Está en la página 1de 183

Metal, Ceramic and

Composite Materials

Edited by
Christian Coddet
Metal, Ceramic and
Composite Materials

Selected, peer reviewed papers from the


2015 International Conference on
Metal, Ceramic and Composite Materials
(ICMCCM 2015),
January 24-25, 2015, Shanghai, China

Edited by
Christian Coddet
Copyright 2015 Trans Tech Publications Ltd, Switzerland

All rights reserved. No part of the contents of this publication may be reproduced or
transmitted in any form or by any means without the written permission of the
publisher.

Trans Tech Publications Ltd


Churerstrasse 20
CH-8808 Pfaffikon
Switzerland
http://www.ttp.net

Volume 1102 of
Advanced Materials Research
ISSN print 1022-6680
ISSN cd 1022-6680
ISSN web 1662-8985

Full text available online at http://www.scientific.net

Distributed worldwide by and in the Americas by


Trans Tech Publications Ltd Trans Tech Publications Inc.
Churerstrasse 20 PO Box 699, May Street
CH-8808 Pfaffikon Enfield, NH 03748
Switzerland USA
Phone: +1 (603) 632-7377
Fax: +41 (44) 922 10 33 Fax: +1 (603) 632-5611
e-mail: sales@ttp.net e-mail: sales-usa@ttp.net
Preface

Dear Distinguished Authors and Guests,

The Organizing Committee warmly welcomes you to 2015 International conference on


Metal ,Ceramic and Composite Materials (ICMCCM-2015)
, 24th-25th , January, 2015, Shanghai, China.

The aim of ICMCCM 2015 is to present the latest research and results of scientists
(professors, students, PhD Students, engineers, and post-doc scientist) related to
Metal ,Ceramic and Composite Materials topics. This conference provides opportunities for
the different areas delegates to exchange new ideas and application experiences face to
face, to establish business or research relations and to find global partners for future
collaboration.

After the peer-review process, the submitted papers were selected on the basis of originality,
significance, and clarity for the purpose of the conference. The selected papers and
additional late-breaking contributions to be presented as lectures will make an exciting
technical program. The conference program is extremely rich, featuring high-impact
presentation. We hope that the conference results constituted significant contribution to the
knowledge in these up to date scientific field.

The proceeding records the fully refereed papers presented at the conference. The main
conference themes and tracks are Metal, Ceramic and Composite Materials and other
correlation technique. Hopefully, all participants and other interested readers benefit
scientifically from the proceedings and also find it stimulating in the process.

On behalf of the organizing committee, I would like to especially thank Anne, Tanja, Dorthe
and all the editors from Trans Tech Publications for their great support to ICMCCM
2015 .Without their excellent editorial work, ICMCCM 2015 will not be published so timely
and successfully.

Finally we wish all the authors and attendees rewarding and enjoyable memory at ICMCCM
2015 in Shanghai,China. We look forward to your participation in the 2nd ICMCCM in 2016.

With our warmest regards,

Tony Sun

Conference Organizing Chair

February 3rd, 2015

BOSI EDU
Committees and Sponsors
Committees

Editorial Chairs
Prof. Christian Coddet Universit de Technologie de Belfort Montbliard France

International Program Committee Chair (India)


Prof. K.B.V.Varma Indian institute of science India

International Program Committee Chair (Sweden)


Prof. Kristiina Oksman Department of Engineering Sciences and Mathematics Lulea
University of Technology

International Program Committee Chair (Tunis)


Prof.Kais Department of Physics Daoudi University of Monasir

International Program Committee Chair (Korea)


Prof.Chong Soo Lee Graduate Institute of Ferrous Technology Pohang Univ.of Science and
Technology

International Program Committee Chair (France)


Prof. Christian Coddet Universit de Technologie de Belfort Montbliard France

International Program Committee Chair (Taiwan)


Prof. Ran Huang Dept.of Harbor and River Engineering,National Taiwan Ocean University

International Program Committee Chair (Taiwan)


Prof. Yen-Pei Fu Materials Science&Engineering

International Program Committee Chair (India)


Prof. Navdeep Malhotra,YMCA University of Science and Technology,Faridabad,India

International Program Committee Chair (India)


Dr. Sona Rani,UIET,Kurukshetra University,Kurukshetra,Haryana,India
Table of Contents

Preface, Committee and Sponsors

Chapter 1: Metal Materials, Alloys and Heat Treatment of Metals

Effect of Heat Treatment on the Creep Properties of Zr-1Nb-0.12O Nuclear Cladding Tubes
I.W. Kim and S.I. Hong 3
Mechanical Performance and Ductility of Cu/Al/Cu Clad Metals
W.N. Kim, I.K. Kim and S.I. Hong 7
Study of Co-Mn/TiO2 SCR Catalyst at Low Temperature
H. Hu, J.L. Xie, D. Fang and F. He 11
Hot Deformation and Cracking during Compression of 21-4N Steel
H.K. Oh, S.M. Park and S.I. Hong 17
The 6N01 Aluminum Components by Orthogonal Design Optimization
F.L. Xu, B. Li, H.B. Wang, J.J. Wu, D.F. Li, C.H. Zhou and X.M. Wang 23
Emission Factors for Dioxin for Coal Burning, Secondary Smelting and Alloying of
Aluminum, and Cement Kiln Processes
Y.Y. Cheng, M.F. Lu and J.J.M. Lin 27
Effect of Shot Peening Coverage on Fatigue Strengths of Steels and Aluminum Alloy
J. Sakamoto, Y.S. Lee and S.K. Cheong 33
Patch Designing for Damaged Metallic Structure Repaired with Adhesively Bonded
Composite Patch
D. Zhang, W.Y. Ding, F. Su and S.J. Jin 37
Application of the Betti Number to Prior Austenite Grain Size Analysis of Repeatedly
Quenched Steel Based on the Homology Method
K. Mizobe, K. Kida, W. Matsuda and K. Nakane 45
Effect of Heat Treatment on the Mechanical Properties and Interface Structure of 3-ply
Ti/Cu/Ti Clad Composite
Y.K. Kim, P.W. Shin and S.I. Hong 51
Effect of Heat Treatment on Galvanic Corrosion of Cu/Al/Cu Clad Soaked in 3.5% NaCl
Brine Solution
G.T. Kang and S.I. Hong 55
Relation between the Betti Number of Fatigue Fracture Surfaces and Stress Intensity
Factors of Low Carbon Steel (JIS, S45C)
M. Ishida, K. Kida, K. Mizobe and K. Nakane 59

Chapter 2: Ceramic Materials and Processing of Ceramics

High Thermal Stability of Mo/Si3N4/Mo/Si3N4/SiO2 Multilayer Solar Selective Coatings


R.H. Yang, J.Y. Liu, L.M. Lin, F.C. Lai, Y. Qu and W.F. Zheng 67
Validation of a Novel Approach for CO2/N2 Gas Separations by Means of a Hybrid Ceramic
Membrane
N.C. Nwogu, M.N. Kajama and E. Gobina 73

Chapter 3: Composites, Advanced Materials and Nano Materials

Baddeleyite Type Monoclinic Zirconium Oxide Nanocrystals Formation


R. Yuvakkumar and S.I. Hong 79
Effects of Selenization Conditions on the Properties of Cu-In-Ga-Se Thin Films Prepared by
Pulsed Laser Deposition
L. Tang, S.Y. Chen, H.Q. Zhang and Z.G. Huang 83
Effect of Nano-Oxide Addition on Mechanical Properties of Alumina-Mullite-Zirconia
Composites
A. Poowancum and S. Watcharamaisakul 87
b Metal, Ceramic and Composite Materials

Hydrogen Separation Using Inorganic Membranes


M.N. Kajama, N.C. Nwogu and E. Gobina 91
Magnesium Phosphate/Polybutylene Succinate Composite for Bone Repair
Q.X. Wang, X.P. Dong, X.F. Tang, Y.Y. Wu, Y.T. Ding, B.Q. Yu, J. Chen and J. Wei 95
Characterisation of Inorganic Composite Ceramic Membrane for Lactic Acid Esterification
Processes
E. Okon, H. Shehu and E. Gobina 99
Fatigue Life of Arenga Pinnata/Epoxy Composites
A.H. Abdullah and M.F. Mat 103
Preparation and Electrical Properties of Graphene/PEK-c Films
F.H. Shi, J. Li, B.Y. Zhang, X.S. Qiu and L.Y. Gu 107
Estimation for the Si-O Structures Based on the Homology Concept
K. Nakane and I. Halasz 113

