Professional Documents
Culture Documents
Editor :
Rafiuddin Syam, PhD Hasanuddin UniversityJapan
Prof. Keigo Watanabe-Okayama University-Japan
Prof. Mitsuhiro Okayasu-Ehime University-Japan
ii
ii
PROCEEDING OF
THE 1ST INTERNATIONAL SYMPOSIUM
ON SMART MATERIAL AND MECHATRONICS
ISBN 978-602-71380-1-8
2014 Graduate School of Mechanical Engineering, Faculty of Engineering, University of Hasanuddin
This work is copyright. no part may be reproduced by any process without prior written
permission from the Editors. Requests and inquiries concerning reproduction and rights should be
addressed to Rafiuddin Syam, PhD Graduate School of Mechanical Engineering, Faculty of
Engineering, University of Hasanuddin Makassar-Indonesia
email to rafiuddinsyam@gmail.com
The intellectual property of each paper included in these proceedings remains vested in the
Authors as listed on the papers.
Published by :
Graduate School Mechanical Engineering Engineering Faculty of Hasanuddin University
Jl. P. Kemerdekaan Km 10 Makassar
Sulawesi Selatan, Indonesia 90221
Telp/Fax : (0411) 586015
Email : teknik@unhas.ac.id
Website: pasca.unhas.ac.id
ii
iv
Foreword
First, we would like to thank all participants who are willing to send the results of scientific
research papers and participated in the International Symposium on Smart Material and
Mechatronics 2014 As the first International Symposium conducted by Graduate School of
Mechanical Engineering, Hasanuddin University, and our challenge theme is Smart Material and
System Mechatronics.
The International Symposium on Smart Material and Mechatronics 2014 presented as gifts
birthday of Hasanuddin University to 58 years old. We hope, in this symposium some steps are to
conduct research and publications acceleration in the field techno-science include Smart Materials
and System Mechatroncis. The both field of science that became one of the sections that need to be
encouraged to become an advanced nation of Indonesia in the field of technology. Furthermore, the
results of research are good input for accelerating industry.
In this symposium we invite on the field of research area, but not limited to:
Metal Material, Smart Material, Concrete Material, Composite Material, Strength and stress
of Material, Structure Analysis, Cad and Cam, Vibration and Acoustic, Transportation
System, Environmental Study, Mining, Chemistry, Naval Architecture, Hydrodynamics,
Machining, Production, Heat and Mass Transfer, Thermodynamics, Fluid Mechanics,
Agriculture Engineering, Education Engineering, conservation energy, new energy and
renewable energy, internal and external combustion engine, Civil Engineering.
Mechatronics, Mobile Robot, Manipulator Robot, Intelligence Systems, Softcomputing,
Artificial Intelligent System, Simulation System, Modeling Systems, Industrial engineering,
Ergonomics, Physics, Applied Mathematics, Computer Science, Information Science,
Smart System on Building, Mechanical System, Design systems, Control System, Control
Practice, Adaptive Control, Sensor Engineering, Electrical and Electronics Engineering,
Material on Electrical and Electronics, Environmental Engineering, Rescue Systems, Smart
Vehicle, Smart Building, Biological Engineering, medical Engineering, Artificial Systems,
Fuzzy Logic Theory and Application.
Thanks to all of my college and all of students of graduate school of Mechanical Engineering
Hasanuddin University.
v
vi
Contents
ST
PROCEEDING OF THE 1 INTERNATIONAL SYMPOSIUM ON SMART MATERIAL AND
MECHATRONICS
No Title Pages
1 Measurement and Control for Unmanned Ground, Aerial and Underwater Vehicles K01
Keigo Watanabe, Okayama University, Japan
2 Mechanical Properties of Composite Materials, 1-6
Mitsuhiro Okayasu, Ehime University, Japan
The Effect of Tool Dimension, Tool Overhang and Cutting Parameters Towards Tool Vibration and
3 Surface Roughness on Turning Process Zuingli Santo Bandaso, Johannes Leonard 7-13
ATIM, Indonesia, University of Hasanuddin, Indonesia
Effect of Water Volume and Biogas Volumetric Flowrate in Biogas Purification Through Water
4 Scrubbing Method Hendry Sakke Tira, Yesung Allo Padang, Mirmanto and Rio Cristovan Mantiri 14-18
Mechanical engineering, Mataram University, Mataram, Indonesia
Effect of Solution Treatment Process on Hardness of Alumina Reinforced Al-9Zn Composite
5 Produced by Squeeze Casting
Dwi Rahmalina, Hendri Sukma, I. Gede E.Lesmana, Asrin HalimPancasila University, Jakarta, 19-22
Indonesia
6 Review of Carbon Fiber Reinforced Polymer Reinforced Material in Concrete Structure 23-30
Ayuddin Gorontalo State University, Indonesia
7 Material properties of various light metals produced by heated mold continuous casting 31-35
Yuta Miyamoto, Mitsuhiro Okayasu, Japan
Microstructure and Mechanical Properties of Al-10Zn-4.5Mg-xCu Turbine Impeller Produced by
8 Investment Casting, 36-38
Muhammad Syahid, Bondan T. Sofyan1,Insani Mukhlisa University of Indonesia, Indonesia
9 Study of Performance Improvement of Various Stoves with Waste Biomass Briquettes Fuel 39-45
Effendy Arif Sallolo Suluh*, Unhas, *ATI Dewantara Palopo, Indonesia
10 Position Control of an X4-Flyer Using a Tether 46-49
Yusuke Ouchi, Keigo Watanabe, Keisuke Kinoshita, Isaku Nagai, Okayama Univercity, Japan
11 Development of a Mobile Robot as a Test Bed for Tele-Presentation 50-58
Diogenes Armando D. Pascua, Sherwin A. Guirnaldo, Mindanao State University Philippines
12 Intelligent Machine Vision for Automated Fence Intruder Detection Using Self-organizing Map 59-66
Veldin A. Talorete, Jr., Sherwin A. Guirnaldo MSUIligan Institute of Technology, Philippines
Simulation and Experimental Works of Quadcopter Model for Simple Maneuver
13 Rafiuddin Syam and Mustari, Hasanuddin University and Dayanu Iksanuddin university, Baubau- 67-71
Indonesia
Design of Wheeled Mobile Robot with Tri-Star Wheel as Rescue Robot
14 Rafiuddin Syam, Wahyu H. Piarah and *Paisal, Hasanuddin University, *State Polytechnic of 72-76
Ambon, Indonesia
Application of Genetic Algorithm for Determining the Optimum Ship Route
15 Faisal Mahmuddin, Rahmad Patarru, Rahimuddin Samad, Hasanuddin University, Makassar, 77-82
Indonesia
16 Fatigue Life Prediction In Journal Bearing 83-86
Irsyadi Yani Hasan Basri and Hafizd Ibrahim Marsil, Sriwijaya University, Indonesia
vii
17 Development of 5-DOF Robot Arm Manipulator 87-89
Ismail Thamrin, Irsyadi Yani, Sriwijaya University, Palembang, Indonesia
18 Kinodynamic Motion Planning for an X4-Flyer Using a 2-Dimentional Harmonic Potential Field 90-93
Kimiko Motonaka, Keigo Watanabe, and Shoichi Maeyama, Okayama University Okayama, Japan
19 Underactuated Control for a Blimp with Four-Propellers by a Logical Switching Method, 94-97
Yoshikazu Nakamura, Keigo Watanabe, Isaku Nagai, Okayama University Okayama, Japan
20 New Waste Beverage Cans Identification Method, 98-101
Firmansyah Burlian, Yulia Resti, Ihsan Budiman, Sriwijaya University,Indonesia
21 Experimental Test of the Thermoelectric Performance on the Dispenser Cooler 102-106
Zuryati Djafar, Amrullah, Wahyu H. Piarah, Syukri Himran, Hasanuddin University, Indonesia
22 Potential Coir Fibre Composite for Small Wind Turbine Blade Application 107-109
Bakri, S.Chandrabakty, R. Alfriansyah, A. Dahyar, Tadulako University, Palu, Indonesia
23 A new development of thermosiphon solar hot water with paralel-serpentine tube configuration, 110-113
Mustofa, Yuli Asmi Rahman, Basri, University of Tadulako, Indonesia
Optimal Design of V-shaped Absorber Plate to the Performance of Solar Water Heater
24 Jalaluddin, Effendi Arief, Rustan Tarakka, Hairul Arsyad, Andi Mangkau, Labusab, University of
Hasanuddin, Indonesia 114-119
vii
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Thirdly, among UUVs, we develop a robotic manta as a kind of fish robot with pectoral
fins. Such a biomimetic thruster is expected to provide noiseless propulsion, and to be
more maneuverable in complex near-shore environments and highly efficient in energy
consumption, compared to the conventional AUVs with a propeller-based thruster.
K-01
.
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Abstract An examination has been made of the mechanical use CFRP seems to be thrown away into landfill without any
and failure properties of several composite materials, such as a consideration of environmental problems [3]. This occurrence
short and a long carbon fiber reinforced plastic (short- and will be a problem in the future, because the amount of waste
long-CFRP) and metal based composite material. The short CFRP will increase [4]. Up to date, several researchers have
CFRP materials were used for a recycled CFRP which investigated the mechanical properties of CFRP including
fabricated by the following process: the CFRP, consisting of recycled CFRP, the information available appears to be
epoxy resin with carbon fiber, is injected to a rectangular plate insufficient.
cavity after mixing with acrylonitrile butadiene styrene resin On the other hand, metal matrix composite material
with different weight fractions of CFRP. The fatigue and (FRM) is also important material as engineering material. This
ultimate tensile strength (UTS) increased with increasing is because of their outstanding mechanical properties. Metal
CFRP content. These correlations, however, break down, matrix composite with silicon carbide particle (SiC) are one of
especially for tensile strength, as the CFPR content becomes the widely known composites, which have high strength, high
more than 70%. Influence of sample temperature on the hardness, high wear resistance and high corrosion resistance
bending strength of the long-CFRP was investigated, and it [5]. Effect of clustering on mechanical properties of aluminum
appears that the strength slightly degreases with increasing the alloy 2024-SiC metal matrix composite has been investigated.
temperature, due to the weakness in the matrix. Broken fiber Fracture toughness and tensile tests were carried out, and their
and pull-out or debonding between the fiber and matrix were mechanical strengths were estimated well by a model [6].
related to the main failure of the short- and long-CFRP Although CFRP and FRM are excellent materials to use
samples. Mechanical properties of metal based composite in various engineering application, there would have still
materials have been also investigated, where fiber-like high technical issue for recycling technique and lack of information
hardness CuAl2 structure is formed in aluminum matrix. regarding their mechanical properties. In this study, our
Excellent mechanical properties were obtained in this alloy, experimental results obtained previously for the material
e.g., the higher strength and the higher ductility, compared to properties of long-CFRP, short-CFRP [7] and FRM [8]-[9]
the same alloy without the fiber-like structure. There are were summarized to consider the mechanical properties of the
strong anisotropic effects on the mechanical properties due to composite materials.
the fiber-like metal composite in a soft Al based matrix.
Index Terms CFRP; Carbon fiber; Mechanical property; II. EXPERIMETNAL PROCEDURE
Crack growth; Failure mechanism II-1. Long-CFRP and short-CFRP
The long-CFRP, consisting of epoxy resin (thermosetting
high polymer) with a volume fraction of 60% carbon fiber,
was used. Fig. 1 shows the photograph of the long-CFRP
I. INTRODUCTION samples showing the carbon fibers and matrix. The short-
In recent years, composite materials have received CFRP samples were made by the following process. The long-
special attention because of the excellent mechanical CFRP was first crushed using a rotating blade to make small
properties. In particular, production amount of carbon fiber fragments for which the average length by width is 3.4 mm
reinforced plastics (CFRP) have been increased due to their 0.4 mm. The crashed long-CFRP pieces were then separated
high strength and low specific weight [1]. CFRP material has individually into fiber and epoxy resin after the ball milling
come into practical use for the aerospace and automotive process. Most part of the surface of separated carbon fibers is
industries, because of their contribution to higher fuel not already coated by epoxy resin, while some fiber bundles
efficiency. In fact, the demand for CFRPs has dramatically were present that contained epoxy resin [3]. After the grinding
increased in recent years [2]. As aerospace and automotive process, it was found that the mean length of the carbon fibers
parts are sometimes employed in atmosphere with high is about 200 m. The short-CFRP samples, consisting of
temperature, examination of mechanical properties of CFRP at acrylonitrile butadiene styrene resin and CFRP pieces, were
high temperature would be required. In addition, development fabricated using standard mixing, grinding and injection
of the recycling technology for CFRP has been significantly molding procedures. In this case, the CFRP pieces were added
important due to their high production amount. Indeed, post- to the ABS resin before the injection process with five
1
International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
different weight fractions of 0 (i.e., pure ABS), 10, 30, 50 and gives a schematic diagram of the heated mould continuous
70 wt.%. The injection molding process was carried out to casting apparatus, consisting of a graphite crucible with runner,
make the short-CFRP with simple rectangular plates 150 mm a graphite mould, a cooling device and pinch rolls for
150 mm 3 mm. withdrawal of the cast metal. The cast samples in the shape of
Dumbbell-shaped specimen and compact tension (CT) a long round bar (4 mm 1 m) was made. The casting
specimen were used in this test, which obtained from the pressure was controlled by the level of molten metal in the
center area of the rectangular plate as shown in Fig. 2. In this crucible, controlled by furnace displacer block. The
case, the rectangular plate cut in two different directions from temperature of the molten metal was maintained at about
the mid-section, either with the loading direction (longitudinal 843K, which is 20K above the melting point of its Al alloy.
axis of the specimen) in the direction perpendicular (Type T) The molten metal was cast through the runner and graphite
or parallel (Type L) to the flow (or carbon fiber) direction. mould before the cooling process. The graphite mould was
The dimension of the parallel area in the dumbbell-shaped heated to approximately 853K, which is just above the
specimen is 7 mm (l) 3 mm (w) 1 mm (t), and that of CT liquidus of the Al-Cu alloy. For the solidification process, the
specimen is W = 24.5 mm and B = 3 mm. The CT specimen aluminum alloy was cooled directly by water flowing to the
was designed based upon the ASTM standard E399 [10]. In exit just out of the graphite mould (see Fig. 3). Interestingly,
the mid-section of the CT specimens, a through-slit (15 mm in with this casting process, a unidirectional growth
length with a V-notch root angle of 45 degrees) was machined. microstructure was created, which could be associated with
metal composite material. Fig. 4 depicts microstructure of Al-
33%Cu sample with the axial and transverse directions. The
primary -Al phase is visible as a dark region. A fine fiber-
like eutectic structure of CuAl2 phases with unidirectional
growth along its axial direction can be observed.
Fig. 5 displays the test specimens formed with a
rectangular shape. Note, in this case, tiny special specimen
was designed to examine the metal composite effect on the
mechanical properties, namely anisotropic microstructual
effects. The specimens are denoted as (i) axial direction (OL)
and (ii) transverse direction (OT), as indicated in Fig. 5.
Fig. 1 Photograph of long-CFRP plate.
(a) Type T
Flow
direction
(b) Type L
Fig. 2 Schematic diagrams showing the test specimens in Fig. 3 Schematic diagrams of the heated mould continuous
the short-CFRP samples [7]. casting system and cooling system [8].
II-2. Metal composite aluminum alloy Because of the tiny specimen, finite element analysis was
In the present study, an attempt was made to create FRM conducted to verify the stress-strain distribution before the
materials via our heated mould continuous casting technology testing. Fig. 5 also indicates the FEA stress distribution on the
(HMC) with a eutectic aluminum alloy. Concept of this loading direction (x-axis). From this result, it is clear that the
technology is as follows: unidirectional microstructure with high stress level is uniformly distributed in the sample of
thin fiber-like phases was created by the unidirectional rapid parallel area. Thus, the material properties can be estimated to
solidification process. In this approach, An Al-33%Cu eutectic understand their material characteristics
alloy was selected to make metal composite Al alloy. Fig. 3
2
International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
0 45
4 mm
90
III. RESULTS
III-1. Long-CFRP materials
Fig. 6 shows representative bending stress - strain
curves for the long-CFRP with different fiber direction. As
seen, different tensile properties are obtained depending on
the fiber direction. It is clear that low bending properties are
obtained as the CFRP with fiber direction of more than 45
against the loading direction, while high mechanical Fig. 7 SEM images showing the fracture surface of the specimen.
properties are detected for the CFRP with 0 fiber direction.
Fig. 7 depicts the fracture surfaces of their specimens after
the bending tests. As seen, fiber surfaces are observed for
the specimens with fiber direction of 45 and 90. Those
3
International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
100C
50C
Vacancy
4
International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
crack growth rates. For instance, a rough fracture surface obviously high in the OL sample compared to the other one
makes low crack growth rate due to the low crack driving due to the fiber-like composite effect. The tensile properties of
force arising from severe crack closure [11]. both Al-33%Cu alloys for OL and OT are UTS = 489.8 MPa
To understand clearly the fatigue behavior of the and f = 4.9% and UTS = 384.1 MPa and f = 4.2%,
short-CFRP samples, the S-N relationships were respectively. Such different tensile properties (OL vs. OT) are
quantitatively evaluated by a power law dependence of reflected to the reinforcement by the fibrous structure. The
cyclic stresses and cycles to failure: fine eutectic structure in the longitudinal direction can
a f Nf-b , MPa (1) enhance the material properties (OL), and that tensile strength
where a is the stress amplitude, Nf represents the cycle is much higher than that for the conventional cast Al alloys.
number to final fracture, f is the fatigue strength On the other hand, the high material ductility obtained for the
coefficient and b is the fatigue exponent. Those values (f OL samples can be explained using the failure mechanism.
and b) were obtained by least square analysis. In this case,
an increased fatigue life is expected for a decreasing fatigue 600
strength exponent b and increasing fatigue strength
200
100
104 cycles 0
0 2 4 6 8
Strain, %
(a) OL (b) OT
5
International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
ACKNOWLEDGMENTS
The author would like to acknowledge many useful
advices and suggestions of Prof. K. Ogi. The author also
appreciates technical support of Mr. K. Ota, Mr. T. Yamazaki,
Mr. R. Sato and H. Iwai.
6
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract Turning process is the removal of metal from the Nowadays, a standard procedure used to avoid vibration
outer diameter of a rotating cylindrical workpiece. Turning is during turning is planning the selection of cutting speed, feed
used to reduce the diameter of the workpiece, usually to a rate, and depth of cut carefully. A method applied is usually
specified dimension, and to produce a smooth finish on the metal. based on an operators experience and also trial and error
This research investigates the effect of feed rate, spindle speed,
method to gain a proper cutting parameters in the machining
tool overhang and tool dimensions toward vibration amplitude
and surface roughness on turning process. This study uses both process.
statistical and graphical analysis of the data collected. The Vibration in machining process occurs throughout the
experimentation was carried out on conventional lathe machine cutting process takes place which is derived from some
with straight turning operation. Material used as workpiece was sources, such as frame structure of the machine, cutting tool
St.60 carbon steel which was turned with HSS tool bit with the type, types of material that are cutting, etc. Vibration on
dimension of 3/8 Inches and Inches. Cutting parameters varied machining process is very complicated because it involves lots
by spindle speed, feed rate, and tool overhang, while the depth of of variables. Nevertheless, at least two kinds of vibrations
cut is maintained at a depth of 0.5 mm. The vibration data of occurred on machining process, this covers both force
cutting tool obtained from a transducer (vibrometer) mounted at
a distance of 10 mm from the tip of the cutting tool during the
vibration and self-excited Vibration. Force vibration is usually
cutting process takes place, whereas the surface roughness data gained from components in the machine itself, for example
obtained from measurements of surface roughness apparatus because there are damaged gear components, imbalances on
after turning process. The results showed that, The effect of feed machine components, misalignment of the shaft, the electrical
rate, spindle speed, tool overhang, and tool dimension motor rotation, and etc. Self-excited vibration which is so-
simultaneously towards vibration amplitude and surface called Chatter caused by the interaction between the release of
roughness has a grater effects on the use of 3/8 inches cutting tool chips and cutting tool which causes interference with the
than inches cutting tool. With the use of the same tool cutting area. Chatter or self-excited always affects on the
dimensions obtained that, The most influential parameters on the
surface roughness of machining product. Therefore a vibration
vibration amplitude is tool overhang while the most influential
parameter on surface roughness value is feed rate.
which is caused by self excited vibration related to the surface
roughness as the result of machining [2]
Key words Turning, vibration, surface roughness, cutting Some previous research, both descriptive and experiment
tool have studied how vibration affects surface roughness towards
surface roughness as a product of machining process
I. INTRODUCTION [1][2][3][4][5]. One of those researches is the effect of spindle
Challenges faced by today's modern machining industry speed, feed rate, and depth of cut towards tool vibration
primarily focused on achieving a high quality product. One of amplitude and surface roughness of the workpiece on the lathe
them are the quality of surface roughness. Surface roughness of machine, in which concludes that spindle speed is the most
a product of machining process can affect some functions of influential towards vibration and so does the surface roughness,
these products such as surface friction, heat transfer, spreading then followed by the feed rate and the last is the depth of cut.
capabilities of lubrication, coating, and others. Thus, in Based on that research, the writer conducted a further research
practical field,the desired of surface roughnes value will be the towards some variables which has not been experimented
reference of cutting parameters selection [1]. Turning is one of before, it is the effect of tool overhang and cutting tool
the main machining processes used in the process of cutting a dimension in which involving the same variation of cutting
rotating cylindrical workpiece. Lots of machinery components conditions that that have been studied before.
made through turning process. Problems are often encountered
in all of the machining process especially on turning, is the
II. THEORITICAL BACKGROUND
vibration during the material cutting process. This vibration
will affect the quality of the products, one of them is the A. Vibration in metal cutting process
surface roughness [1].
Vibration is a back and forth motion about its fixed
equilibrium position. The equilibrium position means a
7
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
condition in which an object is on motionless position if there C. General overview of turning process
is no force acting on the object. Vibration has similar amplitude Turning process is the removal of metal from the outer
(distance deviation furthest to the midpoint) [6]. All objects diameter of a rotating cylindrical workpiece. Turning is used to
that have mass and elasticity are able to vibrate. reduce the diameter of the workpiece, usually to a specified
In the machining process, there are three mechanical dimension, and to produce a smooth finish on the metal.
vibrations caused by insufficiency of dynamic stiffness in Three main parameters in turning operation are cutting
machinery equiptment. The vibrations are free vibration, force speed, feed rate, and depth of cut. Other factors such as
vibration, and self-excited vibration. A tool holder, workpiece workpiece material and type of cutting tool actually has a
and the machine itself are parts of machinery which causes considerable influence. However, the three parameters
vibration. The free vibration is usually caused by shocks effect, mentioned above are parts that can be set by the operator
such as, the presence of impulse waves that are transferred to directly on the lathe machine.
the cutting tool or at the time between the beginning of the Cutting speed (also called surface speed or simply speed)
cutting tool with the workpiece. Force vibration is caused by may be defined as the rate (or speed) that the material moves
periodic force which occurred in the system, for example due past the cutting edge of the tool, irrespective of the machining
to the imbalance of machine components for instance, gears operation used. The equation of cutting speed can be determine
system, spindle or bearing. Self-excited vibration usually from the equation below
occurs as a result of dynamic instability that occurs during
metal cutting processes. As it shows, the self-excited vibration
is the most uncontrolled vibration while two other vibrations (1)
can be controlled through arranging cutting parameters on the
machine [7].
Where Cs = cutting speed; m/minute, D = workpiece
B. Surface roughness diameter /mm, N= spindle speed, revolution / minute. The
Surface roughness is a measurable characteristic based on spindle speed (N) in eq.(1) is a measure of the frequency of a
the roughness deviations as defined in the preceding. Surface rotation. It annotates the number of turns completed in one
finish is a more subjective term denoting smoothness and minute around a fixed axis. The preferred speed is determined
general quality of a surface. In popular usage,surface finish is by working backward from the desired surface speed (sfm or
often used as a synonym for surface roughness. m/min) and incorporating the diameter (of workpiece or cutter).
The most commonly used measure of surface texture is Feed rate, Vf , refers to how fast a lathe-tool should move
surface roughness. Withrespect to Figure 1, surface roughness through the material being cut. This is calculated using the
can be defined as the average of the vertical deviations from Feed per Revolution for the particular material. It is expressed
the nominal surface over a specified surface length. An in units of distance per revolution. Feed rate is determined
arithmetic average (AA) is generally used, based on the based on machine power, material properties of workpiece, tool
absolute values of the deviations, and this roughness value is material, tool shape, and the most important is the expected
referred to by the name average roughness. In equation form surface roughness.
Depth of Cut, the thickness of the material that is removed
by one pass of the cutting tool over the workpiece. [8]
(1)
D. Regression Analysis
Regression Analysis is applied to study and measure the
where Ra = arithmetic mean value of roughness, m (in); y = the scatistical relationship among two or more variables. In simple
vertical deviation from nominal surface (converted to absolute regression analysis two variables are analyzed, whereas in the
value), m(in); and Lm = the specified distance over which the multiple regression analysis more than two variables are
surface deviations are measured. analyzed. In regression analysis, a regression equation was
The AA method is the most widely used averaging method about to set and used to describe a pattern or a function of the
for surface roughness today. Analternative, sometimes used in relationship between variables. Variables to be called the
the United States, is the root-mean-square (RMS) average, dependent variable is usually plotted on y-axis. While the
which is the square root of the mean of the squared deviations independent variable is the variable that is assumed to give
over the measuring length. RMS surface roughness values will effect to the variation in the dependent variable and it is usually
almost always be greater than the AA values because the larger plotted on x-axis.
deviations will figure more prominently in the calculation of Multiple linear regression, On this multiple linear
the RMS value. regression, there are several independent variables
(X1,X2,X3...Xn) which are connected to one dependent variable
(Y), those are parts of multivariate analysis to estimate the
regression coefficient to describe the effect of independent
variable towards dependent variable. In a multiple regression
test, all of the predictor variables are included in the regression
calculation simultaneously.
8
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
If there are two independent variables (X1) and (X2) and the Hypothesis test of multiple correlation using F- test with
dependent variable (Y), then the coefficients of the multiple degrees of freedom (df) consists of:
regression equation is determined by the following equation: df1 = df numerator = k (k = total of independent variable)
df2 = df denominator = n k 1 (n = numbers of pairs of
Y = a + b1X1 + b2X2 +... bnXn (2) data or sample)
E. Correlation Analysis Value conversion of correlation coefficient R to Fcount uses
Analysis of correlation is an inferential analysis used to the following equation :
determine the degrees of freedom or strength of the
relationship, shape or causal and correlation among research R2 / k
Fb (5)
variables. Type of statistical hypothesis testing correlations (1 R 2 ) /( n k 1)
includes simple correlations (bivariate), multiple correlation
and partial correlation. Hypothesis testing criteria, namely:
Pearsons Product Moment Correlation, this correlation is Accept H0 if Fcount< Ftable, dan reject H0 if Fcount > Ftable
used for interval/ratio data in which must meet the following
requirements:
The sample is taken randomly III. RESEARCH METHODOLOGY
Each variable of data is normally-distributed
A linear regression equation A. Experimental set up
Equation : Material used as workpiece was St.60 carbon steel which
was turned with HSS tool bit (The Bohler Super Mo Rapid
n. xy x y Extra 1200 Brand) with the dimension of 3/8 Inches and
rxy (3) Inches. The cutting tool angles used were Side Relief = 11o,
[n x ( x) 2 ][n y 2 ( y)2 ]
2
Front Relief = 8 o, Side Rake = 12 o, Back Rake = 8 o [10]. The
experimentation was carried out on conventional lathe machine
Coefficient of Determination, The Coefficient of and the method of cutting is shown in Fig. 1, referring to the
Determination is denoted as r2. This value states the proportion experimental set up that has been conducted by previous
of the overall variation in the value of the dependent variable researchers [11].
that can be explained or caused by a linear relationship with the Figure 2, shows that workpiece which will be turned then
independent variables, the rest is explained by other variables divided into four segments separated by grooves. The purpose
(errors or other variables). Coefficient of determination of this segmental separation is to minimize the affects of tool
expressed as the square of the correlation coefficient wear which can effects surface quality towards the
r2 x 100% = n% meaning that the value of the dependent effectiveness of measurement. Thus, in collected data
variable can be explained by the independent variables of n%, measurement, tool bit cuts four times before being substituted
while the residual value of (100 - n)% explained by an error by a sharpened cutting tool.
(error) or the influence of other variables.. Meanwhile, for
correlation analysis with more dependent variables, there is a
correlation coefficient which is significantly sensitive with
amount of variables. Usually for multiple correlation analysis,
adjustment coefficients of determination are often used.
Multiple Correlation, multiple correlation is the
correlation between two or more independent variables
together with the dependent variable. Value which shows
directions and strength of the relationship between two or more
independent variables on the dependent variable is called
multiple correlations and denoted as R.
