Mesoscale Model Version 5) modeling system
software. MM5 was developed in cooperation with The
Pennsylvania State University (Penn State) and the
University Corporation for Atmospheric Research (UCAR).

Figure 1. A typical EKMA 2-dimensional depiction of ozone

isopleths generated from initial mixtures of VOC and NOx in air [4]

A constant VOC/NOx ratio = 8/1 is represented by the

straight line in the center of Figure 1, the ozone isopleths
is bisected by this line. Transition from the fairly vertical
lines in the left side of EKMA graph where ozone changes
are fairly sensitive to changes in VOC limited to the mostly
horizontal on the graphs right where ozone changes are
quite responsive to NOx limited.
The VOC limited (VOC sensitive) represents an urban
area with low VOC/NOx ratios. In urban areas, NOx
emission has much greater influence and there is relatively
little biogenic VOC to offset the NOx. In this area, when
reducing VOC, ozone concentrations are most efficiently
lowered. On the contrary, the NOx limited (NOx sensitive)
is typical of less urbanized, more rural air massed where
biogenic VOC are much bigger contribution to VOC
levels. In NOx limited area, when reducing NOx the ozone
concentrations are lower than moving downward to lower
ozone isopleths.
On the basis of these isopleths, the EKMA plot shows
that VOC only control strategies could reduce ozone
concentrations more effectively in low VOC/NOx ratio
areas. Any reduction of NOx initially have an adverse effect

2.2.2. Sparse Matrix Operating Kernel for Emissions (SMOKE)

The Sparse Matrix Operator Kernel Emissions
(SMOKE) Modeling System was created by the MCNC
Environmental Modeling Center (EMC) to allow
emissions data processing methods to integrate highperformance-computing (HPC) sparse-matrix algorithms.
An effective tool for emissions processing in a number of
regional air quality modeling applications is the SMOKE
prototype available since 1996. The support of the U.S
Environmental Protection Agency (EPA) redesigned and
improved SMOKE in 1998 and 1999 for use with EPAs
Models-3 Air Quality Modeling System.
A lot of criteria gaseous pollutants such as carbon
monoxide (CO), nitrogen oxides (NOx), volatile organic
compounds (VOC), ammonia (NH3), sulfur dioxide (SO2),
particulate matter (PM) pollutants such as PM 2.5 microns
or less (PM2.5) and PM less than 10 microns (PM10), as
well as a large array of toxic pollutants, such as mercury,
cadmium, benzene and formaldehyde can be processed by
SMOKE. SMOKE can process no limitation regarding the
number or types of pollutants.
The resolution of the emission inventory data is
converted to the resolution needed by an air quality model
is the purpose of SMOKE.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

2.2.3. Community Multiscale Air Quality (CMAQ)

A third-generation air quality model is the EPA
Community Multiscale Air Quality (CMAQ) modeling
system. CMAQ requires two primary types of inputs:
meteorological information and emission rates from
sources of emissions that affect air quality. The
meteorological model generates gridded meteorology for
input to both CMAQ and the emissions model. The
emission model is required to convert annual, county-level
emissions estimates to gridded hourly emissions formatted
for CMAQ.
The five main CMAQ programs are:
- The meteorology-chemistry interface processor
(MCIP): MCIP is used to preprocess the data from a
meteorological model for CMAQ and SMOKE.
- The initial conditions processor (ICON): a binary
net CDF initial conditions file is created by ICON for input
to CCTM.
- The boundary conditions processor (BCON): a
binary net CDF lateral boundary conditions file is created
by BCON for input to CCTM.
- The clear-sky photolysis rate calculator (JPROC):
Physical information about photoreactive molecules into
clear-sky photolysis rate look-up tables is converted by
JPROC for input to CCTM.
- The CMAQ chemistry-transport model (CCTM):
CCTM run last in the sequence of programs. All of the
other CMAQ programs the emission and meteorological
models are used to prepare the inputs to CCTM. CCTM can
produce estimates of pollutant concentrations, wet and dry
deposition rates and visibility metrics at a time granularity.
Emissions Model
(SMOKE)

Meteorology Model
(MM5 or WRF)

MCIP

CMAQ Programs

ICON
CCTM
BCON
JPROC

Figure 2. The CMAQ modeling system [1]

3. Research Methodology
3.1. Modeling conditions
3.1.1. Study period and domain
There was a rapid ozone formation event on August 23,
2007 and the ozone concentration in Gang-hwa is higher
than Seoul in this day. Therefore, this study selects the
simulation period from August 19 to 25, 2007.
There are 3 model domains in this study: domain 1
includes East Asia, domain 2 includes South Korea,
domain 3 includes Seoul & Gang-hwa.
3.1.2. Meteorological fields
In this study, the MM5 (Mesoscale Model, Version 5)

is used to produce and provide meteorological fields for

CMAQ (wind, temperature, water mixing ratio,
precipitation, surface variables and others). The domain 1
for nesting process is 102102 grid numbers in plane with
27 km grid resolution for East Asia, the domain 2 of MM5
includes 6161 grid numbers in plane with 9 km grid
resolution for South Korea, and there are 5249 grid
numbers in plane with 3 km grid resolution in domain 3 for
Seoul and Gang-hwa.
3.1.3. Emission inventory data
The emission inventory data use for domain 1 is from
INTEX-B emission inventory derived from CGRER
(Center for Global and Regional Environmental Research).
SO 2, CO, NOx, PM10 and VOC emissions are based on
INTEX-B emission inventory. Gridded data from 0.50.5
INTEX-B gridded emissions datais converted into 11
gridded emissions data. The emission data of domain 2
(9km x 9km) and 3 (3km x 3km) are from CAPSS, 2007
(Clean Air Quality Policy Support System in Korea). The
spatial resolution of CAPSS data is 1 km x 1km. CAPSS
data includes point, mobile and area emission sources from
the plants and fugitive dust. VOC emission input includes
anthropogenic emission. The emission inventory data are
sorted according to source classification codes (SCC) for
each pollutant and county. They are converted into the IDA
(InventoryData Analyzer) format.
3.1.4. Air quality model
The chemical transport model used in this study is
CMAQ (Community Multiscale Air Quality Model). The
emission inventory data is converted by SMOKE modeling
system into hourly emission data for CMAQ modeling.
The gridded emission inventory is generated by SMOKE
and meteorological fields are generated from MCIP, the
CMAQ Chemical Transport Model (CCTM) calculates the
chemical reactions, transport and atmospheric deposition
of all participating species according to specified physical
and chemical options.
3.2. Evaluation of model performance
The CMAQ performance is evaluated by comparing the
observation data and the simulated results. Observed
Ozone concentration used in this study is from the National
Institute of Environmental Research in Korea 2007. The
CMAQ output data in net CDF format was statistically
analyzed. The statistical are calculates for 10 sites over the
simulation period (August 19-25, 2007) and each hour over
total sites.The statistical treatments of the data are shown
in Table 1.
Mean bias (MB) can indicate whether the simulations
under or over estimate the concentration at each hour of
each site. As a mean normalized bias (MNB), this
performance statistic averages the model/observation
residual, paired in time, normalized by observation, over
all monitor times/locations, a value of zero would indicate
that the model over predictions and model under
predictions exactly cancel each other out. As a mean
normalized gross error (MNGE), this performance statistic
averages the absolute value of the model/observation
residual, paired in time, normalized by observation, over

Nguyen Phuoc Quy An

all monitor times/locations, a value of zero would indicate

that the model exactly matches the observed values at all
points in space/time. The mean fractional bias (MFB)
normalizes the bias for each model-observed pair by the
average of the model and observation before taking the
average. Correlation coefficient (R) between modeling and
observation concentrations can verify the ability of the
model in predicting the variations of observed
concentrations.
Table 1. The statistical treatment methods for the comparison of data
Mean bias (MB):
,

Mean normalized bias (MNB):

,

,
,

Mean fractional bias (MFB):

,

0.5

, |

Mean normalized gross error (MNGE):

|

,
,

Correlation coefficient (R):

,
,

,
.

4.2. Sensitivity analysis of ozone formation

This study compares the ozone concentration of
average all grids in Gang-hwa and Seoul (including West
Seoul and East Seoul) including base case and 9 scenarios
in 3 days (August 22, 23, 24).
In Gang-hwa, the sensitivity of ozone concentration in
August 23 is highest and much higher than on August 22,
24 because the ozone concentration is highest on this
day.When reducing 50% VOC, ozone concentration
decreases than base case. On the contrary, reducing 50%
NOx, ozone concentration increase than base case. As a
result, ozone concentration in Gang-hwa appears in VOC
limited area of EKMA figure. In addition, reducing 50%
VOC and NOx, ozone concentration decreases than base
case so ozone concentration in Gang-hwa appears on the
left of EKMA figure.
When reducing 50% VOC, ozone concentration in East
Seoul decreases than base case. On the contrary, reducing
50% NOx, ozone concentration increases than base case.
As a result, ozone concentration in East Seoul appears in
VOC limited area of EKMA figure. In addition, reducing
50% VOC and NOx, ozone concentration decreases than
base case so ozone concentration in East Seoul appears on
the left of EKMA figure. The ozone concentration in West
Seoul is similar to that in East Seoul.

Mean absolute gross error (MAGE):

|
,

to 0.580. The correlation coefficient of average 8 hours are

higher than 1 hour, they are from 0.439 to 0.679.

, : Modeling concentrations
, : Observation concentrations

3.3. Sensitivity analysis

The sensitivity analysis of ozone formation evaluates
the impact of changing emission inventory on ozone
formation in the urban and the rural of Korea. This study
evaluates the difference in peak ozone concentrations by
reducing VOC and NOx emission using the same
meteorological input and chemical transport schemes. In
order to analyse the effects of VOC and NOx emissions on
the sensitivity of ozone formation, the study performs 9
scenarios. Because the level of effect of VOC and NOx on
ozone formation is not similar, this study reduces 10%,
25%, 50% of VOC; reduces 10%, 25%, 50% of NOx;
reduces 10%, 25%, 50% of VOC and NOx to analyse the
sensitivity of ozone formation.
4. Results and discussion
4.1. Evaluation of air quality modeling performance
The observation data is compared with the base case
simulated results to evaluate CMAQ performance by using
the algorithms in Table 1. Generally, the mean bias (MB)
is negative, so the observation ozone concentration is
higher than modeling ozone concentration at all sites. The
correlation coefficient (R) between observation and
modeling ozone concentrations for all sites are from 0.347

On August 22, the sensitivity of ozone formation in

Gang-hwa is lowest because this day has the heavier rain
than East Seoul and West Seoul.NOx emission in East
Seoul is higher than in West Seoul but VOC emission in
West Seoul is higher than in East Seoul, so the emission in
East Seoul and West Seoul are similar. However, the
sensitivity of ozone formation in West Seoul is 10.98%
higher than the sensitivity of ozone formation in East
Seoul. The wind direction in Seoul is West North from 1:00
10:00 and East North from 10:00 24:00 so West Seoul
is affected by emission from East Seoul on August 22. As
a result, the sensitivity of ozone formation in West Seoul is
higher than in East Seoul because of the wind speed and
wind direction.
On August 23, the sensitivity of ozone formation of
average all grids in Gang-hwa is highest, it is 34.45%
higher than the sensitivity of ozone formation in East Seoul
and 23.41% higher than the sensitivity of ozone formation
in West Seoul. The sensitivity of ozone formation of
average all grids in West Seoul is 9.84% higher than the
sensitivity of ozone formation in East Seoul. However, the
emission in Gang-hwa is lowest and the emission in East
Seoul and West Seoul are similar in this day. The wind
direction in Seoul is East almost the day, the wind direction
are East North and East South at some hours and the wind
speed in East Seoul is lower than in West Seoul. The
sensitivity of ozone in Gang-hwa is highest because Ganghwa is affected by other areas. The sensitivity of ozone
formation in West Seoul is higher than in East Seoul
because West Seoul is affected by East Seoul.
On August 24, the sensitivity of ozone formation in

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

West Seoul is highest, but the emission in West Seoul is

similar to that in East Seoul. The sensitivity of ozone
formation of average all grids in West Seoul is 25.2%l
higher than the sensitivity of ozone formation in East Seoul
because the wind direction in Seoul are East and East South
with high wind speed almost the day and the wind direction
change many directions with low wind speed some hours.
5. Conclusions
The Community Multi-scale Air Quality (CMAQ)
modeling system has been designed to approach air quality
as a whole by including state-of-the-science capabilities for
modeling multiple air quality issues, including
tropospheric ozone. This study used MM5 - SMOKE CMAQ modeling system to analyse the sensitivity of ozone
formation in urban (Seoul includes East Seoul and West
Seoul) and rural (Gang-hwa). The nine different sensitivity
scenarios (reducing 10%, 25%, 50% of VOC; reducing
10%, 25%, 50% of NOx; reducing 10%, 25%, 50% of VOC
and NOx) were analysed in August 22, 23, 24, 2007. The
average all grids in each region are chosen to analyse the
sensitivity of ozone formation in this study.
From the results of this study, some conclusions can be
brought out: The ozone concentration in Seoul almost
decreases earlier than in Gang-hwa within a day. Ozone
concentration in Gang-hwa, East Seoul and West Seoul
almost appears in VOC limited area of EKMA figure.
Therefore, VOC control strategy could be the best approach
in reducing peak ozone formation in Gang-hwa, East Seoul
and West Seoul. NOx emission in East Seoul is highest,
VOC emission in West Seoul is highest and VOC emission
could be the best way to reduce ozone concentration in
Gang-hwa, East Seoul and West Seoul. For this reason,
reducing VOC emission may reduce the ozone concentration
in East Seoul, West Seoul and Gang-hwa, especially,
reducing VOC emission in West Seoul. The sensitivity of
ozone formation in West Seoul is higher than in East Seoul
because the wind creates the transport of ozone from East
Seoul. On August 23, the emission of Gang-hwa is lowest
but the sensitivity of ozone formation in Gang-hwa is highest
because of the wind direction and wind speed. The transport
of ozone precursors from Seoul can cause significant ozone

production in Gang-hwa on August 23, 2007.

REFERENCES
[1]

Community Modeling and Analysis System Institute for the

Environment University of North Carolina at Chapel Hill, NC,
Operational Guidance for the CMAQ Modeling System v4.7.1, 2010.

[2]

Che-Jen Lin, Thomas C.Ho, Hsing-wei Chu, Heng Yang,

SantoshChandru, NageshKrishnarajanagar, Paul Chiou, Jack
R.Hopper, Sensitivity analysis of ground-level ozone concentration
to emission changes in two urban regions of southeast Texas,
Journal of Environmental Management 75, 315 323, 2004.

[3]

C.Duenas, M.C. Fernandez, S. Canete, J. Carretero, E. Liger,

Analyses of ozone in urban and rural sites in Malaga (Spain),
Chemosphere 56, 631 639, 2004.
Finlayson-Pitts B.J. and N.Pitts, VOCs, NOx and ozone production.
J. Air Waste MgmtAssoc, 43, 1093-1101, 1993.

[4]
[5]

Weimin Jiang, Donald L. Singleton, Mark Hedley and Robert

McLaren, Sensitivity of ozone concentrations to VOC and NOx
emissions in the Canadian lower fraser valley, Atmospheric
Environment Vol. 31, No. 4, 627 638, 1997.

[6]

Li Li, Changhong Chen, Cheng Huang, Haiying Huang, Gangfeng

Zhang, Yangjun Wang, Minghua Chen, Hongli Wang, Yiran Chen,
D.G. Streets, Jiamo Fu, Ozone sensitivity analysis with the MM5CMAQ modeling system for Shanghai, Journal of Environmental
Sciences 23, 1150-1157, 2011.

[7]

G.D. Hayman, J.Abbott, T.J. Davies, C.L. Thomson, M.E. Jenkin, R.

Theford, P. Fitzgerald, The ozone source-receptor model A tool for
UK ozone policy, Atmospheric Environment 44, 4238 4297, 2010.
Veronica Gabusi, Marialuisa Volta, Seasonal modeling assessment
of ozone sensitivity to precursors in northern Italy, Atmospheric
Environment 39, 2795 2804, 2005.

[8]

[9]

Xiao Tang, Zifa Wang, Jiang Zhu, Alex E. Gbaguidi, Qizhong Wu,
Jie Li, Tong Zhu, Sensitivity of ozone to precursor emissions in
urban Beijing with a Monte Carlo scheme, Atmospheric
Environment 44, 3833 3842, 2010.
[10] Yen-Ping Peng, Kang-Shin Chen, Hsin-Kai Wang, Chia-Hsiang Lai,
Ming-Hsun Lin. Applying model simulation and photochemical
indicators to evaluate ozone sensitivity in southern Taiwan, Journal
of Environmental Sciences 23, 790 797, (2011).
[11] Xiao-Huan Liu, YNG Zhang, Jia Xing, Qiang Zhang, Kai Wang,
David G. Streets, Carey Jang, Wen-Xing Wang, Ji-Ming Hao (2010)
Understanding of regional air pollution over China using CMAQ, part
II. Process analysis and sensitivity of ozone and particulate matter to
precursor emissions, Atmospheric Environment 44, 3719 3727.
[12] Pedro Jimenez, OriolJorba, Rene Parra, Jose M. Baldasano (2006)
Evaluation of MM5-EMICAT2000-CMAQ performance and sensitivity
in complex terrain: High-resolution application to the northeastern
Iberian Peninsula, Atmospheric Environment 40. 5056 5072.

(The Board of Editors received the paper on 05/14/2015, its review was completed on 07/03/2015)

Tao Quang Bang, Hoang Van Thanh

STUDY OF DESIGNING AND MANUFACTURING SIMILI FILM-COATING MACHINE

Tao Quang Bang, Hoang Van Thanh
The University of Danang, University of Science and Technology; hvthanh@dut.udn.vn; hoangvanthanh_03c1b@yahoo.com
Abstract - The purpose of this paper is to present research results
of designing and manufacturing of simili film coating machine on
plastic profile to produce wood simili products. Based on advantages
and disadvantages of the products which have been used widely in
the market recently, the researchers decided to choose a method of
using simili film coating on plastic materials to make new products
that can replace others made from wood. The study designed 3D
drawing of Coating machine system thanks to Solidworks software,
including four main units such as stock feeding unit, coating unit,
cutting and heating unit, glue providing and glue drying unit. The
system was fabricated successfully on CNC machine. Experimental
coating process was carried out at QuangThanh Plastic Company,
HoaKhanh industrial Zone, Danang City and the experimental result
shows that recommended coating temperature is about 1220C.
These wood simili products initially showed their effectiveness of
coated-surface quality and easier pressing roller adjustment.

Vietnam as shown in Figure 1. The uPVC material contains

no contaminants, no heavy metals like lead, cadimi,
andharmless substances such as canxi, zinc [1]-[2].
uPVC profile is hard plastic slat manufactured on twinscrew extruder including PVC plastic (K65 K66),
thermostabilizer, lubrication, bulking agent, annealing
color, UV ray resistance.

Key words - Simili film; plastics profile; CAD/CAM/CNC; coating;

heat treatment.

1. Introduction
Wood is raw material widely used in industry,
agriculture, transport, architecture, construction and mine
ores. Nowadays, many consumers have chosen products
made from wood for their families because these wood
products have some advantages such as heat and electricity
insulation, good moisture proof, small heat of expansion,
hardness, high durability and high aesthetic value. However,
wood source in Vietnam as well as in the world is becoming
more and more exhausted since so many natural forests have
been destroyed by illegal wood exploitation. Therefore,
research and application of a new material to replace natural
wood are extremely essential in conserving ecological
environment. One of the materials used to replace natural
wood is uPVC plastic. The uPVC material has many
advantages such as no warping, no shrinkage, high flame
resistance, no oxidization, no limit of age or no discoloration
in any radiant conditions of sun and acid rains in comparison
with wood material. Besides, this material has much lower
cost than the products made from wood material. As a result,
windows, doors, baffles, ceilings made from uPVC material
have been widely used all over the world.
In the world, many companies manufactured equipment to
produce wood replacement products, such as spray system,
coating machine. However, this production is very bulky and
heavy. Moreover, it is difficult to adjust coating temperature,
pressing rollers position and it costs a lot of money. This leads
the researchers to redesign and manufacture simili film
coating machine so that the coating quality may be enhanced
and the cost will be reduced. More importantly, it is easier to
operate the coating machine system.

Figure 1. uPVC plastic profile

2.2. Simili and glue

Simili coated on plastic uPVC profile is provided by Phu
Lam Plastic Industrial Stock Company, Vietnam as shown
in Figure 2. The simili has advantages as follows: moisture
resistance is high, it is easy to clean surface, and there will
be no discoloration. Glue employed in the paper is Nanpao,
Taiwan as shown in Figure 3, [3].

Figure 2. Simili film

2. Experimental Set-up and Method

2.1. Materials
The material used in this study is plastic uPVC slat
obtained from QuangThanh Plastic Company, Da Nang,

Figure 3. Nanpao glue

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

2.3. Designing and Manufacturing Tool

SolidWorks R2011 was used to design simili film
coating machine. All of components were fabricated by
CNC machine as shown in Figure 4 and Figure 5.
Technical requirements were inspected by metrologies
system as shown in Figure 6. Fabrication and inspection
were carried out at Institute of Mechanical Technology and
Automation, Danang University of Technology, The
University of Danang [6]-[9].

In order to define coating temperature, the study

explored glass transition temperature and melting point of
PVC plastic [1].
3. Results and Discussion
The authors researched and designed a simili film coating
system by SolidWorks software as shown in Figure 7. The
system manufactured includes main four units: stock feeding
unit, coating unit, cutting unit, and heating and glue
providing unit as shown in Figure 8. The simili film coating
machine utilizes coating simili film on uPVC profile at
QuangThanh plastic company, HoaKhanh industrial Zone,
Da Nang city, Vietnam. The results show that the quality of
the coated surface is high and the models of each product are
various, and importantly, these products meet the needs of
the consumers as shown in Figure 9.

Figure 4. CNC Milling Machine

Figure 7. 3D Simulation on SolidWorks

Figure 5. CNC turning Machine

Figure 8. Experiment of Coating machine

Figure 6. Coordinate Measuring Machine

2.4. Experimental Method

Thanks to plastic uPVC profile, the study calculated
and designed different impression rollers which pressed
simili film on the plastic uPVC profile. Pressing force of
rollers is then defined and motor power was also
determined. Furthermore, the authors computed and
designed clamping unit and cutting unit of plastic uPVC
profile [4]-[5]-[8].

Figure 9. Simili coated uPVC plastic profile

uPVC profile produced from twin screw extruder is

cleaned, brushed oil-paint, added thermal prior to moving to
coating system. The simili film is simultaneously fed with
uPVC profile through the glue providing system, thermal
addition, and drying fan. The simili film is then coated on
uPVC profile owing to the shaped rollers system arranged
logically. If the consumers have any requirements about the

Tao Quang Bang, Hoang Van Thanh

length of products, this system is automatically controlled to

define the length and carry out cutting process.
Experimental result shows that recommended coating
temperature is about 1220C. When pressing force is
increased, it will enhance the quality of coated uPVC
profile. On the other hand, if uPVC profile is changed with
different shapes, it is necessary to adjust pressing rollers
position as well as select different rollers shaped with
uPVC profile.
4. Conclusion
This paper has introduced a new method to plastic
companies in order that they can manufacture high-grade
interior products of diverse designs from wood
replacement plastic material with high quality. The study
has made a significant contribution to reducing the
customers needs for natural wood which costs a lot to
decorate their houses. Simili film coating machine on
plastic profile will be considered as the one that can
produce wood replacement articles which are popularly
used in Vietnam and all over the world.

Acknowledgment
The authors would express their acknowledgement to
QuangThanh Plastic Company and Chip Viet Technology
Solutions Co., LTD of providing equipment and financial
support for manufacturing the machine.
REFERENCES
[1]

Van der Vegt A. K., "From polymers to plastics", DUP Blue print, 2002.

[2]

Hoang Van Thanh, "Injection Molding Technology of Optical Elements

in LED Illumination", LAP Lambert Academic Publishing, 2013.

[3]
[4]
[5]

GhoshP., "Adhesives and coatings Technology", Tata McGraw-Hill, 2008.

KarwaR., "Machine Design", Laxmi Publication (P) LTD, 2006.
BhandariV. B., "Design of Machine Elements", Tata McGraw-Hill, 2008.

[6]

NarayanaK. L., RamanaS. V., KrishnaP. V., "Production

Technology", Second Edition, I. K. International Publishing House
Pvt. Ltd., 2010.
Nguyn Vn Khi, "Keo dn ha hc v cng ngh", Nh xut bn
Vin Khoa hc v Cng ngh Vit Nam, H Ni, 2006.
Nguyn Trng Hip, Nguyn Vn Lm, "Thit k chi tit my", Nh
xut bn Gio dc, 1998.

[7]
[8]
[9]

Trn Vn ch, Nguyn Trng Bnh, Nguyn Th t, Nguyn Vit

Tip, Trn Xun Vit, "Cng ngh ch to my", Nh xut bn Khoa
hc v K thut, 2003.

(The Board of Editors received the paper on 06/03/2015, its review was completed on 06/11/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

HEAVY METALS ACCUMULATION OF FOOD-CROPS GROWN

IN AGRICULTURAL AREAS AFFECTED
BY INDUSTRIAL WASTEWATER, DA NANG, VIETNAM
Doan Chi Cuong1, Masaki Takaoka2, Vo Van Minh1, Phan Nhat Truong1
1

Danang University; doanchicuong@gmail.com; vominhdn@gmail.com; nhattruong.wbf@gmail.com

Graduate School of Global Environmental Studies, Kyoto University; takaoka.masaki.4w@kyoto-u.ac.jp

Abstract - Heavy metal pollution of agricultural soils resulting from

wastewater irrigation has been causing major concern due to the
potential risk involved. In the present study, food-crops (Chrysanthemum
coronarium; Ocimum basilicum; Coriandrum sativum), irrigation water,
and corresponding agricultural soil were assessed for Pb and As
contamination. Study results reveal that contents of Pb and As in
agricultural soils are lower than their permissible values in comparison
with Vietnam technical regulations on the allowable limits of heavy metals
for agricultural soils. By contrast, the concentration of Pb and As in water
samples are higher than its allowable level for irrigation water. Besides,
most of As values in food-crops grown surrounding Hoa Khanh industrial
zone (IZ) are lower than its permissible level in vegetables. Meanwhile, a
vast fraction of Pb contents in food-crops have exceeded its allowable
level for both leafy and spice vegetables. The associated risk was
assessed using HI index. The results indicate that there is a health risk
associated with the ingestion of contaminated food-crops (HI > 1).
Key words - Hoa Khanh IZ; heavy metals; health risk assessment;
food-crops; Igeo .

1. Introduction
Heavy metals occur as natural constituents of the earth
crust, and are persistent environmental contaminants since
they cannot be degraded or destroyed by natural processes
[1]. Besides, with the rapid development of economy and
society, pollution caused by industrial activities has been and
continues to be a major cause of environmental
deterioration. Slow depletion of heavy metals also takes
place through leaching, plant uptake, erosion and deflation.
This activity adversely affects the quality of environment so
much that it becomes a subject of serious concern worldwide
[2]. Plenty of studies have already concluded that areas in
close proximity to industrial activities are marked by
noticeable contamination of air, soil, and water.
The effects of metal pollution on local environments
and organisms may therefore be substantial and long
lasting in spite of extensive remediation efforts [3]. Heavy
metals can be enriched via the food chain. Once the soil
and water suffer from heavy metal contamination, it is
difficult to be remediated and moreover, it was transferred
and accumulated in food crops [4]. Food consumption
contaminated with heavy metals is a major contributor
pathway to human exposure than any other pathway such
as inhalation and dermal contact. Heavy metal
contamination is a known causative of various disorders
such as genomic instability, endocrine disruption,
neurotoxicity, carcinogenicity, immunological problems
and also impaired psycho-social behavior.
Lead and arsenic are known as two of the most toxic heavy
metals. The most sensitive targets for Pb toxicity are the
developing nervous system, the hematological and
cardiovascular systems, and the kidney. However, due to the

multi-modes of action of lead in biological systems, lead could

potentially affect any system or organs in the body [5].
Consequently, Pb is responsible for elevated blood pressure,
renal and tumor infection, improper hemoglobin synthesis and
reproductive system. Exposure to lead produces various
deleterious effects on the hematopoietic, renal, and
reproductive and central nervous system, mainly through
increased oxidative stress [6]. While, As is a potent toxicant
that may exist in several oxidation states and in a number of
inorganic and organic forms. Acute arsenic poisoning is
associated initially with nausea, vomiting, abdominal pain,
and severe diarrhea. Encephalopathy and peripheral
neuropathy are also reported. Chronic arsenic toxicity results
in multisystem disease. Arsenic is a well-documented human
carcinogen affecting numerous organs [7].
Da Nang is considered as a region of strong
industrialization, especially steel industry. The unplanned
development of this industry has led to many
environmental consequences. Though most of the places
surrounding steel factories were reported being polluted,
the level of the pollution and the impact on humans have
not been assessed. With the expansion of urbanization,
there has been a decrease not only in quality of soil and
water in agricultural areas surrounding the industrial zones
but also in agricultural production acreage. Meanwhile, the
demands of food stuffs such as vegetables, cereals, and
meat for daily meals are more and more increasing in both
quantity and quality. Consuming unsafe products will put
the consumers at risk, thus, an assessment of Pb and As in
food crops is very necessary.
Absorption and elimination of contaminants by the
plants depend on several factors such as climate conditions,
soil characteristics, kind of pollutants, and especially
physiological characteristics of each plant [8]. Therefore,
this study is conducted to evaluate the absorption of heavy
metals in different parts of the food crops (especially the
edible parts). Thence, we did: (1) Measure the Pb and As
contents in soil and some food crops; and (2) Assess the
human health risk caused by heavy metals via using foodcrops in daily meal of local people.
2. Material and methods
2.1. Description of study site
Hoa Khanh IZ was established in 1998 with a total area
of 395.72 hectares, of which 298.25 hectares have been
used for hire. In 2004, expanded-Hoa Khanh IZ - one of the
new industrial zones of the city was built in Lien Chieu
district with a total area of 216.52 hectares; in which 132.6

10

hectares of land are used for factories construction, the rest

are for public buildings, plant trees and services-areas for
the project. Before the expansion, all waste water of oldHoa Khanh IZ was treated by wastewater treatment Plant
with capacity of 2000 m3/day. Although in 2004 city
government decided to extent more, the Wastewater
Treatment Plant did not elevate the treatment capacity. At
the end of 2013, the capacity of wastewater collection
overall Hoa Khanh IZ was 95%. Therefore, about 600 m3
of untreated wastewater pour into surrounding agricultural
areas per day [9].
The farm land vicinity of Hoa Khanh IZ has about 400
households having subsistence mainly relying on
agricultural activities with an area of approximately 52
hectares for cultivation. However, because of the influences
of wastewater from this site in recent years, agricultural land
was seriously polluted, cultivatable area decreased to just
over 25 ha. In the past, this area was paddy field and this was
also the main livelihood of the local people who lived around
Hoa Khanh IZ. Now most of the people have switched to
plant vegetables instead of rice (Oryza sativa) because
agricultural land was seriously polluted.
2.2. Soil sampling and analysis
Soil at agricultural area surrounding Hoa Khanh IZ was
collected by digging a monolith of 10 cm x 10 cm x 20 cm
size by using a plastic scooper. Non soil particles such as
stones, wooden pieces, rocks, gravels, organic debris were
removed. Soil was oven dried and then it was sieved via a
2 mm sieve and stored in the labeled polythene sampling
bags (According to the guideline of TCVN 7538-2:2005).
Soil samples had been natural-dried by sunlight before it
were oven-dried at 600 C by Oven (Model: UNE500) until
constant weight. Take 3 g of soil samples which were
digested by 15 ml tri acid mixture namely, HNO3, HClO4
and H2SO4 at 5:1:1 ratio (volume - mL) at 1800C until the
transparent solution appeared. Water samples were filtered
via Whatman No. 42 filter papers. The filtrate was adjusted
to 100 mL by adding distilled water and stored for further
analysis. Concentrations of Pb and As in the digested
samples were determined using a graphite furnace atomic
absorption spectrophotometer (GFAAS Zenit 700P). Each
preparation of sample was repeated in triplicate.
2.3. Food crops sampling and analysis
Standing food crop samples including Japanese-green
(Chrysanthemum coronarium), Basil (Ocimum basilicum),
and Chinese parsley (Coriandrum sativum) were also
collected from the same sites where soils were collected. At
harvest, plants were divided into leaves and roots, and
properly washed with deionized water to remove all visible
soil particles. Leaves were rinsed briefly in deionized water
while roots were properly washed with tap water and finally
with deionized water to remove all visible soil particles. The
washed plant samples were oven-dried at 60C to a constant
weight. The dried samples were ground and sieved via 2mm
mesh size (According to guideline of TCVN 9016:2011). 3 g
of vegetable samples were subsequently digested in HNO3
and HCl in 3:1 ratio at 1800C. After cooling down, the
suspensions were filtered and filtrate was adjusted to 100 mL
with distilled water. Concentrations of Pb and As in the

Doan Chi Cuong, Masaki Takaoka, Vo Van Minh, Phan Nhat Truong

digested samples were determined using a graphite furnace

atomic absorption spectrophotometer (GFAAS Zenit 700P)
(According to the guideline of TCVN 6496: 2009). With
water-hyacinth samples, the pretreatment and analysis are
conducted like vegetable samples but they are not divided into
leaves and roots.
2.4. Index of geoaccumulation
Originally, the index of geo-accumulation (Igeo) was
proposed by Muller (1969) to assess the heavy metal
contamination in bottom sediments by comparing current
and preindustrial concentrations. Danuta Wiechula (2004)
[10] applied this index to assess the soil contamination by
heavy metal in farming soil. It can be calculated by using
the following equation:
Cn
Igeo = log2
(1)
1.5Bn
Where Cn referred to the total concentration of heavy
metals in soil samples, and Bn referred to the
concentrations of heavy metals in the Earths crust [11].
The Bn values of Pb and As in Earths crust were 20 mg/kg
and 1.5 mg/kg, respectively. Factor 1.5 is used because of
natural fluctuations in background values for a given heavy
metal in the environment as well as very small
anthropogenic influences.
According to Gong Qingjie (2008), Muller has
distinguished six classes of the geoaccumulation index as
in the following table:
Table 1. Igeo values and classes with corresponding to soil quality
Class

Value

Soil quality

Igeo 0

0 <Igeo 1

2
3
4
5
6

1 <Igeo 2
2 <Igeo 3
3 <Igeo 4
4 <Igeo 5
Igeo> 5

practically uncontaminated
uncontaminated to moderately
contaminated
moderately contaminated
moderately to heavily contaminated
heavily contaminated
heavily to extremely contaminated
extremely contaminated

2.5. Health risk assessment

Health risk index (HRI) was used to determine the
health risks of consuming food crops by local inhabitants.
The method was provided by US EPA (2007). HRI is
expressed as the ratio of determined dose of a pollutant to
a reference dose. An index more than 1 is considered not
safe for human health and versus, the exposed population
is unlikely to experience obvious adverse effects if the ratio
is less than 1. The health risk index for Pb and As by
consumption of contaminated food crops was calculated by
the following equation [12]:
HRI =

EF x ED x FI x C x 10-3
RfD x BW x AT

(2)

Where EF (365 days/year) is exposure frequency; ED is

exposure duration (70 years); FI (g/person/day) is vegetable
ingestion (157 g/person/day for adult Vietnamese); C
(mg/kg on fresh weight basis) is heavy metal concentration
in food crops (edible part leaves); RfD (mg/kg/day) is the
oral reference dose; BW is the average body weight, which

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

are considered to be 54 kg and 46 kg for adult males and

females (applied for Vietnamese), respectively; AT is
average time for noncarcinogens (365 days/year number
of exposure years, assuming 70 years in this study). RfD
values for Pb and As were 0.004 mg/kg/day and 0.0003
mg/kg/day), respectively (US EPA, 2000). RfD is measured
heavy metals dose that a person can be exposed in a day
without being at any health risk in the whole life.

11

3. Results and Discussion

3.1. The concentration of Pb and As
The range and mean concentration of Pb and As in surface
soil, water, and water-hyacinth (Eichhornia crassipes) samples
surrounding Hoa KhanhIZ are summarized in Table 2.

Table 2. The concentration of Pb and As in soil (mg/kg), water (mg/L), and water-hyacinth (mg/kg-FW)
Elements

Pb

As

Samples

Minimum

Maximum

Median

Mean

Permissible level

Soil (n=19)

2.04

30.05

6.01

8.881.86

70 (a)

Water (n=5)

0.76

3.13

1.95

1.960.39

0.1 (b)

Water-hyacinth (n=3)

16.21

27.87

17.48

20.523.70

Soil (n=19)

0.00

1.52

0.35

0.440.09

12 (a)

Water (n=5)

0.04

0.58

0.30

0.290.09

0.1 (b)

Water-hyacinth (n=3)

0.22

4.04

1.36

1.871.13

QCVN 03: 2008/BTNMT: Vietnam technical regulation on the allowable limits of heavy metals for agricultural soils
Decision No. 106/2007/Dec-BNN: The permissible level of heavy metal in irrigation water for agriculture

Lead contaminated soil can pose a risk through direct

ingestion or uptake in vegetable gardens and the common
use of Pb makes its concentration in all environmental
media elevated [13]. In this study, the mean value of Pb in
surface soils samples were 8.88 mg/kg with a range of 2.04
mg/kg - 30.05 mg/kg, much lower than the permissible
limit of lead content in Vietnam agricultural soil (70
mg/kg). Meanwhile, the mean content of Pb in water
samples was 1.96 mg/l with a range from 0.76 mg/l to 3.13
mg/l 20 times more than the permissible level of lead
concentration for irrigation water (0.1 mg/l).
Ferric hydroxide generally plays an important role in
controlling the concentration of As in soils and soil
solutions. Significant anthropogenic sources of As are
related to industrial activities (metal processing, chemical
works based on S and P minerals) and to the use of
arsenical sprays, particularly in orchards. The As
contamination of paddy soils resulted from both high
sorption capacity of these soils and the As transportation
through irrigation. In the tested soils of this study, it ranged
from 0 to 1.52 mg/kg with the mean value of 0.44 mg/kg,
very lower than the values in the allowable limits of heavy
metals for agricultural soils of Vietnam (12 mg/kg).
Meanwhile, the mean concentration of As in water sample
(0.29 mg/l) was three times higher when compared to the
permissible level of As in irrigation water for agriculture
of Vietnam (0.1 mg/l).
Farmers in the vicinity of Hoa Khanh IZ not only use
the chemical fertilizers in agricultural production, but also
often have a habit of using organic fertilizer made from
water - hyacinth (Eichhornia crassipes) - the popular plant
in this area is considered as a super-absorbent and hyperaccumulated heavy metal species. This was demonstrated
by the mean Pb and As in water-hyacinth values, those
were 20.5 mg/kg and 1.87 mg/kg respectively, much more

higher than the mean contents in corresponding soil and

water samples.

Figure 1. Indexes of geo-accumulation of Pb and As in soil samples

The Igeo values of Pb and As were mostly negative,

indicated that nearly all the examined soil samples were not
contaminated for Pb and As (Figure 1).
3.2. Pb and As contents in different parts of food-crops
Food-crop samples of this study included 3 vegetables
named Japanese-green (Chrysanthemum coronarium), Basil
(Ocimum basilicum), and Chinese parsley (Coriandrum
sativum). The contents of heavy metals (mg/kg) in different part
of food-crop samples are given in Table 3. The concentrations
of metals varied greatly among food-crop species. In general,
the mean concentrations of Pb and As in roots were found
higher than in corresponding leaves samples. The Pb contents
were at 8.85 mg/kg and 10.11 mg/kg in roots and 5.13 mg/kg
and 5.14 mg/kg in leaves of Japanese-green and Basil,
respectively. The As contents were at 0.11 mg/kg and 0.28
mg/kg in roots and 0.1 mg/kg and 0.05 mg/kg in leaves of
Japanese-green and Basil, respectively.

12

Doan Chi Cuong, Masaki Takaoka, Vo Van Minh, Phan Nhat Truong

Table 3. The content of Pb and As in food-crops

Plant species

Portion

Pb (mg/kg-FW)

As (mg/kg-FW)

Range

Median

Mean

Range

Median

Mean

Chrysanthemum
coronarium(n=11)

Roots

2.62-39.96

6.48

8.853.17

0.00-0.23

0.12

0.110.02

Leaves

0.83-18.06

3.33

5.131.60

0.00-0.21

0.08

0.100.02

Ocimumbasilicum(n=8)

Roots

3.36-42.13

5.35

10.114.62

0.00-1.79

0.03

0.280.22

Leaves

2.64-7.13

5.22

5.140.62

0.00-0.11

0.06

0.05-0.02

Leaves

2.21-43.69

4.11

15.015.66

0.00-0.34

0.06

0.130.04

Coriandrumsativum(n=10)

Permissible level*

1.0

1.0

* Decision No. 106/2007/Dec-MARD: The permissible level of heavy metal in vegetables (Vietnam Agriculture and Rural Development Ministry)

Chinese parsley is the spice food-crop that is used

popularly in daily food processing by Vietnamese in
particular and Asia in general. Study results showed that,
the mean Pb content in leaves (15.01 mg/kg) of Chinese
parsley were 1.5-2 times higher than the ones in Japanesegreen and Basil. On average, the Pb contents in edible part
(leaves) of all food-crops in the current study have
exceeded the Vietnam permissible level of heavy metal in
vegetables (1.0 mg/kg).
According to Xiangdong Li [14], one of major factors
contributing to high metal accumulation in food-crops
could be ascribed to atmospheric deposition. When fuels
were burned, heavy smoke containing various kinds of
heavy metals, metalloids, and organic pollutants were
discharged into the air, and the food-crops growing in those
areas might be the first recipients of Pb and As as well. The
foliar uptake of atmospheric Pb has been proved to be the
dominant pathway for Pb entering the leaf grown in nearby
industrial zone.
3.3. Uptake and translocation of Pb and As in food crops
The capacity of food crops to uptake heavy metals from
soils can be assessed using the bio-concentration factor
(BCF), the BCF values of Pb and As were calculated by
dividing the concentration of a metal in a food crops (dry
weight) by the total soil metal concentration, defined in the
following equation:
BCF =

Heavy metal content in root (mg/kg)

Heavy metal content in soil (mg/kg)

(3)

The higher BCF reflects the poorer retention in soil or

greater efficiency of plants in absorbing heavy metal, and
low BCF reflects the strong sorption of metal to the soil
colloid.
The translocation capacity of Pb and As between the
roots and aerial parts (shoot, leaves or grain) of food-crops
can be defined by the translocation factor (TF), defined in
the following equation:
TF =

Heavy metal content in leaves (mg/kg)

Heavy metal content in roots (mg/kg)

(4)

Translocation factor is one of the main components of

human exposure to heavy metals via the food chain. The
BCF and TF values of Pb and As were calculated and given
in Table 4.
As seen from Table 4, there was a relative difference in
BCF and TF values that was observed among different

food-crops and heavy metals. The BCF values for Pb and

As in Japanese-green were given in descending order as Pb
(0.997) As (0.012) and Pb (1.138) As (0.032) in Basil
respectively.
Table 4. The BCF and TF values across the sampling sites
Food crops

Elements

BCF (soil
to root)

TF (root
to leaves)

Chrysanthemum
coronarium

Pb

0.997

0.579

As

0.012

0.938

Pb

1.138

0.508

As

0.032

0.180

Ocimumbasilicum

The TF values for Pb and As in Chrysanthemum

coronarium were 0.579 and 0.938, whereas the TF values
for Pb (0.508) in Ocimum basilicum were higher than As
(0.18). According to Li, Xiangdong (2011), the TF of
heavy metals can be used to assess the potential capacity of
plants in transportation of heavy metals from soil to plant
or from root to aerial parts. Heavy metals with high TF are
transferred easily from soil to plant and/or from root to
aerial parts of plant than ones with low TF [14].
3.4. Potential risk of Pb and As to human health
HRI is a complex index which has been used for the
evaluation of potential human health risks associated with
long term exposure to pollutants such as heavy metals.
Values of this index for investigated heavy metals Pb and
As were presented in table 6. Individual HRI values of Pb
and As almost exceeded 1 for both males and females in
the study area (HRI values of Pb and As were at 6.21 and
0.93 for males; 7.29 and 1.09 for females respectively).
Table 5.Health risk assessment of Pb and As by HRI index
Inhabitants
Male
Female

Elements

RfD (mg/kg/day)

HRI

Pb

0.004

6.21

As

0.0003

0.93

Pb

0.004

7.29

As

0.0003

1.09

HI
7.14
8.38

According to Hye-Sook (2008), non-cancer risk is

represented in terms of HRI for a single substance; in this
case we assume that the additive effects, HRI can be
summed across constituents to generate a hazard index (HI)
for a specific pathway combination. The HI is a measure of
the potential risk of adverse health effects from a mixture
of chemical constituents. Whether or not a particular

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

chemical mixture poses an additive risk depends on the

targets (tissue, organ, or organ system) and the mechanisms
of action of each chemical [15].
HI values through diet for males and females in the
present study were 7.14 and 8.38, respectively. This suggests
that male and female local people live in the vicinity of Hoa
Khanh IZ may experience adverse health effects. The HI for
males via diet is quite close to that for the females; this
indicates that their levels of health risk are similar.
According to Qichao Wang (2007), the HI is a measure
of the potential risk of adverse human health effects from
a mixture of chemical pollutants. Whether or not a
particular chemical mixture posed an additive risk not only
depends on the targets (tissue, organ, or organ system) but
also on the mechanisms reaction of the individual
chemicals [16].
4. Conclusions

REFERENCES
[1]

Duruibe, J.O., M.O.C. Ogwuegbu, and J.N. Egwurugwu, Heavy

metal pollution and human biotoxic effects. International Journal of
Physical Sciences, 2007. 2(5): p. 112-118.

[2]

Farabegoli, G., et al., Biological treatment of tannery wastewater in

the presence of chromium. Journal of Environmental Management,
2004. 71(4): p. 345-349.
Lade, H., et al., Bioremediation of Heavy Metals from Soil and
Aquatic Environment: An Overview of Principles and Criteria of
Fundamental Processes. Sustainability, 2015. 7: p. 2189-2212.

[3]

[4]

Su, C., L. Jiang, and W. Zhang, A review on heavy metal

contamination in the soil worldwide: Situation, impact and
remediation techniques. Environmental Skeptics and Critics, 2014.
3(2): p. 24-38.

[5]

Registry, A.f.T.S.a.D., Toxicological profile for Lead. 2007, U.S.

Department of Health and Human Services. p. 1-9.

[6]

Malik, R.N., M.U. Khan, and S. Muhammad, Human health risk

from heavy metal via food crops consumption with wastewater
irrigation practices in Pakistan. Chemosphere, 2013. 93(10): p.
2230-8.
Ratnaike, R.N., Acute and chronic arsenic toxicity. Postgraduate
Medical Journal, 2003. 79(933): p. 391-396.

[7]

Based on the heavy metals concentrations, we conclude

that nearly all the contents of Pb and As in agricultural soils
in the vicinity of Hoa Khanh IZ are lower than their
permissible values in comparison with Vietnam technical
regulation on the allowable limits of heavy metals for
agricultural soils. Most of the As values in food-crops
grown around Hoa Khanh IZ are lower than its permissible
level in vegetables. Meanwhile, a vast fraction of Pb
contents in food-crops have exceeded its allowable level
for both leafy and spice vegetables.Food-crops
consumption were identified as the major pathway of
human exposure to heavy metals. Food-crops in the
vicinity of Hoa Khanh IZ are the main source of Pbs and
As intake from foodstuff for adult males ans females. The
HI values of the studied metals > 1 indicated that there is a
health risks associated with the ingestion of contaminated
food-crops.
Acknowledgement
This study was conducted with the financial support of
the GSGES seeds research funding program for overseas
field campuses. We would like to give our thanks to Assoc.
Prof. Izuru Saizen Dr. who connected young researchers of
Danang University with Kyoto Universitys Professors.

13

[8]
[9]

C., A.D., Bioavailability of Trace Metals, in Trace Elements in

Terrestrial Environments. 2001, Springer New York. p. 61-89.
Portal, D.N. Da Nang Industrial Zones information. 2015; Available from:
http://www.danang.gov.vn/portal/page/portal/danang/doanh_nghiep/kcn/hk
and http://www.iza.danang.gov.vn/vie/gioi-thieu/cac-khu-cn-cx.html.

[10] Wiechua, D., K. Loska, and I. Korus, Metal contamination of

farming soils affected by industry. Environment International, 2004.
30(2): p. 159-165.
[11] Taylor, S.R. and S.M. McLennan, The geochemical evolution of the
continental crust. Reviews of Geophysics, 1995. 33(2): p. 241-265.
[12] Wang, Y., et al., Health risk assessment of heavy metals in soils and
vegetables from wastewater irrigated area, Beijing-Tianjin city
cluster, China. Journal of Environmental Sciences, 2012. 24(4): p.
690-698.
[13] O., N.J., A global assessment of natural sources of atmospheric
trace metals. Nature, 1989. 338: p. 47-49.
[14] Li, X., et al., Heavy metal contamination in soils and vegetables near
an e-waste processing site, south China. Journal of Hazardous
Materials, 2011. 186(1): p. 481-490.
[15] Hye-Sook, L., et al., Heavy metal contamination and health risk
assessment in the vicinity of the abandoned Songcheon AuAg mine
in Korea. Journal of Geochemical Exploration, 2008. 96(23): p.
223-230.
[16] Wang, Q., et al., Population health risk due to dietary intake of heavy
metals in the industrial area of Huludao city, China. Science of The
Total Environment, 2007. 387(13): p. 96-104.

(The Board of Editors received the paper on 06/02/2015, its review was completed on 08/19/2015)

14

Do Chi Phi, Le Van Doanh, Phung Anh Tuan

RESEARCH ON MAGNETOMECHANICAL COUPLING RELATION

IN AMORPHOUS METAL CORE TRANSFORMERS
Do Chi Phi2, Le Van Doanh1, Phung Anh Tuan1
1

Hanoi University of Science and Technology, Hanoi, Vietnam; tuan.phunganh1@hust.edu.vn, ledoanhbk@yahoo.com

2
CaoThang Technical College, Hochiminh City, Vietnam; dochiphi@gmail.com

Abstract - The magnetostrictive force is a m ajor cause of noise

and vibration from a transform er. Especially, the m agnetostrictive
force of an amorphous steel core transformer is higher than that of
a silicon steel core transform er. In order to accurately calculate and
evaluate the noise and vibration of a transformer, it is necessary to
fully examine such factors as deform ation, displacem ent, vibration
and magneto-mechanical force. In this paper a generalized
mathem atical model was set up with the above factors taken into
consideration. The mathematical m odel was then developed on a
single-phase amorphous steel core transformer with a capacity of
3,3kVA-220V/115V to result in the degrees of deform ation and
vibration in cases with or without clam ped iron for magnetic legs
and yokes. At the same tim e, these results were evaluated and
com pared with experim ental ones, which helps determine a
reasonable clamping force to minim ize the noise and vibration of
the am orphous steel core transform er.
Key words - transformer; am orphous; vibration; audible noise;
m agnetostriction; m agnetom echanical.

1. Introduction
Amorphous magnetic steel (AMS) has been widely
used in the fabrication of the magnetic core of transformers
thanks to its very low losses. Compared to the traditional
grain-oriented electrical steel, no load loss is decreased to
about 1/5 of silicon steel's [1].
The amorphous alloy is a non-crystal substance
(amorphous state) created by super fast cooling liquids
metal from high temperature. Because there is no time
forcrystal formation or arrangement, the energy loss
(hyste-resis loss) is small when the flux of magnetic
induction passes through the iron core. In addition, eddy
current loss is decreased because the thickness is
approximately 0.03 mm, which is about 1/10 compared to
silicon steel.

In [6], the authors have calculated the vibration of

magnetic cores of power transformers due to
magnetostriction based upon the coupling between the
magnetic field and the mechanical deformation.
Mechanical displacements also have been measured.
However, audible noises and its related sources have not
been investigated.
In the most recent paper [7], the authors have
investigated the audible noises in three-phase three-leg
transformers with different amorphous-cored structures.
The results indicate that amorphous-cored transformer
with a rectangular core has higher vibration intensities, a
toroidal core should have lower core vibrations and audible
noises than the counterparts.
In this paper, the authors have developped a generalised
model for both mechanical and magnetic aspect. Thanks to
this model, mechanical deformation, displacement and
audible noises are linked together. Furthermore, their
contribution to the total noise of the transformer is
quantified. The mechanical parts such as clamping shackle
at yoke and core limb and their influences on the audible
noise have been measured. The optimal clamping force at
each precise position has been pointed out.
2. Mathematical model
The causes of vibrations inside the steel core of
transformers are created by magnetostrictive force. This is
concretized by means of the diagram shown in Figure 1.

Amorphous magnetic steel is very sensitive to the

effects of temperature, deformation or the external
magnetic field. This special material has also a large
magnetostrictive coefficient (20m/m) which causes large
vibration and large displacement in the core of the
transformer compared to the traditional steel core [2], [3].
Because of its energy efficient characteristics,
amorphous magnetic steel core transformers are widely
used in a distribution power system. When these
distribution trans-formers are usually placed in a
residential area, the noise problem caused by
magnetostriction received lots of research interest. In [4],
the authors have used the adaptive active control of
amorphous alloy core transformers in order to reduce
noises. In [5], the authors have presented the results of the
numerical computing of electrical machine vibrations
caused by Lorentz, Maxwell and magnetostriction forces
without any detailed analytic model.

Figure 1. The noise vibration in transformers [8]

Based on [9], [10], [11], we have a substitute equivalent

diagram which is equivalent to mixture electromechanical
systems of the single-phase transformer as follows:

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

15

Magnetic induction

Deformation (caused by an external

magnetic field)

m/m

sH

Elasticity strain factor (depend on H)

m 2/N

Magnetostrictive constant

m/A

Lenght of magnetic circuit

Cross section of steel-core

[m2]

Winding turns number of high-voltage

turn

Velocity of the displacement

m/s

General Calculation Formula:

Figure 2. The structure of survey transformer

ANV

F AB

(1)

B H S

s H

Reviewing the link point CN1, we have:

1
1
1 dFK
FM 11dt
Fb
V1

m11
b11
K 11 dt

1
1
1 dFK
FM 12 dt
Fb

b12
K 12 dt
m12
F1 FM FK Fb FM FK 12 Fb12

F01 FK 12 Fb12 FK 11 Fb11

(2)

11

11

12

(3)

12

11

11

11

12

With X 1 V1 dt the above equation has been rewritten:

Figure 3. The substitute equivalent diagram of transformers in

mixture Magnetomechanical systems
Table 1. Symbols

Name

Physical Description

unit

U0

Max. voltage at high-voltage side

Resistors of the high-voltage coil

Inductance of the high-voltage coil

Li

Inductance components when

considering the magnetostrictive
phenomenon effected to magnetic-leg
and yoke in turns as Figure 2.

Fi

Mechanic-magnetic force (including

magnetostrictive and magnetic force)

M ij

The mass of the leg, yoke and iron

clamps.

Kg

b ij

Damper factor (Viscosity)

N.s/m

Spring constant

N/m

Electromotive induction

K ij

r Absolute

permeability factor

T.m/A

d 2 X1
dX

b11 b12 1 K11 K12 X1

F
m
m

1
11
12

2
dt
dt

(4)
d 2 X1

F01 F1 m11 m12 2

dt

X1 V1dt

With the same above way for the remaining points, we

have the equations of the electromechanical mixture as
follows:
di

U 0 cos(2 f t) iR 2 L dt L1 L 2 L 3 L 4

iN B
lmt

L di NA dB
dt
dt

d 2 X1
dX

b11 b12 1 K11 K12 X 1

F1 m11 m12
dt
dt

d X1
F01 F1 m11 m12
dt

d2X2
dX
b21 b22 2 K 21 K 22 X 2
F2 m21 m22
dt
dt

d2 X2

F02 F2 m21 m22 dt

16

Do Chi Phi, Le Van Doanh, Phung Anh Tuan

dX
d X2
b31 b32 3 K31 K32 X 3
F3 m31 m32
dt
dt

dV

3
F03 F3 m31 m32 dt

2
F m m d X 4 b b dX 4 K K X

41 42
41 42 4
4
41
42

(5)
dt
dt

2
d
X

4
F04 F4 m41 m42

dt

X1 V1dt; X 2 V2 dt; X3 V3dt; X 4 V4dt;

A1 NV1 ; A2 NV2 ;
L2
L1
l1
l2

A
NV

3
3
4 NV4

; L 4
L3
l4
l3

A
B
F

A
B
F3 A3 B; F4 A4 B
;
;

1
2
2
1
Handling the above differential-equation with the
Runge-Kutta method (ODE45) we find out: magnetic
induction
B
(T),
current
intencity
I
(A),
magnetomechanical force F (N) and the displacement
x(m), deformation s(m/m), acceleration a(m/s2) on
magnetic-leg and yoke of the transformer.
2

3. Modeling and Experiments

Metglas 2605SA1 50Hz

3.5
3

1.1

2.5

B-H Curve
Relative permeability-H Curve

0.6

1.5

0.5
0.4
0.3

Relative perm eability

M agnetic induction B [Tesla]

x 10
4

1.3
1.2

0.2
0.1
0
0

0.5
10

20

30

40

50

60

70

80

90

Magnetic field intensity H [A/m]

Figure 4. Magnetization curve and relative permeability factor

of the amorphous materials codes 2605 SA1 50Hz, Hitachi
Metals, USA has been applied to the transformer core
Table 2. Basic electrical parameters and dimension of the
transformer

Item

155/81

Thickness of core BD (m)

0,045

Width of core D(m)

0,142

Height of window H(m)

0,168

Width of window W(m)

0,06

10

Cross-section of core(m )

0,00639

Figure 5. Dimension of the surveyed transformer

Displacement on magnetic leg
Displacement on magnetic yoke

Parameters

Value

No. of Phase

Frequency(Hz)

50

Power (kVA)

3,3

Voltage HV/LV(V)

220/115

Displacement [m]

6,0E-06
4,0E-06
2,0E-06
0,0E+00
0

0,01

0,02

0,03

0,04

2,0E-06

Time[sec]

Figure 6. The displacement on magnetic-leg

and yoke without iron-clamping.

Figure 6 shows an average length of yoke (105mm) and

leg (213mm) then the displacement on magnetic leg and
yoke is xYoke max = 3,15m < xLeg max = 6,84 m.
Displacement on magnetic leg
Displacement on magnetic yoke
7,0E-06

Displacement [m]

2605SA1 50Hz Material, magnetic field intensity

H(A/m) and Magnetic induction B (Tesla) are shown in
Figure 4; Youngs elastic modulus of material
E=120(GPa); Poisson-ratio =0,28; Vacuum permeability
factor 0=4x10-7(T.m/A); Saturation magnetostriction
s=27(m/m).

1
0.9
0.8
0.7

No. of windings HV/LV(turn)

8,0E-06

3.1. Modeling of the mechanical stress and displacement

of the transformer

1.5
1.4

6,0E-06
5,0E-06
4,0E-06
3,0E-06
2,0E-06
1,0E-06
0,0E+00
-1,0E-06 0

0,01

0,02

0,03

0,04

Time[sec]

Figure 7. Ddisplacement of position on magnetic leg & yoke

with iron-clamping

Based on Figure 7, when iron clamping is applied to fix

magnetic-leg and yoke tightly, then xYoke max = 2,9m <
xLeg max = 6,16 m. Thus when the magnetic-leg and yoke
are fixed tightly, the displacement shall be reduced 10% than
no tightly fixed case. It is suitable with Figure 1, when the
magneticleg and yoke are fixed tightly, then the transverse
relative motion of core laminations shall be reduced.
Figure 8 shows the deformation of core without ironclamping Skkmax = 11m and the deformation of core when
applied iron-clamping Sckmax = 10,3m, this reduction is

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Strain [m/m]

6,36%. However, when compared with the result of the

study [6], the displacement and deformation of the
amorphous steel core is higher, which means that the
vibration noise level in the amorphous steel cores would be
higher than that of thd silicon steel core.
1,40E-05
1,20E-05
1,00E-05
8,00E-06
6,00E-06
4,00E-06
2,00E-06
0,00E+00
0

0,2

0,4
0,6
0,8
1
Magnetic induction B [Tesla]

1,2

Figure 8. The deformation of core with and without clamping

From Figure 9 and Figure 10, we recognize that the

max.vibration level when the magnetic leg and yoke have
not fixed tightly is 2,302m/s2, and when the magnetic-leg
and yoke have fixed tightly is 1,963m/s2, this vibration
shall be reduced by 0,339 m/s2 (14,72%).
Core vibration when fixed magnetic leg and yoke

20

Vibration [m/s2]

magnetostrictive force is very strong - approximately

2500N. It is over twenty five times of the magnetic force.
Therefore all of the deformations, displacements and
vibration noises inside the amorphous steel core are mainly
created by the magnetostrictive force. Besides that, the
amorphous steel has a saturation magnetostrictive factor so
big (>20m/m). Thus we can affirm that the amorphous
steel core transformer has a vibration noise higher than that
of the the silicon-steel core transformer.
3.2. Experiments
Similar to the results of the mathematical model, the
experimental measurement results also have shown that if
the magnetic leg and yoke have been clamped through by
torque 4-5 N.m, then their vibration level is the smallest
(Figure 13). When the magnetic leg and yoke have not been
clamped, the vibration level of the steel core is the highest
(Figure 12). The value of amaxkk =3m/s2 and amaxck =2,3m/s2,
the vibration level has reduced by about 23,33%. So when
the magnetic leg and yoke have been clamped tightly, the
transverse relative motion of core laminations shall be
reduced and also the vibration of coil (which has been
created by the magnetic force) shall be reduced much more.
amaxkk [m/s2]: Range of vibration level in case the
magnetic leg and yoke have not been fixed.