Chapter 4: Mechanical Behavior and Characterization of Materials,


Technologies in Materials Engineering

Penetration Performance of Shaped Charge Jets Using Insensitive Munition: Simulative


and Experimental Studies
M. Ahmed, Z.X. Huang, A.Q. Malik, X.D. Zu, Q.Q. Xiao and M. Ahmed 119
Encapsulation of Lanthanide -diketone Hybrids in Organic Oligomer Poly(ethylene glycol)
400
H.T. Li, D. Liu, Y. Li, J.L. Wang and X. Wang 125
Relationship between Solid Lubricant Layer and Friction Coefficient of PPS Races-PTFE
Retainer Hybrid Thrust Bearings under Dry Condition
X.C. Shi, K. Mizobe, Y. Kashima and K. Kida 129
An Image Analyzing Method by a Homology Concept for Fracture Surfaces
A. Takiyama, K. Nakane and K. Kida 135
Dynamic Mechanical Thermal Analysis of Arenga Pinnata Fibre Reinforced Epoxy
Composites: Effects of Alkaline Treatment
A.H. Abdullah, M.S. Abdul Khadir and N.R. Nik Roseley 139
Fabrication of High-Aspect-Ratio Structural Change Microregions in Silicon Carbide by
Femtosecond Bessel Beams
L.L. Zhao, F. Wang, J. Xie and W.W. Zhao 143
Numerical Investigation of the Damage Behavior of S355 EBW by Cohesive Zone Modeling
H.Y. Tu, U. Weber and S. Schmauder 149
Determining the True Strength of the Material of Fiberglass Thick Rings when Stretched
with Half-Disks
R.A. Turusov, A.M. Kuperman and V.I. Andreev 155
The Effects of Ca(OH)2 on the Steam Gasification Process of Acid-Washed and Raw Lignite
C. Shuai, S. Hu, J. Xiang, S. Su, L.S. Sun and Y. Wang 161
The Analysis of the Effect of Surface Stresses at Nanoscale
Z.Y. Ou and Y.W. Wu 169
Experimental Study of the Deformation Behaviour and Mechanical Properties of Fresh
Reinforced Bamboo
L. Gyansah and A.M. Abd El-Aty 173
CHAPTER 1:

Metal Materials, Alloys and Heat Treatment of


Metals
Advanced Materials Research Vol 1102 (2015) pp 3-6 Submitted: 2014-12-26
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.3

Effect of heat treatment on the creep properties of Zr-1Nb-0.12O nuclear


cladding tubes

In Won Kim, Sun Ig Hong*


Department of Nanomaterials Engineering, Chungnam National University, Daejeon, S. Korea
man2049@nate.com, sihong@cnu.ac.kr*

Keywords : Fuel cladding tube, Creep properties, Zr-1Nb-0.12O, Heat treatment

Abstract
Creep properties of one-step and two-step annealed Zr-1Nb-0.12O cladding tubes were studied.
Creep tests were carried out at 450~500 oC with the applied stress between 80MPa and 120MPa
The creep rates of the two-step annealed Zr-1Nb-0.12O alloy were found to be slower than those of
the one-step annealed Zr-1Nb-0.12O alloy. The creep rate decreased with increase of grain size with
annealing for Zr-1Nb-0.12O at intermediate temperatures, suggesting the creep resistance can be
enhanced by the grain size control. The creep life of two-step annealed Zr-1Nb-0.12O increased
over the one-step annealed Zr-1Nb-0.12O by the factor of 18~20 despite the greater initial
instantaneous strain.

Introduction
Zr alloy cladding tubes play important roles not only as the first barrier preventing nuclear
fission products in the UO2 pellet from being released into the reactor coolant, but also as the
conductor transferring nuclear fission-induced heat effectively to the coolant. From the viewpoint of
the nuclear fuel integrity and the reactor safety [1-3], high temperature creep behavior of Zr
cladding tubes along with corrosion behavior is one of the key performance parameters to be
evaluated since they need to sustain the load arising from the pellet-clad mechanical interaction as
well as the load caused by the reactor coolant over pressure acting on the outer surface of the Zr
cladding tubes in the corrosive environment [2]
It was also found that the higher heat treatment temperature after the final pilgering step lowered
the mechanical strength, but enhanced the corrosion and growth resistance [5]. Some cladding tubes
are being heat treated in the stress relieved condition after cold pilgering to render the better
mechanical strength and creep strength while other tubes are being heat treated in the recrystallized
condition to render the better corrosion resistance and growth resistance at the expense of the
mechanical strength. Experimental studies [4] suggest that the creep resistance can be modified not
only by the addition or reduction of a certain alloying elements, but also by the grain size control in
alloys. In this study, the effects of high temperature annealing and the resultant grain growth on the
creep resistance of Zr-1Nb-0.12O cladding tubes were studied.

Experimental Details
In this study, Zr-1Nb-0.12O alloy cladding tubes were processed for studies of creep
characteristics. Zr-1Nb-0.12O alloy cladding tubes were cold-pilgered with two intermediate heat
treatments and finally annealed at 560oC for 5 hrs after final cold pilgering. Some cladding tubes
were additionally annealed at 700oC for 3 hrs to induce the grain growth. The specimens
additionally annealed at 700 oC will be called two-step annealed specimens. In order to examine
the creep characteristics in the circumferential direction, the Zr alloy rings with the width of 4 mm
were cut vertically against the axial direction of tubes. In this study, a special grip with two half-
cylinders with 1 mm wide space between them was used to strain the ring specimen with
maintaining a circular configuration of the ring in contact with the grips all the way during the tests
[1-3,5]. Constant stress creep testers were used to investigate the effect of grain size on creep
deformation behaviors. The creep tests were performed at 450, 475 and 500oC with the applied
4 Metal, Ceramic and Composite Materials

stress between 80 MPa and 120 MPa. For microstructural analyses and fracture studies,
transmission electron microscopy (TEM) and scanning electron microscopy (SEM) were carried out.
Creep Direction

9.5mm

1mm
4.0mm

Creep Direction

(a)Ring specimen (b) grip configuration

Fig. 1. Schematic configuration of (a) ring specimen and (b) grip for creep test

Results and Discussion


Grains of annealed Zr-1Nb-0.12O tubes at 560oC for 5 hrs were too small to be identified by a
typical optical micrographic analysis. Fig.2 displays the typical TEM microstructure of the Zr-1Nb-
0.12O alloy nuclear cladding tubes annealed at 560C for 5 h (a) and additionally high temperature
annealed at 700C for 3 h (b). The regularly annealed (one-step) Zr-1Nb-0.12O tubes displayed
small equi-axed grains with dislocations and dislocation tangles. This means that dislocation
generation and storage also occurs during the pilgering process along with the continuous
recrystallization. In the two-step annealed Zr-1Nb-0.12O tubes, dislocation-free recrystallized
grains with a larger grain size than regularly annealed Zr-1Nb-0.12O tubes. The grain size of
annealed and two-step annealed Zr-1Nb-0.12O tubes was observed to be ~1.2 m and ~5 m
respectively.

Fig. 2. TEM microstructure of the one-step annealed Zr-1Nb-0.12O (a) and two-step annealed Zr-
1Nb-0.12O (b)

Fig. 3(a) and 3(b) exhibit creep curves of Zr-1Nb-0.12O cladding tubes; one-step annealed (a)
and two-step annealed (b). Both one-step annealed and two-step annealed Zr-1Nb-0.12O alloy
exhibit typical creep deformation curves although two-step annealed specimen exhibit less-
pronounced primary period. It should be noted that the creep rates are slower in the two-step
annealed alloy than in one-step annealed specimens at all loading condition. It should be
emphasized that the primary creep period is shorter in the two-step annealed specimen despite the
initially high instantaneous strain upon loading because of lower strength. It was shown that not
only creep rate is faster but also creep life is shorter in the one-step annealed Zr-1Nb-0.12O than in
the two-step annealed Zr-1Nb-0.12O. For example, the two-step annealed Zr-1Nb-0.12O fractured
after 198 hrs at 120 MPa, 500oC whereas the one-step annealed counterpart fractured only after
11hrs under the same loading and temperature conditions. The creep life of two-step annealed Zr-
1Nb-0.12O increased over the one-step annealed Zr-1Nb-0.12O by the factor of 18~20.
Advanced Materials Research Vol. 1102 5

Fig. 3. Creep curves of one-step annealed (a) and two-step annealed (b) Zr-1Nb-0.12O.