The Equation of multiple correlation of two independent
variable X1 and X2 with one dependent variable (Y) as [9] :
Figure 2. Schematic of experimental set-up for turning
r r 2ry1 .ry 2 .r12
2 2
R y.12
y1 y2
(4)
1 r122 B. Method of Collecting data
Measurement of datas is undertaken by a well-trained and
Where Ry.12 = coefficient of multiple correlation experienced lathe operator in the using of vibration
among X1, X2 and Y, ry1 = correlation coefficient between X1 measurement instrument and surface roughness devices.
and Y, ry2 = correlation coefficient between X2 and Y, ry12 = Collecting of measurement datas was arranged as factorial
correlation coefficient between X1 with X2 designed so that the interactions between independent variables
The tested hypothesis is two tailed test: can be observed more effectively. The independent variables in
H0 : y.12 = 0 the study are feed rate(Vf), spindle speed, tool overhang, and
H1 : y.12 0 the tool dimension. While dependent variables is the result of
9
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
cutting tool vibration (Vrms) and the surface roughness (Ra) of obtained the highest vibration on cutting tool, that is
workpiece. The details of cutting condition and the groups of Vrms= 0.65 cms/s while the vibration of 3/8 inches cutting tool
experimental testing are shown in Table 1. is Vrms= 0.78 cms/s at spindle speed 880 rpm, feeding 0.24
mm/rev, and cutting tool overhang 40 mm.
Table 1 Groups of Experimental testing variables
A. Result
The result of cutting tool vibration towards feed rate and
variation of spindle speed and tool overhang can be seen in
Figure 3, where it appears that the use of inches cutting tool
at the same length of tool overhang, addition of feed rate on
the turning process will increase cutting tool vibration, where
the highest vibration value occurs in 0,24 mm/rev feed rate. Figure 4. Comparison plot between vibration and feed rate on variation
Moreover, conditions where feed rate is maintained at a of spindle speed and tool overhang by using 3/8 inches
fixed speed and spindle speed varied at different value, the tool cutting tool dimension
vibration will be higher if spindle speed is increased. However,
The effects of variation on feed rate towards surface
the magnitude of vibration is smaller with a vibration caused
roughness using inches cutting tool as shown in Figure 5,
by variations in the parameters of feeding. On the use of 3/8
indicates that increasing the feed rate will increase surface
inches cutting tool as shown in Figure 4 , vibration value is
roughness on workpiece where the highest value obtained on
higher compared to vibration value generated by the use of
feeding 0.24 mm/rev. On the variation of cutting tool overhang,
inches cutting tool. On the use of inches cutting tool
10
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
the longer the tool bit the higher the surface roughness value regression equation of each variables as shown in Table 2 and
where 40 mm tool overhang has the highest roughness value. Table 3.
By using 3/8 inches tool bit as shown in Figure 6, the
surface roughness value will be greater than the roughness Table 2. The result of correlation statistical analysis and tool vibration
values obtained in the use of inches in every similar cutting regression dan incheses and 3/8 inches.
parameter condition used. On the use of inches tool bit the
highest surface roughness value is Ra= 8.54 m while on 3/8
inches tool bit reaches Ra= 10.34 m at a spindle speed of 170
rpm, feeding 0.24 mm/rev, and tool overhang 40 mm.
B. Discussion
This study shows that there is an effect of feed rate, spindle
speed, tool overhang, and tool dimension toward vibration
amplitude and surface roughness on turning process where the
most influential parameter towards vibration amplitude is tool
Figure 5. Comparison plot between surface roughness and feed rate on overhang, while the most affective parameter towards surface
variation of spindle speed and tool overhang by using 1/2 roughness is feed rate.
inches cutting tool dimension
The experimental result of the effect of independent
variables ( spindle speed, feed rate, tool overhang) towards
vibration amplitude found that, the addition of feed rate on
turning process will increase the tool vibration because the
more feeding given the faster cutting tool cut of the workpiece,
as the result, the frictional forces that occur will be greater due
to the magnitude of compressive force on the tip of the cutting
tool and workpiece. Moreover, a condition where feeding
keeps being constant and the spindle speed varied on different
values resulting the increasing of tool vibration if spindle speed
is increased. However, vibration obtained will not be as higher
as vibration by feeding parameter variation. This vibration is
caused by radial cutting force that occurs as a result of
interraction between the tip of cutting tool and the rotating
workpiece.
The effect of vibration on the cutting tool overhang
variation shows that, the more the tools overhang from the tool
holder, the greater the vibration results. The length of cutting
tool overhang contributes to the deflection which is caused by
cutting force that lead to vibration on cutting tool. At the same
conditions of cutting parameters, the vibration that occurs in
Figure 6. Comparison plot between surface roughness and feed rate on the use of inches tool bit larger than 3/8 inches. This is
variation of spindle speed and tool overhang by using 3/8 caused by the stiffness of the inch tool bit larger than 3/8
inches cutting tool dimension
inches. Material stiffness is determined by volume and
elasticity modulus of material, while the magnitude of the
To determine a detailed characteristics effect between
deflection is inversely proportional to the value of the rigidity,
dependent and independent variables, the data processing is
while the vibration tends to be even greater if the value of
done with statistical methods to determine the correlation and
deflection increases [14].
11
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Statistically, the coefficient of determination (R2y.123 ) for From the results of statistical data processing, for inches
inches cutting tool is 0.9442. This shows that, 94.42% of the tool bit, shows that 95.1% of the variation in surface roughness
variation of the amplitude of vibration is the result of the value is a result of the influence of variable spindle speed, feed
influence of spindle speed, feed rate and tool overhang rate and cutting tool overhang simultaneously. While the
simultaneously. While the rest 5,6% is caused by other factors. remained of 4.9% is the result of other factors. On the use of
Because the R-square value close to 1, then this indicates the 3/8 inches tool bit found that the effect is greater than inches
strength of the relationship between the independent and usage ie 95.9%. From the regression equation generated by the
dependent variables in this case. On the use of 3/8 inches use of both types of cutting tool seen that the coefficient of X 1,
cutting tool shows that R square is 0,963. This clearly reveals X2 and X3 on the use of 3/8 inches tool bit is greater than the
that the effect of spindle speed, feed rate, and tool overhang coefficient of the independent variable on the use of 1/2 inches
variables all at once toward tool vibration amplitude is greater tool bit, in other words, the response to the independent
than the use of inches tool. On the regression equation of variable of vibration amplitude at the use of 3/8 inch tool bit is
turning vibration amplitude on inches tool, the Constants of - much greater than the use of inches tool bit.
0.3844 states that, if the independent variables like spindle The most influential independent variable towards surface
speed, feed rate and tool overhang equals to zero, then the roughness value based on correlation analysis test among
vibration amplitude is -0.3844 m. Regression coefficient (X1) variables, both for inches and 3/8 inches tools are feed rate ,
of 1.351.10-4 states that, the addition of spindle speed of 1 rpm then spindle speed, and the last is tool overhang. Minus sign on
will decrease the vibration amplitude of -1.351.10-4 cm/s. the correlation coefficient on the influence of spindle speed
Regression coefficient (X2) of 0.015488 states that, every states that, the greater the spindle speed is, the smaller the
1mm addition of tool overhang, will increase vibration value of surface roughness (Ra) to be generated. where the
amplitude 0.015488 cm/s. Regression coefficient (X3) of smaller the value of Ra, the smoother is the surface [15].
1.14454 states that, addition of every 1 mm/rev feed rates will
increase vibration amplitude 1.14454 cm/s. From the two C. Conclusion
equations of regression above, it can be seen that coefficient of The effect of feed rate, spindle speed, tool overhang, and
X1, X2 and X3 on the use of 3/8 inches cutting tool is higher tool dimension simultaneously towards vibration amplitude by
than independent variables coefficient on inches, on the using inches tool bit is 94.4 %, while the use of 3/8 inches
other words, response of independent variable towards tool bit is 96.3%. The effect of feed rate, spindle speed, tool
vibration amplitude for the use of 3/8 inches tool is greater than overhang and tool dimension all at once towards surface
the use of inches tool. roughness by using inches tool bit is 95.1%, while the use of
To determine the most influential independent variables on 3/8 inches tool bit is 95.9%. Amplitude correlation of tool
the magnitude of the vibration amplitude, the correlation vibration by using inches tool towards tool overhang
analysis tests is conducted. From the calculation of the 69.68%, towards feed rate 61.82% and spindle speed 27.67%.
correlation coefficient on inches tool shows that the most On the use of 3/8 inches tool bit, amplitude correlation of tool
influential factor on lathe vibration amplitude is tool overhang vibration towards tool overhang 71.8% towards feed rate
with correlation value 0.6968 at 99% convidence level, the next 60.2% , spindle speed 29.3%. Surface roughness correlation by
is feed rate 0.6182 and the last is spindle speed 0.2767. using 1/2 inches tool bit towards feed rate 73.0% , spindle
In addition, correlation analysis on the use of 3/8 inches speed -55.2% , and tool overhang 33.6%. On the use of 3/8
cutting tool, shows that the most influential factor on the tool inches tool bit, correlation of surface roughness towards feed
vibration amplitude is tool overhang on the correlation value of rate 80.9%, spindle speed -42.0% and tool overhang 38.5%.
0,718, then the feed rate 0,602 and the last is spindle speed The most influential parameters on the vibration amplitude is
which reaches 0.293. as all correlation coefficients are positive, tool overhang while the most influential parameter on surface
it means that, by increasing the amount of spindle speed, tool roughness value is feed rate.
overhang, and feed rate will increase tool vibration amplitude.
Nevertheless, value of correlation coefficient on the use of tool
is 3/8 inches which has a higher correlation value than the use REFERENCES
of inches for its all independent variables. [1] Lazuardhy Muchammad T, Endi S, Erwin S. 2012, pengaruh
The test result of independent variable effect (spindle feed motion kondisi chatter terhadap kekasaran permukaan
speed, feed rate, and tool overhang) towards roughness value benda kerja proses bubut, Jurusan Teknik Mesin Fakultas
found that, addition of feed rate on turning process will Teknik Universitas Brawijaya.
increase the value of surface roughness, this is due to the [2] Huang Luke, A Multiple Regression Model to PredictIn-
construction of the pointed end ( very small nose radius) of the process Surface Roughness in Turning Operation Via
cutting tool, thus, if feeding is slowed down then the distance Accelerometer , Journal of Industrial Technology, Vol. 17 No.2
April 2001.
between the grooves cut by a cutting tool tip at one revolution
of workpiece will be even greater. This distance will stimulate [3] M. Siddhp aco anTi, Duan kovaeviura 2011, Wear level
influence on chip segmentation and vibrations of the cutting tool
such serration effect which is distantly spaced and if measured
Tesis, RMZ Materials and Geoenvironment, Vol. 58, No. 1,
with surface tester, so, the distance among these serrations is a pp. 1528, 2011
representation of surface roughness of a turning object. On the
[4] Koten Viktus K. (2006). Analisis Pengaruh Kondisi
variation of tool overhang and tool dimension is seen that Pemotongan Pada Mesin Bubut Terhadap Amplitudo Getaran
surface roughness depends on vibration, thus, the more Pahat Dan Kekasaran Permukaan Benda Kerja, Tesis. Pasca
vibrations the roughness value is greater. Sarjana Universitas Hasanuddin Makassar.
12
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
13
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Hendry Sakke Tira, Yesung Allo Padang, Mirmanto and Rio Cristovan Mantiri
Mechanical engineering, Mataram University
Mataram, Indonesia
hendrytira@yahoo.com, yesungallopadang@gmail.com, mmirmanto@gmail.com
AbstractEnergy supply is a crucial issue in the world in the energy promotion scheme as well as an alternative for
last few years. The increase in energy demand caused by reduction of greenhouse gases emissions.
population growth and resource depletion of world oil reserves Biogas, a clean and renewable from of energy can be a
provides determination to produce and to use renewable
good substitution of conventional sources of energy which are
energies. One of the them is biogas. However, until now the use of
biogas has not yet been maximized because of its poor purity.
causing ecological-environmental problems and at the time
According to the above problem, the research has been carried depleting at a faster rate [3, 4]. Biogas is the combustible gas
out using the method of water absorption. Under this method it is produced through an anaerobic digestion at low-temperature
expected that the rural community is able to apply it. Therefore, and without oxygen. Thus it application includes electricity,
their economy and productivity can be increased. This study heating and cooking. On the other hand, there is lack of good
includes variations of absorbing water volume (V) and input management of the ever-increasing amounts of manure solid
biogas volume flow rate (Q). Raw biogas which is flowed into the and liquid waste in many communities. Most of the rural
absorbent will be analyzed according to the determined communities discharged the manure without treatment directly
absorbing water volume and input biogas volume rate.
onto wasteland or into rivers and streams. This behavior leads
Improvement on biogas composition through the biogas
purification method was obtained. The level of CO2 and H2S was
to unhygienic environment with attendant bad odors and flies
reduced significantly specifically in the early minutes of [5]. With appropriate treatment, the manure can be converted
purification process. On the other hand, the level of CH 4 was into biogas. Biogas is defined as mixture of gases, consisting of
increased improving the quality of raw biogas. However, by the methane (CH4), carbon dioxide (CO2), hydrogen sulfide (H2S)
time of biogas purification the composition of purified biogas was and traces of other gases like nitrogen (N2), oxygen (O2),
nearly similar to the raw biogas. The main reason for this result hydrogen (H2) and ammonia (NH3). The composition of biogas
was an increasing in pH of absorbent. It was shown that higher depends on the organic material as well as on the conversion
water volume and slower biogas volume rate obtained better technology used, varying between 50-75% CH4, 25-45% CO2,
results in reducing the CO2 and H2S and increasing CH4 and 0-20 000 ppm of H2S [6].
compared to those of lower water volume and higher biogas
volume rate respectively. The purification method has a good From the constituents of biogas, CH4 and CO2 are the main
promising in improving the quality of raw biogas and has compounds in determining the quality of biogas. If the level of
advantages as it is cheap and easy to be operated. CH4 is high, the biogas will has higher calorific value. On the
other hand, if the level of CO2 is high, the quality of biogas will
KeywordsBiogas, CH4, CO2, H2S, water absorption be worse, marked by lower calorific value. Therefore, to
improve the calorific value of biogas in order to be used
I. INTRODUCTION effectively as fuel, the level of CO2 should be reduced or
eliminated [6]. On the other hand, H2S, a kind of highly toxic
In the last few years the continuously uninterrupted supply
and corrosive gas, inhibits the biogas process directly, as well
of energy has been a crucial problem in Indonesia. The
as indirectly in the case of higher H2S concentrations in
increase of energy demand which is caused by population
digester. To avoid the negative effects of H2S, a reduction of
growth and acceleration in industries has pressured the
H2S concentration in biogas is required before combustion [6,
government to explore much new and alternative energy to
7].
maintain the development. One of the alternative energy is
Removal of CO2 and H2S from biogas is the main factor
biogas. Biogas is a promising energy among other alternative
to improve the biogas quality [8]. To pursue this aim, a
fuels as it is renewable with abundant feedstock and can be
purification method is required to treat raw biogas. Some
produced in rural area with relatively low operational cost [1,
biogas purification methods have been performed, and water
2]. Therefore, biogas can be the solution for this renewable
scrubbing method can be a solution as it is a simple and cheap
14
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
15
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
This percentage is considered relatively low for heating purified biogas. However, the dissolved rate of methane is
and as a fuel purposes. The experiment results show that the higher compared to those of carbon dioxide and hydrogen
level of methane before purification is 59.36%. However, after sulfide. Methane solubility rate in water at 10 oC is 42 mg/liter
the purification process, the level of CH4 was increased. of water. The solubility of methane in water decreases along
Results show that the concentration of CH4 was improved to with the increase of water temperature [11].
60.4% where reached in the operating condition of 25 liter of The results show that in the early minutes of data
water volume and 1 lt/min for biogas volumetric flow rate collection, from 10th to 18th minute, the methane concentration
respectively. Additionally, the lowest improvement for CH 4, it is relatively low. The air concentration in the top of scrubbing
was of 59.36%, was gained in operating condition of 15 liter unit is still exist at the beginning of measurement. Therefore,
of water volume and 3 lt/min for biogas volumetric flow rate from early minutes to 10th minute, most of the gas measured
respectively (Figure 3). comes from air.
It was shown that the higher water volume, the higher B. Carbon dioxide
content of methane in biogas (Figure 4). It is resulted by the Carbon dioxide is an impurities gas which has relatively
longer contact time occurred between water and biogas high percentage in biogas. Carbon dioxide is an unburned gas
molecule in higher water volume operating condition. The and therefore resulted in the calorific value of biogas reduce
contact between molecules leads to the absorption of the significantly. The more CO2 in biogas, the less calorific value
impurities substances in biogas into water molecule. Under of biogas is. In order to improve the biogas calorific value,
higher water volume, the rate of reaching the higher methane reducing the content of CO2 through biogas purification is a
concentration is faster compared to that of lower water volume must. Carbon dioxide is a water-soluble gas. At room
(Figure 5). temperature, the solubility of carbon dioxide is about 90 cm3 of
CO2 per 100 ml water (Shakashiri, 2014). The application of
water scrubbing method is expected to improve the quality of
biogas by reducing carbon dioxide content.
The research showed that the most effective condition to
reduce CO2 was gained at 25 liter of water volume and 1 lt/min
of biogas volumetric flow rate. While 15 liter of water volume
and 3 lt/min of biogas volumetric flow rate showed the least
effective to improve biogas quality as shown in Figure 6 and
Figure 7.
16
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
17
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
operational cost, durable and easy to be operated. This allows [5] E. D. Aklaku, K. Jones, K. Obiri-Danso, Integrated biological
rural communities to apply this method for gaining their own treatment and biogas production in a small-scale slaughterhouse
energy resources. The research has found that the most in rural Ghana, Water Environment Research, vol. 12, pp.
effective condition to achieve the best results is running the 2335, 2006.
experiment at slower biogas volumetric flow rate and higher [6] H. Naegele, J. Lindner, W. Merkle, A. Lemmer, T. Jungbluth, C.
water volume. At this operating condition the CO2 can be Bogenrleder, Effects of temperature, pH and O2 on the removal
improved up to 10.5%, and H2S up to 27.5%. A further of hydrogen sulfide from biogas by external biological
research is required to make this method more promising and desulfurization in a full scale fixed-bed trickling bloreactor
efficient as the reduction of CO2 and H2S is occurred at the (FBTB), Int Agric & Biol Eng, vol. 6, no. 1, pp. 69, 2013.
early minutes of measurement. Maintaining the pH of [7] W. Barhost, L. Gupta, Benefits of digester gas scrubbing at the
absorbent and eliminating moisture content of purified biogas Dayton WWTP, Ohio, USA, Water Environment Association,
are also a challenge in order to obtain high quality of biogas. 2011.
ACKNOWLEDGMENT [8] D. Shannon, H. Kalipcilar, L Yilmaz, Development of zeolite
filled polycarbonate mixed matrix gas separation membranes,
The authors would like to thank the Department of National department of Chemical Engineering, Middle East Technical
Education of the Republic of Indonesia for supporting this University Ankara, Turkey, 2006.
project under MP3EI research grant with a contract number: [9] A. Dubey, Water scrubbing for carbon dioxide removal from
DIPA-023.04.1.673453, 5 December 2013, DIPA Revision 01
biogas, Annual report of central institute of agricultural
29 April 2014 contract number:
engineering, Bhopal, India, 2000.
012/SP2H/PL/DIT.LITABMAS/V/2014, 05 Mei 2014.
[10] B. Richards, F. Jewell, W. Cummings, R. White, In situ
REFERENCES methane enrichment in methanogenic energy crop digesters,
[1] L. Jian, Socioeconomic Barriers to biogas development in rural Biomass and Bioenergy, vol. 6, no. 4, pp. 275274, 1994.
southwest China: an ethnographic case study, Human [11] M. McGowan, Water Processing. Third Edition, Water Quality
Organization, vol 68, No. 4, 2009 Association. Water Technology Volume 32. International
[2] A. Weiss, V. Jerome, D. Burghardt, L. Likke, S. Peiffer, E. Occupational Safety and Health Centre; University of
Hofsetter, R. Gabler, R. Freitag, Investigation of factors Wisconsin, 2009.
influencing bogas production in a large-scaletThermophilic [12] R. Robert, P. John, P. Brice, The Properties Of Gases and
municipal Biogas Plant, Appl Microbiol Bioethanol, vol 84, pp. Liquids. 4 ed. Boston: McGraw-Hill, 1987.
987-1001, 2009. [13] E. Kovacs, R. Wirth, G. Maroti, Z. Rakhely, K. Kovacs, Biogas
[3] T. Kaosol, N. Sohgrahok, Enhancement of biogas production production from protein-rich biomass: fed-batch anaerobic
potential for anaerobic co-digestion of wastewater using fermentation of casein and of pig blood and associated changes
decanter cake, American Journal of Agricultural and Biological in microbial community composition, PLoS ONE, vol. 8, no.
Sciences, vol. 7, no. 4, pp. 494-502, 2012. 10, 2013
[4] Y. Santosh, T. Sreekrishnan, S. Kohli, V. Rana, Enhancement [14] Y. Lisafitri, Penggunaan Biotrickling Filter Biotrickling Untuk
of biogas production from solid substrates using different Mengatasi Polutan H2S. From http: //
techniques-a review, Bioresources Technol, vol. 95, pp. 1-10. www.academia.edu/3881807/
DOI: 10.1016/j.biortech.2004.02.010. Penggunaan_Biotrickling_Filter_Biotrickling_Untuk_Mengatasi
_Polutan_H2s. Downloaded at 18 August 2014.
18
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
AbstractCharacteristics of aluminium matrix composites The drawback is this material has lower strength than other
reinforced by alumina have been developed to improve commercial material such as cast iron, steel or copper.
mechanical properties. One of the determining factors in the However the strength of aluminium can be increased through
development of this material is parameter of solution treatment alloying, cold working and heat treatment through precipitation
process. This study discusses the performance of the composite
hardening process.
matrix of Al-9Zn-6Mg-3Si reinforced by alumina powder of 5 %
volume fraction. Composite are manufactured by squeeze
casting process with the pressure of 20 Ton in the metal mould. One of the aluminium alloys is Al-Zn, which exhibits the
To improve mechanical properties, the precipitation hardening highest strength and in many cases they have higher strength
process is conducted through variation of temperature of than steel. This alloys is the combination of zinc and
solution treatment of 450, 475 and 500 C and holding time of magnesium that produce the AA7xxx alloy which can be
solution treatment of 30, 60 and 90 minutes. Materials are heated and formed a very high strength. For example, AA 7075
characterized by hardness testing and microstructure with composition of: Zn 5.0-6.0%, Mg 2.0-3.0%, Cu 1.0% -
observation. The results showed that the optimum condition of
2.0% gives specific tensile strength of 580 MPa [3]. Zn
hardness was produced by solution treatment temperature of
500 C and 90 minutes holding time of 86 HRB. element will increase hardness optimally, after the heat
treatment precipitation hardening process [4], The increase in
Keywordsaluminium matrix composite, alumina, squeeze Zn element up to 9% can increase hardness particularly after
casting, hardness precipitation hardening, however Zn content in present study
was limited to 9% because the higher the content, the
possibility for hot cracking to occur is increase [5]. Based on
I. INTRODUCTION this, in order to increase the hardness, Zn and Mg alloy
elements added. Then heat treatment is conducted by
Various engineering components in manufacturing
precipitation hardening process on the composite to optimize
technology need of material with good mechanical
the hardness. This research studies the influence of
characteristics. Many innovations have been made to create a
precipitation hardening parameter such as solution treatment
new kind of lightweight materials with excellent mechanical
temperature and holding time towards hardness of the
properties. Composite materials can be used as one of the
aluminium matrix composite.
alternatives to answer this challenge. Composite material can
combine the excellent properties of consisting constituents to
produce a new material with better characteristic, which offer II. EXPERIMENTAL METHODS
several properties excellence such as high strength and
stiffness, good toughness, good strength in high temperature,
high wear resistance and having high ration strength to weight Material used in this research is Al-3Si ingot added with
[1,2]. alloying element, with 6 wt. % Mg and 9 wt.% Zn. Melting
process conducted in the crucible furnace on the temperature
of 850-870 C. Powder alumina reinforcement with the size of
In the current development, aluminium matrix composite is
0.45 micron was poured with 5 % volume fraction, then stirred
very promising, not only its good mechanical properties but
with the velocity of 7500 rpm. Composite produced by
also relates to its low density. Aluminium alloy is chosen as the
squeeze casting process with the application of pressure of 20
matrix because this metal is light in weight, relatively cheap
ton in a preheated mould, in order to optimize the
and easy to fabricate. Moreover, aluminium is a metal that have
solidification process and minimized defects in the interface.
been produced independently in Indonesia, therefore it can be
The composite was then underwent heat treatment by
developed even further as many application for domestic
precipitation hardening process to enhanced its mechanical
needs.
properties. Precipitation hardening process on cast composite
19
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
plate is started by solution treatment with temperature dendritic structure. The higher solution treatment temperature
variation of 450, 475 and 500 C and variation of holding time the hardening process started earliest and proceeded fastest.
of 30, 60 and 90 minutes; and followed by quenching in water, The figure also shows the squeeze casting technology seen to
and then aging process is performed in temperature of 200 C be succesfully in preventing the occurance of shrinkage due to
for 2 hours. Material characterization is conducted with the solidification process.
chemical composition, microstructure observation with optical
and electron microscope and hardness testing.
a
III. RESULTS AND DISCUSSION
Content (wt. %)
Zn Mg Si Fe Mn Cu Ni Al
9.16 6.12 2.90 0.22 0.003 0.003 0.206 Balance
c
Table 1 shows the chemical compositions of the composite.
As listed in the table, besides the intentional additions, other
elements such as Fe, Mn and Cu are also present. Even small
amount of these impurities causes the formation of a new
phase component [9]. Through casting various intermetallic
phases are formed between aluminium dendrites [9,10]. These
intermetallic phases have different structures, stabilities and
mechanical properties. Based on this reason, the cast
composite require solution treatment to improve mechanical
properties. During this treatment some transformation of
intermetallic phases such as plate-like -Al5FeSi into more
rounded discrete -Al12(FeMn)3Si particles; and dissolution of
-Mg2Si particles [11]. These transformations will give Figure 1: The microstructure of the composite with
maximum hardness after aging process [11,12]. The hardening variation of solution treatment temperature:
process is conducted in order to improve hardness and (a) 450 C; (b) 475 C; (c) 500 C.
toughness of a matrix. The heat treatment process will
decrease Mg2Si on grain boundary and increase volume
fraction of -aluminium on matrix, which then will produce To observe the presence of presipitate more clearly
precipitate MgZn2, which settle in the grain, due to aging SEM/EDS examination was conducted, as seen in Figure 2.
process [8]. Micro analysis using the SEM-EDS result on the Table 2
showed that there was MgZn2 presipitations present in the
The microstructure of the studied alloys is given in Figure matrix that inhibit dislocation process resulting in the increase
1. The heat treatment process will optimize the Zn function in mechanical properties of the Al-9Zn-6Mg-3Si composite
though precipitation mechanism, this could be seen from other [13]. Alumina in the composite showed to be distributed
particle morphology/second phase, distribution and shape. The evenly so that it can be concluded that the stirring process with
figure shows that the matrix have more globular shaped 7500 rpm seen to be distributing the alumina particles evenly.
structure with solution temperature of 500 C. It indicates that
the higher temperature of solution treatment will dissolve
20
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
4
3
3
4
2 3
1
1 1
50m
21
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
process. The authors also thank Ari Noor P, Lutfi Dwi P. and
Fajar H. for assisting the research. [7] M. Noor, Mazlee, et al.: Microstructural Study of Al-Si-Mg
Alloy Reinforced with Stainless Steel Wires Composite via
REFERENCES Casting Technique, American Journal of Applied Sciences
5(6), 2008, p. 721-725.
[8] D. Rahmalina, B. T. Sofyan, Narana Askaningsih, Sigma
[1] _____, ASM Handbook 21: Composites, ASM International, Rizkyardiani., Effect of treatment process on hardness of AL-
The Materials Information Company, 1992. 7Si-Mg-Zn Matrix composite reinforced with Silicon Carbide
[2] J.C. William, Progress in Structural Materials for Aerospace Particulate , Advanced Materials Research Vol. 875-877
Systems (ed. 51st). Acta Materialia. 2003, pp. 57755799. (2014), p. 1511-1515
[3] R. Cobden, Alcan, Banbury TALAT Lecture 1501 Aluminium: [9] r u e ieni s i d ro n t
Physical Properties, Characteristics and Alloys, pp. 60, Basic 137, 2003, 821824.
Level. [10] N.C.W. Kuijpers, W.H. Kool, P.T.G. Koenis, K.E. Nilsen, I.
[4] B.T. Sofyan, S. Susanti, R. R. Yusfranto, Role of 1 and 9 wt.% Todd, S. van der Zwaag, Mater. Charact. 49, 2003, 409420.
Zn in Precipitastion Hardening Process in Aluminium Alloy [11] N.C.W. Kuijpers, W.H. Kool, S. van der Zwaag, Mater. Sci.
AA319, Makara Teknologi, 12 (1), 2008, pp. 48-54. Forum 396402, 2002, 675680.
[5] B.T. Sofyan, D. Rahmalina, Eddy S. Siradj, dan Hery Mochtadi, [12] .K. Gupta, D.J. Lloyd, S.A. Court, Mater. Sci. Eng. A 301,
Effect Zn Alloying Element on Ballistic Performance of Al-Zn- 2001, 140146.