Core vibration when without fixed magnetic leg and yoke

amaxck [m/s2]: Range of vibration level in case the

magnetic leg and yoke have been fixed.

15
10
5
0
-5

0,01

0,02

0,03

0,04

akk [m/s2]: The effective vibration level in case the

magnetic leg and yoke have not been fixed.
ack [m/s2]: The effective vibration level in case the
magnetic leg and yoke have been fixed.

-10

Time [sec]

Figure 9. Vibration level in steel core

Mathematical model result

Core vibration when fixed magnetic leg and yoke

3,0

Vibration [m/s2]

Core vibration when without fixed magnetic leg and yoke

4,0

3,0

Vibration [m/s2]

2,0
1,0
0,0

Experimental measurement

result

2,0
1,0
0,0
-1,0
-2,0

-1,0

-3,0
0,01
-4,0

-2,0

0,015

0,02

0,025

0,03

0,035

0,04

Time [sec]

-3,0
0,01

0,015

0,02

0,025

0,03

0,035

0,04

Time [sec]

Figure 10. Vibration level in steel core in a stabilized state

Magnetostrictive forces

Magnetic forces

Figure 12. Results of the vibration level between experimental

measurement and the mathematical model in case
the magnetic-leg and yoke have not been fixed

Forces total

Mathematical model result

3000

Experimental measurement result

3,0

2500

2,0

Vibration [m/s2]

Forces [N]

17

2000
1500
1000
500
0
0

0,2

0,4

0,6

0,8

1,2

Magnetic induction B [Tesla]

Figure 11. The magneto-mechanical force in the steel-core of

the transformer

Basing on Figure 11, we have recognized that the

magnetic force (which is created by the transverse relative
motion of core laminations) is very small 100N, while the

1,0
0,0
-1,0
-2,0
-3,0
0,00255

0,01255

0,02255

0,03255

Time [sec]

Figure 13. Results of the vibration level between experimental

measurement and mathematical model in case the magnetic-leg
and yoke have been fixed.

Figures 12 and 13 show us that the results of the

vibration level via experimental measurement are higher

18

Do Chi Phi, Le Van Doanh, Phung Anh Tuan

than those of the vibration level via the mathematical

model method. The main reason for the above deviation is
that the experimental measurement method has to mention
the vibration of the transformer coil as well.

Audible noises when fixed magnetic leg and yoke

Audible noises when without fixed magnetic leg and yoke
90

Audible noises [db]

Compared with the results of the mathematical model,

the results of experimental measurements dropped down
more than 6%. It is compatible with theoretical and
experimental calculations.

wrench-force to tighten clampers (to keep the magneticleg

and yoke of the transformer core), the min.value a=1.5m/s2
corresponds to the tightening torque (4-5)N.m. If this
tightening torque is converted into the tightening force, it
is approximately 2600N. equal to the total force that has
been calculated via the mathematical model.

80

70

60

50
0,2

0,4

0,6

0,8

1,2

1,4

Induction B [Tesla]

Figure17. Noise levels before and after the magnetic-leg and

yoke have been fixed tightly with tightening-torque (4-5) N.m
Figure 14. Results of the effective-vibration level in case the
magneticleg and yoke have not been fixed.

Figure 17 shows that if the magnetic leg and yoke have

been fixed tightly with the tightening torque (4-5) N.m,
then the sound intensity is reduced by about 3,23 dB
5. Conclusions
In this engineering article, the authors have developed
a new mathematical model. Through this mathematical
model, they have examined the relationship between
Mechanic and Magnetic which include the relevant
elements such as deformation, displacement, vibration,
magnetomechanic force.

Figure 15. Results of the effective-vibration level in case the

magnetic-leg and yoke have been fixed.

Figure 14 and 15 show the effective-vibration level in

case the magnetic-leg and yoke have not been fixed as
akk =1.7m/s2, and the effective-vibration level in case the
magnetic-leg and yoke have been fixed as ack =1,5m/s2 as
mean its reduced 11,76%.
4. Audible noises in transformer
Based on [7], the noise level Nc(dB) shall been
interdepended by the weight of transformer M(tons), the
average-length of magnetic-core L(meters) and the
magnetic induction B (Tesla). The noise of the transformer
has been measured by [12] as follows:

The experimental results from the measurement of

vibration and noise level with the clamping fixed magneticleg and yoke taken into account are of great importance.
They help us determine the tension force to fix magnetic-leg
and yoke of the transformer so that vibration level and noise
level smallest. In this article so through experimental results
and the results of the recently established mathematical
model, we have noticed the tension force of clamping fixed
magnetic-leg and yoke of the transformer equal with the total
force of magnetostrictive force and magnetic force then the
vibration noise level shall be smallest.
Moreover, the frequency of power sources and the
eigen frequency of transformers are also very important
parameters that cause sound vibrations and noise in the
steel core. This issue will be considered within the
framework of further research papers.

Vibration [m/s2]

REFERENCES

1,5
[1]

1
0,5

[2]

0
0

4
6
Moment [N.m]

10

Figure 16. The vibration level in case of using the wrench-force

to tighten clampers (to keep the magnetic leg and yoke)
with different torques

From Figure 16, we can see that when using the

W. N. Harry, R. Hasegawa, L. Albert, and L. A. Lowdermilk,

Amorphous Alloy Core Distribution Transformers", Proceedings
of the IEEE, vol. 79, no. 11, pp. 16081623, 1991.
F. Alam, Study of Magnetostriction in Iron and Cobalt Based
Amorphous Magnetic Materials", No. December, pp. 1921, 2006.

[3]

C. Hsu, C. Lee, Y. Chang, F. Lin, C. Fu, and J. Lin, Effect of

Magnetostriction on the Core Loss, Noise, and Vibration of Fluxgate
Sensor Composed of Amorphous Materials", vol. 49, no. 7, pp.
38623865, 2013.

[4]

J. Liu, Adaptive Active Noise Control of Amorphous Alloy Core

Transformer", pp. 19461949, 2011.
M. Forces, K. Delaere, M. Iadevaia, W. Heylen, R. Belmans, K.
Hameyer, and K. U. Leuven, Finite Element Analysis of Electrical

[5]

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

[6]

[7]

[8]

[9]

Machine Vibrations caused by Lorenzt, Maxwell and

Magnetostriction Forces, No. 2.
L. Zhu, Q. Yang, and R. Yan, Numerical Analysis of Vibration due
to Magnetostriction of Three Phase Transformer Core, pp. 36,
2012.
Y. Chang, C. Hsu, H. Chu, and C. Tseng, Magnetomechanical
Vibrations of Three-Phase Three-Leg Transformer With Different
Amorphous-Cored Structures, vol. 47, no. 10, pp. 27802783,
2011.
M. K. Master and S. T. Stockholm, Modelling of Core Noise from
Power Transformers Master of Science Thesis Modelling of Core
Noise from Power Transformers, 2011.
Y. Le Bras, A. Lasheras, J. Gutierrez, F. Mazaleyrat, and J. M.

19

Greneche, A New Magnetoelastic Resonance Based Technique to

Determine Magnetomechanical Parameters of Amorphous
Ferromagnetic Ribbons, Vol. 84, pp. 117, 2013.
[10] M. Javorski, Frequency Characteristics of Magnetostriction in
Electrical Steel Related to the Structural Vibrations, Vol. 48, No.
12, pp. 47274734, 2012.
[11] Bohdan T. Kulakowski, John F. Gardner, J. Lowen Shearer,
Dynamic Modeling and Control of Engineering Systems,
Cambrigde University Press, 2007.
[12] Y. Wang, W. Xing, G. Zhang, W. Niu, and D. Han, Experimental
Study of Testing Models for Low Noise Amorphous Alloy Core
Power Transformers, 2010 International Conference on Electrical
and Control Engineering, pp. 37253728, Jun. 2010.

(The Board of Editors received the paper on 09/15/2015, its review was completed on 10/06/2015)

20

Nguyen Thanh Hai, Vo Viet Cuong

IMPROVING STABILITY FOR INDEPENDENT POWER CONTROL

OF DFIG WITH SFOC AND DPC DURING GRID UNBALANCE
Nguyen Thanh Hai2, Vo Viet Cuong1
1

HCMC University of Technology and Education, Vietnam; hai_nguyenthanh2012@yahoo.com.vn

Le Hong Phong High School for The Gifted, Ho Chi Minh City, Vietnam; hoc_vien@yahoo.com.vn

Abstract - This paper presents modified Stator Fed Oriented Control

(SFOC)for Doubly Fed Induction Generator (DFIG) in wind turbines
during grid unbalance,and improves stability by using Notch filter to
eliminate second order harmonic components.The proposed
schemes apply multiple PI controllers with Fuzzy and anti-windup
(PI-F) to obtain commanded rotor currents and also introduce extra
commanded values for rotor currents. Comparison of the proposed
controller with Direct Power Control (DPC) using Notch filters for
improvement during grid voltage unbalance is also included. The
modifications are applied to rotor side converter (RSC) for active and
reactive power controls of wind turbine. The turbine, generator and
control units are also described on MATLAB/SIMULINK. Simulation
results show improved stability of active and reactive powers stator.

Cp (: the aerodynamic efficiency which depends on the

tip speed ratio , and blade pitch angle . According to Betzs
efficiency, the maximum theoretical efficiency is 59.3% [10].
The tip speed ratio is defined as the speed at which the
outer tip of the blade is moving divided by the wind speed.

turb R

(2)

vw

Key words - DFIG; Unbalanced Voltage Dip; PI controller;

Anti-windup; SFOC; Notch Filters; Fuzzy.

1. Introduction
Doubly fed induction generators have been the popular
choice in wind power generation due to the low rating of
power electronic circuit connected to the rotor side of the
generator and the grid [1]. The active and reactive powers
delivered by DFIG can be controlled independently by
Stator Flux oriented Control and Direct Power Control
which are designed for operation with balanced grid
voltage [2]. However, most of the grids experience the
problems of voltage unbalance, which raises the winding
temperature and causes pulsation of torque and power [3].
This paper will investigate the stabilities of active and
reactive powers during transient unbalance of grid voltage
for traditional and modified stator flux oriented control and
direct power control of DFIG. The modifications are
hybrid PI-Fuzzy controller and Sequence Component
controller. The grid unbalance is modeled with a reduction
of 25% of voltage in one phase. Wind speed varies
randomly during the process.
2. Mathematical Model Of Wind Turbine
The model of wind turbine and its formula of shaft
torque, turbine torque, power transferred to generator and
related parameters are presented in this session. Figure
1illustrates the mechanical system of wind turbine which is
often used in large wind turbine systems.
The power extracted from the wind is:
Pturb

1
Avw3 C p ( , )
2

Figure 1. Model of the mechanical part of Wind Turbine [9]

3. Direct Power Control and Stator Flux Oriented

Control Of DFIG
Structure of control method with DPC for DFIG is
shown in Figure 2 and 3. The proposed control structure
with SFOC is shown in Figure 4. Appropriate voltage
vectors for rotor side converter are selected to control
generated active and reactive power in DPC. Converters on
rotor side of DFIG are controlled by stator flux oriented
control to achieve the independent control of active and
reactive powers. Modification of the control system by
using hybrid PI-Fuzzy controller has provided better
performance of the generated powers [5]. However, this is
only verified with balanced grid voltage. To improve
stability of the powers during voltage unbalance, inclusion
of Notch filter has been suggested by [6, 11] and presented
in Figure 3 and 4 to eliminate second order harmonic
components.

dt

(1)

Where: = 31.22 (kg/m3)air density.

A=R2(m2) the cross-sectional area through which the
wind passes.
R(m): length of turbines blades.
vw (m/s):the wind speed normal to the cross-session area A.

Figure 2. DPC for grid-connected DFIG-based wind generator

without Notch filter [11]

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Qs

Lm Vs
3
3
iq r
(vqs id s vd s iqs ) Vs

Ls s Lm
2
2

21

(10)

PI-Fuzzy controllers as shown in Figure 6 are used to

control the errors between the required and actual values of
both the active power and reactive power delivered to the grid
by the generator. The parameters of the PI-Fuzzy are adjusted
by the fuzzy rules to obtain the best output to drive the errors
to zero. The variable parameters of the controllers, which are
fixed in traditional PI controllers, will help to achieve the best
performance of the system. The outputs of these controllers
are commanded values of d-q components of rotor current in
the stator flux oriented reference frame. These commanded
values of currents are used to regulate the RSC for provision
of the rotor phase voltage to DFIG.
Figure 3. The typical configuration of a grid-connected DFIGDPC with Notch Filter

The fuzzy rules for parameters of PI-FUZZY

controllers are presented in Table 1 and Table 2. The rules
are developed by trial and error method. LN, SN, ZE, SP,
and LP represents large negative, small negative, zero,
small positive and large positive respectively. S, M, H
stand for small, medium and high respectively.
Table 1. rule base of Kp [5]
Kp

LN
SN
ZE
SP
LP

LN
H
H
M
M
M

de/dt
SN ZE SP
H H H
M M M
S S S
M M M
H H H

LP
H
H
M
H
H

Table 2. rule base of Ti [5]

Ti

LN
SN
ZE
SP
LP

LN
H
H
H
H
H

SN
H
M
M
M
H

de/dt
ZE SP
H H
M M
S M
M M
H H

LP
H
H
H
H
H

Figure 4. The proposed control scheme for the RSC

of a DFIG using PI+F controller and Notch filters

In both control scheme in Figure 3, Notch filters are

used to eliminate second order harmonic components in
positive and negative sequences of stator voltage. In Figure
4, Notch filters are used with positive sequences of stator
voltage and rotor current.
Figure 5 shows the spatial relationships between the
stationary (,)s reference frame, the rotor (,)r reference
frame rotating at the speed of r, and the dq+ and dq
reference frames rotating at the angular speed of s and s,
respectively. As shown, the d+-axis of the dq+ reference
frame is fixed to the positive sequence stator voltage V+sd+.

Figure 5. Relationships between (,)s, (,)rand

dq+ and dq reference frames [6]

According to Figure 5, the transformations between (,)s,

(,)r and dq+ and dq reference frames are given by
j slip t

I ( ) r e
I dq
I (dq )r e j 2 slip t
r

dqr

I ( )r e

j slip t

( dq )r

j 2 slip t

I (t ) I (t ) I (t )

d dq

j s dq
dt
I d I d e j 2 t [6; 7; 8]
I q I q e j 2 t [6; 7; 8]

Vdq Rs I dq
s

I d I d I d
r

qr

qr

qr

(3)
(4)

Figure 6. PI-Fuzzy controller

(5)

The triangular membership functions of inputs and

outputs of PI-Fuzzy controller are shown in Figure 7, 8:

(6)

slip

(7)

slip

(8)

Active and reactive power of stator:

3
3
Lm
Ps (vd id vq iq ) Vs
iq
2
2
Ls
s

(9)

Figure 7. Membership functions of two inputs of fuzzy bloc

22

Nguyen Thanh Hai, Vo Viet Cuong

Inertia
Inertia of Rotor

Figure 8. Membership functions of two outputs of fuzzy bloc

4. Simulation and Results

Simulation implementation of proposed control method
for 2.3 MW DFIG is carried out, Table 3. The grid voltage
unbalance happens after 35 seconds, the commanded
values of reactive power and active power change at 50s
and 60s respectively. Comparisons of average values of
active and reactive powers in steady state with different
controllers are presented in Table 4 and 5. Both actual
values and percentage of references are shown. Average
electromagnetic torque of the generator is shown in Table
6.The randomly variable wind speed is shown in Figure 9
and Figure 10 is grid unbalance at 35s.

Figure 9. Random variation of wind speed

800
600

Value
159.2 (H)
159.2 (H)
5.096 (mH)
4 (m)
4 (m)
2
100 (rad/s)

400
Vabcs [V]

Symbol
LS
Lr
Lm
RS
Rr
p
S

IWTR

93.22 (kg.m2)
17.106(kg.m2)

The simulation results with different controllers are shown

in figures 11 to 16 for active and reactive output power
respectively. These figures demonstrate the power responses
when voltage unbalance happens and when the commanded
values of powers change under voltage unbalance. Torque
response of the generator is shown in Figure 17.

Table 3. Parameters of DFIG 2.3MW

Parameter
Stator inductance
Rotor inductance
Magnetic inductance
Stator resistance
Rotor resistance
Number of pole pairs
Frequencyof the electric system

Igen

200
0
-200
-400
-600
-800
29.95

29.96

29.97

29.98

29.99

30
30.01
Time [s]

30.02

30.03

30.04

30.05

Figure 10. The grid voltage unbalance happens after 35 seconds

Table 4. Average value of Ps[MW]in steady state for 3 controllers

Grid
Voltage
P sef =2

Mean

Max

Balanced
(11-19s)

2.002
0.1%

Unbanced
(31-49s)

2.001
0.05%

Deviation

DPC WITH NOTCH FILTER

SFOC WITH PI-F & NOTCH

FILTER

Min

Mean

Max

Min

Mean

Max

Min

2.086
4.3%

1.920
-4%

2.002
0.1%

2.08
4.2%

1.919
-4%

2.001
0.1%

2.138
6.9%

1.908
-4.6%

2.113
5.7%

1.904
-4.8%

2.001
0%

2.1
5%

1.915
-4.2%

2.02
1%

2.225
11.3%

1.867
-6.7%

DPC WITHOUT NOTCH

FILTER

( %)

P Psref
Psref

( %)

During the unbalanced voltage, best performances of

active power are observed for DPC with Notch Filter, then
the traditional DPC without Filter. In detail, the lowest

value of PMax for DPC with Notch filters is 5.0% of the

commanded. The highest value of PMin for DPC with Notch
Filter is -4.2% of the commanded value

Table 5. Average value of Qs [MVAR] in steady state for 3 controller

Grid
Voltage
Q sref =1
Balanced
(11-19s)

DPC WITHOUT NOTCH

FILTER
Mean
Max
Min
1.007
1.073
0.928
0.1%
7.3%
-7.2%

Unbanced
(31-49s)

1.051
5.1%

Deviation

( %)

Q Qsref
Qsref

1.09
9%

0.879
-12.1%

Mean
1.00
0%

Max
1.073
7.3%

Min
0.928
-7.2%

SFOC WITH PI-F & NOTCH

FILTER
Mean
Max
Min
1.00
1.114
0.889
0%
11.4%
-11.1%

1.00
0%

1.057
5.7%

0.891
-10.9%

0.997
0.3%

DPC WITH NOTCH FILTER

1.117
11.7%

0.881
-11.9%

( %)

During the unbalanced voltage, best performances of

active power are observed for DPC with Notch Filter, then
the traditional DPC without Filter. In detail, the lowest

value of QMax for DPC with Notch filters is 5.7% of the

commanded. The highest value of QMin for DPC with
Notch Filter is -10.9% of the commanded value.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

23

Table 6. Average value of generators torque in steady state for the 3 controllers

DPC WITHOUT NOTCH

FILTER
Mean
Max
Min
12605
13168
12056
12568
17327
8273.7

Balanced (11-19s)
Unbanced (31-49s)

During the unbalanced voltage, best performances of

active power are observed for DPC with Notch Filter, then
the traditional DPC without Filter. In detail, the lowest value
of TMax for DPC with Notch filters is 15899 (N.m) of the
commanded. The highest value of TMin for DPC with Notch
Filter is 10195 (N.m) of the commanded value.

DPC WITH NOTCH FILTER

Mean
12606
12571
2.3

Ps [MW]

Grid Voltage

5. Discussion

Max
13158
17046

Min
12065
8500

DPC WITH NOTCH FILTER

The results obtained in Table 4 are further

demonstrated in Figure 11. SFOCs active power response
when voltage unbalance happens has higher ripples while
the responses obtained with the two DPC schemes are not
significantly distorted. The responses to change in the
commanded values during the unbalance are good for the
three control scheme as shown in Figure 12. DPC schemes
give faster responses as shown in Figure 13.
DPC WITH NOTCH FILTER

DPC WITHOUT NOTCH FILTER

1.7

1.7

1.7

1.4

1.4

1.4

1.1

1.1

1.1

0.8
49.5

0.8
49.5

50

50.5

2.3

2.3

2.2

2.2

2.2

Time [s]

30
Time [s]

30

40

Time [s]

Figure 11. Active output power of DFIG when voltage

unbalances happen

Ps [M
W
]

DPC WITH NOTCH FILTER

DPC WITHOUT NOTCH FILTER

SFOC WIT H PI+F& NOT CH FILT ER

2.3

2.3

2.3

1.7

1.7

1.7

1.4

1.4

1.4

1.1

1.1

0.8
20

40
Time [s]

60

0.8
20

Time [s]

60

50.5

Qs [MVAR]

DPC WITHOUT NOTCH FILTER

SFOC WITH PI+F&NOTCH FILTER

1.2

1.1

1.1

1.1

0.9

0.9

0.9

0.8
20

30
40
Time [s]

0.8
20

30
40
Time [s]

30

40

Time [s]

Figure 14. Reactive power of DFIG when voltage unbalances happen

2.2

2.2

SFOC WITH PI+F&NOTCH FILTER

2.2

1.9

1.9

1.9

1.6

1.6

1.6

1.3

1.3

1.2

Qs [MVAR]

DPC WITH NOTCH FILTER

.7
20

40
Time [s]

60

DPC WITHOUT NOTCH FILTER

0.7
20

40
Time [s]

60

.07
20

40

Time [s]

60

Figure 15. Reactive power of DFIG during transient states

2.2

2.2

SFOC WITH PI+F&NOTCH FILTER

2.2

1.9

1.9

1.9

1.6

1.6

1.6

1.3

1.3

1.2

DPC WITH NOTCH FILTER

.7
49. 5

50
Time [s]

50.5

DPC WITHOUT NOTCH FILTER

0.7
49. 5

50
Time [s]

50.5

.07
49.5

50

Time [s]

50.5

DPC WITHOUT NOTCH FILTER

FOC WITH PI+F& NOTCH FILTER

20

20

20

18

18

18

15

15

15

12

12

12

0
20

0
20

0.8
20

40
60
Time [s]

80

40
60
Time [s]

80

0
20

40
60
Time [s]

80

Figure 17. Torque of DFIG

1.1

40

50
Time [s]

Figure 16. Dynamic responses of DFIGs reactive power during

the change of commanded value

1.8
20

40

0.8
49.5

1.2

0.8
20

Te [ KN.m]

1.8
20

40

50.5

1.2

DPC WITH NOTCH FILTER

30

50
Time [s]

DPC WITH NOTCH FILTER

SFOC WIT H PI+F& NOTCH FILT ER

2.3

SFOC WITH PI+F& NOTCH FILTER

2.3

Figure 13. Dynamic responses of DFIGs active output power

during the change of commanded value

Qs [MVAR]

Steady state responses of reactive power are also very

good when Notch filters are included. The deviations are
0% and 0.3% respectively for DPC and SFOC. The
deviation is much higher without Notch filter during the
voltage unbalance as shown in Table 5. There is no
significant difference observed between the responses
during the voltage balance, with or without Notch filters.
The fluctuation is observed to be smallest for DPC with
Notch filter. SFOC however gives smallest torque
variation during voltage unbalance as shown in Table 6.

1.8
20

DPC WITHOUT NOTCH FILTER

2.3

Time [s]

DPC has shown good steady state of active power

responses during the voltage balance and unbalance as
shown in Table 4. The deviation of the mean value of active
power from the reference value is almost zero percent with
the inclusion of Notch filter. SFOC also gives good
performance with small deviation (about 1%). The
fluctuation of active power is smallest for DPC with Notch
filter during the unbalance.

SFOC WITH PI-F &

NOTCH FILTER
Mean
Max
Min
12603 13982
11342
12522 15899
10195

40

60

Time [s]

Figure 12. Active output power of DFIG during the transient states

Higher ripples are also observed in reactive power

responses of SFOC when voltage unbalance occurs as
shown in Figure 14. The observation is consistent with
statistics presented in Table 5. Reactive powers in the three
control scheme follow the commanded values under the

24

Nguyen Thanh Hai, Vo Viet Cuong

condition of voltage unbalance as shown in Figure 15. Time

responses of reactive power in DPC control schemes are also
less than those of SFOC as shown in Figure 16.
Torque responses observed in Figure 17 are also
consistent with the statistics shown in Table 6.

edition, USA, 2001.

[3]

Muljadi, E., Yildirim, D., Batan, T., and Butterfield, C.P.,

Understand the unbalanced-voltage problem in wind turbine
generation, Proceeding of IEEE Industry Application Conference,
Phoenix, USA, 1999, pp.1359-1365.

[4]

Baggu, M. M.; Advanced control techniques for doubly fed

induction generator based wind turbine converters to improve low
voltage ride- throught during system imbalances, PhD Thesis,
Missouri University of Science and Technology, 2009.
Pham-Dinh, T., Pham-Trung, H., Le-Thanh, H., PI-Fuzzy
Controller for Doubly Fed Induction Generator Wind Turbine,
Proceedings of ASEAN Symposium on Automatic Control ASAC
2011, Vietnam, 2011, pp.79 81.

6. Conclusion
The proposed SFOC scheme for DFIG with the
inclusion of PI-Fuzzy controllers and Notch filters has
improved the stability of independent control of active and
reactive power during grid voltage unbalance. The
responses of active and reactive power are compared with
a traditional DPC and modified DPC using Notch filters to
increase the stability. The observations are made during the
occurrence of voltage dip in one phase, transient states as
well steady states of the powers under unbalanced
condition. In all the observations, the independent control
of the powers is maintained for the proposed scheme.
However, high fluctuations in active and reactive
powers are present in the responses obtained with the
proposed scheme although lower ripples are observed for
generators torque.
Experimental verification of the new control scheme
should be carried out to validate the results obtained with
simulation.
REFERENCES
[1]
[2]

Ackermann, T., Wind power in power systems, John Wiley and Sons,
USA, 2003.
Leonhard, W., Control of electric drives, Springer-Verlag, 3rd

[5]

[6]

Phan, V. T., Lee, H. H., Chun, T. W.; An Effective rotor current controller
for unbalanced stand alone DFIG systems in the rotor reference frame,
Journal of Power electrionics, Vol.10, No.6, 2010, pp 194-202.

[7]

L. Xu, Y. Wang, Dynamic modeling and control of DFIG based

wind turbines under unbalanced network conditions, IEEE Trans.
Power Syst. 22 (1) (2007) 314323.

[8]

A. Peterson, L. Harnefors, T. Thiringer, Comparison between

stator-flux and grid flux oriented rotor current control of doubly-fed
induction generators, The 35th Annual IEEE Power Electronics
Specialist Conference, vol. 1, 2025 June,2004, pp. 482486.

[9]

Sorensen, P.; Hansen, D.A.; Christensen, P.; Mieritz, M.; Bech, J.;
Bak-Jensen, B.; Nielsen, H.; Simulation and Verification of
Transient Events in Large Wind Power Installation, Project Report,
Ris National Laboratory, Roskilde, Norway; 2003.
[10] Masters, M. G. Renewable and Efficient Electric Power Systems,
John Wiley and Sons, Inc., Publication; 2004.
[11] Jia-bing HU, Yi-kang HE; Modeling and enhanced control of DFIG
under unbalanced grid voltage conditions, Electric Power Systems
Research 79(2009); pp 273-281.
[12] Hai Nguyen-Thanh; Improved Control of DFIG Systems under
Unbalanced Voltage Dip for Torque Stability Using PI-Fuzzy
Controller; International Journal of Electrical Energy, Vol. 2, No.
4, December 2014; pp. 300-307, USA.

(The Board of Editors received the paper on 15/05/2015, its review was completed on 05/07/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

25

EFFECTS OF ETHYLENE GLYCOL, TEMPERATURE AND pH ON THE SIZE

AND THE DISPERSION OF THE PLATINUM NANOPARTICLE CATALYST
ON CARBON SUPPORT
Hoang Anh Huy 1, Ho Thi Thanh Van2
1

Ho Chi Minh City University of Technology; 513050187@hcmut.edu.vn

Ho Chi Minh City University of Natural Resources and Environment; httvan@hcmunre.edu.vn

Abstract - Controlling the size and the dispersion of Pt nanoparticle

structures seems to be one of the most important goals in improving
the electrochemical activity and the durability of an electrocatalyst as a
Platinum nanoparticle catalyst for fuel cells. In this paper, we introduce
a simple method to address the effects of temperature, pH and
ethylene glycol enhancers (EG) on the morphology, the size and the
distribution of the Pt nanoparticle catalyst in the process of fabricating
Pt nanoparticles on carbon support (Pt NPs/C). By means of a facile
method, the Pt NPs/C catalyst was synthesized with and without the
use of EG at room temperature and 60oC in pH = 6.5 and 11 solutions.
The morphology, the size and the dispersion of the platinum
nanoparticle catalyst on carbon support were characterized via X-ray
diffraction (XRD) and transmissionelectron microscopy (TEM). We
have found out that the size, the morphology and the dispersion of Pt
nanoparticles on carbon were strongly affected by adjusting the
temperature, pH and that the presence of ethylene glycol could
enhance the properties of the Pt/C catalyst for fuel cells application.
Key words - Pt/C; Platinum catalyst; Pt nanoparticles; Pt catalyst;
PtNPs/C .

1. Introduction
Pt is catalytically active at room temperature under
electro-oxidation reactions of interest to fuel cell applications.
It is well known that the metal catalytic activity is strongly
dependent on the particle shape, size and the particle size
distribution. The conventional preparation techniques based
on wet impregnation and the chemical reduction of metal
precursors do not provide satisfactory control of the particles
shape and size as well as the distribution of Pt particles on
carbon support [1]. The synthesis of highly dispersed
supporting platinum with a uniform nanoparticle size still
remains a challenge, especially for high metal loading. The
conventional methods for the synthesis of Pt electrocatalyst
are mainly impregnation and colloid methods such as the
sulfite complex route and the colloidal route, the impregnation
method usually produces NPs with large average particle
sizes and broad size distribution while the colloidal route
produces well-homogenized ultrafine Pt electrocatalysts,
however, the complexity of the latter hinders its utilization [2].
Many investigators have contributed many efforts to search
for alternative routes.
Recently, there has been an attempt to develop
alternative synthesis methods based on microemulsions
[3], sonochemistry [4a] and microwave irradiation [4b,c],
all of which are in principle more conducive to the
production of colloids and clusters on the nanoscale,
resulting in greater uniformity.
In this paper, a simple procedure for preparing Pt metal
nanoparticles supported on carbon is reported. The uniform
platinum nanoparticles are supported on carbon with Pt
loading up to 40 wt%, which is a standard amount in order

to obtain higher dispersion and smaller crystallites [5]. This

study lays the foundation for further inspections with a
desire to create the best possible Pt/C catalyst. Through this
work, we have found that the optimal preparation is simple
and fast but it is capable of controlling the particle size and
the distribution of Pt particles on the carbon support and
could be predicted to enhance the activity of the catalyst
for fuel cell applications.
2. Experimental section
2.1. Preparation of Pt/C catalysts
Vulcan XC-72R carbon with the particle size of 60
nm using as a support was purchased from Fuel Cell Store
(USA). All the chemicals were of analytical grade;
Hydrogen hexachloroplatinate(IV) hexahydrate, 99.9%,
(trace metal basis), 38 to 40% Pt (H2PtCl6.6H2O), ethylene
glycol (EG), acetone branded Acros (Belgium), sodium
borohydride (NaBH4), nitric acid (HNO3) (65%-68%)
(China) were used.
Pt/C catalysts particles were synthesized by means of
the following route: first, Vulcan XC-72R carbon powder
was treated to clean the contaminant in the commercial
carbon. (Figure 1) For example,
0.5 g carbon was
dispersed in a round bottom flask with 500 mL of the 5%
HNO 3 solution, the mixture was refluxed for 16 hours at
105 C [6]. The treated carbons were centrifuged with 4500
rpm for several times with 5 minutes each for washing
with the de-ionized (DI) water and acetone (15mL H2O or
Acetone for each centrifuge tubes), then were dried at 105
C in an oven for 10 hours. Second, the Pt particles were
synthesized and dispersed on the carbon supports via the
following process: 50 mg of the treated carbon was
dispersed into the solvent (DI water with and without using
EG), 3.39 mL H2PtCl6.6H2O with concentration 0.05M
into the mixed precursor. The pH of this mixture was
adjusted to 6.5 and 11 by dropping a wise addition of
NaOH 0.1N solution. The mixture was stirred for 5 minutes
and ultrasonicated for 15 minutes at room temperature.
Then an excess amount of reduction agent 6.84 mL NaBH4
0.05M was added and the mixture was stirred by using a
magnetic bar under atmospheric pressure at room
temperature or 60 C for 2 hours (Figure 2).
Finally, the synthesized catalyst particles Pt/C were
washed by DI water, centrifuged and dried for 12 hours at
100 C [7]. All samples were shown in Table 1.
The samples were characterized by using X-ray
diffraction, transmissionelectron microscopy (TEM) in order
to assess the effects of ethylene glycol and the temperature on

26

Hoang Anh Huy, Ho Thi Thanh Van

the morphology, the size and dispersion of platinum

nanoparticles catalyst on carbon support for fuel cells.

Figure 1. Preparation procedure of pre-treated Vulcan XC-72R.

- B = the full width half maximum of the peak

- K = the Scherrer constant; depends on how the width
is determined, the shape of the crystal, and the size
distribution.
- = the wavelength of the radiation used to collect the
data.
We used the Pt (111) plane to determine the average
crystallite size. The FWHM are calculated from the (111)
peak by means of Originlab software. The value of K is 0.9
due to the structures Pt is face-centered cubic and the
wavelength used to be Cu = 1.54 .
Table 2. Average crystallite size calculated from the Pt (111)
using the Scherrer equation

Sample

Average crystallite size (nm)

Pt/C-30_11

7.14

Pt/C-EG-30_11

3.84

Figure 2. The preparation of 40 wt% Pt/C with different

conditions
(pH, temperature, with and without EG).

2.2. XRD and TEM Analysis

X-ray powder diffraction (XRD) patterns were
recorded by using a Cu K radiation source on a D8
Advance Bruker powder diffractometer (University of
Technology - VNU HCM City). The transmission electron
microscope (TEM) were taken by JEM-1400 (JEOL,
Japan), (University of Technology - VNU HCM City).
3. Results and discussion
Table 1 shows the samples of 40 wt% Pt/C catalysts
prepared with various conditions. The effect of such
parameters as EG, temperature as well as pH on the size
and distribution of Pt NPs on carbon were examined. The
samples were synthesized with and without using EG,
carried out at room temperature (30 C) and 60 C. The
influence of pH value was also examined at 6.5 and 11.
Table 1. The samples of 40 wt% Pt/C catalysts
were prepared with various conditions.

Ethylene
glycol (EG)

Temperature pH
(C)

Catalysts

30

11

Pt/C-30_11

EG

30

11

Pt/C-EG-30_11

EG

30

6.5

Pt/C-EG-30_6.5

EG

60

11

Pt/C-EG-60_11

60

11

Pt/C-60_11

3.1. Effect of the presence Ethylene Glycol as a stabilizer

for Pt/C preparation
Scherrer equation:
K
L cos
where L = average crystal size (angstrom or nm)
B 2

Figure 3. X-ray diffraction (XRD) patterns of Pt-30_11

and Pt-EG-30_11 catalysts.

The 40 wt% Pt/C samples are synthesized with or

without the presence of EG at room temperature at pH=11.
X-ray diffraction of these samples is shown in Figure 3. It
indicates that all the broad diffraction peaks of the XRD
patterns at 2 = 39.6, 47.4, 67.1, corresponding to the
reflections (111), (200), (220), respectively, which are
consistent with the face centered cubic (fcc) structure of
platinum, can be assigned to (JCPDS Card 04-0802), thus
demonstrating the presence of crystalline Pt [8]. In
addition, a broad peak at 2 25 was observed but not
clearly due to the fact that the (002) plane of the hexagonal
structure of the carbon support (Vulcan XC-72R) is
amorphous carbon with small regions of graphitic
properties [9]. More importantly, the Pt/C sample with the
presence of EG shows a wider peak than the sample
without using EG in the preparation process, suggesting
that the Pt particles size of sample using EG is smaller than
the one without using EG (3.84 nm compare to 7.14 nm).
(Table 2) The average crystalline sizes of the Pt catalysts
were calculated by the Scherrer formula based on the
diffraction peaks of the Pt (111) plane.
This result is supported by TEM measurement shown
in Figure 4.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Figure 4. (a), (b) are the TEM of Pt-30_11 catalyst;

(c), (d) are the TEM of Pt-EG-30_11 catalyst.

The Pt/C-EG-30_11 catalyst prepared using NaBH4 as

the reducing agent in EG solution at pH=11 has shown that
its particles size is smaller than the size of Pt/C-30_11
(Figure 4. (d) compared to (b)). The presence of ethylene
glycol (EG) supported Pt nanoparticles not only have a
narrow size, but also distribute uniformly on carbon support
(Figure 4. (c) compared to (a)). This has been reported that
the reaction with the attendance of EG and NaBH4 in a
solution will form a complex reducing agent, and EG has
performed roles both as a reducing agent for Pt reduction and
a stabilizing for the reduced Pt nanoparticles [10].

27

at 60C, pH=11 (denoting Pt-EG-60_11). We can clearly

observe that all the samples show the reflections (111),
(200), (220), respectively. However, when the temperature
increases from room temperature to 60C, there is a
significant difference in the crystallinity and the particle
size of Pt on carbon. The Pt-EG-60_11 shows high
crystallinity and a large particle size compared to Pt-EG30_11. For example, the average crystallite size of Pt-EG30_11 is about 3.84 nm and the average crystallite size of
Pt-EG-60_11 is about 5.85 nm (estimated from the
Scherrer formul at Pt (111) peak). (Table 3). This
observation is then confirmed by TEM measurement
(Figure 6). The sample Pt-EG-30_11 shows the good
distribution as well as the small particle size compared to
Pt-EG-60_11, indicates that the temperature strongly
affected the distribution and the size of Pt on carbon. This
phenomena could be explained by the fact that the degree
of supersaturation of a solution decreases when the reaction
temperature increases. At a lower temperature, the
formation of crystal nuclei proceeds more rapidly than the
growth. Therefore smaller Pt particles were obtained at
30C compared to those obtained at 60C [11].

3.2. Temperature effects on Pt/C preparation

Table 3. Average crystallite size calculated from the Pt (111)
using the Scherrer equation

Sample

Average crystallite size (nm)

Pt/C-EG-30_11

3.84

Pt/C-EG-60_11

5.85
Figure 6. (a), (b) are the TEM of Pt-EG-30_11 catalyst;
(c), (d) are the TEM of Pt-EG-60_11 catalyst.

3.3. Effect of the solution pH on Pt/C preparation

Table 4. Average crystallite size calculated from the Pt (111)
using the Scherrer equation

Figure 5. X-ray diffraction (XRD) patterns of Pt-EG-30_11

and Pt-EG-60_11 catalysts.

The effect of temperature is studied by preparing the

sample 40 wt% Pt/C using EG at room temperature, pH=11
(denoting Pt-EG-30_11) and 40 wt% Pt/C sample using EG

Sample

Average crystallite size (nm)

Pt/C-EG-30_11

3.84

Pt/C-EG-30_6.5

7.72

The Pt/C-EG-30_6.5 sample is prepared in the same

procedure as Pt/C-EG-30_11, but with a pH=6.5 solution
differentiating from the pH=11 of Pt/C-EG-30_11 sample.
Compared to Pt/C-EG-30_6.5, the Pt/C-EG-30_11 catalyst
had highly dispersed Pt nanoparticles on the surface of the
carbon support. Bonnemann et al. reported that Pt
nanoparticles are stabilized via the electrosteric repulsion
between the anionic surface of the Pt nanoparticle and the
stabilizer [12]. In the acidic solution, a large number of H+
ions interact with negatively charged Pt particles resulting

28

Hoang Anh Huy, Ho Thi Thanh Van

in the destruction of electrosteric repulsion and leading to

the growth of Pt nanoparticles. In the basic solution, almost
no species would directly interact with negatively charged
Pt nanoparticles, which implied that the electrosteric
stabilization is unbroken [13, 14]. A similar feature has
also been observed in the synthesis of Pt-based metal
nanoparticles using EG as a reducing agent [15]. In our
case, the size of Pt nanoparticles of Pt/C-EG-30_6.5
sample is about 7.72 nm (Table 4) (estimated from the
Scherrer formul at Pt (111) peak). The differences in Pt
particles between Pt/C-EG-30_6.5 and Pt/C-EG-30_11 can
be explained based on the effect of the electrosteric
repulsion. Under high pH conditions, only minor
interaction occurred between H+ ions and stabilizer anions,
yet the stabilizer strongly interacted with the reduced Pt
nanoparticles. Therefore, the growth of Pt particles was
significantly restrained, leading to the formation of Pt
nanoparticles with a smaller size in the Pt/C-EG-30_11
than in the Pt/C-EG-30_6.5 catalyst (Figure 8).

4. Conclusions
The effect of ethylene glycol and temperature on the
size and the dispersion of platinum nanoparticles catalyst
on carbon support has been examined in this research. We
have found out that the temperature, EG agent as well as
pH parameter strongly affected the particle size and the
distribution of Pt on carbon support. The presence of EG
functioning as a weak reducing agent and a stabilizer could
enhance the distribution and make smaller Pt sizes
compared to the sample without using EG. In addition, the
effect of temperature on the Pt/C preparation was studied
at room temperature and 60C. We also found that when
the temperature increases from room temperature to 60C,
there is a significant difference in the crystallinity and the
particle size of Pt on carbon due to the degree of
supersaturation of a solution that decreases with an
increase in the reaction temperature. Finally, the effect of
pH parameter on Pt/C preparation has also been examined
in this work. The experimental results indicate that in the
acidic solution (low pH value), a large number of H+ ions
interact with negatively charged Pt particles, which result
in the destruction of electrosteric repulsion, leading to the
growth of Pt nanoparticles. However, only minor
interaction occurred between H+ ions and stabilizer anions
in high pH conditions. The results of this work showed the
way to control the size and the distribution of Pt catalyst on
carbon support that can be used to enhance the activity of
the Pt/C catalyst with high loading for fuel cell
applications.
Acknowledgment: This work was supported by Ho Chi
Minh City University of Technology, Ho Chi Minh City
University of Natural Resources and Environment.
REFERENCES

Figure 7. X-ray diffraction (XRD) patterns of Pt-EG-30_11

and Pt-EG-30_6.5 catalysts

[1]

Armadi I. S., Wang Z. L., Green T. C., Henglein A., El-Sayed M. A.

- Shape-Controlled Synthesis of Colloidal Platinum Nanoparticles,
Science 272 (1996) 19241926.

[2]

Singh R. N., Awasthi R. and Sharma C.S. Review: An Overview

of Recent Development of PlatinumBased Cathode Materials for
Direct Methanol Fuel Cells, Int. J. Electrochem. Sci. 9 (2014) 5607
5639.

[3]

Liu Z. L., Lee J. Y., Han M., Chen W. X. and Gan L. M. - Synthesis
and Characterization of Ptru/C Catalysts from Microemulsions and
Emulsions, J. Mater. Chem. 12 (2002) 24532458.
a. Fujimoto T., Teraushi S., Umehara H., Kojima I. and Henderson W.
- Sonochemical Preparation of Single-Dispersion Metal Nanoparticles
from Metal Salts, Chem. Mater. 13 (2001) 10571060.

[4]

b. Tu W. and Liu H. - Continuous Synthesis of Colloidal Metal

Nanoclusters by Microwave Irradiation, Chem. Mater. 12 (2000)
564567.

[5]

[6]

Figure 8. (a), (b) are the TEM of Pt-EG-30_11 catalyst;

(c), (d) are the TEM of Pt-EG-30_6.5 catalyst

c. Tu W. and Liu H. - Rapid Synthesis of Nanoscale Colloidal Metal

Clusters By Microwave Irradiation, J. Mater. Chem. 10 (2000)
22072211.
Coutanceau C., Baranton S. and Napporn T. W. - Platinum Fuel Cell
Nanoparticle Syntheses: Effect on Morphology, Structure and
Electrocatalytic Behavior, Nanotechnology and Nanomaterials
(2012) 403426.
Senthil Kumar S. M., Hidyatai N., Herrero J. S., Irusta S., Scott K. Efficient Tuning of the Pt Nano-Particle Mono Dispersion on
Vulcan XC-72R by Selective Pre-Treatment and Electrochemical
Evaluation of Hydrogen Oxidation and Oxygen Reduction
Reactions, International Journal of Hydrogen Energy 36 (2011)
54535465.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1
[7]

29

Joo J. B., Kim P., Kim W., Kim Y. And Yi J. - Effect of the
Preparation Conditions of Carbon-Supported Pt Catalyst on PEMFC
Performance, J Appl Electrochem 39 (2009) 135140.
Shah M. A. - Growth of Uniform Nanoparticles of Platinum by an
Economical Approach at Relatively Low Temperature, Scientia
Iranica F 19 (2012) 964966.

[12] Bonnemann H., Braun G., Brijoux W., Brinkmann R., Schulze
Tilling A., Seevogel K., Siepen K. - Nanoscale Colloidal Metals and
Alloys Stabilized by Solvents and Surfactants - Preparation and Use
as Catalyst Precursors, Journal of Organometallic Chemistry 520
(1996) 143162.

Zhou W. J., Li W. Z., Song S. Q., Zhou Z. H., Jiang L. H., Sun G.
Q., Xin Q., Poulianitis K., Kontou S., Tsiakaras P. - Bi- and TriMetallic Pt-Based Anode Catalysts for Direct Ethanol Fuel Cells,
Journal of Power Sources 131 (2004) 217223.

[13] Hui C. L., Li X. G., Hsing I. -M. - Well-Dispersed SurfactantStabilized Pt/C Nanocatalysts for Fuel Cell Application: Dispersion
Control and Surfactant Removal, Electrochimica Acta 51 (2005)
711719.

[10] Kim P., Joo J. B., Kim W., Kim J., Song I. K., Yi J. - NaBH4Assisted Ethylene Glycol Reduction For Preparation of CarbonSupported Pt Catalyst for Methanol Electrooxidation, Journal of
Power Sources 160 (2006) 987990.
[11] Tian J. H., Wang F. B., Shan ZH. Q., Wang R. J. and Zhang J. Y. Effect of Preparation Conditions of Pt/C Catalysts on Oxygen
Electrode Performance in Proton Exchange Membrane Fuel Cells,

[14] Li X., Hsing I. -M. - The Effect of the Pt Deposition Method and the
Support on Pt Dispersion on Carbon Nanotubes, Electrochimica
Acta 51 (2006) 52505258.

[8]

[9]

Journal of Applied Electrochemistry 34 (2004) 461467.

[15] Zhou Z., Zhou W., Wang S., Wang G., Jiang L., Li H., Sun G., Xin
Q. - Preparation of Highly Active 40wt.% Pt/C Cathode
Electrocatalysts for DMFC via Different Routes, Catalysis Today
9395 (2004) 523528.

(The Board of Editors received the paper on 08/17/2015, its review was completed on 10/16/2015)

30

Dinh Thanh Khan, Nguyen Quy Tuan

A STUDY OF THE CURVATURE OF A THICK ALN FILM GROWN ON A

TRENCH-PATTERNED -Al2O3 TEMPLATE USING X-RAY DIFFRACTION
Dinh Thanh Khan*, Nguyen Quy Tuan
The University of Danang, University of Education; *khannabo86@gmail.com
Abstract - In this article a method using X-ray diffraction for
determining the crystallographic curvature of a thick AlN crystalline
film epitaxially grown on a periodically trench-patterned -Al2O3
template by the hydride vapor phase epitaxy method was studied.
A series of X-ray rocking curve measurements for AlN 0002
reflection was taken at different positions across the surface of the

2. Experiment

thick AlN epitaxial film along the [1100] direction. We introduced a

model for determining the crystallographic curvature and the
curvature radius from X-ray diffraction results. The results clearly
demonstrate that the crystallographic curvature of the film is
convex along the [1100] direction and the radius of crystallographic
curvature of the thick AlN film is estimated to be 3.1 m.
Key words - Curvature; X-ray diffraction; AlN film; trench-patterned
template; strain.

1. Introduction
Aluminum nitride (AlN) has attracted a significant
amount of research interest in undeveloped fields such as
deep ultraviolet (DUV) light emitting diodes, lasers, high
frequency electronic devices because of its wide
bandgap energy of 6.2 eV [1-3]. AlN can alloy with
gallium nitride (GaN) to form compounds such as AlxGa1xN (x = 0 1), which have potential applications in short
wavelength optoelectronic devices. In addition, its
properties such as high hardness, high thermal conductivity
[4] and resistance to high temperatures and caustic
chemicals [5] combined with a reasonable thermal match
with Si and GaAs make AlN an attractive material for
electronic packaging applications. However, due to
difficulties of growing large-area bulk A1N crystals, the
heteroepitaxial growth of thick AlN films on substrates
such as -Al2O3 and 6H-SiC via hydride vapor phase
epitaxy (HVPE) in combination with metalorganic vapor
phase epitaxy (MOVPE) is one of the more promising
techniques being evaluated [6-8]. Unfortunately, lattice
and thermal mismatches between AlN and its substrates are
usually a major impediment to growing high quality
crystalline AlN films because they induce the generation of
crystallographic
defects,
residual
strain
and
crystallographic curvature in such films during growth and
cooling processes [9,10].
Several methods such as double crystal diffraction
topography and two beam laser reflection techniques have
been utilized in order to determine the crystallographic
curvature of films epitaxially grown on substrates [11-13].
However, the experimental setup of these methods are
complex because they require specific devices and
configurations. In this study, we introduce a new method
for determining the crystallographic curvature of the
epitaxial films using rocking curve (RC) measurements of
X-ray diffraction (XRD). The experimental setup of this
method is available in any X-ray diffractometer.

Figure 1. Schematic diagram of the sample fabrication process:

First, (a) A trench-patterned -Al2O3 template was fabricated
from an -Al2O3 substrate using the reactive ion etching
technique; Then, (b) a thick AlN film was grown on the trenchpatterned -Al2O3 template using the HVPE method. (c) Crosssectional SEM image of the thick AlN film grown on the trenchpatterned -Al2O3 template. The white dash line indicates the
interface between the HVPE-grown AlN film and trenchpatterned -Al2O3 template

The sample fabrication process is shown in Figure 1. The

axes of X, Y and Z represent the directions of [1100], [1120]
and [0001], respectively. First, as shown in Figure 1(a), a
trench-patterned template was created on an -Al2O3
substrate using the reactive ion etching technique. The

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

trench direction was [1100] and the pattern was periodic in

the [1120] direction. Trench depth was set at 1.5 m while
terrace and trench widths were both set at 2.0 m. Then, as
shown in Figure 1(b), an 8.6-m-thick AlN film was grown
on this template using a low-pressure HVPE system with
infrared lamps as heaters. The growth pressure was 30 Torr
and the growth temperature range was about 1400 1500C.
NH3, Al, and HCl were used as source materials. N2 and H2
were used as carrier gases. A source of AlCl3 was formed by
the reaction of Al and HCl at 550C in the source zone of the
reactor. AlCl3 was then reacted with NH3 in the growth zone
producing AlN layers on the trench-patterned -Al2O3
template. Figure 1(c) shows a cross-sectional scanning
electron microscopy (SEM) image of the thick AlN film
grown on the trench-patterned -Al2O3 template. Here, it can
be observed that voids form tunnels running along the X
direction over the trenches that were periodically arranged
in the Y direction at 4-m intervals.

31

From the result in Figure 3(a), the incident angle at the

maximum intensity in each RC profile was plotted as a
function of the measured position. The result was shown in
Figure 3(b). It is clearly observed that the incident angle
linearly changes with the position along the X direction. As
schematically shown in Figure 4(a), it can be determined
that the curvature of the lattice planes in the AlN film is
convex when an -incident angle increase is observed by
shifting the X-ray beam in the direction of X. In contrast,
as shown in Figure 4(b), a concave curvature exists when
an -incident angle decrease is observed by shifting the
X-ray beam in the direction of X. An inspection of the
result shown in Figure 3(b) clarifies that the former is the
case for the present AlN film. The convex film curvature
in the [1100] direction is due to the presence of the
compressive strain in this direction [10, 14]. According to
the model shown in Figure 5, the radius of curvature R can
be expressed in the form:
R

L
.

Figure 2. Schematic diagram of XRD from AlN (0002) planes.

K0 and K are the incident and diffracted X-ray beams, respectively.
Red circles indicate sampling positions for RC measurements. is
incident angle of X-ray beam to the film surface

Figure 2 shows schematic diagram of XRD from AlN

(0002) planes. In order to clarify the film curvature in the
X direction, the X-ray incidence was selected so that the
diffraction plane can be determined by the incident and
diffracted vectors can be parallel to this direction. The film
curvature in the X direction was clarified by taking a series
of AlN 0002 RC measurements at different positions across
the film surface along this direction with regular steps of
1 mm. The X-ray beam size was 0.1 mm 0.1 mm. The
X-ray wavelength and penetration depth were 0.15418 nm
and 12.6 m, respectively.
3. Results and discussion
Figure 3(a) shows the result of a series of 0002 RC
measurements taken at different positions with 1-mm steps
in the range of 4 mm along the X direction. It should be
noted that each RC profile consists of a single peak forming
a fairly uniform distribution along the X axis. This
indicates that the crystalline morphology is fairly
homogeneous in the [1100] direction. This homogeneity
leads to the remarkable curvature along the [1100] direction
as a form of macroscopic strain relaxation in this direction.

Figure 3. (a) A series of AlN 0002 RC measurements taken at

different positions with 1-mm steps in the range of 4 mm along
the X direction. (b) Projection of the maximum peak in each RC
profile on the (, X) plane: is the difference between
incident angles of X-ray beam at the positions X = 2 and 2 mm

32

Dinh Thanh Khan, Nguyen Quy Tuan

positions X = 2 and 2 mm, i.e., 0.07 as determined by the

result shown in Figure 3(b). As a result, the curvature
radius R is estimated to be 3.1 m.
4. Conclusion
The crystallographic curvature of the thick AlN film
grown on the trench-patterned -Al2O 3 template was
determined by performing a series of X-ray rocking curve
measurements for AlN 0002 reflection at different
positions across the AlN film surface. The results clarify
that the AlN film is convexly bent along the [1100]
direction. The convex curvature of the AlN film is due to
the presence of compressive strain in this direction.
Acknowledgement
This work was completed with financial support from
The University of Danang.
REFERENCES

Figure 4. Schematic for determining the curvature of the thick

AlN film in the X direction

[1]

Y. Taniyasu, M. Kasu, and T. Makimoto, An aluminium nitride

light-emitting diode with a wavelength of 210 nanometres, Nature
(London), 441, 2006, 325-328.

[2]

H. Hirayama, S. Fujikawa, N. Noguchi, J. Norimatsu, T. Takano,K.

Tsubaki, and N. Kamata, 222-282 nm AlGaN and InAlGaN-based
deep-UV LEDs fabricated on high-quality AlN on sapphire, Phys.
Status Solidi A, 206, 2009, 1176-1182.

[3]

R. McClintock, A. Yasan, K. Mayes, D. Shiell, S. R. Darvish, P.

Kung, and M. Razeghi, High quantum efficiency AlGaN solarblind p-i-n photodiodes, Appl. Phys. Lett., 84, 2004, 1248-1250.
L. M. Sheppard, Aluminum nitride: A versatile but challenging
material, Am. Ceram. Soc. Bull., 69, 1990, 1801-1812.

[4]
[5]
[6]

S. Strite, and H. Morkoc, GaN, AlN, and InN: A review, J. Vac.

Sci. Technol. B, 10, 1992, 1237-1266.
Y. Katagiri, S. Kishino, K. Okuura, H. Miyake, K. Hiramatu, Lowpressure HVPE growth of crack-free thick AlN on a trench-patterned
AlN template, J. Cryst. Growth, 311, 2009, 2831-2833.

[7]

S. A. Newman, D. S. Kamber, T. J. Baker, Y. Wu, F. Wu, Z. Chen,

S. Namakura, J. S. Speck, and S. P. DenBaars, Lateral epitaxial
overgrowth of (0001) AlN on patterned sapphire using hydride
vapor phase epitaxy, Appl. Phys. Lett., 94, 2009, 121906.

[8]

M. Imura, K. Nakano, N. Fujimoto, N. Okada, K. Balakrishnan, M.

Iwaya, S. Kamiyama, H. Amano, I. Akasaki, T. Noro, T. Takagi, and
A. Bandoh, High-temperature metal-organic vapor phase epitaxial
growth of AlN on sapphire by multi transition growth mode method
varying V/III ratio, Jpn. J. Appl. Phys., 45, 2006, 86398643.

[9]

L. W. Sang, Z. X. Qin, H. Fang, T. Dai, Z. J. Yang, B. Shen, G. Y.

Zhang, X. P. Zhang, J. Xu, and D. P. Yu, Reduction in threading
dislocation densities in AlN epilayer by introducing a pulsed atomiclayer epitaxial buffer layer, Appl. Phys. Lett., 93, 2008, 122104.

[10] K. Hiramatsu, T. Detchprom, and I. Akasaki, Relaxation

mechanism of thermal strain in heterostructure of GaN grown on
sapphire by vapor phase epitaxy, Jpn. J. Appl. Phys., 32, 1993,
1528-1533.
[11] C.L. Kuo, P.E. Vanier, and J.C. Bilello, Residual strains in
amorphous silicon films measured by x-ray double crystal
topography, J. Appl. Phys., 55, 1984, 375-377.

Figure 5. Schematic for determining the curvature radius of the

thick AlN film in the X direction

Here, L is the length probed by the X-ray beam on the

film surface along the [1100] direction, i.e., 4 mm. is the
difference between incident angles of X-ray beam at the

[12] J. Tao, L.H. Lee, and J.C. Bilello, Non-Destructive Evaluation of

Residual Stresses in Thin Films Via X-Ray.Diffraction Topography
Methods, J. Electronic Mater., 20, 1991, 819-825
[13] J.F. Geisz, T.F. Kuech, M.G. Lagally, F. Cardone and R. M.
Potemski, Film stress of sputtered W/C multilayers and strain
relaxation upon annealing, J. Appl. Phys. 75 (1994)
[14] G. H. Olsen, and M. Ettenberg, Calculated stresses in multilayered
heteroepitaxial structures, J. Appl. Phys., 48, 1977, 2543-2547.