Creep life has been suggested to increase with increase of grain size because the contribution of
the fast diffusion along grain boundaries decreases. However, dislocation creep process has been
considered to be grain size independent. For example, at intermediate temperatures where dislocation
creep is dominant, that the creep rate of aluminium is grain size independent [6]. However, other
studies have indicated that the creep rate may increase or decrease with decreasing grain size, or be
grain size-insensitive in copper [6]. Lee et al. [7] reported that, in 316LN stainless steel, the creep rate
decreased to a minimum value at the intermediate grain size of 80~130 m and then increased with
further increase of grain size.
Lee et al.[7] suggested that their data agreed with Garofalos model in which grain boundaries act
simultaneously as both dislocation sources and barriers to dislocation movement. Generally, the grain
boundaries which act as barriers reduce the steady state creep rate but they also sometimes accelerate
it, being dislocation sources [6]. For copper and brass, the steady state creep rate is proportional to the
square of grain size at low temperatures. The steady state creep rate of lead is proportional to the
inverse of grain size at high temperatures. Wilshire and Palmer [6] suggested that the contribution of
higher dislocation recovery and rearrangement within the boundary region would increase with
decreasing grain size, resulting in the increase of creep rates with decrease of grain size. Both models
suggest that the effect of grain size varies with experimental temperature and conditions even in
identical materials and the dislocation activity near-grain-boundary regions could increase the creep
rate in the small grain size region, supporting the decrease of creep rate with increase of grain size in
the present study.
Fig. 5 show the SEM images of fracture surfaces of one-step annealed and two-step annealed Zr
alloy crept at 100MPa and 475 C. In both specimens, dimples were observed on the fracture surface.
The presence of dimples suggests a typical ductile fracture in general. In the annealed specimen, some
secondary cracks appeared to be formed along the grain boundaries. In general, with the increase of
temperature, the number and mobility of thermal vacancies increase, causing more vacancies or
impurities to diffuse to grain boundaries.
6 Metal, Ceramic and Composite Materials

Fig. 5. SEM fracture surface images of Zr-1Nb-0.12O at 100MPa and 475 C:


(a) one-step annealed Zr alloy and (b) annealed Zr alloy.

Summary
Creep tests were performed on stress-relieved and annealed Zr-1Nb-0.12O and the results can be
summarized as follows:
1. The creep rate decreased with increase of grain size with annealing for Zr-1Nb-0.12O at
intermediate temperatures, suggesting the creep resistance can be enhanced by the grain size control.
2. The creep life of two-step annealed Zr-1Nb-0.12O increased over the one-step annealed Zr-1Nb-
0.12O by the factor of 18~20.
3. The primary creep period is shorter in the two-step annealed specimen despite the initially high
instantaneous strain upon loading because of lower strength.

Acknowledgment
This work was supported by a grant from the Fundamental R & D Program for Industrial
Technology (2014), funded through the Ministry of trade, Industry and Energy and by a grant from
the Basic Atomic Energy Research Institute Program, through the National Research Foundation of
Korea (NRF) (2014)

References
[1] S. Y. Lee , K. T. Kim, and S. I. Hong, J. Nucl. Mater. 392 (2009) 63-69.
[2] J. M. Lee and S. I. Hong, Mater. Design. 32 (2011) 4270-4277.
[3] S. Ko, S. I. Hong, and K. T. Kim, J. Nucl. Mater. 404 (2010) 154-159.
[4] S. L. Mannan, P. Rodriguez, Mater. Sci. Technol. 17 (1983) 63-69.
[5] W. N. Kim, Y. Choi, and S. I. Hong, Phys. Metals Metallogr. 13 (2014) 1313-1317.
[6] B. Wilshire and C. J. Palmer, Scr. Mater. 46 (2002) 483-488
[7] Y. S. Lee, D. W. Kim, D. Y. Lee and W. S. Ryu, Met. Mater. Int. 7 (2001) 107-114
Advanced Materials Research Vol 1102 (2015) pp 7-10 Submitted: 2014-12-26
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.7

Mechanical performance and ductility of Cu/Al/Cu clad metals

Won Nyeon Kim1, a, In Kyu Kim2, b, Sun Ig Hong1,c


1
Department of Advanced Materials Engineering, Chungnam National University, Daejeon, 305-764,
Republic of Korea
2
POSCO coated & Color steel, Pohang, Korea
a
dnjssus2@naver.com , binom34@poscocnc.com , csihong@cnu.ac.kr

Keywords: tension test, copper, Aluminum, clad, heat treatment

Abstract. The tensile mechanical performance and ductility of Cu/Al/Cu clad composite were
studied. Brittle interfacial reaction compounds were observed at Cu/Al interfaces after annealing at
500oC. The ultimate tensile strength of the tri-layered Cu/Al/Cu clad plates is in close agreement
with those calculated from the UTS data of separated Cu and Al layers using the rule of mixture.
The loss of ductility in Cu/Al/Cu and separate Cu upon annealing at 500oC is associated with the
brittle intermetallic layer. The greater ductility of as-roll-bonded Cu/Al/Cu than those of both Cu
and Al is linked to the excellent interface bonding after roll-bonding. Excellent bonding at Cu/Al
interface is prerequisites for improvement ductility of clad composite.

Introduction

The properties of clad composite material are determined by the selection of component materials
to be joined, the stacking structure and properties between different materials [1-7]. Copper and
aluminum clad composites have been widely studied, because of their advantages associated with
low density and cost-competitiveness over copper and copper alloys [8-10]. For example, a two-
layer clad sheet of Al/Cu can almost reduce 40% in weight, with the equivalent electrical and
thermal conductivity as a copper alloy. But the cost can be reduced to the 60% of a copper alloy.
For these reasons, Al/Cu clad is frequently used for armored cables, yoke coils in TV sets, air-
cooling fin and bus-bar conductor joint. However, producing Cu/Al clad sheet is a great challenge
due to the different chemical and physical characteristics of component metals. The formation of
brittle CuxAly intermetallic compound at elevated temperature weakens seriously the interfacial
bonding. The nature of interfacial intermetallic in Al/Cu composite fabricated by cold rolling is not
completely understood [3]. In this study, the effect of heat treatment on tensile mechanical
performance and ductility of roll-bonded Cu/Al/Cu was investigated. The behaviors of Cu/Al clad
were also compared with those of separated Cu and Al sheets to observe the effect of interface
joining on the deformation behaviors.

Experimental method
The materials used in this study were 3-layerd Cu/Al/Cu clad metal produced by the roll bonding
process. These final clad plates had a total thickness of about 2mm while that of Cu and Al layer
was 0.2mm and 1.6mm respectively. Cu/Al/Cu clad plate was heat-treated at 300 and 500 oC for
3hrs to investigate the effect of the heat treatment and tension tests were performed using a
Universal Materials Testing Machine (UNITED, US/SSTM) at room temperature with the strain
rate of 1 x 10-3/s. The gage length and gage width for of tensile testing specimen were 15mm and
3.4mm, respectively. In order to evaluate the contribution of each layer on the overall mechanical
performance of 3-layerd Cu/Al/Cu, each layer was separated from the clad composite and tension-
tested independently to characterize the mechanical performance. Separate Cu sheets were obtained
by dissolving the Al layer in NaOH solution for 24 hours at room temperature.
8 Metal, Ceramic and Composite Materials

Result and discussions


Fig. 1 shows the optical micrographs of the as-roll-bonded and heat-treated Cu/Al interface at
300 oC and 500 oC for 3hours. For as-rolled Cu/Al/Cu clad and that heat-treated at 300 oC, no
visible reaction layer was observed, suggesting that the bonding interface between Cu and Al is
intact. After 500 oC for 3hours, interfacial reaction layer (29m) was observed.
In fig. 2(a), stress-strain responses of the as-roll bonded clad plate and annealed clad plates at 300
and 500 oC are exhibited. As-roll bonded clad metal showed high yield strength. The decreases of
the yield strength and the increases of the fracture strain upon annealing at 300 oC can be attributed
to the softening associated with recovery/recrystallization in Cu/Al/Cu clad metals. The decrease of
fracture strain and yield strength after annealing at 500 oC, was attributed to the brittlrness of
intermetallics and its harmful effect on the overall ductility [3].
Fig. 3(b) and 3(c) display the stress-strain response of Al sheet and Cu sheet respectively, which
were separated from the as-roll-bonded and annealed clad composite plates. The fracture strain of
the separated Al sheet increased drastically and the flow stress decreased with heat-treatments of
300 oC and 500 oC. The increases of fracture strain the decrease of strength in separated Al sheet are
due to the recovery and recrystallization of aluminum [3]. Stress-strain responses of separated Cu
sheet show the drastic increase of fracture strain after heat-treatments of 300 o and decrease of
fracture strain at 500 oC. The decrease of strain can be attributed to the surface irregularity formed
in the presence of intermetallics.