6Mg Matrix Composite, Proceeding of Insinas Conference, [13] S.A. Balogun, et.al, The Effect of Cold Rolling and Heat
2012, pp. 141-145. Treatment on Al 6063Reinforced with Silicon Carbide
[6] F.L. Matthews, dan R.D Rawlijns, in: Composite Material: Granules, Journal of Materials, Vol 61 No. 8, 2008, pp. 43-47.
Engineering & Science. Chapman & Hall, London, 1994.
22
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract Carbon Fiber Reinforced Polymer (FRP) is a However, the material for structural applications is updated by
material that is lightweight, strong, anti-magnetic and an expert building or construction material in the world along
corrosion resistant. This material can be used as an option with the emergence of the material weakness. Bamboo
to replace the steel material in concrete construction or as materials such as those used for the construction proved to
material to improve the strength of existing construction. have some drawbacks, such as bamboo has a very low
CFRP is quite easy to be attached to the concrete structure resistance, so the potential for bamboo powder beetles
and proved economically used as a material for repairing attacked, so the buildings are made of durable bamboo.
damaged structures and increase the resilience of Therefore the order of the building from bamboo that is not
structural beams, columns, bridges and other parts of the preserved only seen as a temporary building components
structure against earthquakes. CFRP materials can be simply do not hold more than 5 years.
shaped sheet to be attached to the concrete surface. For wood materials are also widely used for the
Another reason is due to the use of CFRP has a higher construction of houses, offices, and bridges in particular in the
ultimate strength and lower weight compared to steel village, but this material was also proved to have weaknesses.
reinforcement so that the handling is significantly easier. The disadvantage is extremely flammable, can be eaten by
Through this paper suggests that CFRP materials can be termites, may expand and creep, stretch for apps roof with
applied to concrete structures, especially on concrete wood construction is often limited due to the size of the timber
columns. Through the results of experiments conducted on the market is only about 4 meters, and other weaknesses in
proved that the concrete columns externally wrapped with terms of procurement, then in the long run the price of wood
CFRP materials can increase the strength. This treatment are increasingly expensive because of the decreasing
is obtained after testing experiments on 130 mm diameter availability of natural wood materials. Other materials such as
column with a height of 700 mm with concentric loading concrete have been widely used in Indonesia, but it also had
method to collapse. The experimental results indicate that shortcomings in terms of usage for construction. Weaknesses
a column is wrapped externally with CFRP materials can were found among forms that have made it difficult to change,
achieve a load capacity of 250 kN compared to the the weak against the strong pull, has a heavy weight, great
concrete columns externally without CFRP material which sound reflections power, and execution of work requiring high
only reached 150 kN. If the column is given internally accuracy.
reinforcing steel and given externally CFRP materials can reach Meanwhile, the material of the steel used for the
270 kN. It shows that CFRP materials can be used for concrete construction of buildings is still relatively large and very
structures can even replace reinforcing steel that has been widely dominant utilized in Indonesia to date. This material has many
used in building construction in Indonesia. advantages such as high tensile strength, not eaten by termites,
Key wordsCFRP material, concrete structure,
able to carry a heavy burden, resistant to high temperatures,
increase strength. low maintenance costs, and easily molded according to the
needs of the construction. However, it turns out, according to
experts looking at the field of construction materials still have
I. INTRODUCTION
a shortage of them can be rusty, weak against the compressive
Technological developments, especially the rapidly force, not as flexible as wood can be cut and shaped a variety
growing field of materials characterized by the appearance of of profiles, not solid, and not resistant to fire, and in the case
materials such as FRP Composite. For building construction of slender structures harmless against buckling. Therefore, the
such as bridges or buildings previously used material of various shortcomings of the steel material and other materials
bamboo, wood, and steel that serves as reinforcement. such as bamboo, wood, and concrete. So the development of
23
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
civil engineering construction material appears more structural element to be attached to the CFRP materials are
promising and superior to other materials that CFRP material. free from contaminants oxide, oil, grease, and dust.
Superiority in terms of stress and strain in the CFRP material In this research, testing the strength of the CFRP material
compared to other materials can be seen in Figure 1 below serves as reinforcement in concrete columns. The columns
(Tamon U, 2004) were tested circular column with a diameter of 130 mm and
length is 700 mm column. The focus of research is directed at
improving the lateral voltage such that it adds strength, slow
collapse process as well as a wide cross-section of the column
is ductile collapse. The results of this study, results in a
significant force in the column. This suggests that the use of
CFRP materials can be used as a function of reinforcement
and become an alternative to steel reinforcement that has been
most widely used in building construction in Indonesia.
II. LITERATURE REVIEW
A. Research results of CFRP materials for Construction
Figure 1. Strength/stiffness and fracturing Elnabelsy G and M Saatciouglu conduct research related to
strain relationship the improvement in the round concrete columns designed with
FRP materials in 2004 in Canada. They perform tests of three
For the construction of the developing world of technology large-scale bridge columns are reported in this paper. All
in the field of civil engineering materials and structural columns had a 508 mm circular cross-section and were
systems running very fast. This is demonstrated by the designed to have predominantly flexural response with a shear
increasing proliferation of research and discoveries are span of 2.0 m Measured to the point of application of load,
oriented to the use of high performance materials coupled with Consisting of 1.7 m of concrete column height and 0.3m of top
the structure of the research system, the better. The loading beam height. They were reinforced with 12 - 19.6 mm
combination of the use of high-performance materials in diameter, 400 MPa grade deformed reinforcement, Equally
structural components is reasonable and in certain distributed along the section perimeter. Each bar was spliced
circumstances can not be avoided anymore. These conditions, near the base with a splice length of 390 mm, corresponding to
among others, due to the demands of mechanical performance, 20 times the bar diameter. Each column had a diameter of 11.3
durability, ease of construction, environmental and economic mm deformed transverse reinforcement, spaced at 300 mm in
aspects. Application of CFRP materials can function as a the form of circular hoops with overlapping ends. The first
repair and strengthening of concrete structures. Retrofitting column tested (BR-C8) was the reference column reflecting
with CFRP system function can improve strength and provide as-built conditions, without any seismic retrofit. The
increased flexural capacity, shear, axial, and ductility. CFRP companion two columns were retrofitted with MBrace CF 130
materials for building construction has many advantages, carbon fiber sheets. Column BR-C8-1 had four plies of CFRP
including high durability CFRP and more economical use in sheets, whereas Column BR-C8-2 had two plies wrapped
corrosive environments than are easily corroded steel material. around the column. The surface of columns was first treated
The use of CFRP is more popular considering the number with an epoxy-based primer. The CFRP sheets
of benefits that can be obtained as the weight of the unit is a were pre-cut to the required length and applied on columns
small, easy to apply and are handled, the cost of installation with epoxy saturant. Coupons were made from CFRP jacket
and low maintenance. Material can provide the most and tested to establish the actual stress-strain relationship of
economical solution in retrofitting problem because it can composite materials. Accordingly, the jackets had
dramatically reduce the cost of labor. CFRP can be used to approximately 60,000 MPa elastic modulus and ultimate
increase the bending and shear capacity of reinforced concrete strength of 700 MPa, with linear elastic behavior. From these
beams, bending plate, push, shear and flexural column. CFRP tests yield data shown in Figure 2, 3, and 4.
in the form of sheet, plate or bar can be mounted on the
surface of the beam or plate having a stretch as a flexure
reinforcement. As the beam shear reinforcement, CFRP sheets
can be glued to the side of the beam. Usage on columns,
CFRP sheets can be placed on the outside of the column to
increase the ductility and strength.
CFRP material that can stick to the structural elements
such as beams, columns, plates, then given the adhesive epoxy
resin has the basic ingredients. This adhesive is made from a
mixture of two components. Its main component is an organic
liquid that is loaded into the epoxy, binding arrangement or
oxygen atoms and two carbon atoms. Was added to the
reaction mixture to obtain the final mix. The surface to be
attached should be prepared to obtain an effective
juxtaposition including cleanup efforts on the surface of the Figure 2. Reference column (non-retrofitted) BRC8,
24
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
reflecting as built conditions amount According to the number of composite layers, the
concrete properties and the cross-section shape. thus the use of
Carbon Fiber Reinforced Polymer is an efficient means of
providing confinement of concrete for strength and ductility
enhancement.
Samdani S and AS Sheikh also conduct research related to
concrete columns given CFRP confinement. Twenty-eight
nearly full-scale concrete columns were tested under
monotonic concentric load at the University of Toronto. The
variables tested in the experimental study included the type of
FRP (glass or carbon), the number of layers of FRP, the
orientation of fibers in the FRP shell and the amount of lateral
reinforcement. All specimens were 356 mm in diameter, 1524
Figure 3. Column BR-C8-1, retrofitted with 4 plies mm high standing. The response of the concrete confined with
of CFRP sheets FRP Showed two slopes of the ascending branch before the
peak stress. The first slope was approximately equal to that of
unconfined concrete. The second slope, being less steep,
started near the peak stress of the unconfined concrete and
continued until the peak. This was Followed by a significant
post-peak response that continued until the FRP shell was
sufficiently ruptured, resulting in a sudden drop of stress in
concrete. Figure 5 shows the axial stress-axial strain curves for
some of Toronto specimens, confined with 1 and 2 layers of
CFRP and GFRP.
25
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
26
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
27
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
III. RESEARCH MODEL shown the maximum load difference of different variations of
In this article the authors also present the results of the test specimen.
research on the use of CFRP materials are applied to the round
TABLE 1. RESULTS OF TESTS ON CIRCULAR COLUMN
concrete columns that serve as external reinforcement in
concrete columns. The work is done through analysis and Maximum Increased
Maximum
experimental studies. The study analysis was conducted to No
Specimen
Load
Load Maximum
study the stress strain models that utilize concrete as a material Code Average Load
(kN)
(kN) (%)
CFRP reinforcement in concrete columns.
1 PS-A 150
Models stress strain of CFRP material is then summarized 150 -
2 PS-B 140
and implemented in the form of a computer program Confined
3 PS-C 160
Column v.1.0 (CC-v.1.0) that have been made to produce a
4 BT-A 230
stress strain relationship chart. The program is used to validate 240 60
5 BT-B 250
the results of experimental studies. 6 BT-C 240
In this study the implementation of the test specimen used 7 B-1 LS-A 280
in the column that is round with a diameter of 130 mm 8 B-1 LS-B 270 270 80
diameter by 700 mm long round columns. In this study 9 B-1 LS-C 260
conducted experiments on plain concrete columns (PS), which Source: Research Results
uses concrete columns internally reinforced steel (BT), and
given a plain concrete column steel reinforcement internally Based on the research that has been summarized in Table 1
and externally CFRP material (B-1 LS). Concrete used is show that the concrete column specimen without restraint (PS)
normal strength concrete with a target compressive strength of capable of withstanding a load of 150 kN, while the concrete
20.75 MPa. For longitudinal reinforcing steel used 6 10 and columns with transverse reinforcement confinement and
ridden spiral reinforcement is 8-50 mm. Furthermore, an longitudinal reinforcement (BT) is able to withstand a load of
analysis and evaluation of the results of testing that has been 240 kN, and the concrete columns with transverse
done to study the behavior of restrained concrete columns with reinforcement confinement and longitudinal reinforcement as
fiber polymer (CFRP) as well as models of effective restraint. well as externally with CFRPconfinement (B-1 LS) capable of
In addition, we will get the formulation / formulas stress strain withstanding a load of 270 kN. This suggests that an increase
constitutive relations that occur due to the confined of the in the strength of the concrete column specimen BT by 60%
fiber polymer (CFRP). The resulting formulation results will when compared to columns that are not confined, while the
be validated using the constitutive equations of the results of test specimen B-1 LS increase is as high as 80% when
other researchers with the help of Confined Column v.1.0 compared to the concrete columns that are not confined. It
program (CC-v.1.0) that have been made. shows that the functioning of the confinement of transverse
and longitudinal steel reinforcement and confinement
IV. RESULT AND DISCUSSION externally with the use of CFRP materials.
Stress strain curves for all test objects can be seen in
Model collapse on circular concrete columns after testing Figure 14 as the comparative column specimens studied.
in the laboratory can be seen in figure 13 below. Concrete column specimens were observed without the use of
confinement, and concrete columns using CFRP restraint. In
figure 14 it can be seen that the presence of transverse and
longitudinal reinforcement (Specimen BT) can improve axial
compressive stress. The most influence on the value of
confinement is the specimen B-1 LS because in addition to
using the transverse and longitudinal reinforcement, also using
CFRP material as an external confinement. Increased
confinement posed CFRP compared to concrete specimen
(a) PS (b) B T (c) B-1LS (BT) is 12.5%. The results showed that with the use of CFRP
materials as an external confinement can increase the capacity
Figure 13. Collapse of Specimen Model PS, BT, of the concrete column. It is appropriate that disclosed by Mac
and B-1 LS Gregor (1997) which states loading triaxial strength of
concrete with concrete (confinement) is greater than the
Based on laboratory test results obtained from the compressive uniaxial loading.
maximum load of each variation of the concrete columns were
tested, as shown in Table 1 in the column of concrete without A. Validation of Experimental Results
confinement (PS), concrete columns with transverse Validation of Value Unconfined Concrete Strength
reinforcement confinement and longitudinal reinforcement The result of the increase in strength of confined concrete
(BT), and concrete columns with confinement transverse and validation (K) to review the model formulation by previous
longitudinal steel reinforcement and external confinement researchers using triaxial test results can be seen in Table 2.
with CFRP 1 (one) layer spacing (B-1 LS). In Table 1 are The model being simulated is a model of Campione and
28
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Miragle (2003), the model of Li et al (2003), and the model of on the results of laboratory testing through system testing
Lam and Eng (2003). using Load Control technique with the speed of movement
(stroke) of 0.012 mm / sec. The results of testing this model
produces a stress strain curve the shape of the ascending
branch. The resulting model is then summarized and carefully
observed the movement of the model curve shape of the
experimental results. The resulting shape of the curve in
general form a parabolic curve with peak coordinates
( ). The results of the model formulation of the stress
strain curve of the experimental results with comparison of
some models of previous investigators are shown in Table 3,
while, for the model of confined concrete stress strain curve of
the experimental results with two models, namely the model in
terms of Li et al (2003) and Model Campione and Miragle
(2003) is shown in figure 15.
TABEL 2. COV VALUE OF PREDICTION RESULT VS Source: Research Result of Lam and Teng, Li et al,
EXPERIMENTAL RESULT and Campione and Miragle Model
29
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
30
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract In the present work, an attempt was made to subjected to ultrasonic waves of different power levels for 5
develop high quality cast aluminum alloys via a new casting min under frequency of about 20 kHz and the maximum
technology, e.g., the heated mold continuous casting (HMC) power of 600 W, in which strong effect on the size and
with ultrasonic vibration (UV) process. With the UV process sphericity of alpha dendrites is obvious. Moreover, high
in the continuous casting process, fine and spherical grains applied ultrasonic power resulted in small, more rounded and
were obtained, where the lattice structure is formed similarly uniformly distributed grain and eutectic particles. Feng et al.
before the UV process while dislocation density increases. The [3] have attempted to treat UV into the melt hypereutectic Al
mechanical properties of the UV-HMC Al alloys are higher 23%Si alloy in a horn crucible.
than those for the related cast Al alloys without UV although From the above previous works, it appears that a number
still high material ductility is obtained. The lattice and of experimental works have been conducted to make high
dislocation characteristics of the continuous cast samples quality cast Al alloys by the UV process [3]. However, the
made with and without the UV processes were analyzed authors believe that there would have still chance to apply the
systematically by the EBSD observations to interrupt clearly UV technology in casting process. This is because, in the
their mechanical properties. previous studies, the ultrasonic vibration is conducted only to
the melt in crucibles and molds, i.e., simple approach.
Index Terms: aluminum alloy; ultrasonic vibration; continuous Moreover, there is apparently lack of the investigations to
casting; mechanical property; microstructural characteristic understand clearly the detailed vibration effect on the material
properties. This is because the previous examination has been
I. INTRODUCTION executed with the limited vibration conditions, e.g., a few
vibration amplitudes and frequencies.
In recent years, high fuel efficiency of automotive is
Thus, in the present study, an attempt was made to propose
required in our society, because of environmental issue. To
a new casting system of a heated mold continuous casting
make this, the reduction in exhaust gases from the automotive,
method with ultrasonic vibration in advance. With this casting
such as carbon dioxide and nitrogen oxide, would be required,
system, mechanical properties of several Al alloys have been
as the number of automotive has been increasing to be more
investigated. To understand clearly the effects of the UV
than 1 billion in the world. The automotive consists of a
process on the material properties of the cast Al alloys, the
number of the related parts, and many of them have been
lattice and dislocation characteristics were scientifically
made by cast irons and steels. It has been expected to replace
analyzed.
Febased automotive parts with more lightweight metals, e.g.,
aluminum alloys. The specific weight of Fe is about 7.8, II. EXPERIMETNAL PROCEDURE
which is more than 2.8 higher than that for Al. Recently, the
production amount of automotive parts, made of Al alloy, has II-1. Material preparation
In the present study, two aluminum alloys (AC4CH and
been increasing gradually.
It is general consideration that small grains with spherical ADC6) and pure aluminum (99.9%Al) were used. In order to
create the high mechanical properties of cast aluminum alloys,
shape are significantly important to make excellent
mechanical properties. To obtain such microstructural a hybrid casting system was originally proposed, where an
ultrasonic vibration (UV) device was added to our original
characteristics, some practical techniques of rapid
solidification, high casting flow and adding fine nucleating heated mold continuous casting system, see Fig. 1. In this case,
a small UV device (PEF-L25A, Sanki Corp.) was employed.
elements are employed. Furthermore, new technologies have
been proposed with mechanical modification, including The specification of this device is as follows: electric voltage:
0240V and frequency: 40400Hz. Such vibration is applied
electromagnetic vibration [1], mechanical vibration and
mechanical shearing processes. directly to the cast sample during the casting process. The
HMC arrangement consists of a graphite crucible in a furnace,
Aghayani and Niroumand [2] have examined the effects of
ultrasonic vibration (UV) treatment on microstructural a graphite mold of 5 mm in diameter, a cooling device and a
dummy rod for withdrawal of the cast sample. The graphite
features and tensile strength. The melt alloy in sand molds was
mold is jointed with the graphite crucible. The cooling system
31
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
was set just out of the mold. The ultrasonic vibration system EDX analysis was carried out to investigate the
was attached near the cooling system, and the vibrations were microstructural characteristics with an acceleration voltage of
executed directly to the casting rod during the casting process. 20 kV a scanning electron microscope. The EBSD analysis
The melts in the crucible were fed continuously into the mold was conducted to observe the crystal orientation
at 1.9 mm/s. characteristics with an acceleration voltage of 15 kV, beam
In the gravity casting (GC) process, the melt was poured current 5 nA and step size 0.520 m. The samples were
directly into a metal mold. Note, the GC process would not be prepared with sectioning to less than 5 mm thick and with
a represent conventional gravity casting process, as our gravity mirror flatness. This EBSD analysis was executed with HKL
casting system does not include sprue, runner and gate. Channel 5 software.
Cooling device
Ultrasonic vibration device III. RESULTS
50 Cast rod
5 Casting III-1. microstructural characteristics
direction
Fig. 2 depicts the optical micrographs for the pure
Molten aluminum, AC4CH and ADC6 alloys produced by GC and
metal
Heated Dummy rod HMC processes. In this case, the HMC process was carried
mold (Stainless steel) out with and without ultrasonic vibration. It can be seen that
40 40 fine -Al phase and tiny eutectic structures are observed in the
Furnace HMC samples compared to their GC ones. In addition, those
Graphite crucible grains seem to be altered slightly to more fine spherical shape
Fig. 1 Schematic illustration for the heated mold continuous casting device of -Al grains with the ultrasonic vibration. Interestingly,
with ultrasonic vibration system. core-like structures can be characterized in the middle of their
grain for the UV pure-aluminum. From the EDX analysis,
II-2. Experimental such core-like structure is related with the iron element (Fig.
Microstructure, lattice structure and strain characteristics 3).
were investigated by various approaches including energy-
dispersive Xray spectroscopy (EDX), electron backscatter
diffraction (EBSD).
Pure Al
20m
Si
ADC6
Mg2Si
Al6(Fe, Mn)
Fig. 2 The optical micrographs for the pure aluminum, AC4CH and ADC6 alloys produced by GC and HMC processes with and without ultrasonic vibration.
32
Proceeding of International Symposium on Smart Material and Mechatronics
Fig. 4 presents the crystal orientation maps (IPF) for pure Al It is interest to mention that even if the UV process conducted
and ADC6, obtained by the EBSD analysis. It is obvious that a strongly, the crystal orientations are still relatively organized.
relatively uniform lattice structure is obtained over a large area However, their lattice structures, i.e., misorientation angle, are
in the HMC samples without ultrasonic vibration, where slightly disordered.
almost perfectly orientated crystal structure, i.e., single
crystal-like formation.
HMCpure Al
SEM image Al-K Fe-K
Fe
-Al
10m
Fig. 3 EDX analysis for HMC samples of pure aluminum without UV.
Misorientation angle
SEM image IPF map
(2-5)
Pure Al
without UV
Pure Al
with UV
ADC6
without UV
ADC6
with UV
111
20m
001 101
Fig. 4 The crystal orientation maps for pure aluminum and ADC6 by HMC with and without ultrasonic vibration.
33
Proceeding of International Symposium on Smart Material and Mechatronics
Fig. 5 shows the electric conductivity (EC) of the cast Fig. 7 depicts the representative tensile stress-versus-strain
samples. Note, in this case, the EC values were measured curves for HMCADC6 alloys with and without UV process.
using the same specimen of 1 mm 100 mm. To understand It is clear that there are different trends of the tensile
EC characteristics clearly, this was also carried out for properties depending on the UV process. Based upon the
commercial wrought pure Cu, wrought pure Al and stressstrain curves obtained, ultimate tensile strength and
continuous cast Al alloys. The data obtained in Fig. 5 is fracture strain are summarized in Fig. 8. The tensile properties
indicated with the rate of the EC value based upon the copper slightly increase for ADC6 with the vibration process. Such
wire. It is clear that the electric conductivity for the increment of the tensile strength would be caused by the
commercial wrought pure Al wire is about 60% of the Cu wire change of the microstructural and lattice structures, as
one. Interestingly, slight improvement of the electric mentioned above. On the other hand, slight high ductility for
conductivity for the HMCpure Al is obvious, which is the UV samples is attributed to the grain refinement and
approximately 15% higher than that for the wrought pure Al spherical structure. Fig. 9 represents the relationship between
wire. This may be attributed to the uniformly organized stress amplitude and cyclic number to final fracture (S-N
crystal orientation, as mentioned in Fig. 4. Furthermore, it is curve) for ADC6 with and without the UV process. It is
obvious that the EC values for the cast Al alloys without UV obvious that, like the tensile properties, the S-N curve for
(AC4CH and ADC6) are about 25 % higher than those for UV. ADC6-UV is located to the higher level compared to the
This is also influenced by the different crystal orientation without UV one, namely the higher fatigue strength for
characteristics. ADC6-UV. On the basis of the above experimental results, it
could be briefly summarized that the UV process is useful to
100
7
1. Pure Al wire improve the mechanical properties of the cast aluminum
2. Pure Al alloys.
Electric conductivity, %
3. AC4CH
80 4. AC4CH (UV)
5. ADC6 350
6. ADC6 (UV)
60 7. Cu wire 300
2
1 Tensile stress, MPa
40 250
3 5
200 With UV
4 6
20 Without UV
150
0 100
Pure Al AC4CH ADC6 Cu
Fig. 5 Rate of the electric conductivity for various metals on the basis of 50
the copper wire one.
0
Fig. 6 displays the Vickers hardness of HMCADC6 alloy 0 10 20 30
as a function of the vibration frequency. As seen, the hardness Strain, %
level of the Al alloy does not change significantly even if the Fig. 7 Representative tensile stress vs. tensile strain curves for ADC6
produced by the HMC process with and without ultrasonic vibration process.
frequency is altered. However, it is obvious that high hardness
value is obtained for the sample with higher vibration (a)
amplitude. The highest hardness by UV is about 7% high 300
compared to that for the cast samples without UV.
Ultimate tensile strength, MPa
79 Amp.=50%
290
77 Amp.=99%
Vickers hardness, Hv
75
280
73
71 270
69
67 260
Without With
65 UV UV
0 100 200 300 400
Frequency, Hz
Fig. 6 Vickers hardness of HMC-ADC6 alloy as a function of the vibration
frequency and vibration amplitude.
34
Proceeding of International Symposium on Smart Material and Mechatronics
(b) ACKNOWLEDGMENTS
24
This work was supported by a grant (Grant-in-Aid for
22 Scientific Research (C), 2014) from the Japanese
Government (Ministry of Education, Science, Sports and
Fracture strain, %
20 Culture).
18
REFERENCES
16
1 C Vivs, Effects of forced electromagnetic vibrations
14 during the solidification of aluminum alloys: Part II.
Solidification in the presence of collinear variable and
12 stationary magnetic fields, Metall. Mater. Trans. B,
27B(1996)457464.
10
Without With
UV UV 2 M.K. Aghayani, B. Niroumand, Effects of ultrasonic
treatment on microstructure and tensile strength of AZ91
Fig. 8 Tensile properties of the HMCADC6 with and without UV process:(a) magnesium alloy, J. Alloys Compd. 509(2011)114122.
ultimate tensile strength and (b) fracture strain.
335.
250
200
150
100
Without UV
50
With UV
0
3
10
1000 104
10000 105
100000 106 10000000
1000000 107
Number of cycles to failure
Fig. 9 S-N curves for the HMCADC6 with and without UV process.
IV. CONCLUSIONS
1) Electric conductivity for the HMCpure aluminum is about
15% higher than that for the wrought pure Al wire. This is
attributed to the uniformly organized crystal orientation.
With the UV process, the EC levels for HMC-Al alloys
decrease about 25% compared to those without UV, which
is affected by the randomly distributed lattice structure
arising from the UV process.
2) The hardness level of the ADC6 alloy is not changed
significantly with increasing the UV frequency. In contrast,
the high hardness was obtained as loaded at the high
vibration amplitude. The highest hardness by UV is about
7% high compared to the mean hardness of the cast
samples without UV.
3) The tensile strength and fatigue strength increase for the
ADC6 alloy with the UV process. In addition, similar to the
mechanical strength, the material ductility is also relatively
increased with the UV process. Such increments of the
strength are attributed to the change of the microstructural
and lattice structures.
35
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract--- Aluminium alloy can be applied for a turbine addition of Zn, Mg and Cu considerably increase the strength
impeller of Organic Rankine Cycle power plant that operates at of Al alloys, due to precipitation of a much dispersed MgZn2
temperatures below 150 oC. Aluminum turbine impeller can and CuAl2 phase during ageing.
enhance efficiency of turbine due to light weight material. Al The aims of this study is to produce turbine impeller for an
7xxxx commonly use for impeller due to good mechanical
ORC power plant by investment casting. The turbine impeller
properties. Turbine impeller commonly produced by machining
but which is time-consuming and less efficient because of material was made from Al-9Zn-4Mg with varied additions of Cu
removal. . It can be replace by Investment casting to produce content. Microstructure and mechanical properties were
impeller turbine due to their complex geometry and precision. identified to show the effect of additional Cu content.
This study identifies effect Cu content of Al-9Zn-4Mg-xCu on the
microstructure and mechanical properties of turbine impeller
produce by investment casting. The study also identifies casting II EXPERIMENTAL METHOD
defect of turbine impeller. The structures consisted of -Al, The gating system and ceramic mold is used refer to
MgZn2, CuMgAl2 and CuAl2. Higher Cu content is the higher
previous research [7]. An Al ingot, Mg and Cu ingots were
hardness value due to CuAl2 phase. Visual examination showed
that the turbine impeller was free of macro defects and misruns. used as master alloys. Mg content was 4 wt. % and Cu content
varied at 1, 3, and 5 wt %. The alloys were melted in a graphite
Keywords: investment casting, aluminum alloys, turbine crucible. Degassing was conducted by using argon gases to
impeller, organic rankine cycle avoid gas porosity. The nominal composition of the cast alloys
is presented in Table 1.