(The Board of Editors received the paper on 05/25/2015, its review was completed on 11/12/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

33

THE APPLICATION OF FUZZY ENTROPY TO SELECTING FEATURES

OF PARTIAL DISCHARGE IN HIGH VOLTAGE CABLE JOINTS
Nguyen Tung Lam
The University of Danang, University of Science and Technology; tunglam87@gmail.com
Abstract - Partial discharge (PD) measurem ent is one of the most
im portant diagnostics m ethods of insulation systems in high
voltage equipment.PD activities may stem from various kinds of
defects, and its characteristics correspondingly behave differently.
In this study, 104 features of partial discharge are collected through
a series of experiments in laboratory, which are large
dimensionality data set.However, not all of features are useful for
classification and recognition, so the problem needed to solve is
the selection of the relevant features and elim ination of nonim portant features.The fuzzy entropy algorithm was applied to find
out features owning characteristics for distinguishing the defects in
high voltage cable joints.
Key words - high voltage cable joint, partial discharge, feature
selection, Fuzzy entropy, recognition

1. Introduction
Underground cables are a key link in metropolitan power
grids. Hence, any cable accident can lead to serious
economic losses and disruption of service to customers.
Despite the strict quality controls for the cable production
process in a plant, potential defects can occur in cable
accessories during installation [8]. Although the degradation
mechanism and identification process of cable joints have
not been fully cleared yet, it is deserving to conduct an
investigation into the prevention of unexpected failure of
cable systems [6]. Power cable system basically consists of
cables themselves and their accessories. Cable accessories
include joints and terminations.Statistically, the accidents
caused by partial discharge mostly occur at cable joints [7].
Compared to many protection methods in power system,
partial discharge is considered as one of the most
promising measures for monitoring and detecting possible
faults in the system before they occur.One of the undoubted
advantages of a computer-aided measuring system is the
ability to process a large amount of information and
transform this information into an understandable output
[4]. In this study, phase-resolved data are acquired from
digital PD measurement systems during tests. The phase
resolved data consists of a 3D discharge pattern: phase
angle discharge magnitude - discharge rate (q--t) at a
specified test voltage.
There are many kinds of defects in cable joints and each
defect own specific characteristics. Different kinds of
defects create different partial discharge signs and the
extents of damage are not the same. Based on the
investigation into partial discharge from defects, the type
of defects could be recognized, and from that the states of
cable joints can be evaluated appropriately.In this study,
104 features of partial discharge are collected through a
series of experiments in laboratory, which are large
dimensionality data set. However, not all of features are
useful for classification and recognition, so the problem
needed to solve is the selection of the relevant features and

elimination of non-important features. In addressing this

problem, different methods of data reduction have been
used and managed to eliminate the redundancy and
non-important features present in the data sets. Among
them feature selection has been shown to be a powerful
approach of dealing with high dimensional data by
selecting relevant features from data set and at the same
time removing irrelevant and redundant features that harm
the quality of the results, and therefore builds a good
learning model. A good feature selection techniques will
be able to detect and model the noisy and misleading
features from the domain problem and help to get minimal
feature subsets but still keep the important information
present in the original data [5]. This research proposes
Fuzzy entropy method to evaluate the contribution of each
feature to classification. It shows that not all the features
have one and the same discriminatory power [1]. As a
result, the crucial features are identified by using fuzzy
entropy.
2. Partial discharge data acquisition and analysis
Partial discharge could be defined as an electrical pulse
or discharge in a gas filled void or on a dielectric surface
of a solid or liquid insulation system. This pulse or
discharge only partially bridges the gap between phase
insulation to the ground, and phase to-phase insulation.
A full discharge would be a complete fault between line
potential and ground. These discharges might occur in any
void between the conductor and the ground. The pulses
occur at high frequencies; therefore, they attenuate quickly
as they pass through a short distance. The discharges are
effectively small sparks occurring within the insulation
system. Therefore, it can deteriorate the insulation and can
eventually result in complete insulation failure.
A set of PD measurement tests were carried out at the
High Voltage Laboratory of National Taiwan University of
Science and Technology (NTUST) based on the standard
IEC60270 [3].

Figure 1. Experimental setup for PD measurement

34

Nguyen Tung Lam

2.1. Partial Discharge Data Acquisition

PD measurements were generated and recorded from
laboratory tests. During the experimental process, all of the
measuring analog data was converted to digital data in
order to be stored in computer. After that process, these
data was transformed into q--t format (discharge
evaluation pattern). This data format is called as Phase
resolved partial discharge data. Figure 2 illustrates general
PD data acquisition scheme.
The most basic quantities of PD activities are apparent
charge, q, apparent charge number n, and phase position of
PD pulses with respect to the applied test voltage, ,
interpretation purposes of the under test insulation system.

by a defect present in the insulation on a 3D phase-resolved

pattern, representing a one second (40 cycles) snapshot of
PD activity. This is achieved by plotting each pulse, or in
the case of the IEC data the peak amplitude of the apparent
charge, on a three-dimensional axis consisting of the
pulses relative amplitude, the cycle number on which the
pulse appears and the phase position of the pulse on the
voltage cycle. An example of a phase-resolved pattern,
which represents three kinds of defect PD activities, can be
seen in Figure 5. The pattern is in the form of a 40x600
matrix of floating points that represent the PD activity in
40 consecutive cycles across 600 phase windows of the
voltage cycle; with the positive half cycle appearing first,
between 0 and 180 and then the negative half cycle
between 180 and 360.

Figure 2. A general PD data acquisition scheme

However, the above mentioned quantities cannot be

sufficient for a perfect diagnostics. So, heuristically there
have been introduced lots of features derived from basic
quantities termed as deduced and statistical operators, which
can be used for defect identification and evaluation.
Therefore acquiring PD data and extracting statistical feature
from acquired data benefit us for reliable PD monitoring.

Figure 3. Partial discharge signal measurement

- Basic quantities, which are quantities observed

during one voltage cycle.
- Deduced quantities, which are integrated values of
basic quantities from the first group observed
throughout several voltage cycles.
- Statistical operators, which are operators for the
statistical analysis of the deduced parameters.

Figure 4. 3D q--t transformation

This process data analysis can be a good indicator for

ambiguous PD patterns to diagnose as it presents
distinctive features of each PD defect pattern which has
been accumulated for a longer time than PD real-time data.
For the convenience of statistical analysis, the 3D patterns
were decomposed into two 2D distributions by projecting
it into the two axes - phase and magnitude.
Statistical analysis is performed separately for those two
distributions. Also, statistical analysis is performed
separately for phase angles from 0 to 180 (positive PDs),
for phase angles from 180 to 360 (negative PDs), and
on the difference between positive and negative PDs. For
each of the distributions, two types of statistics, names
amplitude statistics and shape statistics, are calculated. The
statistical descriptors are mean, standard deviation,
skewness and kurtosis. In addition, overall maximum
magnitudes of positive and negative PDs and discharge
phase region PD patterns are also calculated as features.
To diagnose a fault from the PD data it first needs to be
transformed into a generic workable format. One way of
displaying the data is to plot consecutive pulses generated

Figure 5. Defect type A 3-D q--t pattern

Using data in this form, the knowledge-based system

offers an automated approach to defect classification and
offers an explanation of the reasons for its conclusion. This
ability offers a physical explanation for the automatic
classification sets.For further statistical analysis, the 3-D
patterns are decomposed into two 2-D distributions by
projecting it into the two axes - phase and magnitude.
Figure 6 shows 3D q--t pattern decomposition that is

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

transformed into 2-D distribution pattern that includes two

axes - phase and magnitude.

Figure 6. 3-D q--t decomposition

Having basic PD quantities at hand, through statistical

operators, 104 statistical features (positive and negative)
were extracted from the four major PD quantities related to
phase and height distributions. These 104 statistical
features are also called PD fingerprints.
The PD-fingerprint in this work is a histogram
combination of statistical features of a PD signal. The
shape of the histogram provides information about the
nature of the PD signal [9]. The features of a histogram are
statistical characteristics, where the histogram is used as a
model of the probability distribution of a pattern. These
statistical features provide us with the characteristics of a
PD pattern.
Statistical methods for extracting PD features are based
on phase resolved PD patterns. By applying statistical
computation on PD patterns, different distributions can be
characterized as statistical features. The mean value,
standard deviation, skewness and kurtois values are
calculated according to statistical formulas. Discharge
phase region is also calculated.
2.2. Experimental setup
In this research, the experimental objects are cable joint
defect models. Two types of relevant models are well
designed, based on investigations of numerous power
equipment failures. Two defect types are described, as
follows:
1. Defect type A: remove a part of insulation. According
to the criterion, the length of insulation of two cable sides
is the same and complies with standard. In this type, a part
of one insulation side was cut out.
2. Defect type B: gap between insulation, a hole made in
insulation belong to the part of cable inside the joint.
3. Feature selection based on fuzzy entropy
Feature selection is a process of choosing small subset
of features out of a set of candidate features based on
certain criteria. Feature selection plays an important role in
classification for several reasons. First it can simplify the
model and in this way computational cost can be reduced
and also when the model is taken for practical use fewer
inputs are needed which means in practice that fewer

35

measurements from new samples are needed. Second by

removing redundant features from the data set one can also
make the model more transparent and more
comprehensible, providing better explanation of suggested
diagnosis. Feature selection process can also reduce noise
and in this way enhance the classification accuracy.
The key of PD classification problem like in any
classification systems is a set of high quality features. These
features should represent the characteristics of PD signals.
More importantly, these features must possess strong
discriminant power so that the classifier designed based on
those features can give desired performance. Since PD is a
stochastic process, namely, the occurrence of PD depends on
many factors, such as temperature, pressure, applied voltage,
and the test duration, and since PD signals contain noise and
interference, PD measurements corresponding to different
insulation conditions are almost indistinguishable, i.e., PD
diagnosis is a complex classification problem. Thus finding
a set of high quality features that give more accurate and
reliable classification is even more critical in design of PD
diagnostic systems.
In this paper, Fuzzy entropy method is proposed to
reduce dimension of partial discharge features.
For a classification system, the most important
procedure is partitioning the pattern space into decision
regions. Once the decision regions are decided, we can
apply these partitioned decision regions to classify the
unknown patterns. The partition of decision regions is part
of the learning procedure or training procedure since the
decision regions are decided by the training patterns. In
fuzzy entropy classifier, decision regions are enclosed by
the surfaces produced from each dimension. The surfaces
are determined by the distribution of input patterns.
Entropy is a measure of the amount of uncertainty in the
outcome of a random experiment, or equivalently, a
measure of the information obtained when the outcome is
observed [2]. In this paper, a fuzzy entropy measure which
is an extension of Shannons definition will be proposed.
The fuzzy entropy can discriminate the actual
distribution of patterns better. By employing membership
functions for measuring match degrees, the value of
entropy not only considers the number of patterns but also
takes the actual distribution of patterns into account.
The fuzzy entropy reflects more information in the
actual distribution of patterns in the pattern space. Since
the fuzzy entropy can discriminate pattern distribution
better, we employ it to evaluate the separability of each
feature. Intuitively, the lower the fuzzy entropy of a feature
is, the higher the features discriminating ability is. The
procedure for computing the fuzzy entropy of each feature
is described as the flowchart in Figure 7. This process
includes four main parts:
-

Determine the number of intervals.

Determine the interval locations.
Assign a membership function for each interval.
Compute the fuzzy entropy of each feature via
summation of the fuzzy entropy of all intervals.

At first, assume the number of interval I equal to 2 which

36

Nguyen Tung Lam

is the smallest number of interval. Then increase I until the

total fuzzy entropy of I intervals is less than that of I - 1
intervals. The final fuzzy entropy is computed with I-1
intervals.
Set initial number
of interval I=2

Set initial centers of

intervals c

Assign interval label to

each element

Yes

Recompute the
cluster centers

Check: Does any center

change?

Table 2. Values of Fuzzy Entropy

A1, B1
Value

No.
Feature

83

0.69525

0.67481

0.75072

30

0.72506

2
104

0.74844
0.75031

35

0.89264

11

0.94105
0.95920
0.96313

the total fuzzy

entropy of I intervals
is less than that of I 1 intervals

True

I=2

I=I+1

The number of
interval I=I-1
& The fuzzy
entropy is
computed with
I-1 intervals

False

Figure 7. Flowchart of calculating value Fuzzy entropy

4. Results
As mentioned in the previous chapter, each defect was
tested on 3 cable joints.
Table 1. Set of PD datas class

PD Defect Types
Remove a part of insulation
Gap between insulations

0.72590

14

0.78717

13

0.79021

46
47

45
82

0.76027
0.76313

45
63

0.79145
0.81147

45
38

0.96674
0.98958

27

0.76987

23

0.81163

16

0.80721

0.99492
0.99593

0.81381

0.81554
0.82831

2
33

46

46
11

34

0.99628

A2, B2
No.
Feature

Value

Value

A2, B3
No.
Feature

Value

83

0.43104

97

0.78922

0.91810

46
3

0.48254
0.55353

98

0.79779

46

0.93049

92

0.95837

45

0.93576

0.55498

21

0.96971

2
88

0.55773
0.55843

22

0.97513

2
8

0.94505
0.94838

45

0.97986

10

0.95489

45
61

0.56206
0.56995

46
101

0.98135
0.98343

18

0.96062

17

0.96299

60

0.56995

0.96787

0.57423

0.98379
0.98403

15

58

2
19

41

0.97356

A3, B1

A (A1, A2, A3)

B (B1, B2, B3)

After finishing all the tests, partial discharge data of

defects was collected. In this study, data obtained in the
results of experimental works was considered and
transformed
into
statistical
features
and
these
considerations are explained in PD Data Acquisition and
Analysis section. These statistical features including
skewness, kurtosis, standard deviation, mean, DPR,
Q_sum, Q_num, Q_max, Q_ave are all
calculated based on the PD signals. Statistical features
consist of 104 features numbered from 1 to 104. In this
study, 34 kV PD experiment data that include 120 sample
data for each defect model is used. Apply Fuzzy entropy
theory to all features of partial discharge, with inputs to
Matlab program as values of all features and types of
defects corresponding. As a result, the fuzzy entropy value
of each feature is computed; features with higher fuzzy
entropy are less relevant to classification goal. Totally 9
pairs of defects (each pair includes 1 defect type A and 1
defect type B) were conducted to calculate fuzzy entropy
values. Table 2 shows the features owning smallest values
of fuzzy entropy of each case.

Value

0.75572
0.75999

No.
Feature

Class

No.
Feature

2
14

No.
Feature

I>2

Value

29

A2, B1

Assign membership
function for each
interval.

A1, B3

No.
Feature

No

Centers of intervals
are determined

Compute the total

fuzzy entropy of
all intervals

A1, B2

A3, B2

Value

No.
Feature

83

0.69525

30

0.72506

29

0.72590

2
14

0.75572
0.75999

45
82

0.76027
0.76313

27

0.76987

16
46

A3, B3

Value

No.
Feature

Value

47

0.15372

35

0.14152

50

0.20041

46
40

0.21575
0.22609

46
31

0.14490
0.14748

2
44

0.22911
0.24992

45
53

0.14959
0.15461

42

0.25441

2
10

0.15624
0.16715

0.27248
0.27675

0.17647

0.80721

45
61

38

0.19347

0.81381

93

0.28323

34

0.20328

As a result, it can be clearly seen that the feature

number 2, 45 and 46 are always in the top of features
having the smallest values of fuzzy entropy. As mentioned
in the theory, those features impact significantly on
classifying defects in cable joints. We can use three
features instead of all 104 features to recognize not only
more accuracy but also less time of computing.
Table 3. Selected features

No.

Features

Total number of partial discharge in all circles

45

Height

distribution

average

partial

discharge

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

REFERENCES

magnitude kurtois values

46

Height distribution average

maximum kurtois values

37

partial

discharge

5. Conclusion
Based on the preprocessing stage, it is necessary to
gather the database from conducted PD tests with two high
voltage cable joints including prefabricated defects. Phaseresolved PD data was successfully evaluated and processed.
The data gathered from the selected databases is connected
to MATLAB software where the data is processed. The
second part of the analysis system is based on feature
selection algorithm. This takes Fuzzy entropy method as a
medium to process the PD data. Feature selection techniques
were applied to the PD data and the characteristic points of
interests are being selected by computing fuzzy entropy
value of each defect. The data provides meaningful
information for the classification of PD defects.
The measure of uncertainty is adopted as a measure of
information. Hence, the measures of fuzziness are known
as fuzzy information measures. The measure of a quantity
of fuzzy information gained from a fuzzy set or fuzzy
system is known as fuzzy entropy. In this study, the fuzzy
entropy algorithm was applied to find out three features
owning most useful characteristic for distinguishing the
defects in high voltage cable joinst.

[1]

G. MacLachlan, "Discriminant Analysis and Statistical Pattern

Recognition", Willey-Interscience, pp. 389-398, 2004.

[2]

Hahn-Ming Lee,Chih-Ming Chen, Jyh-Ming Chen, Yu-Lu Jou, "An

Efficient Fuzzy Classifier with Feature Selection Based on Fuzzy
Entropy", IEEE Transactions On Systems, Man, And Cybernetics
Part B: Cybernetics, Vol. 31, No. 3, pp. 426-432, 2001.
IEC 60270, "High-voltage test technique-Partial discharge
measurements", 2000.

[3]
[4]

N. Sahoo, M. Salama and R. Bartnikas, "Trends in partial discharge

pattern classification: a survey", IEEE Transactions on Dielectrics
and Electrical Insulation, Vol. 12, No. 2,pp. 248 - 264, April 2005.

[5]

Q. S. Jensen Richard, "Computational intelligence and feature

selection: Rough and Fuzzy Approaches", IEEE Press Series on
Computational Intelligence, 2008.
Tokunaga. S, Tsurusaki. T, "Partial Discharge Characteristics till
Breakdown for XLPE Cable Joint with an Artificial Defect", Proceedings
of the 7th International Conference on Properties and Applications of
Dielectric Materials, Vol.3, pp 1206-1209, Nagoya, 2003.
Wenhu Yang,Yanqun Liao, Yang Xu, Xiaolong Cao "Analysis of
Partial Discharge Measured on Field for Cable Joint", International
Conference on High Voltage Engineering and Application, pp. 408411, Chongqing, China, November 9-13, 2008.

[6]

[7]

[8]

Wu Ruay-Nan, Chang,Chien-Kuo, "The Use of Partial Discharges as

an Online Monitoring System for Underground Cable Joints", IEEE
Transactions On Power Delivery, vol. 26, pp. 1585-1591, 2011.

[9]

Yu-Hsun Lin, Ruay-Nan Wu, I-Hua Chung, "Novel trend of "l" shape
in PD pattern to judge the appropriate crucial moment of replacing
cast-resin current transformer", IEEE Transactions on Dielectrics and
Electrical Insulation, vol. 15, no. 1, pp. 292-301, 2008.

(The Board of Editors received the paper on 07/10/2015, its review was completed on 09/27/2015)

38

Tran Huu Nam, Vu Minh Hung, Vo Quoc Thang

A STUDY OF IMPROVING PROPERTIES OF ALIGNED MULTI-WALLED

CARBON NANOTUBE/ EPOXY COMPOSITES
Tran Huu Nam, Vu Minh Hung, Vo Quoc Thang
PetroVietnam University, Long Toan, Ba Ria, Ba Ria - Vung Tau; namth@pvu.edu.vn
Abstract - Composites made of an epoxy resin film and differently
stacked aligned multi-walled carbon nanotube (MW CNT) sheets
have been developed using hot-melt prepreg processing. The
horizontally aligned 20-ply MW CNT sheets were created from
vertically aligned MW CNT arrays using solid-state drawing and
winding techniques. However, wavy and poor-packed MW CNTs in
the sheets have restricted their load-transfer efficiency in the
composites. Therefore, mechanical stretching was used to straight
the wavy MW CNTs and to increase the dense packing of MW CNTs
in the sheets. In this article, improving the composite properties
through mechanical stretching of the MW CNT sheets was studied.
Mechanical stretching of the MW CNT sheets considerably improved
the mechanical properties of the composites. The improvement of
the composite properties is derived from the straightening of wavy
MW CNTs and the increase of MW CNT dense packing is caused by
mechanical stretching.The decrease of the wavy MW CNTs is more
efficient than the enhancement of MW CNT dense packing.
Key words - aligned carbon nanotubes;prepregs;
com posites; m echanical stretching; m echanical properties.

nano

1. Introduction
Carbon nanotubes (CNTs) were discovered by Iijima in
1991 [1]. They have attracted interest in extensive research
because of exceptional mechanical, electrical, and thermal
properties [2, 3]. The excellent properties make CNTs an ideal
reinforcement in high-performance composites. Most studies
of CNT-reinforced polymer composites have focused on
dispersing CNTs intopolymer matrices [4]. However,
mechanical properties of such composites fall far short of the
corresponding properties of high-performance structural
composites. Therefore, great efforts have recently been
undertaken to synthesize aligned CNT arrays vertically [5] for
production of long-aligned CNT sheets. The easiest way of
processing aligned CNT sheets from the aligned CNT arrays
is the use of solid-state drawing and winding techniques [6].
The aligned MWCNT sheets can be used to fabricate
advanced
composites
with
desirable
structural
characteristics [7]. Although those composites contain
aligned MWCNTs, their mechanical properties are
inadequate partly because of wavy and poor-packed
MWCNTs. Therefore, stretching has been applied to the
aligned MWCNT sheets to improve composite properties
[8, 9]. Results in my earlier report [9] show that the
stretching of the MWCNT sheets with 50 and 100 plies is
less efficient than that of 20-ply MWCNT sheet.
Consequently, in this study 20-ply aligned MWCNT sheets
were used to develop laminated epoxy composites. Effects
of mechanical stretching the 20-ply MWCNT sheets on the
composite properties were studied.
2. Materials and Methods
2.1. Materials
Vertically aligned and spinnable MWCNT arrays with
approximately 0.8 mm height were grown on a bare quartz

substrate using chloride-mediated chemical vapor

deposition [5]. As-grown MWCNTs used in this study
have mean diameter of 38 nm [10]. The MWCNT diameter
in the sheets varies from about 20 nm to 55 nm. B-stage
epoxy resin films covered with release paper and plastic
film were obtained from Sanyu Rec Co. Ltd. (Osaka,
Japan) with the recommended cure condition of 130C for
2 h. The real weight of the epoxy resin sheet with density
of 1.2 g/cm3was controlled approximately 12 6 g/m2.
2.2. Methods
2.2.1. Processing of horizontally aligned MWCNT sheets
Solid-state drawing and winding techniques were
applied to transform a vertically aligned MWCNT array
into horizontally aligned MWCNT sheets. The MWCNT
webs are drawn from vertically aligned MWCNT arrays
and are wound on a rotating spool to create horizontally 20ply aligned MWCNT sheets. Detailed procedures for the
fabrication of multi-ply MWCNT sheets are depicted in the
literature [710]. Although most MWCNTs are aligned,
many wavy and entangled MWCNTs are visible in pristine
sheets [79]. In this study, the aligned 20-ply MWCNT
sheets were used for laminated composite fabrication.
2.2.2. Mechanical stretching of the MWCNT sheets
To straight wavy MWCNTs, mechanical stretching
with a ratio of 2% was applied to pristine MWCNT sheets.
Stretch ratio was calculated using the following equation.

L2 L1
L1

(1)

Therein, L1 and L2 are segment lengths of the MWCNT

sheets between the clamped grips before and after
stretching, respectively. More details about the stretching
device and processing were presented by Nam et al. [9].
2.2.3. Fabrication of aligned MWCNT/epoxy composites
The composites made of an epoxy resin film and stacked
MWCNT sheets were fabricated using hot-melt prepreg
processing with a vacuum assisted system (VAS)
[10].Figure 1 portrays the schematic views of stacking the
MWCNT sheets on an epoxy resin film to form the
composite laminates. Firstly, aligned MWCNT/epoxy
prepregs were prepared by stacking1, 5, and 10 nonstretched (pristine)or stretched 20-ply MWCNT sheets with
20 mm in width and 40 mm in length on an epoxy resin film.
The prepregs were set in two release films (WL5200;
Airtech International Inc., CA, USA) and were pressed
under 0.5 MPa pressure for 5 min at 100 C using a test press
(MP-WNL; Toyo Seiki Seisaku-Sho Ltd., Tokyo, Japan).
Subsequently, the prepregs were peeled off from the release
films. Finally, the prepregs were cured at 130 C for 2 h
under 2 MPa in the VAS to produce the composites. The

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

non-stretched and stretched composites are assigned

respectively as NCom-X and SCom-X, in which X
corresponds to the MWCNT sheets in number (1, 5 and 10).

39

loss of MWCNTs (mf), epoxy resin (mm), and the

composite (mc) as follows.

mf

mm mc

m f

(2)

The MWCNT volume fraction (Vf) was ascertained

from the MWCNT mass fraction, epoxy resin density (m),
and the density of the composite (c) as follows.

Vf 1

1 m
f

(3)

The MWCNT volume fractions of the composites are

presented in Table 1. The MWCNT volume fraction of the
composites increases by increasing the aligned MWCNT sheets.
Table 1. The MWCNT fractions estimated from TGA results
Non-stretched
composites

Materials

Epoxy
MWCNT
MWCNT sheet MWCNT sheet
resin

Figure 1. Schematic views of stacking the MWCNT sheets on an

epoxy resin film to form the composites.

2.2.4. Thermogravimetric analysis (TGA)

The thermal degradation of epoxy resin, MWCNTs,
and their composites was analyzed up to 800 C in argon
ambient at a flow rate of 300 ml/min using a
thermogravimetric analyzer (DTG60A; Shimadzu Corp.,
Kyoto, Japan). About 5 mg of each specimen was loaded
for each measurement at a heating rate of 10 C/min.
2.2.5. Microstructural characterization and testing
Tensile tests were conducted for the aligned MWCNT
sheets and composites in the laboratory environment at room
temperature (RT).Tensile specimens with 610 mm gauge in
length and 35 mm in width were tested on a testing machine
(EZ-L; Shimadzu Corp., Kyoto, Japan) with a load cell of 50
N and a crosshead speed of 0.1 mm/min. Specimens width was
measured using an optical microscope (SZX12; Olympus
Corp., Tokyo, Japan), whereas their thickness was measured
using a micrometer with 0.001 mm accuracy (Model 102-119,
Mitutoyo Corp., Kanagawa, Japan). The strain of tensile
specimens was measured using a non-contacting video
extensometer (TRIViewX; Shimadzu Corp., Kyoto, Japan)
with two targets. Mean tensile properties were obtained from
at least five specimens for each MWCNT sheet and composite.
The microstructural morphologies of MWCNTs in the sheets
and composite fracture surfaces were observed using field
emission scanning electron microscopy (FE-SEM) (SU8030;
Hitachi Ltd., Tokyo, Japan). Polarized Raman spectra were
measured to determine the degree of MWCNT alignment in
the composites using Raman spectroscope with laser excitation
of 532 nm (XploRA-ONE; Horiba Ltd., Kyoto, Japan).
3. Results and Discussion
3.1. MWCNT volume fraction of the composites
MWCNT volume fraction of the composites was
determined through TGA results. The respective mass loss
of MWCNTs, epoxy resin and the composites were
measured between 150oC and 750 oC. The MWCNT mass
fraction (mf) of the composite was calculated from the mass

Stretched
composites

10

10

Mass loss
(%)

87.9

2.59

74.4 45.4 35.7 71.6 42.7 33.4

MWCNT
mf (%)

15.8 49.9 61.2 19.1 53.0 63.9

MWCNT
Vf (vol. %)

10.1 37.4 48.7 12.1 40.4 51.5

3.2. Properties of MWCNT sheets and their composites

Thickness of the pristine 20-ply MWCNT sheets was
measured as 12 m. The mechanical properties of epoxy
resin, pristine 20-ply MWCNT sheets, and composites
were measured using tensile test. The epoxy resin film and
pristine MWCNT sheets respectively showed mean tensile
strength of 64.4 and 96.8 MPa, elastic modulus of 2.55 and
7.34 GPa, and strain at maximal stress of 4.84 and 2.11%.
Typical stressstrain curves of epoxy resin, pristine 20-ply
MWCNT sheets, and the composites are depicted in Figure
2. As observed in Figure 2, the composites indicated a
linear stress-strain relation until the specimen fractures
with no bending of the curves at high loads. Typical stressstrain curve of pristine MWCNT sheets showed that the
stress is increased up to the maximum by increasing strain
to approximately 2%. In this stage, the wavy CNTs are
straightened under the tension. Above 2% strain, the stress
decreases concomitantly with enhancing strain up to the
specimen fractures. The reduction of the stress is attributed
to the MWCNT sliding during the tensile testing, as
presented by Inoue et al. [6]. Consequently, mechanical
stretching of the MWCNT sheets was conducted with a 2%
ratio in the laboratory environment at RT.
The properties of the non-stretched and stretched
MWCNT/epoxy composites are given in Table 2. The
mechanical properties of the non-stretched and stretched
composites increase by increasing the MWCNT sheets
(volume fraction). Tensile strength and elastic modulus of
the non-stretched composites enhanced strongly whereas
fracture strain increased only slightly. The NCom10 and

40

Tran Huu Nam, Vu Minh Hung, Vo Quoc Thang

SCom10 respectively exhibited an increase in tensile

strength by 241.6% and 204.2%, in elastic modulus by
204.3% and 181.8%, and in fracture strain by 10.7% and
7.3% compared with the NCom1 and SCom1. The

enhancement in the mechanical properties of the

composites is attributed to increased MWCNT volume
fraction (Table 1).

Table 2. Properties of non-stretched and stretched MWCNT/epoxy composites

Property

Non-stretched composites

Stretched composites

1 sheet

5 sheets

10 sheets

1 sheet

5 sheets

10 sheets

Thickness (m)

67

11 13

16 18

56

10 12

15 17

Density (g/cm3)

1.28

1.50

1.59

1.30

1.52

1.61

430.3 49.4

548.5 52.5

Tensile strength (MPa) 105.6 10.1

258.1 29.4

360.6 31.1 180.3 16.1

Elastic modulus (GPa)

23.1 2.9

54.5 6.0

70.3 8.0

35.6 3.6

79.3 7.4

100.4 11.3

Fracture strain (%)

0.46 0.08

0.48 0.05

0.51 0.04

0.51 0.07

0.54 0.04

0.55 0.03

and pressed composites. As observed in Figure 4, the

alignment of MWCNTs in the composites is maintained
during resin impregnation using hot-melt prepreg
processing. The non-stretched MWCNT/epoxy composites
contained many wavy and entangled MWCNTs (Figure 4a).
The stretched MWCNT/epoxy composite showed marked
straightening of wavy MWCNTs caused by mechanical
stretching of the MWCNT sheets (Figure 4b).

Figure 2. Typical stressstrain curves of epoxy resin, pristine

MWCNT sheet, and the composites

3.3. Evaluating of MWCNT alignment and straightening

Microstructural morphologies of the aligned MWCNT
sheets before and after mechanical stretching are shown in
Figure 3. Although most MWCNTs in the sheets are selfaligned in the drawing direction, the wavy MWCNTs can
be clearly seen in the non-stretched samples (Figure 3a).
After 2% stretching, the wavy MWCNTs were reduced
considerably (Figure 3b). The wavy MWCNTs are selfassembled and are straightened along the load direction
during stretching.Therefore, the dense packing of
MWCNTs in the stretched sheets (Figure 3b) became more
compact than that in the non-stretched sheets (Figure 3a).

Figure 3. FE-SEM micrographs showing microstructural

morphologies of (a) non-stretched and (b) stretched MWCNT sheets

FE-SEM micrographs taken from polished surfaces of

the non-pressed and pressed composites reinforced with 10
MWCNT sheets are presented in Figure 4. Those images
showed in-plane MWCNT distribution in the non-pressed

Figure 4. FE-SEM micrographs showing in-plane MWCNT

distribution of (a) non-stretched and (b) stretched composites
reinforced by 10 MWCNT sheets

The straightening and alignment of MWCNTs after

mechanical stretching can be examined using polarized
Raman spectroscopy [11, 12]. Typical polarized Raman
spectra in the range of 10002000 cm-1 are presented in
Figure 5. Raman spectroscopic measurements were
conducted with incident light normal to the composite
samples,
which
was
polarized
parallelly
and
perpendicularly to the MWCNT alignment (see Figure
5inset). Raman spectra for all samples show two main
peaks located at approx. 1350 cm-1 and approx. 1580 cm-1,
which are attributed respectively to the disorder-induced D
band and the graphic-like G band. Compared with the nonstretched samples, the stretched ones showed a higher
intensity of D and G bands at 0 and lower D and G band
peaks at 90. The G band peaks decreased greatly for the
stretched composites at the angle of 90, which proves that
the MWCNT alignment in the composites was improved
considerably after stretching the MWCNT sheets.
In addition, the ratio of G-band intensity in the parallel
configuration to
the perpendicular configuration
was
used
to characterize the degree of
R

I
/
I
G G

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

MWCNT alignment [6]. The higher MWCNT alignment

produces the higher G-band intensity ratio, because Raman
scattering is more intense when the polarization of the
incident light is parallel to the axis of a MWCNT [12]. The
G-band intensity ratio R of the non-stretched composites
reinforced by 10 MWCNT sheets was 1.33, as presented in
Figure 5. After stretching, the R value of the stretched
composites reinforced by 10 MWCNT sheets was
markedly enhanced to 1.95. The marked enhancement in
the R is ascribable to the better alignment of MWCNTs in
the composites caused by mechanical stretching. Therefore,
the mechanical stretching considerably improved the
MWCNT alignment in the stretched composites.

41

the axial loading direction (Figure 4a). Therefore, just a

small fraction of MWCNTs in the non-stretched
composites carries load effectively during tensile testing.
The wavy MWCNTs are straightenedduring the stretching
of the MWCNT sheets (Figure 3b). The straight MWCNTs
have a larger fraction of their length aligned with the
loading direction, which resulted in improved mechanical
properties of the stretched composites [8].

Figure 6. Mechanical properties of the composites versus

MWCNT volume fraction

Figure 5. Polarized Raman spectra of the non-stretched and

stretched composites reinforced by 10 MWCNT sheets at 0 and
90 (0 and 90 directions correspond to configurations where
the polarization direction of the laser light are, respectively,
parallel and perpendicular to the direction of CNT alignment)

3.4. Effects of stretching onthe composite properties

As Table 1 shows, MWCNT volume fractions of the
stretched composites are higher than those of
corresponding non-stretched ones. The increased MWCNT
volume fraction of stretched composites is explained by the
decrease of the composite thickness (Table 2). The
reduction of the composite thickness is attributable to
straightening of wavy MWCNTs and dense packing of
MWCNTs in the sheets caused by stretching (Figure
3).The MWCNTs in the stretched sheets tend to contract in
the directions transverse to the stretching direction [9].
Therefore, the thickness ofstretched sheets became thinner
than that of the non-stretched ones. The mean thickness of
the stretched composites reinforced by 1, 5, and 10
MWCNT sheets respectively reduced by 8.1%, 13.6%, and
16.4% compared with that of the non-stretched ones.
Enhancement of the mechanical properties of the
composites as a result of stretching the MWCNT sheets is
presented in Figure 6. The mechanical properties of the
stretched composites are significantly higher than those of
the non-stretched ones. The stretched composites
reinforced by 1, 5, and 10 MWCNT sheets respectively
exhibited an increase in tensile strength by 70.8, 66.7, and
52.1%, in elastic modulus by 54.4, 45.5, and 42.9%, and in
fracture strain by 9.7, 14.2, and 6.4% compared with
corresponding non-stretched ones.The non-stretched
composites evidently showed many wavy MWCNTs along

The increased mechanical properties of the stretched

composites are probably derived from enhancing the
MWCNT volume fraction (Table 1) and from reducing the
wavy MWCNTs (Figure 3). To assess the effects of these
two factors, the respective percentage increases of elastic
modulus and MWCNT volume fraction in comparison
between stretched and non-stretched composites were
analyzed, with results presented in Figure 7. The percentage
increases of the MWCNT volume fraction are markedly
much lower than those of elastic modulus. Therefore, the
percentage increase of elastic modulus by the enhanced
MWCNT volume fraction is lower than that coming from
reduction of the wavy MWCNTs. The increased MWCNT
volume fraction of the stretched composites is attributed to
the decreased composite thickness caused by the dense
packing of MWCNTs (Figure 3). Generally, the increase of
MWCNT volume fraction is less efficient than the decrease
of wavy MWCNTs. Moreover, the percentage increase of
elastic modulus showed a reduced trend by increasing the
MWCNT sheets (Figure 7).

Figure 7. Percentage increase of MWCNT volume fraction and

elastic modulus of stretched composites compared with nonstretched ones

The fractured surface morphology of the non-stretched

and stretched composites reinforced by 5 MWCNT sheets
is depicted in Figure8. High-resolution micrographs in
Figure 8 show that epoxy resin was infiltrated well between
the MWCNTs. However, many pulled-out MWCNTs with
length of a few micrometers are apparent on the fractured
surface of the composites (Figure 8). For the stretched

42

Tran Huu Nam, Vu Minh Hung, Vo Quoc Thang

MWCNT/epoxy composites, the MWCNT bundles can be

seen on the fracture surfaces (Figure 8b). The bundled
MWCNTs formed by stretching are evidently observed on
surface morphologies of the stretched MWCNT sheets
(Figure3b). The MWCNT bundles indicated the dense
packing of MWCNTs in the stretched sheets.In general,
mechanical stretching of the MWCNT sheets enhanced
their composite properties considerably.
a

1 m

REFERENCES
[1]

Iijima S, Helical microtubules of graphitic carbon, Nature, 354,

NPG, 1991, 5658.

[2]

Ruoff RS and Lorents DC, Mechanical and thermal properties of

carbon nanotubes, Carbon, 33(7), Elsevier, 1995, 925930.

[3]

Ebbesen TW, Lezec HJ, Hiura H, Bennett JW, Ghaemi HF, and
ThioT, Electrical conductivity of individual carbon nanotubes,
Nature, 382, NPG, 1996, 5456.

[4]

Thostenson ET, Ren Z, and Chou TW, Advances in the science and
technology of carbon nanotubes and their composites: a review,
Composites Scienceand Technology, 61(13), Elsevier, 2001, 1899
1912.

[5]

Inoue Y, Kakihata K, Hirono Y, Horie T, Ishida A, and Mimura H,

One-step grown aligned bulk carbon nanotubes by chloride mediated
chemical vapor deposition, Applied Physics Letters, 92(21), AIP,
2008,213113.

[6]

Inoue Y, Suzuki Y, Minami Y, Muramatsu J, Shimamura Y, Suzuki

K, and et al., Anisotropic carbon nanotube papers fabricated from
multiwalled carbon nanotube webs, Carbon, 49(7), Elsevier, 2011,
24372443.
Ogasawara T, Moon SY, Inoue Y, and Shimamura Y, Mechanical
properties of aligned multi-walled carbon nanotube/epoxy
composites processed using a hot-melt prepreg method, Composites
Science Technology, 71(16), Elsevier, 2011, 18261833.
Wang X, Bradford PD, Liu W, Zhao H, Inoue Y, Maria JP, and et
al., Mechanical and electrical property improvement in CNT/Nylon
composites through drawing and stretching, Composites Science and
Technology, 71(14), Elsevier, 2011, 16771683.

MWCNT bundles

1 m

Figure 8. FE-SEM micrographs showing the fracture surfaces

of (a) the non-stretched and (b) stretched composites reinforced
by 5 MWCNT sheets

[7]

4. Conclusions
The composites based on epoxy resin and stacked
aligned 20-ply MWCNT sheets were developed using hotmelt prepreg processing with the VAS. The mechanical
properties of the composites enhanced gradually by
increasing the MWCNT volume fraction. Mechanical
stretching of the MWCNT sheets decreased the composite
thickness and increased MWCNT volume fraction.
Mechanical stretching the MWCNT sheets with a 2% ratio
considerably improved the composite properties. The
improved mechanical properties of stretched composites
proceeded from decreased wavy MWCNTs and from
increased dense packing of MWCNTs caused by stretching.
The reduction of the wavy MWCNTs is more efficient than
the enhancement of MWCNT dense packing.
Acknowledgements. This research is funded
PetroVietnam University under grant code GV1514.

by

[8]

[9]

Nam TH, Goto K, Oshima K, Premalal EVA, Shimamura Y, Inoue

Y, and et al., Effects of stretching on mechanical properties of
aligned multi-walled carbon nanotube/ epoxy composites,
Composites Part A, 64, Elsevier, 2014, 194202.

[10] Nam TH, Goto K, Yamaguchi Y, Premalal EVA, Shimamura Y,

Inoue Y, and et al., Effects of CNT diameter on mechanical
properties of aligned CNT sheets and composites, Composites Part
A, 76, Elsevier, 2015, 289-298.
[11] Liu W, Zhang X, Xu G, Bradford PD, Wang X, Zhao H, and et al.,
Producing superior composites by winding carbon nanotubes onto a
mandrel under a poly (vinyl alcohol) spray, Carbon, 49(14), Elsevier,
2011, 4786-4791.
[12] Ji J, Sui G, Yu Y, Liu Y, Lin Y, Du Z,et al.,Significant improvement
of mechanical properties observed in highly aligned carbonnanotube-reinforced nanofibers, The Journal of Physical Chemistry
C, 113(12), ACS, 2009, 4779-4785.

(The Board of Editors received the paper on 12/04/2015, its review was completed on 12/25/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

43

PHYSICAL LAYER SECRECY PERFORMANCE ANALYSIS

OF TAS/ MRC SYSTEM OVER RAYLEIGH/ NAKAGAMI FADING CHANNELS
Nguyen Van Tho1, Van Phu Tuan1, Vo Tan Loc2, Ha Dac Binh1
1

Duy Tan University; nguyenvantho@duytan.edu.vn

2
Pham Van Dong University

Abstract - The broadcast nature of radio propagation m akes

wireless comm unication extremely vulnerable to eavesdropping
attack. In this paper, we investigate the physical layer secrecy
performance of multiple-input multiple-output (MIMO) system with
transm ission antenna selection (TAS) and receiver maxim al-ratio
com bining (MRC) in the presence of a single antenna passive
eavesdropper over dissim ilar fading channels. W e consider two
scenarios: 1) The legal / illegal channels are subject to Rayleigh
/Nakagam i fading, respectively; 2) The legal /illegal channels
undergo Nakagam i /Rayleigh fading, respectively. Especially, the
exact close-form expressions for the probability of non-zero
secrecy capacity and the secrecy outage probability using
statistical characteristics of the signal-to-noise ratio (SNR) of these
scenarios is derived. These expressions allow us to assess the
security capability of the considered system. The numerical result
discussion provides practical design of the effect of various system
param eters, such as average SNRs, Nakagam i fading m odel, and
num ber of transmission antennas on the secrecy performance of
the considered system .
Key words - physical layer secrecy; secrecy capacity; TAS/RMC
system ; Rayleight fading; Nakanami fading.

1. Introduction
The increase in exchange information demand becomes a
motivation for development of wireless communication.
Because wireless communication is a flexible data communication, it leads the explosive growth in recent decades.
However, the broadcast nature of wireless medium makes the
security risk always be challenges. In recent years, physical
layer (PHY) security has become an attractive topic due to its
low complexity, latency and ability to combine with other
mechanisms in order to improve a capability of overall
ensuring security. Shannon [1], Wyner [2], and Leung-YanCheong [3] were pioneers in the research on physical layer
secure communication. There are many extensive works
aimed at im- proving the secrecy performances of wireless
communications by exploiting the multiple antennas. Some of
them are [4][10] that present a quasi-static Rayleigh fading
wiretap channel multiple antenna devices. In [4], the authors
have investigated the PHY secrecy performance of a
communication scheme consisting of a multiple antenna
transmitter using TAS and a single antenna receiver in the
presense of a multiple antenna eavesdropper. Their results
show that high levels of security can be achieved when the
number of antennas at transmitter increases, even when
eavesdropper has multiple antennas. The authors in [5]
analyze the impact of antenna correlation on secrecy
performance of MIMO wiretap channels where transmitter
employs transmission antenna selection while receiver and
eavesdropper perform MRC with arbitrary correlation. Nan
Yang et al. [6] analyzed secrecy performance of MIMO
wiretap channel in Nakagami-m fading environments with
non-identical fading parameters for the main channel and the
eavesdroppers channel. The authors in [7] proposed an

opportunistic scheduling with TAS to enhance physical layer

security. At the transmitter, a single antenna is selected to
maximize the instantaneous SNR of the main channel, while
at the receiver and the eavesdropper, MRC or selection combining (SC) is applied. They can also conclude that the secrecy
outage probability is almost independent of the number of
antennas and eavesdroppers in high SNR region. The physical
layer security performance of MRC systems under two-waves
with diffuse power fading channels is analyzed in [8]. Two
practical scenarios are taken into account, depending on
whether or not the channel state information (CSI) of the
eavesdropper is known at the transmitter. For the first scenario
where eavesdroppers CSI is not known, the expressions for
the exact and asymptotic average secrecy capacity are derived.
For the second scenario where eavesdroppers CSI is known,
the authors derive the expressions for the exact and asymptotic
secrecy outage probability. Based on these, we show that the
secrecy diversity order is solely dependent on the number of
receive antennas at the legitimate receiver and independent of
the number of antennas at the eavesdropper. The PHY secrecy
performance of multiple-input single-output (MISO) UltraWideband (UWB) system with TAS is evaluated in [9] and
the time-reversal technique is used to improve the secrecy
capacity in MIMO UWB system [10].
From above studies and to the best of our knowledge,
most of previous works on PHY security consider the
similarity between legal channel and illegal channel.
However, due to the mobility of mobile devices, the
difference in fading characteristics between two channels
must be examined, practically. In this paper, we investigate
the physical layer secrecy performance analysis of MIMO
system using TAS/MRC in the presence of a single antenna
passive eavesdropper over dissimilar Rayleigh/ Nakagami
fading channels. The main contribution of this paper
resides in the derivation of the exact closed-form
expressions of the probability of non-zero secrecy capacity
and the secrecy outage probability overmixed Rayleigh/
Nakagami fading channels.In addition, we also show the
results of simulation and analysis to clarify the secrecy
performance of this considered system.
The rest of this paper is organized as follows. Section
II presents the system and channel model. Physical layer
secrecy performance of the considered system is analyzed
in Section III. In Section IV, we show the numerical
results. We conclude our work in Section V.
2. System and channel model
We consider the system illustrated in Figure 1. Alice
and Bob are two legitimate users equipped with Na and Nb
antennas respectively while Eve is a single antenna passive

44

Nguyen Van Tho, Van Phu Tuan, Vo Tan Loc, Ha Dac Binh

eavesdropper which tries to extract information sent from

Alice without active attack. Let H denote the NbN a
channel matrix between Alice and Bob. Its entries are the
fading coefficients hij; 1iNb, 1jNa. An Nb1 vector h,
which is a column of H, is used to denote the channel
between the single selected transmission antenna and Nb
reception antennas. The single selected transmission
antenna NK; 1 K Na which maximizes the total received
signal power, is determined by
Nb
2

K argmax C j h ij
1 j N a
i 1

MN , j 1
b

f M , j M , j

( N b ) MN b

M , j
M

(4)

Where denotes the Gamma function.

The transmiter chooses the best antenna which achieves
the highest SNR by using (1). The instantaneous SNR of
TAS/MRC system is M max [ M , j ]. The PDF of M
1 j N a

has the following form

f M M

(1)

N a MN b 1
( N b )

N a 1

Nb
M

M
M

i0

N 1
a
i

M
M

(5)
i
N b 1 1 k
M


k 0 k ! M

Eve is capable of eavesdropping the signal sent by

Alice. The received signal z(t) at Eve is as follows
z(t) = hwx(t) + nw

(6)

where hw is the Nakagami fading coefficient between

the selected transmission antenna at Alice and the reception
antenna at Eve, nw is zero-mean AWGN with power Nw.
P | hW |2

The instantaneous SNR at Eve is W

the average SNR is W

Figure 1. System model

PE [| hW |2 ]
NW

NW

, while

. The PDF of W

We consider two scenarios: The legal/ illegal channels

respectively, are subject to 1) Rayleigh/ Nakagami fading;
2) Nakagami/Rayleigh fading.

is

A. The legal/ illegal channels are subject to Rayleigh/

Nakagami fading

B. The legal/ illegal channels are subject to Nakagami/

Rayleigh fading

The legal channel is assumed to undergo Rayleigh

fading, while the eavesdropper experiences Nakagami
fading. Alice sends the signal x(t) on the jth antenna, the
received signal at Bob y(t) = [y1, y2,, yNb ]T has the
following form

The legal channel is assumed to undergo Nakagami

fading, while the illegal channel is assumed to undergo
Rayleigh fading. Similarly, the PDF of M , ij is as follows

y(t) = hM,jx(t) + nM

(2)

where hM,j = [hM,ij, hM,2j,, hM,Nbj ]T is the jth column of

H, nM = [nM,1, nM,2,, nM,Nb ]T is the zero-mean additive
white Gaussian noise (AWGN) vector at Bob with power
NM, and superscript (.)T denotes the transposition
operator.
The instantaneous SNR and the average SNR at ith
P | hM , ij |2
antenna
at
Bob
are
and
M , ij
NM
2
PE[| hM ,ij | ]
M ,ij
respectively. P is the average
NM
transmission signal power at Alice. Assuming that M ,ij of
each link from Alice to Bob has the same value M . The
probability density function (PDF) of M , ij is
f i

M ,j

1
i
M,j

i
M
,j

M

(3)

The received signals at Bob are combined by using

MRC. Let M ,ij

M i 1 | hM ,ij |
Nb

be the instantaneous

SNR at Bob when using MRC. The PDF of M; j has the

following form

f W W

f i

M ,j

Mi , j

mm

Wm m

Wm 1 e

mm
Mi , j
m Mm

m W

(7)

m 1

i
m M
,j

(8)

The PDF of M , ij has the following form

f M , j M , j

m mN b MmN, bj 1
mN b MmN b

m M

(9)

The PDF of M is given by

f M M

im M

N a m mN b MmN b 1

mN b

mN b
M

mN b 1 1 m k
M

k 0 k ! M

m M

N a 1

1
i0

i
N 1
a
i

(10)

The PDF of W is as follows

f W W

W
W

(11)

3. Secrecy capacity analysis

A. Preliminaries
Channel capacity of link between two legitimate users is
C M log 2 (1 M )

(12)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

(13)

predetermined secrecy rate of transmission RS (RS > 0).

The secrecy outage event occurs when transmission rate is
below RS. In other words, at this time we cannot ensure the
secure transmission.

(14)

The legal/ illegal channels are subject to Rayleigh/

Nakagami fading: The secrecy outage probability of
Rayleigh/ Nakagami fading channels can be calculated as
follows.

Channel capacity of link to illegitimate user is

CW log 2 (1 W ).
The instantaneous secrecy capacity is given by
C S [C M CW ]
1 M

log 2 (
), M W

1 W
0, M W
B. Probability of Non-zero Secrecy Capacity

RS P CS RS 1

1) The legal/ illegal channels are subject to Rayleigh/

Nakagami fading: Assuming that the main channel and the
eavesdropper channel are independent of each other, we
can derive the probability of a non-zero secrecy capacity as
follows

f M M

f M M

N a 1 m 1

Wm m

m W

N 1
a
i

i 0 l 0

d W d M

N b 1 1 k
M e M 1 d M
k0 k ! M

1 Na u1 1 !mm 2jR bjl l m1!

1
u j
j! m Mj Wm2lm
i 0 p i j 0 l 0 (Nb ) i 1

i 1
M

i0

pk i

(16)

(m N b ) Mu2 3u2

kp

, and 3

k
0 k mN b 1

i 1 m
M

m .
W

C. Secrecy Outage Probability.

The secrecy outage probability can be defined as the
probability that the achievable secrecy rate is less than a

1
i 1 2
and 2
m ..
2R s
RS

' RS P ' C S R S

N a 1

u 2 1

i0

pk i
0 k mN b 1

j 0 l 0

f M W M , W d M d W

N a u 2 1 !

(m N

) i 1

u2 j

( b 4 ) l m 2 Rs

W l ! 4j 1 M

i
j

N ai 1

l p0 ,..., p m 1

R
pk
m i 11 2 S

1
M
e
0 k mN b 1 k !
where 4

pk
i

p0 ,..., p N b 1 0 k mN b 1 k !

where u2 mN b

(17)

The legal/ illegal channels are subject to Nakagami/

Rayleigh fading: Similarly, the secrecy outage probability
of Nakagami/ Rayleigh fading channels is given by

0 k N b 1

N 1
a
i

N a m u2 u2 1 !

pk
i112 S
i

j
1 M
p ,..., p e
Nb 1
l 0
0kNb 1 k!

Where b

f M W M , W d W d M

N 1
a
i

N a 1

u1 1 j

i!
.
and

,...,
p
p
!
!...
p
p
pN 1 !
N 1
0
2
0
2) The legal/ illegal channels are subject to Nakagami/
Rayleigh fading: This process is similar to the previous
one, we derive the probability of a non-zero secrecy
capacity as
P ' CS 0 P ' M W

Na 1

u1 1

k
0kNb1

pk
i

N ai 1

p0 ,..., p N b 1 0 k Nb 1 k !

pk
i

p0 ,..., pNb 1 0kNb 1 k!
i112RS

j
e M W b Wm1eW2 dW

(15)

N 1
a
i

N a m l u1 l 1 !

0 k N b 1

(Nb )

p0 ,..., pNb 1 0kNb 1 k!

N a ml

where u1 N b k pk , 1

a
i

k
0kNb1

( N b )l ! Mu1 Wl 1u1 l

pk i
0 k Nb 1

fM M fW W d MdW

1 Na u1 1!mm 2jR
1
u j
j! m Mj Wm
i 0 p i j 0 (Nb ) i 1

pk
i

1 u l 1

0 M1 e M 1 d M
p0 ,..., p N b 1 0 k Nb 1 k !

N a 1 m 1

pk i
0kNb 1

Na 1

( N b )l ! Mu1 Wl

pk i
0 k Nb 1

N a m l MN b l 1

( N b )l ! MNb Wl

N 1
a
i

N a 1 m 1
i 0 l 0

i 0 l 0

m 1
W

1 Na N 1

Na 1
i 0

Mu11 i1MM
e
dM fW W dW
y Nb
M

f W d W d M
m

P CS P M W
1

45

m i 1 2

RS

(18)

1 .
W

4. Numerical Results
In this section, we discuss some results based on the
theoretical analysis and Monte-Carlo simulations of the
probability of existence of non-zero secrecy capacity and the
secrecy outage probability of considered system in the effect of
various system parameters, such as average SNRs, Nakagami

46

Nguyen Van Tho, Van Phu Tuan, Vo Tan Loc, Ha Dac Binh

secrecy capacity and the secrecy outage probability for

Rayleigh/ Nakagami and Nakagami/ Rayleigh fading,
respectively with different shape parameter m for
W 10dB , Na= Nb =2. We can see that the secrecy

fading model, and number of transmission antennas.

A. Effect of average SNR

performance is better with increasing m when M

W.

Figure 2. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Rayleigh/ Nakagami, m=2,
Na=Nb=2, RS=1 bit/s/Hz)
Figure 4. The probability of non-zero secrecy capaciy and the
secrecy outage probability (Rayleigh/ Nakagami, W 10dB ,
Na=Nb=2, RS=1 bit/s/Hz)

Figure 3. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Nakagami/ Rayleigh, m = 2,
Na=Nb=2, RS=1 bit/s/Hz)

Figure 2 and Figure 3 show the probability of non-zero

secrecy capacity and the secrecy outage probability in two
scenarios: Rayleigh/ Nakagami fading P CS , RS
and Nakagami/

Rayleigh fading

P C , R ,
'

'

respectively, versus M for different W with the shape

parameter m=2, the number of transmission antennas
Na = 2 and the number of reception antennas Nb = 2. In
these figures, P (CS) and P(CS) increase, while O(RS) and

Figure 5. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Nakagami/ Rayleigh, W 10dB ,
Na=Nb=2, RS=1 bit/s/Hz)

C. Effect of the number of antennas

O(RS) decrease when Bobs SNR M increases, on the

contrary, P(CS) and P(CS) decrease, while O(RS) and
O(RS) increase with increasing W . These assessments are

resonable because when M increases, the received signal

at Bob is better than that at Eve so that the capacity of
legitimate users will be larger than the capacity of
illegitimate users. From these two figures, we can see that
the secrecy performance over Rayleigh/ Nakagami fading
channels is worse than Nakagami/ Rayleigh fading
channels. In other words, the secrecy performance is better
when the Nakagami fading is on the main link due to the
Line of Sight (LOS) component.
B. Effect of Nakagami fading model
Figure 4 and Figure 5 depict the probability of non-zero

Figure 6. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Rayleigh/ Nakagami, m = 2,
W =10dB, Nb=2, RS=1 bit/s/Hz)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

47

this considered system we can increase the number of

transmission antennas or the number of reception antennas
of legal devices.
As it can be observed clearly from above figures, the
secrecy performance is improved with: the increase in SNR
at Bob receiver or the decrease in SNR at Eve or the
increase of the number of antennas at Alice and Bob. The
good agreement between analytical and simulation results
verifies the correctness of our analysis.

Figure 7. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Nakagami/ Rayleigh, m = 2,
Nb=2, W =10dB, RS=1 bit/s/Hz)

5. Conclusion
In this paper, we focus on PHY secrecy performance
analysis of MIMO system using TAS/MRC in the presence
of a single antenna passive eavesdropper in two scenarios:
the main channel undergoes Rayleigh fading, while the
eavesdroppers channel is subject to Nakagami fading and
vice versa. The exact closed form expressions of
probability of non-zero secrecy capacity and the secrecy
outage probability have been derived and validated by
Monte-Carlo simulations. In addition, our results show that
the secrecy performance of the Nakagami/ Rayleigh fading
channels outperforms that of the Rayleigh/ Nakagami
fading channels due to the LOS component. Our results
also show that increasing the number of transmission
antennas or the number of reception antennas can improve
the secrecy performance of the considered system.
REFERENCES

Figure 8. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Rayleigh/Nakagami, m = 2, Na=2,
W =10dB, RS=1 bit/s/Hz)

[1]

C. E. Shannon, Communication theory of secrecy systems, Bell

Syst. Technol. J., vol. 28, pp. 656715, Oct. 1949.

[2]

A. D. Wyner, The wire-tap channel, Bell Syst. Technol. J., vol.

54,no. 8, pp. 13551387, Oct. 1975.
S. Leung-Yan-Cheong and M. Hellman, The gaussian wire-tap
channel, IEEE Trans. Inf. Theory, vol. 24, no. 4, pp. 451456, July
1978.

[3]

[4]

H. Alves, R. D. Souza, M. Debbah, and M. Bennis, Performance of

transmit antenna selection physical layer security schemes, in IEEE
Signal Process. Lett., vol. 19(6), 2012, pp. 372375.

[5]

N. Yang, H. A. Suraweera, I. B. Collings, and C. Yuen, Physical layer

security of TAS/MRC with antenna correlation, IEEE Transactions
on Information Forensics and Security, vol. 8(1), pp. 254259, 2013.
N. Yang, P. L. Yeoh, M. Elkashlan, R. Schober, and I. B. Collings,
Transmit antenna selection for security enhancement in MIMO
wiretap channels, IEEE Transactions on Communications, vol.
61(1), pp. 144 154, 2013.
A. P. Shrestha and K. S. Kwak, Performance of opportunistic
scheduling for physical layer security with transmission antenna
selection, EURASIP Journal on Wireless Communications and
Networking, vol. 2014:33, pp. 19, 2014.

[6]

[7]

Figure 9. The probability of non-zero secrecy capaciy and the

secrecy outage probability (Nakagami/Rayleigh, m = 2, Na=2,
W =10dB, RS=1 bit/s/Hz)

Figure 6, Figure 7, Figure 8 and Figure 9 illustrate the

variation of the probability of non-zero secrecy capacity
and the secrecy outage probability with respect to the
number of transmission antennas Na and the number of
reception antennas Nb in two approaches: Rayleigh/
Nakagami and Nakagami/ Rayleigh respectively. When Na
or Nb increases, the secrecy performance becomes better.
Obviously, in order to enhance the secrecy performance of

[8]

L. Wang, N. Yang, M. Elkashlan, P. L. Yeoh, and J. Yuan, Physical

layer security of maximal ratio combining in two-wave with diffuse
power fading channels, IEEE Transactions on Information
Forensics and Security, vol. 9(2), pp. 247258, 2014.

[9]

D.-B. Ha, N. G. Nguyen, D.-D. Tran, and T.-H. Nguyen, Physical layer
security in UWB communication systems with Transmit Antenna
Selection, in The 2th IEEE International Conference on Computing,
Managements and Telecommunications 2014 (ComManTel 2014),
DaNang, Vietnam, April 27-29, 2014, pp. 280285.

[10] V. T. Tan, D.-B. Ha, and D.-D. Tran, Evaluation of physical layer
security in MIMO ultra-wideband system using time-reversal
technique, in The 2th IEEE International Conference on Computing,
Managements and Telecommunications 2014 (ComManTel 2014), Da
Nang, Vietnam, April 27-29, 2014, pp. 7074.

(The Board of Editors received the paper on 07/09/2015, its review was completed on 10/23/2015)

48

Hoang Van Thanh, Luu Duc Binh

STUDY OF STUDENT ENGAGEMENT LEVELS IN CLASS ACTIVITIES

Hoang Van Thanh, Luu Duc Binh
The University of Danang, University of Science and Technology; hvthanh@dut.udn.vn

Abstract - This paper presents the results obtained from

application of active learning m ethods into lectures to students of
Mechanical Engineering classes. The traditional methods of
teaching make students become m ore and more passive when
they are in class. This leads to serious consequences that directly
affect students achievem ent. Firstly, students will easily fall asleep
during the lesson or lecture. Secondly, students find it very difficult
to understand the lessons and concentrate on their study. Finally,
students will lack comm unicative com petence after graduating
from university. In order to make students get involved in class
activities, the author will suggest some useful techniques, including
group discussion, learning by teaching and giving marks to
students who have good answers and presentations in front of the
class. To examine the reliability of these techniques, one survey
was conducted with a questionnaire. The result shows that group
discussions and interactions between students and teacher are two
im portant factors for student engagement in the classroom with
assent of about 87,5% of total number of students taught.
Key words - student engagem ent; active learning; think-pairshare; comm unication skill; educational changes.