Fig. 1. Interface region of the as-roll boded (a) and heat-treated Cu/Al/Cu composite at 300oC (b) and 500 oC (c) for
3hrs.[7]

The comparison of stress-strain responses of as-roll-bonded Cu/Al/Cu clad composite in Fig.2(a)


to those of separate Al and Cu revealed that the decrease of strength with heat treatment in
Cu/Al/Cu is close to that of separated Al layer in fig. 2(b) although the strength level of Cu/Al/Cu is
higher than separated Al. On the other hand, the increase of ductility after 300 oC heat treatment and
decrease after 500 oC heat treatment in Cu/Al/Cu clad composite is similar to those of separated Cu
layer in fig.2 (c). The intermetallic compound layer as thick as 29m in Fig. 1(c) strongly suggests
that the loss of ductility in Cu/Al/Cu and separate Cu could be associated with the presence of
brittle intemetallic compounds.
In fig. 3 (a) and (b), the ultimate tensile strength (UTS) (a) and fracture strain (b) of as-roll-
bonded and annealed Cu/Al/Cu clad composites are shown as well as those of separated Al and Cu
sheets. In this figure, experimental UTS values and those calculated using the rule of mixture for
Cu/Al/Cu clad composite using the data of separated Cu and separated Al are plotted. The decrease
of the UTS of clad sheet with heat treatment is thought to be caused by softening of Cu and Al. The
calculated strength is closer to that of Al because the volume fraction of Al is 80% in Cu/Al/Cu clad.
The experimental UTS of clad composite are in good agreement with calculated UTS using the data
of separated Al and Cu sheet. It is apparent that UTS of Cu/Al/Cu sheet follow the rule of mixture.
Advanced Materials Research Vol. 1102 9

Fig. 2. Stress-strain curves of as-roll bonded and heat-treated Cu/Al/Cu clad composite (a) [7], separated Al plate (b),
separated Cu plate(c)

In fig.3 (b), the fracture strain of Cu/Al/Cu clad composite and those of separated Al and Cu are
exhibited. The calculated fracture strain of Cu/Al/Cu based on the rule of mixture is also shown.
The experimental fracture strain of as-roll-bonded and annealed at 300 oC Cu/Al/Cu clad shows
higher value than the calculated fracture strain from the data of separated Al and Cu sheet using the
rule of mixture. The experimental fracture strain of annealed at 500 oC of Cu/Al/Cu clad has a
similar value to that calculated fracture strain using the rule of mixture.
It should be noted that the fracture strain of as-roll-bonded Cu/Al/Cu is greater than those of both
Cu and Al and, needless to say, that predicted from the rule of mixture. After 300 oC heat treatment,
the fracture strain of Cu/Al/Cu is smaller than that of separated Cu, but far higher than that of
separated Al and that predicted from the rule of mixture. Jin and Hong [11] reported that the
enhanced ductility of the clad composites is due to the suppression of the localized neck and the
enhancement of homogeneous deformation driven by the mutual constraint imposed by an adjacent
parallel layer. They [11] suggested that excellent interface bonding and absence of brittle
intermetallics at the interface are prerequisites for the enhanced ductility of clad composites because
constraint by the adjacent joined metal layer can be made by the stress and strain transfer through
the interface. The as-roll bonded Cu/Al/Cu has an excellent interface bonding strength with the
presence of negligible intermetallic layer, ensuring the better ductility of Cu/Al/Cu over the
separated Al and Cu.
10 Metal, Ceramic and Composite Materials

Fig. 3. Ultimate tensile strength (UTS) (a) and fracture strain (b) of as-roll-bonded and annealed Cu/Al/Cu clad
composite plotted along with those of separated Al and Cu sheets, UTS and Fracture strain values calculated using the
rule of mixture for Cu/Al/Cu clad composite using the data of separated Al and separated Cu are also plotted.

Summary

In this study, the influence of intermetallic compounds on the ductility in Cu/Al/Cu clad was
examined. The following conclusions were reached;
1. The UTS of the Cu/Al/Cu clad metal is close to the calculated UTS using the rule of mixture
from the data of separated Cu and Al sheets.
2. The fracture strain of as-roll-bonded Cu/Al/Cu is greater than those of both Cu and Al. The
improved ductility of clad composite is linked to the excellent interface bonding after roll-bonding
3. The experimental fracture strain of Cu/Al/Cu clad annealed at 500 oC has a similar value to that
calculated fracture strain using the rule of mixture.

Acknowledgement
This work was supported by the 2nd phase of the Fundamental R&D Programs for Core Technology
of Materials funded by Ministry of Trade, Industry and Energy (2014-2015).

References

[1] I. K. Kim, S. I. Hong. Mater Des. 47 (2013) 590-598.


[2] I. K. Kim, Hong SI. Metall Mater Trans A. 44 (2013) 3890-3900.
[3] I. K. Kim and S. I. Hong. Mater Des. 57 (2014) 625-33.
[4] A. Khosravifard, R. Ebrahimi, Mater Des. 31 (2010) 4939.
[5] I. K. Kim, S. I. Hong, Mater Des. 49 (2013) 93544.
[6] J. S. Ha, S. I. Hong, Mater Des. 51 (2013) 293-9.
[7] H. Kim, S. I. Hong, Mater Des. 67 (2015) 42-49
[8] J. Y. Song, S. I. Hong, Mater Des. 32 (2011) 1790-95
[9] J. S. Ha, S. I. Hong, Mater Sci and Eng A. 565 (2013) 9-12.
[10] S. I. Hong, M. A. Hill, Scripta materialia. 42 (2000) 737-742.
[11] J. Y. Jin, S. I. Hong, Mater Sci and Eng A. 596 (2014) 1-8.
Advanced Materials Research Vol 1102 (2015) pp 11-16 Submitted: 2014-12-29
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.11

Study of Co-Mn/TiO2 SCR Catalyst at Low Temperature

Hua Hu1,a, Junlin Xie1,b, De Fang1,c, Feng He1,d *


1
Wuhan University of Technology, Province Hubei, China
a
huhua6682153@163.com, b xjlclxy@126.com, c fangde94@163.com, d he-feng2002@163.com

Key Words: Co doping, MnOx/TiO2, SCR, CoMn2O4

Abstract: The Co-Mn/TiO2 catalysts with different contents of Co were prepared by the
impregnation method. Catalytic activity tests showed that the addition of Co could obviously
enhance the SCR performance of Mn/TiO2 catalyst, and the NO conversion of 0.05Co-Mn/TiO2
samples could reach 94.03% at 180 . At the same time, with the help of XRD, BET, H2-TPR and
NH3-TPD, the various physical and chemical characteristics as well as phase structures of catalysts
were characterized systematically. According to the results, the formations of crystalline CoMn2O4,
Mn3O4 and amorphous Mn2O3 were detected. Furthermore, Co doping could reduce the reduction
temperatures of catalysts and enhance the adsorption capacities of NH3, resulting in the excellent
SCR performance at low temperature.

1. Introduction
Its known that a large number of NOx will destroy the nitrogen balance of nature, causes the
phenomena, such as acid rain, photochemical smog [1]. In recent years, people have paid close
attention to environment problems, many measures used to reduce the NO emission are studied [2].
Relative to other flue gas denitration technology, selective catalytic reduction (SCR) at low
temperature has the characteristics of low cost and high efficiency, and thus receives much more
attention [3,4]. Researches have shown that Mn/TiO2 catalyst showed good activity at low
temperature and had been widely researched [5]. Furthermore, most transition metals have
changeable valence state which could improve the catalytic activity of Mn/TiO2 catalysts [6-8].
Currently, in the research field of Co-Mn/TiO2 catalyst, a few scholars have had made some
progress, Xu et al. [9] prepared Co/TiO2 catalyst doped Mn to improve the SCR performance, but
the catalyst only could work at a high temperature range 250-350 . And Kims [10] research
confirmed that CoOx/TiO2 catalyst surface might form CoTiO3 or Co2TiO4 species. However, the
research failed to reveal the relationship of material composition, structure and performance.
In this work, various Co-Mn/TiO2 catalysts were prepared by a simple impregnation method and
were applied to NH3-SCR of NOx in the temperature range of 80-180 . A systematic investigation
on physical and chemical characteristics, phase structures and surface properties of the Co-Mn/TiO2
catalysts was carried out by means of XRD, BET, H2-TPR and NH3-TPD. And the inclusion of a
mixed-oxide phase, CoMn2O4, played a key role and showed the significant synergistic effect with
MnOx for higher activity of low temperature NH3-SCR of NOx.