I. INTRODUCTION
Table 1: Composition of cast alloys
Process technology to produce turbine impeller is
commonly by machining and investment casting because the Alloys Composition (wt%)
geometry of turbine impeller is more complex, need highly Zn Mg Cu Fe Mn Al
precise, and the tips of the blades are very thin. Fabrication by I 10.73 4.48 1.02 0.21 0.006 Balance
machining produces is high precision but the cost is high due to II 11.21 4.45 3.18 0.25 0.009 Balance
lengthy process and material removal. Alternatively, III 9.6 4.39 4.94 0.22 0.011 Balance
investment casting can produce a highly precise parts
characterized with more complex geometry. It is cheaper than The melting alloys was pouring into ceramic mold at
machining since the material removal is not required [1]. temperature 750 oC while preheating temperature for the
However the microstructure and mechanical properties of cast ceramic shell mold was 730 oC. After solidification, the
products highly depend on alloys, ceramic shell, gating system ceramic shell mold was broken. The turbine impeller casting
solidification process. The failure rate in production of the product was characterized by visual examination.
impeller by investment casting has been 30-40% [2]. Some Microstructural characterization by Optical microscopy and
defects commonly appear in products of investment casting: scanning electron microscopy (SEM). Mechanical properties
these include misruns, inclusion, macro and micro porosity and was identified by hardness testing were cut from the hub of the
hot cracks[3,4]. impeller. The hardness based on the Rockwell B scale
Al 7xxxx commonly use for impeller due to lightweight,
III RESULTS AND DISCUSSION
high strength to weight ratio, high corrosion resistance and
excellent mechanical properties. Compressor turbocharger that A. Visual inspection.
operates on conditions similar to ORC turbine impeller, used Turbine impeller casts were free of macro defects, such as
Al-(1-2) Mg-(2-3) Cu (wt %) with addition of titanium [5]. misruns, macro porosity, and surface cracks. Smooth surfaces
Wallace et al [6], produced turbocharger impeller that operates were obtained for all cast specimens. The successful
at temperature of 90 C by using thixocasting method. The elimination of defects resulted from a good design of gating
composition of the alloy is Al-6Si-3Cu-0.35M (wt. %) with system, ceramic shell molds, and casting parameters (i.e.,
tensile strength at 400 MPa and the elongation of 7.7 %. The
36
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
30 mm
b
Figure 2 Microstructures of the hub of turbine impeller
a
of Al-10Zn-4,5Mg-xCu with (a, b) 1 (c, d) 3 and
(e, f) 5 wt. % Cu produced by investment casting.
30 mm
B. Effect of Cu on Microstructures
Figure 2 shows microstructures of the hub of impeller with
varied Cu content. Microstuctures of the Al-10.7Zn-4.48Mg-
1.02Cu alloy are shown in Figures 2a and 2b. It can be seen the
dendritic grain structucture with the second phase around the Figure 3 Backscettered SEM of a tip of impeller of Al-
grain boundaries. The second phase commonly found in Al-Zn- 11.21Zn-4.45Mg-3.18Cu alloy.
Mg-Cu alloys are MgZn2, Mg3Zn3Al2 , Al-Mg3Zn3Al2 dan Al-
MgZn2. The addition of Cu form CuMgAl2 and CuAl2 phase. Table 2 Elemental composition on the structure in the alloy Al-
The microstructures of the Al-11.21Zn-4.45Mg-3.18Cu alloys 11.21Zn-4.45Mg-3.18Cu at position shown in
are shown in Figures 2c and 2d. The structure is relatively the Figure 3
same as that of the Al-10.7Zn-4.48Mg-1.02Cu. But it has a Rata-rata unsur (wt.%)
finer grain size and the second phase in the grain bondary look No Phase may form*
Zn Mg Cu Al
thicker. The microstructures of the Al-9.6Zn-4.39Mg-4.94Cu
1 4.48 2.70 1.52 91.29 Al (matriks)
alloy are shown in Figures 2e and 2f. The grain size more finer
than the others. The second phase in the grain bondary also 2 3.80 1.91 2.09 92.20 CuMgAl2
thicker than the others. The CuAl2 and CuMgAl2 formed in this 3 5.49 2.91 0.62 90.98 Mg3Zn3Al2
alloys due to higher Cu content. Different amount of in CuAl2 4 4.15 2.49 9.75 83.61 CuAl2
and CuMgAl2 phase caused significant differences in
mechanical properties.
37
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Figure 3 shows the FE SEM of Al-11.21Zn-4.48Mg-1.02Cu higher the Cu content the higher the hardness value due to finer
alloy. The -Al matrix can be seen (position 1), and the grain size and promote more CuAl2 and CuMgAl2 phases
CuMgAl2 phases (position 2), the thicker phase around which is hard and tough.
boundary is MgZn2, Mg3Zn3Al2 , Al-Mg3Zn3Al2 or Al-MgZn2
and it is clear that the white phases are CuAl2 (position 4)), ACKNOWLEDGMENTS
The alloys and the corresponding elemental analysis is provide The research was funded by the PUPT 2014. The authors
in Table 2. Phase formations on as-cast 1 Cu and 5 Cu would like to thanks also to PT. Metinca Prima for facilitating
containing alloys were similar except for quantities. The investment casting process. MS is grateful for the provision of
presence of a second phase MgZn2 and CuAl2 will increase the scholarship by BPP-DN Dikti .
hardness as cast and are expected to be precipitates after heat
treatment [7].
REFERENCES
78
[1] Jones, S. and Yuan, C., Advances in shell moulding for
Hardness Value
74 [2] Lia, D. Z., Campbell, J., Li, Y. Y., Filling system for
investment cast Ni-base turbine blades, J. Mat. Proc.
72 Tech., Vol. 148, 2004, pp. 310316.
70 [3] Gunasegaram, D. R., Farnsworth, D. J., Nguyen, T. T.,
Identification of critical factors affecting shrinkage
1 3 5 Cu(wt. %) porosity in permanent mold casting using numerical
Figure 4 Effect of Cu on hardness of the Al-10Zn-4.5Mg- simulations based on design of experiments, J. Mat. Proc.
xCu alloys Tech., Vol. 209, pp.1209-1219
[4] Jovanovic, M.T, Dimcic, B., Bobic, I., Zec, S.,
C. Effects of Cu on hardness value Maksimovic, V., Microstructure and mechanical
Figure 4 shows the effect of Cu on hardness of the Al- properties of precision cast TiAl turbocharger, J. Mat.
10Zn-4.5Mg-xCu alloys. The addition of Cu significantly Proc. Tech., Vol. 167, 2005, pp.14-21
increases the hardness value. The highest hardness value was [5] Furukawa-Sky Aluminum Corporation Chiyoda-Ku
obtained by 5Cu of 77 HRB. The alloy contain 5 wt. % Cu has Tokyo (Jp), Cast Aluminum Alloy Compressor Wheel
finer grain than 1 and 3 wt. % Cu caused highest hardness For A Turbocharger, European Patent Application, 2005
value, also the higher Cu content lead to more phase CuAl2 and [6] Wallace.G, Jackson.A.P, Midson. S. P. and Zhu.Q,
CuMgAl2 which is hard and though. The presence of phase High-Quality Aluminum Turbocharger Impellers
CuAl2 and CuMgAl2 leads to significantly higher hardness. Produced By Thixocasting , Trans.Nonferrous
MgZn2 phase also contribute to increase mechanical properties Met.Soc.China, vol 20, 2010, pp 1786 1791
of the alloys [8]. [7] Muhammad Syahid, Bondan T.Sofyan, Singgih Giri
Basuki, Bayu Adam, Characterization of Al-7Si-Mg-Cu
IV CONCLUSION Alloy Radial Inflow Turbine Blade produced by
Investment Casting, Advances Material Research,
The turbine impeller produce by investment casting show
vol.789, 2013, pp 324-329
free from macro defect such as misrun, macro porosity,
[8] Zang jin-xin, Zhang kun, Dai Sheng-long, Precipitation
shrinkage and others surface defects. It is indicate that gating
behavior and properties of a new high strength Al-Zn-Mg-
system design, casting parameters and ceramic shell mold work
Cu alloy, Tran. Nonferrous Met. Soc. China, Vol. 22,
optimally.
2012, pp. 2638-2644
Microstructures of turbine impeller made of Al-10Zn-
4.5Mg-xCu alloys fabricated by investment casting mainly
consisted of -Al, MgZn2, CuAl2, and CuMgAl2. Alloy
containing 5 wt. % Cu achieved the hardness of 77 HRB. The
38
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
AbstractThe utilization of effective and efficient biomass Some types of energy sources which can be renewed and
briquettes is strongly influenced by the type of stoves used and developed are solar energy, ocean thermal energy (OTEC) and
how to use them. This study aims to make coconut shell charcoal biomass energy. These biomass or organic materials can be
briquettes as stove fuel, to conduct the proximate, heating value processed as alternative fuel, such as briquette.
and physical property tests, and then to test the performance of
Coconut shell charcoal is the product obtained from the
temperature, water boiling ability and the efficiency of three types
of stoves before and after modification. The method used is an incomplete combustion of coconut shell. Charcoal gives a
experimental method by utilizing waste coconut shell charcoal higher heat and less smoke and can be smoothed and then
briquettes as stove fuel on three different types before and after compressed into briquettes in a variety of forms, in which the
modification. The results of the analysis of the chemical use of briquettes will be more practical, efficient and
composition test with average proximate analysis were 1.67 % of economical as well as easy to get than fire wood.
water content, 13.03 % of ash content, 28.69 % of volatile matters, The studies carried in connection with coconut shell
and 56.69 % of fixed carbon. Heating value was obtained in 4949 briquettes that have more heating value than other biomass
kcal/kg, compressive strength 0.489 gr/cm2 and density 0.705 briquettes i.e. : Siti Jamilatun (2008 and 2011), found that
gr/cm3. The results of combustion tests on three different types of
stoves before and after modification indicated that the stove
coconut shell briquette was the most optimum and economic
burners K1 and K3 were two types (codes) of the most superior alternative fuel, which was quite high in heating value 5779.11
briquette stove in terms of the boiling ability and modifications to kkal/kg. Herotje Siwi (2010) obtained the heating value of
suit the combustion efficiency (K1 for 71.30 % and K3 for 70.73 coconut shell of 4569.22 kcal/kg and Meli and Muslimin (2010)
%). Both of these stoves have their respective advantages. K3 obtained 5410.77 kkal/kg. The heating value difference in some
stove, in particular, can be locally produced (from clay) and previous research were probably because of the briquettes
affordable in price. The three stoves had significant increase in different manufacturing process and material composition.
performance improvement after modification. Esmar Budi (2011) found that coconut shell charcoal has
carbon content of 76.32 % which was potentially good as fuel.
Key words: coconut shell charcoal briquettes, chemical
Based on the above considerations, biomass energy in the
composition, calorific value, physical property, and stove efficiency
form of coconut shell briquette used as fuel in various briquette
I. INTRODUCTION stoves, and their modification to improve performances in
increasing the effectiveness and efficiency of alternative fuels
Indonesia was formerly known as one of the OPEC countries, a
to ease the burden of government, especially to the people who
world oil-producing organization. However, since 2003
had a hard time finding kerosene. Therefore, it is necessary to
Indonesia has turned into an oil-importing country. In 2005,
study performance improvement possibility of various stoves
Indonesias energy consumption was approximately 700 barrels
with waste biomass fuel briquettes to reduce the dependence on
of oil equivalent (BOE) per year. This amount, approximately
petroleum, especially kerosene, and to look for a more
57% of energy comes from oil, 24 % gas, 13 % coal, and the
economic alternative energy.
remainders are from hydroelectric, geothermal, and so on.
Due to the impact of the prolonged economic crisis, the II. THEORETICAL BACKGROUND
conditions change drastically when the fuel subsidies were
gradually being phased out. Several layers of society, not only Briquette stove a cooking appliance that uses fuel from
the lower class and the middle class but also the domestic briquette, which was a solid material that has been processed
industry, began to feel the weight of the fuel subsidy removal. either with or without carbonization process derived from coal
Facts and data show that the use of fossil fuel was reaching to biomass or the like. Nowadays, the use of briquette is not
the end, because the amount of oil reserves were running low. unfamiliar anymore, because of the recommendation of
High oil prices are unstable and continue to rise. The issues that government to diversify energy. Moreover, Indonesias coal
fossil fuels cause environmental damage already started to reserves were very abundant, as well as biomass. Materials used
prove. Along with the growth of the world population which in producing the stove affect the appearance, durability, and
continues to increase, people are encouraged to find alternative quality of heat utilization. The types consist of:
sources of new energy by utilizing renewable energy sources.
39
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
40
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
effiency by getting a high temperature and to minimize heat 2. KM stove (K2) was distributed by UD Barokah at Jalan
loss. The process of making briquettes and combustion test was Ketilang number 9 Makassar, South Sulawesi. This stove
carried out in Production Process Engineering Laboratory was made from steel plate. This stove would be modified
UNHAS Makassar, proximate and heating value tests were on the exterior walls of the air hole stove. By adding 8
carried out in Animal Feed Chemical Laboratory, Faculty of pieces of air hole which were originally 11 mm in diameter.
Animal Husbandry UNHAS Makassar and physical test was And with asbestos insulation and zinc bound with wire. The
carried out in Laboratory of Metal Science, Faculty of distance between each air hole to others was 31 mm. With
Engineering UKIP Makassar. All data including combustion this addition and insulations, more air was expected to enter,
time, flame temperature, water temperature, and ambience so the combustion could produce higher temperature and
temperature displayed on this study derived from measurements shorter cooking time. Photo K2 stove and photo
on experiments conducted in the laboratory, while the formula modification of K2 stove can be seen in Figure 4 below.
used to calculate the thermal efficiency was obtained from
several reference books.
A. Materials and Equipment
Materials and equipment used in this study were as follows:
1. Material:
a. Coconut shell
b. Tapioca starch
c. Clay
d. Water
2. Equipment:
- Briquette press machine
- Coffee grinder a. K2 stove before modification b. Modifield of K2 stove
- Carbonization drum
Gambar 4. Photos of K2 stove
- Sieves (40-60 mesh)
- Briquette stove 3. K3 stove was made by Takalars pottery center with wood
- Thermocouple charcoal as fuel and as a substitute for kerosene stove. This
- Scales stove was made of clay (pottery) which was very easy to
- Water find in traditional markets. This stove would be modified by
- Bomb calorimeter adding aluminum plate cylinder with a thickness of 0,9 mm,
- Aluminum pot a diameter of 90 mm,a high of 140 mm by making one row
- Beaker glass of air holes surrounding the cylinder. The diameter of each
B. Modified Stoves air holes 10 mm and the distance between the air holes was
The method used in this study was an experimental method, 20 mm. The purpose was to add insulation to reduce the
which was to modify 3 different types of stoves. heat loss to the walls in radial direction. Photo K3 stove and
1. Britubara Stove (K1), made by PT Britubara Indoraya photo additing cylinder could be seen in Figure 5 below.
Indonesia, made of porcelain-coated steel plate, heat-
resistant (1500 oC), which was a surface heat-resistant
cylinders. This stove modified by adding aluminum plate
cylinder with a thickness of 0,9 mm, a diameter of 90 mm, a
high of 140 mm by making one rows of air holes surrounding
the cylinder. The diameter of each air holes 10 mm and the
distance between the air holes was 20 mm. The purpose was
to add insulation to reduce the heat loss to the walls in radial
direction. Photo K1 stove and photo additing cylinder
modification can be seen in Figure 3 below.
41
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
A. Results of Research
1. Produced briquetees with specification :
a) Cylinder-shape briquette (bee nest)
Briquettes produced had average diameter dimension of (d) =
65 mm, high (t) = 45 mm, hole in the center hole (d1) = 15 mm
and around (d2) = 8 mm (four holes).
b) Briquettes mass and volume
The briquette mass was 93.3 gr
42
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
43
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
temperature was 255 oC. And compared in Figure 11 the K3 K1 stove before modification only two times water boiling
stove after modification (K3) could boil water 5 times at thewith maximum temperature of 309oC and efficiency 24.92%.
27th, 58th, 93rd, 157th, and 227th minute (blue line), could only
And after the modification, K1 stoves water boiling had
heat the water until 87oC (red line), where the maximum flame suffered almost as many as 5 times with temperature of 393 oC,
temperature was 373oC. efficiency was 71.30%. This was because the addition of a
b. Combustion efficiency modified aluminum cylinder plate hole above the line that was
able to maintain the briquettes fire heatfor almost more than 5
Example for thermal efficiency calculation was for K1 stove
boiling water for 2 times with maximum flame temperature of hours.
309 oC with briquette burning time for 240 minutes (4 hours), K2 stove before modification produces 3 times water boiling
and spending briquette 0.22 kg. Furthermore, the data could bewith temperature of 451oC and efficiency of 36.69%. K2 was
seen as follows: also the best before modification because the preparation of the
Calculation efficiency was : briquettes in the combustion chamber is arranged horizontally.
And after the modification of 8 holes variation with asbestos
insulation and zinc bound with wire, that had K2 stove. So, the
performance improvement could be obtained twice better than
previous modification with the process of water boiling to 4
times.
K3 stove before modification only boiling the water one
time with flame temperature of 255 oC and efficiency 15.70 %.
After modification of K1 stove, water boiling ability was
The comparison of the three types of stoves data testing obtained 5 times with not too high temperature of 373 oC and
before and after modification in water boiling ability, efficiency of 70.73%. This was due to the additional
maximum flame temperature, and thermal efficiency of the modification of one top row aluminum cylinder plate which was
burnt-out briquettes presented in the Table 3 below. also able to sustain the flame briquette for 335 minutes.
A. Conclusion
Briquettes had been made in the form of bee nest, with the
diameter of 65 mm, height of 45 mm, a 15 mm hole, and 4
holes with diameter of 8 mm. The average proximate test result
showed water content of 1.67 %, ash of 13.03 %, volatile
matters of 28.61 %, and fixed carbon of 56.69 %. Heating value
K1 stove before modification had maximum flame temperature, was obtained in 4949 kcal/kg, compressive strength 0.489
water boiling ability, thermal efficiency and burnt-out briquette gr/cm2 with the density of 0.705 gr/cm3. The overall results of
mass respectively as : 309 oC, 2 times within 240 minutes, and proximate and physical property tests did not meet the existing
24.96%. After modification of K1, the maximum flame quality standard of briquettes, except for water content and
temperature was 393 oC, 5 times boiling within 345 minutes volatile matters.
and efficiency 71.30%. K2 stove before modification was The results of performance testing of 3 types of stove before
respectively 451oC, 3 times within 215 minutes and efficiency modification obtained : that K2 stove was superior in terms of
36.69 %. And after modification of K2 stove was 4 times water boiling ability, maximum flame temperature, and thermal
within 240 minutes, 398 oC and 54.17 %. K3 stove before efficiency of each of 3 times within 215 minutes, 451oC and
modification produced 255oC, 5 times within 180 minutes, and 36.69%. Followed by K3 stove with water boiling 2 times
15.07%. K3 stove after modification was 5 times within 335 within 240 minutes, 309 oC and 24.92 %. And the lowest was
minutes, 373oC, and 70.73 %. K3 stove which was to boil water as much as one time within
B. Discussion 180 minutes, 255 oC and 15.70 %. Water boiling ability and the
Briquettes was burned in a bee nest shape with diameter best thermal efficiency after modification were produced by
dimension of 65 mm, height of 45 meter, one 15 mm hole and K1 and K3 stoves which was 5 times, where K3 stove was 10
4 small holes with diameter 8 mm. The average proximate, minutes faster than K1 stove with each efficiency 71.30% and
heating value and physical property test results were water 70.73 %. Of the overall testing both before and after
content of 1.67 %, ash content 13.03 %, volatile matters 28.61 modification, the most superior are K1 and K3 stoves. Both of
%, fixed carbon 56.69 %, heating value 4949 kkal /kg, density these stoves had their respective advantages, which K1 stove
0.705 gr/cm3and compressive strength 0.489 gr/cm2. The result was slightly more superior in terms of thermal efficiency and
of these tests indicate that they did not meet with the briquette maximum flame temperature. Nevertheless, K3 stove excels in
standard, except for the water content and volatile matters. ignition time and also had the additional advantage as it could
Although quality improvement efforts by improving the be produced locally (from clay) and affordable in price. The
charcoaling process (reducing the ash content) and drying best efficiency improvement was obtained by K1 stove for
process (reducing the water content) as well as the optimal use 46.34 % (from 24.96 %-71.30%), K2 stove for 17.485 (from
of particles of 40-60 mesh sieve have been done.
44
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
B. Suggestions
Improved performance had been successfully performed
on three different types of stoves. But there were still
deficiencies found in this research that the process of water
boiling (cooking time) was quite long and less amount of
materials cooked. Therefore, it was necessary to conduct
further research, especially in terms of designing the optimal
stove with more fuel capacity and more ingredients that could
be cooked in relatively short time. Emission testing was also
necessary to conduct.
REFERENCES
[1] Kuncoro Heru dan Damanik Ladjiman (2005), Kompor Briket
Batubara Tanpa BBM Dan Hemat Biaya, Penebar Swadaya,
Jakarta 2005.
[2] MM Faozi., 2008, Peluang Pasar Produk dari Kelapa Indonesia,
Analisa Dampak Menipisnya Cadangan Minyak Bumi Dan
Perubahan Iklim, Http://www.mmfaozi.com/peluang_pasar,dll.
Diaskes 11 Februari 2009.
[3] Arif, E,.2008, Pemanfaatan Briket Limbah Biomassa Sebagai
Sumber Energy Alternatif, Laporan Penelitian Fakultas Teknik
Universitas Hasanuddin Makassar
[4 Mahadir Sirman., 2013, Peningkatan Kualitas Briket Campuran
Limbah Ketam Kayu Merbabu, Sekam Padi dan Tongkol Jagung
Pada Berbagai Komposisi, Laporan Penelitian Fakultas Teknik
Universitas Hasanuddin.
[5] Syahrir M., 2011, Limbah Batang Jagung Sebagai sumber Energi
Alternatif. Laporan, Penelitian FakultasTeknik Universitas
Hasanuddin.
[6] Mangkau A (2011), Karakteristik Pembakaran Briket limbah
Tongkol Jagung dan Sekam Padi Dengan Berbagai Perbandingan
Tongkol Jagung Dan Sekam Padi, Laporan Penelitian Fakultas
Teknik Universitas Hasanuddin.
[7] Jamilatun S., 2011, Kualitas sifat-sifat dari pembakaran tempurung
kelapa, briket serbuk gergaji kayu jati, briket sekam dan briket
batubara, Prosiding seminar Nasional Teknik KimiaKejuangan
Universitas Ahmad Dahlan Yogyakarta.
[8] Jamilatun S., 2008, Sifat-Sifat Penyalaan dan Pembakaran Briket
Biomassa, briket batu bara dan Arang Kayu, Jurnal Rekaya
proses., Vol. 2, no. 2, 2008.
[9] Siwi H (2010), Pemanfaatan Limbah Tempurung Kelapa dan
Enceng Gondok Sebagai Sumber Alternatif, Laporan Penelitian
Tesis Fakultas Teknik Universitas Hasanuddin
[10] Esmar Budi., 2011, Tinjauan Proses Pembentukan dan
Penggunaan Arang Tempurung Kelapa sebagai Bahan Bakar,
Jurnal Penelitian Sains Vol. 4, No. 3 (B), Oktober 2011.
[11] Meli dan Muslimin (2010), Pengaruh Dimensi Arang
Tempurung Kelapa Terhadap Mutu Briket, Skripsi S1 Teknik
Mesin Fakultas Teknik Universitas Hasanuddin.
[12] Arianto (2010), Daun Kering Kakao dan Daun Kering Kayu Jati
Dijadikan Sebagai Energi Alternatif, Skripsi S1 Teknik Mesin
FakultasTeknik Universitas Hasanuddin.
45
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
AbstractIn Japan, aging of infrastructures, such as roads, as a tunnel or under a bridge, where the GPS signal does not
bridges, and water and sewer services, etc. poses a problem, and reach to or is weak. In addition, the manual operation by a
it is required to extend the life-span of such infrastructures by joystick etc. is difficult when affected by disturbances such as
maintenance. Among infrastructures, especially bridges are wind etc. Lupashin and DAndrea [5] have proposed a method
periodically inspected by short range visual observations, which
for controlling an X4-Flyer using a tether, not relying on the
check the damage and deterioration of the surface. However,
since there are some cases where the short range visual operation of a joystick or the use of GPS. However, this
observation is difficult, an alternative method is required so as to method only controls the tilt of the aircraft towards the tether,
replace the short range visual observation with it. So, "X4-Flyer" so that it is impossible to vary independently the altitude and
is very attractive because of realizing a movement at high altitude position of the X4-Flyer, respectively. Therefore, it is thought
easily. The objective of this study is to develop a tethered X4- to be difficult to be used in an inspection of infrastructures,
Flyer, so that the conventional short range visual observation of such as bridges etc., as it is.
bridges is replaced by it. In this paper, a method for the In this study, it aims at developing the X4-Flyer with tether
measurement and control of the position is described by using a to replace the short range visual observation of infrastructure.
tether for controlling the position of the X4-Flyer. In addition, it
is checked whether the tethered X4-Flyer can control the position
The position and attitude are controlled by the inertial sensor
using the proposed method or not, letting it fly in a state in which and the altitude sensor that are attached on the airframe of the
a tether is being attached. X4-Flyer, and by a tether attached at the bottom of the airframe.
In this paper, we first describe the summary of the X4-Flyer
Index Terms Aerial Robotics, Unmanned Aerial Vehicles, and a controller for the position and attitude. Then, a method is
vehicle dynamics, Control. explained for measuring the position of the X4-Flyer by
applying a tether. In addition, it is checked whether the tethered
I. INTRODUCTION X4-Flyer can control the position using the proposed method or
In Japan, there are about 700,000 bridges whose length is not, letting it fly in a state in which a tether is being attached.
2[m] and more. Aging of such bridges is a serious problem [1], II. OVERVIEW OF AN X4-FLYER
because those about 50[%] and more exceed 50 years in 2030,
which is the life of a bridge. As a general rule, an inspection f2 f1
period of the bridge is determined to be five years and less by
the short range visual observation to cracks and corrosion [1]. X
The short range visual observation by a human is widely used
to evaluate the degree of damage and understand the damage I
status of concrete structures such as bridges, because it is
possible to check the deterioration and damage of the surface f3 {B} f4
cracks, etc. However, there exists a case where the bridge
inspection vehicle cannot be used due to an insufficient space
under the bridge digit, and also exists a case where a close
visual inspection is difficult because of a large-scaled traffic \
control, a necessity of installation of scaffolding, etc. For this Y
reason, it needs a substitutive method of the short range visual Ex
T
observation. So, an aerial robot to move at high altitude easily Z {E}
is very attractive. In particular, an "X4-Flyer", which is a kind Ey
of VTOL type aerial robot, has high maneuverability,
compared to conventional VTOL aerial robots possessing other Ez
rotor arrangements [2]. Therefore, it is expected to be used in
various applications, such as security, pipe inspection, etc. [3] Fig. 1. Definition of the coordinate system for the X4-Flyer.
[4]. It needs to control the position and attitude of the X-4 Flyer,
if it is used for the inspection of infrastructures, such as bridges Fig. 1 shows the coordinate systems and the appearance of
etc. Although the position control using the GPS is common, it X4-Flyer, respectively. The body coordinate system of the X4-
is difficult to use such a control method in environments, such Flyer is denoted by B and the inertial coordinate system is E ,
46
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
U2 k1 (I Id ) k 2I Fig.2 shows a situation where a device for measuring the
inclination of the tether is attached to the airframe. This device
U3 k3 (T T d ) k 4T consists of a gimbal mechanism equipped with potentiometers.
This gimbal mechanism is composed of orthogonal two axes,
U4 k5 (\ \ d ) k6\ which can incline in any direction respectively. The inclination
of the X4-Flyer can be known by measuring the slope of each
B. Controller of the Position axis, because two axes move in any direction.
The position control of the X4-Flyer is performed by B. Position Measurement Using the Tether
changing the attitude of the airframe. It is found from Fig.1 that
the X4-Flyer can move X -direction and Y -direction by tilting
the airframe to the direction T and I , respectively.
Therefore, the X4-Flyer in this paper is controlled to X -axis za
and Y -axis directions by changing the target value T d and Id
J
in Eq. (4), respectively. That is, a feedback-roop is constructed E
to generate and change the target values of attitude angles of Ex
D
the airframe, by using the errors from the current position to
the target position of the airframe. Here about the position
control, a PD controller is assumed to be used as the control at l
the attitude angles. When defining the P gains of the controller xa
as k 7 and k 9 , the D gains of the controller as k 8 and k10 ,
and the target values of attitude of the airframe as x d and y d ,
the PD position controllers are given by c
Td k 7 ( x x d ) k 8 x . Ey
ya
Id k 9 ( y y d ) k10 y .
Ez
The X4-Flyer is equipped with a tether, maintaining in the Fig. 3. The tether tilt and the airframe positions.
state where it is stretched. Then, the control input U 1 is set to
be constant so as to generate a constant thrust to the height of Fig.3 shows the relationship between the inclination of the
Z -axis direction. tether and the airframe position. Let the E x -, E y -, and E z -
axis positions of the airframe be x a , y a , and z a . c denotes
the distance from the origin of the coordinates to a point at
which a perpendicular line given from the center of the
47
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
airframe intersects the E x - E y plane, and l is the length of performing the attitude control are set to k1 4.5 , k 2 1.5 ,
the tether. The slopes of the tether against the perpendicular k 3 4.5 , k 4 1.5 , k 5 1.2 , and k 6 0.4 . The constant
line directed to E x -axis and E y -axis are defined by D and gains in the PD controller for performing the position control
E , respectively, and the J is a slope of the tether. Then, the are set to k 7 0.12 , k 8 0.8 , k9 0.18 , and k10 0.2 .
airframe position in the E x -axis is given by X4-Flyer
PC
xa z a tan D
Fixed point
Furthermore, the airframe position in the E y -axis is reduced to
ya z a tan E
0 10 20 30 40 50
input U 1 in the height direction to perform an operation such Time [s]
as takeoff etc. The constant gains in the PD controller for Fig. 6. Position in the X-axis direction of the airframe.
48
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
0.8
VI. CONCLUSION
In this paper, a method for measuring and controlling the
0.6
position of an X4-Flyer has been described by using a tether.