1. Introduction
Education is considered as the shortest way to success.
It can be seen that in some developed countries education
attracts huge investment. As a result, their economy has
been developing dramatically. The purpose of education is
to bring success to students. A question is posed is that who
is responsible for their success. Obviously, educators,
government, employers and students all are factors that
accelerate the development of education system. Before the
international integration, domestic and international
employers often seek to hire staff with 'strong interpersonal
skills' - they want people who will work well in a team and
be able to communicate effectively with colleagues,
customers and clients. What should we do to help these
students succeed in their study? As a lecturer, the author
always work out new techniques to improve the quality of
teaching and learning at university, especially enhance
student engagement in the classroom.
With credit training programs, students have fewer
classes than before. Therefore, teachers only have enough
time to provide students with fundamental knowledge at
school and students have to spend a lot of time studying at
home. Since new teaching program was applied into
specialized classes, however, our observation reveals that
few students pay attention to their study. According to
traditional methods of teaching, especially teachercentered method, teachers only provide knowledge for
students and students only copy what they can get from
their teachers. As a result, students are becoming more and
more passive when they are in class. They do not want to
make questions or raise their voices in front of their
teachers and their classmates.
Student engagement has been studied for more than

twenty years. Engagement refers to a students

willingness, need, desire and compulsion to participate in,
and be successful in, the learning process.
Disengagement revolves around minimal participation,
motivation and misbehaviour amongst students as well as
limited group interaction [1]. Previous studies that show
the way to encourage student engagement in tertiary
institutions are dependent on specific majors, learning
environment, gender as well as kinds of students. Russell
B. et. al. employed a conceptual model with four factors
including motivation and agency, transactional
engagement, institutional support and active citizenship
to analyse data of student engagement. The results show
that the first three factors were significant in student
learning [2]. Schuetz has mentioned student motivation
and their effort as a main strand in engaging students [3].
Some reported that student engagement was related to
student, teacher, school and family variables and their
combination. The study indicated that engagement levels
among young females are higher than those among males.
Furthermore, parental education played an important role
in engagement. Students who have better levels of
parental education were higher in engagement than those
from families with worst levels of parental education [4].
Engagement is an important issue in teaching because
it affects how much students learn on a daily basis.
Quality learning depends greatly on students interest and
teachers ability to create a motivating environment. It
can be seen that many new teaching techniques have been
applied to teaching in the classroom not only in the world
but also in Vietnam so far. The reality shows that these
techniques are not well-applied to teaching and learning
of teachers and students at Mechanical Engineering
classes. Many teachers admitted that they had trouble in
explaining the lessons to students in the classroom
because these students did not concentrate on their study
and teachers explanations. Even they do not want to go
to class. Another aspect indicates that students are in the
class but they have anxiety when teacher asks questions.
It is obvious that to attract students attention to class
activities is a big challenge to teachers at Mechanical
Engineering Department at DUT. This paper examines
the effectiveness of some new techniques applied in
teaching and learning such as Think-Pair-Share,
Group discussion, Learning by teaching and
interaction between teacher and students.
2. Literature review
2.1. Think-Pair-Share
This technique will be presented in Table 1

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Table 1. Think-Pair-Share technique

What teachers do?

What students do?
Topic selection: Choose a
situation where the process of
reflection
and
shared
discussion would bring deeper
understanding, and insert a
brief Think/Pair/Share activity
into the lesson at that point.
Group
selection:
With Remember their numbers
students seated in teams of 4,
number them from 1 to 4
Entrance: At the end of this Listen
to
teachers
class today, we are going to directions
do a TPS activity
and
ask
their
Instruction: Here is your Listen
for
further
challengeRecite
the teachers
problem statement: Example: information
Students explain working Read the text
principle of the diagram
- Pause for question or confusion
- Deliver worksheets to students
Introduction to TPS technique Listen and ask their teacher
for more information
Send them on their way:OK- Prepare for TPS
Are you ready?-OK-Here we
go-Think-Pair-Share
Give students at least (1- Formulate thoughts and
3min.) of think time to ideas, write themdown if
THINK of their own answer necessary to prepare for
sharing with a partner.
Using
student
numbers, Practise good active listening
announce
discussion skills when working in pairs,
partners. (Example: For this using techniques such as
discussion, Student #1 and #2 paraphrasing what the other
said,
asking
for
will be partners. At the same has
time, Student #3 and #4 will clarification, and orally
clarifying their own ideas.
talk over their ideas.)
Observation:
Look to see how the
conversation is going.
Move towards those students
who are not engaging.
Ask them the question personally.
Give positive comments:
Thats a good point, Good
idea and Well-done.
Exit: OK.Times up
Reflection: Id like to hear Pinpoint any information
from a few group, students that is still unclear after the
Finally, randomly call on a pair discussion, and ask the
few students to SHARE their class and teacher for
clarification
ideas with the class
Take notes
Give them remarks
Transfer: We will have a
more interesting TPS activity
next week.

49

2.2. Learning by teaching and interaction between

teacher and students
According to J. P. Martin (2008), learning by teaching is
a method where students learn by teaching each other.
Students actively research a topic and prepare the
information so that they can teach it to the rest of the class
[5]. After Think-Pair-Share Activity, teacher has to check
students understanding about the given topic. Normally,
some students do not understand the lesson clearly.
Therefore, teacher should raise a question Who can explain
the lesson againto help other students understand. This
technique is called learning by teaching that demonstrated
specifically as shown in Table 2.
Table 2. Learning by teaching technique

Teacher activities
Students activities
Teacher chooses a topic Listen to teachers directions
related to the lesson and
deliver the topic to students
Introduce
Learning
by Listen
teaching method and objective
Give students about 5-7 min Students study and prepare
information according to the
topic requirement
Select volunteered students Be willing to present
for teaching the topic to the
rest of class.
Give students remarks
Write down something
3. Method
This study was carried out at classes belonging to
manufacturing technology, mechatronics and transportation
mechanical engineering. Author applied active learning
techniques into teaching in the classroom, especially
Think-Pair-Share, Group discussion, Learning by
teaching and interaction between teacher and students.
In each session, the teacher selects a topic and requires
students to answer questions related to the issue. In this
case, the teacher can pick students randomly, asking them
to stand up and deliver their speech. This activity enables
every student to prepare and think about the topic. In order
to stimulate students to give opinions, teacher gives extrapoint to students who have good ideas. With this way,
students have good chances to interact with instructor so
that they can develop their communication and criticism
skills. To confirm how to get student engagement, a
survey was performed at mechanical and transportation
mechanical engineering classes.
3.1. Research question
It is clearly seen that there are a lot of factors affecting
student engagement in the classroom such as motivation,
methods of teaching, family, school, social environment
and others. In this study, the author only focused on
learning and teaching activities in the class. Therefore, my
main research question here was Which factors affect
student engagement in the classroom?.
3.2. Participants
The author selected two third-year student classes for

50

Hoang Van Thanh, Luu Duc Binh

my research. The students in this grade were mostly twenty

years old. Two groups of students that belong to
mechanical transportation engineering department and
mechanical engineering department, DUT were invited to
take part in the survey as informants. 64 students coming
from naval engineering class called group 20, and 57
students from mechanical engineering one called group 05.
3.3. Data collection
At the end of the semester, the author delivered a survey
sheet to the students of two groups: group 20 and group 05.
The survey sheet focused on student engagement in
teaching techniques. There were four teaching techniques
selected for this study including only chalk, only slides
show, chalk & slides show combination, chalk & slides
show & group discussion & teacher and student interaction
combination. In each teaching technique, depending on
agreement or disagreement, students can mark x in
respective columns as depicted in Table 3.

student interaction combination have had a large

contribution into increasing student engagement in the
classroom.
The following pictures describe different student
engagement levels in teaching and learning. Firstly, teacher
represented a diagram, after completely drawing the
diagram on the blackboard, teacher only explained some
basic signs and required student to state working principle
regarding to the diagram. Teacher raised a question, Can
anybody explain working principle of this model. The
result shows that no student raised their hand although
teacher had stimulated three times as shown in Figure 2.

Table 3. Survey sheet of student engagement

Figure 2. Nobody raised their hands

4. Results and Discussion

Figure 3. Think-pair stage used

Figure 1. Survey results of teaching methods

affecting student engagement

This result was done with third-year students in

mechanical and naval engineering at Department of
Mechanical Engineering and Transportation Mechanical
Engineering, Danang University of Science and
Technology. Figure 1 shows the relationship between
number of students and teaching techniques with group 20
and 05. The vertical axis indicates the number of students
in each group. The horizontal axis represents the teaching
techniques. The result shows that techniques of teaching
with chalk, slides show, group discussion and teacher and

Figure 4. Only two students raised their hands

After using TPS technique, the teacher encouraged

students to think about the given issue individually within
one minute and then asked two students to have small
discussion within three minutes as shown in Figure 3. In
general, most students did well in this stage. However, some
students were not familiar with this method. That teacher
should do in this situation is to come to those students and

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

51

stimulate them to follow this method. After this step, teacher

made the same question again and there were two students
who raised their hands immediately as shown in Figure 4.

Figure 9. Six students volunteered using learning by teaching

Figure 5. A large group discussion

It can be seen that the number of students who raised

their hands was too few compared to the total number of
64 students in that class. In order to enlarge students
interaction ability, a large group discussion which took
place as depicted in Figure 5 represented the increase in
number of volunteered students (Figure 6). Thumbs up,
thumbs down and thumbs side were used to check students
understanding level as shown in Figure 7. The result of this
step shows that there were three students who did not
understand the lesson as represented in Figure 8. The next
technique employed was learning by teaching. Teacher
asked the students who understood the matter to teach this
topic to students who did not have clear understanding.
There were only 06 students raising their hand as shown in
Figure 9. It is clear that not all students who understood the
matter could explain it again to other students. This proved
each student had different understanding level.
5. Conclusion

Figure 6. A lot of Students had their opinion

Figure 7. Understanding level checking of Students

Figure 8. Three students understood the lesson

This paper presented the results of active teaching

activities and reported responses from a questionnaire
about student engagement. The work was carried out with
the 3rd and 4th year- students of manufacturing,
mechatronics and naval engineering that belong to
Mechanical Engineering and Mechanical Transportation
Engineering Departments at DUT, Vietnam. Interviews
with students at the beginning of this semester show
students in the class are very passive. They just go to class
for attendance and only copy what their teacher writes
down on the blackboard. One thing needs to be considered
is that they are not willing to be at school. Active learning
used at mechanical classes has attracted more students
attention to class activities. The number of students coming
to class and their understanding level also increase
considerably. Teaching method with group discussion and
interaction between students and teacher plays an
important role in student engagement in the classroom with
agreement accounting for 87,5% of total number of
students taught in semester II, schoolyear 2014-2015.
Furthermore, interaction between students and students,
students and teacher and learning by teaching are the
factors which help students improve their communicative
skills and academic achievements at university.
Acknowledgment: The author would express their
acknowledgement to classes of group 20, 05, 02, 03,

52

Hoang Van Thanh, Luu Duc Binh

semester II, year 2014-2015 from Department of

Mechanical Engineering and Mechanical Transportation
Engineering, DUT for doing the study.
REFERENCES
[1] Bomia L., Beluzo L., Demeester D., Elander K., Johnson M., Shedon
B., "The impact of teaching strategies on intrinsic motivation, ERIC
Clearinghouse on Elementary and Early Childhood Education,
Champaign, IL (1997).
[2] Russell B., Slater G. R. L., Factors that encourage student

engagement: Insight from a case study of first time students in a

New Zealand University, Journal of University Teaching &
Learning Practice, Vo. 8, Issue 1, Article 7 (2011).
[3] Schuetz P., A theory-driven model of community college student
engagement, Community College Journal of Research and Practice,
32, 305-324 (2008).
[4] Bempechat J., Shernoff D. J., Parental Influences on Achievement
Motivation and Student Engagement, Handbook of Research on
Student Engagement, 315-342 (2012)
Martin S., An Analysis of Collaborative Models for Learning to
Teach Science, World of Science Education: North America. Sense
Publishers (2008).

(The Board of Editors received the paper on 06/01/2015, its review was completed on 06/23/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

53

OPTIMIZING P.I.D PARAMETERS IN CONTROL ACCELEROMETERS

AND GYROSCOPESIN SELF - BALANCING QUADROTORS
Vu Van Thanh, Huynh Thanh Tung
The University of Danang, University of Science and Technology; httung@dut.udn.vn, vuvanthanh85@gmail.com

Abstract - The algorithm that calculates PID controller consists of

three separate parameters, so sometimes it is also calledthree stage
control: theproportion, integral and derivative values, referred to as
P, I, and D.The proportion value determines the impact of the current
uncertainty, the integral value determines the total impact of past
errors and the derivative value determinesthe value of the differential
impact of error variable speed. Total short of three effects are used
to adjust the process via a control element such as the position of the
control valve or the source of the heating element [1].In this paper
the authors find and optimize 3 constants in the algorithm of the PID
controller.The controller can be used in the designs that have special
requirements. The response of the controller can be described in
terms of the sensitivity of the controller error. The error values are
compared with setpoint value of the controller and the value of
fluctuations of Quadcopters.
Key words - PID digital; self-balancing robots; Quadrotor; IMU; optimize.

design a completed discrete digital control system that will

provide the necessary stability.This paperal so designs a
control system to balance the quadrotor using a 6-axis IMU
sensor (MPU-9150) and TivaC SeriesTM4C123GXL
microcontroller applied to PID control algorithm with
optimal parameters.
The rest of the paper is organized as follows. Section 2
will describes the design of the quadrotor. Design of
control unit for quadrotor are presented in section 3.
Section 4 will study how to optimize PID parameters.
Finally, section 5 provides some final conclusions
2. Designing quadrotor model

1. Introduction
In recent years, quadrotor and mobile robotics
technology has gained popularity in both commercial and
military use.There are a lot of techniques suggested to
increase robotic mobility on dynamic environments. In
particular, the most common techniqueis used to provide
greater mobility to a robot platform based on inverted
pendulum model.Quadcopter is operated by thrust that is
produced by four motors that are attached to its body. It has
four input forces and six output states (x, y, z, , , ) and
it is an under-actuated system, since this enables
Quadcopter to carry more load [1]. Quadcopter has the
advantages over the conventional helicopter because the
mechanical design is simpler. Besides, Quadcopter
changes direction by manipulating the individual
propellers speed and does not require cyclic and collective
pitch control [1],[2]. Nowadays, the research related to
Quadcopter covers the areas of design, control, stability,
communication systems and collision avoidance.
Reference [3] focused their study on the 3-DOF attitude
that control free-flying vehicles. The characteristic is heavily
coupled with inputs and outputs, and the serious non-linearity
appears in the flying vehicle and due to this non-linear control,
appears multi variable control or optimal control for the
attitude control of flying Quadcopter. Reference [4] worked
on intelligent fuzzy controller of Quadcopter. A fuzzy control
is designed and implemented to control a simulation model of
the Quadcopter. The inputs are the desired values of the
height, roll, pitch and yaw. The outputs are the power of each
of the four rotors that is necessary to reach the specifications.
Simulation results prove the efficiency of this intelligent
control strategy. References [5], [6] have done research to
analyze the dynamic characteristics and PID controller
performance of a Quadcopter.
This paper will provide the techniques involved in
balancing an unstable robotic platform.The objective is to

Figure 1. Main block diagram of Quadrotor

From Figure 1, the important parts of Quadrotor are

included: Frame (includes motor and fans), controller,
signal transceivers and battery source. With the target of
designing a Quadrotot that is able to carry 2kg of load,
flight time of at least 15 minutes, the mechanical
structureof the frameis designedas follows:

Figure 2. The forces applying on Quadrotor

From Figure 2: The main system gravity P = mg, and

M isrotation momentum of motor. Force from propellers
while rotatiing:
TMT=2
:

air density, S (

(N)
) Area of propellers

(1)

54

Vu Van Thanh, Huynh Thanh Tung

Load each propeller can carry:

With

TMT = WP =

(2)

while m is the weight, g l earth gravity (g=9.8).

Based on the principles of aerodynamics we can
calculate Quadrotor condition to lift off the ground.Area of
propellers must conform to lift the plane dressed. We have
an area of propeller S = *(D2/4), with D as rotor diameter.
Choose D = 0.33m to meet our design [9].
To satisfy the given parameters and calculations, we
would choose the following components: Engine Tarot
4006-620KV as Figure 4a, with the given parameter.
Speed:620 rpm/v
Power: 1000W
Battery: 4 or 5cell of lipoly at least 19V supplied

sensor and motor driver ESC IMU sensor which consists

of accelerometer and gyroscope gives the reference
acceleration and angle with respect to ground (vertical
direction), and the encoder which is attached to the motor
gives the speed of the motor. These parameters are taken
as the system parameter and determine the necessary
external forces to balance the quadrotor.
In this paper, to control Quadrotor altitude motion, PID
controller has been developed and embedded in Tiva C
Series TM4C123GXL microcontroller. PID control will
maintain a stable equilibrium for Quadrotor when flying in the
air, orbe affected by external forces such as winds ... based on
the read value through the sensor MPU9150. MPU 9150 will
provide Accel, Gyro, Mag to controllers to calculate the 3
values of angles Roll, Pitch, Yaw. Figure 6 shows ablock
diagram of the control system for Quadrotor

Maximum current: 30A

Figure 4. a) Engine HP2217-930KV, b) Propellers for the Quadrotor

Eliminating torque by the rotary engine, we produce 1

pair of clockwise rotation and 1 pair of counter clockwise
rotation motor as Figure 4b. So we chose two types of
structure opposite wing. Wings are called pros and cons
propellers. The frame is made of 2 aluminum 10mm x 15mm
and 1mm thickness with high strength properties. Moving
quadrotor safely, we must ensure the gap between the
propellers. So the length of the aluminum bar must be greater
than (Dpropeller// 2) with D as rotor blades (254mm length).
So we choose the 550mm for 2 aluminum bars[9].

Figure 6. Block diagram of the quadcopter hovering system

3.1. The PID theory

From these requirements and reliability we have the

following design parameters:
-Aluminum frame cross 550mm x 550mm length.
-Impeller type 10x4.5with pros and cons 2 wings each.
- The square phip substrate size100mm x 100mm.
- Triangular tripod 50mm x 100mm square size.
- The maximum weight of Quadrotor <2kg.
-Tarot 4006-620KVengine.

Figure 7. The PID Theory

The PID controller algorithm involves three separate

constant parameters, and is accordingly sometimes called
three-term control: the proportional, the integral and
derivative values:
PID=P+I+D
P depends on the present error; I depends on the
accumulation of past errors; D is a prediction of future
errors, based on current rate of change.
PID can be described by equation:
t

u(t ) K pe(t ) Ki e( )d Kd
0

a)
b)
Figure 5. a) Design patterns in solid, b) Final Model Quadrotor

3. Designing control unit for Quadrotor

The basic mechanical design includes invenSense
MPU-9150: 3-axis gyro, 3-axis accelerometer, 3-axis
compass 4 Bruhshless DC motor, one Tiva C Series
TM4C123GXL microcontroller, IMU (inertial mass unit)

d
dt

e(t )

(3)

The weighted sum of these three actions is used to

adjust the process via a control element
Kp: Proportional gain, a tuning parameter; Ki: Integral
gain, a tuning parameter; Kd: Derivative gain, a tuning
parameter
e: error; t: Time or instantaneous time (the present);
: Variable of integration; takes on values from time 0 to
the present (t)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

55

3.2. The influence of P.I.D gains in Quadrotor

3.2.1. Proportional Gain
Proportional control applies an effort in proportion to
how far you are from the set-point. Its main drawback is
that the closer you get to the set-point, the less it pushes.
Eventually it does not push hard enough to move the
variable, so the process can run continuously close to the
setpoint, but is not quite there.
3.2.2. Integral Gain
Integral control tries to even out the difference of the
time spend on both sides of the line. If youve spent a
minute running at 98%, it will try to push you over to 102%
for similar amount of time. This action compensates for Ps
inability to make that last effort
3.2.3. Derivative Gain
Derivative acts as a brake or dampener on the control
effort. The more the controller tries to change the value, the
more it counteracts the effort. In our example, the variable
rises in response to the set-point change, but not violently.
As it approaches the set-point, it settles in nicely with a
minimum of overshoot
3.3. Building algorithm chart for the controlling quadrotor

Figure 9. Time interrupt PID function

4. Result
In order to find the optimal value for quadrotor, we
change the 3 values Kp, Ki [8] based on The ZieglerNichols closed loop method.This is based on experiments
executed on an established control loop (a real system or a
simulated system. The tuning stepsare as follows:
Bring the process to (or as close to as possible) the specified
operating point of the control system to ensure that the
controller during the tuning is feeling representative process
dynamic and to minimize the chance that variables during the
tuning reach limits. You can bring the process to the operating
point by manually adjusting the control variable, with the
controller in manual mode, until the process variable is
approximately equal to that of the setpoint.
Turn the PID controller into a P controller by setting set
Ti = and Td = 0. Initially set gain Kp = 0. Close the
control loop by setting the controller in automatic mode.
Increase Kp from 0 to a critical value Kpu at which the
output first exhibits sustained oscillations with period Pu.
(Pu is measured in sec.)
Measure the ultimate (or critical) period Pu of the
sustained oscillations (In this paper, we chose Pu <2s).
Calculate the controller parameter values according to
Table 1, and use these adjustment parameters in the
controller to optimize the system.
Table 1. Formulas for the controller parameters in
the Ziegler-Nichols closed loop method.

Kp
PID

Figure 8. The main control program

Figure 8 shows the flowchart for the main program with

the main purpose to initialize the values of PWM
(originalpulse valueformotor control), declare I2C standard
to connect with MPU 9150 sensor, initialize the library for
sensors, and initialize data converters by Direction Cosine
Matrix (DCM).The program will beallowed to interrupt
timer A after sampling time t to perform PID function.
Infinite loop will be performed involving waiting time to
read data from sensors to calculate the angles Roll,Yaw,
Pitch to supply the PID function. The Interrupt TimerA
function of PID is given in Figure 9, with the main task to
update velocity values of 4 motors to balance Quadrotor.

Ti

0.6Kpu

~1
2
According to the results described below

Td
~0.25

4.1. All gains to 0 (Kp=Kd=Ki =0)

This condition means that there is no PID control to quadrotor.

Figure 10. All gain to 0(Kd=Ki=Kp=0)

56

Vu Van Thanh, Huynh Thanh Tung

Observation: the max value is 6 and the min value is

around -6 degree. However, the trend of this oscillation
value makes the system one side deviated, this also makes
Quadrotor fall down.
4.2. Increase the P gain until the steady oscillations occurs.
During increasing Kp from 0 to stable oscillation value,
the most striking points are 2 values Kp= 1.5 and Kp=2.
The comparison andevaluation of those results are
described at Figure 10 below.
With Kp=2 the Quadrotor oscillates heavily from the
equilibrium point. (Around -40 to 40). Using this
result,Quadrotoris strongly shaking, but still remains balanced.
At Kp=1.5 the oscillation of Quadrotor is more stable
at 5 degree from -10 to 10. This value is the most suitalble
for Kp parameter, though it still has one side inclined.

Figure 12. Increase the Kd to 1

Observation: the max value is 80 and the min value is

around -80 degree.
Table 4. Result of Kd=1

Overshooting

N/A

Setting Time

N/A

Error

Maximum 80

From the Figure 12 and Table 4, we realize that the

value Kd=1 is too high, so we decrease 10% of the last
value. However, the oscillations still exist. We decrease it
to just 0.05 that means 5% of the last value.At this point
the oscillation does not disappear but with Kd = 0.03, we
get the best performance of quadrotor.
4.4. Increase theI gain until it brings you to the set point
with the number of oscillations desired
We start by putting Ki = 0.5. However the angle error is
notreduced to zero, besides the Quadcopter oscillates again.

Figure 11. At Kp= 2 andKp=1.5

Observation: with Kp=2 the maximum value is 60 and

the minimum value is around -40 degree. This is strong
oscillation, so that we can set the maximum 40. The result
is described below. Moreover, with Kp=1.5: the maximum
value is 5 and the minimum value is around -15 degree and
the motion is quite harmonic.
Figure 13. Increase the Ki value to 0.5

Table 2. Result of Kp=2 and Kp=1.5(Kd=Ki=0)

Kp=2

Kp=1.5

Observation: The angle error is stable at around -12 degree.

Table 5. Result of Ki=0.5

Overshooting

20 degree

Setting Time

Around 10 time unit Around 15 time unit

Overshooting

-14 degree

Maximum 40

Setting Time

20 time unit

Error

Maximum -17

Error

15 degree
Maximum 15

From above result, we notice that at Kp=1.5 oscillation

is quite steady
4.3. Increase the D gain until the the oscillations go away.
We set the Kd to 1 and the Quadcopters behavior is
unpredictable.

4.5. The optimum parameter adjustment:

Therefore we put Ki just less than Kd, Ki = 0.01. The
Quadcopter is stable at the angle error around 2.3
degree,which is very good.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

57

and setting time is not so quick, these gains are our best effort,
and we can control the balance of quadrotor quite well. In
the future,we can apply this method not only in selfbalancing quadrotors but also in balanced auto robots
REFERENCES

Figure 14. The optimum parameter

[1]

A. Z. Azfar and D. Hazry, Simple GUI Design for Monitoring of a

Remotely Operated Quadcopter Unmanned Aerial Vehicle,
Proceeding of the 7th International Colg and its Applications
(CSPA), 23-27.

[2]

K. W. Weng, Quadcopter, Robot Head to Toe Magazine, Vol. 10,

2011, pp. 1-3.
D. Park, M.-S. Park and S.-K. Hong, A Study on the 3-DOF
Attitude Control of Free-Flying Vehicle, Proceeding of the IEEE
International Symposium on Industrial Electronics (ISIE), Pusan,
12-16 June 2001, Vol. 2, pp. 1260-1265.
M. Santos, V. Lpez and F. Morata, Intelligent Fuzzy Controller of
a Quadrotor, Proceeding of the IEEE International Conference on
Intelligent Systems and Knowledge Engineering (ISKE), Hangzhou,
15-16 November 2010, pp. 141-146.

[3]

Table 6. Result of optimum parameter Kp,Kd,Ki

Overshooting

1 degree

Setting Time

22 time unit

Error

Maximum -2

From the achieved results, Quadrotor has stable

equilibriumin flight with minimum vibration though a
slight drift caused by the offset between the system and
ground of 2 degrees. This case can be overcome by
utilizing the GPS data to update the coordinates to find the
suitable position against the drift when Quadrotor is flying.
5. Conclusion
This paper proposes the method of adjusting the value in
optimizing P.I.D controller in gyroscopes and
accelerometers by applying The Ziegler-Nichols closed
loop method in the experiment.With method of trials and
errors, we come out of three P.I.D gains Ki = 1.5, Kd = 0.03
and Ki = 0.01. Despite the fact that the error is still not zero

[4]

[5]

I. Morar and I. Nascu, Model Simplification of an Unmanned

Aerial Vehicle, Proceeding of the IEEE International Conference
on Automation Quality and Testing Robotics (AQTR), Cluj-Napoca,
24-27 May 2012, pp. 591- 596

[6]

[13] J. Li and Y. T. Li, Dynamic Analysis and PID Control for a

Quadrotor, Proceeding of the International conference on
Mechatronics and Automation (ICMA), Beijing, 7-10 August 2011,
pp 573-578.

[7]

P. Cominos and N. Munro, PID controllers: recent tuning methods

and design to specification, IEE Proceedings - Control Theory &
Applications 149, 2002, pp. 4653.
D. Wang and Qing-Guo, "PID tuning for improved performance."
Control Systems Technology, IEEE Transactions on 7.4, 1999. pp.
457- 465.

[8]

[9]

A. L. Salih, M. Moghavvemi, H. A. F. Mohamed, and K. S. Gaeid,

Flight PID controller design for a UAV quadrotor, Scientific
Research and Essays, vol. 5, pp. 33603367, 2010.

(The Board of Editors received the paper on 10/20/2015, its review was completed on 26/11/2015)

58

Pham Xuan Thuy

A METHOD FOR FRUITS RECOGNITION USING IMAGE PROCESSING TECHNIQUES

Pham Xuan Thuy
Hanoi Le-Quy-Don Technical University; thuy.phxuan@gmail.com
Abstract - Pattern recognition has been, and continues to be, the
subject for extensive research and development due to its wide range of
applications in the real life. In this paper, we introduce a fruits recognition
system that can recognize some types of fruits, which are common food
in our life, and make corresponding sounds. By analyzing image
processing algorithms and taking into account the features, which are
extracted from the images for classification purpose, we establish a vision
system using a few feature set. The output of our system is a saying,
which matches the result of the classifier a correct kind of the fruits. In
order to validate the effectiveness, we have tested the designed system
with some plant fruits such as apple, kiwi, lemon, orange, strawberry, and
tomato in the supermarket with a success rate of around 91%, and haved
compared this system with other works in recent years.
Key words - recognition system ; im age processing algorithm s;
m inim um distance classifier; extracted features; sounds

1. Introduction
In recent years, several researchers have implemented
fruit recognition systems. Harsh S Holalad, Preethi Warrir
and Aniket D Sabarad [1] developed a fruit identification
system based on FPGA technology. They built a system
that identified fruits such as apple, banana, sapodilla, and
strawberry. However, they used offline images for training
and testing because of no camera interfacing. In addition,
three features including mean, variance, and shape were
used to characterize the fruits with an accuracy of 85%
only. Moreover, the output of the system was LEDs. Woo
Chaw Seng, Seyed Hadi Mirisaee [2] proposed a
classification system for fruits spherical pattern. In this
system, the combination of three different features,
including color, shape, and size,was designed to perform
the sequential pattern recognition. The results of
experimentation were greatly affected by the fruit size
scalar values, which were selected by users. The accuracy
of the recognition mainly depended on the number of fruit
images for each type of fruit, collected and used to test the
system. Dr Vishwanath. B. C, S. A. Madival, Sharanbasa.
Madole [3] proposed a methodology for recognition and
classification of fruits in fruit salad image samples. The
fruits including apple, chikku, banana, orange and
pineapple, were considered. A K-mean classifier was used
and had the classification efficiency of around 98%. The
features using for classifying the different kinds of fruits
were color and texture the property that represents
surface and structure of an image. R. M. Bolle, J. H.
Connell, N. Haas, R. Mohan, G. Taubin [4] presented an
automatically-produce-ID system, intended to ease the
produce checkout process. In this system, a variety of
features such as color, texture (shape, density) was
extracted and integrated to classify the products. The
experiments were performed in several supermarkets and
grocery stores, where are hostile, robust, and rugged
environments in terms of images. S. Arivazhagan, R.
Newlin Shebiah, S. Selva Nidhyanandhan, L. Ganesan [5]
also proposed an efficient fusion of color and texture

features for fruit recognition. This was done by using the

minimum distance classifier based upon the statistical and
co-occurrence features derived from the Wavelet
transformed sub bands. The method used to analyze images
has many potential applications for automated agricultural
tasks. Jyoti A Kodagali and S Balaji [6] presented the
recent development and application of image analysis and
computer vision system in an automatic fruit recognition
system. The paper revealed that still much of work needed
to be concentrated on the fruits recognition, for instance,
the quality of the lighting system, image processing, the
recognition performance, and ease of use. A. M. Aibinu,
M. J. E. Salami, A. A. Shafie, N. Hazali and N. Termidzi
[7] proposed a method for the development of automatic
fruit identification and sorting system by using both
artificial neural network (ANN) and Fourier descriptor
(FD) techniques. The features that were used to recognize
and sort different fruits were color and shape.
From the published research works, it is observed that
color, shape, and texture are frequently used and have high
accuracy of the recognition system. However, the solution
of designing a fruit recognition system still needs to be
improved. In this work, we have performed a fruit
identification system. We have considered six types of
fruits, including apple, kiwi, and lemon, orange,
strawberry, and tomato as typical fruit samples. The system
recognizes given 2D query fruit image by extracting
features, including color, shape, texture, size (perimeter
and area) and computing their values to measure the
distance between the computed values of the query image
with the stored mean values of training fruits. A minimum
distance classifier based on the Euclidean distance is then
constructed. Finally, the sound that corresponds to the
certain kind of fruits is spoken out.
The rest of the paper is organized as follows: Section 2
describes the implementation of the system, which gives
some basic information about the system, the proposed
methodology, feature selection, feature extraction,
designing classifier and sound player. Section 3 gives the
results of implementing system, including the results of
different stages and the final output. The discussions and
conclusion of this work are given in section 4 and section 5,
sequentially.
2. The implementation of fruits recognition system
2.1. Hardware description
The system of our design included five components:
lighting system, a camera, personal computer (including
loudspeaker), real-time controller, and software, which were
connected to each other with proper settings (Figure 1).
The first component is the lighting system. The system
was carefully designed and set up because it mainly affects

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

59

Then, almost small particles and the part of the image

connected to border are removed so that the clean
area fruit images (Figure 2), which is one of the
sources to calculate necessary features, can be given.
Finally, the clean edge fruit images the second
type of images needs to be extracted - are produced
by using an Edge Detection algorithm. Moreover,
the gray scale fruit images the last source - arealso
extracted by using the Mask function (Figure 2).

the quality of the whole system it helps to provide a

consistent picture, eliminate the appearance of variations
such as shadow, distort colors and low contrast images.

2.2.2. Features for classification

Figure 1. Overall structure of the system

The second component is a camera. The image sensor uses

solid state charged coupled device (CDD) technology. The
speed of image acquisition is up to 30 frames per second.
The third component is CompactRIO, which provides
the flexibility, capability, and ease of deployment of image
processing program.
The fourth component is the personal computer. The
rest of the system, the software is programmed and run on.
For communication issues between components to
exchange data, the computer is connected to the real time
controller via the Ethernet port and the camera linked to the
personal computer via the USB port.
2.2. The fruits recognition algorithm
We propose the algorithm of fruits recognition
including six stages: image acquisition, segmentation,
feature extraction, classifier, sound player, and evaluation.
Most of the stages are developed in the way that they can
be implemented in embedded platforms the further step
of our research. The designed process is strictly serial order
in the sense that there can be feedback from later stages
back to earlier ones.The realization of each stage is pointed
out so that the software can be developed by benefiting
multithreading programming techniques to reduce the
processing time of the whole system.

For the design of the fruit classification system, the

issue is noticing not only which features color, texture,
and shape are the obvious choices but also how to tailor
the features and their representations to suit the application
[4, 7]. The representation for a fruit should be invariant
with respect to rotation, translation and the number of fruit
presented. Secondly, because it will be necessary to train
the system, the representation and the classification
mechanism should be simple.
As a result, there are five features selected for fruit
recognition, including color (average and variance
intensity), texture, and size (perimeter and area).
2.2.3. Feature Extraction
The developed algorithms are used to extract five
features mentioned above from fruit sample images. All
calculations are performed by using the real time
controller, CompactRIO.
a. Color Feature Extraction

In the image acquisition stage, image capturing is

designed to continuously transfer pictures from digital camera
to the personal computer via the USB port. The next stage is
segmentation, which extracts the fruitfrom the background.
The feature extraction stage determines which calculations
have to be performed on the calibrated data coming from the
camera. During the classifier design stage, an algorithm is
devised to compare the feature vector calculated from the
previous stages with the feature vectors, defined by the
specific fruits. Finally, based on the result from the previous
stages , the specific sound will be produced.
2.2.1. Segmentation
The goal of this stage is to extract three kinds of images.
The algorithm - edge detection-based algorithm - is
implemented based on the following steps:
The color images are first converted to gray scale images.
After that, the enhancement of the image is performed.
The edges of the filtered (enhanced) images are then
extracted by using Canny Edge Detection algorithm.
The images are converted to binary images by using
Morphology and ConvexHull functions.

Figure 2. Source images in segmentation stage

The
The
The
The
The
The

image 1 - the original image

image 2 - the extracted hue image
image 3 - the edge image after using the canny filter
image 4 - the area image
image 5 - the final extracted hue image
image 6 - the final edge image

60

Pham Xuan Thuy

Color captures a salient aspect of the appearance of

fruit, and does not depend on the position or orientation of
the fruit. Many color descriptors (spaces) can be found in
the literature, including:
The Hue/Saturation/Intensity (HSI) space [8].
The opponent color space [9].
The Red/Green/Blue (RGB) space [8].
Our system builds its single color from the threedimensional HSI space. While Saturation is the depth or
strength of the color, Intensity is the gray level.
However, Hue is spectral shade, which varies continuously
from red through green to blue. Moreover, the most
profound difference between fruits is in the Hue
component. Two parameters, average intensity and
variance intensity, have a substantial impact on the
efficiency and are simple to implement as follows.
For a given set of input data:x = [x(1,1), x(1,2), ..,
x(m,n)] , according to Harsh S Holalad [1], these
parameters are calculated with the following formulas:
1 m n
(1)
average
x i, j
mn i j
variance

2
1
x i, j average
mn i j
m

(2)

Where, mn the size of the image.

x(i,j) the gray intensity of the pixel in the image.
a. Perimeter and Area Features
Area and perimeter of an object are convenient
measures of the object and easily computed during
theprocess of segmentation. The area depends on the
boundary of the object and a measurement of area
disregards variations of graylevel inside the image. In
addition, the perimeter of an object is particularly useful
for discriminating among objects with either simple or
complex shapes. Based on the computing method of area
and perimeter mentioned in [10], we propose a new one,
which is much easier to process as follows.
area
perimeter

intensity of pixels in area fruit frame

mn
intensity
of
pixels
in edge fruit frame

b. Texture Features

mn

(3)
(4)

Texture is important for classifying fruits, because

many fruits cannot be reliably discriminated by color.
Texture is a visual feature that is much more difficult to
describe. The authors in [11] developed a method of
measuring the inhomogeneity of the distribution of the grey
values on the surface - image. This measurement is defined
as the lumpiness of the image data the ratio between
uniformity of the image intensities and their mean value .
Adapting from that, we propose a formula for calculating
texture as below:
1 m n
2
texture
x i, j
divides

mn i j
(5)
2
1 m,n

average
x
i
j
,

m n i, j

2.2.4. Classifier Design

Classifier uses color (intensity and variance), size
(perimeter and area), and texture parameters. For the sake
of simplicity, the ease of implementation and processing
speed, we opted for a minimal distance classifier.
Classification algorithms typically use two phases of
processing: training and testing.
In the training phase, when an image of fruit is
captured, the characteristic properties of typical image
features are separated and a training class is created. After
collecting the feature data, the average value vector for
each kind of fruit is defined by applying approximately
scale values and then used to classify unknown image data.
In the testing phase, the distance between the input
feature vector and the feature vectors of the defined fruits
is calculated to classify unknown image data.
The distance classifier that has been implemented
employs the modified Euclidean distance given by,

distance

xt i, j x0, t

(6)

Where, xt(i,j) the feature of the tth class from test sample
and x0,t the feature of the tth class from the center is obtained
by the test samples. t- the posistive constants,which are
selected by the trial and error method, making the
contribution of each feature to the classifier effective.
Training Algorithm
Step_1 - Consider an image of a fruit belonging to a class.
Step_2 - Extract the Hue, perimeter, and area
components of the images.
Step_3 - Transfer three images simultaneously from
the computer to the real time controller.
Step_4 - Compute the mean, the variance, and
texture of Hue component using Equation 1, 2 and
5. Then, calculate the area and perimeter of a fruit
image using Equation 4,5.
Step_5 - Repeat the steps from 1-4 for different
images of a class.
Step_6 - The central values are found in each feature
(color, texture, and size). Then, the mean of each
feature (means of 200 images of a class) is calculated.
All values are automatically saved to a standard
format file such as excel, text and binary file.
Step_7 - Repeat the same for the remaining classes.
Testing Algorithm
Step_1 - All Steps1_6 from the training algorithm
without saving values to excel files are performed.
Step_2 - The Euclidean distance between the test
values obtained from Step_2 and that of the already
stored center values obtained from training are
calculated using Equation 4.
Step_3 - Find out the minimum distance among all
the distances (each class) and assign the test image
to the class with minimum distance.
2.2.4 Sound Player
The sound player is designed to play a specific sound

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

file (*. wav), which is the corresponding output of the

classifier. More specifically, the content of the sound file
that issaying carries information about a certain fruit that
is more convenient for users. For example, This is apple
is content of the sound file corresponding to the output of
the classifier: an apple.
2.2.5 The evaluation of fruits recognition system.
For evaluation, the system was tested byusing many
fruits in the supermarket. Proper evaluation of the system
could not be done with the test set obtained during data
collection, so a new set of fruits is required for two reasons.
Firstly, the test set has already contained the calculated
features for every fruit. Secondly, the test set only contains
data from the fruits that also provide the training set.
Two characteristics of the system are evaluated: error
recognition rate and recognition speed. The recognition
success rate was evaluated for six types of fruits
simultaneously. In addition, the recognition speed is
evaluated based on the number of frames processed per
second. Excellent recognition speed is expected as the
system can process at the speed of 24 frames per second.
3. Results
3.1. Recognition success
Table 1 summarizes the recognition results of fruits
recognition system on the fruit images that are being sent
in as input images during testing the system. The table lists
out the test results of the system, including the fruit name,
the numbers of test fruits, and the test result.
The overall efficiency of recognition and classification
of fruits is found to be around 91%. The results are greatly
affected by the fruits, which are selected by users. If the
fruits are more carefully selected, there will have an
increase in the recognition and classification efficiency.
Table 1. The recognition results on the test fruits of fruits
recognition system
Fruit

The number of
input fruits

Testing results
(Correct)

Apple
Kiwi
Lemon
Orange
Strawberry
Tomato
Total

30
30
30
30
30
30
180

28/30
27/30
29/30
28/30
25/30
27/30
164/180
(91.11%)

3.2. Recognition speed

The fruit recognition system can process the average
seven frames per second. The processing speed depends on
some aspects, for instance, the speed of image acquisition,
the speed of image segmentation, the speed of feature
calculations, and the duration of classification. In addition,
the synchronization of different parts and transferring data
between the personal computer and real time controller
also affects the overall speed.

61

As can be seen from the Table 2, the speed of the

system is greatly affected by the segmentation. Because the
duration of the synchronization and the time of transferring
data (using TCP/IP protocol) are non-deterministic, the
speed of the whole system is done by measuring the
duration of the whole system - around 7 frames per second
(120 ms).
Table 2. The durations of different processing stages
The stages
Image acquisition
Image segmentation
Feature calculations
Classification

The average time (ms)

20
80
70
50

4. Discussion
4.1. Tuned parameters
The recognition success rate can be affected by the
scale values in feature calculation stage. The scale value
for each feature needs to be chosen carefully. If the balance
of contribution of each feature is not equal, some features
will have less influence than the others will. As a result, in
some situation, some features do not play any role at all. In
the future, we will investigate the way of finding out
optimal values for those parameters.
4.2. Adding features
One of the clearest improvements is to use additional
features for our fruit recognition. At this moment, only five
are used. As seen earlier, this makes distinguishing certain
fruits more error-prone. By understanding physical
meanings of the features mentioned in [7] and using trial
and error method, we can select additional features in the
classification in order to improve accuracy.
4.3. Classification methods
Because the classifier, which is used in the fruit
recognition system, is developed by the minimal distance
method ,which is easy to implement but has some
limitations. For instance, if two kinds of fruits have
different values of features, but the same distance, the
incorrect recognition will occur.
In order to reduce the classification error rates, some
methods can be used to make improvement such as tree
decision method [12]. Any decision tree will progressively
split the set of training examples into smaller and smaller
subsets. For each branch, the decision is to continue to split
and accept, or select another property and grow the tree
further. As a result, the quality by applying the distance
classifier can be improved.
4.4. Choosing fruits
By observing the fruits during the tests, we notice that
the mistaken recognition usually happens when the fruits
are much different from others in the same type. So, to
increase the success rate, fruits should be pre-classified
beforehand.
5. Conclusion
We have set out to implement an automatic visual

62

Pham Xuan Thuy

recognition of fruit with high accuracy. By properly setting

up designed lighting and carefully designing edge based
detection algorithm, a precise segmentation of fruit from the
background is done. From these segmented images,
recognition clues such as color, size, and texture are
extracted and calculated. According to the results, the
success rate is higher than 91% with the speed of seven
frames per second. In comparison with other works, such as
[1, 2, 13], our designed system has the same as, or even
higher accuracy. In addition, this is a completed system and
is much more convenient for users. This system could be
used for multiple purposes, in education or for blind people.
REFERENCE
[1]

[2]

[3]

Harsh S Holalad, Preethi Warrier, Aniket D Sabarad,: An FPGA based

Efficient Recognition System Using Minimum Distance Classifier.
Journal of Information Engineering and Applications, 2012.
Woo Chaw Seng, Seyed Hadi Mirisaee: A new method for fruits
recognition system. Electrical Engineering and Informatics, 2009;
ICEEI 09 International Conference on, pp. (130-134), Aug. 2009.
Dr
Vishwanath.B.C,
S.A.Madival,
Sharanbasava.Madole:
Recognition of fruits in fruit salad based on color and texture
features. International Journal of Engineering Research and
Technology (IJERT), 2012.

[4]

R.M.Bolle,
J.H.Connell,
N.Haas,
R.Mohan,
G.Taubin.:
VeggieVision: A Produce Recognition System. IEEE Computer
SocietyWashington, DC, USA1996.

[5]

R.Newlin Shebiah, S.Selva Nidhyanandhan, L.Ganesan: Fruit

recognition using color and texture features. Journal of Emerging
Trends in Computing and Information Sciences, 2010.

[6]

Jyoti A Kodagali and S Balaji: Computer Vision and Image Analysis

based techniques for automatic characterization of fruits review.
International Journal of Food Science and Technology (IJFST), 2012.
A. M. Aibinu, M. J. E. Salami, A. A. Shafie, N. Hazali and N.
Termidzi: Automatic Fruits Identification System Using Hybrid
Technique. Electronic Design, Test and Application (DELTA), Sixth
IEEE International Symposium on, pp. (217-221), 2011.

[7]

[8]

D.H.Ballard, C.M.Brown: Computer Vision. Prentice Hall, 1982.

[9]

R.W.G. Hunt: Measuring Color (Second Editor). Ellis Horwood,

West Sussex, England, 1991.

[10] Kenneth R.Castleman: Digital Signal

Education, Second Impression (2008).

Processing.

Pearson

[11] Maria Petrou, Pedro Garcia Sevilla.: Image Processing: Dealing

with texture, Wiley, March 2006.
[12] Richard O.Duda, Peter E.Hart, David G.Stork.: Pattern
Classification (Second Edition). Prentice Hall.
[13] Saswati Naskar, Tanmay Bhattacharya: A Fruit Recognition
Technique using Multiple Features and Artificial Neural Network,
International Journal of Computer Applications, Vol 116 No. 20,
April 2015.

(The Board of Editors received the paper on 08/14/2015, its review was completed on 12/22/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

63

RESEARCH ON REDUCING RESIDUE OF FIPRONIL PESTICIDE

BY VETIVER GRASS IN CONSTRUCTED WETLAND A CASE STUDY IN QUANG NAM PROVINCE, VIETNAM
Pham Phu Song Toan
The University of Danang, College of Technology; ppstoan@gmail.com

Abstract - In this study, a survey of pesticides in Quang Nam

Province illustrates that the best-selling and the highest frequency
of use is Regent 800WG, of which the main component is Fipronil,
with 95% and 82% respectively. Thus, that the residue of Fipronil,
a chemical substance in Group C that can cause cancer in
humans according to the U.S. EPA, in the soil and groundwater is
inevitable. The results have proven that the concentration of
Fipronil in groundwater at farmers households, which are closed
to the paddy field, is around 1.0 mgL to 2.0 mg/L. Therefore, this
study aims to examine the absorptive capacity of constructed
wetland system in which vetiver grass is the selective plant. The
pilot of constructed wetland is designed in 36 L tank with the
density of vetiver grass of 100 plants/m 2. The vetiver grass is 20 cm
long with bushy roots. The results show that the fipronil, which
reduces efficiency of the constructed wetland by vetiver
grass,reaches 72% with the optimal velocity and hydraulic
retention time (HRT) of 3.30*10-6 m/s and 7.58 hours, respectively.
Key words - Fipronil; constructed wetland; Regent 800WG;
insecticide; vetiver.

1. Introduction
Paul Truong et al. showed the special attributes and
effective applications of vetiver grass for wastewater
treatment and phytoremediation [1]. Xindi Liao et al.
demonstrated that vetiver significantly influenced the
removals of COD, BOD, NH3-N and total P from pig farm
wastewater [4]. Furthermore, Yahua Chen et al. proved that a
large amount of heavy metals (approxi-mately 99.7% of Pb,
93.8% of Cu, 95.6% of Zn and 98.2% of Cd) in the leachate
solutions were absorbed by the soil matrixes under the
influence of the vetiver growth in the soil columns [5]. So, can
toxic substances from pesticides be handled by vetiver? This
paper will provide an overview for Fipronil pesticide reducing
ability by vetiver grass in Quang Nam Province.

hazardous [6] and a chemical substance in Group Cpossible human carcinogen according to the U.S. EPA [7].

Figure 1. The molecular structure of Fipronil [3]

2.1.2. Vetiver grass

The scientific name of Vetiver is Vertiveriazizanioides
which is an Indian species of perennial Poaceae grass.
Vetiver can grow in the harsh environment such as drought
and high temperature (22 0C 55 0C), high salinity, a wide
range of pH (3.3 - 12.5). It can also recover very fast after
suffering from environmental damage. Vetiver is known as
a multipurpose grass for slope protection and
environmental treatment and control as well [1]. The shape
of vetiver grass is described in Figure 2 below.

2. Materials and methods

2.1. Materials
2.1.1. Fipronil pesticide
Fipronil - C12H4Cl2F6N4OS -, whose molecularis are
shown in the Figure 1, is a broad-spectrum phenylpyrazole
insecticide. The International Union of Pure and Applied
Chemistry (IUPAC) named for fipronil is ()-5-amino-1(2,6-dichloro-,,-trifluoro-p-tolyl)
4trifluoromethylsulfinylpyrazole - 3 - carbonitrile. Fipronil
was first registered fonnr use by the United States
Environmental Protection Agency (U.S. EPA) in May
1996 [8]. In agriculture, Regent 800WG is not only used
for preventing thrips, stinkbugs, leaf rollers, stem borers
and planthoppers in paddy fields but also for killing worms,
aphids, maggots in vegetable, watermelon, cucumber and
gourds [7]. The WHO IPCS hazard class for fipronil
technical active ingredient is Class II, or moderately

Figure 2. Vetiver grass [2]

2.2. Methods
2.2.1. Surveying methods
In this study, 100 chemical shop owners and 250
farmers in five cities and districts of Quang Nam Province
were interviewed to find out what insecticides have been
sold and used popularly. The questionnaire and the direct
questionable methods were used for the chemical shops
and the users as farmers respectively.

64

Pham Phu Song Toan

2.2.2. Constructed wetland

The constructed wetland system has been used for
treatment of domestic, seafood, industrial and farming
wastewater. Accordingly, aquatic plants absorb pollutants
in wastewater during their growth stages [1].In this study,
vetiver grass was used as an aquatic plant growing in the
constructed wetland system (Figure 3).
= 49 mm

Vetiver

reaching the outlet. ( = 49 mm and = 21 mm). The

control system and wetland system were both operated
simultaneously to examine and eliminate the absorption of
Fipronil by the soil colloidal particles.
High-pressure tank
Vetiver grass
(constructed wetland)

Control wetland
(without vetiver)
Sand
Gravel

Output, = 21mm

Figure 3. Design model for constructed wetland

The pilot of constructed wetland system was built by

waterproofing material with the cubic rectangular tank
(400 mm x 600 mm x 500 mm) as in Figure 4. The water
collecting pipe, is made of plastic with 49 mm in diameter,
is at the bottom of the tank. At the output, there was a valve
( = 21 mm) to control the flow. The gravel layer
(h = 400 mm) was the bottom layer, next to the sand layer
with 80 mm in depth. Two layers aimed to reduce the
obstacle of flow in wetland system and created favorable
conditions for root growth as well.
Output
( = 21mm)

Figure 5. Constructed wetland and control models

2.2.3. Analytical methods

Ground water was sampled and stored in accordance
with Vietnamese standards such as TCVN 6000:1995 (ISO
5667-11: 1992); TCVN 5992:1995, and TCVN 5993:1995
(ISO 5667-3: 1985).
15 ground water samples were taken at already
available pumping wells in 5 districts and cities in Quang
Nam Province, as Table 1 show.
Table 1. Sampling places in QuangNam Province

Hoi An City
HA1 Truong Le Village, Cam Chau Ward
HA2 Tra Que Village, Cam Chau Ward
HA3 4th Village, Cam Thanh Commune
Dien Ban Town
DB1

Dien Phuoc Commune

DB2

Dien An Commune

DB3

Dien Hong Commune

DL1

Dai Hiep Commune

DL2

Dai Chanh Commune

DL3

Dai Dong Commune

TB1

Binh Nam Commune

TB2

Binh Phuc Commune

TB3

Binh Trieu Commune

Dai Loc District

Valve
(control flow)
( = 21mm)

Thang Binh District

Tank
(waterproofing
material)

outlet pipe
( = 49mm)

Gravel layer

Figure 4. Constructed wetland model

In this study, two wetland pilots were designed next to

each other as in Figure 5. The first one was a constructed
wetland with 400 mm thick of gravel and soil layers and
the vetiver density of 100 plants/m2. After 45 nourishment
days, the vetiver grows around 700 mm tall and 300 mm
root. In this period, only water was supplied for wetland
system to grow vetiver grass. The other system was a
control wetland pilot which was the same one without
vetiver.
In test period, Fipronil was mixed with fress water and
stored in a blue container (V = 200 L) and was distributed
into the wetland systems by PVC pipe systems ( = 21
mm). Fipronil was absorbed by the vetiver roots before

Nui Thanh District

NT1

Tam Xuan 1 Commune

NT2

Tam Hoa Commune

NT3

Tam My Dong Commune

The water samples were taken and stored in dark grass

bottles. Fiponil was determined by the method of 28 TCN
180: 2002 - Organochlorine pesticides and PCB-congeners
in fish and fishery products a method for quantitative
analysis by Gas Chromatography (GC) at The
Environmental Research Center of Danang University. The
samples were extracted by Soxhlet Extraction system and
analyzed by GC with Electron Capture Detector (GCECD). These methods had high reliability with0.3% of the

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

3. Results and discussion

3.1. Survey results
The survey results from 100 chemical shops show that
there are 24 kinds of pesticides that are generally traded in
Quang Nam Province. However depending on the purposes
of use for farming and plants, the type and amount of
chemicals are sold more or less. The bar chart in Figure 6
shows the ten mostly traded pesticides in 100 chemical
shops in Quang Nam Province in April of 2015.
The Ten Mostly Traded Pesticides in Quangnam Province

Shops

Regent
800WG

Dylan
2EC

Padan
95SP

Kasumin
2L

Motox
5EC

Kasai
16,2 SC

Basa
50EC

Eagle
50EC

Actana
25WG

Angun
5WG

100
90
80
70
60
50
40
30
20
10
0
Pesticides

10

avai labl e
commerce

Figure 6. The mostly traded pesticides

in Quang Nam Province

It can be clearly seen that Regent 800WG, Dylan 2FC

and Padan 95SP are the three highest selling pesticides
with 92%, 87% and 82%, respectively. This means that the
use demand for these pesticides is so high during
agricultural cultivation.
In the other survey, 250 farmers who directly use
pesticides were asked about the spraying frequency and the
amount of pesticides for a hectare in a crop. The survey
results are shown in Figure 7.
Using Status of Pesticide in Quangnam Province
%
90
80
70
60
50
40

20
10

EC

od
nv

40
u

as

User

an

,2

75
n

16
H

op

sa

ai

le

as
K

N
D

SC

C
50

P
Ea
g

25

2L
A

lo
h

m
in

as
u
K

SP
n

95
M

ot
ox

5E

C
2E

da
Pa

G
W
00

yl
an

t8

Pesticides

en

eg

3.2. Concentration of Fipronil in ground water in Quang

Nam province
To assess the transportation of Fipronil from paddy
field to residential areas, 15 samples are taken at 15 wells
in farmerss gardens which are around 50 m close to the
fields in five districts of Quang Nam Province. The
analysis results are shown in Table 1.
Table 2. Concentration of Fipronil in ground water in several
areas in Quang Nam Province (mg/L)

Places
Hoi An City

Dien Ban
Town
Dai Loc
District
Thang Binh
District
Nui
ThanhDistrict

Sampling sites [Fipronil] (mg/L)

HA1

1.56

HA2

1.87

HA3

1.42

DB1

1.91

DB2

1.53

DB3

1.84

DL1

1.16

DL2

1.55

DL3

1.02

TB1

1.34

TB2

1.06

TB3

1.75

NT1

1.56

NT2

1.72

NT3

2.05

We can see clearly that the concentrations of Fipronil

in 15 samples are around 1.0 mg/L to 2.0 mg/L. Although
Fipronil is not mentioned in the Vietnamese standard for
ground water quality, it is said to be as "Group C - Possible
human carcinogen" by the U.S.EPA [8] and classed as
"Class II" of hazardous chemicals by the WHO IPCS.
Therefore, it is necessary to minimize residues of Fipronil
on the paddy fields and prevent the transmission of Fipronil
from fields to residential communities.
3.3. The optimal velocity and hydraulic
time(HRT) of constructed wetland

30

higher than Dyland 2FC and Padan 95SP (Group III WHO). Therefore, the residue of Regent 800WG in the soil
and groundwater is at an alarming rate

Concentration of Fipronil in ground water in Quang Nam

Province (mg/L).

standard deviation (CVS), 93 of the recovery and 0.99 of

the correlation coefficient of linear regression (R2).

frequency/crop

Figure 7. The using status of pesticides

in Quang Nam Province

The results showed that Regent 800WG, Dylan 2FC

and Padan 95SP are mostly used by 250 people who are
interviewed with 83%, 59% and 51%, respectively.The
spraying frequency of Regent 800WG is up to 4 times for
one crop. The mount of spraying depends on the weather
and different growth periods of plants.
In addition, according to the classification of toxicants
by WHO, toxicity of Regent 800WG (Group II - WHO) is

65

retention

At the beginning, we mix Fipronil (Regent 800WG

pesticide) and water together (following the instruction on
packing) in the high tank with a concentration of 80 mg/L.
After that, input valve is adjusted to supply Fipronil into
the system. After 30 minutes of retention and stabilization,
samples are taken to measure concentration of Fipronil at
the output. At the same time, the velocity of flow is defined
by velocity equipment.
In the next measurements, the output valve is tuned on
litle by litle to gradually raise the flow velocity. At every
change, when the water flows continuously out of the
outlet, the sample is taken casually to measure
concentration of Fipronil.
Thereby, the relationship between the velocity of flow,

66

Pham Phu Song Toan

Hydraulic Retention Time (HRT) and the concentration of

Fipronil is established and shown in Figure 8.
[Fipronil] mg/L
90
80

vetiver is entirely able to reduce Fipronil pesticide residue

by constructed wetland system.
Treating Performance of The Constructed Wetland by Vetiver

The Optimal Velocity and HRT of Constructed Wetland

80

70

64.12

60

66.13

63.51

80

67.28

70

50

P e fo r m a n c e ( % )

60

40
30
20
10

10.52

5.95 5.72

15.47

19.36

3 .8 5 *
1 0 ^ -6

4 .1 3 *
1 0 ^ -6

4 .9 6 *
1 0 ^ -6

5 .4 5 *
1 0 ^ -6

6 .5 3 *
1 0 ^ -6

6 .9 4 *
1 0 ^ -6

V (m/s)

7.58

6.49

6.05

5.03

4.59

3.83

3.6

HRT (h)

Constructed Wetland

40
30
20
10

3 .3 0 *
1 0 ^ -6

50

V (m/s)
0
Treating peformance

0
0

3.30*10^-6 3.85*10^-6 4.13*10^-6 4.96*10^-6 5.45*10^-6 6.53*10^-6 6.94*10^-6

72.7125

72.2375

71.9375

70.6125

66.2875

63.325

59.9

Control System

Figure 8. The optimal velocity and HRT

of constructed wetland

The line chart shows that the constructed wetland

system with vetiver grass can reduce Fipronil effectively.
Specifically, at the velocity 3.30*10-6 m/s (corresponding
with HRT = 7.58 hours) the concentration of Fipronilin
constructed wetland system decreases from 80 mg/L to
5.95 mg/L. Besides, in control system (without vetiver
grass), the concentration of Fipronil decreases slightly to
64.12 mg/L. Therefore, the role of vetiver in constructed
wetland is important.
The faster velocity is, the shorter HRT is and the lower
removal efficiency of vetiver reaches. Therefore, there is a
a significant increase in concentration of Fipronil from
5.95 mg/L to 19.36 mg/L at output, corresponding to the
flow velocity increase from 3.30*10-6 m/s to 6.94*10-6 m/s.
A similar trend also occurs at the control system.
Overall, the optimal velocity of water flow in
constructed wetland by vetiver is 3.30*10-6 m/s,
corresponding to the optimal HRT of 7.58 hours.