2. Experiment
2.1 Catalysts preparation
Various kinds of catalysts were prepared by the impregnation method. And TiO2 powder (P25,
Degussa, 80% anatase and 20% rutile) was used as support material to prepare these catalysts while
12 Metal, Ceramic and Composite Materials

the molar ratio of Mn/Ti is 0.4. The Mn/TiO2 catalysts doped with Co were denoted as
xCo-Mn/TiO2 (x=0, 0.025, 0.05, 0.10, 0.15, 0.20 and 0.25) in this paper, where x represented the
atomic ratio of Co to Ti. Acetic acid manganese (Mn(AC)24H2O) and nitric acid cobalt
(Co(NO3)36H2O) were used as the precursors of manganese and Co. In the process of preparation,
the required amounts of precursors were added to a 300 mL beaker containing 4.0 g of support in
200 mL deionized water, and the mixture was stir at room temperature for 2 h. Then, the mixture
was put in a drying oven at 110 for about 16 h to remove excess water. Finally, the materials
were ground and sieved (80-120 mesh) to obtain homogeneous powder. Prior to the reaction studies,
the powder was calcined in a tubular oven at 500 for 3 h.
2.2 Catalytic experiments
Catalyst performance of catalyst samples were tested in the fixed bed reaction, and reaction gas
composition: 720 ppm NO, 800 ppm NH3, 3 vol% O2 and N2 as balance. In the test, through
adjusting the temperature step by step, the NO initial maximum concentration ([[NO]in) and NO
concentration of outlet that could indicate the reaction level of that corresponding temperature were
recorded ([NO]out). The temperature of reaction bed was ranged from 80 to 180, and the NO
conversion could be calculated by the formula: =([NO]in-[NO]out)/[NO]in100%.
2.3 Characterization of Catalysts
X-ray powder diffraction patterns had been recorded on a Germany Bruker diffractometer (D8
Advance) using Cu K (wavelength 0.15418 nm) radiation source. With the help of the Thermo
thermogravimetric analyzer, weight losses along with the changes of the samples were recorded.
The specific surface area and pore volume of catalysts were measured by Micromeritics 3020
specific surface area analyzer. H2-TPR and NH3-TPD tests were performed using an automated
catalyst characterization system (Thermo TPDRO1100 Serie).

3. Results and discussion


3.1 Performance of xCo-Mn/TiO2 catalysts
As shown in Fig. 1, seven samples of Co-MnOx/TiO2 catalysts were used to study the effect of
different Co loadings on NO conversions with respect to the temperature (80-180 ) in the SCR of
NO by ammonia. The NO conversion of MnOx/TiO2 catalysts is about 82% at 180 , and
0.05Co-Mn/TiO2 catalyst always maintains the highest denitration rate and can reach to 94.03% at
180 , indicating that different Co loadings can indeed influence the SCR activity of
Co-MnOx/TiO2 catalyst. But this promotion does not increase with the increase of Co content, on
the contrary, the SCR activity of the catalyst has a varying degrees of decline when Co/Ti0.10, for
example, 0.25Co-Mn/TiO2 catalyst only has the NO conversion about 88.75% at 180 .

Fig. 1. NH3-SCR performances of xCo-Mn/TiO2 catalysts


Advanced Materials Research Vol. 1102 13

3.2 Phase analysis for xCo-Mn/TiO2 catalysts


XRD tests of xCo-Mn/TiO2 catalysts calcined at 500 are shown in Fig. 2. Diffraction peaks of
anatase TiO2 (JCPDS#84-1826) and rutile TiO2 (JCPDS#88-1172) are observed in the samples due
to the support TiO2 (P25). For Mn/TiO2 catalyst, the strong characteristic peaks of Mn3O4 phase
(JCPDS#89-4837) can be detected, which indicates that the main manganese phase in the catalyst
might be Mn3O4. When Co/Ti = 0.025, Co species are not found due to the low content of Co
loadings. Meanwhile, Co might enter other phase or in the form of amorphous lattice. But with the
increase of Co content to Co/Ti = 0.05, the strength of the diffraction peaks of the Mn3O4
significantly reduces, suggesting that crystal phase Mn3O4 content gradually reduces. Furthermore,
there are obvious CoMn2O4 crystal diffraction peak (JCPDS#1-1126), and this similar kind of phase
is beneficial to improve the performance of catalyst because of electron transfer become easier [11].

Fig. 2. XRD patterns of catalysts.


a: Mn/TiO2; b: 0.025Co-Mn/TiO2; c: 0.05Co-Mn/TiO2; d: 0.10Co-Mn/TiO2;
e: 0.15Co-Mn/TiO2; f: 0.20Co-Mn/TiO2; g: 0.25Co-Mn/TiO2.

Fig. 3. TG-DSC tests of xCo-Mn/TiO2 catalysts


TG curves of the 0.025Co-Mn/TiO2 samples are shown in Fig. 3, there are three obvious
endothermic peaks at 38.1, 527.8 and 719.7. As support, TiO2 is stable and would not
decompose, even if the phase transformation happen, it wouldn't appear a large weight loss, so the
diagram indicates that the above three endothermic peak are assigned to the removal of free/bound
water, the transformation of MnO2-Mn2O3, and the transformation of the MnO2-Mn3O4, respectively.
14 Metal, Ceramic and Composite Materials

And it could be calculated, the transition of MnO2-Mn2O3 has a weight loss 1.80%, so as the
transition of Mn2O3-Mn3O4 (1.19%). Moreover, with the increase of Co content (Fig. 2. b-d), the
samples (Co/Ti = 0.05, 0.10, 0.20) have similar results, but the transition of MnO2-Mn2O3 has a
decreasing weight loss, oppositely the transition of Mn2O3-Mn3O4 has a increasing weight loss,
which suggests that amorphous MnO2 and Mn2O3 exist in xCo-Mn/TiO2 catalysts, and the latters
content has a increase with the increase of Co. Moreover, the transform temperature of
Mn2O3-Mn3O4 in xCo-Mn/TiO2 catalysts decreases distinctly, which indicates that the chemical
activity of Mn2O3 is increasing along with the content increase of Co.
3.3 BET surface area measurements
xCo-Mn/TiO2 series catalysts BET test results are shown in Table 1. In general, the samples
prepared do not have a good specific surface area and pore volume. It could be seen that
0.025Co-Mn/TiO2 catalyst has a large specific surface area and the pore volume. 0.05Co-Mn/TiO2
catalyst has the largest pore volume, and the specific surface area is smaller but larger than the
others. Obviously, compared with the results of Fig. 1, 0.05Co-Mn/TiO2 and 0.1Co-Mn/TiO2
catalyst with larger specific surface area and pore volume have the very good SCR activity. With the
increase of doping Co, catalysts (Co/Ti=0.15, 0.25) specific surface area and pore volume decrease,
but 0.20Co-Mn/TiO2 catalyst has an outstanding specific surface area (38.37 m2/g) with a poor pore
volume. Compared Fig. 3, this should be due to the formation of new phase CoMn2O4, which
causes the change of the catalyst pore structure characteristics.
Table 1. BET measurements of xCo-Mn/TiO2 catalysts
Co/Ti BETm2/g Pore volumecm3/g
0 34.77 0.2241
0.025 38.87 0.2895
0.05 36.23 0.3337
0.10 37.15 0.2921
0.15 32.08 0.2644
0.20 38.37 0.2416
0.25 32.85 0.2690

3.4 H2-TPR, NH3-TPD


For an adequate determination of effect of Co on the reduction profile of the Mn/TiO2 catalysts,
temperature programmed reduction (H2-TPR) experiments were carried out, the corresponding TPR
profiles are presented in Fig. 4. Two different reduction peaks are observed for the
0.025Co-Mn/TiO2 catalyst. But, it had reported that the reduction peaks in Mn/TiO2 catalyst at
303(T1), 392(T2), and 463(T3) could be interpreted as a stepwise reduction of MnO2 to
Mn2O3, Mn2O3 to Mn3O4 and Mn3O4 to MnO, respectively [12]. Meanwhile, oxides consume the
amount of H2 and its reduction peaks have close relationship with the position the oxides existed. It
is hard to explain the peaks in Fig. 4. However, there still are some finds according to the position
and height of H2 consumption. With the increase of Co content, the reduction temperatures of all the
titania-supported mixed oxides are decreased to varying degree. Two reduction peaks of
0.20Co-Mn/TiO2 catalyst are shifted from 307 to 270 and 460 to 405 , respectively.
Therefore, the adding Co has a very positive and effective interaction between oxide to improve the
activity of oxygen, and the SCR activity is promoted. It also can be seen from Fig. 4., 0.15Co-
Mn/TiO2 and 0.25Co-Mn/TiO2 catalyst have similar H2-TPR curves, which illustrate that they have
approximate oxidation ability. And the results are in keeping with table 1 (the specific surface area
and the pore volume of them are so close.) With the increase of Co content (Co/Ti0.05), the
Advanced Materials Research Vol. 1102 15

reductive peak at about 610 appears, but its position does not change very much, just the
intensity increases, the peak should be a corresponding reduction of Co oxide (Co2O3-CoO).