0.4 Furthermore, the proposed method was verified using a real
Position Y [m]
49
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract In this paper a human-sized tracked wheel robot by seating in a control station-a computer with a joystick and a
with a large payload capacity for tele-presentation is presented. webcam. The wide angle display of the robot cameras presents
The robot is equipped with different sensors for obstacle the doctor with the view of the robots environment for
avoidance and localization. A high definition web camera navigation and to examine patients and converse with the
installed atop a pan and tilt assembly was in place as a remote
hospital staff. Blackwells QB on the other hand is a general
environment feedback for users. An LCD monitor provides the
visual display of the operator in the remote environment using purpose remote telepresence robot. It is a Wi-Fi enabled,
the standard Skype teleconferencing software. Remote control vaguely body-shaped wheeled robot with an ET-looking head
was done via the internet through the free Teamviewer VNC that has cameras for eyes and a display in its chest that shows
remote desktop software. Moreover, this paper presents the an image of the person it's standing in for. You can slap on
design details, fabrication and evaluation of individual virtual-reality goggles, sensor gloves, and a backpack of
components. Core mobile robot movement and navigational electronics to link to it over the Internet for an immersive
controls were developed and tested. The effectiveness of the telepresence experience. Or you can just connect to the robot
mobile robot as a test bed for tele-presentation were evaluated through your laptop's browser.
and analyzed by way of its real time response and time delay
effects of the network.
Telerobots, teleoperators, and remotely operated
vehicles belong to a class of machines used to accomplish a
Index Terms telepresence, teleoperation, remote sensing, task remotely, without the need for human presence on site.
Skype, Teamviewer. They are typically used in situations that are too hazardous to
human health or survival, like deep water, outer space, or toxic
I. INTRODUCTION environments. A growing number of telerobots is used for
A.BACKGROUND applications where it would be too expensive or too time-
consuming to send humans, for example in telemedicine or
Telepresence has been the hype of the modern day
tele maintenance that requires highly trained individuals with
fictional films like the Surrogate, Avatar, Sleep Dealer and
special skills. Sheridan [11] defines a telerobot as a machine
Gamer. In Surrogate for example, a person will be assigned to
with sensors of the environment and devices to perform
a robot replica of oneself which one has to control via his
mechanical work. The human operator supervises the telerobot
mind in a recliner seat at home. The much younger, stronger,
through a computer intermediary. The operator communicates
better looking replica will then do your biddings in the world.
to computer information about goals, plans, and orders relative
The real technology behind such fantastical fiction is
to a remote task, getting back integrated information about
grounded both in far-out research and practical robotics. In the
accomplishments, difficulties and sensory data. The telerobot
present day real world telerobots is presently limited to
executes a task based on the information received from the
physical interfaces-through wireless internet connections,
human operator plus its own artificial sensing and intelligence.
video cameras, joysticks, and sometimes audio. Humans move
The cost of a telerobotic system can be considerably
robots around at the office, in the operating room, underwater,
reduced by using personal computers and prevailing standard
on the battlefield, and on Mars. Examples of advances in
software for most of the computing tasks[1]. Another way to
todays telepresence robots are the RP-7 of InTouch Health
reduce costs for some applications is to use the free Internet
Systems and QB of Blackwell's Anybots [5]. RP-7 is a
for communication between the computer that the operator
mobile robotic platform that enables the physician to be
interacts with and the computer that controls the robot.
remotely present. Through the integration of key technologies,
Telepresence robots on the other hand are specialized types of
RP-7 can remove time and distance barriers and effectively
networked telerobots that offer the operator some form of both
extend the physician's reach to manage patient care. The visual and tactile feedback giving him a sense of as if he is at
Robots visualization system consists of a camera,
the actual site of the robot.
microphone, and a speaker. Mobility and navigation is
Compared to plain robotic systems, in which a robot
possible via a holonomic drive system and an array of infrared
executes a motion or other program without further
sensors. Physicians teleoperate the robots through the internet
consultation of a user or operator, telerobotic systems provide
50
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
information to and require commands from the user. Their robot come in kits that can be obtained in local stores. So in
control architectures can be described by the style and level of general, a telepresence robot that can be built using modular
autonomy. components was developed. In terms of robot control
Using standard internet technology for telerobotic software, most of the previously reviewed robots have their
applications offer the advantage of low cost deployment. own proprietary software and most of them are closed source.
There is no longer a requirement for expensive purpose built In this particular telepresence robot, it was decided to use free
equipment at each operators location. Almost every computer and open source software. This decision was based on the
connected to the internet can be used to control a teleoperable notion that open source and free software can contribute to the
device. The downside is the limitation of the varying robots low price and a generalized robot can be synthesized
bandwidth and the time delays. by just anyone with the right means. The previously reviewed
The internet offers the infrastructure for telepresence robot are considered complex in terms of control
communication but still the operator requires software that and software and the designed robot however was not.
displays the user interface and communicates with the Moreover, the teleconferencing component of the robot came
telerobot over the internet. Powerful browsers are freely from the freely available Skype software. Overall, a
available and often updated with increased functionality. User telepresence robot designed with modular components and
interfaces can be developed using a web browser only can be free software was developed. This means that anyone with
achieved. Operator visual feedback can also be realized using sufficient knowledge in electronics and computer
readily and freely available Skype software, freeing the programming can develop their own telepresence robot since
developer from developing its own visual feedback. the components can be obtained readily.
Existing commercial telepresence robots are way far To solve the problem stated thus we have done the
expensive ranging from 6000 to 15000 dollars. Some have following:
their own dedicated web servers to manage control, and visual 1.Designed and constructed a web controlled
feedback commands thus adding to the cost overhead by telepresence robot platform with the following
charging users a monthly subscription fee[5]. This research features
was conceptualized bearing in mind the cost savings when an a. A tracked wheel differential drive motion
elementary telepresence robot will be developed from off the capable robot with motion control system
shelf components and doing away with a dedicated web incorporating two sets of optical wheel encoders
server. for dead reckoning linear displacement
Having a dedicated web control infrastructure measurements and speed control using pulse
however leads to a much effective telepresence robot by width modulation.
ensuring an almost zero communication downtime, thus a b. Incorporated a webcam, microphone sets and an
significant reduction in delays, by efficiently rerouting IP LCD display for two way audio and video
packets to other routes. Since the robot to be designed do transmission between the robot and remote
away with this type of infrastructure, a controller/robot controller.
communication system must be developed with the ultimate c. Employed three sets of ultrasonic distance sensor
goal of minimizing the effects of transmission delays. for the robots obstacle avoidance system.
Telepresence robots are useful in various areas such as d. Incorporated on the robot a digital compass for
remote presentation, teleconferencing, telemedicine: remote direction sensing based on the 4 cardinal
diagnosis, remote treatment of patients with the aid of a directions.
medical staff and remote consultation, military applications, e. Incorporated on the robot a 3 axis accelerometer
remote surveillance, teleoperation , advertising and remote for inclination measurement on the three
education Due to these benefits, a low cost telepresence robot cartesian axis.
from off the shelf components was developed. A heightened f. Use of the Skype teleconference software for the
perception of the presence of the speaker in the remote area audio and video information transmission
was sought and for safety purposes, a control scheme with between the robot and the remote controller.
minimal delay effects was developed. g. Low cost and sourced from Commercial off the
Most of the existing telepresence robots are either Shelf (COTS) materials.
private projects or commercial ones. Each of the robots have
their own advantages and disadvantages and this study would 2. Developed a purely local robot control using a
like to fill in the gaps left by these telepresence robots. In this processing GUI whose control is ported to the
study, a robot was developed that somewhat mimics the operator via TEAMVIEWER VNC.
properties of these established telepresence robots but has 3. Tested the effectiveness of the robot as a telepresence
some inherent characteristics that somehow complements the agent based on the evaluation of the time delay
shortcomings of these previously reviewed robots. First of all between the transmission and execution of control
a robot made up of locally available materials was built and commands.
thus leads to a low cost platform. The blueprint for the robot 4. Developed and optimized control algorithms for the
was laid out as simple as possible so that a generalized model robots navigational control.
can be easily duplicated. The design of the hardware was open
sourced and the sources of the components were easily
obtained locally. The electronic controls components of the
51
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
II.SYSTEM DESIGN
To reduce complexity, the telepresence robot was a passive
terminal with minimal autonomy. A little intelligence however
was incorporated through the sensors in a way wherein the
robot can have control in case sensors detect obstacles.
Mobility was limited to positioning controls, although the
prototype can be designed for much general and complex
teleoperation task, modelling, calibration and testing for the
prototype would be done for remote tele-presentation only. The
software used for teleconferencing was limited to Skype and
robot control was achieved through forwarding the remote
desktop to the controlling computers screen using Teamviewer.
The Processing programming language was adopted in
developing the robot controllers GUI. Figure 2.The robot Graphical User Interface
The choice of the type of robot was based on some existing A. Hardware Design
commercial telerobot designs. The robot first and foremost The hardware of the robot platform is shown in Fig. 3 and
must be able to move around in the remote environment thus a
Fig. 4. The remote robots skeletal system was made up of a
suitable mobile platform was designed. Since the robots task combination of 1x1 square aluminum bars comprising the robot
was to interact with other people on the remote site as
frame and a main vertical post made up of a 15/8 by 1 5/8 14
naturally as possible, a human sized robot was designed. gauge slotted angle bar. The post acts as a backbone to support
The robot can be remotely controlled by a remote PC. Fig.
the visual feedback system consisting of the 14 inch LCD
1 shows the overall scheme of the system. Basically, the monitor and the camera system. Nuts and bolts of various sizes
system consist of two computers, one top of the robot and
were utilized as fasteners. Acrylic plastic sheets with thickness
other will be the remote controller station. These PCs were of 2.5 mm and 3 mm was used as coverings. A tracked
both connected to the net and communicates with each other
industrial platform acts as the robot base. The main controller
via Skype and Team viewer VNC package. of the robot was a microcomputer with an Intel-based
processor. This computer processes the video feed from the
robot mounted camera, processes the audio feed from the
microphone , outputs the received video information from the
controlling PC to a robot mounted LCD, outputs the received
audio signal from the controlling PC to the robot speakers,
control and monitor the communication link between the
controller and the robot, and perform motion and tactile
commands to the robot as well as process the robots sensors
relaying feedback signals to the controlling PC. The robot
moves via a differentially driven track wheels which are in turn
webcam
Figure 1.The telerobotic system showing the telepresence robot and its LCD
remote controller station
Video, audio data were relayed from the remote PC to the Speakers
controlling PC in the robot and vice versa. In this setup, the Digital Compass
52
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
US-100
ultrasonic
Buzzer
Wifi to sensors
webcam
Serial x3
Module slave mcu 1
Digital digital inputs
Intel based
Compass master mcu
Microphone Laptop digital input
Computer UART
3-Axis interrupt digital output
Accelerometers soft serial Atmega 8l
analog inputs soft serial slave mcu 2
LCD Screen Wireless Optical digital input digital output Atmega 8l
adapter
Encoder
Atmega328 soft serial
x2
12V digital output
inverter 2 nissan 6A Motor
High power Controllers\
speakers geared x2 Tilt Servo Pan Servo
Motors Motor Motor
53
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
54
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
For this example, the robot selects the clockwise movement System
Initialization
for it entails lesser angular displacement to be covered. Routine
The actual heading angle was measured by a digital
compass sensor whose output is a serial data sent to the robot
controller. The actual desired angular headings of the robot are
displayed on the GUI using a circular gauge with two dials. Wait for Serial
The red dial is for the actual heading and the blue dial is the Commands
desired heading. Whenever the robot moves, this actual
heading dials location is updated. One can also manually
update the actual angular position by clicking the
no
READCOMPASS button. The Robots Pan and Tilt Camera Command
can be controlled via the Tilt Angle and Pan Angle sliders in Recieved?
the GUI. The Pan angle slider value corresponds to the amount
of angular degrees the camera pans. These angular values can yes
range from 0 to 180 degrees. The tilt angle slider values
correspond to the amount of degrees the camera tilts. These
Process Commands
angular values can range from 0 to 180 degrees. By clicking
the neutral button, one can reset the camera to point to a
default viewing position which is the usual orientation of a
webcam. Neutral position corresponds to a pan angle of 85
degrees and a tilt angle of 90 degrees. An overall GUI for the
robot controller was employed on the robot. This GUI serves Figure 9. Flowchart for the main controller loop
as a frontend to a telnet session between the robot PC and The
Wi-Fi to serial converter server. The Wi-Fi to serial converter
55
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
The initial step in the main microcontroller code was the III.TESTING AND EVALUATION
initialization routine. Here every piece of hardware interface The main goal of this study was to develop a basic
was initialized, I/O pins data direction are set depending on telepresence robot whose affectivity was dependent on how
the peripheral connected to it. Figure 10. is the flowchart for accurately it performs the tasks assigned by the controller. The
the initialization routine. The initialization is implemented on study implements the objective method of evaluating the
the setup( ) function of the main microcontroller code. performance of the telerobot through a series of tasks that it
has to complete at the shortest time possible, thus task
completion time and reaction time to a remote stimuli were the
metrics used. Task completion time was measured and
averaged while the reaction time was measured by getting the
time difference between the completion time in the local
station and the time of completion as perceived in the remote
station. Two tasks have been designed, one is the traversal of a
square path and the other one is the traversal of a straight path.
The traversal of the square path employs the shared
continuous means control. In this method a command and
wait strategy was implemented by which the robot was sent
commands one at a time. One command must be finished first
before another command was sent. The robot on the other hand
monitors obstacles along the path, and once an obstacle was
detected, it stops the present maneuver of the robot. The next
step for the robot was then to wait for the next command and
execute it. The path was traversed 10 times and the time of
completion was recorded through Wireshark, an open source
network sniffing software. The perceived time of completion
was also measured on the controlling computer through
Wireshark by looking at the timestamp on the TCP
transmission of the last command coming from the remote
machine. The testing area would be a tiled surface with an
outline drawing of the path to be traversed. Two paths have
been defined for testing, one is a square path with a side
dimension of 1.2 m and the other is a straight line with a
Figure 10. . Flowchart for the initialization routine
length of 1.6m.
In the experiment, the robot performs its usual startup
routine whereby it connects to the controller station via
The bulk of the main loop of the master controllers firmware
internet. Skype and Teamviewer were run on both the robot
was a blocking wait loop. The program waits for serial data
and the controlling station. The robots controller GUI was
coming from the wifi serial module. The data coming from the
then forwarded to the controller screen so that it can remotely
wifi module were actually commands coming from the main
control the robot. Skype provides the visual feedback to the
graphical user interface. As mentioned previously, the
controller PC so that it can effectively maneuver the robot.
graphical user interface processes commands entered by the
Once connection was established, the robot then traverses the
user as relayed through the internet via Teamviewer. The
paths. While the robot was traversing the paths, Wireshark
commands received by the GUI were then encapsulated into a
monitors the wireless data packets leaving and entering the
TCP packet which was then relayed to the serial wifi module.
robot PC. These data packets are then sent remotely to the
The serial wifi module then decapsulates the commands from
controller stations PC for monitoring. Wireshark in the
the TCP packet into a stream of serial data. These serial data
controller station also monitor its wireless data packets
will now be the serial commands the main controller loop
particularly the timestamps on each TCP packet. Through the
processes. Commands are formatted as a series of 4 bytes The
timestamps in the TCP packets, the time of completion and
wait loop in the main loop waits for this and upon reception,
time delays was calculated and recorded. In calculating the
the first byte then was used as an argument in a switch case
time delay, the tasks starting time and ending time were
block in the code. Different commands then are executed
extracted from the TCP stream trace and packet display list of
based on the decoded first byte. Likewise if a particular
the TCP transmission between the robot PC and the wifi
command needs a response from the robot, the said response
module. Fig. 11. shows software modules and the data flow
was sent to the wifi serial module as a serial character stream.
during the experiments.
This serial stream was then converted to a TCP packet which
will then be transmitted to the GUI client on the controlling
PC. An example of this was the string ready which was
transmitted every time a command was successfully executed
by the robot.
56
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Internet
teamviewer control
signals
TCP and UDP
Local Area Network
Controller PC Robot PC
Teamviewer control
commands passed on
Putty is used to signal the
the Java GUI
robot that the last manuever Teamviewer
was observed to have been
Teamviewer
accomplished
Wireshark WinPCap
Figure 11. Software modules and the data flow during the experiments
57
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
58
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
59
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
The designed details of Classifiers development is systems classification technique, it is classified as may or may
described in Figure 2. not be an intruder. Similarly, Level one (L-1) intruder and
Level two (L-2) intruder is an object classified as human and
specifically located in a certain point of a fence, refer to Figure
5.
a b c
Figure 5: Relative intruder position (a) Potential
Intruder, (b) L-1, (c) L-2
60
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Based on the calculated Gaussian value, individual pixel of that have a pixel value of 0, surrounded by pixels that have a
an input data was being group into background pixel or pixel value of 1. Unwanted blob was identified by counting
foreground pixel. The background subtraction process was the number of ones within the detected blob or by considering
implemented. . the minimum blob area. Human blob had greater numbers of
Below was the procedural approach used in foreground ones or greater area compared to unwanted blob which made it
detection: easier to identify and eventually removed from target object.
For each pixel in a video frame: Object contour extraction technique involved data conversion
1. For every N values taken from the pixel or transformation. The method was used to extract the object
2. Find the K Gaussians and weights that best fit to contour in preparation to the object classification. The objects
sample N values using Expectation Maximization outline details were the only considered data, instead of using
(EM) algorithm the entire object blob as object representation, shown in Figure
3. Find the Gaussian with the largest weight and 7(b). Furthermore, silhouettes or object contour was the
store its mean as the value of the background refined object representation taken from object blob.
image for that pixel. Object Silhouettes in Figure 7 (b) is defined by pixel
4. Subtract the background image from the frame. location: P(1,2), P(1,3), P(1,4), P(1,5), P(2,3), P(2,4), P(3,2),
P(3,5), P(4,2), P(4,5), P(5,2) P(5,5), P(6,2), P(6,5), P(7,3)
In the resulting difference image, any value larger than and P(7,4).
three standard deviations from the mean was considered
foreground, and any other value was considered background.
Feature vector extraction and formulation was the final
stage of eliminating unwanted data, refining and finalizing the
target object using several image segmentation techniques. The
technique used in this stage is depicted in Figure 6.
a b
Figure 7 Object Silhouettes extraction
Feature Vector was the data structure that contained all the
unique details of the target object, needed in object
classification. Feature Vector dimension was formulated using
To human Pose the equation 5 and equation 6.
Intruder Classifier ( )
(5)
Figure 6: Feature vector extractions (6)
For simplification, the input data was converted into binary Where: ( )
image, 0 and 1. Zero value was used to represent the unwanted
pixel and 1 for a pixel of interest that composed the structure
of the target objects. Blob detection was applied in input Furthermore, the Feature Vector dimension can be reduced
images to detect points or region that differ in properties like by half of its size, using the distance formula from the equation
brightness or color compared to the surrounding. In this study, of a line (Equation 7) to calculate the unique attribute of the
the value of individual pixels and the minimum blob area, object.
which was 250 unit pixels were the properties used as a basis (7)
for region detection. Blob area analyses were realized by * +, where aiwas the
counting the number of touching pixels or pixels with adjacent individual value represented the object details, shown in
side and have a pixel value of 1. Image re-composition aimed Figure 8 (f) (page 47) the red doted color pixels.
to recreate a whole new image representation concerning only * +, where bi was the
the detected region of interest which generated form blob individual value taken from the point of reference, refer to
detection stage. At this point the processed image still Figure 3.18 (f) the blue doted color pixels.
contained the unwanted data and still considered as noisy The number of raw a from Feature Vector dimension
image but much refined compared to the data produced after ( ) was taken from the total number of object details
applying region filtration. The image morphology correction from Equation 7.
technique used in the project development was a collection of Figure 8, illustrated the logical representation of
non-linear operations related to the shape of target image. The object details reduction, from originally acquired image to
goal of this technique was to correct the target object process or converted image.
imperfections or distorted pixels. Even though, input image Three classifiers were used in the designed project
already undergone pixel correction, still there were parts of the development such as Human Pose Intruder Classifier, Day-time
image that needed for further refinements that were not meet Intruder Localization Classifier and Night-time Intruder
in pass pixel distortion correction technique. Filling of holes Localization Classifier.
was one of the concerns. Image hole was referred to pixels
61
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
The data that were used in mapping were the same The first experiment was done using 1000 image for
data that were used during map development. The result below both human and not human data samples. Human sample was
was depicted according to its level as well as to its horizontal acquired using the same acquisition device used in actual
distance from the camera. experiments and all of the representations contain the pose of
The intruder hit sample result depicted a clear boundary human intruder in different level and different distance. The
between three intruder level classifications. As expected, level data used to represent not human pose were taken from
1 intruder classification would be closer to potential intruder internet. The data were composed of different basic
level and level 2 intruders. Level 2 intruders and potential geometrical shapes and different animal shapes that were
intruder level as shown in Figure 10 and 11, was isolated by a already in masked format. Human Pose classifier that was
clear boundary distinction. The result implies that a potential trained using 10 thousand epochs, painted a hit rate of 0.884
intruder classification could hardly be identified as intruder and a false rate of 0.116. Hit rate value () depict a result of
level 2 and vice versa, but it could possibly be classified as 0.8942857 and false hit rate of 0.105714286 () for
intruder level 1. However, there were few occurrences of Introduction Localization Classifier using 30 thousand epoch.
classification fluctuations as shown in Figure 10 (c), this was Hit rate has a value ranges between 0 1 (0<=<=1). Value
usually happened when the noise reduction process fails to close to 1 means high hit ratings, 0.5 values was the neutral hit
preserved some important object details. rating of the system, implies that system has accuracy of 50
percent (50%). Figure 13 presented the sample images that
were false identified as not human and Figure 14 were the
sample image of falsely identified as human from a set of not
human data samples.
c
Figure 10: Night-time (a) L-1 (b) L-2 (c) Potential Intruder Figure 13: ID as human
Similar to Night-time hit sample results, level 2 intruder Figure 12 and Figure 13 were the result sample during
classification was expected to be the neighbor of intruder level simulated experiment using 10 x 10 map dimension that
1 classification and intruder level 1 classification to potential trained 10 thousand times (10K epoch). To aid the deficiency
intruder classification. Level 2 intruder could possibly be of Human Pose Intruder Classifier, Intruder Localization
classified as intruder level 1 or vice versa. Additionally, Classifier was used to confirm the position of the identified
intruder level rarely could be classified as potential level human. In connection with this, mostly non human objects
intruder, based from Figure 11, hit sample result. have the position located below or distance from the fence and
the human object usually located near or climbing in the fence.
In intruder localization classification, potential identifier
and level one intruder were the false classification possibly
occurs, because of the location which was near to identical.
The reason stated was also true to Level 1 Intruders and
Level 2 Intruders. Classifier test result summary using 30
thousand epoch and 10x10 map dimension was shown in
Table 1 using Human Pose Classifier, Table 2 for Intruder
Localization Classifier Night-time and Table 3 for Intruder
a b Localization Classifier Day-time.
63
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
No. of Positive Negative Hit False Accuracy device. It was because; the acquisition device could not have a
Epoch ID ID Rate Rate (%) clear view or good image representation of the given potential
() () intruder level, due to the physical setup of the acquisition
3K 1519 481 0.7595 0.2405 75.95 device, as shown in Figure 6. The system can only have a
5K 1525 475 0.7625 0.2375 76.25 clear view to the intruder when the intruder starts to climb on
10K 1768 232 0.884 0.116 88.4 the fence, and nearly the situation could be possibly identified
as Level 1 Intrusion.
TABLE 2: INTRUDER LOCALIZATION CLASSIFIERNIGHT- Red line was the target system performance, blue curve was
TIME the approximation of the actual performance of the system,
Result
TABLE 3: USING INTRUDER LOCALIZATION CLASSIFIER and green curve was the average system performance.
Used 700 data sample per Level
DAY-TIME
No. of Positive Negative Hit False
Accur
Result
Epoch ID ID Rate () Rate
acy
Used 700 data sample per Level () (%)
Epoch
10K Positive 195 Hit Rate
1905Negative 0.9071 False0.09 Accuracy
90.71
() Rate ()
No.30K ID 1878 ID 222 0.8942 0.10 (%)
89.42
10K 1828 272 0.8704 0.1295 87.0476
30K 1806 294 0.86 0.14 86
64
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
References
[1] T. Kohonen, The Self-Organizing Map, Institute of
Electrical and Elictronics Engineers, Vol. 78, No. 9, 1990
[2] K. Bapat and S. Mantri, Neural Network Based Face
Recognition Using Matlab, International Journal of Soft
Computing and Engineering,Vol. 1, No. 1, Pp 6-9, 2011.
Figure 18: Intruder L-1 Detection Result of Experiment 3 [3] N. Gupta et al., To Reduce the False Alarm in Intrusion
Detection System Using Self Organizing
65
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Ma.International Journal of Soft Computing and [20] X. Bai et al., Skeleton Pruning by Contour Partitioning
Engineering,Vol. 1, No. 2, 2011. with Discrete Curve Evolution, M.S. theses, China:
[4] J. T.Jimenez et al., Analysis of Performance Metrics Huazhang University of Science and Technology, 2005.
from a Database Management System Using Kohonens [21] P. Kumar et al., Study of Robust and Intelligent
Self Organizing Map, WSEAS, Vol. 2, No. 3. 2003. Surveillance in Visible and Multimodal Framework M.S.
[5] P. A. Aguilera et al., Assessment of Groundwater theses, India: Indian Institute of Technology, 2006.
Quality by Means of Self-Organizing Maps: Application [22] L. Zhao, Liang, Dressed Human Modeling, Detection,
in a Semiarid Area, Environmental Management,Vol. and Parts Localization, Ph.D. dissertation, United States:
30, No. 5, 2002. The Robotics Institute Carnegie Mellon University
Pittsburgh, PA 15213, 2001.
[6] A. Iroala. (2009). Skeleton based visual pattern
recognition: application to table top interaction, Ph.D.
dissertations, Donostia San Sebastian: University of
the Basque, 2009.
[7] A. Mitrokotsa et al., Detecting Denial of Service Attacks
Using Emergent Self-Organizing Maps, M.S. theses,
Greece: University of Piraeus, 2005.
[8] A. Balaz et al., Intrusion Detection System Using Self
Organizing Map, M.S. theses, Slovak Republic:
Technical University of Kosice, 2006.
[9] C. Charnsripinyo et al., Real-time Intrusion Detection
and Classification, M.S. theses, Thailand: University of
Technology Thonburi.
[10] S. Kaski, Sami, The Self-Organizing Map (SOM), M.S.
theses, Finland: Helsinki University of Technology,
(1999).
[11] E. Weippl and K. I. Ismail, Content-based Management
of Document Access Control, M.S. theses,
Autria:Software Competence Center Hagenberg, 1999.
[12] M. Kumar et al., Feature-Based Alert Correlation in
Security Systems Using Self organizing Maps, Ph.D,
Karachi Pakistan: NED University of Engineering and
Technology,2009
[13] S. Kaski and Kohonen, Exploratory Data Analysis by
the Self-Organizing Map: Structures of Welfare and
Poverty in the World Research Project, Finland:
University of Technology Rakentajanaukio, 1995.
[14] I. Yoo, Visualizing Windows Executable Viruses Using
Self-Organizing Maps. M.S. theses,Switzerland:
University of Fribourg, 2004.
[15] M. Kolemainen et al., Reducing energy consumption by
using self-organizing maps to create more personalized
electricity use information, M.S. theses, Finland:
University of Kuopio, 2008.
[16] Moore, David, A real-world system for human motion
detection and tracking, M.S. thesis,California: California
Institute of Technology, 2003.
[17] S. V. Ioannou et al., Emotion reconition through facial
expression analysis based on a neurofuzzy network, M.S.
theses,Greece: National Technical University of Athens,
2005.
[18] N. Aggarwal et al., Vision Based Hand Gesture
Recognition, M.S. theses, World Academy of Science,
Engineering and Technology, 2009.
[19] E. BosKurt and O. S. Mehmet, Ambient Intelligence in
Home Environments, M.S. theses, Turkey:
Telecommunications Engineering Sabanci University,
2004
66
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
AbstractThis study aims to create a simulated and experimental II. DYNAMICS MODEL
of aircraft movements for multirotor quadcopter. The research
method is theoretical and experimental methods. For theoretical In general, a quadcopter described simply as four rotors
method consists of calculating the dynamics and kinematics. that are in a cross configuration. Vertical movement is
While the experimental method consists of the aircra obtained by adding or reducing the speed of the rotor with all
the same value. This movement produces a vertical force U1
(N) to the body frame which will be up or down the
1.751 m/s2, y = 2.038 m /s2 = 1.6371 m dan
quadcopter. Roll motion is obtained by increasing or reducing
the maximum error between the theoretical and the actual
movement is ex = 0.682 m; ey and ez = 0.353 m = 0.546 m. the rotor speed of the left and at the same time reducing or
Theoretical movement pattern already resembles the actual increase the right rotor speed. Pitch motion is obtained in the
movement.. same way on both the other motors.
Front and rear motors turn to the the anti-clockwise
Keywords: model simulation, quadcopter, dynamic control, direction while the two other motor turn to the clockwise, so
kinematics control
that the direction of the yaw motion/ anti-clockwise obtained if
the front-rear speed propeller increase/ decrease and the left
I. INTRODUCTION
and right speed of propeller decrease/ increase.