Figure 9. The treating performance of

constructed wetland by vetiver

4. Conclusions
This study has achieved an overview of the current state
for trade and use of pesticides in Quang Nam Province and
the reducing capacity of Fipronil pesticide residue by
constructed wetland with vetiver is as follows:
(1) Regent 800WG (Fipronil pesticide)is mostly sold
by 95 % of surveyed shop owners and used by 82 % of
interviewed farmers.
(2) The Fipronil treatment performance of the
constructed wetland by vetiver is 72.71 % at 3.30*10-6 m/s
of velocity and 7.58 hours of HRT.
REFERENCES
[1]

[2]

Paul Truong, Tran Tan Van, Elise Pinners The Vetiver System for
improving water quality: The prevention and treatment of
contaminated water and land, Published by The vetiver network
international (2008), pp. 2-15, pp. 37-56.
R. R. Rao and M. R. Suseela Vertiveriazizanioides multipurpose
eco-friendly grass of India, Pacific Rim Vetiver Network (2009), pp.
439 442.
Rachael Cull, Heather Hunter, Malcolm Hunter, Paul Truong Application of Vetiver Grass Technology in Off-Site Pollution
Control II. Tolerance to herbicides under selected wetland
conditions, 2nd IVC proceeding (2000), pp. 296-302.

3.4. Treating performance of constructed wetland with

vetiver

[3]

According to the results presented in Figure 8, the

treating performance of constructed wetland with vetiver
grass is shown in Figure 9.

[4]

Xindi Liao, Shiming Luo, Yinbao Wu and Zhisan Wang - Studies on

the Abilities of Vetiveriazizanioidesand Cyperusalternifolius for Pig
Farm Wastewater Treatment, ICV3-Proceedings (2003), pp 174-181.

[5]

Yahua Chen, Zhenguo Shen, Xiangdong Li - The use of vetiver grass

(Vetiveriazizanioides)
in
the
phytoremediation
of
soils
contaminated with heavy metals, Applied Geochemistry 19 (2004),
pp. 15531565.

[6]

WHO - The WHO Recommended Classification of Pesticides by

Hazard and guidelines to classification 2009, International Program
on Chemical Safety, 2009, pp.2
Amrith S. Gunasekara and TrescaTroung - Environmental Fate of
Fipronil: Toxicity, USA (2011), pp.16-22.

In overview, the Fipronil removal efficiency of

constructed wetland with vetiver is about 60% to 70%
corresponding to the flow velocity from 3.30*10-6 m/s to
4.96*10 -6m/s. The highest reducing performance of
Fipronil by the constructed wetland with vetiver is 72.71 %
at 3.30*10-6 m/s of velocity and 7.58 hours of HRT.
Rachael Cull et al (2000) demonstrated that vetiver is
unlikely to be adversely affected by runoff containing
residues
of
Atrazine
(C8H 14CIN 5)
or
Diuron
(C9H 10Cl2N 2O) which are herbicides [3]. However, they
have not found the absorbption ability of these herbicides
by vetiver. Meanwhile, this study initially proves that

[7]
[8]

US.EPA Fipronil. Human Health Risk Assessment Petition to

Support and Maintain the Established Rice Grain Tolerance for
Imported Rice, 2009, pp.4

(The Board of Editors received the paper on 11/21/2015, its review was completed on 12/13/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

67

DESIGN OF SILICON WIRES BASED DIRECTIONAL COUPLERS FOR

MICRORING RESONATORS
Le Trung Thanh1, Nguyen Canh Minh, Nguyen Van Khoi2, Bui Thi Thuy3, Nguyen Thi Hong Loan3
1

Vietnam National University, Hanoi (VNU); thanh.le@vnu.edu.vn

2
University of Transport and Communications, Hanoi, Vietnam
3
Hanoi University of Natural Resources and Environment, Hanoi, Vietnam
Abstract - In this paper, we investigate the design of directional
couplers and m icroring resonators based on silicon wires. The aim
is to design the directional coupler for high performance microring
resonators. The effect of microring radius, gap, silicon waveguide
width on power transmission ratios is analyzed by using the 3D
Eigenm ode Method (EME). The behavior of m icroring resonators
using the investigated directional coupler such as finesse (F),
Q-factor and free spectral range (FSR) is investigated. The effect
of the waveguide width variation on the finesse, Q-factor is also
studied. The FDTD simulation shows a very good agreement with
the proposed design approach.
Key words - integrated optics; coupled resonators; integrated
optics devices; silic guidewaves; EME sim ulation method.

1. Introduction
Photonic Integrated Circuits (PIC) based on silicon on
insulator (SOI) platform are very attractive in recent years.
The SOI platform can provide advantages of high index
contrast,creating photonic devices with compactness, CMOS
compatibility and large scale production technology [1].
Microring resonators are versatile building blocks for
photonic circuits especially for compact photonic structures.
There are many photonic devices based on microring
resonator structures such as optical filter [2], optical
multiplexer [3], optical switch [4], optical delay line [5],
modulator [6], differential equation solver [7], optical
interleaver[8], fast and slow light [9], optical biosensor [10,
11], optical gyroscopes [12], generation of optical signals
[13], etc. For accurate operation of these devices, power
coupling ratios and loss of the coupler used in microring
resonators are crucial and need to be carefully designed. The
coupling element determines the device performance. For
example, designing couplers with desired coupling power
ratios is extremely important for high Q-factor and high
performance for biomedical sensing based on SOImicroring
resonators [10], for obtaining exact free spectral range (FSR)
of dense wavelength division multiplexing and interleaver
[3, 8], for high bandwidth photonic signal processing [7, 14].
Although silicon wire based directional couplers have
been used in many microring resonator structures, it still
lacks a detailed and accurate analysis and design for
microring resonators based on the directional coupler.
Silicon wire based directional couplers with two parallel
waveguides has been analyzed [15, 16], but it is not suitable
for applying to almost all microring resonator structures
including a straight waveguide coupling with a ring
waveguide. Recently, quality factor, finesse and fabrication
tolerances of directional couplers used for microring
resonators have been investigated [17, 18]. However, such
analyses can only applied to SOI rib waveguides and
microring resonators based on InGaAsP/InP.

In this paper, we present a detailed analysis and design

of silicon wire based directional coupler consisting of a ring
waveguide coupled with a straight waveguide used for
microring resonators. The effect of silicon waveguide width,
gap, and radius of the ring waveguide on the power coupling
ratios is investigated. The Q-factor and finesse of microring
resonators using the directional coupler is analyzed. The
fabrication tolerances and wavelength sensitivity are also
studied. We use the 3D EME for modeling and analysis of
directional couplers and the FDTD method then is used for
microring resonators to compare the results.
2. Modeling of the SOI directional coupler
The directional coupler based on the silicon wire used
for microring resonator structures is shown in Figure1. In
this work, we designed for a 1550nm center wavelength
and TE polarized light. The directional coupler is made
from a straight waveguide coupled with a ring waveguide
with radius R. For our investigations, the height of the
silicon wire is usually chosen to be hco 220 nm and the
thick of the buried oxide (BOX) is 3m. g is the gap of the
directional coupler. The waveguide core Si (n Si 3.45) is
covered by the SiO2 cladding (n SiO 1.46 ).
2

Figure 1. Schematic of the investigated SOI directional coupler

for microring resonators (a) top-view and (b) cross-section view

The directional coupler can be characterized by a

transfer matrix [19, 20]

M
j

(1)

Pout
is the power transmission coefficient
Pin
2
and | | is the power coupling coefficient of the coupler.
Where

Figure 2(a) and (b) show the dispersion characteristics of

the silicon wire at different waveguide widths of 400, 450,
dN eff
480 and 500nm, where the group index N g N eff
,
d
N eff is the effective index calculated by the 3D EME.

68

Le TrungThanh, Nguyen Canh Minh, Nguyen Van Khoi, Bui Thi Thuy, Nguyen Thi Hong Loan

microring waveguide can be coupled to one or two

waveguides. Here we employ the transfer matrix approach
to transfer the design parameters of the directional coupler
on microring resonators.

Figure 2. Effective index and group index of the silicon wire

Transmission coefficients between straight waveguide

and ring waveguide have been plotted in Figure 3 and 4 for
R ranging from 5m to 50m and g is from 80nm to 200nm
with different waveguide widths. The insets are field
propagation over the coupler at Wa 480 nm , R 50 m
and gap g=120nm and 200nm, respectively.

Figure 5. Wavelength dependence of the power transmission

coefficient at different width and gap

Figure 3. Power transfer between the straight and ring

waveguides dependence gap and waveguide width, R=50m

Next we investigate the wavelength sensitivity of the

coupler. The transmission coefficients are shown in Figure 5
for R=5, 20m and gap g=87.5nm, 120nm respectively at
the width Wa=400nm and wavelength ranging from 1550nm
to 1600nm. The transmission coefficients decrease linearly
when the wavelength increases. The decrease in
transmission coefficient is proportional to the increase in
radius. For R=5 m, the coefficient 2 can be expressed by
2

= -1.4368 + 2.8362 and 0.4635 1.5627 with

R=20 m. For a wavelength range of 25nm, the deviation of
the transmission of the coupler is within 4%.
2

Figure 6. Two microring resonator configurations used in this study

Based on the transfer matrix theory, the transmission of

the microring resonators can be achieved. The quality
factor Q of the microring resonator of Figure 6(a) and 6(b)
can be derived as follows[20]:
Q1

N g L
1

1 4 4 2
Ng L
)
Q2
arcc os (

2 2

(2)
1

(3)

Where is the resonance wavelength of the resonators,

L is the ring cavity length, L 2 R , is the single pass
field strength attenuation which is related to the power
attenuation a(1/cm) and round trip length as... For a
standard silicon wire, the attenuation factor is a=2 3
dB/cm [22].
Another important parameter for microring resonators
is the finesse F, which is defined and calculated for the
single and ad-drop microring resonators by

F1

FSR
FWHM
F2

Figure 4. Power transfer between the straight and ring

waveguides dependence gap and waveguide width

3. Modeling of Microring resonators

Microring resonator is modelled by the transfer matrix
approach as reported in [19, 21]. To further investigate the
behavior of microring resonators using the coupler
designed in Section 2, we examine two configurations of
microring resonator structures shown in Figure 6. The

(4)

(5)
1 2
Where FWHM is the resonance full-width-at-halftmaximum and FSR is the free spectral range. The FSR is
identical for two microring resonator configurations.
Figure 7 plots the FSR at different R and Wa. The FSR
increases by deceasing R. We see that the FSR=2.3 nm for
R=5m, and FSR is equal to 0.25 to 2.3nm for R ranging
from 5m to 50m. There is a slight increase in FSR when
Wa decreases from 500nm to 400nm.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

69

to a new waveguide width expressed by W Wa Wa.

Adding to the change of the transmission coefficient, the
deviation of the waveguide width also leads to the change in
effective index. For a positive Wa , the effective index is
increased. For any gap and radius, a positive Wa leads to a
decrease in the transmission coefficient.

Figure 7. FSR dependence on ring radius and width of the

silicon wire

The finesses F of the single and add-drop microring

resonators for gap g ranging 80nm to 200nm, R from 5 to
40m and Wa=400nn are plotted in Figure 8(a) and (b).
The finesse increases monotonically by decreasing the gap.
Figure 9 shows the Q-factor for two configurations for
gap g ranging from 80nm to 200nm and Wa=400nm. We
see that the Q-factor of both configurations is proportional
to the radius R when the gap g>110nm. For g<110nm and
R>30m, the Finesse F and Q-factor Q are nearly
independent of the radius, but for R<30 m, the Q and F
depends on the radius. For example, g=200nm, when
radius increases by 5m, the finesse F decreases by 8 while
the Q-factor increases by 6 x10 3 . Increasing in the gap leads
to increase in Q-factor but decrease in the finesse. In
addition, from the simulations of Figure 7 and 8, the finesse
and Q-factor are larger for the single microring resonator
than for the add-drop resonator configuration.

Figure 10. Change of the transmission coefficient and the

deviation from the calculated value at Wa=450nm as effect of
the width variation

For W a 10 nm , the transmission coefficient is

decreased by 0.044 for g=120nm and 0.037 for g=130nm
at the same width Wa=450nm and radius R=10m.
While this coefficient is decreased only by 0.012 if the
ring radius R=5m. As a result, the transmission coefficient
of the coupler is quite stable for a smaller ring radius and
larger gap. For a width variation within 20nm, a deviation
of the transmission coefficient of 13% can be obtained.

(a)
(b)
Figure 11. Calculated effects of the deviation of waveguide
widths on the behavior of the microring resonator (a) change in
the finess F and (b) finesse deviation from the normalized value

(a)
(b)
Figure 8. Finesse F for single and add-drop microring resonators

(a)
(b)
Figure 12. Calculated effects of the deviation of waveguide widths
on the behavior of the microring resonator (a) change in the
Q-factor and (b) Q-factor deviation from the normalized value
Figure 9. Q-factor for single and add-drop microring resonators

4. Fabrication tolerance analysis

The simulation results for the deviation of the transmission
2
coefficient depending on the waveguide width variation
Wa are shown in Figure 10. Due to the manufacturing
tolerances, the variation in waveguide width occurs and leads

The calculated results for the change of the finesse F

and Q-factor as well as the relative deviation from the
normalized values in % are presented in Figure 11 and 12.
The graphs are plotted over the waveguide width variation
Wa for different radius and gap. For a positive Wa , both
finesse and Q-factor increase. For a waveguide width
variation of 10nm, the maximal finesse change of 16%

70

Le TrungThanh, Nguyen Canh Minh, Nguyen Van Khoi, Bui Thi Thuy, Nguyen Thi Hong Loan

and Q-factor change of 27% is achieved.

5. FDTD Simulation
To verify the accuracy of the transfer matrix analysis,
we compare the results obtained with the FDTD. For our
FDTD simulations, the radius of the microring resonator is
to be R 5 m , the waveguide width is W a 400 nm , the
gap between the microring waveguide and the straight
waveguide is chosen to be g=160nm in order for the power
2
transmission coupling to be 0.9.
In our FDTD simulation, we take into account the
wavelength dispersion of the silicon waveguide as shown in
Figure 2. We employ the design of the directional coupler
presented in the previous section as the input for the FDTD.
A Gaussian light pulse of 15fs pulse width is launched from
the input to investigate the transmission characteristics of the
device. The grid size x y 0.02nm and z 0.02nm
are chosen in our simulations. The simulation results for
field propagation over the single microring resonator based
on the designed directional coupler are shown in Figure 13.
We see that the FDTD simulation has a very good agreement
with the proposed analysis.

Vietnam National University, Hanoi (VNU) under project

number QG.15.30.
REFERENCES
[1]

M. Jamal Deen and Prasanta Kumar Basu, Silicon Photonics:

Fundamentals and Devices: Wiley Series in Materials for Electronic
& Optoelectronic Applications, 2012.

[2]

C. Vzquez, S. Vargas, J. M. S. Pena et al., "Tunable optical filters

using compound ring resonators for DWDM", IEEE Photonics
Technology Letters, vol. 15, pp. 1085-1087, 2003.
D. T. H. Tan, A. Grieco, and Y. Fainman, "Towards 100 channel
dense wavelength division multiplexing with 100GHz spacing on
silicon", Optics Express, vol. 22, pp. 10408-10415, 2014.

[3]

[4]

Sang-Yeon Cho and Richard Soref, "Interferometric microring-resonant

22 optical switches", Optics Express, vol. 16, pp. 13304-13314, 2008.

[5]

F. Liu, Q. Li, and Z. Zhang et al., "Optically tunable delay line in silicon
microring resonator based on thermal nonlinear effect", IEEE Journal of
Selected Topics in Quantum Electronics, vol. 14, pp. 706 - 712, 2008.

[6]

Yingtao Hu, Xi Xiao, Hao Xu et al., "High-speed silicon modulator

based on cascaded microring resonators", Optics Express, vol. 20,
pp. 15079-15085, 2012.
Ting Yang, Jianji Dong, Liangjun Lu et al., "All-optical differential
equation solver with constant-coefficient tunable based on a single
microring resonator", Scientific Reports, vol. 4, July 2014.

[7]

[8]

J. Song, H. Zhao, and Q. Fang et al., "Effective thermo-optical

enhanced cross-ring resonator MZI interleavers on SOI", Optics
Express, vol. 16, pp. 21476-21482, 2008.
[9] Chris Fietz and Gennady Shvets, "Simultaneous fast and slow light in
microring resonators", Optics Letters, vol. 32, pp. 3480-3482, 2007.
[10] C. Ciminelli, F. DellOlio, D. Conteduca et al., "High performance
SOI microring resonator for biochemical sensing", Optics & Laser
Technology, vol. 59, pp. 60-67, 2014.
[11] G Testa, G Persichetti, and R Bernini, "Optimization of a Hybrid
Silicon-Polymer Optical Ring Resonator", Sensors, Springer, vol.
319, pp. 259-263, 2015.
[12] Mario N. Armenise, Caterina Ciminelli, Francesco Dell'Olio et al.,
Advances in Gyroscope Technologies: Springer, 2011.

Figure 13. FDTD simulation of the single microring resonator

compared with the transfer matrix approach integrated with the
proposed analysis

6. Conclusion
We have presented a detailed analysis and design of the
directional coupler based on silicon wires. We have focused
on the behavior of the directional coupler used for microring
resonator structures. The parameters of the directional
coupler are taken into consideration in order to achieve the
desired design characteristics. The transmission coefficient
of the coupler, finesse, Q-factor of the investigated
directional coupler based microring resonators have been
investigated. In addition, the sensitivity of the coupler to
variations of the waveguide width, gap and wavelength is
also presented. Our method approach can be applied and
extended to the design of microring resonator structures
using directional couplers based on the silicon wires.
Acknowledgements
This research is funded by Vietnam National
Foundation for Science and Technology Development
(NAFOSTED) under grant number 103.02-2013.72" and

[13] IS Amiri, A Shahidinejad, and J Ali, "Generating of 5761 GHz

frequency band using a panda ring resonator", Quantum Matter, 2014.
[14] Fangfei Liu, Tao Wang, Li Qiang et al., "Compact optical temporal
differentiator based on silicon microring resonator", Optics Express,
vol. 16, pp. 15880-15886, 2008.
[15] Yu-jun Quan, Pei-de Han, Qi-jiang Ran et al., "A photonic wirebased directional coupler based on SOI", Optics Communications,
vol. 281, pp. 31053110, 2008.
[16] Vittorio M.N Passaro, Francesco Dell'Olio, Branislav Timotijevic et
al., "Polarization-insensitive directional couplers based on SOI wire
waveguides", The Open Optics Journal, vol. 2, pp. 6-9, 2007.
[17] C. Ciminelli, V. M.N. Passaro, F. Dell'Olio et al., "Quality factor
and finesse optimization in buried InGaAsP/InP ring resonators",
The Journal of European Optical Society, vol. 4, pp. 09032-, 2009.
[18] Andreas Prinzen, Michael Waldow, and Heinrich Kurz, "Fabrication
tolerances of soi based directional couplers and ring resonators",
Optics Express, vol. 21, pp. 17212-17220, 2013.
[19] A. Yariv, "Universal relations for coupling of optical power between
microresonators and dielectric waveguides", Electronics Letters,
vol. 36, pp. 321322, 2000.
[20] Ioannis Chremmos, Otto Schwelb, and Nikolaos Uzunoglu (Editors),
Photonic Microresonator Research and Applications: Springer, 2010.
[21] D.G. Rabus, Integrated Ring Resonators The Compendium:
Springer-Verlag, 2007.
[22] Yurii Vlasov and Sharee McNab, "Losses in single-mode silicon-oninsulator strip waveguides and bends", Optics Express, vol. 12, pp.
1622-1631, 2004.

(The Board of Editors received the paper on 05/03/2015, its review was completed on 06/26/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

71

HUMAN - COMPUTER INTERACTION BASED ON IMU SENSORS

Huynh Thanh Tung, Vu Van Thanh
The University of Danang, University of Science and Technology; httung@dut.udn.vn, vuvanthanh85@gmail.com

Abstract - In this paper we introduce how to establish, collect and

process data from IMU sensors. The sensor we use here is MPU
6050. Data received from the sensor is transmitted to the
microcontroller through I2C protocol where data will be filtered
through a smooth filter. Data from MPU6050 is 6 axis data
consisting of acceleration meter and gyroscope which after being
filtered and reducing noise at high frequency will be transmitted to
com puter through UART protocol. On the com puter, we build
programs in C# to receive data from the microcontroller and
process data to send dummy command to control application on
computer. From characteristics of data about axis, we can build a
state stable algorithm , determine movem ent trajectory to control
the state of the computer mouse [2], [3], which facilitates interaction
between people and com puters.
Key words - IMU sensor; smooth filter; com puter-m ouse for the
disabled; digital m otion processor-DMP; UART;

1. Introduction
Today, the development of electronic devices is going
fast, which requires more information exchange and the
interaction between humans and machines is not limited to
mechanical motion (button,). Electronic devices can
recognize actions, gestures of humans, then process, encrypt
and transmit them to computer in the form of digital data or
dummy command. Some devices can be used as acceleration
sensors, ultrasonic transceivers, infrared sensors In this
paper, we use acceleration sensors MPU to get the data to
perform communication in computer.
Acceleration sensors have been widely used in recent
years. Some devices such as: smart phones, aircraft
controls, gaming equipment are using acceleration
sensors. Using accelerometer with 1 axis, we can make the
system self-balance; with 3 axes (acceleration) we can
control the mouse; with 6 axes (acceleration + gyroscope)
we can determine angle 3 axes to control planes, play
games or motion tracking. Using acceleration sensors we
can detect motion and hand gestures (left move, right
move, left rotate, right rotate). Using parameter of 6 axes
to detect hand states, we can use those data to control the
application (mouse, plane).
Inertial measurement unit (IMU):
Inertial measurement unit (IMU) is integrated devices
which consists of two types of sensor: accelerometer sensor
and gyro sensor, placed on three perpendicular axes to
track the position and orientation. IMU combines the
advantages of two systems described in above applications
as low latency, high frequency, self-contained, small, and
lightweight.
Theory acceleration sensor and gyro sensor:
Acceleration sensor:
An accelerometer is a simple object, attached to a spring
with constant elasticity coefficient k. Displacement of
objects with mass m blocks from the center to the location

to be measured is x. Using Hooke's law and Newtons law

we can determine the acceleration as follows:
Hookes Law

F = kx

(1)

Newtons law:

F = ma

(2)

From (1) and (2) we have inferred acceleration:

(3)
From formulas(3), it is easy to see that the calculation
is simple acceleration, but the actual springs in linear
ranges around a point called the original position and it will
generate errors in the read values acceleration when there
is a relatively large impact force on the spring. Thus, the
construction of enhanced sensor accuracy is required, with
the aim to make the object always keep the position at its
original location. This is done by closed-loop system with
the power and range shifts in the magnetic field.
Acceleration can be determined by the amount of power
generated electromagnetic force to keep the object lying in
bed. This method is usually constructed using microelectromechanical systems (MEMS) [4], [7].
Gyro sensor:
The term of Gyroscope immediately appeared from the
mid-nineteenth century, and in recent decades it has been
widely used and is replaced globally with the gyro. The
original theories of gyroscopes are used to explain the
motion of an object such as the Earth turns. And gradually
gyro has been developed and widely applied in many
fields, especially in the inertial navigation system INS. The
gyro was the first practical application to assist in
determining the direction of ships, submarines and aircraft
by determining the roll, pitch, yaw from the frame of
reference of a particular principle .Traditional gyro system
is called Gimball quarterly. However, the gyro system is
too bulky and heavy to be extended to used for other
purposes such as monitoring humans and robots.
Everything changed when the micro-electromechanical
systems (MEMS) was born, allowing the implementation
of small, light and cheaper gyro called Coriolis vibrating
gyroscopes (CVG). The gyro has many advantages such as
frequency response of thousands of Hz, the low noise
"sliding phase" (jitter).
2. The device structure and operation principle
2.1. Introduction to device
2.1.1. About MPU6050
MPU6050 is the worlds first integrated 6 axis motion
sensor. It combines one 3 axis accelerometer and one 3 axis
gyroscope, and it has its own digital motion processor
(DMP) which can process the motion data with its inside
algorithm. The chip itself has an internal of 16 bit analog
to digital converter (ADC), so the output data are 16 bit

72

Huynh Thanh Tung, Vu Van Thanh

digital values. There are 117 registers in total inside the

chip and all of the registers are 8 bit, so it needs two
registers to hold the value for one axis data. The detection
range of the accelerometer is +2g, 4g, 8g, 16g and that of
the gyroscope is +250, +500, +1000, 2000/s, the range can
be chosen by setting the corresponding registers[1], [8].

computer through UART protocol. Figure 3 shows the

connection between MCU, MPU6050 and PC.

2.1.2. Introduction to MSP430

The TIs MSP430 is a very clean 16-bit byte-addressed
processor with a 64K unified address space, and memorymapped peripherals. The MSP430 excels where low power
consumption is important. Many applications such as water
meters are currently achieving more than 10 - year
operation from a single button cell battery. It programs
very well in C, making assembly language programming
unnecessary. There is no memory bank switching to make
the compiler's life difficult; it uses a normal RAM for its
stack; it has a clean 16 bit instruction set. In fact, it is
somewhat like an ordinary desktop RISC processor, but
requires very little power [9].
2.1.3. Introduction to I2C protocol

Figure 3. System implementation

IC is a multi-master protocol that uses 2 signal lines.

The two IC signals are called serial data (SDA) and
serial clock (SCL). Virtually any number of slaves and
any number of masters can be connected onto these 2 signal
lines and communicate between each other.
The data rate has to be chosen between 100 kbps, 400
kbps and 3.4 Mbps, respectively called standard mode, fast
mode and high speed mode.

2.3. MCU Algorithm

Start

Initialize I 2C,
UART, IN/OUT

Setup MPU6050
sensor

Get values and

filter

Send filtered
values to PC

Figure 4. Processing steps in MSP430

Figure 4 shows that the task of this overall block is to

get data values from MPU6050, then process data through
filters and finally send the filtered data to computer using
UART protocol.

Figure 1. I2C bus with 2 devices connected

2.2. The device structure

I2C Protocol

MPU6050

USB to UART

MSP430G2553

Computer

Figure 2. Overall block diagram

MPU6050 is communicated with MSP430G2553

through the I2C data. MSP430G2553 is connected with

Initialize I2C, UART, IN/ OUT: prepares for

connection between MPU6050 and computer.
Setup MPU6050 sensor: Working like other
difference modules. This sensor module needs a
setup step. The main task of this function is to select
the range and frequency operation of acceleration
and gyroscope.
Get and filter values: We get digital signal from
MPU6050. The accelerometer can detect
acceleration in accuracy, but the results can suffer

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

from the vibration error. Especially when human

hand is moving the accelerometer, the data would
be unstable, so the smooth filter is applied to the
accelerometer data. It is filtered by the equation x.
Send filtered values to PC: Filtered values are sent
to computer through UART and will be processed
in PC application.
2.4. The smooth filter
In this project, we use the moving average filter smooth.
A moving average filter smoothes data by replacing
each data point with the average of the neighboring data
points defined within the span. This process is equivalent
to low-pass filtering with the response of the smoothing
[5], [6] given by the equation:

73

Command Mapping: translates data from MSP430.

Send Command: sends command to control
application. Application can be mouse move event,
right click, left click
MSP430G2553 sends data to computer through USB to
UART; computer using software C# to receive data from
USB port. User interface would be controlled to connect
port. Data from buffer would be read and processed. Here
dummy command to control applications will be mapped
onto corresponding data from MSP430.
3.2. Data processing on computer
Software C# is used to receive data and process it [10].

(4)

Where,
x[k]: the kth smoothed value.
From (4) with N = 4 so we get:

(5)

The smooth filter is applied to both accelerometer and

gyroscope data.
3. Application program and result
3.1. Application program

Figure 5. Computer application program

Figure 5 shows the diagram of this application.

Computer would receive data from USB port, then process
it and send command to control application.
USB to UART: board is converted from USB to
TTL to transmit data through UART (Universal
Asynchronous Receiver/Transmitter) protocol.
Data buffer: data received from USB is sent to data buffer.
User Interface: the program interface and algorithm
are designed by using C# on Window Form.

Figure 6. Receiving and processing algorithm

3.3. Result
Figure 7 shows waveform of accelerometer data when
MPU6050 rotates right following X axis with
accelerometer angle of 900. When MPU6050 is rotated,
accelerometer data will change. We can determine a
threshold value which changes.

74

Huynh Thanh Tung, Vu Van Thanh

From their characteristics, accelerometer and

gyroscope are used to detect, recognize hand motions.
Then we use it to control application. For example, when
user rotates hand right, value accelerometer will change,
and respectively command right move cursor.
4. Conclusion and perspective
Figure 7. Waveform of accelerometer data

Figure 8 shows waveform of accelerometer data when

MPU6050 rotates left following X axis with accelerometer
angle of 90 . We can see that accelerometer value is less
than zero. Both Figure 1 and Figure 2 accelerometer value
are stable and it will change when MPU6050 rotates its
original state.

This paper shows that acceleration sensors can detect

action, movement of objects (hand, plane). From data
receiver, we can control the computer with a simple
operation without interacting it directly. The result of
filtered data is stable and accurate. It can be applied to
control planes, game devices, track movements of objects.
In the next research, we will use Kalman filter and wireless
connection to connect hardware and computer. This
method is more stable and reliable, more convenient and
flexible for users.
REFERENCES
[1]

Liqiang Du, Design and implementaion of home use porttable

smart electronics, Michigan Technological University.

[2]

E. Foxlin: Chapter 7. Motion Tracking Requirements and

Technologies. 2002.
G. Welch, E. Foxlin: Motion Tracking: No Silver Bullet, but a
Respactable Arsenal. IEEE Computer Graphics and Applications,
November/December, 2002.
Norhafizan Ahmad, Raja Ariffin Raja Ghazilla, and Nazirah M.
Khairi, Reviews on Various Inertial Measurement Unit, in
International Journal of Signal Processing Systems Vol. 1, No. 2
December 2013.

[3]

Figure 8. Waveform of accelerometer data

Figure 9 shows waveform of gyroscope data when

MPU6050 rotates right or left following X axis. Gyroscope
value will change when MPU6050 rotates right or left.
Gyroscope value will return to its original value when
MPU6050 is in stable state.

[4]

[5]

Savitzky, A., Golay, M.J.E. Smoothing and differentiation of data

by simplified least squares procedures, Analytical Chemistry,
36(2), p.1627, (1964).

[6]

Jos Luis Guin, Emma Ortega, Jos Garca-Antn, Valentn

Prez-Herranz, Moving Average and Savitzki-Golay Smoothing
Filters Using Mathcad, International Conference on Engineering
Education ICEE 2007.
Diego E. Serrano, Design and Analysics of MEMS
accelerometers, IEEE Sensors 2013.

[7]
[8]

InvenSense Inc, MPU6000 and MPU6050 Product specification

Revision 3.1, [online]: http://www.elecrow.com/download/PSMPU-6000A.pdf

[9]

John H.Davies, MSP430 Microcontroller Basics, 2008.

[10] John Sharp, Microsoft Visual C# 2013 step by step, 2013.

Figure 9. Waveform of gyroscope data

(The Board of Editors received the paper on 11/02/2015, its review was completed on 11/17/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

75

APPLYING MODEL-VIEW-CONTROLLER MODEL (MVC MODEL)

IN BUILDING A DORMITORY MANAGEMENT INFORMATION
SYSTEM FOR THE UNIVERSITY OF DANANG,
UNIVERSITY OF SCIENCE AND TECHNOLOGY
Dang Hoai Phuong1, Nguyen Van Dong1, Dinh Thi Hong Huyen2, Pham Thi Thu Hien1, Lam Thi Hong Tam1
1
The University of Danang, University of Science and Technology
dhphuong@dut.udn.vn; nvdong@dut.udn.vn; thuhien06185@gmail.com; lthtam@dut.udn.vn
2
QuyNhon University; honghuyendhqn@gmail.com

Abstract - The article presents the method of building a dormitory

management information system for the University of Science and
Technology, University of Danang by using the modern web
technology. It is a website manager that is built on the Model-ViewController model (MVC model) using ASP.NET MVC technology. In
this paper, the authors have clearly identified functions of the MVC
model in the current website management systems. They have
analyzed two patterns of MVC model, which are Supervising
Controller and Passive View. The Passive View pattern is selected
to build the dormitory management information system because it
can manage to separate tasks into dependent and special
components easily, which helps the system to operate better and
more effectively.
Key words - Dormitory m anagem ent inform ation model; MVC
m odel; Supervising Controller pattern; Passive View pattern;
Dormitory managem ent website; ASP.NET technology.

1. Introduction
Currently, in the country and in the world, there is much
commercial software designed to serve boarder
management. However, in universities, the adaptability of
these applications is not high actually. In addition, the cost
of owning the copyright of such software is quite high.
On the other hand, demand for accommodation of
university students across the country is great. The
dormitory management with large numbers of students
needs much effort, time and cost. Recognizing this, some
universities have developed a website system to manage
operations of their dormitories.
In the dormitory website of National University of Ho
Chi Minh City [5], students can only read the news, search
and look up but they cannot register/extend dormitory
rooms. Therefore, the website is only used for post news,
announcements, but it is not as alternative for the manual
work in dormitory management activities.

application for the University of Science and

TechnologyUniversity of Danang is an urgent requirement
to make the dormitory management more convenient and
more accurate. Moreover, it could support boarders to get
information in a quick and timely way.
2. Dormitory management model
We have based on the MVC model [1], [3], [4] using
ASP.NET MVC technology [2] to build the Dormitory
Management Model of the University of Science and
Technology, University of Danang.
2.1. Model-View-Controller (MVC model)
MVC model (Figure 1) plays an important role in the
construction, development, operation and maintenance of
a system or a software application. It creates a 3-layer
Model View Controller isolation and interacts with
each other [3], making the exchange and processing of
transactions quick and easy. MVC model mainly applies to
the software systems of website and is used with the Web
programming popular language such as PHP, ASP.NET,
JSP and so on.
Model: the component that contains all logic
operations, processing methods, databases accessing, data
description object such as processing function, etc.
View: undertakes the display of information,
interaction with the users, which contains all GUI objects
as textbox, images, etc. in a simple way. It is a set of form
or HTML file.
Controller: takes the responsibility of redirection for
user request and revokes the proper methods to handle
them. For instance, this component will receive URL/form
requests to operate directly with the Model.

With the dormitory website of Pedagogical University,

the University of Thai Nguyen [6], students cannot perform
the following affairs:
- Interacting directly with dormitory management staff
via website;
- Asking for registering or extending dormitory rooms;
- Sending requests to the dormitory management staff.
Most of dormitory management websites, which are
still superficial, are only used for introduction and
promotion donot fully meet the needs and the purpose of
boarders and managers.
Therefore, using information technology to develop an

Figure 1. MVC Model

76

Dang Hoai Phuong, Nguyen Van Dong, Dinh Thi Hong Huyen, Pham Thi Thu Hien, Lam Thi Hong Tam

We have chosen the MVC model to develop Dormitory

Management Website because of the following reasons:
- This model does not depend on environment, platform
and/or programming language;
- Arranging functions/classes in separated Controller
Model View. Therefore, it would be easy to build,
manage, operate and maintain a project. Making it clear in
the process of project development as well as controlling
the processing flows and creating the components which
could handle specialized services;
- The MVC model is not a language, but when different
participants (programmers, managers, investors, etc.) take
the look together they would understand what it is, then
they can exchange requests and discuss their work.
This is a standard model and the most optimal
compared to others. It is used in many projects and areas,
particularly in the technology of application-software
production. We use the standard model MVC to distribute
and transfer technology easily.
There are two versions in the MVC model. They are the
Passive View and Supervising Controller patterns.
2.1.1. Supervising Controller pattern
Firstly, the View will take events from user transactions
and transmit them to Controller where they can be handled.
To update changes of Model, the View will take databinding for simple handlings. For complicated ones, the
Controller will retrieve data from the Model and transfer
updates to the View.

In this pattern (Figure 3), the View does not process as

well as interact directly with Model components. It
transfers handlings to the Controller to undertake. It will
interact with the Model and update changings to View.
Thus, the Controller could play a role as intermediate
component that communicates View and Model.
The deployment process is vertical in the Passive View;
.i.e. data will go from View to Controller then come to
Model. When the model changes data, the Controller will
receive the modified data via link between them.The
updated data will continue to fetch the View from the
Controller. The Passive View pattern was selected to build
the dormitory management information system because it
can manage to separate tasks into specific ones easilyas
well as make the system operate coherently and effectively.
2.2. ASP.NET MVC Technology
This technology allows us to distinct the task of
application (logical input, logical business and logical
interface). Thus, the testing is taken easily. All principal
features of MVC model are installed based on interface and
are tested by using mock objects (simulation of real objects
in the application). The unit-test could be applied without
running Controller in the ASP.NET MVC process. This
leads to the fast and convenient deployment of unit-test [1].
MVC is an extensible and pluggable platform. The
design of ASP.NET MVC components make them easily
changeable or adjustable. Therefore, the View engine,
URL routing mechanique, action-method parameter
extraction and others could be embedded.
The ASP.NET MVC also supports Dependency
Injection (DI) and Inversion of Control (IoC). DI permits
us to bind objects to a class instead of obeying that class
has to initiate ones. IoC defines that if an object requires
another,then the first object will get the second one from a
foreign source such as configuration file. As a result, the
use of DI v IoC will make testing easier.
There is a powerful URL mapping component in
ASP.NET MVC which allows us to build applications
including contensive URLs as well as search them easily.

Figure 2. Supervising Controller pattern

2.1.2. Passive View pattern

The ASP.NET MVC platform contains outstanding

feature that is lighweigt and easy testing for interface while
comparing it with Web Forms application. Moreover, it
integrates available features of ASP.NET. The definition
of ASP.NET MVC platform has been taken by the name
space System.Web.Mvc [2].
We have implemented the version 4.0 of ASP.NET
MVC since it appends ASP.NET Web APIsthat support
mobile devices better.
In this system, we have not used complicated SQL
instructions. The querying of data has been done via
LinQToEntities.
2.3. System analysis and design
2.3.1. Actors
The system consists of tow user groups (Figure 4):

Figure 3. Passive View pattern

1. Boarders: use the system to exploit information and

interact with managers in the boarding process.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

2. Management staff: use the system that serves the

management tasks with appropriate decentralization, i.e.
there is different authority for each employee.
Dormitory
management system
Management staff

Boarders

Figure 4. The system actors in Dormitory

managementinformation system

Management staff not only manage the system but also

edit news as well as activities before publishing them on
the website.
2.3.2. Use case diagram of Dormitory management
information system
The system has to ensure to provide enough procedures
that attend to management activities manually (Figure 5).
Registration and
extending management
for staff

facility management
Registration
service

management.

Boarders

Management
staff

Sent
requirements
Receiving and Resolving
of student requirements

<<extend>>

<<extend>> <<extend>>

Statistical process
and report status of
boarders

Other

<<extend>>

Change
room

Check
out

c. Use case Examination of dormitory

(manually and automatically)

room

With manual examination function, management staff

can permit the students to join in the rooms selected by
them or assign them into other rooms for any reason.
Moreover, management staff can delete invalid requests.
With the automatic examination function, the system
will rely on the student priorities as the first parameter. The
second parameter is the registration time to assign students
into appropriate rooms according to their registers.
Management officers can manage the facilities of the
whole dormitory or managed by each room, make statistics
of the number of assets that are damaged or need to be
repaired or replaced.

for Students

Dormitory service

After being logged in successfully, the system will list

the qualified rooms for non-resident students. They can
choose the rooms that they would like to join in. For
boarders, the system allows them to extend their room or
choose another one. While selecting the rooms and sending
information,
the
registration/extending
may
be
cancelled/suspended if the management staff has not
examined it yet. When the request has been approved, the
students will receive an anouncement containing the full
information to let them do boarding procedures.

d. Use case of dormitory facility management

Registration and
Extending Management

Dormitory

77

Repair

Figure 5. Use case diagram of Dormitory management

information system

a. Use
case
Registration
and
extending
management for staff
Before starting new curriculum, management staff
login to make a schedule the registration and extending of
boarding in the dormitory. Then they are going to inform
students about it. After its deadline, the system will close
registration/extending automatically. Management staff
could assign conditional students into rooms manually and
automatically based on priorities and registration time.
When the assignation completes, the system will send
students a confirmation and quotations to let them pay it
and take the boarding.
b. Use
case
Registration
and
Extending
Management for Students
In order to register or extend room in the dormitory,
firstly, students have to log in the system before the
deadline as mentioned above in the schedule. Non-resident
students can only register rooms whereas boarders have the
only option of extending their rooms.

When repair requests are sent by the students via

system, management officers will approve and send a staff
to do the repair. After completing the repair, repair
personnel will report to management officials about
updated assest status and repair activities.
e. Use case of dormitory service management
About Internet service of the dormitory, this kind of
service is room service and only need one student
representative to register.After registering and being
approved, students will receive notice and come to
management office to finish the procedures, then
management officers will send a staff to install internet. The
service charge will be notified by email to each student.
f. Use case Receiving and Resolving of student
requirements
Boarders can submit their requests to dormitory
management via messages on the system: request for room
transfer, check, request for repairs and other requirements.
With each boarder, the request for repairs and other
requirements are not made simultaneously with the request
for transfer and check. After boarders send request,
management officials will review to make a decision to
approve or not approve of the requirement via system.
g. Use case Statistical process and report status of
boarders
After logging in the system, managers can select the
information required for the statistics to view and export
the results. There are two statistical functions: quick
statistics and detailed statistics.
Quick Statistics provide general data about students,
facilities and boarding fees. Detailed statistics provide

78

Dang Hoai Phuong, Nguyen Van Dong, Dinh Thi Hong Huyen, Pham Thi Thu Hien, Lam Thi Hong Tam

specific data for each item. All statistical results can be

exported to the forms and have an attached specific list,
making the report quick and accurate.
2.3.3. General structure of system
The system structure includes the following principal
functions (see Figure 6):
Dormitory
management system

beds, boarding fees Fixed assets in rooms include the

code of assets, name of assets, price, use status Common
assets of dormitory: include the code of assets, name of
assets, price, use status, etc.
Registration or extension of student: Student ID
number, registration room, time, approved room, the status
of the registration or extension.
Request: Student ID number, type of request, content,
time and condition.
Services: type of services, service charges, and form of
payment.

Management staff

Boarders

Rooms management

STUDENTS
- idStudent
- studentNumber
- firstName
- lastName
- class
- adrress
...

registration rooms

Registration and
extending management

Extending rooms
Requirement

Students/ Staffs
management

Room information

Requirement /Notification/
Message management

0..n

Rent rooms
management
Figure 6. The structure of system.

User gesture

:CONTROLLER

REGISTRATION_ROOM

1..1
0..n

0..n

- idRenewal
- idRentRoom
- idRoom
- time
- statusOfRenewal
REQUEST

- idAsset
- name
- quantity
- statusOfAsset
- unitPrice

Change
Database

- idRequest
- idRentRoom
- type
- content
- statusOfRequirement

Return

1..1

SERVICE_IN_ROOM

Return
result
1..1

Figure 7. Sequence diagram of the dormitorymanagement

informationsystem.

0..n

RENEWAL_ROOM

ASSETS
ExecuteOfRequest()

1..1

1..1

1..1
0..n

CheckData()

- idRentRom
- idAccount
- idRoom
- statusOfRentRom
0..1

- idAssetOfRoom
- idRoom
- idAsset
- quantity
- statusOfAsset

Invoke Action

- idRegistration
- idAccount
- idRoom
- time
- statusOfRegistration
RENT_ROOM

1..1

ASSETS_IN_ROOMS

:MODEL

0..n

1..1

1..1

Sequence diagram of the dormitory management

information system in MVC model (Figure 7):

- idAccount
- idStudent
- password
0..n

- idRoom
- RoomNumber
- quantityOfStrudent
- boardingFees
- statusOfRoom
0..n

Statistical

VIEW

0..1

ROOMS

facility management

Management
staff/ Boarder

ACCOUNTS
1..1

- serviceInRoom 1..1
- idService
- idRoom
- dateOfDeadLine

SERVICES
0..n

- idService
- nameService
- price

Figure 8. Class diagram of system.

2.3.4. Analysingdata system

The main database of system management includes
(Figure 8):
Boarder: Student ID, accounts and related personal
information.
Facilities of Dormitory: include rooms, fixed assets
and common assets of the dormitory. Rooms in the
dormitory include number of rooms, blocks, the number of

Figure 9. Registration and extending management functions

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Dormitory management information system is

implemented in details according to the system design,
including the management part for staffs and the using part
for students. Following are some screen images of the
dormitory management website: Dormitory registration
and extending request activation for the manager (Figure
9); Room information for students to register (Figure 10).

79

Department of student affairs and dormitory of University

of Science and Technology, The University of Danang.
The experimental results show that Dormitory
management information system has achieved the
objective and the requirements of the Dormitory
management department.
However, to make the use of Dormitory management
information system more flexible and convenient, in the
near future, we continue to develop a version for mobile
devices like smartphones and tablets.
REFERENCES
[1]
[2]
[3]
[4]

Figure 10. Function of room information viewing

3. Conclusions
The Dormitory management information system has
been running tests with real data provided by the

[5]
[6]
[7]

Adam Freeman, Pro ASP.NET MVC 5 (Expert's Voice in ASP.Net),

5th Edition, Apress, 2013.
Jeffrey Palermo, ASP.NET MVC 4 in action, Manning Publications
Co., 2008.
Martin Fowler (2006), The evolution of MVC and other UI
architectures, http://martinfowler.com/eaaDev/uiArchs.html
Simple Example of MVC (ModelViewController) Design Pattern
for Abstraction, 2008,
http://www.codeproject.com/Articles/25057/Simple-Example-ofMVC-Model-View-Controller-Design
Dormitory Management Center's website of National University, Ho
Chi Minh City, Internet: http://ktx.vnuhcm.edu.vn.
The dormitory website of Pedagogical University, the University of
Thai Nguyen, Internet: http://dhsptn.edu.vn/kytucxa/index.php/vi.

(The Board of Editors received the paper on 12/10/2015, its review was completed on 12/16/2015)

80

Tran Thi Ngoc Trang, Le Viet Man, Nguyen Minh Duc

REACTIVE MULTI-AGENT MODEL FOR COLLABORATIVE FILTERING BASED RECOMMENDER SYSTEMS

Tran Thi Ngoc Trang1, Le Viet Man2, Nguyen Minh Duc2
1

Hue University; mantrang27@gmail.com

College of Economics, Hue University; lvman@hce.edu.vn, ducnm@hce.edu.vn

Abstract - In recent 20 years, using multi-agent models has been

developed in many research fields, especially in social science.
These multi-agent m odels allow sim ulating and studying a com plex
part of real world by performing insilico test, or called real
sim ulation. Recently, some research has also proposed m ultiagent model for Information Retrieval problems and has achieved
some remarkable results. In this paper, we introduce a reactive
m ulti-agent m odel as a new approach for recommender systems in
order to overcome some comm on limitations of recommender
system s, especially recomputation problems when new data is
added to the system. Experimental results also indicate that the
proposed m odel can be applied for recomm endation problem s and
our model perform s more stably than collaborative filtering based
recomm ender system s.
Key words - Collaborative filtering; recomm ender systems; m ultiagent system s; reactive multi-agent model; reactive agent;
attractive force; repulsive force.

1. Introduction
In daily life, people usually rely on recommendations
from other people by spoken words, reference letters, news
reports from news media, general surveys, travel guides,
and so forth. Recommender systems (RS) assist and
augment this natural social process to help people sift
through available books, articles, webpages, movies,
music, restaurants, jokes, grocery products, and so forth to
find the most interesting and valuable information for
them. The most common technique used for
recommendations is collaborative filtering (CF). CF-based
RS predict user preferences for products or services by
learning past user-item relationships from a group of user
who share the same preferences and tastes. Although
owning many advantages in comparing to other techniques,
CF has been facing many problems needed to be solved,
such as data sparsity, scalability, similar items, grey-sheep,
black-sheep, false recommendations, privacy,.
Until now, there have been many methods proposed to
tackle all the problems of CF approach, such as hybrid RS
[15], graph-based RS [11], especially multi-agents based
RS [2, 7]. In this research, we propose a reactive multiagent model for RS in which user-rating list and the
methods for computing similarity are used based on Itembased CF technique. This solution is an new approach for
RS which offers precise recommendations based on
particular preferences of users with better performance
than CF- based RS.
The rest of this paper is organized as follows. In section
2, we review some existing works about CF approach and
multi-agent systems. Next, in section 3, we first give an
overview of proposed model, reactive agents and then the
method for determining attractive and repulsive forces as
well as self-organized model. The results of an

experimental evaluation are presented in section 4 with the

use of a movie database called MovieLens 100K. The
paper ends with a discussion of the limitations of the work
and an outlook on possible directions for future work.
2. Related works
2.1. Collaborative filtering-based recommender systems
Most of RS basically rely on three methods: contentbased, knowledge-based and CF-based where CF is the
approach which has been used most widely. CF-based RS
provide personalized recommendations according to user
preferences. They maintain data about active users
purchasing habits or interests and use this data to identify
groups of similar users. They then recommend items liked
by similar users. CF systems offer two major advantages:
Firstly, they do not take into account content information,
and secondly, they are simpler and easier to implement.
Further, ignoring content information allows CF systems
to generate recommendations based on user tastes rather
than the objective properties of domain items. This means
that the system can recommend items very different from
those that the user had previously shown a preference.for.
Mathematically, CF algorithms represent a user as an
M-dimensional vector of items, where M is the number of
distinct catalog items. By computing the similarity of users,
a set of nearest neighbours whose known preferences
correlates significantly with a given user are found.
Preferences for unseen items are predicted for an active user
based on a combination of the preferences known from the
nearest neighbours. Filtering these neighbours is equivalent
to computing the distance among M-dimensional vectors.
Accordingly, CF algorithms are categorized as memorybased filtering and model-based filtering. Memory-based
filtering computes distance between vectors by using
Euclide distance, Pearson correlation, whereas modelbased filtering is considered as an approach to solve some
limitations of memory-based filtering, especially scalability
problem. In particular, machine learning techniques (such as
PCA Principal Component Analysis [10], MDS Multidimensional scaling [14] or SOM SelfOrganizing Maps
[9]) are used in model-based filtering in order to map
M-dimensional vectors into 2 or 3-dimensional space in
order help the process of computing distance, clustering or
classification to be easier. Despite getting some effective
results, these techniques still have some disadvantages, such
as data sparsity, data change, computing complexity, decline
of recommendation quality,
Recently, a new approach about user preference data in
RS has been proposed by representing user preference
matrix in form of graph and using graph theory to solve some

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

problems about computing the similarity between users [11].

Also, with graph-based approach, ODonovan [13] draws a
graph of user preferences in 2-dimensional (2D) space and
recommendation is operated by computing the distance
between user nodes in the space. In spite of reducing
computation complexity, this system still uses memorybased and model-based techniques, thus it also faces
common problems of CF algorithms. However, the idea
about drawing a graph and computing similarity between
users/items in 2D space in [8, 11] will be aslo applied for
computing the similarity between items in our system.
2.2. Multi-agent systems
Muti-agent systems (MAS) refer to a computer research
domain that addresses systems which are composed of
micro level entities (agents), which have an autonomous
and proactive behaviour and interact through an
environment (either virtual environment or real
environment), thus producing the overall system behaviour
which is observed at the macro level [6]. Until now, MAS
have been considered as an interesting and convenient way
of understanding, modeling, designing and implementing
different kinds of (distributed) systems [5]. Futhermore,
MAS also represent a very interesting modeling
alternative, compared to equation based modeling, for
representing and simulating real-world or virtual systems
which could be decomposed in interacting individuals [4].
There are many types of agents used in MAS, such as
assistant agents, collaboration agents, mobile agents and
reactive agents where reactive agents have widely used in
many fields, especially information retrieval. Two typical
systems which use reactive multi-agents are presented in [3,
12]. Particularly, in [12], Renault used dynamic attractive
and repulsive multi-agent model which aims to organize
emails in a 2D space according to similarity where each
email is represented by an agent and there is no need to
specify axes as well as how to organize information. The
model allows agents to communicate with each other
through virtual pheromones and collectively auto-organize
themselves in a 2D space. Without much constraints, the
system can organise (like clustering/classification)
information and let the user intuitively interact with it.
Based on the idea of Renault, Cao Hong Hue et al. [3]
presented a new model for image browsing and retrieval
which uses a reactive multi-agent system supporting
visualisation and user interaction. Each agent represents an
image. These agents move freely in the space which their
routes are not predefined. They just react to external stimuli
sent by other agents. Each agent interact to others through
forces, either attractive forces or repulsive forces. These
forces are generated by the visual and textual similarities
between an agent and its neighbours. Thus, the agents are
attracted by similar agents and repulsed by dissimilar
agents. This model is operated according to loop steps by
the time. In each loop step, agents change their position in
the model. Forces between agents or neighbours cause
these changes. Selecting neighbours in each time step

Item mentioned here is an item in RS

81

makes this model operate really slowly. That is the main

limitation of this model.
The multi-agent systems proposed in [3, 12] are
equivalent to the core idea of RS which use the similarity
among agents to organize data. Also, RS use similarity
between users or items to extract a list of recommendation
items.
However,
in
CF-based RS, selecting
recommendation lists usually uses complex computing
formulas whereas, using attractive and repulsive forces
between agents will help computing process become easier
by just finding neighbours (in 2D space) of each agent.
This is the main idea used for our proposed model.
3. Reactive multi-agent model for CF method
Gii thch: Trong phn ny, m hnh a tc t phn ng
vi mi trng c th hin kh r qua cc phn nh
m chng ti nu bn di. Vic trnh by phn ton
hc ca m hnh ch yu xoay quanh vic tnh ln ca
lc v hp lc tc ng ln mt agent. Theo , vic tnh
ton ln ca lc c ni r trong phn 3.2. Cn i
vi hp lc tc ng ln mt agent, r hn, chng ti
c b sung mt phn ghi chv vic tnh tng hp lc tc
ng ln mt agent da trn cc lc tc ng ln mt agent
v cc lng ging ca n (Figure 3).
3.1. Model overview
The proposed model uses reactive agents in which each
agent represents an item1 and actions of each agent depend
on list of users ratings for that item. The agents move freely
in a 2D environment which has no pre-defined axes or
meaning (Figure 1). They are reactive and only react to
outside stimuli sent by other agents. Each agent interacts
with others through forces (either attractive forces or
repulsive forces). Forces originating between agents are
computed based on the similarity. Two agents attract each
other when their similarity is high and repulse each other
when their similarity is low. According to the sum of
attractive and repulsive forces acting upon an agent, these
agents will move to the new position in the space. Ihere,
agents interact to new agents and then continue moving. The
movement of agents will be ended when they reach to stable
status. This helps to create a self-organized model in 2D
space. At steady status, two closed-agents are similar to each
other and they can be used for recommendation process.

Figure 1. The environment of agents. Each agent is represented

by an image which corresponds to a poster of a movie

82

Tran Thi Ngoc Trang, Le Viet Man, Nguyen Minh Duc

As presented above, at each time step, an agent interacts

with its neighbours, gets forces from them and moving
reactively. Hence, computing forces only can be done
when we get list of neighbours for an agent. In our model,
neighbours can be chosen according to four methods
including proximity, sampling, random and defined area
(Figure 2).

is the average rating value of all users for item j.

Force direction is characterized by the type of forces
(either attractive forces or repulsive forces). These forces
show that the behavior of an agent is toward or away from
other agents. In local level, agents behavior is determined
by the similarity or dissimilarity among agents.
Accordingly, force direction is determined as follows:
- If two agents are similar then they will attract each
other. It means that they tend to be closer.
- If two agents are not similar then they will repulse
each other. It means that they tend to be separated.

Figure 2. Methods for choosing neighbors. Proximity: Choose

neighbors in fixed-radius; Sampling: Choose randomly some neighbors
in the list of closed-area; Random: Choose randomly all agents in the
model; Defined area: Choose agents from a specific area

Force magnitude belongs to the similarity and the

distance among them is combined to form the force
characteristic. However, in practice, it is difficult to define
exact value of the similarity and distance between forces.
Thus, we determine force magnitude according to
continuous approach as showed below (Figure 4).

From experimental process, in local level, we choose

neighbours by using proximity approach which allows
selecting agents in fixed radius. And in global level, we
select agents according to random approach which
randomly pick agents from all agents in the model.
Once the neighbour list for an agent is known, then this
agent can simply compute the forces received from all
these neighbours and react according to them (Figure 3).
Figure 4. Force characteristic and magnitude basing on
similarity and distance (continuous approach)

As clearly seen from the Figure 4, there is always a

neutral threshold of forces. This threshold is the basic to
determine force types:

Figure 3. An example of reaction of an agent toward two

neighbours. This image shows the rule for summing forces, each
agent interacts toits neighbours. The force generated from these
interactions will be combined for making the final global force.
This final global force for an agent is simply the vectorial
summation of all forces between that agent and its neighbous

3.2. Attractive and repulsive forces (item-based forces)

A force applied between two agents can be attractive or
repulsive and is characterised by a vector with direction
and magnitude. However, firstly, we need to determine the
similarity between agents. In item-based CF method, the
similarity between agents is usually computed by using
Pearson correlation. Implementation results obviously
show that this method is widely used in the CF research
community and gives better results than other methods
[13]. The similarity between items is computed according
to the following formula:
,

,
,

where wi,j is the similarity between item i and item j, U is

the set of users rating for both item i and item j, ra,i is rating
value of user a for item i and is the average rating value of
all users for item i, ra,j is rating value of user a for item j and

If results are higher than neutral threshold, we have

repulsive forces which are computed by:

If results are lower than neutral threshold, we have
attractive forces which are computed by:

where w is the similarity between two agents; ,

,
are respectively mean value, maximum value and
minimum value for active agents neighbors; d is the
distance between two agents computed by Manhattan [22].
3.3. Self-organized model
During the evolution of the model, agents gradually
move to a status position with indefinite route. Thus, our
model is similar to self-organized model in machine
learning. However, also unlike the model proposed by Cao
Hong Hue et al. [5], our model uses two levels: local level
and global level.
Local level: Agents choose their neighbors according
to proximity approach which divides the space into
separate areas for operating independently. Local force
generated from agents helps to create clusters which are
disposed sparsely in the space. However, local level does

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

not offer the high accuracy for the model. So, a global level
is needed to break down the local connections and collect
small groups together in order to enhance the accuracy of
the model.

83

After having the prediction for items which maybe

liked by active user, the system collects all the films which
are unseen by active userwith highest predicted ratings.
The list of recommended films is described in Figure 7.

Global level: Agents choose their neighbors with

random positions in a large area. Force originated in this
level are called global force which is combined to a local
force to form an associated force (according to force
association show in Figure 3. above).

Figure 5. Simulation on local level (a) and global level (b)

Figure 7. List of recommended films with predicted ratings

(according to measurement scale from 1 to 5)

4.2. System evaluation

4. Experimental results
4.1. System implementation
Gii thch: Theo yu cu ca phn bin, phn ny
chng ti b sung thm mt kt qu ca qu trnh ci t
thc nghim trn h thng t vn film nhm m t trc
quan kt qu ca qu trnh t vn (Figure 7). Hnh ny m
t danh sch cc b phim m mt ngi dng no c th
thch, km theo l gi tr d on cho tng b phim .
System is built by using Objective C and Open Graphics
Library (OpenGL). To evaluate the performance of the
system, we use dataset MovieLens 100K including 100000
ratings (with scale from 1 to 5) from 943 users for 1682
movies. Each user rated at least 20 movies and supplied
demographics information (age, gender, occupation,).
Begin

Pick list of items rated by active user

Choose k nearest neighbors with
item needs to be predicted

After offering prediction value for active user, we

compute prediction accuracy (MAE) for five testing data
sets. This result is illustrated in Figure 8. below:

Figure 8. MAE for five testing data sets

The Figure 8. shows that attractive and repulsive multiagent model give accurate prediction with the average of
MAE of 0.724. Meanwhile, this value for item-based CF
proposed by Badrul Sarwar et al. [2] is 0.723. It can be seen
obviously that recommendation results offered by our
proposed model and previous CF methods are equivalent.