Fig. 4. H2-TPR profiles of xCo-Mn/TiO2 catalysts.


NH3-TPD test results are shown in Fig. 5, it could be seen that a strong peak appears at about
176 . With the increase of Co content, the integral area of the peak has the trend of increasing first
and decreasing later. 0.025Co-Mn/TiO2 has the smallest peak area and 0.05Co-Mn/TiO2 reaches the
maximum. As it mentioned above, the abundant pore structure and larger specific surface area are
conducive to gas adsorption, the results are consistent with its pore structure change rule. At the
same time, the amount increase of NH3 adsorption reveals that Co doping could really promote the
catalyst adsorption of NH3, and lead to the high dispersion of the active ingredients on the carrier
surface. Earlier studies had proved that Co doping could promote the formation of acid site and
enhance the surface acidity of catalysts. Therefore, the effects are more conducive to NH3
adsorption and activation [13]. As we know, NH3 adsorption is an important step in catalytic
reaction, so a larger amount of NH3 adsorption causes the 0.05Co-Mn/TiO2 catalyst has a good SCR
performance.

Fig. 5. NH3-TPD profiles of xCo-Mn/TiO2 catalysts.

4. Conclusions
Samples with the same atomic composition can have very different properties, depending on the
Co content and on their chemical properties. The 0.05Co-Mn/TiO2 catalyst samples are the best one
16 Metal, Ceramic and Composite Materials

for the low temperature SCR of NO with NH3, having a NO conversion about 94.03%. By the XRD
and TG tests, crystalline CoMn2O4, Mn3O4 and amorphous Mn2O3 are found. Moreover, their
formations reduce the reduction temperatures of catalysts and enhance the adsorption capacity of
NH3, according to H2-TPR, NH3-TPD tests. The formation of the CoMn2O4 has a certain influence
on the catalyst pore structure and the SCR activity.

Acknowledgment
The authors wish to acknowledge financial support from the National "Twelfth Five-Year" Plan
for Science & Technology Support (No. 2011BAE29B02).

Reference
[1] W. Zhou, X.S. Wang, Y.H. Zhang, et al. Current status of nitrogen oxides related pollution
condition in China and integrated control strategy. J. Peking. Univ., 44 (2008) 323-323.
[2] K.Q. Sun, Q. Zhong. Technologies and application of flue gas denitrification in thermal power
plants. Chem. Ind. Pre., Beijing, 2006.
[3] Y.J. Gao. Research on the preparation and denitrification performance of low temperature SCR
catalyst. Nanjing Univ. Sci. & Tech., 2009.
[4] B.J. Wu, S.G. Wang, Z.X. Fang, et al. Technologies of flue gas denitrification and their chemical
reaction aspects. J. Therm. Power Gener., 35 (2006) 59-64.
[5] Y.F. Jiao, X.T. Wang, S.L. Zhuang, et al. Research progress in Mn-based composite metal oxide
catalysts for low-temperature selective catalytic reduction of NOx. J. Boi. Tech., 5 (2012) 66-66.
[6] H.D. Xu, C.T. Qiu, Q.L. Zhang, et al. Influence of tungsten oxide on selective catalytic
reduction of NOx with NH3 over MnOx-CeO2/ZrO2-TiO2 monolith catalyst. Acta. phys-Chim. Sin.,
26 (2010) 2449-2454.
[7] J.L. Xie, D. Fang, F. He, et al. Performance and mechanism about MnOx species included in
MnOx/TiO2 catalysts for SCR at low Temperature. Catal. Commun., 28 (2012) 77-81.
[8] D. Fang, J.L. Xie, H. Hu, et al. Influence of the valence state of manganese ion on the activity of
Mn supported catalysts for denitration. J. Wuhan Univ. Technol.-Mater. Sci., 35 (2013) 37-40.
[9] W.Q. Xu, J. Zhao, H.R. Wang, et al. Catalytic oxidation activity of NO on TiO2-Supported
Mn-Co composite oxide catalysts. Acta. phys-Chim. Sin., 29 (2013) 385-390.
[10] M.H. Kim, K.H. Choo. Low-temperature continuous wet oxidation of trichloroethylene over
CoOx/TiO2 catalysts. Catal. Commun., 8 (2007) 462-466.
[11] D. Fang, J.L. Xie, D. Mei, et al. Effect of CuMn2O4 spinel in Cu-Mn oxides catalysts on
selective catalytic reduction of NOx with NH3 at low temperature. Rsc. Adv., 4 (2014) 25540-2551.
[12] Y. Doggali, Y. Teraoka, R. Mungse, et al. Combustion of volatile organic compounds over
CuMn based mixed oxide type catalysts supported on mesoporous Al2O3, TiO2 and ZrO2. J. Mol.
Catal. A: Chem., 358 (2012) 23-30.
[13] W.S. Kijlstra, D.S. Brands, H.I. Smit, et al. Mechanism of the selective catalytic reduction of
NO by NH3 over MnOx/Al2O3. Catal., 171 (1997) 219-230.
Advanced Materials Research Vol 1102 (2015) pp 17-21 Submitted: 2014-12-30
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.17

Hot deformation and cracking during compression of 21-4N steel


Ho-Kyun Oha Sang Min Parkb and Sun Ig Hong*c
Department of Advanced Materials Engineering, Chungnam National University, Daejeon, 305-764
Republic of Korea
a
maxycol@gmail.com, b kazia2040@naver.com, *c sihong@cnu.ac.kr

Keywords: Austenite stainless steel, Hot deformation behavior, Compression test, Dynamic
recrystallization, Strength

Abstract.

Hot compressive deformation and cracking behaviors of cast 21-4N steel were investigated over a
temperature range between 900~1250. The strength of the specimen at the outer surface of
ingot is greater than that in the inner part because of the initial rapid solidification and
corresponding fine structure at the outer surface. The initial flow stress hump in the stress-strain
curves can be attributed to the dynamic recrystallization, which requires the initial strain energy
build-up before the activation of grain nucleation process. With increase of temperature the critical
stain for dynamic recrystallization decreases and at high enough temperatures, no critical strain for
the build-up of strain energy is needed. At temperatures between 1000 and 1200, homogeneous
deformation occurred and no cracks were observed even after 50% strain. The cracks observed
below 1000 could be associated with the more difficult plastic flow in the presence of carbides
and nitrides. The presence of fatal crack at 1250 and the loss of hot ductility can be attributed to
the partial melting in austenite grain boundaries at high temperatures.

Introduction

Nickel has been the most important alloying element in the production of austenitic stainless steels.
However, the high and unstable nickel price has pushed many steel makers to find and develop
alternative austenitic stainless steels containing no or little nickel content. Nitrogen is not subject to
price volatility. And nitrogen is an enormously useful and cost-competitive element which is
effective in the stability of austenite phase and rendering the excellent corrosion resistance and
mechanical properties. In addition, nickel has is known to be an allergen, which also favors the
spreading use of nickel-less steels. In this respect, nitrogen-containing austenite stainless steels with
little or no nickel is expected to be more widely used in the future.
21-4N is an austenite steel alloy with high nitrogen and low nickel content. It is commonly used for
engine valve because it has high temperature mechanical stability and excellent corrosion properties.
The demand for 21-4N steel as automobile components has been increasing. However, it is known
that lack of toughness can make cracks due to difference of solubility of N and unbalanced diffusion
of alloying elements, and worsen properties of fatigue during shearing process. Also 21-4N treated
with solid solution heat treatment can cause surface problems like bulging by grain coarsening
during hot-working [1]. When 21-4N has fine grain size and homogeneous precipitations, improved
toughness and high quality of surface can be attained.
Lack of data related to hot deformation processes and phase transformations during heating and
thermomechanical treatment restricts the use of 21-4N to a large extent [1-3]. In current study,
compressive deformation characteristics of 21-4N steel at high temperatures were investigated.