67
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
F4 F1
4 1
.. .. It appears that MH is a diagonal matrix and a is a
constant. C n i a C i is matrix,
.. 0 0 0 0 0 0
2 F3 z 1 F2
3 .. 0 0 0 0 0 0
.. x 2 3x3 3x3 0 0 0 0 0 0
y C =[ ]= 0 0 0 0 zz yy
(9)
3x3 S( . A )
0 0 0 zz yy xx
3 4 .. [0 0 0 yy xx 0 ]
Fg
Gravitational vector,
Figure 3. Freebody diagram of quadcopter with four rotors. 0
0
Figure 3, shows the dynamics of the movement of a - g
quadcopter. If the aircraft is given an acceleration of the =[ ]= (10)
0
aircraft will change its position and velocity. If 3x1
0
is a ai n nsis ing n n s (x ,y , z , , , ) [ 0 ]
, the new velocity vector at time t is obtained by the
acceleration vector mengintegral quadcopter against t. While If JTP [N m s2] is the total rotational moment of inertia
the new position vector at time t is obtained by integration of h axis and [ ad / s] is h angu a iy
the velocity vector respect to time t. Percepatan yang bekerja the propeller whole, then the propeller gyroscopic matrix,
pada pesawat diakibatkan oleh gaya yang diberikan pada
pesawat. 0 0 0 0
Then, the equilibrium of the dynamics model of 0 0 0 0
quadcopter will be shown in this part. Then it is important for 0 0 0 0
= (11)
analyzing the dynamic equilibrium in the plane so it will know
the forces and moments that can work on the plane.
Furthermore, of the forces and moments acting on the aircraft [0 0 0 0]
will be known acceleration occurs.
If the moment of inertia matrix MH, acceleration If lift force factor b [N s2] and d [N m s2] is the drag
matrix , S n i a C i is matrix CH, velocities matrix factor, then the movement matrix can be shown below,
, GB gravitational vector and the action vector of the
general movement of an aircraft dynamics model of
quadcopter can define in the following matrix form [1],[2]:
MH . + CH . GH = (5)
68
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
0 0 0 0 yy zz U2
0 0 0 0 = +
xx xx xx
= (12) zz xx U3
0 0 = +
yy yy yy
0 0 U4
xx yy
[ d d d d] = +
zz zz
If U1 [N], U2 [N m], U3 [N m] and U4 [N m] are the lift
force, roll torque, torque pitch and yaw torque respectively, On equation (18) shows the mathematical model of
then lift factor b [N s2] and the drag factor d [N m s2 ] then quadcopter.
action vector relative to frame-B,
III. DYNAMICS MODEL OF QUADCOPTER
0
0
0
0 2 2 2 2 This section will be calculated dynamics of the movement of
U1 (1 + 2 + 3 + 4 )
2 aircraft with specifications as shown in Table 1 The calculation
U = = = (24 22 ) (13)
U2 process will be assisted with Matlab program.
U3 (23 21 )
[U4 ] d (2 + 2 2 2 )
[ 2] 4 1 3 Tabel 1.Spesifikasi Pesawat
Para-
Deskripsi Nilai
If rotation matrix then relative action vector aksi relatif meters
L arm 240 mm
respect to frame-H,
b lift coefficient 72,081.10-6Ns
d drag coefficient 3,433.10-6Nms
(s s + s )U1 m mass of quadcopter 1,576 kg
( s +s s )U1 Ixx body inertia moment for axis-x 15,293.10-3Nms
3x3 Iyy body inertia moment for axis-y 15,293.10-3Nms
U =[ ]U = ( )U1 (14) Izz body inertia moment for axis-z 20,307. 10-3Nms
3x3 3x3 U2 JTP rotational moment inertia 0,103. 10-6Nms
U3
[ U4 ] An example of output program using matlab program:
If ck = cos k, sk = sin k then matriks matrix, | 0.1243|
| 0.1397|
s + s s s s + s | 26.0946|
Vektor Aksi Pergerakan, U_H =| -0.0484|
| 0.0003|
=
s +s s s s +s s (15) | -0.0002|
69
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
6
Translation B-F Translation L-R Vertical Translation Actual
Aktual
L-R Theoritical
Teoritik
5
x (m)
3
1
time (s) time (s) time (s)
0
waktu (s)
time (s)
4.5
Actual
Aktual
4
Theoritical
Teoritik
3.5
3
time (s) time (s) time (s) 2.5
y (m)
1
movements
0.5
waktu (s)
time (s)
6
Actual
Aktual
5
Theoritical
Teoritik
a
x
z (m)
i
s 2
z 1
-1
0 0.5 1 1.5 2 2.5 3 3.5
ax waktu (s)
time (s)
i
s
y Figure 6. Position (m) vs time t(s)
x
axi s
In Figure 6, shows the graph plane distance (m) and time (s) in
the x, y and z both theoretical and experiments using matlab
Figure 5.Simulation of aircraft movement program.
70
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
V. CONCLUSIONS
A a i n du h h us an is a g n ugh ha x =
1.750 / s2, y = 2.0377 / s2 and z = 1.6371 / s2
each mileage is x = 0.5565 m , y = 0.6307 m and z = 1.9700 m.
Thus indicating that the aircraft is quite agile in maneuvering.
Quadcopter movement simulation shown that the maximum
error position between the theoretical and the actual movement
is ex = 0.683 m; ey and ez = 0.353 m = 0.546 m. Movement
patterns and the theoretical value of the error is influenced by
the control limit membership function of the fuzzy logic
control design. Theoretical movement pattern similar to the
actual movement.
REFERENCES
71
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
72
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
(8)
Torque at the axis can be shown,
(3)
Torque of motor is,
Robot linear velocity v(t) and the angular velociti of
the robot (t) can be determined based on both the linear
speed of the wheel. In detail, can be shown in matrix presented
in equation below,
(4)
(10)
73
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
(15)
(16)
(21)
A
Lengan
Acceleration of point A obtained by derivate velocity point
C
A respect to time t.
(a) (b)
Figure 6. (a) Model of gears of Tri-Star Wheel (b) Moving at flat surface
When moving on a flat surface then, there are two wheels in (22)
contact with the runway, each wheel just spins on its axis and
each arm is not moving, see figure 6.B. Then, comparison of For dynamic analysis of tri-star wheel by using complex
the angular velocity of the wheels of their gears are numbers, the first step is to describe the free-body diagram.
The second step is to create equilibrium equation based on free
(17) body diagrams. Furthermore, outlining all relevant vectors in
the form of complex numbers.
Consider the free body diagram in Figure 7 the equilibrium
(18) equation is given as follows
74
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
(27)
(28)
(29)
(30)
Torque
Angle vs Torque
75
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
IV. CONCLUSION
WMR robot by tri-star wheel can perform experiments for
uneven surfaces. In addition, the tri-star robot successfully run
straight, turn and climbs stairs. Meanwhile, the average of
velocityof the Tri-star robot is 0.1 m / s. Furthermore, the
controller of this system is done by open loop method using
infrared control system.
ACKNOWLEDGMENT
Thanks to DIKTI (DGHE, Directorate General Higher
Education) Government of Indonesia (GOI) for the
Competitive Grant Universities 2014 and LP2M of Hasanuddin
University.
REFERENCES
[1] Jufri, Sialana , Design Rescue robot for searching, Hasanuddin
University, Makassar, 2010. (Original in Bahasa)
[4] Sakti, Indra et. al, Driving force mechanism for soccer robot, Journal
Information Technology (Journal in Bahasa), Vol. 5 No. 1, 2004
76
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Rahimuddin Samad
Department of Naval Architecture, Engineering Faculty
Hasanuddin University
Makassar, Indonesia
rahimnav@gmail.com
AbstractDetermining the most optimum route is a crucial ships, dangerous coral reefs, or bad weathers in certain area of
factor in ship operation in terms of technical and economical the ocean.
points of view. In the present study, genetic algorithm (GA) is In order to determine the most optimum route of a ship with
applied to determine the most optimum route of a ship with
several destinations, genetic algorithm is adopted as the main
several destinations to visit. The problem and solution are
considered to be similar with the travelling salesman problem
computation method in the present study. The method is one of
(TSP). In GA implementation, the fitness of each individual is popular methods in optimization research field. The flexibility
measured from the distance of its origin and final destination of the method makes it convenient to be adopted in numerous
points. Moreover, several standard genetic operators are also research fields. For the purpose of finding the shortest path, it
implemented in the present study such as selection, mutation and was also implemented by several researchers to the vehicles
crossover. An interactive program is developed with visual basic routing problem [1, 2, 3].
programming language to demonstrate and simulate the As described previously, in the present study, it is assumed
practicability and effectiveness of the method to solve the that the ship has several destinations to visit. The ship needs to
problem. Three cases are simulated in order to analyze the
visit each destination location at least once. Therefore, the
performance of the solution method which are simulations with
varying number of population, probability crossover and problem and solution method in the present study are similar to
mutation crossover. From the computation results, it can be the travelling salesman problem (TSP). The implementation
shown that even though the average fitness of the population is and solution of TSP with GA can also be found in Larranaga et
fluctuating, the most optimum route still can be obtained. al. [4] and Grefenstette et al. [5].
In order to define and solve the problem, ship sailing area is
Index TermsGenetic algorithm, optimum ship route, divided into a grid to make it easier to define the location and
travelling salesman problem, visual basic simulation. distance of the ship inside sailing area. The fitness of an
individual is measured from the distance the ship sails starting
I. INTRODUCTION from origin to the final destination points. In the present study,
As fuel price increases, the ship sailing operation costs also even though it is assumed that the origin point can be from any
get higher. This is not a favorable condition either for ship position inside the grid, it would be easy to extend the problem
operators or for ship passengers in terms of ticket prices and with fixed origin point.
sailing time. Therefore it is an important and crucial task to Moreover, a computer program is developed to simulate the
make the ship operation effective and efficient. process of determining the most optimum route using Visual
There are several ways to address the problem such as Basic programming language. With this program, the user can
installing a fuel-efficient engine, utilizing cheap and renewable define the number and position of all locations that ship needs
energies, designing ship hull forms with less resistance, etc. to visit. For this purpose, the user can determine them by
Another way to resolve the problem is to sail following the manually clicking inside the grid or requesting the application
shortest route. As a result, the engine fuel and sailing time can to randomly determine certain numbers of locations to visit.
be saved. In order to demonstrate the applicability and performance
In the present study, the shortest route is assumed to be the of the solution method and program, simulations are performed
most optimum route. It is also assumed that the ship needs to by varying 3 (three) variables, which are:
visit or reach several locations before the ship arriving at a final Case 1: varying initial population
destination. The reason of the ship to visit or reach several Case 2: varying crossover probability
locations can be to load/unload passengers and cargos at other Case 3: varying mutation probability
ports or simply to avoid static obstacles such as anchoring
77
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
78
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
d12 x2 x1 y2 y1
2 2
where
(x1, y1) = ordinate of the first position
(x2, y2) = ordinate of the second position
Besides the introduction text and input fields, there are also
3 buttons. One button is to define the position labelled
Position Definition, another one is to start the simulation
labelled Start Simulation and third button is in the lower left
corner of the window which is used to quit the program. In
order to start the simulation, the user firstly needs to define the
position by clicking Position Definition button. The program
also shows a message below the Positions Definition button
says Positions not defined when the user is still not defined
the positions.
When the user clicks the Positions Definition button, the
following window will be shown.
Fig. 5. Example of random locations to demonstrate the program.
79
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
As shown in the figure above, the number of positions are In Fig. 7, it can be observed that the average fitness is
chosen to be 20. The positions are created randomly by fluctuating in all 4 values of PopSize. This could mean that not
clicking Random Position button. all individuals in the population change to be the optimum
In order to analyze the effect of genetic variables and ones. The results might be caused by the type of encoding used
performance of the method, 3 (three) cases are analyzed in the in the present study. The integer encoding limits the number of
present study which are varying the number population, chromosome to be modified, hence limits the identities to be
crossover probability and mutation probability. introduced in the population as well.
Moreover, from Fig. 7, it can also be observed that the
1) Varying Number of Population average fitness of the population has the same trend with best
The first case is simulated by computing 4 different fitness where higher value of PopSize will converge faster than
numbers of population which are 5, 20, 50 and 70. The lower one.
crossover probability and mutation probability are fixed to be
75% and 30%, respectively. The number of maximum 2) Crossover Probability effect
generation is chosen to be 1000 generation. The computation The second case is simulated by varying the crossover
results are shown below probability to be 50%, 80%, 90% and 100%. In this case, the
number of population and mutation probability are fixed to be
Best Fitness of Case 1 50 and 30%, respectively. The computation results in terms of
160
best fitness are shown in the following figure
140 Pc = 75 %
Pm = 30 %
Best Fitness of Case 2
120 160
PopSize = 5
PopSize = 20 150
PopSize = 50 140 PopSize = 50
100 Pm = 30 %
PopSize = 75 130
120
Fitness
Pc = 50 %
80 110 Pc = 80 %
100 Pc = 90 %
Pc = 100 %
90
Fitness
60
80
70
60
40
0 200 400 600 800 1000
Generation
50
Fig. 6. Best fitness with varying population size
40
It can be seen from the figure above that optimal results can 0 200 400 600 800 1000
Generation
be obtained when the number of population is 5, 20 and 50.
Moreover, it can also be noted that for these 3 values of Fig. 8. Best fitness with varying crossover probability
PopSize, the solution method can obtain the most optimal
solution faster as the number of population higher. From the figure above, a different trend with the previous
The average fitness for the first case are shown in the case can be observed. In this case, the most optimum solution
following figure is obtained with crossover probability equals to 80% while
higher crossover probability (Pc=90%) obtains the optimum
solution slightly with more number of generations. Moreover,
Average Fitness of Case 1
160 the highest crossover probability (Pc=100%) converges with
Pc = 75 %
140 Pm = 30 % optimal solution lower than results from other 3 values of Pc.
The results show that in genetic algorithm, the higher crossover
PopSize = 5
120 PopSize = 20 probability will not mean the best choice to use in the
PopSize = 50
PopSize = 75 implementation. The best value of crossover probability will
100
depend on the problem we are solving.
The average fitness of the population for 1000 generations
Fitness
60
40
0 200 400 600 800 1000
Generation
80
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Fitness
80 80
60
60
40
0 200 400 600 800 1000 40
Generation 0 200 400 600 800 1000
Generation
Fig. 9. Average fitness with varying crossover probability
Fig. 11. Average fitness with varying mutation probability.
Fig. 9 above shows that the average fitness have the same
trend with the best fitness one as shown in the previous figure. From the figure above, it can be noted that the average
It is also interesting to see that a relatively high peak occurs fitness for the value of Pm=90%, 75%, and 50% have lower
when the crossover probability equals to 5% and 100%. The average fitness compared to their best fitness. The results
peak can be attributed to the randomization implemented in the suggest that many individuals in the population which have
present study. poorer fitness even after many generations. The trend is
different to all other cases previously. The results support the
previous statement that the genetic operator of mutation has
3) Mutation Probability effect
made the convergence of the solution method slower.
The final case is simulated by varying the mutation
From all computations, the most optimum solution is found
probability to be 10%, 50%, 75% and 90%. The computation
to be 51,344. However, there are several routes which have the
results in terms of best fitness are shown in the following figure
same fitness value. One of the route is shown in the following
Best Fitness of Case 3
figure.
160
150
PopSize = 50
140 Pc = 75 %
130
120 Pm = 10 %
Pm = 50 %
110 Pm = 75 %
100 Pm = 90 %
90
Fitness
80
70
60
50
81
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
computation results, it can noted the genetic algorithm method [3] C. W. Ahn and R. S. Ramakrishna, A Genetic Algorithm for
is viable to be implemented in this problem. Moreover, it is Shortest Path Problem and the Sizing of Populations, IEEE
also shown that even though the solution can be obtained, the Transaction on Evolutionary Computation, vol. 6, no. 6 pp. 566-
average fitness of a population is still fluctuating which can be 579, December 2002.
contributed to type of chromosome type of the individual. [4] P. Larranaga, C. M. H. Kuijpers, R. H. Murga, I. Inza, and S.
Another important point in the present study that mutation Dizdarevic, Genetic Algorithms for the Travelling Salesman
Problem: A Review of Representation and Operator, Artificial
probability gives a negative effect to the convergence of the Intelligence Review, vol. 13, pp. 129-170, April 1999.
solution method.
[5] J. Grefenstette, R. Gopal, B. Rosmaita, and D. V. Gucht,
REFERENCES Genetic Algorithms for the Travelling Salesman Problem,
Proc. of the First Int. Conf. on Genetic Algorithms and Their
[1] B. M. Baker and M. A. Ayechew, A Genetic Algorithm for Applications, pp. 160-168, New Jersey, 1985.
Vehicle Routing Problem, Computers and Operations
[6] D. A. Coley, An Introduction to GA for Scientists and
Research, vol. 30, no. 5, pp.787-800, April 2003.
Engineers, World Scientific Publishing, 1998.
[2] S. B. Pattnaik, S. Mohan, and V. M. Tom, Urban Bus Transit
[7] S. N. Sivanandam and S. N. Deepa, Introduction to Genetic
Route Network Design Using Genetic Algorithm, J. Transp.
Algorithms, Springer, 2008.
Eng., vol. 124, no. 4, pp. 368-375, July 1998.
82
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract Failure of fatigue is damaged materials where caused an iterative design cycle is centered on the construction of
frequent load. Fatigue owing to some factors, which is Stress real prototype components. This inhibits the ability to
concentration on fatigue, Stress life, Effect size and surface, and develop new concepts and reduces confidence in the final
Change properties of surface. The fatigue failure of a material is product due to a low statistical sample of tests. It is also
dependent on the interaction of a large stress with a critical flow. In
essence, fatigue is controlled by the weakest link of the material, with
common to find early products released with known defects
the probability of a weak link increasing with material volume. This or product release dates being delayed whilst durability issues
phenomenon is evident in the fatigue test results of a material using were addressed.
specimens of varying diameters. From this research we can get effect A more desirable approach is to conduct more testing
of concentration stress on strength fatigue with S-N method. On this based on computer simulations. Computational analysis can
method only count fatigue life or endurance limit from Journal be performed relatively quickly and much earlier in the design
bearing housing. By Finite Element Analysis, it is not so easy to cycle.
determine fatigue life. When we find the first yield point, it means Confidence in the product is therefore improved because
this point is in the highest stress state. Then we can refer S-N curve.
more usage scenarios can be simulated. It is not
In this paper, the effect of bearing and housing elasticity on the stress
field, which could result in surface fatigue in journal bearing, has recommended, however, that these simulations completely
been investigated. This condition is proved with occurred slip lines replace prototype testing. It will always remain desirable to
on surface of specimen. These slip lines are caused on some have prototype signoff tests to validate the analysis performed
thousands stress cycles. Additional crack is happened immediately and improve our future modeling techniques. However, the
and finally long enough crack. So that formed unstable crack that number of prototype stages, and hence the total
caused fracture of brittleness or fracture of toughness because section development time, can be reduced.
of specimen cannot keep The following subsections are including: bearing in
general, journal bearings, thrust bearing, other types of
Index Terms journal bearing, numerical, fatigue life
bearings, rotor-bearing system. Coupled thermomechanical
non-linear finite element models have been developed to
I. INTRODUCTION study 2D and 3D rolling, and rolling plus sliding contact
While many parts may work well initially, they often fail in problems. The various less or more realistic material
service due to fatigue failure caused by repeated cyclic loading. constitutive models have been used to model behavior of
Characterizing the capability of a material to survive the many bearing materials. The contact stress fatigue is considered as
cycles a component may experience during its lifetime is the a primary wear mechanism. The damage process under
aim of fatigue analysis. In a general sense, Fatigue Analysis contact loading such as, for example, is the cracking,
has three main methods, Strain Life, Stress Life, and Fracture spaling, and tribological reaction, can be study by the finite
Mechanics element method. We can study the mechanics of the sub-
According to independent studies carried out by the Battelle surface or near surface modes of rolling contact failure.
group in 1982, between 80-90% of all structural failures occur In this paper we overview the physical behavior
through a fatigue mechanism and the estimated annual cost of responsible for fatigue stress from initiation to final
fracture and fatigue to the US was 4.4% of GDP. component failure of journal bearing.
Furthermore the Battelle Study concluded that this could be
reduced by 29% by application of current fatigue analysis
technology. II. JOURNAL BEARING
In the past, fatigue analysis was largely the domain of the The bearings are important enough to be studied because if
development engineer, who used measurements taken from the shafts orbit is not stable, or t he bear ing is no t wel l
prototype components to predict the fatigue behavior. This des igned , cont act between the shaft and the bearing will
gave rise to the traditional Build it, Test It, Fix It approach appear
to fatigue design. This approach is known to be very costly as
83
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
The plain journal bearings are fully used in hydraulics due in the notch root. Different methods can be used for this
to their small size, low price, and its capability of carrying conversion, depending on the stress state in the notch root and
load. the applied loading. The most well-known approach is that due
The journal bearing appears in the finite element equations to Neuber which relates nominal elastic values to notch root
as a spring and damper. It appears at one node, linking the stress and strain as = K t n E = C, where C is a
shaft to a rigid structure. It has 4 degrees of freedom.
constant and en is nominal elastic strain. Plane stress is
assumed in this analysis, which can be shown not to be the
case for a circumferentially notched bar. A method for general
stress states is outlined later.
The fatigue strength of a welded component is defined as
the stress range which fluctuating at constant amplitude
Fig 1. FE Equation of Journal Bearing causes failure of the component after a specified number of
cycles (N). The stress range is the difference between the
The first thing we need to do is determine the static maximum and minimum points in the cycle. The number
(mean) position of the shaft in the bearing of cycles to failure is known as the endurance or fatigue life.
m
The expression linking N and R can be plotted on a
logarithmic scale as a straight lineand is r efer red to as a n
S-N curve. The relationship holds for a wide range of
endurance. It is limited at the low endurance end by static
failure when the ultimate material strength is exceeded. At
endurances exceeding about 5-10 million cycles the stress
ranges are generally too small to permit propagation under
constant amplitude loading. This limit is called the non-
propagating Stress.
Finite element technique involves element-modeling
Fig 2. Static Bearing Calculation discretion, which is defined through a displacement function of
each node.
Two important parameters are obtainedfrom the bearing
test: (i) bearing yield (sb,yield) and (ii) bearing ultimate {F}= [k]{D} (1)
(sb,ult) of the material; where bearing stress is defined
with the following relation: sb=P/(Dt). The Handling Finite Element Analysis stress requires a good
yield parameter is defined as the stress at a 2% permanent understanding of the stress-concentration effect, quantified as
hole deformation, which is a definition comparable to the a factor Kt. The theoretical stress-concentration factor is
tensile yield. Bearing ultimate is defined as the first maximum based on a theoretical elastic, homogeneous, isotropic
load peak, which generally was the maximum stress reached. material and can be expressed as:
For the material model, it is assumed that it behaves as an k t = max / nom (2)
isotropic material with isotropic hardening. Uniaxial tensile
test data are simplified into a trilinear behavior, consisting of Handling FEA fatigue stresses correctly also requires
(i) an elastic part, (ii) plastic part up to necking (15% good understanding of fatigue stress-concentration factor
p l a s t i c s t r a i n ) w i t h s t i f f n e s s e q u a l t o 1467.67 By Finite Element Analysis, it is not so easy to determine
MPa, which is followed by (iii) a description of the necking fatigue life. When we find the first yield point, it means this
behavior. point is in the highest stress state. Then we can refer S-N
Fatigue is defined as Failure under a repeated or otherwise curve.
varying load which never reaches a level sufficient to cause
failure in a single application. Fatigue cracks always develop
as a result of cyclic plastic deformation in a lo ca l iz ed area.
This plastic deformation might arise through the presence
of a small crack or pre-existing defect on the surface of a III. RESULT AND DISCUSSION
component, for both cases it is practically undetectable From this research we can get effect of concentration
and unfeasible to model using traditional Finite Element stress or Kt on strength fatigue with S-N method. On this
techniques. method only count fatigue l i fe o r e n d u r a n c e l i m it
The fatigue life prediction follows the strain life approach f r o m Journal Bearing. The lubricant was assumed to supply at
used for notched geometries. Surface g r o o v e s a r e ambient pressure via a full width line groove in the upstream
t r e a t e d a s microscopic notches, where elastic stresses groove. The Reynolds equation was solved using the Gauss-
and strains are converted to local plastic stresses and strains Seidel iterative method with over relaxation factor. The
84
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
a) P = 0 at = 0.
b) P = 0 at = /2 - , and P = 0 at = 3/2 -
c) P = dP/dX = dP/dZ = 0 at = 2 = +
d) P = 0 at z = + b/2.
IV. CONCLUSION
A general approach to modelling the durability of Journal
Bearing has been developed. The approach removes the
requirement of rebuilding FEM models in order to capture the
Fig 4. FStress Consentration for Journal Bearing important stress raising features which significantly affect
fatigue life predictions.
The method is ideally suited for predicting data for fatigue
life calculations in Journal Bearing.
Example applications have been presented
demonstrating some of the capabilities of the method. The
most important point addressed in this work is that the method,
which uses, can be implemented easily to predict fatigue life of
journal bearing by Finite Element Analysis. Moreover, the
algorithm provides robust and fast results because the proposed
method avoids the extra computational burden for
85
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
preprocessing since only stress concentration of the [4] Calculating and Displaying Fatigue Result, Raymond
components are used to predict the fatigue stress of the journal Browell, Al Hancq, Development Engineer ANSYS, Inc,
bearing. The proposed method can identify stress March 29, 2006
concentration, fatigue stress and calculate the fatigue life of the [5] L. Roy and S. K. Kakoty, Optimum Groove Location of
journal bearing. Hydrodynamic Journal Bearing Using Genetic Algorithm,
Advances in Tribology, Article ID 580367, Hindawi
REFERENCES Publishing Corp, 2013.
[1] Fatigue Life and Crack Growth Prediction Using FEM [6] W.A. Gross, et. all., Fluid Film Lubrication, John Wiley
Data, Robert A Adey, John M. W. Baynham, Sharon & Sons, New York, 1980.
Mellings, Tom Curtin, Computational Mechanics BEASY, [7] F.A. Martin, C.S. Lee, Feed Pressure Flow in Plain
Ashurst Lodge, Southampton, Hampshire, SO40 7AA, UK Journal Bearings, Trans. ASLE. Vol. 26, No. 3 (1982)
[2] Fatique Life Prediction of machined components using 381-392.
finite element analysis of surface topography, S. K. As, B [8] E.S.D.U. Item No. 84031, Calculation Methods for
Skallerud, BW Tveiten, B Holme, Science Direct, Steadily Loaded Axial Groove Hydrodynamic Journal
Elsevier, International Journal of Fatigue, 27 (2005) 1590- Bearings, E.S.D.U., London, 1984.
1596 [9] F.P. Brito, J. Bouyer, M. Fillon, A.S. Miranda, Thermal
[3] Finite element analysis of machine elements, A behaviour and performance characteristics of a twin axial
bibliography (1977 1997), Jaroslav Mackerle, groove journal bearing as a function of applied load and
Engineering Com, Vol. 16 No. 6, 1999, MCB University oil supply temperature, TRIBOLOGIA - Finnish Journal
Press. of Tribology 3 vol 25 (2006) 24-33.
86
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract This research is designed and made prototype Unilateral action gripper; contact with the surface
articulated type robot arm with five degrees of freedom (5-DOF). of the object. (Method: vacuum, magnetic and
Where the end-effector is a gripper is used that has two claws. This principles of adhesion
robot arm joint rotation has 5 pieces / revolute which serves to
connect the links that one with the other link. Each joint is restricted Bilateral action gripper; contact with the surface of
1 degrees of freedom. It aims to facilitate the mechanical and control the object 2. (Method: clamp)
system of the manipulator. As the actuator (driver) joint use stepper Multilateral action gripper; more than 2 contacts
motors, which required the driver to control (control circuit) with the surface of the object. for example, multijointed
input data in the form of computer commands communicated with the fingers (method: gripping).
system working port. Parallel communication or control system of
the manipulator is to adjust the angular position of each joint with Mechanical system illustrates how the shape of the robot
computer program created with Visual Basic 6.0 language, which is
hand and the type of components to be used.
working parallel to the driver circuit so that the driver circuit
generates pulses to drive the stepper motor. Mechanical design is the principle that determines the
skill/dexterity of a robot arm. 3 basic aspects that must be
Index Terms degree of freedom, robot arm, manipulator considered in designing a robot arm:
The number of rods (link)
I. INTRODUCTION Number of connections (joint linear / prismatic and
revolute joint)
Products rapidly evolving field of technology today is
The size and movement of each rod (degrees of
robotics. The reason is because the robot can be designed to freedom / DOF)
work on the process of working with high accuracy (quality) as
well as having more job productivity (quantity) that can help End Effector
Joint
increase the volume of production.