Get mean rating value of k nearest

neighbors
End

Figure 6. Prediction algorithms with input is the item which

needs to be predicted

After the operation in 300 time steps, we recognize that

forces acting upon agents are gradually decreasing to 0,
agents do not move any more, the distances among agents
do not change as well. At that time, the model reaches
stable status. Because the proposed model is a
self-organized model of agents in the space, the similarity
of agents is shown exactly in this model. Hence, the result
of prediction will be the rating values for nearest movies to
the one needed to be recommended in the space. The
prediction algorithms is illustrated in Figure 6.

Figure 9. MAE values for the proposed model

and traditional CF technique

Otherwise, the quantity of selected neighbors

significantly influences on the MAE value. Experimental
result (Figure 2) denotes that if the number of selected
neighbors is under 50 then MAE value is quite high, if the
quantity of selected neighbors is over 50 then MAE value
is quite stable and decreases regularly. This proves that our
model works more stably than traditional CF techniques.

84

Tran Thi Ngoc Trang, Le Viet Man, Nguyen Minh Duc

What is more, our proposed model also overcomes the

common limitations of traditional CF methods related to
computation time and scalability when new item or new
user is added to the system. Indeed, adding new objects to
our system means that adding agents to the model, then
computation is processed and agents will move in the space
until they find the exact position. That is the main
advantage of our model.
5. Conclusion and future works
The paper proposed a reactive multi-agent model for
item-based RS. With MovieLens 100K dataset,
recommendation movies are the acquired result based on
the analysis rating values of hundreds of former users.
Experimental results also indicate that attractive and
repulsive multi-agent model can be used as an alternative
approach for CF techniques with more stable performance.
Moreover, the model solves problem of recomputing when
a new item is added to the system. This research is the basis
for future works of reactive multi-agent model for RS with
many improvements in the performance, the ability of
visualization and interaction so as to enhance
persuasiveness, transparency and satisfaction for
explanations in RS. Furthermore, by combining item
agents and user agents in the environment, supplementing
knowledge/content into agents will help to give more
intelligent and exact recommendation results.
REFERENCES
[1]

Badrul Sarwar, George Karypis, Joseph Konstan, John Riedl, ItemBased Collaborative Filtering Recommendation Algorithms,
Proceedings of the 10th international conference on World Wide
Web, (2001), pp. 285-295.

[2]

Bsarry Smyth, Brynjar Gretarsson, Svetlin Bost, Tobias Hllerer,

Peerchooser: visual interactive recommendation, Proceedings of the
2008 Conference on Human Factors in Computing Systems,
Florence, Italy, (2008), pp. 1085-1088.

[3]

[4]

[5]

[6]

[7]

Cao Hong Hue, G., A., A multi-agent model for Image browsing and
retrieval, Studies in Computational Intelligence Volume 457,
(2013), pp. 117-126.
F. Klugl, C.Oechslein, F.Puppe, andA.Dornhaus, Multi-agent
modelling in comparison to standard modelling. In F.J. Barros and
N. Giambasi, editors, Proceedings of AIS 2002: Artificial
Intelligence, Simulation and Planning in High Autonomous
Systems, San Diego, CA, (2002), pp. 105110.
F. Zambonelli and H.V.D. Parunak, From design to intention: signs
of a revolution, inProceedings of the first international joint
conference on Autonomous agents and multiagent systems, ACM
Press, (2002), pp. 455456.
Fabien Michel, Jacques Ferber, and Alexis Drogoul, Multi-Agent
Systems and Simulation: a Survey From the Agents Communitys
Perspective, CRC Press, LLC, (2001).
Guerra-Hernndez, Alejandro, Amal El Fallah-Seghrouchni and
Henry Soldano, Learning in BDI Multi-agent Systems,
Computational Logic in Multi-Agent Systems, (2004).

[8]

J. O'Donovan, B. Smyth, B. Gretarsson, S. Bostandjiev, and T.

Hllerer., PeerChooser: visual interactive recommendation,
presented at the CHI '08: Proceedings of the SIGCHI Conference on
Human Factors in Computing Systems, New York, New York, USA,
(2008), pp. 10851088.
[9] Laaksonen, J., Koskela, M., Oja, E., PicSOM Self-organizing
image retrieval with MPEG-7 content descriptors, IEEE
Transactions on Neural Networks 13(4), (2002), pp. 841-853.
[10] Moghaddam, B., Tian, Q., Lesh, N., Shen C., Huang, T.S., Visualization
& User-Modeling for Browsing Personal Photo Libraries, International
Journal of Computer Vision 56(1/2), (2004), pp. 109130.
[11] Nguyen Duy Phuong, Le Quang Thang and Tu Minh Phuong, A GraphBased Method for Combining Collaborative and Content-Based
Filtering, Trends in Artificial Intelligence, Springer, (2008), pp. 859-869.
[12] Renault, V., Organisation de Socits d'Agents pour la Visualisation
d'Informations Dynamiques, PhD thesis, University Paris 6
(France), (2001).
[13] Ricci F., Rokach L., Shapira B., Kantor P.B, Recommender Systems
Handbook, 1st Edition, Springer, 845 p. 20 illus., Hardcover, ISBN:
978-0-387-85819-7, (2011).
[14] Rubner, Y., Guibas, L.J., Tomasi, C., The earth movers distance,
multi-dimensional scaling, and color-based image retrieval, In
APRA Image Understanding Workshop, (1997), pp. 661-668.
[15] Xiaoyuan Su and Taghi M. Khoshgoftaar, A survey of collaborative
filtering techniques, Advances in Artificial Intelligence, (2009), pp. 120.

(The Board of Editors received the paper on 04/17/2015, its review was completed on 06/22/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

85

FDI ATTRACTION FOR INDUSTRIAL DEVELOPMENT EXPERIENCE

FROM DEVELOPED COUNTRIES AND APPLICATION LESSONS FOR LAOS
Nguyen Ngoc Anh1, Lamngeun Xayasene2
1

University of Economics, The University of Danang; ngocanh14204@yahoo.com

2
PhD Course 21, The University of Danang; ngeunh@gmail.com

Abstract - This paper includes experience in attracting FDI into the

industrial branch of some countries in the world such as newly
industrialized countries (NICs), China and ASEAN countries
(Malaysia, Thailand and Vietnam). These countries have the sam e
starting point as Laos in terms of natural, economic and social
characteristics, and they have been very successful in FDI
attraction to develop their industry. On this basis, with reference to
the specific characteristics of Laos econom y, the paper draws
som e lessons for Laos to attract more FDI into its industrial sector,
including the im provem ent of the investment environm ent,
administrative procedure reform, the opening of some fields, the
construction of industrial zones (IZs), export processing zones
(EPZs), investm ent incentives, investm ent prom otion, hum an
resource developm ent, infrastructure and support industries (SI).
Key words - industry branch; export processing zones; industrial
zones; investm ent environm ent; FDI.

1. Introduction
The economy of Laos and its industry are still behind
the development due to the restriction of FDI companies
and the weakness of domestic companies. Therefore,
learning from other countries in the region, particularly
ASEAN countries is useful for the improvement of the
competitiveness of FDI attraction into Laos. These are
countries with similar charactertistics in geography, natural
resources and society. Moreover, the industrialization in
these countries began from their industrial agricultural
structures. However, they have reformed, constructed legal
bases and created attractive investment environments for
FDI. These lessons will be selectively applied into Laos to
attract more FDI for its industrial development.
2. Expericences in attracting FDI to develop the
industry from some countries in the world
2.1. Experiences from NICs
NICs are idential to Laos in the first stage of the
industrialization, in which the economy was unbalanced,
underdeveloped and exshauted after the war; income per
capita was low; natural resources were poor; the weather was
bad; small land was occupied by a large population; main
advantages inluded sea ports and cheap labour. However,
with appropriate policies in economy development and
attracting FDI, these countries became newly industrialized
countries with high ecocomy growth rates, specifically in
Singapore, this rate was 12.2% p.a (19651980), industry
constituted 29.1%GDP in 1990; in Korea, the growth rate
was 16.6% p.a (19651980), industry accounted for 40% of
GDP in 1990; in Taiwan, these figures were above 7% p.a
(19881991) and 42.3% GDP, respectively. Export
turnovers of industrial products in these countries during
2000-2003 were more than 60% GDP on average, in which,
more than 50% was exported to the American market [5].
Experiences that can be applied to Laos are as follows:

- Maitaining the stability of policical and social situations.

In the Cold war time between Socialism and Capitalism, when
many countries fell into war or strained relationships with
capitalist countries, these countries took the chance, kept the
policical and social situations stable in order to attract FDI.
Thus, this is a precious experience for Laos.
- Culivating the strength in labour and natural
resources. Abundant and cheap labour is the advantage of
these countries in attracting FDI. Nevertheless, they are
different in natural resources and geographical locations.
For example, exploiting its advantage of locating in the
centre of an international marine route, Singapore becomes
an ideal transhipment port and an atttractive general trade
zone for investors.
- Enhancing the infrastructure to create a convenient
environment for FDI to be used effectively.The weakness
of the infrastructure not only discourages foreign investors
but also destroys the advantages of FDI. Therefore, the
government ought to much attention to investing in its
infrastructure and education system. The investment in
infastructure in Korea (19601990) was 8.7%/GDP, in
Taiwan (19701990) it was 9.5%/GDP [6].
- Successfully forming EPZs to attract FDI.These countries
regard EPZs as an effective solution for new technology
diffusion into the country and promoting the right orientation
of other ecomomics. In Taiwan, FDI reached 13.8 billion USD
(19511993), in which, electronics constituted 54.7%,
manufacturing 35.5%, non-metal production and chemical
17.9% in 1970. Korea attracted 67 billion USD until 2002,
most were manufacturing, electronics (46% industrial
products). Singapore attracted more than 41 billion USD
(2002-2003), mainly electronic parts manufacturing (37%
industrial products). Hongkong got 13.5 billion USD up to
1993 (electronics was more than 53% industrial products) and
became the 10th FDI attraction point in the world [5]. The
experience revealed that many EPZs might not be effective but
the determination and selection of the right partners in finance,
engineering and technology was much more important.
- Actively exploiting knowledge and technology from
foreign countries through FDI. This is considered as the
key for the rapid development of the economy in general
and the industry in particular. The transferring of
technology not only depends on the investors but also the
specialists in the countries; hence, these countries focus on
three directions in exploiting the technology factor of FDI:
(1) determining and searching for solutions so that the flow
of FDI into the domestic country is the most and the fastest;
(2) establishing a mechanism for the investors to transfer
the technology; and (3) form an expert team, domestic
workers qualified to receive foreign technology.

86

- Establishing a flexible mechanism in the relationship

between various economic sectors to mobilize domestic
capital and FDI for industrial development. State-owned
enterprises in these countries account for only a small
percentage, yet they have important roles, which focus on
sectors where privatization has not been done, such as the
industries of petrochemicals, steel, ship-building, and raising
private capital, FDI to develop other sectors. FDI enterprises
have different degrees of importance in the NICs economy:
this rate in Taiwan, Korea is low, but that in Singapore and
Hong Kong is quite high; enterprises with 100% foreign
capital accounted for 16.7 % of the base, 55.9 % of
production and 42.7 % of workers in Singapore in 1981 [5].
In
summary,
besides
appropriate
ways
of
industrialization, NICs countries have exploited FDI
factors effectively to provide them a push for taking off,
and become a phenomenal growth in the industry.
2.2. Experiences from China
The Chinese industry has thrived since the Foreign
Investment Law was issued (1979). Until 2002, 414,000 FDI
projects were licensed with 816.66 billion USD registered
capital and 438.78 billion disbursements. China became the
second receiver of FDI in the world (1993-2001) and the
largest in 2002 (more than 50 billion USD, surpassing the
US) [7], and reached 100 billion USD in 2005. Besides its
large-scale advantage, China has made efforts to improve
and enhance the competitiveness of the investment
environment. Lessons learned from China were as follows:
- Increasingly investing and developing the
infrastructure. Infrastructure investment in China (2003 2004) accounting for 8%/GDP on average [6] has formed
a clear infrastructure system, connecting ICs, IZs and EPZs
with the national infrastructure system. During the 90s,
China constructed IZs, EPZs in areas with low levels of
industrialization, far from the urban but still attractive to
FDI because of the convenient connection with the national
infrastructure system, which help reduce costs and time.
Localities with better infrastructure (the coastal cities)
attract more FDI [8].
-Improving the regulatory environment. China is
always concerned about the interests of foreign investors
by regularly amending and supplementing the Law on
Foreign Investment. Many provisions are removed to
match international practices. In 2001, foreign investment
guidelines were issued, in which, many projects related to
the liberalization of the services sector (banks, insurance,
commerce, international trade, tourism, communications,
transportation, accounting services, finance and regulation)
were proposed to conform with WTO rules. The
mechanisms and policies to attract FDI have now become
more open after three times of amendments (1979-1990;
1991-2000; and 2001 till now).
- Developing high-quality workforce by positively
reforming education and training system at all levels,
sending people for training abroad since 1978 and
attracting foreign experts to meet the high-quality
employment requirements of investors. The number of
people trained abroad in 1990 was 280,000, in OECD
countries in 1998 was 86,000, in 2001 was 124,000, in the

Nguyen Ngoc Anh, Lam ngeun Xayasene

US in 2007 was 54,000 [8]. 480,000 foreign experts and

290,000 experts from Hong Kong, Macao, Taiwan came to
work in these countries in 2003 and 2004.
- Gradually expanding the category for incentives. The
innitial list of 186 priority industries was expanded to 262 and
the limit investment industries were reduced from 112 to 75 [3].
- Cautiously testing and successfully operating special
economic zones. These economic zones are convenient for
economic exchanges with the outside (Shenzhen, Shantou
bordering Hong Kong, Zhuhai bordering Macau, Xiamen
near Taiwan and Hainan island). The special preferential
regime and the syncly invested infrastructure in these zones
have attracted FDI with modern technology, which boost
exports, especially FDI from overseas Chinese.
Approximately 50% of FDI in China from the Chinese in
Hong Kong, Taiwan, Macau and Singapore was invested
in these zones in 2004.
- Effectively using preferential tax policies. Besides
policies on supporting infrastructure construction, land rent
reduction, relocation costs and clearance, tax policies are
very preferential and specific for each branches, regions
and forms of investment. Five economic zones, 14 coastal
cities and disadvantaged areas have very preferential tax
policies. Corporate income tax of 15% in the sector is
encouraged, investment in disadvantaged areas is exempt
from tax for 2 years after earning interest and pay 50% only
in the next 6 years.
- Encouraging the development of SI. SI such as
automobiles, textile and electronics developed rapidly
thanks to the development strategy of SI flexibility,
namely: (1) creating market conditions such as calling for
FDI assembly, which created cost advantages for attracting
big businesses and promote these businesses find local
partners, (2) building the conditions for the development of
domestic industrial enterprises (SMEs), and (3)
establishing the support of information, human resources,
and legal to form and develop links and components
market [1]. In addition to financial support, technology and
training for SMEs, China encouraged FDI to be invested in
the manufacturing industry and SI through incentives and
proactive solicitation to companies capable of technology
and market, and IZs establishment.
- Encouraging FDI investments into exports and
disadvantaged areas. FDI enterprises which directly
exported products manufactured abroad were classified
into corporates and encouraged to invest. If export sales in
these corporations accounted for 70% of sales, tax on sales
would be reduced to 10% [8]. Moreover, foreign investors
had more preferential tax policies for 10 years if the
investment projects were carried out in disadvantaged
areas. Foreign investment projects on the approved and
limited investment list will benefit significantly when
investing in these regions.
In short, China is considered as a typical example of
business environment improvement so that Laos can learn
from its experiences to increase competitiveness in
attracting FDI for industrial development in the current
trend of international economic integration.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

2.3. Experiences fromASEAN countries

2.3.1. Experiences from Malaysia
Malaysia has adopted many policies and measures to
encourage FDI for development industry, specifically:
- Continuously completing the law system and improving
the investment environment. Important legal documents
regulating relations related to FDI and creating a solid legal
basis for business activities were enacted and amended such
as the Companies Act (1965), Industrial Coordination Law
(1975), Investment Promotion Law (1968, 1986), Law on
Industrial Assurance (1978), Income Tax Law (1967),
Excise Tax Law (1967), Sales Tax (1972), Excise (1976),
Commercial Law (1972), and Customs Law (1967).
Furthermore, the investment environment in the country and
the necessary conditions for investment activities have
always been emphasized, done regularly and synchronized
with a clear mandate by the government [4].
- Issuing many preferential FDI policies specifically
and transparently, including policies to liberalize shares
ownership, applying fiscal incentives (from 1986), FDI
incentives through tax exemptions, tariff protection,
encouragement of FDI in finance, banking, insurance,
loosening regulations on real estate ownership to
foreigners, development of road, rail, sea and air
infrastructure. In addition, there were incentives for large
investment, high-tech, export-oriented projects and FDI in
underdeveloped areas.
- Actively promoting investment with various forms.
Regularly holding investment seminars at home and abroad
to help foreigners know about investment policies, sending
high-level delegations to introduce opportunities, mobilize
investment abroad (the target countries), publishing
multilingual publications, in conjunction with industrial
and commercial organizations in other countries to
introduce investment opportunities [2].
2.3.2. Expericences from Thailand
Thailand has adopted policies and laws to encourage
FDI, which are expressed in three specific laws:
- Foreign Business Act (1972) defined the rights of foreign
investors. The percentage of foreign capital contribution must
be more than 50%; foreign investors were permitted to invest
into the sector of group C (list of branches was divided into 3
groups namely A, B, C; investment into group B was possible
with the permission of the Thailand Investment Commission).
This law amendment (1999) allowed foreign investors to
invest in sectors such as construction, advertising and
accountancy; legal consultancy and capital contribution ratio
was decreased. In the early stage, Thailands attitude towards
FDI was different. The government did not provide incentives
to FDI in areas that local firms have strength. However,
Thailand then started to pay more attention to FDI by
changing share ownership restriction ratio of foreign
investors, allowing foreign companies to own 100% equity
when exporting 100% of production, companies that exported
at least 20% of production were eligible for duty free.
- Investment Promotion Law (1977, amended in 1991
and 2001) in order to create incentives for domestic and
foreign investors in Thailand. This law was divided into 3

87

preferential regions: region 1 included Bangkok and five

developed provinces surrounding Bangkok, region 2 is 12
provinces around Bangkok and region 3 is the remaining
provinces. Enterprises investing in region 1 were tax-free
for 3 years, reduced 50% of equipment import duty.
Enterprises investing in region 2 were tax-free for 3 years,
and 7 years for projects in the technology. Enterprises
investing in region 3 were excempted from tax for 8 years,
tax-free import of equipment and materials for
manufacturing for the purpose of exports for 5 years.
- Industrial Zone Act (1979) creating incentives for
investors in the IZs. This law grouped IZs into 2 categories:
popular IZs to serve common industrial projects, and EPZs to
serve the direct industry and other industries to export with
more incentives. Besides, in order to improve the
competitiveness to attract more FDI, Thailand provided
equipment import tax exemption for 61 operations, 8-year
income tax exemption for 19 SI sectors, duty free imports of
raw materials or income tax for a number of projects in region
1 and 2 which previously not entitled, signed investment
protection agreements with 21 countries, signed the agreement
on avoidance of double taxation with 40 countries [8].
2.3.3. Experiences from Vietnam
Vietnam has been very successful in attracting FDI in
recent years and the lessons to be drawn are as follows:
- Actively improving the legal environment towards
facilitating FDI. Vietnam Foreign Investment Law (1987) has
been amended several times. The first time (1990) allowed
private economic organizations in the country to have the
legal status to do direct business cooperation with foreign
partners, expanded the form of joint venture (with multiple
stakeholders), allowed joint enterprises to benefit from
financial incentives as manufacturing export goods
enterprises. The second time (1992) allowed private
businesses to cooperate with foreigners, regulated investment
into EPZs, added BOT (specific rules of management and
finance); the operational period was extended to 70 years;
foreign-invested enterprises were entitled to open loan
accounts in foreign countries. The third time (1996) adjusted
the field and areas of priority, the role of technology
assessment organizations, new regulations on the transfer of
capital; FDI enterprises could choose to apply accounting
regulations in accordance with the law of Vietnam or
international standards; joint-venture businesses could
forward losses by using profits in the next year (no more than
5 year); regulated capital proportion of foreign investors in the
joint venture was not less than 30% of the legal capital; the
legal capital of enterprises with foreign investment must be at
least 30% of the capital; enterprises with foreign investment
were allowed to transfer earnings abroad with tax rates from
5% to 10%. To restrict foreign workers and managers, the
government adopted income regulatory tax (0% - 50%). In the
fourth time (2000), in order to create favourable conditions for
attracting FDI, the government allowed foreign-invested
enterprises to open accounts in foreign banks, 100% foreigninvested enterprises might transfer losses, change the
decentralization in FDI for the local and branches. In the fifth
time (2005), the investment law was formed to unify the
investment environment for domestic and foreign investment.

88

- Developing policies encouraging FDI that is competitive

compared to other countries in the region. Corporate income
tax in Vietnam was quite low in the region (25% popularity,
preferentials from 10%, 15%, and 20%, the period of tax
exemption was long, which can be up to 8 years). Import
duties exemption for goods imported were applied to create
fixed assets (construction materials equipment which cannot
be produced domestically). FDI enterprises could forward
losses but not exceed 5 years from the time of incurred losses.
Investment projects in the encouraged regions, manufacturing
of components, mechanical, electric and electronic parts were
excempted from income tax for 5 years. Import tax exemption
for materials and raw materials were used to produce export
products, 0% value-added tax rate was used for enterprises
with foreign investment selling manufacturing products to
foreigners in Vietnam and appointed to deliver in Vietnam,
land tax exemption for investment projects came under the
statute BOT, BT and BTO.
3. Lessons drawn for Laos
Laos industry was formed in the 1970s but developed
mainly in Vientiane and produced goods to primarily serve
domestic consumer demand. In 1975, the nationalization of the
factory hindered the industry growth due to lack of capital and
technology. In 1986, Laos advocated to opening and
innovation. However, serious imbalance, unstable finance, high
inflation, difficult living conditions, low GDP, low investment
efficiency made the economy in general and Laos industry in
particular fell into a serious crisis. To rapidly mobilize domestic
and foreign capital, change the economic structures and
industrial structures, Laos specified a renovation policy by
enacting the Law on Foreign Investment (1988). Nevertheless,
FDI activities were still new and did not attract foreign investors
to the potential of Laos. From the successful experiences of the
countries, lessons for Laos to increase the competitiveness to
attract FDI in the industry are as follows:
3.1. Improving the investment environment
A stable and healthy investment environment is the
most important factor to attract FDI because it nourishes
the intention, forms and decides the existence of
investment decisions. To maintain a stable investment
environment, it is crucial to:
- Maintain the stability of the security, political, social
environment and enhance the role of the government.
These are considered to be important conditions to ensure
the effectiveness in the investment actitives. This lesson is
also shown quite clearly in Thailand about the decline of
FDI due to political instability in recent years.
- Ensure macroeconomic stability. This is a crucial
factor in attracting FDI because it creates a stable
investment environment and avoids risks for investment.
- Improve and change the regulatory system to be more
flexible, simple in encouraging FDI policies and
institutional frameworks to be closer to the developed
countries, reducing the vague distance to avoid risks for
foreign investors. Tax is among incentives to be regarded
as an important tool to mobilize resources for economic
development and create a fair competitive environment.
Besides, it is necessary to use other policies such as land

Nguyen Ngoc Anh, Lam ngeun Xayasene

leasing policy, clearance policy flexibly and reasonably.

These policies should be transparent and specific.
- Carefully prepare the financial foundation such as the
state budget, residence capital mobilization, outside capital
to invest in infrastructure and create a favourable
investment environment. This is the lesson from China in
attracting FDI in disadvantaged areas.
3.2. Enhancing administrative reform
Reforming the administrative procedures towards
creating favourable conditions for investment activities is
considered an urgent requirement to improve the business
environment. This is the lesson learned from the
transformation economies like China and Vietnam. They
have created FDI attraction thanks to positive reforms of
administrative procedures on the following aspects:
- Robustly Reforming appraisal and licensing,
improving the service efficiency of government agencies,
focusing on after-licensed services, removing barriers,
equal treatment, guaranteed and transparent commitment
and government regulation.
- Strengthening the role of the authorities in foreign
investment is needed to create an attractive investment
environment, ensuring the correct orientation of foreign
investment, efficiently exploiting the resources, protecting
environmental resources, employees and sustainable
development. All levels of the government examine,
supervise and accompany foreigners to guide, facilitate,
remove obstacles, comply with their commitments, and
resolutely retrieve delayed and suspended projects.
3.3. Gradually and rationally opening each
optimizing the structure of foreign investment

area,

This is the policy that China has successfully applied. The

government initially allowed foreign investors in laborintensive sectors (textiles, garments, processing), then
gradually expanded the scope of the market to energy, raw
materials, tried to improve the structure of FDI and
encouraged foreign investors to invest in high-tech industries,
unified regulatory environments, provided incentives for
domestic and foreign investment, opened little by little
investment conditions (export, localization, development of
resources), allowed foreign investors to invest in sectors that
law did not restrict, diversified forms of FDI, eliminated
capital restrictions. Therefore, Laos should have policies to
encourage FDI in each area, each period and improve FDI
structure towards encouraged sectors and high-tech industries.
3.4. Gradually and firmly building up economic zones,
and the key areas
Attracting further FDI through constructing and
developing centralised IZs and EPZs are to be done firstly in
key economic areas, then this should be spread to other areas.
This is the lesson successfully applied in Vietnam and China.
These countries formed IZs and EPZs in coastal regions,
where there were favourable conditions for production and
business activities of foreign investors. Furthermore,
attractive particular mechanisms and policies in these zones
helped foreign investors establish factory easily and
conveniently as well as reduce operating costs. Thus, these
areas attracted considerable FDI of the whole economy.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

3.5. Implementing incentives to attract more FDI is

competitively and reasonably
Pioneering countries carried out effective policies in
attracting FDI such as tax incentives, land rent,
administrative reform, employee training costs support,
infrastructure investments, tax support for projects using
high technology, special tax rates for encouraged
investment areas and protectionist policies elimination.
These are tools applied in most countries to promote
investment. However, in the conditions of current
globalization, Laos needs to consider the advantages and
disadvantages of each sector and region in comparison with
other countries to offer competitive, specific and
transparent preferential policies.
3.6. Actively pursuing investment promotion
Countries like Malaysia, Thailand and Vietnam have
actively organized investment promotion, established
representative offices in some key countries and held
national and international conferences and seminars to
introduce the advantages of the country as well as sectors,
industries and projects needed priority in attracting FDI.
Thereby foreign investors were able to obtain more
information and support during their investment.
3.7. Developing the human resources
NICs highly appreciate education of humans because
they are the subject of all creation. Moreover, humans are
important objects in receiving, transferring technology
between foreign investors and host countries. Therefore, it
is necessary to improve the quality of human resources,
develop and basically innovate the education and
vocational training in the direction of meeting labour
market needs, foreign investors, the economy and
employment requirements for workers. Developing and
improving the labour market should approach international
standards on education, vocational training, labour and
employment consistent with international practices and
commitments of the country in the integration process. In
cases no training could be applied, policies are required to
attract human resources. Besides, it is necessary to raise the
level of public authority staff, especially project appraisal
and foreign investment management officers.
3.8. Investing in infrastructure development
Experience from China shows that they succeeded in
attracting FDI due to their infrastructure development. With
a low base, the infrastructure investment needs of an
underdeveloped country like Laos was very large. whereas
the budget is limited; hence, in order to solve the conflict
between budget limits and requirements of infrastructure
rapid growth, there should be a mechanism to mobilize all
resources in the country and abroad. With limited financial
resources, the state's investment focused on key
infrastructure projects, which motivated regional
development, then expanded to private investments in the
sector, and important projects with difficult capital recovery
and less attractive to investors. In addition to actively
seeking funding from foreign governments to invest newly
and modernly in important transport infrastructure such as

89

highways, airports, ports, railways, it was crucial to have

policies encouraging private and foreign investment in this
field on the basis of resolving the difficulties of investors.
3.9. Developing SI
With the trend of business globalization, MNEs only
keep important stages and intermediate products provided
by the outside. So international production location is
determined by factors that create global business activity
efficiency. Investors will choose a favourable location for
intermediate products as well as SI is the key to attract FDI
into the industry. Experience from China was focusing on
attracting FDI through simple procedures, taxes and
reducing ownership restrictions accompanied with
localization rateregulations. These policies helped the
domestic industry receive technology transfer from FDI.
However, regulations on localization rate are difficult to
implement today due to the violation of rules for commerical
liberalization. Therefore, some countries focus on the
development of IZs support, business incubators and ICs.
IZs support are being developed in Malaysia and Thailand.
4. Conclusion
This paper outlines the lessons learned to attract FDI
into the industry of some countries such as NICs, China
and ASEAN countries. Thereby, lessons to attract FDI into
Laos industry have been drawn as follows: (1) improving
the investment environment through maintaining stable
security, political and social environment, creating
macroeconomic environment stability and reforming the
legal system; (2) reforming administrative procedures; (3)
gradually and reasonably openingthe industries and
sectors; (4) investing in building IZs and EPZs in
advantaged areas; (5) effectively applying investment
incentives; (6) enhancing investment promotions; (7)
developing human resources; (8) investing in infrastructure
development; and (9) developing SI./.
REFERENCES
[1]

[2]

[3]

[4]

[5]
[6]

[7]

[8]

Hong Vn Chu (2011), Policies to develop support industries in

Vietnam up to 2010, State-level Scientific Research Work, Code:
KX.01.22/06-10.
Nguyn Tin Ci (2008), Malaysias FDI attraction policies in the
process of international economic integration - status quo,
experience and applicability to Vietnam, Ph. D Dissertation in
Economics, Hanoi National Economics University.
Trn Th Tuyt Lan (2014), Foreign Direct Investment in the
direction of sustainable development in the key economic areas of
Northern Vietnam, Ph. D Dissertation in Economics, Hochiminh
National Institute of Politics, Hanoi.
Phng Xun Nh (2000), Foreign Direct Investment for
industrialization in Malaysia Experience for Vietnam, World
Publishing House, Hanoi.
Nguyn Minh Phong (1999), Lessons of experience for attracting
FDI in the world, World Publishing House, Hanoi.
Nguyn Xun Thnh (2013), Attracting investment to local
economies: international experience, Proceedings of Central Coast
investment promotion conference in Danang.
Central Research Institute of Economic Management (2004),
Policies for economic development: experience and lessons from
China. Transport, Publishing House.
Vng c Tun (2007), Perfecting policy mechanisms to attract
Foreign Direct Investment in Hanoi in the period 2001 2010, Ph. D
Dissertation in Economics, Hanoi National Economics University.

(The Board of Editors received the paper on 12/06/2015, its review was completed on 12/14/2015)

90

Nguyen Thi Xuan Trang

A REVIEW OF SOCIETAL MARKETING

Nguyen Thi Xuan Trang
The University of Danang, University of Economics; trangntx@due.edu.vn
Abstract - The term societal m arketing has become more and
more familiar in marketing literature. However, there are still
confusions in understanding and using this term in practice
because there are also other marketing terms related to the society
such as social marketing, socio-cultural m arketing, nutri-marketing,
or environm ental marketing. This misunderstanding m ay lead to
difficulties in doing research on marketing as well as recognizing
necessities of each concept. This paper will go m ore detailed about
societal m arketing concept in terms of definition, significance,
differentiation with social marketing, requirem ents and its practical
expressions. The paper will be a useful reference to clear up
existing confusions.
Key words - societal m arketing; social m arketing; socio-cultural
m arketing; nutri-m arketing; environm ental m arketing.

1. Introduction
Philip Kotler, "the world's foremost expert on the
strategic practice of marketing according to the hail of
Management Centre Europe, gave definition of marketing
in the first edition of his textbook in the year of 1967. At
that time, marketing was defined as the analyzing,
organizing, planning and control of the resources, policies,
and activities of the firm in order to satisfy the needs and
wants of targeted customer groups (Kotler 1967, p. 12).
Then, in 1972 Kotler added the concept of societal
marketing that was mentioned on the necessary
combination of customer satisfaction and consumer
welfare in the long run in marketing.
2. Why is societal marketing important?
The term of marketing was firstly used on a lecture at
the University of Michigan in the United States in 1902.
Until now, it has become more and more popular all over
the world with different renewed definitions by Kotler and
other authors. In general, marketing can be defined as a
process by which companies create value for customers
and build strong customer relationships in order to capture
value from customers in return (Armstrong et al. 2009, p.
7). Therefore, marketing programs and marketing
strategies are designed based on customers needs and
wants. Marketing has proved that it is not only an
indispensable tool to connect businesses and consumers,
but also it is appropriate to apply in different types of
organizations such as production, commercial and service
companies, or profit and nonprofit organizations.
However, it also receives a lot of social criticisms due to
its negative effects on individual consumers, the whole
society and other businesses (Armstrong et al. 2009).
In terms of the benefits of individual consumers, many
critics charged that individual consumers have to incur high
prices of commodities due to high costs of distribution, high
advertising and promotion costs and excessive mark-ups
(Armstrong et al. 2009). When companies want to be more
competitive, they might spend more on advertising their
products; that will boost selling prices. Thus, in 2013, the

National Assembly of Vietnam added to the Enterprise

Income Tax Law with a clause related to marketing
expenditure. They would only accept marketing
expenditures, including costs of advertising, promotion,
brokerage commission, and others, that do not exceed 15%
of the total deductible expenses of the company in the
process of calculating income tax the enterprise should
incur. Although until now this clause has been faced by
several disagreements of firms, it shows that in some
aspects, marketing tool is criticized in the market of
Vietnam. In addition, marketing is accused of persuading
people to buy products that they do not think they will buy;
meanwhile, producers are always in the posture of accepting
the obsolescence of their products if there are any changes
in consumers behaviors. For instance, fashion marketing
may convince customers to buy fashionable clothes easily
and quickly; in the meantime, it puts pressure on producers
to find more new models before their products become
obsolescent. Finally, another disadvantage of marketing to
individual consumers is that some firms use it as a tool to
provide harmful or unhealthy products such as fast food or
tobacco to consumers.
Regarding the whole societys benefits, critics argued
that the marketing system creates a trend of materialism in
the society. A successful person may be assessed by having
a huge number of luxury assets like luxury homes, cars or
jewelry although he is poor at spiritual assets such as
compassion or kindness to others. Another drawback is that
cultural pollution happens through marketing system.
Advertisements will appear at any time people surf
information on the internet, or in any newspaper and
magazines people read. This is gradually changing the way
of thinking of citizens about materialism, relationships or
status over time (Armstrong et al. 2009).
Additionally, critics also argued in favor of harmful
impacts of marketing on other businesses such as acquiring
competitors, preventing new firms from penetrating the
market or even destroying competitors by cunning
(Armstrong et al. 2009). A typical example of acquisition is
the situation of Google Company. Google has been
acquiring other firms since 2001 and in some periods,
acquisitions happened every week. An interesting thing is
that Googles acquired firms can be in various business
sectors such as SlickLogin, a company developing soundbased password alternatives, or Nest, a company that
produces thermostats and smoke alarms, or Makani Power
that manufactures airborne wind turbines (Stunt, 2014). This
fact shows that Google Company prefers expanding by
acquiring competitors to launching its own new products.
From that, competition will be reduced and the risk of young
competitors being absorbed in the market will be high.
Responding to above social criticisms of marketing,

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

there have been appearances of two movements,

consumerism and environmentalism with the former
protecting the rights and power of buyers in the relationship
with sellers and the latter improving peoples living
environment. Before the drastic reactions of these two
movements, most companies realize that their marketing
strategies should not only enhance customer value, but also
reinforce customer relationships. In order to achieve this
objective, companies may choose the policy of enlightened
marketing to follow, that can put in the best long-term
performance of the marketing system based on five
principles: consumer-oriented marketing, innovative
marketing, value marketing, sense-of-mission marketing and
societal marketing (Armstrong et al. 2009, p. 563). Hence,
societal marketing is one of five principles that enlightened
marketing may be used (Figure 1) so that businesses will be
able to establish effective marketing systems.
Consumeroriented
marketing
Societal
marketing
Enlightened
marketing
Customer
value
marketing

Figure 1. Five principles of enlightened marketing

3. Difference between societal marketing and social marketing

Schwartz (1971), Kotler (1972) and Baker, M. (2003)
had arguments on societal marketing in common. All of
them agreed that societal marketing requires marketers to
find proper ways in setting up marketing strategies and
programs so that the impartiality between benefits of
customers and of the society as a whole will be guaranteed.
Table 1. Main differences between societal marketing
and social marketing
1. Scope of term
2. Essence of term

Societal marketing

Social marketing

Narrow

Wide

A Principle

A Process

3. Types of products sold goods and services

desired behaviors

4.Primary beneficiary corporate shareholders

society

5. Competition

can be summarized in the Table 1.

Firstly, the scope of social marketing is much wider
than that of societal marketing because in essence, while
social marketing is a process, societal marketing is only a
principle. Specifically, social marketing is a process by
which organizations influence public behaviors in order to
deliver positive benefits for society such as improving
health, preventing injuries or protecting the environment
(Lee and Kotler 2008, p. 7 - 8). In the meanwhile, societal
marketing is a principle of enlightened marketing strategy
that commercial sector marketing should go towards in the
process of creating value for customers and building strong
customer relationships. This can be illustrated in Figure 2.
Social marketing

Commercial sector marketing

Enlightened
marketing strategy
Societal
marketing
principle

Innovative
marketing

Sense-ofmission
marketing

91

with other organizations with current or preferred

offering similar goods behaviors of the target market
and services

6. Who
applies Organizations
in - Public sector agencies (such
societal marketing or commercial sector
as
World
Health
does social marketing
Organizations, Department
of Health, or Department of
Wildlife and Fisheries)
- Nonprofit organizations and
foundations
- Professionals working in a
for-profit organization
- Organizations engaged in
social marketing campaigns

Societal marketing should be distinguished with social

marketing. The main differences between these two terms

Figure 2. An illustration shows the scope of societal marketing

in comparison with the scope of social marketing

Secondly, desired behaviors are types of products sold

in social marketing to bring better societal gains than those
made by current or preferred behaviors in the market,
whereas societal marketing is established to sell goods and
services with the target of financial gains for corporate
shareholders and competes with other organizations
offering similar goods and services.
Furthermore, social marketing is a process being gone
through by public sector agencies (such as World Health
Organizations, Department of Health, or Department of
Wildlife and Fisheries), nonprofit organizations and
foundations, professionals working in a for-profit
organization, and organizations engaged in social
marketing campaigns (Lee and Kotler, 2008). On the
contrary, societal marketing is a principle that is suggested
to apply in organizations operating the commercial sector.
4. How to apply societal marketing?
Societal marketing concept holds that it is necessary for
companies to balance three factors, including customer
needs and wants, profits of the companies and human
welfare of the society in designing marketing strategies
(Figure 3).
Society
(Human welfare)

Societal
marketing
concept

Company
Consumers
(Profits)
(Want satisfaction)
Figure 3. Three considerations underlying the societal
marketing concept (Armstrong et al. 2009, p. 15).

92

Nguyen Thi Xuan Trang

In order to apply societal marketing, marketers need to

find ways not only to please consumers but also to enhance
long-term consumer benefits or the societys welfare.
Therefore, the first important thing for societally-oriented
marketers is delivering immediate satisfaction for
customers. Satisfaction is defined as feelings of pleasure or
disappointment a person gets when he compares the
products performance with expectations he held (Kotler
2003, p.61). Nowadays, most companies are raising
expectations of customers about their products or services;
at the same time, they try to match produced outcomes with
those expectations. Companies use different methods to
track customer satisfaction such as complaint and suggestion
systems, customer satisfaction surveys, ghost shopping and
lost customer analysis (Kotler 2003, p.64). For instance,
Saxon Business Systems provides Customer First Program
that guarantees the company to deliver services according to
expectations of their customers (Saxon Business Systems,
2014). Then, a comparison of customer satisfaction between
of the company and of competitors can be done to create
favorable conditions for marketers in setting up more
effective strategies in the future.
The second essential requirement for societally-oriented
marketers is offering the best long-term well-beings of
customers and communities. This requirement leads to the
fact that social and ethical considerations must be taken into
account into marketing practices (Kotler 2003, p.27). It
requires companies to impose stricter requirements when
they construct marketing programs through marketing mix.
In other words, corporate marketing ethics policies which
cover distributor relations, advertising standards, customer
service, pricing, product development and general ethical
standards (Armstrong et al. 2009, p.566) need to be
developed in companies.
IMMEDIATE SATISFACTION

LONGTERM
CONSU MER
B ENEFIT

Low

High

High

Salutary
products

Desirable
products

Low

Deficient
products

Pleasing
products

Figure 4. Societal classification of products

(Source: Armstrong et al. 2009, p.565)

Finally, producing desirable products is an optimal

approach towards societal marketing because Armstrong et
al. (2009) suggested that desirable products would be the
best combination of immediate pleasure and the long-term
benefits for consumers in comparison with deficient
products that are characterized by both low immediate
pleasure and long-term benefits, salutary products that
benefit consumers in the future despite low satisfaction at
the beginning, and pleasing products that can hurt
consumers in the long run but have high appeal when
Armstrong et al. (2009, p. 565) classified products into four
categories according to the relationship between
immediate satisfaction of consumers and the benefits
consumers receive in the long term (Figure 4). For

example, washing machines of Miele Company are

desirable products. The Miele Company has been so proud
of this product with the average 20 years of normal use
owing to this products overwhelming advantages in
saving energy and natural resources for the society. This
might prove that the company has applied societal
marketing that has created more competitive advantages
for the company.
5. Different expressions of societal marketing
5.1. Nutri-marketing
The first form of societal marketing can be mentioned
is nutri-marketing. Szakly and Berke (2004) proposed that
nutri-marketing is a type of marketing that tries to
communicate nutritional information to target groups of
consumers. In their research, from the fact of low life
expectation and high death rate in Hungary compared with
the average levels of Western European developed
countries, they realized that the poor heath status was a
main reason for this fact, and the structure of the Hungarian
food consumption would play an important role in
explaining the poor heath status in this country. Thus, one
of the best ways to improve living conditions would be
producing and consuming health-productive foods or foods
with high quality to citizens. Then, a quality model of food
was built up from the viewpoint of customers. This module
includes five factors, namely safety, usage, enjoyment,
nutritional and symbolic components. It can be seen from
this model that consumers highly appreciate nutritional
components when they think about the quality of the
product. However, in the research of Szakly and Berke
(2004), it was found that Hungarian consumed more
traditional foods than organic and functional foods because
of difficult availability, high prices and their insufficient
knowledge in terms of nutrition foods. From that, Szakly
and Berke (2004) supported nutri-marketing in setting
marketing strategies of companies.
Nutri-marketing is designed based on customer needs
and wants of health-productive foods with nutritional
components. Strategic foods involve organic and functional
foods that not only receive immediate satisfaction from
consumers, but also benefit their health in the long run. Thus,
nutri-marketing can be utilized as a tool for companies to
build up relationships with customers in the balance between
the profits and the whole societys benefits.
5.2. Socio-cultural marketing
Socio-cultural marketing may be also considered as an
expression of societal marketing with the application in the
tourism industry. In the research of King et al. (2000), they
examined the effectiveness of using societal marketing
approach to tourism planning in South Pacific Island
countries. They insisted on the connection of marketing and
socio-cultural issues in the process of planning. In order to
create national tourism development, each country should
produce desirable products, that are called as community
tourism products based on its societal and ecological
resources. These community tourism products will help to
connect benefits of consumers, who are visitors, of local
residents and of the host community (Figure 5).

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Host
community

Consumers
(visitors)

Local
residents

Figure 5. Three considerations underlying the societal

marketing concept to national tourism planning

5.3. Environmental marketing or green marketing

In conformity with societal marketing concept,
environmental marketing is defined as a process by which
companies identify, anticipate and satisfy the requirements
of the consumers and society in relationship with
profitability they achieve (Peattie, 1995). Krn et al.
(2003)
researched
on
social
responsibility
in
environmental marketing planning of European companies
in forest industry. They classified marketers based on
social responsibility values into three types involving
consumption marketers, reactive green marketers and
proactive green ones. While consumption marketers are
traditional marketers focusing on providing with
customers' satisfactions better than competitors do for the
purpose of sales and profits, both reactive and proactive
green marketers put emphasis on sustainability; however,
the former emphasizes on reaching sustainability under
governmental balancing and the latter
reaches
sustainability in a free market system.
Finally, they found that the necessity of environmental
emphasis in marketing strategies, structures and functions
were admitted in most companies in four European
countries; however, environmental issues in marketing
planning were emphasized more by proactive green
marketers than by consumption marketers or reactive
green marketers. Hence, they suggested that in order to
gain
competitive
advantage
towards
sustainable
development in business and society, marketers should
become more proactive and attempt to deliver
environmentally friendly products to consumers.
6. Conclusion
In conclusion, this paper analyzed the importance,
prerequisites and different forms of societal marketing.
Nowadays, societal marketing is an essential principle that
marketers should apply in the process of setting marketing
strategies and building marketing programs as the two
movements, consumerism and environmentalism, are
growing strongly. Societal marketing concept holds that a
company needs to reconcile interests of consumers, of the
company, and of the whole society. This can be done by
producing desirable products which are not only pleasing
but also beneficial to customers in the long run. This concept
is mostly used in commercial sector marketing with the
target of selling products and services; so it should be
distinguished with the terms of social marketing which is a

93

process related more in public sectors. Finally, this paper

mentioned various applications based on societal marketing
concept such as nutri-marketing, socio-cultural marketing
and environmental marketing. These applications have been
made by several companies in the food industry, tourism
industry and forest industry respectively, and receiving
generous support from the society.
This paper is expected to be a useful reference to clear
up most aspects of societal marketing consisting of the
definition, significance, requirements and its practical
expressions, as well as to differentiate between this term and
the term of social marketing. This is a theoretical paper and
it may encourage researchers to conduct empirical research
on the extent of applying societal marketing in corporations
in Viet Nam, a developing country in Asia, in the future.
REFERENCES
[1]
[2]
[3]

[4]

Armstrong, G., Kotler, P., Harker, M., & Brennan, R. (2009),

Marketing an introduction, Pearson Education Limited, England.
Baker, M. (2003), The Marketing Book (Fifth edition), Butterworth
Heinemann, Burlington MA.
Krn, J., Hansen, E. & Juslin, H. (2003), Social responsibility in
environmental marketing planning, European Journal of
Marketing, Vol. 37, No. 5/6, pp. 848-871.
King, B., McVey, M. & Simmons, D. (2000), A societal marketing
approach to national tourism planning: evidence from the South
Pacific, Tourism Management, Vol. 21, pp. 407-416.

[5]

Kotler, P. (1967), Marketing Management: Analysis, Planning, and

Control, Prentice Hall, Englewood Cliffs.

[6]

Kotler, P. (1972), Marketing Management: Analysis, Planning, and

Control (2nd edition), Prentice Hall, Englewood Cliffs.

[7]

Kotler, P. (2003), Marketing Management (eleventh edition),

Prentice Hall, United States.

[8]

Lee, N. R. & Kotler, P. (2008), Social marketing: influencing

behaviors for good (third edition), Sage Publications, Los Angeles.

[9]

Miele
Company
(2014),
Quality,
available
at:
http://www.miele.com/international/enint/company/quality_13384.
htm (accessed 20 October 2014).

[10] National Assembly of Vietnam (2013), Law No. 32/2013/QH13 Law amended to the law on enterprise income tax, available at:
http://www.itpc.gov.vn/investors/how_to_invest/law/Law_No.32_2
013_QH13/mldocument_view/?set_language=en
(accessed
23
October 2014).
[11] Peattie, K. (1995), Environmental Marketing Management, Pitman,
London.
[12] Saxon Business Systems (2014), Customer First Program,
available at: http://www.saxon.net/PageDisplay.asp?p1=14608
(accessed 24 October 2014).
[13] Schwartz, G. (1971), Marketing: the societal marketing concept,
University of Washington Business Review, Vol. 31, pp. 31-38.
[14] Stunt, V. (2014), Why Google is buying a seemingly crazy
collection of companies, CBC News, 19 February, available at:
http://www.cbc.ca/news/technology/why-google-is-buying-aseemingly-crazy-collection-of-companies-1.2537110 (accessed 23
October 2014).
[15] Szakly, Z. and Berke, S. (2004), The connection between
nutrition, quality and marketing in case of foodstuffs, in Bercs, J.,
Lehota, J., Piskti, I. & Rekettye, G. (Eds.), Marketing theory and
practice: A Hungarian perspective, Budapest University of
Economic Sciences and Public Administration International Studies
Center, Budapest, pp. 380 402.

(The Board of Editors received the paper on 06/11/2015, its review was completed on 09/28/2015)

94

Pham Dang Tram Anh, Luu Quy Khuong

APPLYING GROUP WORK ACTIVITIES TO IMPROVE ENGLISH SPEAKING

SKILLS FOR NURSING STUDENTS AT DANANG UNIVERSITY OF MEDICAL
TECHNOLOGY AND PHARMACY
Pham Dang Tram Anh1, Luu Quy Khuong2
1

Danang University of Medical Technology and Pharmacy; tichanhvn@gmail.com

University of Foreign Language Studies, University of Danang; lqkhuong@cfl.udn.vn

Abstract - Challenges in teaching and learning English for specific

purposes (ESP) in Vietnam ese universities have attracted m uch
attention and have been mentioned in a variety of scientific papers
as well as in national and local newspapers. Teachers and students
of Danang University of Medical Technology and Pharm acy
(DUMTP) are facing the same problem . This paper is to investigate
DUMTP nursing students drawbacks in their learning English
speaking skills. The data for the study was based on
questionnaires forDUMTPs 250 nursing students, and 11 English
teachers, and the pilot application of some group activities to
teaching English speaking skills to 50 nursing students.
Throughout the analysis and conclusion, the authors introduced
som e useful group work activities and how to use them effectively
in order to enhance students English speaking com petence. Also,
the researchers have m ade some suggestions for the government,
the m anagement and the teachers on how to improve English
speaking skills for nursing students of DUMTP in specific and of
Vietnam ese m edical universities in general.
Key words - group work; pair work; English for specific purposes;
speaking skills; nursing

1. Introduction
Every student wishing to get the benefits from modern
education, research, science, trade, etc., knows that it is
impossible without a working knowledge of the English
language and good communication skills. A person
without oral communication skills will suffer in this era of
competition and may find it difficult to achieve a higher
position (Morozova, 2013).
In Vietnam, enhancing quality of teaching and learning
English for Specific Purposes (ESP) in universities and
colleges in national wide has been a great concern. It is
almost a must for graduates to communicate well in English
if they want to succeed in their career on every field.
However, not many of them can do this. Like in various
countries, communicative skills in English, especially in
ESP are the biggest challenge for every teacher and student
in Vietnam in general and of the DUMTP in specific. One
reason why the problem of speaking skills in Vietnam is
crucial is that Vietnamese students have not been provided
appropriate textbooks as well as have not applied active
learning methods right from the start of learning English at
primary schools. As a result, their reaction to English
speaking gradually disappear over the years due to the fact
that they have not been given chance to speak inside and
outside the class, even they do not know how to pronounce
a word correctly. Another reason is that studentsEnglish
knowledge has been assessed based on writing papers
without any oral test so students set up wrong purpose and
do not have any motivation of speaking English when
learning it. These are the main reasons why speaking English
in nursing profession is almost an unfeasible mission

towards nursing students of the DUMTP. To illustrate this

point, let us take an example in one speaking hour, the
teacher raised a question what is the nurses role? and
asked the students to work in a group of six to give the
answer. It took the students about 30 minutes to practise the
pronunciation of only the followings: checking
temperatures, taking pulse rates, taking respiration rates,
checking blood pressures, changing dressings, giving
injections and removing sutures. However, when the time
was out, no students had been confident enough to speak out
until the teacher called them and they gave no completely
correct answer due to wrong pronunciation.As most
problems have a variety of causes, there are likely many
solutions rather than one possible solution to this problem.
From this perspective, this article focuses on
investigating the difficulties in speaking English of the
nursing students of the DUMTP and the effectiveness of
applying group work to improve their speaking skills as
well as tries to suggest a package of complementary
solutions to the problem.
2. Previously works related to the study
Meng (2009) showed that the advantages of group work,
the application of the various group work activities, the
viewpoint of learners proficiency for speaking, feasible
conditions for speaking activities in groups allcontribute a
lot to improve speaking skills for learners in a large class.
A number of researchers investigated this field and came
to the conclusion about students' low level of speaking
ability and their inability to speak confidently and fluently.
One among the many reasons to take into consideration
might be the lack of confidence and anxiety about making
errors as stated by Trent (2009) and in other related studies.
Most college students are not confident in their ability to
learn to speak; teachers must help them overcome their
reluctance in order to change this situation.
Lo (2012) carried out a research on ESP towards
vocational high school students of tourism in Taiwan. The
study drew a conclusion of both theoretical and
pedagogical implications for ESP education for vocational
high school students.
Do and Cai (2010) analyzed the current situation and
some challenges which learning and teaching ESP is
facing. They suggested some feasible solutions in order to
reach the goal set up by institutions. They also added that
effective, long-term vision, consistent and comprehensive
managerial solutions to some extent helped improve
learning and teaching ESP in the new current situation.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

Do (2006) stated that the improvement of teaching and

learning English as a foreign language (EFL) or English for
Specific Purposes (ESP) is drawing a lot of concern. This
article reflected the result of a research project on teaching
and learning EFL and ESP in accordance with the social
needs. It alsopointed out that learners needs analysis was a
crucial element to the planning of strategic policies for more
fruitful training of EFL and ESP at Hue University.
International and national research has shown that
teaching and learning English opens many opportunities
but a number of challenges. However, most research
touched on general issues about English for specific
purposes, while little attention has been paid to
recommending some implications for learners to improve
their English for specific purposes, especially improve
English speaking skill for nursing students.
3. Some theoretical concepts
3.1. Group work and pair work
3.1.1. Group work
Groups may be defined in many ways, indeed providing
an absolute definition of a group, as with much of the
theory around group work, is highly problematic and
contestable.
In group work, the teacher divides the class into small
groups to work together (usually four or five students in each
group). All the groups work at the same time (it is sometimes
called simultaneous group work). (Doff, 1988)
Gibbs and Habeshaw (2001) recommended the
following simple techniques to improve the quality of
small group work:
First, ground rules help the group work better. These
ground rules should state at the start what the teacher wants
these rules to be.
Second, paying attention to the tone can help the
group to work better. The quantity and quality of
interaction between people in a small group is greatly
influenced by its atmosphere and tone. Third, structure
is as important for small group work as it is for lectures. If
the teacher prepares group work activities in a wellorganized way and has a clear statement of aims and
objectives for the small group work, students will feel
happy and respond positively in their learning group.
Fourth, variety sustains interest and involvement. There
are many different ways of working in small groups which
may be used to facilitate the achievement of an equally
wide range of goals. As Doff (1988) mentioned group work
is not a teaching method, it is a way of organizing the class.
Group work can be used for a substantial number of
different activities, and is naturally suitable for some
activities than for the others. Fifth, students can do it
themselves. Small groups can operate effectively without
the teachers presence since many of the roles which the
good facilitator performs in a group can be performed
adequately by students if they are clearly briefed.
Finally, if students are helped to make effective
presentations in small groups it has great benefits for their
self-esteem and for their future work.

95

3.1.2. Pair work

In pair work, the teacher divides the whole class into
pairs. Every student works with his or her partner, and all
the pairs work at the same time (it is sometimes called
simultaneous pair work). Remember that this is not the
same as public or open pair work, with pairs of students
speaking in turn in front of the class. (Doff, 1988)
Penny Ur (2000) considers pair work to be a kind of
collaboration. She has listed types of interaction patterns in
order from the most teacher dominated to the most student
active: teacher talk, choral response, closed ended teacher
questioning, open end teacher questioning, student initiates
and teacher answers, full class interaction, individual work,
collaboration, group work.
Apart from open pairs, where students talk to one
another across the class under your control, there are
according to Byrne (1989) two main kinds of pair work.
These are fixed pairs and flexible pairs. Fixed pairs are
originated when the students work with the same partner
(usually the student on the left or the right) in order to
complete a task of some kind. In flexible pairs, the students
keep changing partners. Teacher must decide whether he
can let the students stand up and move around the
classroom freely. This will make the activity more
interesting for them because they can choose the person
they want to talk to. If the classroom is too small, the
students will be able to interact with those around them
without getting up.
3.2. Advantages and disadvantages of using group work
and pair work
According to Bilash (2009), the differences between
pair work and group work are:
Table 1. Pair work and group work
Type

Benefits

Challenges

Pair

Students have the

chance to work with
and learn from their
peers; struggling
students can learn from
more capable peers; it
is especially useful for
students who prefer
interpersonal learning
settings.

If students are not matched up

well (i.e. low students
together, high students
together, a higher student with
a low student but they dont
work well together, etc.) pair
work wont be useful; the
ability of the students to work
in this way needs to be taken
into consideration.

Group Group work provides

more opportunity for
practice, an increased
variety of activities is
possible, increased
student creativity, the
Zone of Proximal
Development increases.

As with pair work, the groups

must be carefully selected to
ensure students can work
productively; not all students
are able to work to their full
potential in this situation;
assessment of student
progress can be challenging.