EXPERIMENTAL PROCEDURE

21-4N steel with the chemical composition shown in Table 1 was cast and the hot compression
samples were cut from the 21-4N ingot. The primary objecvtive of this study is to investigate hot
18 Metal, Ceramic and Composite Materials

compresive deformability and cracking behavior of 21-4N steel. Compressive specimens taken from
the outer surface and inner (central) part of the ingot was compressed to 50% strain. To investigate
the effect of homogenization, some specimens were heat-treated at 1473K for 7 hours under
vacuum atmosphere before compression. (Group 1 : non-heat treated samples / Group 2 : Heat
treated samples).
The hot compression test was performed using a hot deformation simulator from Fiji Denpa at the
temperature ranging from 1173K(900) to 1523K(1250) at the displacement rate of 100mm/min.
After compression test, we got the images from an OM(Optical Microscope) and a SEM(Scanning
Electron Microscope). The surface and cross sections of compressed samples were observed using
OM and SEM.

Table 1. The chemical composition of 21-4N steel used in the present study (wt%)
Composition C Si Mn P S Ni Cr Mo Cu Al N
21-4N 0.520 0.040 9.450 0.031 0.020 3.53 20.28 0.08 0.09 0.014 0.403

RESULTS AND DISCUSSION

Hot compresive stress-strain responses of as-cast specimens taken from the outer surface (a) and
the inner (central) part (b) of ingot is presented in Figure 1. The stress level of curves decreases
with increase of temperature. The strength of the specimen at the outer surface of ingot is greater
than that in the inner part because of the initial rapid solidification and corresponding fine structure
at the outer surface. Despite the increasing cross sectinoal area with strain, the flow stress initially
reached the maximum and then decreased at the strain of 0.1. The initial flow stress hump can be
attributed to the dynamic recrystallization [4~6], which requires the initial strain energy build-up
before the activation of grain nucleation process. At 900, most pronounced flow stress hump was
observed because more strain energy build-up for the activation of grain nucleation process is

Fig 1. Compresive stress-strain responses of as-cast specimens taken from the outer surface (a) and
the central part (b) of ingot. The yield strength and UCS(Ultimate Compressive Strength) of as-cast
specimens taken from the outer surface (c) and the inner part (d) of ingot are also plotted as a
function of temperature.
Advanced Materials Research Vol. 1102 19

needed at lower temperatures. With increase of temperature the critical stain for dynamic
recrystallization decreases as shown in Fig.2 and at high enough temperatures, no critical strain for
the build-up of strain energy is needed. At 1523K(1250), the stress-strain curves of the outer
surface are similar to that of the inner section.
The compresive stress-strain responses of homogrnized specimens taken from the outer surface
(a) and the inner part (b) of ingot is presented in Figure 2. In general, the strength level after
homogenization decreased irrespective of location. The initial flow stress hump is also observed,
suggesting that the dynamic recrystallization occurs also after homogenization. The critical stain for
dynamic recrystallization is slightly greater because more strain energy build-up is needed because
of the lower strength after homogenization.

Fig 2. Compresive stress-strain responses of homogenized specimens taken from the outer surface
(a) and the central part (b) of ingot. The yield strength and UCS of homogenized specimens taken
from the outer surface (c) and the inner part (d) of ingot are also plotted as a function of temperature.

(a) (b)
Fig. 3 Cracks observed in the center of the specimen surface compressed to 50% strain at 900 in
the outer surface region (a) and the central part (b) of ingot. Surface bulging and roughening was
more pronouncedly developed in the center region of ingot.

The general cracking behavior is not greatly modified after homogenization and cracks observed
in the as-cast specimens will only be shown in this report. Fig. 3 shows the cracks observed on the
20 Metal, Ceramic and Composite Materials

specimen surface of the specimens compressed to 50% strain at 900 in the outer surface region (a)
and the central part (b) of ingot. The size and shape of cracks both in the surface and the inner
region appear to be similar. However, surface bulging and roughening was more pronouncedly
developed in the center region of ingot probably due to the more chemically segregated structure in
the inner part. The cracks become more fatal and developed more pronuncedly close to the upper
and lower anvil of the compression die in the specimens from both outer surface region and the
central part of ingot. Fig. 4 shows the cracks observed in the top or bottom of the specimen surface
in contact with the compression die. The specimens were compressed to 50% strain at 900 in the
outer surface region (a) and the central part (b)of ingot. Surface bulging and roughening developed
more pronouncedly in the top and bottome region of the specimen in contact with the compression
die.

(a) (b)
Fig. 4 More fatal cracks observed close to the top or bottom of the specimen surface in contact with
the compression die. Specimens were compressed to 50% strain at 900 in the outer surface region
(a) and the central part (b) of ingot.

(a) (b)
Fig. 5. Surface appearence of the specimens deformed at 1100 (a) and 1200(b).

(a) (b)
Figure 6. Surface appearnce (a) and the fatal crack (b)observed in the specimen deformed to 50%
strain at 1250
Advanced Materials Research Vol. 1102 21

At temperatures between 1000 and 1200, more homogeneous deformation occurred and no
cracks were observed even after 50% strain. Fig. 5 shows the surface appearence of the specimens
deformed at 1100 (a) and 1200(b). The cracks observed below 1000could be associated with
the more difficult plastic flow in the presence of Cr23C6 carbides and Cr2N nitrides which were
reported to be stable at below 1000 [3].
Figure 6 shows the surface appearnce and the fatal crack observed in the specimen deformed to
50% strain at 1250 Surface bulging and roughening is not as pronounced as in the specimen
deformed at low temperature (900) and cracks were not frequently observed. However, the cracks
once formed developed into large fatal cracks, resulting in the mechanical instability. The presence
of fatal crack at 1250 and the loss of hot ductility can be attributed to the partial melting in
austenite grain boundaries at 1250 C [1].

CONCLUSIONS
Hot compressive deformation and cracking behaviors of cast 21-4N steel were investigated over a
wide temperature range and the following conclusions were obtained:

1. The strength of the specimen at the outer surface of ingot is greater than that in the inner part
because of the initial rapid solidification and corresponding fine structure at the outer surface.
2. The initial flow stress hump in the stress-strain curves can be attributed to the dynamic
recrystallization, which requires the initial strain energy build-up before the activation of grain
nucleation process. With increase of temperature the critical stain for dynamic recrystallization
decreases and at high enough temperatures, no critical strain for the build-up of strain energy is
needed.
3. The strength level after homogenization decreased irrespective of location. The critical stain for
dynamic recrystallization is slightly greater because more strain energy build-up is needed because
of the lower strength after homogenization.
4. The size and shape of cracks both in the surface and the inner region appear to be similar.
However, surface bulging and roughening was more pronouncedly developed in the center region of
ingot probably due to the more chemically segregated structure in the inner part.
5. At temperatures between 1000 and 1200, more homogeneous deformation occurred and no
cracks were observed even after 50% strain. The cracks observed below 1000could be associated
with the more difficult plastic flow in the presence of carbides and nitrides which were reported to
be stable at below 1000
6. The cracks, although they were not frequently observed, developed into large fatal cracks once
they were formed, resulting in the mechanical instability. The presence of fatal crack at 1250 and
the loss of hot ductility can be attributed to the partial melting in austenite grain boundaries at high
temperatures.

References

[1] E. Erisir, U. Prahl, and W. Bleck, Metall Mater Trans A, 44 (2013) 5549-5555.
[2] H. Bern, ISIJ International, 36 (1996) 909-914
[3] R. L. Plau, C. Herrera, D. M. Escriba, P. R. Rios, A. F. Padilha, Materials Research, 10 (2007)
453-460.
[4] F.J. Humphreys : Mater Sci and Tech, 37 (1999) 125-139
[5] R.A. Petcovic, M.J. Lutin and J.J. Jonas : Acta Metall, 27 (1979) 1633-1648
[6] T. Sakai and J.J. Jonas : Acta Metall, 32 (1984) 189-209
Advanced Materials Research Vol 1102 (2015) pp 23-26 Submitted: 2015-01-02
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.23

The 6N01 Aluminum Components by Orthogonal Design Optimization

Fenglin Xu1,a, Bo Li2,b, Hongbo Wang1,c, JinJin Wu1,d, Dongfeng Li1,e,

Chenghou Zhou1,f, Xiaomin Wang3,g*


1.
CSR Nanjing Puzhen Co., Ltd., Nanjing 210031;
2.
Materials Science and Engineering College of Southwest Jiaotong University, Chengdu 610031
3.
School of Life science and Engineering of Southwest Jiaotong University, Chengdu 610031
a
xfl615@163.com; bsanpolincoln@163.com; c wangjelly1@163.com; dwujinjin19850616@163.com;
e
15950592340@163.com ; f zhouchenghoujiayou@163.com; g xmwang991011@163.com

Keywords: orthogonal design; 6N01 aluminum alloy; mechanical properties; microstructure

Abstract: In this study, we designed four components of 6N01 aluminum alloy by orthogonal design,
and analyzed the effect of the different elements for the mechanical properties and microstructure of
the designed material after casting, homogenization annealing, extrusion treatment and natural aging.
Considering the above series of tests of the data indicators, and thus orthogonal filter out two
materials that all aspects of performance are more excellent.