One of the advantages of the robot is able to be placed for
the operations that are considered at risk (high-risk) to be done Links
by humans, such as in the area of high temperature, hazardous
chemicals, toxic gases, radioactive risks, including a mission in Fig 1. Schematic Diagram of Robot Arm
outer space. In addition, the robot is able to do continuous
process repeatedly deemed too tedious to be done by humans. Robot arm control system in general is a microprocessor-
The work process includes a series of robot control system that based electronic circuit that functions as a regulator of all
enables a robot to take orders, run the program works, and components in shaping the work function.
produces both interrupt and process repeatedly. One of the
The control system generally has three functions:
robot controller is a computer that is often used.
The robotics system is being developed not to replace the entire Opening and closing movement of the manipulator
human role, but rather help people to more easily, quickly and components at the desired stage and specific point.
comfortably in his work. Storing data in memory positions and phases.
Allow the robot to be able to interface with a particular
II. ROBOT COMPONENT device.
The main components of the robot arm generally has three Broadly speaking, the motion control system in robotics have
important parts, namely manipulators, mechanical systems, and structured parts, which include:
control systems. Driver
Possition end-
Manipulator is a mechanical part that can be used to move, effector Processing
Input Output
lift, and manipulate the work piece. Function depends on the (Volt) (Radian)
type of robot end-effector mounted on the manipulator. End-
effector can be a device welding, painting, machine tools or a Fig 2. Control System of Robot Arm
gripper which has a movable jaw opens and closes. Gripper can
be classified into 3 types:
87
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Robot
kontroller (elektronik & mekanik)
ri + e u y
G(s) H(s)
-
88
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
The magnitude of the maximum displacement distance of Table 3 Results of Measurement Circuit Voltage Driver
each link: condition Voltage
Saturation 0V
Cut Off) 21 V
Voltage in 21,5 V
In the same way it will be obtained:
Analysis: From the results of stress measurements obtained
If R2 = 100 mm, then the S2 = 0.209 m
parallel port voltage value logic 0 and logic 1 are the normal
If R3 = 109 mm, then S3 = 0.285 m limits, ie, logic 0 in the range 0 s / d 0.8 V and logic 1 in the
If R4 = 130 mm, then S4 = 0.170 m range of 2.4 s / d 5 V. therefore concluded parallel port can
If R5 = 25 mm, then the S5 = 0.039 m work well.
Voltage standard deviation value of less than 5%, so that
The amount of energy the maximum movement of each link: the circuit can be concluded worked well.
E=FxS=mxgxS Test Results: Form programs appear and function properly.
Analysis: Program under normal circumstances, because if it
E1 = 0,164 Joule E4 = 0,291 Joule detected an error in the program after the Run, then the form
E2 = 0,188 Joule E5 = 0,036 Joule and a visual basic program that has been made cannot be
E3 = 0,710 Joule displayed.
89
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
AbstractIn this research, we present a control method using gradient which is calculated from the HPF. The system input
kinodynamic motion planning based on a harmonic potential field U = [U1 U2 U3 U4 ]T , which is constructed by nonholonomic
(HPF) for an X4-Flyer moving in a 3-dimensional space. In the control input uc and control input u based on the gradient
previous research, it was confirmed that a controller using two of the HPF, is as follows:
HPFs generated on the X-Y and X-Z planes was able to guide the
X4-Flyer to the arbitrary target point in a 3-dimensional space. U = uc + u. (1)
In this paper, the previous method is extended to the case where
three HPFs generated on the X-Y, X-Z, and Y-Z planes are used, Here, U1 is a control input for acting on each translational
and it is verified that the X4-Flyer can move efficiently by using motion, and U2 , U3 and U4 are control inputs for acting on
the proposed method through some simulations. roll angle , pitch angle and yaw angle respectively. In
the following subsections, we describe the control input based
I. I NTRODUCTION on nonholonomic control uc and the proposed control input
Recently, there are many researches on the autonomous u based on the gradient of an HPF.
locomotion for an X4-Flyer[1]-[3]. An X4-Flyer is a vertical
takeoff and landing (VTOL) aerial robot with four rotors, and A. Nonholonomic Control Input
it has received attention in recent years as search and rescue
robots because of its highly maneuverability and hovering The control input uc = [uc1 uc2 uc3 uc4 ]T is added for z
ability. In traditional motion planning, dynamic constraints and direction and three attitude angle and given as follows[7]:
kinematic constraints are generally solved separately. In that
uc1 = cos mg cos cos cos
mU1
case, dynamic constraints are mostly solved by designing the
control input according to the result of kinematic constraints. uc2 = Ilx ( T ) k1 (2)
There is kinodynamic motion planning that is aimed at I
uc3 = ly ( T ) k2
solving these two constraints simultaneously, for designing the
control input from the current state[4]. Kinodynamic motion uc4 = Iz ( T ) k3 .
planning can design the control input in one-step, and therefore
Here, U is :
it has an advantage of being able to decide the control input
simply, compared to existing motion planning. There are many U = k4 (z zT ) + k5 z. (3)
techniques for kinodynamic motion planning, and kinodynamic
In these equation, k1 -k3 are positive constant gains, and zT is
motion planning based on using Harmonic potential field
an arbitrary altitude and (T , T , T ) are the desired angles.
(HPF) was proposed as one of them[5][6].
In this research, a control method is proposed for guiding
B. Added Control Input
an X4-Flyer to an arbitrary target point in a 3-dimensional
space. The present controller guides an X4-Flyer by appropri- In this subsection, an added control input u is described
ately switching the HPF generated in a 2-dimensional plane. for the translational motion. Kinodynamic motion planning
It was already confirmed in the previous research that a which is proposed in the previous research can only guide an
controller using two HPFs generated on the X-Y and X-Z X4-Flyer on the X-Y plane, because it uses an HPF generated
planes can guide the X4-Flyer to the arbitrary target point in the X-Y plane. The proposed method can control the Z
in a 3-dimensional space. In this paper, the previous method direction of an X4-Flyer by switching an HPF including the
is extended to the case where three HPFs generated on the Z direction and an HPF in the X-Y plane.
X-Y, X-Z, and Y-Z planes are used. Moreover, it is verified
by simulations that the X4-Flyer can move to the arbitrary For the control in the X-Y plane, the HPF on the X-Y
target point efficiently by using the present method based on plane including the current position of the X4-Flyer is used.
2-dimensional HPFs. At that time, using the gradient of the HPF on the X-Y plane,
an added control input u is designed by
II. K INODYNAMIC M OTION P LANNING FOR AN
bc x kv V(x) kc FC (x, x)
X4-F LYER
u = if < (x xT )2 + (y yT )2 (4)
In the proposed method, kinodynamic motion planning is
b d h(x, x) kv V (x)
achieved by combining nonholonomic control input and the otherwise,
90
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
h(x,
[ x)
Z
= nT xn+ Current position of an X4-Flyer
( ) ]
V (x)T V (x)T
V (x) x (V (x)T x) V (x) ,
An HPF on the X-Y plane
FC (x, x)
= (xT x) ( |xT x|) (xT (xT x)).
Y
where bc , bd , kv and kc denote a positive constant gain,
h(x, x) = [0 h(y, y) h(x, x) 0]T , x = [0 y x 0]T , V (x) =
X An HPF on the Y-Z plane
[0 fy fx 0]T , and FC (x, x) = [0 FC (y, y) FC (x, x) 0]T . An HPF on the X-Z plane
Note that, fx and fy mean the gradient of HPF parallel to the
direction of x- and y-axis respectively. Fig. 1. A schema of generating 3-directional gradient using two HPFs.
For the Z-directional motion control, the HPF on the X-Z
or the Y-Z plane including the current position of the X4-Flyer
is used. In this control, the X4-Flyer moves in the Z-direction
by subtracting the Z-directional gradient fz calculated with the 50
HPF from the current altitude z and regarding it as the target Obstacle
40
altitude zT in Eq. (3), so it follows that
30
Z [m]
U = k4 {z (z fz )} + k5 z. (5)
20
10
planes including the current position of the X4-Flyer as shown 50 Current position
40
(45, 25, 10) 35
in Fig. 1. By giving the X-, Y-, and Z-directional gradients 40 30
Y [m]
to the X4-Flyer, the controller guides the X4-Flyer in the 3- 25
Z [m]
30
15
1 are X-, Y-, and Z-directional gradients, respectively. 10
40
50
10
0 30
0 5
10 20
20 10 Y [m]
As an illustration, consider the case where the X4-Flyer X [m]
30
40
50 0
0
0 10 20 30
X [m]
40 50
moves from the current position (x, y, z) = (45, 25, 10) to the
(a) X-Y plane including the (b) The gradient of the HPF
target position (xT , yT , zT ) = (5, 5, 10) in the environment
current position on the X-Y plane
shown in Fig. 2. In this case, the HPF on the X-Y plane
including the current altitude z = 10 (see Fig. 3 (a)), the HPF
on the X-Z plane including the current y-position y = 25 (see Fig. 3. The HPF on the X-Y plane.
Fig. 4 (a)), and the HPF on the Y-Z plane including the current
x-position x = 45 (see Fig. 5 (a)) are generated. In order to
generate the HPFs, (xT , yT ) = (5, 5) is set as a target point
in Fig. 3, (xT , zT ) = (5, 10) is set as a target point in Fig. 4, 50
45
and (yT , zT ) = (5, 10) is set as a target point in Fig. 5. Then, 40
25
Z [m]
30
X4-Flyer toward the target point. 20
20
15
10 40 10
0 5
20
0 10 0
III. S IMULATIONS 20 30 40
Y [m] 0 10 20 30 40 50
X [m] 50 0 X [m]
In this section, it is confirmed that the proposed method (a) X-Z plane including the (b) The gradient of the HPF
current position. on the X-Z plane.
can guide the X4-Flyer 3-dimensionally in the environment
shown in Fig. 2, by simulations in MATLAB. Additionally,
the proposed method is compared with the previous method
Fig. 4. The HPF on the X-Z plane.
to verify that the proposed method is more efficient than the
previous one.
91
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
50
45
40
50
35
40 Current position
(45, 25, 10) 30
Z [m]
30 25
Z [m]
20
20
15
10 50 10
5
0
0 10 20
Y [m] 0
30 40 0 0 10 20 30 40 50
50
X [m] Y [m]
(a) Y-Z plane including the (b) The gradient of the HPF
current position on the Y-Z plane
Fig. 6. The X4-Flyer.
2 4 1 3
= 2 + 4 1 3 .
20
Here b is the thrust coefficient and d is the drag coefficient.
The parameters used for this simulation are shown in Table 10
92
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
using the proposed method, and the dotted blue line denotes the
trajectory when using the previous method. Fig. 8 shows the
change of the Euclidean distance between the current position
Distance between the current position
25
proposed method and the target position, and Fig. 9 and Fig. 10 show the
and the target position [m] previous method change of the attitude angles (, , ) when using the proposed
20
method and the previous method respectively.
C. Discussions
15
It is found from Fig. 7 that the X4-Flyer was able to
reach the target point from the initial state while avoiding
10 the obstacle in the 3-dimensional environment by using the
proposed method or the previous method. Moreover, the X4-
Flyer was able to move 3-dimensionally by using the gradient
5
of HPFs in spite of only using the 2-dimensional HPFs.
However, the trajectory when using the previous method takes
0 a long detour to upside of the obstacle. On the other hand,
0 50 100 150 200 250 300
the trajectory when using the proposed method takes a shorter
Time [s]
route from the side of the obstacle and it seems that an efficient
trajectory was chosen.
Fig. 8. Distance from the target position.
As shown in Fig. 8, the X4-Flyer was able to keep its
attitude on the target point after reaching the target point
even if which controller was used. However, the proposed
method was able to guide the X4-Flyer to the target point
about two times faster than the previous one. Small steady-state
0.05 errors remained after the X4-Flyer reached the target point,
irrespective of the control method adopted.
0.04
0.03 All the attitude angles of the X4-Flyer always fell within
0.02 0.04 [rad] as shown in Figs. 9 and 10. This result is acceptable
Angles [rad]
0.04
IV. C ONCLUSIONS
0.05
0 50 100 150 200 250 300
In this paper, a method for guiding the X4-Flyer to an
Time [s] arbitrary target point in a 3-dimensional space is proposed by
switching the HPFs, which are generated on 2-dimensional
Fig. 9. Angles of the X4-Flyer using the previous controller. planes. The usability of the proposed method was confirmed
in simulations by comparing it with the previous method.
In future works, we are going to implement the proposed
method to an actual machine, and make experiments in actual
environments.
0.05 R EFERENCES
0.04 [1] Y. Bouktir, M. Haddad, T. Chettibi, Trajectory planning for a quadrotor
0.03 helicopter, in Proc. of 16th Mediterranean Conf. on Control and
Automation, pp. 12581263, Jun. 2008.
0.02
[2] P. Bristeau, F. Callou, D. Vissiere and N. Petit, The navigation and
Angle [rad]
0.01 control technology inside the AR.Drone micro UAV, in Preprints of the
0 18th IFAC World Congress, pp. 14771484, 2011.
0.01 [3] Y. Suh, Robust control of a quad-rotor aerial vehicle, Int. J. of Applied
Electromagnetics and Mechanics, vol. 18, pp. 103114, 2003.
0.02
[4] B. Donald, P. Xavier, J. canny, and J. Reif, Kinodynamic motion
0.03
planning, J. of the ACM, vol. 40, no. 5, pp. 10481066, 1993.
0.04 [5] A. Masoud, Kinodynamic motion planning, IEEE Robotics & Automa-
0.05 tion Magazine, vol. 17, no. 1, pp. 8599, 2010.
0 50 100 150 200 250 300
Time [s] [6] K. Motonaka, K. Watanabe, and S. Maeyama, Offline Optimization of
Gains for Kinodynamic Motion Planning for X4-Flyer, SICE Annual
Conference, 2013.
Fig. 10. Angles of the X4-Flyer using the proposed controller.
[7] B. Samir, M. Pierpaolo and S. Roland, Design and control of an indoor
micro quadrotor, IEEE Robotics and Automation, Vol. 5, pp. 43934398,
2004.
93
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
I. INTRODUCTION
In recent years, unmanned aircrafts are expected to play an Fig. 1. Definition of the coordinates
important role in observing vegetation and gathering
information on disaster sites etc.[1] where it is hard for human When the airframe is inclined, the righting moment prevents
to enter. Especially, airships that can float by its own buoyancy the X4-Blinp from being controlled, because the conventional
are attractive for good energy efficiency to travel for long time. X4-Blimp has the center of gravity and the center of buoyancy
However, a big airship requires a wide space and cost for in different point. This paper proposes a new X4-Blimp, which
maintenance. Thus, small airships which are called blimp is symmetric in structure. The new X4-Blmip can fly stable
have been developed, because it is easy to maintain and use it. because the new X4-Blmip has the center of gravity and the
Most of existing airships have propellers and rudders for center of buoyancy in the same point. A method for controlling
controlling them. In this control method, the airframe is the X4-Blimp by switching two controllers is adopted, one of
controlled by the rudders, using the airflow flowing on its which is designed by using a model that includes nonlinear
surface. Such a method has a problem that if the traveling speed parts or a model that only includes linear parts, where those are
is slowed down, then the operability is deteriorated because of separated from the derived dynamical model. The effectiveness
the weak airflow. Thus, it is desired to develop a blimp of the proposed method is verified by some simulations.
controlled without using rudders.
II. OVERVIEW OF THE X4-BLIMP
In this research, a controller method is proposed for an X4-
Blimp where the airframe is controlled by only four propellers A. Structure of the X4-Blmip
without any rudders. Since the X4-Blimp can control the The X4-Blimp proposed in this research is composed of
positions and attitudes in three-dimensional space by regulating envelopes, a gondola and propellers as shown in Fig. 1. The
the output of the propellers, it can realize high operability, envelopes is filled with helium gas to balance airframe mass
irrespective of its traveling speed. However, it is not easy to with the buoyancy. The envelope form is a spheroid to decrease
control the X4-Blimp, because it is an underactuated system. air resistance for traveling direction. The gondola includes
From an actual experiment, we have found that it was hard for batteries and controllers, and it is placed on the center of the
a conventional X4-Blimp [2], in which the envelope is placed airframe. The gondola form is a rectangular solid to maintain the
at the upper part of the airframe whereas the gondola is placed space for the controllers etc. and simplify a calculation of the
at the lower part of the airframe to fly downward. moment of inertia. The four propellers are attached on up, down,
left and right sides of the gondola with the same distance from
the center of the gondola. This airframe is designed
symmetrically at a point C so as to be controlled easily.
94
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
95
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
i = i ( = 1,2,3,4), = ( = 5,6) (22)
1
The Eq. (22) is rewritten as follows
1 = 2 (23)
)
2 = (1 + 2 2 1 2 1 (24)
36 = (1 + 2 ())36 + 2 (25)
where 36 = [3 , 4 , 5 , 6 ] . Here, 1 , 2 () and B are
denoted by
0 1 0 0 0
0 0 0 0
1 = [ 0 ] , = [ 1]
0 1 1 0
12 0 0 1 0
0 0 0 0
0 0 0 0] Fig. 3. Structure of energy regions
2 () = [ 0 0 1 0
(1 + 2 ) 0 0 1 V. ENERGY REGION BASED SWITCHING METHOD
where = 21 + 1 . The controllability of [1 , ] is confirmed. Switching the two partial underactuated controllers for 4
A controllable matrix is represented as [ 1 12 13 ]. It is inputs 10 states is considered to control an underactuated system
regular because 1 > 0 . Since 1 + is controllable, the with 4 inputs 12 states. However, if input chattering phenomena
feedback gain = [1 , 2 , 3 , 4 ] is calculated to make matrix occur when controllers are switched, an excessive burden is
1 + as the Hurwitz matrix by the pole placement method. placed on motors. Therefore, a switching method[5] that has
The control input 2 is denoted by multiple boundary regions is used to prevent the chattering
phenomena.
2 = 36 = 1 3 + 2 4 + 3 5 + 4 6 (26)
The energy is defined from the errors of generalized
Thus, since it can be stabilized to the origin, the control input for
coordinates. Since the state x is doubly generated from the set of
the chained form are derived as follows
(x, , y) and (x, , z), and similarly the corresponding attitude
1 = (1 + 2 ) 1 2 (27)
angle is also doubly generated from the set of (, ) and (,
2 = 1 tan + l2 + 3 + 4 (28) ), the errors for the stabilization to the origin are directly
cos 2 represented by , y, and z because both partial underactuated
In this way, the controller for 2-input/6-state partial model for x, controllers always stabilize the state x and the angle to the
and y is designed. Next, the controller for the 2-input/6-state origin. Then, the energy based on the errors is defined as
partial model for and is designed by a linear feedback such follows:
as 1 = 2 + 2 (35)
2 = 1 2 (1 , 2 > 0) (29) 2 = 2 + 2 (36)
3 = 3 4 (3 , 4 > 0) (30) In Fig. 3, a twodimensional plane is represented by 1 and 2 ,
The partial underactuated controller 1 for a model with 4 input and hysteresis like boundary lines 1 and 2 to separate the
and 10 state is designed by combining the controllers for x, energy plane are represented respectively by
and y with the controller for and . 1 (1 ) = 1 1 (37)
Similarly, the partial underactuated controller 2 is designed 2 (1 ) = 21 (38)
by combining the controller for the 2-input/6-state partial model In Fig. 3, the partial underactuated controller 1 is used on the
for x, and z with the controller for the 2-input/4-state partial region 1 , whereas the partial underactuated controller 2 is used
model for and . When the partial model for x, and z is on the region 2 . Considering an overlapped region, switching
transformed to a chained form, the input transformation is rules are decided as follows:
denoted by Rule 1:
1 = (1 + 2 ) 1 2 If 0 < 2 1 (1 ) then =
tan Rule 2
2 = 1 ( ) + 2 ( ) + 3 + 4
cos cos cos 2 If 1 (1 ) < 2 < 2 (1 ) and 1 = then =
The control inputs based on the chained form transformation is Rule 3:
denoted by If 1 (1 ) < 2 < 2 (1 ) and 1 = then =
1 = 1 (31) Rule 4:
2
2 = cos cos 2 2 tan 2
(32) If 2 (1 ) < 2 then =
The 2-input/4-state partial model for and is derived by a Where st represents the controller used for each rule. When
linear feedback such as = , the partial underactuated controller 1 is used, whereas
2 = 1 2 (1 , 2 > 0) (33) when = , the partial underactuated controller 2 is used. st1
4 = 3 4 (3 , 4 > 0) (34) represents the controller used before one-sampling time.
The partial underactuated controller 2 for the model with 4 According to this switching rule, the partial underactuated
inputs and 10 states is designed by combining the controller for controller 2 is used to control the state z. Similarly, the partial
x, and z with the controller for and . underactuated controller1 is used to control the state y. It should
be noted that, in this switching rule, the chattering phenomena
96
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
97
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract The primary emphasis of this work is on the Automated sorting systems are classified into mechanical
development of a new waste beverage cans identification method and optical systems. Since 1932 to 2009, different mechanical
for automated beverage cans sorting systems known as the SVS and optical sorting methods have been developed to fill the
system. The method described involved window-based demand of object sorting. Mechanical sorting cannot achieve
subdivision of the image into X-cells, construction of X-candidate
commercially viable throughputs and accuracy. The popularity
template for N-cells, calculation of matching scores of reference
templates for the N-cells image, and application of matching of optical sorting systems has increased because of inadequate
score to identify the grade of the object. The SVS system throughput of mechanical sorting systems. The greatest
performance for correct beverage cans grade identification is advantages of optical sorting systems include the following:
95.17% with estimated throughput of 21,600 objects per hour consistent and reliable production efficiency with a relatively
with a conveyor belt width of 18. The weight of the throughput high hit rate and purity; and low operational cost because of
depends on the size and type of the objects. fewer manual workers on the production line.
The main objective of the research is to develop a smart
Index Terms Identification, SVS, beverage cans, sorting
vision sensing (SVS) system for automatic recyclable waste
system
beverage cans sorting. More specifically, the aims is To select
the best features and classifier for the smart vision sensing
I. INTRODUCTION (SVS) system for automatic recyclable solid waste sorting.
Computer vision (CV) deals with extracting meaningful
descriptions of physical objects from images (Ballard & Brown
1982, Brosnan & Da-WenSun 2003, and Kulkarni 2001). Due II. THE SVS SYSTEM FOR SOLID WASTE SORTING
to low cost powerful solutions, the applications of CV have Figure 1 illustrates the block diagram of the proposed
increased tremendously in diverse fields such as medical intelligent computer vision system for automatic sorting of
diagnostic imaging, food industry for quality evaluation, recyclable beverage cans and a picture of the actual systems.
factory automation, robot vision, object identification, military The computer vision process consists of three parts: perception,
reconnaissance, remote sensing, mineral exploration, cognition and action. The perception or image acquisition
cartography, and automated object grading and sorting. The portion of this vision system consists of a commercially
aim of this motivation is to realize the necessity of the available webcam and a special lighting arrangement. The
automated solid waste sorting system and justify the main responsibility of the action component of the vision
development of a smart vision sensing (SVS) system for system is to segregate waste beverage cans into different types
automated recyclable waste beverage cans sorting using state- based on the command of the cognition part of the vision
of-the-art of the CV. system. Mechanical system are used to segregate and to pile
The primary challenge in the recycling of beverage cans is different type of object according to their respective waste bins.
to obtain raw material with the highest purity. In recycling, In this research, we emphasize a beverage cans type
highly sorted stream facilitates high quality end product, and identification system, which covers the perception and
save processing chemicals and energy because various grades cognition components of the proposed system.
of beverage cans are subjected to different recycling processes. In this proposed system, 640480 RGB images are
In addition, the amount of sludge and rejects generated in captured from inspection zone on the conveyor belt by using
recycling processes is decreased for the utilization of sorted Logitech QuickCam Pro 4000 Web Camera [46], [47]. The
object in recycling as well as reduces the amount of energy specifications of the webcam are CCD Optical sensor, color
needed to produce recycled cans. In this work, the type of a Camera, CCD Image Sensor Lens Construction support
beverage cans is based on weight, color, usage, raw material or Manual Focus Adjustment, 4 pin USB Type A Interface for
a combination of these factors.
98
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
USB Expansion / Connectivity with Computer Interface. In provided any improvement over RGB. Thus, the RGB color
webcam properties setting, the brightness, contrast and space is considered in this research. For color images, each of
saturation are adjusted at 50%, 50% and 100% of their the three color components red, green and blue are
respective scales. considered separately. For gray scale image, standard grayscale
. transformation is obtained from the original RGB image.
Identification is primarily based on the dominating color level
of the objects. In the feature selection process, special emphasis
is placed on those features, which provides significant
information regarding the dominant color level. Initially,
seventeen first order features, such as mean, standard deviation,
skewness, kurtosis, dispersion, lowest color level, highest color
level, mode of the color level, entropy, energy, lower quartile,
upper quartile, histogram tail length on dark side, histogram tail
length on light side, median color level, range of the color
level, and inter-quartile range, are extracted from the image
using equations to determine the significant features in
identification.
To calculate the first order features, the gray level
histogram of the image is calculated first. The histogram, h(x),
Figure 1 Block diagram of the intelligent computer vision
is a one dimensional array that represents the number of pixels
system for automatic sorting of Beverage Cans
in the image with a gray level of x. The x parameter can take
any value between 0 and Z-1, where Z is the number of gray
In this experiment, it is observed that the performance of
levels in the image. For color images, three histograms are
the vision system is extremely influenced by the lighting
calculated for the three color components: red, green and blue.
arrangement. For calibration and adjusting the lighting, three Z 1
Z
illumination, and diffuse front lighting are considered in this
research as shown in Figure 2. In both front lighting- Z 1
[2]
Standard Deviation, x 0
the images from the inspection zone show some reflection Z
problems such that the object on the conveyor. Moreover, the
Z 1
reflection from the surface of the object is not uniform. It is
important that the texture information of the objects is
( h( x ) ) 3
[3]
Skewness x 0
[4]
lighting. In front lighting-directional-darkfield illumination, Kurtosis x 0
Z 4
image from inspection zone is distinctive for texture analysis
and the object surface is illuminated uniformly. Moreover, Z 1
front lighting-directional-darkfield illumination is widely used Dispersion h( x) [5]
in surface scratches or texture analysis (Pham D.T., & Alcock, x 0
UpperLimit
Figure 2: Lighting Techniques: (a) brightfield illumination, h(x).log 2 (h(x))
[7]
(b) darkfield illumination and (c) diffuse frontlighting Entropy x LowerLimit
tpixelsp
III. FEATURE EXTRACTION Where tpixelsp is the total number of pixels used to
In the feature extraction phase, both color and gray scale calculate entropy
UpperLimit
images are considered. Brunner et al. (Brunner, C. C.,
Maristany, A. G., Butler, D.A., Leeuwen D.V., & Funck, J.W.,
x .h( x)
2
[8]
Energy x LowerLimit
1992) converted the usual RGB color space into other tpixelsg
potentially more useful color spaces, but they found that none
99
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
100
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
REFERENCES
[1] J. Petek and P. Glavic, An integral approach to waste
minimization in process industries, Resource,
Conservation and Recycling, Elesevier, 17: 169188,
1996.
[2] WasteCap. (2008), WasteCap of Massachusetts, 68
Hopkinton Road, Westboro, MA 01581, March 2008,
available
at:[http://www.wastecap.org/wastecap/commodities/paper
/paper.htm]
[3] Paper Grades (2009), available at:
V. CONCLUSION [http://www.paperonweb.com/ppmanf.htm], (accessed
The primary emphasis of this work is on the development February 2009).
of a new waste beverage cans identification method for [4] M. K. Ramasubramanian, R. A.V enditti, C. M.
automated beverage cans sorting systems known as the SVS Ammineni and M. Mallapragada, Optical Sensor for
system. Noncontact Measurement of Lignin Content in High-
Another important idea that has been implemented is the Speed Moving Paper Surfaces, IEEE Sensors Journal,
adaptability to new subcategories of the primary Object grades. 5(5), pp. 1132-1139, 2005.
The wide range of subcategories of object grades is used to [5] A. G. Doak, M. G. Roe, and G. R. Kenny, Multi-Grade
train the system to recognize new subcategories, and as a result Object sorting system and method, US Patent
the system is scalable and able to provide robust decisions for No.7173709, (2007).
object identification tasks. Besides, the method was trained [6] A. G. Doak, M. G. Roe, and G. R. Kenny, Multi-Grade
with many reference templates using different lighting Object sorting system and method, US Patent No.
US2007/0002326, (2007).