4. Advantages and disadvantages of using group work to

teach English speaking skills for nursing students of DUMTP
There is quite a large input level disparity among the
nursing students because their general English backgroundis
different from area to area. Before being enrolled in this
university, some have learned English for 7 years, others have

96

Pham Dang Tram Anh, Luu Quy Khuong

done it only for 3 years. Moreover, the 4 language skills are

variously focused in different schools. It is advantageous to
use group work with these students who will support,
cooperate with each other. Instead of sitting alone trying to
understand something difficult, the less able students can rely
on their more able friends who will be able to solve a problem
so both may feel a sense of achievement while completing a
task. Though it is more demanding and time consuming for
the teacher to prepare appropriate tasks for different students
levels, it gives a lot of satisfaction and motivates the students.
Another advantage of applying group work to improve
speaking skills for the nursing students is that they will try
their best in groups to practice as much speaking as possible
because all speaking tasks are about their nursing profession.
They are eager to practice what they have learned in their
mother tongue. They find it exciting to learn their
professional knowledge in English one more time. That they
set up a clear objective of learning speaking English is the
most beneficial factor because they do know they will
certainly be using it in their career for the whole life.
Nevertheless, there are some drawbacks facing the teacher
and the students. First, most of the students are from rural
areas where English was not taught in a proper way and in less
amount of time compared with those in urban areas. The
development of the language skill for the students was limited
because the teachers focused on teaching reading and writing
skills without paying much attention on speaking and listening
skills. The speaking level of those students is almost zero.
Second, based on the curriculum framework stipulated by the
Ministry of Education and Training, the amount of time
allocated is limited with only 75 periods for ESP module (one
period is 45 minutes). The teacher is not able to give the
students enough oral practice through the whole class work.
Third, the teacher is facing with large class which causes such
problems as physical constraints, lack of individual attentions
or difficulty on assessment. Finally, teaching facilities are
short of audio-visual laboratory. The language classrooms
close to other classes make the nursing studentsspeaking
practice restricted due to the noise.
5. Research methodology
From the beginning of the ESP module which covers
15 nursing professional units of the book Nursing 1 by
Tony Grice delivered within 75 periods of real learning in
class, the authors have applied group work to improve
speaking skills for third-year nursing students of DUMTP.
This study has been carried out based on an instrument
of 250 questionnaires including 22 questions each
distributed to 250 third-year nursing students randomly
picked up from 500 peer students as a whole of DUMTP as
soon as they completed the final test of their ESP module.
The data has been calculated percentage and analyzed to
find out difficulties the students face as a basis to make
some suggestions to teachers on improving speaking skills
for nursing students at DUMTP.
The authors have also collected the opinions from all
English
teachers
of
DUMTP
about
nursing
studentsEnglish learning.
A pilot teaching of some group work activities has also

been conducted on 50 nursing students to take into

consideration the efficiency of these activities.
The study is quantitative and qualitative.
6. Findings
6.1. The nursing students background
More than half of the nursing students come from rural
areas where English has not been focused as much as in
urban area because of a variety of reasons. However, 84%
of participants have learned English for 7 years before they
enroll in DUMTP. That is they follow the same curriculum
and textbook at high schools.
Their purpose of learning ESP is unclear, 85% said they
study it just because it is a compulsory subject. A few
participants said their purpose is to communicate ESP while
the others said they want to read medical texts or materials.
They do not make much progress in ESP partly because
they do not spend enough time on practicing it. 47% said
they spend less than 3 hours a week, 41% said from 3 to 6
hours a week and the rest said they spend more than 6 hours
learning English every week. A substantial number of
participants (81%) said they answer the teachers questions
only when they are called, only 15% volunteers to raise
their hands and give the answers, the rest 4% said I do not
have the answer. when the teacher called them. Another
reason is that there are a large number of students in a class.
While an ideal class of learning a language has around 20
students, the class of the participants has more than 50,
making correcting mistakes for each individual impossible.
ESP is even more difficult for the participants if they
do not have motivation to learn it. Only 37% answered they
like the subject, 44% said they neither like nor dislike, 9%
said they do not like it meanwhile only 10% said they like
it very much.
Regarding the English input level, 83% assessed their
English as unsatisfactory at the start of the ESP course. Only
17% thought their English is satisfactory. No one said they
are very good at general English. This demonstrates that the
nursing students must spend sufficient time and make big
efforts so that they can fulfill ESP modules requirements.
65% said that they did not acquire such basic knowledge as
grammar rules, vocabulary, daily conversations when they
learned general English at high schools. Therefore, they
found it too tough to catch up with ESP course.
6.2. Application of group works to improve the nursing
students speaking skills
It can be clearly seen that most nursing students of
DUMTP have low background of English. Over several
years of teaching English for specific purposes, the authors
understand their drawbacks in learning English. One of the
biggest barrier for them to reach a good English level is the
speaking skill.
Before starting ESP course, 84% of the participants
knew what group work is but only 46% have been taught
group work before. This means group work teaching
method is still quite new to a majority of the students.
However after the course, 64% said they like group work
teaching method; 11% like lecturing and 25% like the

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

method of teacher asks, students answer. Applying group

work in the ESP module has been a great success because
98% stated that it helps them improve their speaking skill.
47% said they are now more confident to speak English
out, 38% said they have practiced and adjusted their
pronunciation during group work time. 15% of the
participants answered that another advantage of group
work is to reduce the pressure due to large class.
Nevertheless, there are some drawbacks of applying group
work reported by the participants. 35% participants said
students speak Vietnamese more than English. Some students
do not speak out because they are so shy and afraid of making
mistakes. 15% said group work is applied just for good
students. A lot of noise made during group work time is also
inconvenient upon implementation of this activity.
48% agreed that the teacher should create a friendly
atmosphere, give compliments or award to those who speak
frequently and fluently when working in small groups. This
encourages not only good but also weak students to get
involved in the activity. 37% said the teacher should go
around to provide them with guidance or correction when
necessary. It takes a large amount of time for students to look
up new medical terms and pronounce them so the teacher
should give them sufficient time to fulfill the tasks.
The participants found the most difficulty in speaking
English is that they are short of vocabulary, not only
medical terms but also general vocabulary. They said it is
very demanding for them to remember professional words
because they are completely new and multi-syllable. The
second most difficulty is that they do not know how to
pronounce a word correctly even if they look at its phonetic
symbol. It is obviously due to their previous learning
method. This is the key factor explaining why so many
Vietnamese students cannot speak or communicate simple
English conversations though their writing and reading are
satisfactory. The most interesting find-out is that many of
them said they have never spoken out a single English
sentence. Most of these students are from rural areas where
English was not set priority so their English level is low as
a whole but their speaking skill is the worst. They also
reported they do not know what to speak because they do
not understand what the others say, that is, their listening
skill is weak, too. Some other factors that make the nursing
students dumb during group work are that they do not
understand the question, the medical context, the grammar
as well as they do not master their professional knowledge.
They also stated they lack confidence and motivation of
learning English for specific purposes.
7. Comments from English teachers of DUMTP about
the nursing students English learning
All the teachers participants said the nursing students
rarely prepare the lessons before class. The reason might
be that the nursing students have more pressure on time
than those of other major due to the fact that they are on
duty in hospitals at night time. 90% of the teachers said the
students never speak out their answer until they are called.
Only 1 teacher said about 3 good students in a class often
contribute their ideas during all sessions. A majority of the

97

students are inactive or passive in their English learning,

said all interviewees. The students English input level,
especially speaking skill, is very low. They added most of
the students show their aspiration for ESP module because
of its practicability but admitted that learning ESP is an
onerous task for them.
When being asked about applying group work activities
to enhance English speaking skills for the nursing students,
11 teachers agreed that its a necessary and effective
method, 3 of them wondered how to proceed it in class.
They suggested that the teachers should go around
constantly to correct their mistakes and pronunciation, give
more pair work and group work to improve the students
speaking skill.
In addition, they have other recommendations:
equiping more facilities such as 2 microphones for 1 class
(3 teachers), strenghthening class observation to learn
experience from each other (2 teachers), organizing
workshops on group work and pair work (5 teachers),
training teachers (11 teachers).
8. Pilot teaching
To find out solutions for improving nursing students
speaking skills, the researchers decided to apply some
group work activities to teaching speaking skills for thirdyear nursing students.
Lesson plans including group work activities have been
designed to improve their English speaking skills from the
beginning of the ESP course. The researchers have
introduced group work as a main activity in teaching
speaking skills during the course. The teachers have
explained to the students how groups would operate and how
they would be graded individually and in groups as well as
established ground rules for participation and contributions.
The pilot teaching has been applied to a class of 50
students within 75 learning periods.
At the start of the module, the International Phonetic
Alphabets and the way to pronounce words correctly by
looking at their phonetic symbols were introduced to the
students. They were given about 20-30 minutes to practice
speaking in small groups every five-period session.
Some group work activities, introduced by Grice
(2001), were given to the nursing students to improve their
speaking skills as a pilot:
1. Stump the partner. The task was: Students A
describe their picture and listen to students B describe their
picture without looking at each others pictures, find ten
differences between them. Speak out ten differences in
front of the whole class.

Figure 1. The nurse is taking care of the patient

98

Pham Dang Tram Anh, Luu Quy Khuong

2. Think-pair-share. The teacher posed a question: Five

minutes ago a poisonous snake bit a child. The parent phones
an emergency helpline and explains the situation to a
helpline nurse. Suppose you are a nurse working on a
telephone helpline. Use the notes and discuss with your
group to tell the caller what to do and answer any questions.
Students took a few minutes to think through an
appropriate response, then turned to a small group and
shared their responses. The teacher took this a step further
by asking students to find someone who arrived at an answer
different from their own and to convince their partner to
change their mind. Finally, student responses were shared
with the entire class during a follow-up discussion.
3. Fishbowl debate. The researchers asked the students
to sit in groups of three and described a hospital admission
procedure. Student A gave the order of a hospital
admission procedure in her/his opinion. Student B gave a
different one. Student C took notes and decided which side
was the most convincing and provided an argument for his
or her choice. The teacher debriefed by calling on a few
groups to summarize their discussions.
4. Work in small groups. A list of medical terms about
elderly care (incontinence, signs of confusion, cognitive
function, confinement, mobility aid, shuffling, gait,
degenerative, frail) were given, the students took turns to
say these terms in a small group of 5. The best student of
the group decided which was more correct and provided an
accurate pronunciation by looking at phonetic symbols and
tried to adjust his/her friends pronunciation. The teacher
debriefed by calling on a few students to say these medical
terms aloud.
5. Pyramid. The teacher asked the students to work
alone to say hospital jobs by looking at pictures
(anaesthetist, pharmacist, cardiologist, physiotherapist,
consultant, porter, lab technician, radiologist, midwife,
receptionist, scrub nurse, surgeon, paediatrician,
paramedic), put them in pairs to discuss what these people
do, then in fours to describe routines and current activities.
Finally, a full group discussion was held to compare the
conclusions drawn.
6. Leave the room. The teacher asked the students to
work in groups of 5 to speak out as far as they knew about
the nursing profession, briefed them and left the room. The
teacher came back after 30 minutes and asked a few groups
to present their discussion to the entire class.
The following form was used to get feedback from the
students after pilot application. They were asked to rate
from A to E (from the best to the worst) for the six different
group work activities and give their comments.
Table 2. Feedback form
FEEDBACK FORM
Group work activities
Stump the partner
Think-pair-share
Fishbowl debate
Work in small groups
Leave the room

Rating

Comments

A majority of students have chosen grade A forWorking

in small groups because they have been able to get support
from their friends and felt more confident working with
their friends and easy to accept their friends correction or
criticism. Surprisingly, the second largest group of students
has given grade B to the activity of Leave the room. They
enjoyed the opportunity of uninterrupted discussion and
said that this activity had involved more of the students
than when the teacher was present. Think-pair-share and
stump the partner have been mostly given grade D and E.
They found it difficult if two students had a big gap of
English abilityto get involved in these two activities
because they did not understand what their partner said so
they could not give the appropriate answers.
9. Conclusion and suggestion
After having taken all the pros and cons of group work
activities into consideration, it is advisable to use these
techniques to improve English speaking skills for nursing
students. As long as the teacher is properly prepared to
introduce and facilitate group involvement and
participation, group work can produce very positive and
lasting results. The researchers would recommend the two
group work activities of Work in small groups and Leave
the room to be applied to nursing students who are old and
independent enough to take responsibility for their study.
As can be seen in the findings, a wide range of students
have never worked in small groups before or lack the skills
to work with others. Below are some suggestions to
improve the situation:
First, in order to design and implement group work
effectively to improve English speaking skills of the
nursing students, the teacher should make sure that each
student understands the assignment. Before group work,
students should know the purpose of the project, the
learning objective as well as know how the assignment is
related to the course content and what the final product is.
Second, the teacher needs to reinforce listening skills as
most nursing students cannot speak anything if they do not
understand the question or they cannot continue the
conversation if they do not know what was mentioned
previously.
Third, as to create motivation for students to speak, the
teacher should have a clear idea of how he/she wants to
evaluate the group work of speaking English for specific
purposes. Some faculty members assign each member of a
group the same grade, which may promote unhappiness if
some members devote more time and effort to the group
and get the same grade. Some instructors assign each group
member an individual grade, which may or may not foster
competition within the group and may undermine the group
solidarity. Because this is speaking class, in-class process
evaluation is recommended.The teacher should assess the
work individually and give compliments to good speakers.
The important thing in the aspect of assessment is that the
students know and understand how they will be evaluated.
Fourth, the quality of teaching ESP at universities is not
effective partly because of English teachers level of
proficiency. They are unqualified both in the English language

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

itself and the pedagogical method. Organizing training

courses in order to foster English teachers who are teaching
ESP or sending them abroad to learn about technical terms and
teaching methodology are strongly suggested.
Finally, in order to improve speaking skills for nursing
students, language classrooms should be placed in quiet
buildings separately from other classrooms. In meantime,
audio-visual labs should be equipped to help students
practice the language among their peers.
REFERENCES

99

English for Specific Purposes, Bulletin of Science, Hue College of

Foreign Languages, vol. 2, p.1 10.
[5]

Graham G.; Trevor H. (2001), Preparing to teach, The Cromwell Press.

[6]

Grice T. (2007), Oxford English for Careers: Nursing 1, Oxford

University Press.

[7]

Haines, S. (1995), Pair Work in Modern English Teacher, vol.1,

Cambridge University Express.

[8]

Trent, J. (2009). Enhancing oral participation across the curriculum:

Some lessons from the EAP classroom, Asian EFL Journal, vol.
11(1), p. 256-270.
Watcyn-Jones, P. (2002), Pair Work 2, Penguin Books.

[9]

[10] Watcyn-Jones, P.; Howard-Williams, D. (2002), Pair Work 1,

Penguin Books.

[1]

Byrne, D. (1989), Techniques of Classroom Interaction, Longman

Group UK Limited.

[11] http://www.cte.cornell.edu/teaching-ideas/engagingstudents/collaborative-learning.html (retrieved on August 14, 2014)

[2]
[3]

Doff, A. (1988), Teach English, Cambridge University Express.

Th Xun Dung; Ci Ngc Duy Anh (2010), Teaching and
learning ESP in the new setting: challenges and solutions, Journal
of Science, Hue University, vol. 26(60), p.31 - 37.
Th Xun Dung (2006), Apply Task-Based Learning to teaching

[12] http://www.ccsenet.org/journal/index.php/elt/article/view/3720
(retrieved on October 2, 2014).

[4]

[13] http://www.teacherstandards.aitsl.edu.au/pilots
(retrieved
on
January 5, 2015).
http://www.educ.ualberta.ca/staff/olenka.bilash/best%20of% 20bilash/pai
rwork.html# (retrieved on January 9, 2015).

(The Board of Editors received the paper on 02/11/2015, its review was completed on 04/01/2015)

100

Nguyen Thi Minh Hanh, Nguyen Thi Quynh Hoa

POSITIONS AND STRATEGIES OF SELF-REPAIR IN ENGLISH AND

VIETNAMESE CONVERSATIONS
Nguyen Thi Minh Hanh1, Nguyen Thi Quynh Hoa2
1

Master Student of the English Language Course 31 (2015-2019)

The University of Danang, University of Foreign Language Studies; hoadng@dng.vnn.vn

Abstract - Self-repair is very necessary for the speakers to keep

their conversations go on and gain efficiency in communication.
There have been a lot of surveys on this matter, but the
researchers have inclined to do the researches on other-repair or
both. Schegloff et al. [5] mentioned self-repair with such features
as positions and strategies of correcting the m istakes causing
communication breakdowns from the speakers. This is one of the
most valuable researches contributing to helping the speakers in
the conversations feel more confident and help them to know how
to control their conversations so that they can gain their
comm unicative purposes. This paper presents and analyzes the
positions and strategies of self - repair in English and Vietnamese
conversations, based on the theory of Schegloff et al.[5] with the
aim of discovering the similarities and differences in the positions
and in using the strategies to repair the mistakes in the two
languages. The results will provide learners with some useful
ways to m ake effective conversations.
Key words - self-repair; positions; strategies; conversations;
theory of Schegloff et al.

1. Introduction
Conversation - the verbal communication between
people is indispensable in daily life. However,
interlocutors cannot always understand all the utterances
from the speaker. Sometimes the speakers cannot clearly
express his thoughts in their utterances due to some
reasons. For example, the speakers are searching for
suitable words to talk or are looking for a correct word in
a specific situation. As a result, they make the hearers
misunderstand or not understand their sayings. All of
those problems are factors that disrupt the conversations
and lead to the lack of communication efficiency. To
overcome
the
factors
causing
communication
breakdowns, the speakers must use some strategies.
However, this matter has not been properly studied in
Viet Nam, so this is the first time the positions and
strategies of self-repair in English and Vietnamese
conversations have been studied. This article presents the
results of the research with the hope of having a small
contribution to teaching and learning speaking skill.
2. Theoretical Background
2.1. Emanuel Abraham Schegloffs theory of repair
Shegloff et al. [5] investigated repairs because of their
roles as a self-righting mechanism for the organization of
language in social interaction. They focused on, for the
first time, self-repair and other-repair with the positions
and strategies. After observing and analyzing the data
from a large corpus of audio tapes, video tapes, naturally
occuring conversations and transcripts of those tapes, they
find that self-correction occurs more frequently than
other-correction. Since then, many studies have applied
the theory of self-repair from Schegloff et al. to carry out

their conversational analyses, for example, Lu Qu

Khng & Nguyn Th Minh Hnh [1]; Rieger, [2];
Rogers [3] ... In this paper, we apply this theory to
discover and compare the positions and strategies of selfrepair in English and Vietnamese conversations.
2.2. Repair in Conversations
According to Schegloff et al. [5, p361], repair is the
treatment of trouble occuring in interactive language use
or a mechanism that operates in conversation to deal
with problems in speaking, hearing, and understanding
the talk in conversation.It includes processes for mutual
comprehension such as word search as well as a
replacement or correction on hearable errors or mistakes.
Schegloff et al. [4] gave out two types of repair: selfinitiated repair and other-initiated repair. Each type
consists of three components, namely trouble - source,
repair initiator and repair completion.
- Trouble source refers to any elements in a
conversation which cause communication breakdown. It
could be found anywhere during conversational
interactions and any element could be repaired by
participants in conversations, even it is grammatically
correct form or pragmatically appropriate expression.
- Repair Initiation: A trouble-source can be identified
by the speaker in his / her own talk (i.e., self-initiated) or
by the recipient in someone elses talk (i.e., otherinitiated), and then the speaker (self-initiator) or the
listener (other-initiator) initiates the repair strategies to
repair it or to ask for repairing.
- Self: Self is the party who produces the trouble
source in his/her talk.
-Repair strategies include the interlocutors verbal
responses to another interlocutors wrong, incomplete, or
silent responses. They also include interlocutors
responses that either repair the trouble directly in the
same turn or initiate repair that another interlocutor will
complete the repair.
-Repair completion: After a trouble source is
identified, it can be repaired by the self-initiator or by the
other-initiators.
2.3. Self-Repair
According to Sack, Schegloff and Jefferson [5], selfrepair consists of two types: self-initiated repair and
other-initiated repair. Self-initiated repair ordinarily
involves the speaker of the trouble source initiating repair
and prosecuting it to conclusion in the same turn; Otherinitiated repair generally involves a recipient of the
problematic talk initiating the repair, but leaving it for the

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

speaker of the trouble-source to deal with the trouble

himself/herself in the following turn. So, self-repair
ordinarily involves the speaker of the trouble-source
initiating repair and prosecuting it to conclusion in the
same turn or the listener will help him/her to repair.
In example (1), the trouble source is caused by the speaker:
(1) It's important for us to accelerate ... accumulate
the working experience.
[http://www.sil.org/liguistics/GlossaryOfLinguisticTerms
/WhatIsSelfinitiatedRepair.htm/retrieved on June 27, 2010]
This example shows that the speaker produces the
trouble source accelerate. Then he/she finds out the
trouble source and initiates repair by using the word
accumulate.
(2) , ... chng ti ... xin mt t x phng, qun xin
ngm nc ung, kht qu.
[Hi c mt binh nh, Nguyn Th Tng]
In (2), the S produces the trouble source x phng. It
is a lexical problem, so in the same turn, S initiates repair
by replacing this word.
Schegloff et al. [3] investigated and systematically
described self - repair in conversations. This type of self repair is conducted in two positions within the same turn;
after the next turn (third turn repair and third position
repair) and it has a particular set of repair strategies such
as error correction, searching for a word, hesitation
pauses, immediate lexical changes, false starts, and
instantaneous repetitions.

in the wrong tense, so he/she immediately corrects it.

(4) Bc ... bc hi ai ? A, c phi bc l ... l ch Hoi
khng ?
[Ma l rng trong vn, Ma Vn Khng]
In this example, S produces the trouble source bc
and initiates repair in the same turn after recognizing H
(the hearer) is a familiar person.
4.1.2. After the Next Turn
There are different types of repair that occur after the
next turn. They can be divided into two types depending on
their position and sequential development in relation to the
trouble source: third turn repair and third position repair.
a. Third Turn Repair
(5) S: Its just about three oclock, so shes probably
free. Ill call her now.
H: What time is it?
S: Three, isnt it?
H: I thought it was earlier.
S: Oh, two, sorry.

[3, p.377]

As an example of third turn repair, Ss first turn Its

just about three oclock makes H know that S is giving
the wrong time. However, S does not know that. In the
second turn, S answers the question What time is it? from
H with Three, isnt it?, H feels unsure about his/her telling
the time in the first turn but S only initiates repair in the
third turn when H says I thought it was earlier.
(6) S:Hiu qu lm! Hiu qu lm!
H: My ni g th?

3. Methodology

S: Tao bo ging ca ca my hiu qu gh lm.

This study adopts descriptive and analytic methods.

Three criteria to select samples have been established:
1-the samples have to be conversations, 2-they have to
contain self - repair, 3- the sources have to be reliable.
The sources from 30 English stories and 30 Vietnamese
stories in the 1970s. Each of them contains from 200 to
500 pages.The year of publication is carefully chosen
with the hope that the utterances in conversations are
modern and similar to what people say in daily
conversations at the present time. Some data are from the
internet. With such set of criteria, 300 English
conversations and 300 Vietnamese conversations have
been collected to build up the corpus. The numbers of
conversations like this can be reliable enough and various
to represent the features of selfrepair.

H: Tao ch hiu g c.

4. Findings and discussion

4.1. Positions of Self -Repair in English and Vietnamese
Conversations
4.1.1. Within the same turn: the speakers realize the
trouble source in their utterance and they initiate repair
by themselves.
(3) "I am - I was - I am his son."

101

S: C g u khng hiu! Ny nh, Hin Ha lu nay

nh mt mnh, khng c th gi qut dn nh ca nn
chut gin, su b, mi mt ln vi trng lc nhc khp
ni, mua thuc dit cn trng v xt cha chc ht.
Nay my rng ln nh th, tt nhin bn chng s chy
vt gi ln c khng cn mt mng, ch hiu qu qu x
qu xa l g!
[Knh vn hoa, Nguyn Nht nh]
S produces the trouble source Hiu qu lm which
confuses H because H does not know what makes hiu
qu. In the second turn, S explains but it is not explicit
enough for H to understand. H continues asking for
explaining and in the third turn, S initiates repair by
explaining everything in details.
b. Third Position Repair
(7) S: He came from somewhere.
H: You mean his mind came from somewhere.
S: I mean the whole package. I think you're going to
end up replying on this boy, betting all our lives on him,
and I think you need to know what's going on in his genes.
It's a poor second to knowing what's happening in his mind,
but that, I suspect, will always be out of reach for you

[Enders Shadow, Card. O.S]

[Gentlemen in their season, Gabriel Fielding]

In the example above, S (he speaker) has the trouble

source in his/her saying but he/she can repair it in the same
turn of speaking. S knows that he/she uses the verb to be

(7) is an instance of third position repair. Repair is

triggered by the recipients misunderstanding displayed in
the next turn following the trouble source. A common type

102

Nguyen Thi Minh Hanh, Nguyen Thi Quynh Hoa

of third position repair occurs when an interlocutor displays

misunderstanding of the speakers prior talk. That is, one
participant produces a turn in a conversation and a listener
produces a sequentially appropriate response in the
following turn, based on her/his own understanding of the
prior turn. However, the initial speaker finds this
understanding problematic and thus may undertake a repair
on the misunderstood prior talk. In the above example, in
the first position of the conversation, S says He came from
somewhere, in the second position of the conversation, H
finds it difficult to understand what or who he is from Ss
utterance, H tells about his/her idea of Ss saying. In the
third position of the conversation, S recognizes that H
misunderstood Ss prior talk, so S self-repairs the prior talk.
(8) S: Cht ri!

[http://www.sil.org/liguistics/GlossaryOf LinguisticTerms
/WhatIsSelf-initiatedRepair.htm/retrieved on June 27, 2010]
The speaker produces a trouble-source happinest and
he/she knows that it is a phonological trouble source, so
he/she initiates a repair.
There are no phonological
Vietnamese conversations.

trouble

source

S: Em bo anh Qu cht ri! Cht nhn ra mnh va

ni cu xi quy, n vi vng cha li em mun ni
nu bit anh Qu l Bnh Minh, bn b trong lp s tha h
chc gho nh, nht l ti t quy! Ti mnh chc chn
s khng nh yn!
[Knh vn hoa, Nguyn Nht nh]
H does not know what S is saying, so H asks for
information. After answering for H, S recognizes that S
makes a trouble source. S continues repairing the
misunderstanding of the prior talk.
Table 1 below is the summary of the positions of self repair and the frequency of its presence from 113 English
conversations and 120 Vietnamese conversations.
Table 1. Positions of Self-Repair in English and Vietnamese
Conversations
English

Vietnamese

in

b. Lexical Trouble source Repair

Many English words have nearly the same sounds, so
the speakers sometimes produce the trouble sources
during their conversations.
(10)
It's
important
for
for
us
accelerate...accumulate the working experience.

H: Em bo ci g cht ri?

Positions

(9) My happinest my happiest memory is related to

my friends

to

[http://www.sil.org/liguistics/GlossaryOfLinguisticTerms
/WhatIsSelf-initiatedRepair.htm/retrieved on June 27, 2010]
This example shows that the speaker produces the
trouble source accelerate. Then he/she finds out the
trouble source and initiates repair.
(11) , ...chng ti...xin mt t x phng, qun xin
ngm nc ung, kht qu.
[Hi c mt binh nh, Nguyn Th Tng]
In (11), S produces the trouble source, x phng. It is
a lexical problem, so in her/his same turn, he/she initiates
repair by correcting the word.
c. Morphological Trouble source Repair (including
repairing word forms, tenses, and so on.)
(12) In my in everyday life, she always help me helps me
[http://www.sil.org/liguistics/GlossaryOf LinguisticTerms
/WhatIsSelfinitiatedRepair.htm/retrieved on June 27, 2010]

Occurrence

Occurrence

Within the same

turn

53

46.9

30

25

Third turn
repair

15

13.3

22

18.3

Third
position
repair

45

39.8

68

56.7

S knows that good without -s will confuse the listener,

so he/she initiates repair by saying goods.

113/300

100

120/300

100

There are no morphological trouble

Vietnamese conversations.

After
the
next
turn

Total

After saying, the speaker finds out that he/she

pronounces the verb wrongly, so he/she initiates repair.
(13) I consider the price and the quality of the good goods.

The statistics indicates that the position of self-repair

within the same turn in English conversations is higher
than that in Vietnamese ones.The reason is that there are
more factors causing trouble in English conversations
such as phonological, morphological and lexical
problems. However, with the position of After the next
turn, more Vietnamese people tend to repair in the third
position than English people.
4.2. Strategies of Self-Repair in English and Vietnamese
Conversations
4.2.1. Trouble source Correction
a. Phonological Trouble source Repair (speakers
may find they do not pronounce some sounds
correctly which may cause misunderstanding, so they
go back and correct them.)

[http://www.sil.org/liguistics/GlossaryOfLinguisticTerms
/WhatIsSelfinitiatedRepair.htm/retrieved on June 27, 2010]

sources in

4.2.2. Searching for a word

(14) I need a new bolt for my oil fit - um PAN
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
In (14), S cannot think out a word he/she wants for the
first time of speaking, but in the same turn after the
hesitation marker um, the speaker can find the right word
Pan and initiates the repair of his/her utterance.
(15) Ni thit vi ch, mi ch xung chi cho bit
nh bit ca. Sau na, ti c chuyn ny tnh k ch nghe.
Nu khng thch, ch b ngoi tai, ch vi ti, n ... - anh
dng li tm ch- ... n thing ling lm!
[Nhng ngi th nn, Hong nh Trung]
In this example, S hesitates because he is choosing the

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

103

word to express his feelings. After his moment of

thinking, he finds the word and initiates repair.

speaker initiates a repair just after saying it out.

4.2.3. Hesitation Pauses

(22) I dont know. The school - school uh, bookstore

doesnt carry anything anymore [1, p.91].

Hesitation pauses occur when the speakers cannot

express or find it difficult to express their thoughts at the
moment of speaking; or when they want to mention
another thing.
(16) Its like some sort of ... I dunno ... like some sort
of rally said Mr. Roberts. They all seem to know each
other. Like a big party.
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
(17) You...you... Wormtails voice suddenly
sounded hoarse, as though his mouth had gone very dry.
You...are going...to kill me too?
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
In (16), S hesitates because he does not know how to
express what he sees. After lengthening his silence by
..., he initiates the repair in his same turn with like some
sort of rally, like a big party. In contrast, in (17),
S predicts that something dangerous will happen to him,
so his hesitation pause ... shows his frightening and his
difficulty to express his ideas. The speaker also initiates
the repair in the same turn.
(18) Tri i!...con...con hy tha th cho m...m Thu
ca con y...V tng lai ca i con...m khng mun
cho mi ngi bit con c ngi m b hi nn ra i t
lc con ln hai...Tri i...Thu Trang...cha ca con u
ri?
[Ngi tnh ca cha, T Nguyn Tnh]
In this example, the speaker hesitates because she is
too moved to say anything fluently. Her hesitation helps
her keep calm and gain some more time to continue her
utterance.
(19) S:...oa...oa!
H: Hoa?
S:...oa...oa!
[Knh vn hoa, Nguyn Nht nh]
In this example, S has trouble with the pronunciation,
so H identities the word ...oa from S as the trouble
source, H repeats Ss prior talk with the rising intonation
to confirm Ss utterance. In the third turn, S self
completes her utterance and confirms that the name H
mentions is right.
4.2.4. False start repairing
(20) Talking about breathing in front of me! she said
shrilly, and her voice echoed loudly around the bathroom.
When I cant ... when I havent...not for ages...
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
In this example, S also starts his/her saying wrongly
with I cant and initiates a repair at once in the same turn.
(21) Ci xe ... Ti cung cung ngi thc xung, ln
m ci van. - Chu...ci xe...lm sao y.
[Chng trai bn xe i, H Anh Thi]
S knows he/she is wrong when beginning with a
different subject which he/she wants to mention, so the

4.2.5. Immediate Lexical Changes

S finds out that the word school is not the word

he/she wants to mention, so bookstore is immediately
used to replace school in the same turn of speaking.
There are no immediate lexical changes in Vietnamese
conversations.
4.2.6. Repetitions
Repetition is one of the strategies including a lot of
subcategories.
a. Repetition of Personal Pronouns
The repetition of personal pronouns occurs when the
speakers want to postpone a possible transition-relevance
place and to gain additional time [1].
(23) I - I thought she might be useful, My Lord -
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
In this example, the pronoun I is repeated so that the
speaker can have more time to think about what he/she is
going to say.
(24) C...c...c gip anh mt tay, gi m tr ra
ngoi ny chi (nhng ti ny khng tr ni th n lm
lon). T qu! T qu mt thi!
[Chuyn thng gp, Nguyn Th o Tnh]
The personal pronoun c is repeated when the
speaker feels puzzled to ask the hearer to do something.
However, after the repetition, the speaker self completes
his utterance.
b. Repetition of Conjunctions
The repetition of conjunctions such as but and and
serves to delay the production of the next lexical item or
to postpone a possible transition-relevance place.
Conjunctions are most often repeated without any other
elements, but occasionally pauses, lengthening of sounds,
quasi-lexical and lexical fillers are uttered between the
first and second use of the conjunction [1]
(25) But they get paid? she said. They get holidays,
dont they? And - and sick leave, and pensions, and
everything?
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
(26) But - but you said its illegal, Professor, said
Herminone uncertainly as Moody cleared away the desks
with a sweep of his wand, leaving a large clear space in
the middle of the room. You said - to use it against
another human was-
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
In (25) and in (26), the speakers repeat the conjunction
and, but without adding any other elements. In (25),
the speaker wants to have more time to think out some
things related to the salary. In (26), the speaker wants to
postpone producing the lexical item although the speaker
knows what he/she is going to talk about.
(27) S: nh c v khng tin em!

104

Nguyen Thi Minh Hanh, Nguyen Thi Quynh Hoa

H: Ghen h?
S: Ni ghen cng cha ng. Nhng ... nhng
nh c v khng thch ch Phng.
[Qung i m p, D Ngn]
The conjunction nhng is repeated in the above
example because S wants to gain more time to think out
what he/she is going to say. Then, he/she self completes
his/her utterance in the same turn.
c. Repetition of Indefinite articles
(28) "Ss! Shut up! Listen!" They were mouse-still.
"I thought I heard a - a sort of knocking," said the bull.
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
d. Repetition of Prepositions
(29) But youre behind - very behind - on dealing
with curses, said Moody. So Im here to bring you up to
scratch on what wizards can do to each other. Ive got one
year to teach you how to deal with Dark -
[Harry Porter and the Goblet of Fire, Rowling, J.K.]
There are no repetitions of indefinite articles and
prepositions in Vietnamese conversations.
Table 2 below is the summary of strategies of self repair and the frequency of its presenting from 113
English conversations and 120 Vietnamese conversations.
Table 2. Strategies of Self-Repair in English and Vietnamese
Conversations
English

Strategies

Vietnamese

Occurrence

Occurrence

1. Trouble
source correction

13

11.5

10

8.3

2. Searching for
a word

12

10.6

15

12.5

3. Hesitation
pauses

24

21.2

48

40

4.False start
repairing

7.1

5. Immediate
lexical changes

5.3

6. Repetitions

50

44.3

47

39.2

113/300

100

120/300

100

Total

gain more time to think and express their thoughts more

clearly. The strategies of Trouble source correction and
Searching for a word are also used a lot in two languages.
There are no morphological trouble source repair, false
start repairing, immediate lexical changes, repetitions of
indefinite articles and repetitions of prepositions in
Vietnamese, due to linguistic features.
5. Conclusion
The results of the survey show that English and
Vietnamese conversations contain nearly all of the
features of positions and strategies of self-repair
according to the theory of Schegloff et al. There are two
positions namely within the same turn and after the next
turn (including third turn repair and third position repair)
and six strategies such as trouble sourcecorrection,
searching for a word, hesitation pauses, false starts
repairing, immediate lexical changes and repetitions of
self - repair. However, there are no false start
repairing,immediate lexical changes, no phonological
and morphological trouble sources in Vietnamese
conversations. The similarity of positions of self-repair in
the two languages indicates there is a unity of mechanism
for the organization of language in social interaction. The
differences in strategies show that each language always
has some its own features of formation and grammar.
The results of the research can be applied to teaching
English and Vietnamese conversations. The features of
the positions and strategies of self - repair help learners
have a good look at how to solve the problems causing
communication breakdowns from the speakers.
REFERENCES

The table shows that two strategies such as Hesitation

pauses and Repetitions are used more frequently than
other strategies. Both of these strategies help speakers

[1]

[2]

[3]
[4]

[5]

1. Lu Qu Khng & Nguyn Th Minh Hnh (2012), Nghin

cu v tr v chin lc xut sa cha cc yu t gy tr ngi
hi thoi t ngi nghe trong hi thoi ting Anh v ting Vit,
Ng hc ton quc 2011, Hi Ngn ng hc Vit Nam- Trng
i hc Ngoi ng, HN, tr.377-383.
2. Rieger, C.L. (2003), Repetitions as self-repair strategies in
English and German conversations, Journal of Pragmatics 35,
p.47-69.
3. Rogers S. (1978), Self-initiated corrections in the speech of
infant school children, Child. Language, 5: 365-371.
4. Sack, H., Schegloff, E. & Jefferson, G. (1974), A simplest
systematics for the organization of turn-taking in conversation,
Language 50/4: p.693-735.3.
5.Schegloff, E.A., Jefferson, G., & Sacks, H. (1977), The
preference for self-correction in the organization of repair in
conversation, Language, 53, p.361-382.

(The Board of Editors received the paper on 12/14/2015, its review was completed on 12/28/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

105

APPLYING GRAMMATICAL METAPHOR TO ACADEMIC WRITING

Phan Van Hoa1, Nguyen Van Vui2
1

The University of Danang, School of International Education; hoauni@gmail.com

2
Vo Giu High School; ngvui.edu@gmail.com

Abstract - The term Gramm atical Metaphor (GM) was first

launched by Halliday (1985) in An Introduction to Functional
Grammar. GM has been studied and commonly used in various
types of discourse. The more GM is employed in writing, the more
formal and academ ic the writing is. Language learners often face
long and complex nominal phrases. To comprehend them is not
easy, and to use them in writing is more difficult. This paper is to
present some basic constructions of GM based on the theory of
Systemic Functional Grammar. Thus, firstly, the paper gives a brief
account of GM so that we can have an overview of the term GM.
Secondly, types of GM are provided to establish the theoretical
frame for the paper. Lastly, some common constructions of GM are
suggested to help writers or language learners master them and
use them in writing most efficiently.
Key words - gramm atical m etaphor;
Halliday; clause; m etaphorical.

Functional

Grammar;

1. Rationale
Written language displays a much higher frequency of
lexical items or content words to the total amount of words
used than grammatical items. The two following sentences
have the same meaning but the frequency of lexical items
and grammatical ones is employed differently.
1.a: Investment in a rail facility implies a long-term
commitment. And
1.b: If you invest in a rail facility, this implies that you
are going to be committed for a long term.
[5, p.61]
The sentence 1.a displays seven lexical items, covering
70% the total words and three grammatical ones,
accounting for 30%, while the sentence 1.b displays seven
lexical items (35%) and thirteen grammatical one (65%).
The two sentences above show that language in a
written text is more condensed than that in a spoken one.
In other words, the difference between written and spoken
language is density. What makes a written text so dense is
GM in which nominalization is the most powerful device
in creating grammatical metaphor [6]. The more
grammatical metaphor a discourse has, the more formal
and academic it is. Halliday [5, p. 94-95] claims that
written language is associated with the use of GM and is
more characteristic of written language than spoken
language. He points out that GM includes two types:
Metaphor of mood (including Modality), and Metaphor of
Transitivity. Semantically, these are respectively
Interpersonal Metaphor and Ideational Metaphor.
However, in this article we just focus on metaphor of
transitivity where nominalization is the key for creating
grammatical metaphor.
GM occurs popularly in documents for language
learners such as in IELT, CAE documents, etc. in rewriting
the sentence without changing the original one, in many
other academic documents. How do English learners apply
GM to their academic writing? This paper is to suggest

some common constructions to write an academic paper

most effectively.
2. Grammatical Metaphor
The notion of grammatical metaphor is described by
Halliday in Introduction to Functional Grammar in
Chapter 10 - Beyond the Clause: Metaphorical Modes Of
Expression. He places his introduction of the term GM in a
more general framework outlining traditionally recognized
types of rhetorical transference or figure of speech;
metaphor, metonymy and synecdoche.
Halliday [6, p.342] defines grammatical metaphor as
variation in the expression of a given meaning. The term
metaphor in grammatical metaphor is different from
metaphor in traditional view that is used for something
resembling that which it usually refers to. Metaphor in the
traditional view is usually described as variation in the use
of words: a word is said to be used with a transferred
meaning. However, we are looking at it from the other end,
asking not how is this word used? but how is this
meaning expressed? [6, p.341].
A meaning may be realized by a selection of words that
is different from that which is in some sense typical or
unmarked.
Ravelli [1, p.40], when exploring the role of GM in
written discourse, claims [] one choice in the semantics
may have two (or more) lexicogrammatical realizations
and GM is an alternative lexicogrammatical realization of
a semantic choice.
Moreover, Thomson [15, p.165] defines GM as the
expression of a meaning through a lexico-grammatical
form which originally evolved to express a different kind
of meaning. The expression of the meaning is
metaphorical in relation to a different way of expressing
the same meaning which would be more congruent.
Thomson further points out that there is no basic
difference between lexical and grammatical metaphor.
Lexical grammar then can be considered to be a subcategory of GM. He assumes that the metaphorical way
of expressing the meaning is intuitively closer to the
events in the external world: instead of representing the
process of comparing as a thing from which something
else can emerge of its own volition, it represents it as an
event involving human participants which results in those
participants understanding a phenomenon. The essential
distinction between congruent and incongruent is that the
former is closer to the state of affairs in the external
world than the latter.
Martin (2004) places GM into analyzing discourse of
text and considers GM as a tool for packing meaning
[11, p. 401]. GM is the real gatekeeper in which meanings

106

Phan Van Hoa, Nguyen Van Vui

are packed for grammatical realization. Martins most

noteworthy contribution to GM is adding textual metaphor
as the third type of GM to Hallidays original types of
ideational and interpersonal metaphor.

In ideational GM, the shift does not only fall into any
isolated elements but it always occurs in clusters. For
example, when the type 2.i occurs (from verb to noun: from
Event to Thing), it entails other changes such as from noun
to possessive deictic, from adverb to adjective, etc. In this
article we summarize these 13 types of GM in grammatical
shift as follows:
from
1

to

Adjective Noun

Verb

Ty Grammatical shift
pe from
to
8

9
Noun
(i.Event
Thing,
ii.Auxiliary
Thing,
iii.Catenative
Thing)

Preposition Verb
(al phrase) (i.Minor
process
Process;
ii.Location,
Extent &c
Process)
Conjunction Verb

Conjunction noun

([x] the fact/

Phenomenon

Verb

Adjective
12 + Verb
(i.Event
Epithet/Classifier;
ii.Auxiliary
Epithet/Classifier,
iii.Catenative
Epithet/Classifier)

13 i. Noun
Preposition Adjective
(alphrase) (i.Minor
ii. Adverb
Epithet/Classifier,
ii..Auxiliary
Prepositional
Epithet/Classifier,
phrase
iii.Catenative
Adverb
Epithet/Classifier)
Preposition
al phrase

(O)

(Adv)

Part.

Pro.

Part.

Cir.

Metaphorical

Part.

Function
Type:
I&III

V-lex

(Adj)

Pro.

(Prep.
+ O)

NML

Epithet Thing

Cir.

Range
12i

13ii

2i

Notes: S: Subject; V: verb; O: Object; Adv: Adverb;

Adj: Adjective; Part.: Participant; Pro.: Process;
Cir.: Circumstance; NML: Nominalization
(1)We are no longer interested in unidirectional
importation of goods and machinery from Europe. [13, p. 336]
Congruent

(Adv)

Adj

Part.

Pro.:
relational

Cir.

Epithet

(Prep.) Possessive

(Adj.)

NML

Possessive

Epithet

Thing

13i(b)

13ii

Function
Type:
I&II

=> We no longer take any interest in unidirectional

importation of goods and machinery from Europe.
(2)The regular annual meeting of the middle regatta
Association will be arranged at the New York Athletic
Club Ma.

(i.+verb;
ii. +(causative
& c)verb
iii.+ (phasal &
c) verb2)

=>The arrangement of the regular annual meeting of the

middle regatta Association will be made at the New York
Athletic Club Ma.

(various)
(a) Qualifier,
(b) Possessive
Deictic,
(c) Classifier,

=> Roentgen was awarded the first Nobel Prize in

physics in 1901, but The Times made no further mention
of it until the publication of his obituary in 1923.

adjective,
adjective,
(various)
(various)

Function

Metaphorical
Preposition(al
phrase)
(i.Conjunctive
Minor process,
ii.Conjunctive
Location,
Extent & c)

11 + Noun

Congruent

Function

10 Conjunctio
PrepositionNoun
(al)
(i.Minor process
n
Thing
ii.Location,
Extent & c
Classsifier)

Positive GM means the original form of process/quality

isnominalized (NML) through using derivation of the
process/quality. We here give five basic structures.
1) Pattern 1:

3. Types of Grammatical Metaphor

Type Grammatical shift

3.1. Positive GM

Conjunction Adjective

However, we also recognize that the major shift takes

place at two types: Positive GM and Negative GM.
Halliday [7, p.250] states that the pervasive of the types of
GM listed in the table above are type 1 and 2: quality =>
thing and process => thing . Some of the shifts can
occur independently while others occur only under the
driving force of the shift of type 1 and 2.

(3) Roentgen was awarded the first Nobel Prize in

physics in 1901, but The Times further didnt mention it
until his obituary was publicized in 1923.

In this structure, the lexical verb is a verb has no

meaning but serves as a grammatical word in the sentence.
Each noun has its lexical verb, so learners must learn them
by heart.
2) Pattern 2:
Congruent
Function

(O)

Part.

Pro.

Part.

Metaphor
There
ical
Function
Type:
I&III

(Adv)
Cir.

be

(Adj)

NML

Prep.
+O

Pro.

Epithet

Thing

Cir.

2i

13i(a)

12i

13ii

(Prep.
+ O)
Cir.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

(4) The cost of living has increased considerably in

the last few years.

107

5) Pattern5:
Congruent

S1

V1

S2

V2

O2

=>There has been a considerable increase in the cost

of living in the last few years.

Function

(5) The value of Sterling has fallen considerably in the

past week.

Metaphorical

=>There has been a considerable fall in the value of

Sterling.

Function

(6) Next month the ticket will undoubtedly be not as

cheap as they are now.

Type:
I&II

=>There is no doubt that tickets will be more

expensive next month.

This pattern often occurs in case of the first clause with

a reporting verb or with projection sequences. However, it
also occurs in any types of verbs, transforming two or more
clauses into one clause.

In the structure There be + NML, not only are

verbs nominalized but both adjectives and adverbs are
nominalized as well. Subject now is turned into Object
after a preposition. Thus, learners must choose an
appropriate preposition after the Object.
3) Pattern 3:
(7) I was amazed when Eva got into university, given
the fact that she hadnt worked hard at school. [1, p.277]
=>To my amusement Eva got into university despite
the fact that she hadnt worked hard at school.
(8) On a recent visit to Kew Gardens (Sept) I was
surprised to see four parrots flying free in the gardens.
=>On a recent visit to Kew Gardens (Sept), to my
surprise, four parrots were flying free in the gardens.

Part.
S1
Part.

Pro.
V(1)
Pro.

Part.

Pro.

Posses
sive

NML

Posses
sive

Thing

13i

1/2i

Part.
(Prep.
+ O2)

(13) Judge Richard Leon declared that the mass

collection of metadata probably violates the fourth
amendment, which prohibits unreasonable searches and
seizures, and was "almost Orwellian" in its scope.
=>) Judge Richard Leon declared the violation of the
mass collection of metadata probably against the fourth
amendment, which prohibits unreasonable searches and
seizures, and was "almost Orwellian" in its scope.
(14) Its possible that the burglars got into the building
by forcing open a fire exit.
=> The burglars may have gained access to the building by
forcing open a fire exit.
(15) They shredded the documents before
departed for the airport.

they

=>Pauls frequent absence cost him his chance of

passing the examination.

=>They shredded the documents before

departure for the airport. (10i + 13i(b) + 2i + 13ii

their

The subject in the structure 3 now is transformed into a

possessive adjective to modify the NML from a verb or an
adjective.

It can be seen that the structure 5 allows learners to

shorten two or more clauses into one sentence. The shift is
rather more complex. It requires learners to understand the
meaning of the clauses clearly to make the sentence have
the same meaning as the original one.

(9) By being absent so often, Paul failed the examination.

4) Pattern 4:
Congruent

Conjunc
tive

Function
Metaphorical
Function
Type:
I&III

Prep.
phase
Relator
10i

(O)

(Adv)

Part.

Pro.

Part.

Cir.

Posses
Adj.
sive

(Prep.
NML
+ O)

Posses
Epithet Thing
sive
13i(b)

Cir.

2i

10) Naturally, his farther agrees, though he feels

rather doubtful about the limits of the form.
=>Naturally, his farther agrees, despite his doubts
about the limits of the form.
(11) As soon as we receive your check, we will send
the goods to you.
=>On receipt of your check, we will send the goods to you.
(12) Although Jimmy was the stronger of the two, his
attacker soon overpowered him. [24, p.161]
=>Despite his superior strength, Jimmys attacker
soon overpowered him.

3.2. Negative GM
Negative GM means the metaphoric shift from different
elements (an adverb, a conjunction, modal adverbs, a
preposition, etc.). In negative GM form, another noun as a
Thing that is somehow synonymous with the different
elements is used to replace them to express the same
meaning as that of the original form.
1) Adverb NML

(13ii)

(16) They also needto keep the money rolling in.

Understandably, perhaps, they sometimes overstate their
arguments.
=>They also need to keep the money rolling in.
Understandably, perhaps, there are some occasions they
overstate their arguments
The adverb sometimes is synonymous with
occasionally that is derived from the noun occasion. Here
the use of the determiner some and the noun occasionto is
made to express the meaning of the adverb sometimes.
2) A Verb
(17)

NML(borrowed) (2i)

You could be arrested for not giving a breath

108

Phan Van Hoa, Nguyen Van Vui

sample to the police.

=>Refusal to give a breath sample to the police could
led to your arrest
The verb phrase not giving does not have a noun form
from which the noun Refusal is borrowed to be
nominalized to express the meaning of not giving. We
consider some more examples below.
(18)

Our plane takes off at 10:00 a.m.

=>Our planes departure time is 10:00 a.m.

(19)The committee said they liked the first proposal best.
=>The committee expressed a preference for the first
proposal best.
3) A Phrase

NML (x)

(20) Something must be done quickly to solve the

problem of homelessness.
=>Urgent action must be taken to solve the problem
of homelessness.
(21) Australia's free trade pact with China won't feel
like freedom for workers.
=> Australia's free trade pact with China won't be in
the right mood for freedom for workers.
In this structure, NML is not derived from a verb, an
adjective or an adverb but from the whole meaning of the
phrase or a clause. A nominal phrase that has the same
meaning as the phrase or the clause is borrowed to fulfill
the meaning.
4) A Preposition

NML (10i)

=>The reasons were varied; some of the women

alluded to demanding work schedules that made having a
child undesirable, others to the fact that they simply have
no desire for having a child.
4. Conclusion
Briefly, owing to the extensive change of the structures
and the big operation of nominalization as a grammatical
metaphor device, we can generate various constructions with
nominalization to make different expressions that have the
same meaning as the meaning of the original form. English
learners enjoy benefit of the power of nominalization to
make an effective use of GM in their writing.
Besides positive GM, negative GM takes an important
role in transforming a word, a phrase and a clause into
another structure without changing the basic meaning of
the original form. Thanks to negative GM, English learners
can use another noun to replace other elements to do their
writing more academically and metaphorically.
Apparently, the list of GM types listed in the above
table, to some extent, does not exactly fit the variation of
expression of a given meaning in reality. And the patterns
that we have suggested above do not cover all the
variations and all the metaphorical shifts of language.
Language learners and users can consider those
constructions to be a way of applying GM in their writing
conveniently and effectively.
REFERENCES
[1]

Anne-Marie, S.-V., Taverniers, M. & Ravelli, L. (2003),

Grammatical Metaphor, John Bejamins Publishing Company.

[2]

Butt, D. (2002), Using Functional Grammar: An Explorers Guide,

Macquarie University.

[3]

Gerard, J. S. (2007), Finding Metaphor in Grammar and Usage,

John Bejamins Publishing Company.

(23) Apart from Philip, everyone else in the meeting

was a party member.

[4]

Haliday, M.A.K (1985), An Introduction to Functional Grammar.

London: Edward Arnold.

=>With the exception of Philip, everyone else in the

meeting was a party member.

[5]
[6]

Besides the structures given above, we also find out that

GM occurs in various structures in sentence transformation.

[7]

Halliday, M.AK. (1989), Spoken and Written Language, London: Oxford.

Haliday, M.A.K (1994), An Introduction to Functional Grammar,
Second edition, London: Edward Arnold.
Haliday, M.A.K & Matthiessen, C. M.I.M (1999), Construing
Experience through Meaning: A Language-Based Approach to
Cognition. London: Cassel.

(22) By the 1920s architects throughout Europe were

against the conditions created by industrialization.
=>By the 1920s architects throughout Europe raised
objection over theconditions created byindustrialization

(24) The more proficient he is, the more he likes the games.
=>His fondness for the game increases with his
proficiency.
(25)English spellings dont match the sounds they are
supposed to represent. Its time to change.
=> English spellings dont match the sounds their duty
is to represent. Its time to change.
(26) John was shocked to hear that he had failed his
driving test.
=>It came as shock to John to hear that he had failed
hisdriving test.
(27) The reasonswere varied; some of the women
alluded to demanding work schedules that made having a
child undesirable, others to the fact that they simply dont
feel like having a child.

[8]
[9]

Halliday, M.A.K. & J.R. Martin (1993), Writing Science: Literacy and
Discursive Power. London, Washington D.C.: The Falmer Press.
Koptjevskaja-Tamm, M. (1993), Nominalizations, Intyre London Ltd.

[10] Martin, J.R. (1991), Nominalization in Science and Humanities:

Distilling Knowledge and Scaffolding text. Eija Ventola (ed),
Trends in linguistics: Functional and System Linguistics, New York:
Mouton de Gruyter, p. 307-334.
[11] Martin, J.R. (2004), English Text, John Bejamins Publishing Company.
[12] Ravelli, L.J. (1988), Grammatical metaphor: An initial analysis. In:
E. Steiner & R. Veltman (eds.) Pragmatics, Discourse and Text:
Some Systemically-Inspired Approaches. London: Pinter Publishers.
[13] Ravelli, L. J. and Ellis, Robert A. (2005), Analysing Academic
Writing, Continuum.
[14] Taverniers, M. (2002), Systemic-Functional Linguistics and the
Notion of Grammatical Metaphor: A Theoretical Study and a Proposal
for a Semiotic-Functional Integrative Model, University of Ghent.
[15] Thomson, G. (1996), Introducing Functional Grammar, London:
Hodder Education.

(The Board of Editors received the paper on 08/23/2015, its review was completed on 09/28/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

109

LINGUISTIC FEATURES OF ASSERTIVES

IN PRESIDENTS WELCOME MESSAGES ON OFFICIAL WEBSITES
OF COLLEGES AND UNIVERSITIES IN THE USA
Nguyen Thi Quynh Hoa1, Doan Thi Hong Nguyen2
1

The University of Danang, University of Foreign Language Studies; hoadng@dng.vnn.vn

2
Master Student of the English Language Course 27 (2013-2015).

Abstract - Pragmatically, an assertive is one type of illocutionary

speech acts. The typical function of the assertive is to assert or claim
some statement or fact. Noticeably, assertives occur so often in
Presidents W elcome Messages (PW Ms) posted on the official
websites of the colleges and universities in the USA. These assertives
help to highlight the purpose of PW Ms -attracting more and more
learners to the higher education institutions. This paper attempts to
identify and analyze the assertives used in PW Ms in terms of lexical,
syntactic and semantic aspects. The findings help to illuminate the
linguistic features of assertives commonly found in PW Ms.
Key words - assertive; Presidents W elcom e Messages (PW Ms);
illocutionary speech act; lexis; syntax; semantics.

1. Rationale
It is obvious that higher education is expanding more
rapidly than ever before. One important task of colleges
and universities is to enroll students. For learners, the more
education institutions develop, the more choices and
opportunities they have in choosing an appropriate
institution. Moreover, it is vital for them to make the right
decisions on which programs to attend and which tertiary
institutions to enroll in.
On one hand, training service providers have to
establish an impressive image for their institutions. On the
other hand, potential learners have to make wise choices on
applying for the appropriate higher education. PWMs on
the official websites of American colleges and universities
optimally fulfill the need of building tertiary institutions
images and of learners decision-making. It is noticeable
that PWMs contain many assertives stating the institutions
strengths and reputations. Therefore, this research is
conducted with a view to clarifying the linguistic features
of assertives used in PWMs.
It is hoped that the results of the study can be beneficial,
to some extent, to the teaching and learning of English. This
study may also be of some help to those who are supposed
to write welcome messages for the purpose of attracting as
many learners to their training institutions as possible.
2. Theoretical Background
2.1. Searles Taxonomy of Illocutionary Speech Acts
In this paper, to identify assertives in PWMs, the
researcher employs Searle [1975]s Taxonomy of
Illocutionary Speech Acts, which includes five categories:
Representatives or Assertives, Directives, Commissives,
Expressives, and Declaratives. The first reason is that this
classification system is more popular than those by other
linguists such as Austin [1962], Bach and Harnish [1979],
Yule [1996]. The second reason is that in order to classify
speech acts, Searle presents twelve linguistically significant

dimensions of variation whereby illocutionary acts differ

from one another and he builds most of his taxonomy around
three primary dimensions. These are the illocutionary points,
the direction of fit, and the sincerity condition.
Table 1. Five General Functions of Speech Acts (Searle 1975)

Speech act
type

Direction of fit

S=speaker
X=situation

Representatives make words fit the world S believes X

/ Assertives
Expressives

make words fit the world S feels X

Directives

make world fit the words S wants X

Commissives

make world fit the words S intends X

Declaratives

words change the world

S causes X

In the above table, Direction of fit is the relationship

between the propositional contents and the referred world;
words are utterances that convey the proposition contents of the
illocution; the world is the existing state of affairs in the world.
In addition, Searle maintains that there are four
directions of fit in language:
a. Word-to-world, where the utterance fits an
independently existing state of affairs in the world. A
statement of fact exhibits this direction of fit.
b. World-to-word, where the world is altered to fit the
proposition content of the illocution.
c. The double direction of fit is when the world is altered
to fit the propositional content of the utterance by being
represented as so altered.
d. The null direction of fit is where there is no question
of achieving success of fit between word and the world.
Searle also states that a speech act is an assertive if its
purpose is to represent some truth, if it has a word-to-world
direction of fit (the word matches the world), and if the
speaker believes what he says.
In addition, the study also makes use of the definition
of representative by Yule [1996, p. 53] because its content
is similar to that of an assertive speech act. He listed 4 types
of representatives or assertives namely statements of fact,
assertions, conclusions, and descriptions, whereby the
speaker represents the world as he or she believes it is.
2.2. Definitions of Terms
Below are the working definitions of the key terms
employed in this research:
- Welcome means to greet somebody in a friendly
way when they arrive somewhere or to be pleased that

110

Nguyen Thi Quynh Hoa, Doan Thi Hong Nguyen

somebody has come or has joined an organization or an

activity. In PWMs, the presidents welcome the expected
learners to their institutions through the PWMs.
- Presidents Welcome Message is a letter to the
audience by the president of a college or university that
describes the institutions strengths, identifies what is
unique or special about the institution, and invites readers
to seek more information about the institution.
- Assertive in PWMs is a speech act that the president
produces to assert or claim some statements about the
institutions status and describe the institutions strengths to
the audience for the purpose of persuading and attracting
more and more learners to the presidents tertiary institution.
3. Research Methods
The descriptive method is used to describe, give details,
explanations and clarify the characteristics of assertives in
PWMs in terms of lexis, syntax and semantics.
Then the inductive method is employed to draw out
conclusions on the common linguistic features of assertives
in the PWMs. The analysis of assertives in this research is
based on Searles theory.
In order to collect the data for the studys analysis, the
researcher has built up some criteria for choosing samples
as follows:
As regards Presidents Welcome Message, a PWM
must be from the President of a higher education institution
and available on the official website of an American
college or university; it must be in the written form and
from 200 to 400 words in length. This is because through
the process of collecting data from many official websites
of American colleges and universities in the USA, the
researcher found that most PWMs are in letter forms with
a variety of lengths. Then the researcher divided these
messages into five groups based on their lengths and
picked out the group accounting for the vast majority
which were to be chosen for investigation.

available on the official websites of American colleges and

universities. The data shows that on average there are from
five to fifteen assertives in one PWM.
4. Findings and Discussion
Through the process of collecting and analyzing data,
the researcher found that in PWMs, among 5 categories of
illocutionary speech acts, assertives predominate the others
with 429 instances, accounting for 67,45% whereas there
are small proportions of 117 expressives with 18,40%,
70 directives with 11,01% and 20 commissives with
3,14%. Furthermore, the researcher did not find any
declaratives in these PWMs. This shows that assertives
play the most important role in asserting some information,
especially in PWMs.
Table 3. Frequency of Five Speech Act Categories in PWMs

Categories

Occurrence

Rate (%)

Assertives

429

67,45

Expressives

117

18,40

Directives

70

11,01

Commissives

20

3,14

Declaratives

636

100

Total
500

429
Assertives

400

Expressives

300

Directives

117

200

70

100
0

20

Commissives
0

Declarations

Figure 1. Distribution of Categories of Illocutionary Speech

Acts Taxonomy in PWMs

4.1. Lexical features of assertives in PWMs

4.1.1. Adjectives

Number of
messages

Percentage

In PWMs, descriptive adjectives, extreme adjectives

and superlative ones are exploited most in assertives in
PWMs in order to describe, compare and state something
about the tertiary institutions.

0 200 words

24

19,83

Table 4. Frequency of Adjectives in Assertives in PWMs

200 400 words

77

63,64

400 600 words

17

14,05

Descriptive Adjectives

345

76,7

600 800 words

0,83

Extreme Adjectives

56

10,9

800 1.000 words

1,65

Superlative Adjectives

49

12,4

121

100

Total

450

100

Table 2. Length of Welcome Messages in Groups

Number of words in
PWMs groups

Total

As far as Assertives are concerned, the sample of an

assertive speech act must be a sentence or a sequence of
sentences which conveys information about some state of
affairs of the world (i.e. tertiary institutions) from the writers
(i.e. presidents) to the readers (i.e. perspective learners).
Assertives in PWMs are based on Yules classification
[1996] which shows 4 subtypes namely assertion,
description, conclusion and statement of fact.
With the above set criteria, 50 PWMs have been
collected to build up the data. All these PWMs are

Adjectives

Occurrence

Rate (%)

From the data, the descriptive adjectives are used with

a surprising number to describe the strengths of the
American colleges and universities. The superlative
adjectives and extreme adjectives also frequently appear in
the discourse because they help to assert the strong
positions of the institutions with their distinctive capacities
in the tertiary education system. Here are some examples:
(1) Our outstanding faculty are not only excellent teachers
and scholars, they are dedicated to providing a uniquely
engaging, collaborative, hands-on learning experience.

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

(2) We are justifiably proud of our safe, comfortable,

and attractive campus, with its cutting-edge technology,
modern and convenient library, spectacular athletic
facilities, and updated facilities for science and the arts.

these assertives belong to 4 types of sentences: simple

sentences, compound sentences, complex sentences and
compound-complex sentences. These types of sentences
are distributed in PWMs with different proportions:
Table 5. Sentence Types of Assertives in PWMs

4.1.2. Noun phrases

The data shows that there is a variety of noun phrases
used in the assertives to denote the modern facilities and
infrastructures of the American colleges or universities:
(3) Of course, the best way to experience Western's
vibrant learning environment is to visit our state-of-the-art
facilities in the foothills of the Cascade Mountains on the
Puget Sound, between Seattle and Vancouver, Canada.