101
International Journal of Smart Material and Mechatronics Vol.1 No.1 2014 ISSN: 9772356-531002
Abstract This study aims to find out of the cooling performance Seebeck's discovery inspires Jean Charles Athanase
of thermoelectric coolers with single, double series, and double Peltier to examine the opposite of the phenomenon. He flows
parallel circuit. The experiment was conducted in the Cooling the electric discharge on two metal pieces glued together in a
and Heating laboratory of Mechanical Engineering Department, series. When electrical power is applied, the heat absorption
Hasanuddin University, Makassar. The data taken were hot side
occurs at the junction of the two metals and heat release in
temperature, cold side temperature, water temperature, and
ambient temperature. Data analysis was carried out on water another connection. This heat release and absorption revert
temperature, temperature difference, absorbed heat, and COP each other when the current is reverted . The discovery which
with some variations of thermoelectric circuit and DC electric occurred in 1934 then known as the Peltier effect [3]. Seebeck
voltage in 360-minute period. The result reveal that the best and Peltier effect is then the basic for the development of
module was the double thermoelectric arranged with a series thermoelectric technology.
circuit in the voltage of 10 V. This could be seen after 360 Simple mode of cooling is by using a thermoelectric
minutes with cold water temperature of 12 oC, temperature device. However, due to the limits of thermoelectric materials
difference of 28oC, absorbed heat of 19.52810 and COP of performance, one degree of the thermoelectric cooler machine
1.25268.
can only be operated with a small temperature range. If the
Keywords: thermoelectric cooler , water temperature, DC electric temperature ratio between the heatsink and cooling is large,
voltage. then the coolant engine with one degree of thermoelectric will
lose its effectiveness. Thus, the application of thermoelectric
I. INTRODUCTION with two or more levels are combined in the coolant engine is
an important method to improve the performance of
The national need for energy is increasing along with thermoelectric [4].
the growth of national economy that needs the efforts to
ensure the continuous availability of energy in sufficient
quantity and quality at a reasonable price level. With the II. THEORETICAL FOUNDATION
decreasing amount of energy derived from fossil, humans are A thermoelectric device works by converting heat
trying to find new sources of alternative energy. One of the energy directly into electricity (thermoelectric generators), or
solutions that can be used to generate energy and is otherwise, the electricity generating cold (thermoelectric
environmentally friendly is by using thermoelectric. coolers). Thermoelectric modul composed by semiconductor
The selection of the thermoelectric module material arrangement (usually Bismuth Telluride) which uses
specification is based on the heat load, the temperature three principles of thermodynamics, known as Seebeck effect,
difference and the electrical parameters used. Thermoelectric Peltier and Thomson. Its construction consists of a pair of P-
cooler has several advantages including no noise, easy type semiconductor material and N-type which forms
maintenance, environmentally friendly and does not require a thermocouple like a sandwich between two thin ceramic
lot of additional components. In addition, another benefit of wafers [5].
thermoelectric cooler as the engine is able to reduce air Thermoelectric cooler (TEC), which is a
pollution and Ozone Depleting Substances (ODSs) because it semiconductor circuit by utilizing the Peltier effect has been
no longer uses Hydrochlorofluorocarbons (HCFCs) and used as a cooling device on some mini cooling system. In
Chlorofluorocarbons (CFCs) known as Ozone Depleting which cooling has become a necessity in modern society that
Substances (ODSs) [1]. has been proven to improve the quality in terms of taste and
Thermoelectric first discovered in 1821 by the hygiene of food and beverages [6].
German scientist Thomas Johann Seebeck. He connected Generally, thermoelectric module, has a measurement
copper and iron in a circuit. Between the two metals are then of 40mmx40mm or smaller and has less than 4 mm thick. Age
placed compass needle. When the metal is heated, it turns the of a thermoelectric module in accordance with the industry
compass needle move. Later known, it happens because standard is about 100,000-200,000 hours and more than 20
electricity that occurs in metals cause the magnetic field. This years when used as a coolant, and by the number and voltage
magnetic field that moves a compass needle. This which is appropriate with the characteristics of each module
phenomenon known as the Seebeck effect [2]. [5].
102
International Journal of Smart Material and Mechatronics Vol.1 No.1 2014 ISSN: 9772356-531002
Test using thermoelectric cooler module is the resistance (R) and assumed to be divided toward the cold side
application of the Peltier effect to move heat. Thermoelectric and hot side.
cooler which is used consists of a number of pairs of P-type (3)
and N-type semiconductor connected in series and parallel
thermal electricity. Heat which is pumped directly can be
changed by changing the pole which is flowed by DC Heat absorbed at the cold side of the thermoelectric module
electricity. The thermoelectric semiconductor material
composed of N-type made from a mixture of bismuth- (4)
telluride-selenium (BiTeSe) and P type made from a mixture
of bismuth-antimony-telluride (BiTeSb). The use of bismuth Heat released at the hot side of the thermoelectric module
telluride on thermoelectric cooler based on some studies that
suggest that bismuth telluride is a material that has the best (5)
performance even though it has limitations on the heat
temperature [7]. As described above, to determine the absorbed
In this study we want to know the performance of calorific value (qc) and the released heat (qh) on thermoelectric
cooler using single, double series, and double parallel can be written in equation (4) and (5), where the first term is
assembled thermoelectric cooler. given electrical energy, the second term is the heat energy
In analyzing the performance of thermoelectric transmitted by conduction, and the third term is the loss of
modules can be observed in Figure 1, the heat transfer occurs heat due to electrical current.
from the heat load to the cold side of the thermoelectric Based on the type of thermoelectric modules used, TEC1-
module can be determined from the amount of heat that is 12706, number of connection elements (N) is 127 so that the
pumped by the Peltier effect, heat moves from the hot side to thermoelectric module is twice of the number of connection
the cold side because the thermal conductivity of elements (2N).
thermoelectric materials, and partly of the total Joule heating Seebeck coefficient value element (m), thermal conductivity
effect generated by the electric current to thermal resistance element (Km), and the thermal resistance elements () usually
[8]. can be seen from the data vendors or from the corresponding
equations form a thermoelectric material, in this case the
material used is Bismuth Telluride.
Seebeck coefficient
Value of the Seebeck coefficient () is determined by
the value of the Seebeck coefficient element (m) and the
number of elements on the thermoelectric modules.
(6)
Electrical resistivity
(2) The amount of electrical resistance (R) is determined
by electrical resistance elements (), the geometry factor (G),
Joule heating effect generated by the electric current
and the number of elements on the thermoelectric modules.
Joule heating effect (qJ) is the heat loss that occurs as
a result of electrical current which can be determined from the
value of the square of the electric current (I) and electrical (10)
103
International Journal of Smart Material and Mechatronics Vol.1 No.1 2014 ISSN: 9772356-531002
(13)
The electric power supplied to the thermoelectric module III. RESEARCH METHODOLOGY
The amount of electrical power supplied to the The research method used is the experimental
thermoelectric module influenced the size of the electric method. Thermoelectric performance testing carried out by
current (I) and the amount of the electrical resistance (R). variation of the DC power supply given,which is 8 V, 10 V, 12
V and variation of module by using single module, multiple
(14) series module, and multiple parallel module with a 360-minute
Energy equilibrium long test as shown in Figure 1 and Figure 2 .
According to the working principle of thermoelectric Data collection was performed by measuring the cold
based on Peltier effect, heat is absorbed from the cold side by side, hot side , water temperature, and ambient temperature
qc and the heat released to the environment by qh. The using a thermocouple and a temperature controller.
difference between the two is the amount of electrical power Determination of the value of the element geometry factor (G)
required or Pin = qh-qc [9] so that the thermoelectric energy is using the AZTEC software; version 3.1 [10]. Data
equilibrium can be written in the following equation: processing done by calculating the calorific value absorbed,
(15) heat removed, the electrical power used, figures of merit, and
COP.
Figure of merit
Figure of merit (Z) is the default for determining the Single Thermoelectric Testing Installation
efficiency of thermoelectric materials. If the value of Z
increases the capability of thermoelectric materials also
increased. Figure of merit value varies depending on the needs
of the thermoelectric material temperature [7].
(16)
104
International Journal of Smart Material and Mechatronics Vol.1 No.1 2014 ISSN: 9772356-531002
thermoelectric, hot side temperature will increase by the temperatures of 29C and the water temperature can reach
addition of voltage. 28C only.
105
International Journal of Smart Material and Mechatronics Vol.1 No.1 2014 ISSN: 9772356-531002
Symbol
Cp = Specific heat of water [J/kgK]
Figure 5. Graph of electricity voltage to the electrical power supplied with the G = The geometry factor [cm]
circuit variation on 360 minutes. I = Electric current [A]
K = Thermal conductivity [W/K]
Coefisien of Performance (COP) Km = Thermal conductivity of elements [W/cmK]
COP value is a measure of the efficiency of a m = Mass of water [kg]
thermoelectric cooler that can be seen from the comparison of N = Number of elements on the thermoelectric
the amount of heat absorbed at the cold side (qc) to the Pin = Electric power [W]
amount of incoming power (Pin). Now, thermoelectric cooling = Heat absorbed at the cold side of the thermoelectric [W]
still has low COP value that it can not compete with vapor = Heat released at the hot side of the thermoelectric [W]
compression cooling system [9]. = Heat absorbed from the water [W]
R = Electrical resistance []
Tc = Cold side temperature [K]
V = Electrical voltage [V]
Z = Figure of merit [K-1]
= Seebeck coefficient [V/K]
m = Seebeck coefficient of the element [V/K]
= Resistance of electric element [cm]
T = Temperature difference [K]
Tw = Temperature difference of water [K]
t = Time difference [s]
REFERENCES
Figure 6.Graphs of the relationship of voltage applied to the COP on 360 [1] Tellurex. 2010. Introduction to Thermoelectric. 1462 International
minutes Drive.Traverse Citi,MI. (http://www.tellurex.com).
[2] California Institute of Technology.2013.Brief History of
Figure 5 is a graph of the relationship between the Thermoelectrics.(Online).
(http://thermoelectrics.caltech.edu/thermoelectrics/history.html).
COP to the applied voltage and variation of circuit on 360 [3] Chakib Alaoui. 2011. Peltier Thermoelectric Modules Modeling and
minutes. The amount of COP influenced by heat absorbed at Evaluation. International Journal of Engineering (IJE), Volume (5) :
the cold side and the amount of electrical power used. For Issue (1).
voltage of 8V, the highest COP values shown in multiple [4] Jincan Chen, Yinghui Zhou, Hongjie Wang, Jin T. Wang. 2002.
Comparison of the optimal performance of single- and two-stage
series, by 1.55046. For voltage of 10V, the highest COP thermoelectric refrigeration systems.
values shown in multiple series circuit is equal to 1.25268. For [5] Riffat, S.B; Ma X. 2003. Thermoelectrics: a review of present and
voltage of 12V, the highest COP value is shown in a series of potential applications. Applied Thermal Engineering 23 913
multiple series that is equal to 1.09192. It shows that voltage 935.Pergamon-Elsevier Science Ltd.
[6] Hendi Riyanto, Sigit Y. Martowibowo.2010. Modeling and
variation given, double series circuit showed the highest COP Prototyping a Mini Portable Thermoelectric Beverage Cooling
value compared to other circuit variations. Device. ICCHT2010 - 5th International Conference on Cooling and
From the three variations of the thermoelectric circuit, best Heating Technologies.
known performance is a thermoelectric module which is [7] Christopher M. Jaworski. 2007.Opportunites for Thermoelectric
Energy Conversion in Hybrid Vehicles. The Ohio State University.
arranged in double circuit with the voltage of 10 V as it can Department of Mechanical Engineering.
achieve the lowest water temperature, the lowest power [8] Rehab Noor Mohammed Al-Kaby. Study Of Thermal Performance of
consumption, and best cooling speed. Thermoelectric Cooling System. Mechanical department, Babylon
University-College of Engineering.
II. CONCLUSION [[9] Yunus A. engel and M. A. Boles. 2006. Thermodynamics: An
Engineering Approach, 5th ed, McGraw-Hill.
From the calculation results and discussion can be concluded [10] Laird Technologies. 2010. Thermoelectric Handbook.
as follows: (http://www.lairdtech.com, diakses pada 1 Agustus 2012).
1. After analyzing the performance of single
thermoelectric, best known performance generated by
giving voltage of 8 V.
2. After analyzing the performance of double
thermoelectric with series assembled, the best
performance is produced by giving voltage of 10 V.
106
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Abstract Natural fibers have been developed as Coir fiber composite have been developed to some
reinforcement of composite to shift synthetic fibers. One of applications like automotive and structure. In this paper, the
potential natural fibers developed is coir fiber. This paper aims potential application of this composite was presented for small
to describe potential coir fiber as reinforcement of composite for wind turbine blade and it was also evaluated effect of
small wind turbine blade application. The research shows that
weathering time to the mechanical properties.
mechanical properties ( tensile, impact, shear, flexural and
compression strengths) of coir fiber composite have really similar Using of fiber composites in wind turbine blade was
to wood properties for small wind turbine blade material, but applied with glass fiber as reinforcement [10]. Glass fiber
inferior to glass fiber composite properties. The effect of composite for small wind blade have been applied and
weathering was also evaluated to coir fiber composite in this compared to flax fiber composite. Flax fiber as natural fiber
paper. can replace possibly glass fiber for reinforcement composite
[11]. The designed small blades will be subjected to load when
Index TermsCoir fibers, composites, wind turbine blade. operation, therefore they need good strength, stiffness and tip
deflection.
I. INTRODUCTION For application in wind blade, weathering will affect to the
Natural fibers have been applied in composite materials. materials of wind blade. Some literatures explained the effect
Related to this, natural fibers have beneficial properties over of weathering to the natural fiber composites. Kenaf high
synthetic fibers like high specific strength and modulus, low density polyethylene (K-HDPE) composite has been tested for
density, low cost and abundant in many countries [1, 2, 3]. durable behavior towards weather effect. The result shows that
Some natural fibers used as reinforcement composite are coir, composite obtained brittleness proportional to the amount of
flax, jute, and ramie fibers. weathering time [12]. Then, outdoor weathering affected
Coir fiber is a natural fiber which has been used for tensile and moduli of the banana/phenol formaldehyde
reinforcement of composite. Coir fiber composite has been composite, and alkali treatment of fiber can improve tensile
developed in India and Brazil. Some researches of mechanical strength if exposure to outdoor weathering [13]. Mechanical
properties of coir fiber composite were done. Flexural strength properties (including impact, tensile and shear strengths) of
was obtained for coir fiber/polyester composite really similar to coir/epoxy composites were influenced by weathering when
the conventional materials [4]. Meanwhile, impact strength of composites were placed in outdoor for 10 days, 20 days and 30
coir fiber composites is higher than jute and kenaf composites days [14] [15]. These effect can be seen in Table 1 where we
Alkali treatment of coir fiber increases its bonding with were published.
polyester matrix. Coir fiber composites show tensile strength
improves when fibers is soaked in 2% alkali prior to mixing TABLE 1. THE EFFECT OF WEATHERING TIME ON
polyester and flexural strength improves when 5% alkali [5]. IMPACT,TENSILE AND SHEAR STRENGTH OF COIR FIBER
COMPOSITES
This result was supported by another reseacher that states
tensile strength of coir fiber composites increased when fibers Tensile Shear Impact
are soaked with alkali prior to binding with matrix. This is Specimens
Strength Strength Strength
Treatment Time
because good adhesive between fibers and matrix after alkali (MPa) (MPa) (kJ/m2)
treatment [6]. Tensile, impact and flexural strengths of Without treatment
17.56 14.57 384.99
coir/epoxy composites were evaluated with the average values (WT)
of 17.86 MPa, 11.49 kJ/m2 and 31.08 MPa respectively. These
10 days
values have lower than glass reinforced composite laminate 17.37 14.42 328.13
[7]. The tensile strength of coir reinforced composites was also
tested and found lower its value. But, their impact strength 20 days
16.38 13.83 307.22
was found higher which have potential for application in
automotive that require impact resistance [8]. The impact
strength of coir fiber composites was also reported that its 30 days 16.40 13.63 296.00
value is higher than other natural fiber composites [9].
107
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Shear
Tensile Flex. Compr
Stre.
Materials Stre. Stre. Stre. Sources
(MPa
(MPa) (MPa) (MPa)
)
Fig 2. Specimens of compression testing (standar ASTM D695) GFRP 826 90.9 14.04 [11][16]
Laminated
veneer
19.1 28.7 [17]
lumber
Timberstra
n wood 5.1 [17]
108
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
days. The possible cause of decreasing its strength was due to REFERENCES
solar radiation and high humidity.
[1] S. V. Joshi, A. K. Mohanty, and S. Arora, Are natural fiber
35 composites environmentally superior to glass fiber reinforced
composites?, Compos. Part, vol. 35, pp. 371376, 2004.
30 [2] Z. Li, X. Zhou, and C. Pei, Effect of Sisal Fiber Surface
Compression Strength (MPa)
109
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
Basri
Mechanical Engineering Department
Tadulako University
Palu, Indonesia
basrist@yahoo.com
Abstract The new configuration of the heat pipe copper within flat plate and sinusoidal wave configuration and with
absorber fluid in solar water heater (SHW) has been developed serpentine-parallel fluid tubes above the plates. The water
in an aluminium collector box that aims to improve the efficiency output temperature in pipe parallel is high enough till about
and especially the outlet temperature of the fluid prior to 80C, while water temperature in the reservoir averaged less
entering the stainless steel hot water. Parallel-serpentine pipe
50oC. The reservoir water temperature is not enough last long
configuration is used in this study in which 5/8" copper tube
installing on the inlet header and header outlet of parallel due to insulating ability of fiberglass reservoir material. In
configuration. The header outlet is connected with 3/8" other words, it is needed to be modified the tank material and
serpentine configuration tube before entering water tank. tubes configuration.
Meanwhile collector panels mounted on a north-facing slope of Thus, this research has focus on modification of some
20o to get a lot of sunlight. From the observations show that the collectors materials such as fiberglass reservoir to be stainless
temperature of the water out of the collector is around 90o C with steel, wood collector box becomes aluminium ones. In
thermosiphon SHW efficiency of approximately 25%. addition, to accelerate the circulation of hot fluid entered into
the reservoir, the configuration of the serpentine-parallel tubes
Key words: solar hot water, parallel-serpentine, tube configuration
is reversed into a parallel-serpentine in serial connecting. Water
flows from the reservoir made by stainless steel to fistly
I. INTRODUCTION parallel tube configuration and then serpentine within hot fluid
Solar hot water (SHW) is cheaper application of solar back into the tank in means of natural convection
collector energy use than air ones. This can be seen a plenty of (thermosiphon) [3]. That happens due to the water mass
water heater technology are manufactured with various models difference within solar collector box that is exposed to north at
according to customer needs. Recently, one obstacle water 20o slope angle. The slope angle () is defined as the angle
heating technology products available in the market is the between the plane of the collector and the horizontal. The
selling price that is not affordable all society levels. Therefore azimuth angle ( ) is defined as the displacement angle
this study is aimed at producing water heating collectors that between the projection on a horizontal plane of the normal to
the collector surface and due north. The incidence angle, , is
are less expensive, easy installed and environmentally friendly.
Thermosiphon solar hot water collectors become interesting the angle between the direct radiation on a surface and the
topic for research and even more manufactured with various normal to that surface. For maximum direct radiation, the
modification leading to increasing efficiency. Mostly, tubes incidence angle should be a minimum [4]. Fig. 1 shows these
configuration of current SHW are parallel and/or serpentine angles.
laying out independently. Both configurations have advantages
and disadvantages for certain solar radiation hours in one day.
For instance, parallel tube configuration is more useful form
midday till afternoon, whereas serpentine ones is between
morning and midday for water solar heating. Therefore,
Mustofa dkk. [1] and Mustofa et al., [2] has been designed and
made a prototype of a hybrid model both serpentine and
parallel tube configuration in one single collector box leading
to minimizing the disadvantages fo SHW. Thermosiphon
principle is applied for collector plate heat with fluid absorber
Figure 1. Major angles in solar applications
110
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
III. METHOD
The research begins with searching solar hot water collector
materials and some tube configurations topics, their advantages
Figure 2. Prototype SHW parallel-serpentine configuration
and disadvantages. Mustofa et al., [2] used serpentine-parallel
in of the tubes diameter. It takes 3 till 4 hours to reach
For equator line like in Palu, Central Sulawesi, the slope
maximum hot fluid in the tank, while high temperature within
angle is less influence on solar intensity. Thus, 20o is accepted.
the reservoir is not long lasting. Therefore, research method is
started by turning the configuration to parallel-serpentine with
II. COLLECTOR DESIGN 3/8 and 5/8 tubes diameter and so on. Testing has been done
with 2 mass flow rate (0.035 and 0.189 ltr/s) on clear weather
As can be seen from Figure 2, collector box is mainly made condition. Data collected from 9.30am till 3pm each 30
from aluminium with dimensions 128 x 78 x 10 cm comprising minutes both fluid temperatures and solar intensity. The
copper tubes on flat and wave sinusoidal plates. The tube and intesity was recorded with Solarimeter Tenmars as shown in
plates were painted by double Arclic Lacquer Black Metallic Figure 3, while fluid temperature is noted by thermocouple
leads to increasing heat absorption from the Sun. There are two digitals in Figure 4.
differences tube diameters for parallel configuration. For
parallel configuration tubes, they are 3/8 and 5/8. Header
parallel tube is 5/8 and the rest of tubes dimension is 3/8
including serpentine configuration. Insulator sponge 3 cm thick
is under the flat and wave sinusoidal plates.
For stainless steel tank is located above the collector in
which cool and hot water are circulating through the tubes. The
tank size itself is 80 x 30 cm with copper taps on two sides as
for output and input water circulating.
Based on those design, instantaneous equations associated
with collector performance are given as follows:
Qu Figure 3. Solarimeter Tenmars
(1)
IAc
While Qu is divided by 2 correspondences formula, For ambient temperature is measured by mercury
Q par AparS UApar (Tpar Ta ) (2a) thermometer. All testing had been conducted at outdoor
Mechanical Engineering laboratory in Tadulako University.
For parallel configuration tubes.
Data are tabulated and calculated based on formulas from (1) to
Qser Aser S UAser (Tser Ta ) (2b) (5).
For serpentine configuration, where Ta is ambient
temperature and S is energy from the Sun that can be
absorbed by collector.
The useful gain given collector to fluid:
qu m C p T (3)
Which m is equal
Figure 4. Thermocouple digitals type-K
m A( par ser ) / 2 v (3a)
111
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
IV. RESULT AND DISCUSSION rate of 0.035 is more stable in low efficiecny than that 0.189
kg/s.
A. Solar Intensity
C. Water Outlet Temperature For Parallel-Serpentine
Figure 7 depicts that water outlet temperature (ToutletParl) in
parallel tube heat absorber is lower than that of serpentine
ones (ToutletSer). The highest temperature of serpentine is about
90oC. This temperature will heat water in the tank. As a result,
NOMENCLATURE
Apar parallel collector area m2
Aser serpentine collector area m2
Cp water specipic heat J/kgoC
I solar intensity W/m2
Figure 6. Total SHW efficiency with 0.189 and 0.035 kg/s mass flow rates Qu useful heat gain from collector W
The graph indicates that collector efficiency for 0.189 kg/s m mass flow rate kg/s
mass flow rates is higher than that of 0.035 kg/s. However, the U collector heat losses coeficient W/m2K
112
Proceeding of International Symposium on Smart Material and Mechatronics ISBN 978-602-71380-1-8
113
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
AbstractSolar energy is known as an environmentally efficiency. Various techniques to enhance the thermal
friendly energy source and wide range of applications. This efficiency in solar water heater have been reviewed by
energy is utilized in various applications such as domestic and Jaisankar et al. [5]. Solar water heating system is an effective
industrial water heating, refrigeration, cooking, power technology to convert solar energy into thermal energy. The
production and water pumping etc. The present study analyzes efficiency of solar thermal conversion is around 70% when
absorptivity of flat-plate absorber and various V-shaped compared to solar electrical direct conversion system which
absorber plates. Analytical investigation of absorptivity of the has an efficiency of only 17%. Summary on the development
various V-shaped absorber plates and comparison with that of
of various system components that includes the collector,
the flat-plate absorber was carried out. The result shows that V-
shaped absorber plate with angle of = 21 0 (V-shaped dimension
storage tank and heat exchanger are discussed in [6].
of t = 4 cm and l = 4 cm) has a better absorptivity compared with The present work analyzes absorptivity of flat-plate
that of the flat-plate absorber and others V-shaped absorber absorber and various V-shaped absorber plates. Absorptivity
plates. Improving the absoptivity of absorber plate enhances of flat-plate absorber and V-shaped absorber plates with
thermal performance of solar water heater. Utilization of V- various angle and dimensions were investigated.
shaped absorber plate will increase the efficiency of solar water
heater due to increasing its absorptivity of absorber plate. II. SOLAR WATER HEATER
Key wordsSolar water heater, V-shaped absorber plate, Thermosyphon solar water heaters which its components
Efficiency includes the collector, storage tank and heat exchanger are
characterized by converting solar energy into thermal energy.
I. INTRODUCTION The system performance is affected by absorber plate and its
design, selective coating, thermal insulation, tilt angle of the
Solar energy is a renewable energy source with wide range
collector and working fluid. The schematic layout of a typical
of applications such as domestic and industrial water heating,
thermosyphon solar water heater is shown in Fig. 1.
refrigeration, cooking, power production and water pumping
etc. Utilization of solar water heaters is increasing in the world
due to their simplicity and reliability. Thermosyphon solar
water heating system is now widely used in domestic as well
Storage
as industrial sector due to its ease of operation and simple
tank
maintenance. Solar water heating system proves to be an
Inlet pipe
effective technology for converting solar energy into thermal
Outlet pipe
energy. Development of various system components that
includes the collector, storage tank, and heat exchanger is
interest subject to enhance thermal performance of solar water
heater. Several investigations have been reported to address
these issues. Collector efficiency of air solar heater has been
investigated [1]. Effect of glass cover of the solar water heater
collector has been reported in [2]. Ayompe and Duffy [3]
have investigated the thermal performance of solar water Solar
heating system with 4 m2 flat-plate collectors in Dublin, collector
Ireland. The results show that solar fraction was 32.2 %,
collector efficiency was 45.6 % and system efficiency was
37.8%. Optical analysis, experimental study and cost analysis
of the stationary V-through solar water heater system have Fig. 1. The schematic layout of a typical thermosyphon solar water heater
been studied by Chong et al. [4]. They proposed a stationary
V-through solar water heater with the maximum solar
concentration ratio of 1.8 suns to improve the thermal
114
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Design and thermal properties of the absorber plate III. ABSORPTANCE OF V-SHAPED ABSORBER PLATE
contribute to the performance of solar collector. The cross Two types of absorber plates which are flat-plate and V-
sectional view of a solar water heater with flat-plate absorber shaped plate are investigated to study the absorptivity of the
is shown in Fig. 2 (a) and V-shaped absorber plate is shown in plates. Increasing the absoptivity of the absorber plate lead to
Fig. 2 (b). enhance the collector efficiency. The solar energy is converted
to usefull energy by absorbed it in the absorber plate and
a) Flat-plate absorber
transfered it to working fluid. Various types of V-shaped
Solar radiation absorber plates are V-shaped plate with angle of = 21 0 (t = 4
Glass cover cm), 32 0 (t = 4 cm), 41 0 (t = 4 cm), 27 0 (t = 3 cm), 40 0 (t = 3
Water pipe Plat plate absorber cm), 49 0 (t = 3 cm). Analytical method is applied to calculate
the absorptivity of absorber plates. Cross-sectional view of ray-
tracing result of flat-plate absorber and V-shaped absorber
plates are shown in Appendix B.
Solar radiations with incident angle of = 0 0, 30 0, 60 0
b) V-shaped absorber plate
reach the surface of absorber plates. Solar radiations reach the
absorber plate perpendicular ( = 0 0) in the daytime, with =
Solar radiation 60 0 in the morning and afternoon. Absorptivity of absorber
Glass cover
Water pipe V-shaped plate plate is calculated by absorbed solar radiation in the plate based
absorber on its incident angle. The absorber plates are assumed as a
black surface. In the case of V-shaped plate, solar radiation is
reflected by several times. Solar radiation is absorbed by the
plate in each reflection based on its incident angle.
Absorptance of black flat-plate with various incident angles is
t shown in Appendix A [7]. Absorptivity of flat-plate absorber
and V-shaped absorber plates are shown in table I.
l
Fig. 2. The cross sectional view of a solar water heater
60 0.981
5. V-shaped = 490 0 0.966
30 0.966 0.965 3 and 2
60 0.963
6. V-shaped = 400 0 0.979
30 0.960 0.968 3 and 3
60 0.966
7. V-shaped = 270 0 0.970
30 0.968 0.970 3 and 4
60 0.970
115
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Figure 3 shows absorptivity of flat-plate absorber and others V-shaped absorber plates. Utilization of V-shaped
various V-shaped absorber plates. Absorptivity of the flat-plate absorber plate will increase the efficiency of solar water heater.
absorber decreases in increasing the incident angle. However,
absorptivity of V-shaped absorber plates changes slightly with ACKNOWLEDGMENT
increasing the incident angle. This fact indicates that V-shaped This work was supported and financed by LP2M
absorber plates provide a better absorptivity compared with the Hasanuddin University under contract:
flat-plate absorber. Several reflections of the solar radiation on 16187/UN4.42/PL.09/2014.
the V-shaped absorber plates increase its absortivity.
REFERENCES
1.000
[1] Jalaluddin and A.E.E. Putra, Studi Teoritis Tentang Efisiensi
0.975 Kolektor Surya Pemanas Udara dengan Pelat Absorber
Gelombang, Jurnal Penelitian Enjiniring Fakultas Teknik,
0.950 Universitas Hasanuddin, 2002.
[2] Jalaluddin, Analisis Perbandingan Prestasi Kolektor Surya
0.925 Pemanas Air dengan 1 (satu) dan 2 (dua) Penutup di Kota
Absorptivity ()
116
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Appendix A
114
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
Appendix B
Cross-sectional view of ray-tracing result of plat and V-shaped absorber plate
Flat-plate absorber
115
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
116
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
117
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
118
Proceeding of International Symposium on Smart Material and Mechatronics ISBN: 978-602-71380-1-8
119