Sentence Types

Occurrence

Rate (%)

Simple

160

37,30

Compound

19

4,43

Complex

231

53,85

Compound-complex

19

4,43

Total

429

100

(4) We are conducting leadingedge research in

production of ethanol from grass, E20 ethanol conversion
kits, and other alternative energy products and sources.

4,43; 4%
37,30; 37%

Mottos play an important role in the manipulation and

development of the educational institutions. They convey
the educational visions and missions of the institutions as
well as help the educational institutions to be widely
publicized. Therefore, in PWMs, such mottos are included
in assertives for the purpose of capturing readers attention.
As a result, these assertives will carve the institutions
images in the readers hearts and minds.
The data of the research shows that the motto is one of
the common characteristics found in the assertives in most
PWMs. This can be atributed to the fact that the mottos are
brief, succinct, impressive and memorable. In PWMs, the
mottos are structured by various patterns such as
prepositional phrases, noun phrases and verb phrases to
highlight the function of assertives. For example:
(5) They form an inclusive community centered on this
campus but extending throughout the nation and around
the globe, where its members strive to live up to the
University's informal motto: to be "in the nation's service,
and in the service of all nations."
(6)
You
have
found
a
special
place,
where Faith, Wisdom, and Service are the hallmarks of
the Huntingdon College tradition.
(7) Atlanta Metropolitan State College is the place to
BELIEVE, BEGIN and BECOME.
The mottos: "in the nation's service, and in the service
of all nations.", Faith, Wisdom, and Service,
BELIEVE, BEGIN and BECOME are embedded in
these assertives to highlight the crucial missions of the
institutions which are closely attached to the learners
benefits, to denote the striking characteristics of the
institutions and to affirm their strong determination in
realizing their set goals.
4.2. Syntactic features of assertives in PWMs
4.2.1. Sentence Types
With 429 occurrences of assertives found in 50 PWMs,

1
2

4.1.3. Mottos
In this study, it can be seen that mottos facilitate the
management of the institutions. Higher education
institutions employ mottos to make the audience remember
their beliefs or purposes more easily.

111

3
53,85; 55%

4
4,43; 4%

Figure 2. Distribution of Sentence Types of Assertives in PWMs

As can be seen from Table 4, complex sentences are

most commonly used for assertives in PWMs, which
account for 53,85% with 231 instances. The second most
popular type is simple sentences which make up 37,30%
with 160 instances. Compound and compound-complex
sentences occupy the lowest rates with more than four
percent in total.
From the data, we find that complex sentences are used
for making some statements of fact and providing
information on the higher education institutions. In PWMs,
one of the striking structures of complex sentences of
assertives in PWMs is the presence of the clause of
concession. Below are some examples:
(8) The liberal arts have long been at the heart of a
Linfield education, whatever major a student may choose,
and I strongly believe that the liberal arts provide the best
preparation for life and a career in any field.
(9) No matter what type of student you are professional worker, stay-at-home mom, career seeker or
military service member - the convenience and flexibility
of online learning can help make your goals attainable.
4.2.2. Voices
The data analysis shows that the sentences in the active
voice appear with a much higher frequency than the
sentences in the passive voice, as tabulated below:
Table 6. Distribution of Active and Passive Sentences of
Assertives in PWMs

Voice

Occurrence

Rate (%)

Active

391

91,14

Passive

38

8,86

Total

429

100

Significantly, as can be seen from Table 5, assertives in

the form of active sentences predominate in PWMs. With
391 instances occupying 91,14% of the total, the active
voice is given top priority in forming assertives in PWMs.
Orwell [2003] claims that the active voice is shorter

112

Nguyen Thi Quynh Hoa, Doan Thi Hong Nguyen

and more forceful. This has proved to be true in the case

of assertives in PWMs because the active voice helps to
make assertives more direct, succinct and powerful.
(10) One of our primary focuses is to make a quality
college education affordable to you.
In the following examples, the active voice is used
because it is more suitable than the passive voice in
expressing the information the writers intended to convey:
(11) We offer a wide range of experiential
opportunities that are designed to deepen student learning,
broaden perspectives and provide valuable, real-world
experience.
(12) Recent campus facility upgrades include the
construction of the Margaret R. Preska Residence
Community, remodeling of Trafton Science Center,
completing of an Outdoor Recreation project that included
four new softball fields, a new outdoor track with a soccer
infield, along with other facilities, and remodeling of the
Centennial Student Union Ballroom.
4.2.3. Parallel Structures
Through the scrutiny of the collected data, it can be said
that parallelism is thoroughly employed in many assertives
of the PMWs. The analysis has found out some parallel
structures as follows:
a. Parallel Structures with Verbs
(13) Linfield emphasizes lifelong learning, embraces
diverse cultures and international study, and recognizes
moral principle and the freedom of conscience.
b. Parallel Structures with Adjectives
Not only verbs but also adjectives are used in parallel
to reinforce the various and fantastic characteristics of
something or someone. The following assertive serves as
an example of this kind:
(14) We also offer unparalleled occupational skills
training in numerous technical, business, and health-care
fields geared toward preparing students for high-wage,
high-skill, high-demand occupations.
c. Parallel Structures with Clauses
In an effort to draw attention to a particular part of the
message and make it stand out from the rest of the writing,
the writers have made good use of parallel structures to
express their ideas in a series of similar structures. This is
to emphasize that these ideas are on equal rank, which
helps to make the message more memorable.
Especially, parallelism at the sentence level is
employed most in order to emphasize the strengths of the
institutions. Below is an example:
(15) We have an excellent faculty, and our campus
boasts a number of state-of-the-art facilities to enhance our
living and learning environments. We have an impressive
graduate program as well, offering technical, professional
and scholastic skills and competencies in teaching,
research, and public service, including extension. We
provide our graduates with marketable skills that are
among the nation's best.
Besides, the writers prefer to use conditional sentences

so commonly to present the supposition in the if-clause and

to affirm something in the main clause. This emphatic form
stresses the important information or key figures expressed
in the assertives. The following sentences are the most
impressive structures that the researcher found during the
data analysis. This structure has a profound effect on the
audience because they are kept unchanged in many
successive utterances:
(16) If you are wondering whether FLC is the right fit,
please think about the following points. If these points are
important, then this is the place for you.
If you want a world-class education on one of the most
beautiful campuses in the nation, Fort Lewis College is for you!
If you want to have your pick of more than 100 majors,
minors and certificates in four different schools, Fort
Lewis College is for you!
If you are interested in learning from the best
professors in small classes where youre more than just a
number, Fort Lewis College is for you!
If you want a diverse student body, Fort Lewis College
welcomes all students of goodwill, and FLC's commitment
to Native American education dates back a century.
As can be seen from the above example, the writers
repeatedly use the clause If you to picture a great deal
of the learners demand. This series of parallel sentences
implies that the institution recognizes all the concerns of
the learners and then shows them that this institution is the
place where their desires and their future plans can be fully
satisfied. That is why the clause Fort Lewis College is for
you! is continuously repeated at the end of the sentence.
In the following example (17), by choosing four
concise nouns faith, service, responsibility and initiative to
be used in four parallel structures in the corresponding
assertives, the writer has been successful in asserting his
belief in his fabulous institution. This way not only makes
the writing credible but also arouses readers attention.
(17) This a community of faith.
This is a community of service.
This is a community of responsibility.
This is a community of initiative.
4.2.4. Sentence Structures Commonly Used in PWMs
a. Complex Sentences with Whatever/Whichever/
Whether
As specified in Table 4, complex sentences predominate
the other sentence types. The researcher found that the
complex sentences with concessive clauses of the patterns:
Whatever / Whichever / Whether (+ Noun) + Subject +
Verb + Main clause are mostly used. Particularly, these
clauses take the first position in the sentence that reveals the
writers purpose of introducing something first, then
concluding or recommending appropriate things at the end.
They are demonstrated as follows:
The first pattern is Whether you are + Noun phrases as
Subject Complements + Main clause:
(18) Whether you are a new student seeking to get
started, a current student accessing course or college

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

information, or a community member inquiring about

WWCC, you should be able to find needed information here.
The second pattern is You + Will + Bare Infinitive Verb:
This second common structure of assertives in PWMs
is worth being considered because of its high degree of
assertion. In contrast with the concessive clauses, the
pattern You + Will + Bare Infinitive Verb directly
addresses what the writer wants to say and assures the
readers of his uttered things. Here is an instance:
(19) You will find that Linfield is committed to
educating the whole person, both within and outside the
classroom, and that it offers something that students crave
but rarely find elsewhere in todays America: a sense of
genuine community and civil intellectual exchange.
It is obvious that through these sentences, the
Presidents of the American colleges and universities show
their strong determination to meet the perspective learners
relating to the aspects of academic training courses,
facilities, faculty staff and the institutions output.
b. Conditional Sentence Type 1
Assertives are also formed in the structures of
conditional clauses in the PWMs. Once again, this structure
not only requires a person to implement one thing before
obtaining the other thing but it also wants to confirm or
assert that the things they are wondering is no matter.
Conditional Sentence Type 1 is highly preferred as
illustrated in the following example:
(20) If you are a student, you will find the information
you need to progress smoothly to graduation and enhance
your college experience.
4.3. Semantic features of assertives in PWMs
According to Richards and Platt [1992, p.164], the
semantic field is defined as the organization of related
words and expressions into a system which shows their
relationship to one another.
As Riemer [2010, p. 2] claims, any attempt to
understand the nature of language must try to describe and
explain the ways in which linguistic expressions have
meaning and he concludes that semantics is the study of
meaning in language.
Moreover, in the article Semantic Analysis of English
Performative Verbs, MacQueen [2] asserts:
The nature of the primary speech acts that are
performed in the use of a natural language is determined
by the semantic structure of that language. Actual natural
languages such as English and French have in their
lexicon a large number of speech act verbs whose
meanings serve to determine the possible illocutionary
forces of the utterances of their sentences.
In this study, based on Richards and Platt [1992]s
view, Riemer [2010] and especially Yule [1996]s
classification of representatives, through the data analysis
of assertives in PWMs, the researcher found 4 types of
assertives in PWMs. They are assertions, descriptions,
statements of fact and conclusions. Here are some
examples:

113

Assertion: (21) We believe this unique combination of

classroom learning, active service, and spiritual formation
creates a transformational education that produces
graduates who think critically and creatively, lead with
high ethical and professional standards, embody the
character and compassion of Jesus Christ, and who
thereby are prepared to impact the world.
Description: (22) We have high quality transfer
courses and curricula for students who wish to pursue their
first two years of a four-year degree at SMC before
transferring to a four-year college or university.
Statement of fact: (23) As a matter of fact, the College
is approved by SACS to offer 100% of the Business
Administration and General Education Programs and 70%
of all courses online.
Conclusion: (24) Its clear that our special, focused
approach to learning leads to effective problem-solving
collaborations and creative new thinking.
The statistics results of the types of assertives found in
PWMs are shown in the following table and figure:
Table 7. Types of Assertives in PWMs

Types of assertives

Occurrence

Rate (%)

Assertions

238

55,48

Descriptions

88

20,51

Statements of fact

84

19,58

Conclusions

19

4,43

Total

429

100

250

238

200
150
100
50

88

84
19

0
1

Figure 3. Types of Assertives in PWMs

As shown in Table 7, while there appear 4 types of

assertives in PWMs, just three of them account for the vast
majority of the total number. They are assertions with 238
instances making up 55,48%, descriptions with 88
instances making up 20,51%, and statements of fact with
84 instances making up 19,58%. These assertions,
descriptions and statements of fact, which are in the wordto-world direction of fit, are the important types of
assertives in PWMs inasmuch as they function in asserting
the high positions of the institutions in the academic world,
stating the best quality of their training services and
providing as much as necessary information about the
institutions faculty and facilities for the audience.
The remaining type of assertives in PWM - conclusions
- is sometimes used for the purpose of persuading the
learners to believe in the institutions capability as well
recommending the learners some ideas useful for their future
study. Following is the clarification of the above 4 types:

114

Nguyen Thi Quynh Hoa, Doan Thi Hong Nguyen

4.3.1. Assertions
Assertions occur with the highest rate - 238 instances
occupying 55,48% and are mostly expressed with the verb to
be and to believe. They function in asserting or claiming
the mission, vision, and commitment of the colleges and
universities. Take the following assertives as examples:
(21) Our students are the pride of BGSU.
(22) The heart and soul of our University is our faculty.
(23) Huntingdon is a comfortable place where every
student is known and where relationships with faculty,
staff, and students form lifetime bonds.
4.3.2. Descriptions
Descriptions occur with the second highest
proposition. They are to provide more information about
the capacity of the colleges or universities including
training courses and modes, types of degrees, modern
facilities and high qualified faculty and staff. Here are
some examples:
(24) Thats why we offer NCAA Division II Athletics
programs, many visual and performing arts activities,
research and outreach endeavors, and a host of student
programming and volunteer opportunities.
(25) We also have a strong global presence with our own
facilities in Delhi, Beijing, London, Paris, and Singapore,
active educational programs in the College in locations
throughout the world, and the ongoing interaction of scholars
and students at work with colleagues from around the globe.
4.3.3. Statements of facts
Statements of facts rank third from the analyzed data.
A statement of fact exhibits word-to-world direction of fit,
where the utterance fits an independently existing state of
affairs in the world. In PWMs, they mainly focus on the
admirable and amazing achievement of institutions.
(26) And our Virtual Campus was recognized as the
Best of the Best in the Education and Academia category
of the Computerworld Honors Program in June 2009.
(27) We also continue to be the university of choice for
many of the states best and brightest high school
graduates: this year, 127 of 211 Montana high school
graduates offered Montana University System Honor
Scholarships, or 60 percent, chose MSU.
4.3.4. Conclusions

research and outreach endeavors, and a host of student

programming and volunteer opportunities.
5. Conclusion
This article is aimed at investigating assertives employed
in PWMs. The results from the analysis of 50 PWMs show
that the assertives outnumber other illocutionary speech acts.
In terms of the lexical features, most of the assertives
are composed of descriptive adjectives, superlative
adjectives and extreme adjectives. Besides, there appears a
variety of noun phrases denoting the modern facilities and
infrastructures of the US colleges or universities.
Additionally, the mottos which are in forms of
prepositional phrases, noun phrases and verb phrases are
frequently found in theses assertives.
With regard to the syntactic features, the active voice
outweighs the passive voice. Interestingly, although
assertives in PWMs are found in four sentence types:
simple, complex, compound and complex-compound,
complex sentences occupy the highest proportion.
Especially, the complex sentences with concessive clauses
are very common. Furthermore, the parallel structures of
verb phrases, adjective phrases and clauses, the structure
You will .. and Conditional Sentence Type 1 are also
commonly used.
As regards the semantic features, in PWMs, there
appear four forms of assertives namely assertion,
description, statements of fact and conclusion, which can
be considered as the semantic fields of assertives in PWMs.
In a word, assertives as a type of illocutionary speech
act have proved to be one striking linguistic feature of
PWMs, which play a significant role in rendering the
contents of the PWMs.
REFERENCES
[1]

Austin, J.L (1962), How to Do Things with Words, London: Oxford

University Press.

[2]

MacQueen, K. Semantic Analysis of English of Performative Verbs,

http://www.uqtr.ca/~vandervk/english_performative_verbs _ch6.pdf
Richards, J. C., Platt J. and Platt H. (1992), Longman Dictionary of
Language Teaching and Applied Linguistics, London: Longman.

[3]
[4]
[5]

Riemer, N. (2010), Introducing Semantics, New York: Cambridge

University Press.
Robert M. Harnish (1979), Linguistic Communication and Speech
Acts, Cambridge, MA: MIT Press.

Conclusions rarely occur in PWMs but they also

contribute to the asserting function of an assertive speech
act. Once again, they confirm the strong capacity of the
institution among many other colleges and universities.

[6]

Searle, J.R (1969), Speech Acts - An Essay in the Philosophy of

Language, London: Cambridge University Press.

[7]

Searle, J.R (1975), A Taxonomy of Illocutionary Acts, Cambridge:

Cambridge UP.

[8]

(28) From models of urban sustainability to the impact of

scientific exploration and invention, GW is making a difference.

Searle, J.R (1979), Expression and Meaning: Studies in the Theory

of Speech Acts, Cambridge: Cambridge University Press.

[9]

Searle, J. and Vanderverken, D. (1985), Foundations

Illocutionary Logic, Cambridge: Cambridge University Press.

(29) Thats why we offer NCAA Division II Athletics

programs, many visual and performing arts activities,

[10] Yule, G. (1996), Pragmatics, Oxford: Oxford University Press.

(The Board of Editors received the paper on 06/19/2015, its review was completed on 06/27/2015)

of

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

115

AN INVESTIGATION INTO SEMANTIC FEATURES OF HYPERBOLE USED IN

FOOTBALL COMMENTARIES IN ENGLISH AND VIETNAMESE NEWSPAPERS
Luu Quy Khuong1, Cao Nguyen Minh Thong2
1

University of Foreign Language Studies, The University of Da Nang; lqkhuong@cfl.udn.vn

2
Huynh Thuc Khang Highschool, Khanh Hoa; thongcaonguyen214@gmail.com

Abstract - This paper investigates the semantic features of

hyperbole used in football commentaries in English and Vietnamese
newspapers to find out the similarities and differences between them.
To conduct the study, both the descriptive and contrastive methods
are used.The data consists of 600 hyperboles in English and
Vietnamese (300 for each) online newspapers. The results shows
that most of hyperboles used in the two languages are the same in
some typical topics. Along with the similarities, there are some
differences resulted from cultural differences as well. For example,
the same image can be employed to show quite opposite meanings
in the two online newspapers. In general, the paper is expected to
help Vietnamese learners of English achieve comprehensive
understanding of hyperboles in football commentaries in English and
Vietnamese newspapers. Also, those who are interested in writing
this genre can find the paper a good source of reference.
Key words - hyperbole; sem antics;stylistics; football comm entary;
newspaper.

1. Introduction
Hyperbole, one of the popular types of stylistic devices,
plays a crucial part in everyday language use.
Speakers/writers find it a powerful tool to make the
discourse more impressive and to effectively attract the
listeners/ readers attention.This kind of stylistic device
appears frequently in conversations, literature, short
stories, as well as in other discourse types, especially, in
football commentary where all scales of emotion seem to
be all expressed. Nevertheless, in order to catch the ideas
shadowing in the expressions or to have a thorough
knowledge on hyperbole is not easy for Vietnamese
learners of English as well as for foreigners who study the
Vietnamese language and culture. Take the following
sentences in English and Vietnamese as examples:
(1) The Spaniard fired a precise shot through a sea of
bodies and beyond a stranded Boaz Myhill.
(goal.com/en-/match/newcastle-united.../report)
(2) Than Qung Ninh c c hi cc k nguy him pht
36 khi Quang Hi t ph trong phm vi hp trc mt
rng cu th ri c th mn T Vnh Li.
(bongdaplus.vn/.../thua-trn-b-bd--hai-phong)
From the above examples we can see that the
hyperbolical expression in English is different from the
Vietnamese one in meaning but they are used with the
same purpose to emphasize the difficulty the players face
in front of the opponents goal. More interestingly, the
same hyperbolical image is utilized in different functions
as follows:
(3 Ghosting in at the far post, he gained the crucial touch
on Fabregas' delivery to register his first Premier
League goal for the club.
(goal.com/en/match/chelsea-vs-arsenal/.../report)

(4) Trn hng tin v, Fabregas vt v nh bng ma trn

sn. (bongdaplus.vn/tin-bai/36/131539/Chelsea)
English writers make use of the image of ghost as a
hyperbole to talk about someones skill, they find ghost
a special image that a normal person cannot catch sight of
or keep up with to denote Chambers skill, no Arsenal
players could keep up with his performance as in example
(3). On the contrary, the Vietnamese ones often utilize
ghost to denote someone that can do nothing or show the
weakness of someone as we often hear vt v nh bng
maas in example (4).
Obviously, there is a wide variety of semantic features
and uses of hyperbole both in English and Vietnamese
which may cause difficulties to learners. Therefore, this
paper can help Vietnamese learners of English raise the
awareness of the differences between the two languages in
terms of using this stylistic device to enhance their
communicative competence.
2. A review of previous studies related to the topic
Up to now, there have been a number of studies on
hyperbole. Galperin [5] provided the definition of
hyperbole and some functions of hyperbole. Carter and
McCarthy [1] succinctly explained that hyperbole
magnified and produced a contrast with reality. Cutting
[4]introduced some cases in which the speaker flouted
Grices maxim of quality by exaggerations such as I could
eat a horse. Colston and Keller [3], Colston and OBrien
[2] presented modern theories of hyperbole and related
tropes. Hong Tt Thng [10] gave his point of view on
hyperbole and mentioned two functions of hyperbole
which are cognitive and expressive. inh Trng Lc [7],
[8] showed the main purpose of hyperbole and two levels
of hyperbole. Phan Hng Lin [9] assumed that
exaggeration is somewhat related to simile. V Th Kiu
Loan [6] investigated linguistic features of hyperbole in
literature. In general, hyperbole so far has been
investigated in the settings of novels, poems or short
stories.Meanwhile, this paper mainly focuses on semantic
features of hyperbole used in football commentaries in
English and Vietnamese newspapers.
3. Some theoretical concepts
3.1. Stylistic Devices
Galperin [5, p.26] defines that a stylistic device (SD) is
a conscious and intentional literary use of some of the
facts of the language(including expressive means)in which
the most essential features (both structural and semantic)of
the language forms are raised to a generalized level and
thereby present a generative model.

116

3.2. Definitions of hyperbole

According to Galperin [5, p.173], hyperbole is a
deliberate overstatement or exaggeration, the aim of which
is to intensify one of the features of the object in question
to such a degree as will show its utter absurdity.
For Dinh Trong Lac, [7, p.46] Hyperbole (also called:
excess, bombast, overshooting or exaggeration) is the use
of word(s) or expression(s) to multiply the characters of an
object or a phenomenon in order to highlight the nature of
the mentioned object and make a strong impression on the
reader or listener. Hyperbole is different from a lie is that
its nature, motivation and purpose do not blow up or
distort the truth to trick the reader or listener. It does not
make him / her believe in what the speaker says, but makes
him / her understand what is said.
Based on the above definitions, in this paper hyperbole
is understood as a lexical stylistic device which
significantly contributes to creating the wonder of any
language in common and literary works in particular by
using words or expressions several times to multiply the
properties of the object or phenomenon. It aims is to
highlight the nature of the object which needs describing,
and to produce a particularly strong impression or to draw
attention of the reader or make him pay more attention to,
and understand more deeply what the author
writes/speaks. This working definition satisfies the view on
hyperbole of Galperin [5] in English, and the view on
hyperbole of inh Trng Lc [7] in Vietnamese.
4. The corpus
The paper mainly concentrates on the analysis of the
semantic features of hyperbole used in football commentaries
in English and Vietnamese newspapers. The corpus consists
of 300 examples in English and 300 in Vietnamese randomly
taken from online sports newspapers such as goal.com,
nydailynews.com, nytimes.com, and the guardian.co.uk in
English, along with bongda.com, bongdaplus.com.vn,
bongda24h.com, and tinthethao365.com.vn in Vietnamese.
5. Findings and discussion
5.1. Semantic features of Hyperbole used in English and
Vietnamese football commentaries
The data analysis has detected 21 topics divided into 3
groups of hyperboles often used in football commentaries
in online newspapers in the two languages. The first group
is about emotions consisting of joy, sadness, anger, worry,
frightfulness, regret and shock. The second group is on the
nature of a football match such as tension, excitement,
boredom, victory, loss, danger and beauty. The last one is
of elements related to people such as strength, speed,
accuracy, talent, weakness, determination and great ability.
The most noticeable topic in the first group may be joy.
Joy is a feeling of great happiness an inner state of well
being which enables us to profit spiritual values in our
life. In English football commentaries, as happy as kids,
as if it were a World Cup winner, erupt, into a
frenzy and so on are used as hyperboles to express the
state of joy. Like in (5), the writer overshot their happiness,
their joy as happy as kids on summer recess to make a

Luu Quy Khuong, Cao Nguyen Minh Thong

strong impression on readers. Like English writers,

Vietnamese ones usually employ quite a large number of
hyperbolical phrases such as nh mt a tr, nh m
c trong bng, nh n tung to multiply joy in
Vietnamese as in (6).
(5) They spent most of 90 minutes at Ottawas
Landsdowne Stadium lookingas happy as kids on
summer recess, and the upshot was their happiest result
of this World Cup.
(nydailynews.com/.../u-s-soccer-team-road-women-w... )
(6) Ti 8.4, HLV giu c tnh L Thy Hi n n ci
rt ti, nhy ln m chm ly tng ngi nh mt a
tr khi Abass ghi bn san bng t s 1-1 cho Becamex
Bnh Dng (B.Bnh Dng).
(tinthethao365.com.vn/news/tag/l+thy+hi/)
Along with the happiness, sadness is a feeling that
always appears in each football match. Both English and
Vietnamese writers employ hyperbole skillfully to show
this feeling in many ways. As we know that Newcastle
United sleepwalk their way in (7) is an unreal thing, but
we can get the sadness shadowing it or khc thm as in
(8) is, too.
(7) So Arsenal continue to keep Manuel Pellegrini and
perhaps even Jos Mourinho sweating while Newcastle
United sleepwalk their way towards the end of a deeply
unsatisfactory season?
(theguardian.com Soccer Newcastle United 1-2
Arsenal | Premier League)
(8) Ch 1 tun trc y thi, nhng tri tim Arsenal cn
rn rng ht vang bi ca chin thng Wembley, khi
h h b nh v ch Chelsea, khin cho HLV
Mourinho tc n mu mt. Mc d vy, gi y, h
li ang khc thm ngay mi nh Emirates, trc
i th yu hn rt nhiu, West Ham.
(bongda365.com.vn/bong.../arsenal-thua-soc-...)
Similarly, anger, worry, frightfulness and shock exist
during and even after a finished match, and they are
exaggerated tactfully to show anger as tc n mu mt
in (8), or to denote worry, frightfulness and shock as the
following examples:
(9) Newcastle's hearts were in their mouths when Stewart
Downing raced into the penalty area 23 minutes in.
(goal.com/en-/match/newcastle-united.../report.)
(10) Trong nhng pht cui cng thng, nu Casillas khng
xut sc cn 2 pha nguy him mi mi th by Kn
kn m hn thin thu.
(bongda365.com.vn /.../real-madrid-3-4-schalke-)
(11) He talked of a psychological pressure on his players
yet they had the extra man; even though his defenders
frozeat key moments.
(goal.com/en/match/chelsea-vs-arsenal/.../report)
(12) D HLV Trn Bnh S nhn nh rng Than Qung
Ninh l i th rt mnh, nhng cc hc tr ca ng ln
khn gi ch nh vn phi chong vng khi h thua
qu nhanh trong cuc so k chiu nay.
(bongdaplus.vn/.../thua-trn-b-bd-mat-ngoi-dau-)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

In fact, feeling dizzy is a normal status but in this example

(12) it is an exaggeration of shock because no one felt dizzy
at all and the match were happening as it used to be.
In general, hyperboles used for the first group are very
multifarious and diversified and we can find them in Table
1 bellow.

ever, a wave of noise breaking around its cantilevered

sides as Bayern pressed aggressively at the start.
(theguardian.com Soccer Bayern Munich 3-2
Barcelona /Champions League)
(16) Trn t kt lt v Cp nh Vua khi u khng
th hp dn hn.

Table 1. Hyperboles of the first group

English

Vietnamese
1. Hyperboles of Joy

as happy as kids, nh mt a tr, n

erupt, into a frenzy tung, nc lng, v a,
rocking,as if it were a dy sng, ngt ngy,
World Cup winner, a nh m c trong bng,
khng ng, nh tr v t
dream start
ci cht
2. Hyperboles of Sadness
brokenhearted, spirit khc thm, cht lng,
felt broken, remained sp , cn c mng,
unmoved, were stunned
into silence, sleepwalk
their way
3. Hyperboles of Anger
want
to
fight n ti nut sng,tc n
anyone,death
stares, mu mt
The anger was stamped
all over his face
4. Hyperboles of Worry
hearts were in their bu tri lo lng, tri tim
mouths, nervously over nh mun nhy khi lng
ngc
their shoulders,
5. Hyperboles of Frightfulness
froze,
a
attack,almost
silenced

heart t
cng,
ni
m
nh,phi nn th, bt mt,
nh cht lng,

6. Hyperboles of Regret
waste or miss a golden b l/ b qua c hi ngn
opportunity,squander/ vng / khng th ngon n
spurn a gilt-edged chance hn, m hn thin thu
7. Hyperboles of Shock
huge shock

chong vng

In the second group, aspects of a football match such as

tension, excitement, boredom, victory, loss, danger and
beauty are hyperbolized as follows:
(13) This match began with both teams pressing high and
it was breathless. (theguardian.com Soccer Bayern
Munich 3-2 Barcelona /Champions League)
(14) Tottenham tn cng nght th trong 4 pht b gi
hng nu ko li 1 im nhng bt thnh.
(bongda.com.vn/ tin bai/.../real/rodriguez...)
Clearly, breathless and tn cng nght th are of the
same meaning when used to highlight tension in football.
Excitement is a pleasant topic to write or talk about in
football.
(15) For all that, the Allianz Arena was as boisterous as

117

(tinthethao365.com.vn/./Atletico...)
Besides, there are different hyperboles used in English
and Vietnamese to highlight victory and loss as follows:
(17) Sergio Aguero equalled a Premier League record by
scoring five times as Man, City returned to winning
ways with a 6-1 demolition of Newcastle.
(goal.com/en-/match/.../ newcastle/.../rep...)
(18) Lc ny, Cordoba v trn, tinh thn suy sp, ri vo
th thua nh ni .
(bongda365.com.vn/.../msn-toa-sang-barcelona)
Danger in football is rather different from other fields.
It does not cause death like others, but it is about dangerous
situations leading to goals or loss. Nevertheless, English
writers take advantage of such phrases as deadly,
venomous, or lethal to overshoot the dangerous situations
in football.
(19) Despite Athletic's stubborn defending in the third
encounter between the sides this month, Barca took the
lead early in the second half through Suarez's
venomous strike.
(goal.com/en-/match/athletic...barcelona/.../report)
Besides, beauty is also one of the most common
subjects to be hyperbolized. When talking about beauty,
people often think of the appearance of someone or thing
whereas beauty in football can be considered as an art,
therefore, it is hard to hyperbolize it. Dream may be the
most appropriate word to describe beauty and is employed
in most discourse types. Along with nh m(dream),
Vietnamese writers often use tuyt, siu or comparing to
drawing as we sometimes hear p nh v/ nh tranh as in
example(20)to magnify beauty.
(20) Nhng n pht 25, li ca i khch phi rung
ln sau mt pha phi hp nh v ca T.LA.
(bongdaplus.vn/.../danh-bai-hai-phong-1-0-)
Other hyperboles of the second group are summarized
in Table 2 below.
Table 2. Hyperboles of the second group

English

Vietnamese

1. Hyperboles of Tension
palpable, frantic,
breathless,
wild,
ruthlessness,starved
of oxygen

tn cng nght th, p sn

kinh hong, trn Derby ny
la, khng cho mt giy
pht th, chng mt

2. Hyperboles of Excitement
burst into life, as cc k, khng th hp dn
boisterous as ever, hn, hp dn hn c mong
surreal moment full i, siu kinh in
voice, an exhibition,
emotions ran high on and

118

Luu Quy Khuong, Cao Nguyen Minh Thong

English

Vietnamese

off the field

3. Hyperboles of Boredom
y ht nh mt trn giao hu, v
hn
4. Hyperboles of Victory
thriller,
demolition,cruelly
finished off

chin
thng
in
r
nht,chin thng hy dit,
cuc tn st,v thm st
kinh hong,nghin nt,
mn hy dit, cn a
chn, nh mt h en khng
l trong v tr nut chng di
ngn h

5. Hyperboles of Loss
thua nh ni , b mng,
cht, b h nhc
6. Hyperboles of Danger
deadly, venomous, cht ngi, him a, or
or lethal
cc k nguy him
7. Hyperboles of Beauty
dreamy

p nh m, siu phm,
pha phi hp nh v

The third group expressing strength, speed, accuracy,

talent, weakness, determination and great ability may be a
rich field for hyperboles to be employed. In fact, we can
find a wide variety of hyperboles used in this area. Let us
consider the following examples;
(21) But Real were soon back in front through Marcelo's
thunderous shothigh into the net after a one-two with
Ronaldo.
(goal.com/en-gb/match/espanyol-vs-real.../report)
(22) C i cnh nhanh nh ct ca Messi v Neymar c
th khng khin Juve bt ng, v l bi tn cng quen
thuc ca h, nhng hai pha xm nhp ca Iniesta v
tin v Croatia l khng th ngn chn.
(bongda365.com.vn / tam-biet-pirlo)
(23) Getafe's lead lasted just six minutes, however, as
Ronaldo took centre stage once more with a fizzing
free-kick that curled around the wall and into the
bottom-left corner with laser-like accuracy.
(goal.com/en-india/match/real...getafe/.../report)
(24) La Liga, ch c 2qui vt Ronaldo v Messi l c
s ng chuyn thnh bn nhiu hn anh.
(bongdaplus.vn/tin-bai/19/.../quai-vat-ghi-ban-cr7-)
(25) Luis Enrique described Messi as player from another
dimension, noting that while others watch him weekly,
he has the pleasure every day.(theguardian.com
Soccer Barcelona|Champions League)
Besides, writers sometimes use comparing as a strategy
to hyperbolize weakness of someone.
(26) They kept making chances but that high-pressing
backline still looked as brittle as a four-day old pretzel.
(theguardian.com Soccer Swansea Manchester.)

(27) Bn g n khin khng kh ti Etihad nh hi sinh, tt

c ang mong ch cuc li ngc dng ngon mc th
Clichy ln ting vi th vng th hai ri sn, cm hng
ca i ch nh tt xung nh bnh a ngm nc.
(bongda365.com.vn /.../suarez-lap-cu-dup...)
(28) The United striker gave away possession in the leadup to Swanseas equaliser and, running as if he had
lead in his boots, ... (theguardian.com Soccer
Manchester United /Premier League)
Great ability is a topic attracting writers as much as other
subjects. English writers use saviour or talisman to
overstate peoples ability. Moreover, Vietnamese writers
sometimes employ siu nhn or they convey other
famous characters to beautify this topic as in (29).
(29) C chch bng c li gn ca anh hnh nh l cc kh
vi nhiu th mn, nhng Buffon, nh mt Spider-man
bc ra t nhng trang truyn ca Marvel, t chi tt
c. (bongda365.com.vn /tin-chinh/tam-biet-pirlo)
Table 3 below gives a wide range of hyperboles
belonging to group 3.
Table 3. Hyperboles of the third group

English

Vietnamese

1. Hyperboles of Strength
hammer,
thump,
slam,
or,
bullet,
rocket,
cannon,
orstorm,thunderous

cng nh k ch, nh
ba b, or nh tri ph,
n rocket, nh n i
bc, nh v bo

2. Hyperboles of Speed
electricity,lightning,
or blink

nhanh nh chp, nh
ct, nhanh nh mt con
sc, nh nhng con
thoi, nh x gi, thn
tc, nh mi tn, tc
tn la, nh cn lc,
nhanh nh in,

3. Hyperboles of Accuracy
inch-perfect,
placed cc k chnh xc, chnh
finish,
laser- xc tng mm, chnh xc
like accuracy, or like tng centimet, nh t
there was a computer chip in
his boot
4. Hyperboles of Talent
catlike reactions,
a
phenomenal
machine,
stunning,another
dimension,
or
ghost

nh ln ng, khng
goal- tng, xut thn, thn
su, ma thut, siu,
qui vt, thin ti,v
even tinh t trn Di thin h,
sinh vt ngoi hnh tinh,
n t hnh tinh khc

5. Hyperboles of Weakness
nightmare performance,
sleepy, virtually nonexistent,dreadful
defending, or as if he had
lead in his boots, as brittle

cc k m nht, siu ti
t, gn nh mt tch,
nh m ng, nh nhng
k nghip d, thiu cht
xm, vt v nh bng ma,

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

English

Vietnamese

as a four-day old pretzel thi u nh eo ch, nh

bnh a ngm nc
6. Hyperboles of Determination
covered mile after mile,
frantic,
stifle,
rejuvenate, non-stop,

cy tng cm, nh mt chin

binh, in cung tn cng,
cht b, or nh mun n
ti nut sng i th

7. Hyperboles of Great Ability

saviour, talisman
super-human

, ng cu
nhn

sinh,

siu

From the examples above, it can be found that

hyperboles are really multifarious and diversified. Each
writer has different ways to make fantastic and creative
hyperboles. The more skilled the writer is, the more
interesting the comment is.
5.2. Frequency of English and Vietnamese hyperboles
used in sports online newspapers
Table 4. Relative frequency of hyperboles in English and
Vietnamese used in football commentaries

Topics

English

Vietnamese

Occurrence

Occurrence

[1] joy

14

4.7

22

7.3

[2] sadness

17

5.7

11

3.7

[3] anger

1.7

0.7

[4] worry

2.3

11

3.7

[5] frightfulness

1.7

1.7

[6] regret

15

5.0

1.0

[7]shock

0.3

3.0

[8] tension

32

10.7

18

6.0

[9] excitement

2.3

18

6.0

0.0

2.3

6.3

11

3.7

0.0

1.3

[10] boredom
[11] victory

19

[12] lose
[13]danger

2.0

1.0

[14] beauty

0.3

12

4.0

[15] strength

65

21.7

21

7.0

[16] speed

19

6.3

26

8.7

[17] accuracy
[18] talent

17
24

5.7

7.0

8.0

21
41

13.7

[19] weakness

32

10.7

36

12.0

[20] determination

3.0

14

4.7

[21] great ability

1.7

1.7

300

100 %

300

100 %

TOTAL

Table 4 indicates that hyperboles expressing strength is

the most frequently employed in English with 21.7% and
talent is the most in Vietnamese with 13.7%. While
hyperboles expressing beauty and shockare perhaps the
least-commonly used in English with 0.3%, Vietnamese
writers find them good topics to hyperbolize in their
commentaries and on the contrary, anger, regret and

119

danger are rarely utilized in Vietnamese with 0.7% and

1.0% but they are common fields to be exaggerated in
English. Interestingly, both English and Vietnamese
hyperboles share the same distribution for frightfulness and
great ability: 1.7%. The results also show that there is no
English hyperboles expressing boredom, loss. From the
table, we can find that most English and Vietnamese
writers consider the third group of elements related to
people as fascinating features for hyperboles to be
employed and English writers seldom pay attention to the
second group, which is contrary to Vietnamese ones. The
other discrepancies are not remarkable.
6. Conclusion and suggestions
6.1. Conclusion
In brief, the analysis above shows that English and
Vietnamese sports journalists employ quite a large number
of hyperboles when they want to highlight a certain feature
of the mentioned object, to create a strong impression on
readers or to make their writings more dramatic and
interesting.
a. Similarities
Hyperboles used in football commentaries in English
and Vietnamese newspapers have many things in common.
They express most of the basic aspects of football in3
groups of 21 topics. All scales of emotion or feelings can
be found there. Other features related to football such as
tension, excitement, strength, victory, speed, accuracy,
beauty, talent, and so on are also denoted. More
importantly, both hyperboles in English and Vietnamese
are the same in some typical topics. For example, both
English and Vietnamese writers use dream to denote
beauty; erupt for joy or break ones heart instead
of sadness. Most English and Vietnamese writers consider
cannonorthunder as hyperboles for strength or
lightning instead of speed. Moreover, genius and
sleepy may be the popular hyperboles replacing talent
and incapability, and so on.
b. Differences
There are some differences between English and
Vietnamese hyperboles used in newspaper football
commentaries. Firstly, Vietnamese writers sometimes
consider boredom of a football match and loss or a lost
game as interesting topics to hyperbolize; , these aspects , on
the other hand, are rarely paid attention to by the English ones.
Secondly, while English writers tend to utilize hyperboles to
promotestrength, the Vietnamese ones find talent a good
topic for hyperboles to be employed. Finally, the same idea
but different words are employed in English and Vietnamese;
when talking about accuracy, while inch is used in
English, centimeter or millimeterare resorted to in
Vietnamese. Clearly, there is a slight difference between the
underlying cultural features. English writers make use of the
image of ghost as a hyperbole to talk about someones
talent, they find ghost a special image that a normal
person cannot catch sight of or keep up with to show skilled
players. On the contrary, Vietnamese ones often utilize
ghost to denote the weakness of someone as we often hear
vt v nh bng ma.

120

Luu Quy Khuong, Cao Nguyen Minh Thong

- To students:

6.2. Suggestions
a. Problematic Situations in Foreign Language
Teaching and Learning
In reality, learning to interpret and express
exaggeration seems to be a difficult task for Vietnamese
learners of English. Moreover, hyperboles represent a part
of English largely untaught, whereas they exist in almost
every aspect in our daily life. Therefore, in order to
understand and use hyperboles correctly, learners should
first equip themselves with a wide knowledge of cultural,
religious and social aspects.
b. Suggestions
Teaching

for

Language

Learning

and

In order to help Vietnamese learners of English to use

hyperbole better, suggestions for learning and teaching this
stylistic device are provided below.
- To teachers:
Firstly, hyperboles occur in most aspects in our life and
sometimes cause some difficulties for Vietnamese learners
of English. Therefore,teachers should draw their attention
to the differences between what is stated and what is really
meant in both languages.
Secondly, fluency in a language includes mastering
how to produce and understand words or phrases with
particular meanings. Consequently, foreign language
teachers should supply learners with knowledge of
hyperboles and stimulate them to beautify their own
language.

Students should master features of hyperboles in the

two languages, which will make it easier for them to
understand and use hyperbolical expressions. Also,
students need to improve their knowledge of language to
be able to understand hyperboles in reading texts, and make
daily conversations more communicative with hyperboles.
REFERENCES
[1]

Carter, R. A., & McCarthy, M. J. (2004). "There's millions of them:

Hyperbole in everyday conversation." Journal of Pragmatics 36:
149-184.

[2]

Colston, H. L. & OBrien, J. (2000), Contrast and Pragmatics in

Figurative Language: Anything Understatement Can Do, Irony Can
Do Better, Journal of Pragmatics, 32, 1557-1583.
Colston, H. L. & Keller, S. B. (1998), Youll Never Believe this: Irony and
Hyperbole in Expressing Surprise, Journal of psycholinguistic Research
27, 499-513.

[3]

[4]

Cutting, J. (2002), Pragmatics and Discourse, London and New

York: Routledge.

[5]

Galperin, I.R. (1971), Stylistics,Higher School Publishing House,

Moscow.
Vo Thi Kieu Loan (2008), The Syntactic, Semantic and Pragmatic
Features of Hyperbole in English and Vietnamese Literature, Master
Thesis, University of Danang.

[6]

[7]
[8]
[9]

inh Trng Lc (1995), 99 phng tin v bin php tu t ting Vit,

NXB Gio dc.
inh Trng Lc (1999), Phong cch hc ting Vit, NXB Gio dc.
Phan Hng Lin (2002), ting Vit ngy cng trong sng, NXB
Vn hc.

[10] Hong Tt Thng (1993), Phong cch hc ting Vit hin i, NXB
Trng i hc Tng hp Hu.

(The Board of Editors received the paper on 11/18/2015, its review was completed on 11/23/2015)

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

121

AN INVESTIGATION INTO SEMANTIC FEATURE OF ENGLISH SLANG

CONTAINING WORDS DENOTING ANIMALS
Dinh Quang Trung
Ton Duc Thang High School, Khanh Hoa
Abstract - Vietnamese learners of English are likely to get
confused when they confront English slang - a linguistic modality
confined to spoken language. They find it difficult to explain and to
grasp slang as a unitary phenomenon. One difficulty that users of
English face when dealing with slang that contains words denoting
animals is to understand their meanings because a slang in
general, as well as a slang containing words denoting animals in
particular is a word or a sequence of words whose meaning can
not be predicted from the m eanings of the words themselves.
Therefore, the purpose of this investigation is to give users of
English different semantic fields when dealing with slang that
contains words denoting anim als. Collected from m any different
sources, 359 slang words and ewxpressions containing words
denoting animals are categorized and arranged into each sem antic
field logically.
Key words - language; slang; English users; animal; semantic field.

1. Statement of the problem

Slang is a language which rolls up its sleeves, spits on
its hands- and goes to work Carl Sandburg quotes [12].
Slang is full of vivid reality that almost any language
has, whether wanted or not. The writer Victor Hugo took
notice of using slangs in his work Le dernier Jour d'un
Condamne (1828). He even devoted his entire volume VII
in the fourth section of the massive popular novel Les
Miserable (1861) to discuss the slang. These are identified
by Victor Hugo: All profession, all business, any
random social system and all the forms of intellectual have
their own slangs. On the pure literary, slang can be studied
much more interestingly than other sciences.
Slang is language deliberately selected for its striking
informality and is consciously used in preference to proper
speech (or, more rarely, writing). It usually originates in
small social group. For these groups, it is a private code
that embodies their particular values and behavior and
reinforces their exclusivity. As a result, understanding and
using slangs correctly and effectively is not an easy task.
Sometimes, non-native speakers of English find it difficult
to comprehend what is meant by English slangs as well as
what structures the slangs belong to. Moreover, slangs
denoting animals are various in the ways of expressing
their meanings because, in some cases, we cannot
understand some expressions by basing on their literal
meanings. As a matter of fact, you frequently hear such
slang containing words denoting animals in your daily life:
- My God, he made a real dogs breakfast of that
presentation.
[8, p.131-132]
- We aint gonna walk no eight milestonight. My
dogs are burned up. [1, p.11]
- Man, who was that fox I saw you with? [7, p.78]
- Hey, monkey. Wanna go to a house party Saturday night?
[13, p.59]

- Plinio, the barman with ducks disease came running up.

[1, p.11]
Very often we understand every individual word in this
sentence but still fail to grasp what the sentence is all about.
Thus, if we do not learn slang in general and slang containing
words denoting animals in particular, we cannot understand
their exact meanings. It means that it is so difficult for us to
master true and full meanings of such phrases as dogs, dogs
breakfast, cat fight, mouse potato and bird food if you are
not under the light of slang, especially slangs containing
words denoting animals. In consequence, we lose interaction
in communication because we cannot understand speakers
intentions and opinions.
For the above reasons, the topic An Investigation into
Semantic Features of English Slang expressions containing
Words Denoting Animals has been chosen for this article.
2. Problem solution
2.1. Definition of slang
In Wikipedia-The Free Encyclopedia [10], slang is
defined as the use of informal words and expressions that
are not considered standard in the speaker's dialect or
language It is also used to identify with one's peers.
According to Yule [9, p. 211], he defines slang in his workThe Study of Language as follows slang describes words or
phrases that are used instead of more everyday terms among
younger speakers and other groups with special interests.
In The Concise New Partridge Dictionary of Slang and
Unconventional English by E. Partridge and P. Beale [5],
slang is a peculiar kind of vagabond language, always
hanging on the outskirts of legitimate speech, but continually
straying or forcing its way into the most respectable
company. As noted by Dalzell and Victor (2008: xv),
slang is easy enough to use, but very hard to write about
with the facile convincingness that a subject apparently so
simple would, at first sight, seem to demand.
Slang is a language which rolls up its sleeves, spits on
its hands- and goes to work Carl Sandburg- an American
poet once described [12]. As to Chesterton G.K, he referred
to slang as the one stream of poetry which is constantly
flowing [2, p. 61].
In the definition of Online Oxford Dictionary from the
website: http://oxforddictionaries.com, [11] slang is a
type of language consisting of words and phrases that are
regarded as very informal, are more common in speech
than writing, and are typically restricted to a particular
context or group of people. Another definition of slang is
that slang is words or expressions that are very informal
and are not considered appropriate for more formal
situations. Some slang is used only by a particular group of

122

Dinh Quang Trung

people in Macmillan English Dictionary For Advanced

Learners of American English [4, p.354].

- Make one move, and youre dog meat. [7, p.61]

2.2. Sampling

- The crocodile rows of little children [1, p.55]

Data needed for the study are taken from various

sources like books, dictionaries and web pages on the
internet. The samples should include 359 English slang that
contain words denoting animals.
2.3. Sample Analysis
This study was carried out with descriptive, analytical
and qualitative methods.

(a dead person)
(a group walking two by two in a long file)
3.2.2. Sex/sexual orientation (23/57)
Generally speaking, sex and sexual orientation are
mentioned much when people in general or teens in
particular use slangs: dog style, pig party, sheeps eyes,
birds and bees, fish-kiss, monkey spank, snake, turtle, pig
meat, cat lapper...

3. Results and discussion

- He fish-kissed me, then ran back to his car.[7, p.74]

The implication which the English want to mention

through slangs containing words denoting animals is very
meaningful. In this study, the implications are discussed
and analyzed in the following 12 main semantic fields and
sub-fields which are based on John Aytos classification of
slang [1].

(to kiss someone with puckered up lips)

3.1. The body and its functions (48/359)

3.1.1. Parts of the body/ bodily functions. (39/48)
In this semantic field, it occupies a much more
proportion than the rest. The implications of slangs show
different parts of the body and its functions. For example:
dog-log, shoot the cat/whip the cat, feed the fish, step on a
duck, see a man about a mule, drain the dragon, snakes,
lay hot snakes, hang a rat...

- He did not look like a wolf, but he did look like a

young man with an eye for a girl. [1, p.67]
(a promiscuous man)
- He goddam tried to snake my old lady. [8, p.405]
(to seduce and/or have sex with)
3.2.3. Crime/prison (5/57)
Only 5 slangs that are often used by criminals show
crime and prison. For example: tiger cage, rabbit season,
cat walker, wolf-pack, mule.
- I bought from him a copula times. He was a mule,
Dad. That means he pushed to other kids. [1, p.98]
(a smuggler)

- I think I stepped in a pile of dog-log. Yuck!

3.2.4. Police/military/maritime/air force (11/57)

(a section of dog feces) [7, p.61]

For example: dog driver, bull, pig heaven, bear, bear in

the air, duck, dragonfly, turkey shoot

- You bird-watchers should just mind your own

business! [7, p.19]
(a girl watcher, usually a man who enjoys watching
women go by)
- Ive been drinking too much; I have to go drain the
dragon. (urinate) [18]
3.1.2. Illness/physique. (9/48).
Not many slangs containing words denoting animals
implies illness and physique. For example: crook as a dog,
sick as a dog, goat knee, monkey bite, pig, whale out,
beached whale
- Jodie whaled out when she was in college, but
slimmed down when she went back to work. [17]
(gain weight usually)
- Fred: I cant come in to work today, Im sick as a dog.
Jane: Too badhave a nice trip to the beach! [16]
(trying to get out of something such as work or school)
3.2. People and society (57/359)
3.2.1. People/group (18/57)
They tell us not only about ethnic, national groups or
people but also their status and relation. For example: dog,
be dog meat, lucky dog, mouse potato, sitting duck, fox
hunter, turkey, pig Island, crocodile, bigger fish to fry,
frog, ...
- Pig Island is no place for the likes of you.[1, p.374]
(New Zealand)

- Theyve got a bear in the air on duty in northern

Indiana. [7, p.14]
(a police officer in an airplane or a helicopter)
3.3. Animals (3/359)
Animal that is one of the semantic fields has the least
quantity of slangs. For example: pig-dog, horse hockey,
rats with wings.
- You dont see horse hockey in the streets anymore.
(horse dung) [6, p.112]
3.4. Sustenance and intoxication (40/359)
3.4.1. Tobacco/alcohol/drugs (31/40)
Almost slang in this field is about tobacco, alcohol and
drugs. For example: horseshoe, monkeys nuts, dog juice,
bite the dog, sheep-dip, drink like a fish, goose eye,
monkey swill...
- A bee is what he calls his habit; its always stinging
him to get a fix. (a drug addiction) [3, p.12]
- Where did you get this monkey swill? This would kill
a monkey anyway. (strong liquor) [7, p.139]
- Ironbark went into the poison shop. Old Nick handed
him a glass of snake juice. [1, p.144]
(poor-quality drink)
3.4.2. Eating/drinking (9/40)
For example: street dog, mousetrap, hen-fruit, rabbits
food, pig out, rats coffin, cat-soup

ISSN 1859-1531 - THE UNIVERSITY OF DANANG, JOURNAL OF SCIENCE AND TECHNOLOGY, NO. 12(97).2015, VOL. 1

- Do you want some cat-soup on your burger?

(ketchup) [6, p.39]
- Although often dismissed as mousetrap, Cheddar
is much the most popular cheese in Britain. [1, p.136]
(cheese)
- You can both come back with me and eat pounds of
rabbits food. (salad) [1, p.136]

123

- If Sally insists on being a dogs mother on this matter,

Ill tell her what I think of her. [7, p.61]
(a bitch)
- She is such a fox! (a sly person) [1, p.291]
3.8. Thought and communication (39/359)
3.8.1. Thought (17/39)

3.5.1. Tools/ implements/containers (4/11)

For example: cat off, birdbath, silly goose, mule head,

chicken-head, bird-brained, mad cow, chicken oriental,
have snakes in ones head, sacred cow

For example: dog and cat, cat pan, lions share,

monkeys island.

- She totally cowed out the day before her Chem

midterm. (to lose control and go crazy) [13, p.33]

- Up on the monkeys island he had realized there

would be no power for the lights. [1, p.391]

- Hadnt it been plain all along that there was a streak

of madness in the old boy? ... He had done a spell in the rat
house and was only out on sufferance.
[1, p.305]

3.5. Articles and substances (11/359)

(parts of a ship)
3.5.2. Clothing/accessories (7/11)
For example: whale tail, turtles, bull and pants, ducks
ass/behind, ...
- Oh, my God! Look at that butt, her whale tail is
showing so high. [15]
(the shape formed when a g-string rides up high over a
womans trousers)
- He looked like another sub-Elvis, smooth flesh and
ducks ass hair. (coiffure) [1, p.175]
3.6. Money, commerce and employment (29/359)
3.6.1. Money (14/29)
The examples of slangs denoting money are bad dog,
fat cat, fish-skin, whale in the bay, goose and duck,
elephant bucks, ...

( a mental hospital)
3.8.2. Communication/education (22/39)
Just look at such some examples as dog it, frog slicing,
cow college, dog on, safety monkey, bull session, goat
mouth, bull butter
- When I tripped, my friends all dogged on me.
(to put down, to mock) [13, p.36]
- Im sure theyll be saying terrible things about us, but
quite frankly I couldnt give a monkeys. [8, p.295]
(a damn)
3.9. The arts, entertainment and the media (12/359)
3.9.1. Cards/gambling (9/12)
For example: cats eyes, fish and chips, duck bucket,
snake eyes, rabbit ears, rats and mice

- My client is not the kind of man to be satisfied with a

mere monkey for his services. (500) [8, p.295]

- We used to play dice with themRats and mice the

game was called. (game of dice) [1, p.360]

- In peacetime, officers in the British Army were men

of independent means to whom their Army pay was
chicken-feed.
(a small amount of money) [1, p.183]

3.9.2. Sports/performance (3/12)

3.6.2. Work/Business/commerce (15/29)

Here are some examples: chase the dog, eat (ones) own
dog food, lion, goat roper, cowboy, turkey bacon...
- She gave him the bird finally and for good.
(dismiss/be dismissed) [1, p.204]
3.7. Behavior, attitudes and emotions (73/359)
3.7.1. Good/positive/suitable feelings (27/73)
For example: wolf bait, silver fox, finer than frog hair,
cats ass, go on the lamb, curiosity killed the cat, pig in shit,
goat heaven, hen party
- They were having a whale of a good time as they
helped each other set up their cots. They were horsing
about. (to enjoy oneself/have fun) [1, p.236]
3.7.2. Bad/negative/unsuitable feelings(46/73)
For example: dogs mother, pig it, dog someone
around, cook ones goose, foot dragon, dogs breath, fox,
cat fight, rabbit killer
- What's the matter, are you chicken? [8, p.87]
(a coward)

Just 3 slang words are used to refer to sports and

performance and they, of course, account for only 15%. For
examples: chicken run, frog show, frog salad.
3.10. Time and temperature (4/359)
They are cold like dog nose, monkey bath, donkeys
years, until the cows come home.
- It was donkeys years since he had been in an English
train. (a very long time) [1, p.361]
3.11. Location and movement (20/359)
3.11.1. Places (6/20)
Here are some examples: dog-ear, pigs ballroom,
Mickey mouse, rat-run, snake room, frog and toad.
- Im off down the frog and toad for a pint of pigs.
( a road) [8, p.170]
- Drinking and curling are synonymous and many a
good rink has lost a crucial Brier game or two in a hotel
snake room. (a bar) [3, p.170]
3.11.2. Movement/vehicles (14/20)
For example: pig off, pigmobile, turn turtle, horses,
panda car, mules ear, wolf in the pack...
- Look out! Here comes the pigmobile.

[7, p.160]

124

Dinh Quang Trung

(a police car)

9. The arts, entertainment and the media

12

3,3%

- How many horses does this thing have? [7, p.112]

Cards/gambling

2,5%

(horse power as in a engine)

Sports/performance

0,8%

3.12. Abstract qualities and states (23/359)

3.12.1. Good qualities/states (6/23)

10. Time and temperature

1,1%

11. Location and movement

20

5,6%

They are queen bee, donkey-click, pull a tiger, duck

soup, on like Donkey Kong

Places

1,7%

Movement/vehicles

14

3,9%

12. Abstract qualities and states

23

6,4%

Good qualities/states

1,7

- Man! That guy really pulled a Tiger. [14]

(the public revelation
infidelities)

of Tiger Woods marital

Bad qualities/states
Total

3.12.2. Bad qualities/states (17/23)

For example: dog-box, pigs breakfast, eat like a horse,
the dog dead, work like a dog
- Youve made a right pigs breakfast of that.
(a mess, a complete disaster) [8, p.334]
Table 1. Summary of semantic fields of slang containing words
denoting animals
Semantic fields (359 slangs)

Quantity Percentage

1. The body and its functions

48

13,4%

Parts of the body/ bodily functions

39

10,9%

Illness/physique

2,5%

2. People and society

57

15,9%

People/group

18

5,0%

Sex/sexual orientation

23

6,4%

Crime/prison

1,4%

Police/military/maritime/air-force

11

3,1%

3. Animals

0,8%

4. Sustenance and intoxication

40

11,1%

Tobacco/alcohol/drugs

31

8,6%

Eating/drinking

2,5%

5. Articles and substances

11

3,1%

Tools/ implements/containers

1,11

Clothing/accessories

1,99%

6. Money, commerce and employment

29

8,1%

Money

14

3,9%

Work/Business/commerce

15

4,2%

7. Behavior, attitudes and emotions

73

20,3%

Good/positive/suitable feelings

27

7,5%

17

4,7

359

100%

4. Conclusion
In conclusion, on the semantic fields, slang containing
words denoting animals reflects mostly all the fields of the
life. People and society and behavior, attitudes and
emotions are 2 semantic fields that rank the 2 highest
positions. Based on this investigation, we are able to see
the whole of semantic fields of slang containing words
denoting animals used in daily life.
REFERENCES
[1]
[2]
[3]
[4]
[5]

[6]

Ayto G. (1998), Oxford Dictionary of Slang, Oxford University

Press, New York.
Chesterton G.K (2008), The Defendant, Nu Vision Publication, LLC.
Dalzell T. and Victor T. (2008), Vice Slang, Routledge, London and
New York.
Macmillan (2002), English Dictionary for Advanced Learners of
American English, Macmillan Publishers Limited.
Partridge E. and Beale B. (2008), The Concise New Partridge
Dictionary of Slang and Unconventional English, Routledge,
London and New York.
Spears R.A (2007), Mc Graw-Hills essential American slang
dictionary, 2nd edition, Mac Graw Hill.

[7]

[7] Spears R.A. (2007), Mc Graw-Hills super-mini American slang

dictionary, 2nd edition, Mac Graw Hill.

[8]

[8] Thorne T. (2007), Dictionary of Contemporary Slang, A& C

Black Publishers Ltd, London.

[9] Yule G. (2006), The Study of Language, 3 rd edition, Cambridge

University Press.
[10] https://en.wikipedia.org/wiki/British_slang
[11] https://www.oxforddictionaries.com/definition/english/slang
[9]

[12] http://www.iwise.com/jXMIe
[13] www.linguistics.ucla.edu/publications/opl_12.pdf
[14] http://onlineslangdictionary.com/meaning-definition-of/pull-a-tiger

Bad/negative/unsuitable feelings

46

12,8%

8. Thought and communication

39

10,9%

[15] http://www.urbandictionary.com/define.php?term=W hale+Tail

[16] http://www.urbandictionary.com/define.php?term=Sick+as+a+dog

Thought

17

4,7%

[17] http://onlineslangdictionary.com/meaning-definition-of/whale-out

Communication/education

22

6,2%

[18] http://www.urbandictionary.com/define.php?term=drain+the+dragon

(The Board of Editors received the paper on 09/25/2015, its review was completed on 10/15/2015)