Introduction
Because of its light weight, good anti-corrosion and good tensile properties, aluminum alloys
has become a preferred material in the manufacture of high-speed train [1]. Currently, the aluminum
alloy for high-speed train body hollow sections mainly include 5000 series, 6000 series, 7000 series
alloy [2]. 6N01 aluminum alloys belonging to 6000 series ,which is part of Al-Mg-Si system, and
which can be heat strengthened and has a moderate strength, good formability and machinability
and other [3, 4] features.
Today, China is implementing the localization of the train bodys material, based on the
material from abroad. We have done research and development, mainly based on the German
standard, Japanese standard, GB and other relevant standards. But through research and literature
review we found that what they provided is so wide, and the material properties is not stable enough,
failed to reach the optimal level. So we made reference to the relevant range of each standard,
adjusted 6N01 aluminum alloy element content, reducing the standard range, to find the best alloy
composition.

Experimental material and procedure


Tab.1 The element content of 6082 aluminum alloy
Factor Level 1 Level 2
Mg 0.5-0.650.6 0.65-0.80.8
Si 0.4-0.60.4 0.6-0.80.7
Mn 0.025-0.150.05 0.15-0.250.2
Cr 0.005-0.0150.01 0.05-0.100.08
24 Metal, Ceramic and Composite Materials

Selecting the main alloying elements Mg, Si and major trace elements Mn, Cr of 6N01 materials for
content design. With Mg, Si as the factor 1, Mn as a factor 2, Cr as a factor 3, the content of three
factors designed to two levels, as shown in Table 1-1.
After orthogonal designed, the following results were obtained, as shown in Table 1-2, a total of
four design results obtained, namely four experimental materials were named 6N01-1 ~ 6N01-4.
Tab.2 6N01 aluminum orthogonal design
Factor 1 2 3
Classification Mg,Si Mn Cr
Experiment 1 0.5-0.650.6 0.025-0.15 0.005-0.015
0.4-0.6 0.4 0.05 0.01
Experiment 2 0.5-0.650.6 0.15-0.25 0.05-0.10
0.4-0.6 0.4 0.2 0.08
Experiment 3 0.65-0.80.8 0.025-0.15 0.05-0.10
0.6-0.80.7 0.05 0.08
Experiment 4 0.65-0.80.8 0.15-0.25 0.005-0.015
0.6-0.80.7 0.2 0.01
After pre-ground, taking the sample of finished product to do mechanical polishing and etching
with a mixed acid, then use the ProgRes C5 Zeiss microscope to observe the microstructure. Finally,
taking the four components of material after aging to conduct the tensile test, impact test and
hardness test.

Experimental Results and Discussions


microstructure morphology
Four kinds of alloy microstructure after homogenization morphology shown in Figure 1. Figure a, b,
c, d respectively represent 1 #, 2 #, 3 #, 4 # four components of the alloy. We can see the black
phase Mg2Si almost disappeared, dissolved back into the aluminum matrix, forming a
supersaturated solid solution and the white mesh structure is destroyed in a chain, a white granular
precipitate was fine, dispersed precipitates fishbone eutectic morphology gradually disappear,
intragranular segregation and grain boundary segregation basically eliminated, improving the alloy
mechanical properties, corrosion resistance strengthened. Further, it can be seen from the figure, the
first and third component reticular white slightly less than the other phase two components.
The microstructure after extrusion is shown in Figure 4. the a, b, c, d respectively represent 1 #, 2 #,
3 #, 4 # of four components of the alloy.
As can be seen from Figure 2, the 1st and the 2nd coarse grain is obvious, a # coarse grain thickness
is about 148.7m, 2 # is 245.4m, 3 #, there are coarse-grained layer having a thickness of 48m,
significantly less than the 1 # and 2 #, 4 # no coarse-grained layer. Contrast four kinds of material
composition can be obtained, 3 # Cr element is high, the 4# element for high levels of Mn, 1 # two
elements are low, 2 # are high. Mn elements can significantly improve the recrystallization
temperature to avoid recrystallization reduce coarse grain layer. Cr in a significant grain refinement
effect, it is these two elements can reduce the probability of formation of coarse-grained layers, but
the two elements are at high levels may occur when both the interaction should be reduced effect.

Mechanical Properties
Four different components of the aluminum alloy 6082 tensile strength, yield strength, elongation,
bending and impact toughness properties as shown in Table 3.
Advanced Materials Research Vol. 1102 25

a) 6N01S-1 aluminum alloy b) 6N01S-2 aluminum alloy

c) 6N01S-3 aluminum alloy b) 6N01S-4 aluminum alloy


Fig. 1 6N01 aluminum alloy microstructure after homogenization

a) 6N01S-1 aluminum alloy b) 6N01S-2 aluminum alloy

c) 6N01S-3 aluminum alloy d) 6N01S-4 aluminum alloy


Fig.2 6N01 aluminum alloy extrusion microstructure after aging
Compared the mechanical properties of the four components, we can conclude that the first
component of 6N01 aluminum tensile strength, yield strength, elongation, etc. Although not the best,
but its performance is superior to the other three summary materials. The fourth materials, while
elongation poor performance, but it showed excellent tensile strength and yield strength. Point of
26 Metal, Ceramic and Composite Materials

view of the element, a first are low, the fourth of Mg, Si, Mn are high, Cr is low. Further, we need to
elect our optimal material by orthogonal, Therefore, we calculated the value of each element K and
R values as Table 4 shows.
Tab. 3 Mechanical test results of 6082 aluminum alloys
No. b/MPa s/MPa L/% Bending performance
1# 264.50 248.41 10.43 Qualified
2# 234.87 210.97 10.67 Qualified
3# 290.07 275.67 8.4 Qualified
4# 309.24 300.80 6.23 Qualified

Tab. 4 The range analysis of the effect of different element on the mechanical properties
Tensile strength Yield strength Elongation rate
A(Mg,Si) B(Mn) C(Cr) A(Mg,Si) B(Mn) C(Cr) A(Mg,Si) B(Mn) C(Cr)
K1 499.37 554.57 573.74 459.38 524.08 549.21 21.1 18.83 16.66
K2 599.31 544.11 524.94 576.47 511.77 486.64 14.63 16.9 19.07
R 99.94 10.46 48.8 117.09 12.31 62.57 6.47 1.93 2.41
Filter Screening protocols A2B1C1 Screening protocolsA2B1C1 Screening protocols B1C2
the Factors A>C>B Factors A>C>B FactorsA>C>B
results The material 6N01S-4 The material6N01S-4 The material6N01S-2

From the screening results, we can draw, the final selection of the material is the6N01S-2 and
6N01S-4, but 6N01S-2 tensile strength is not enough, so removed from the Filter the results.

Conclusion
(1) Mg, Si element are the main elements to affect the strength of 6N01 aluminum alloy. High
levels of Mg, Si element is often expressed as high level of material strength.
(2) Mn alloy elements are key elements to affect 6N01 coarse-grained layer, Mn elements can
significantly improve the recrystallization temperature to avoid recrystallization reduce coarse
grain layer.
(3) When Mg, Si elements, Mn element is low, high levels of Cr element, the material elongation is
poor, or even not reach the basic requirement.
(4) The tensile strength, yield strength, elongation, and microstructure morphology as assessment
indicators, the final out of the6N01 aluminum alloy is preferably 1 # alloy and 4# alloy.

References
[1]. S. G. Yao, P. J. Xu: Traffic Transp. Eng, Vol. 4 (2004) No. 2, p. 40-44.
[2]. J.M. Dai: Light Alloy Fabrication Technology. Vol. 23 (1995) No. 5, p. 72-76.
[3]. M. Ericsson, R. Sandstrom: International Journal of Fatigue, vol. 25 (2003) no. 12, p.
1379-1387.
[4]. P.M.G.P. Moreiraa, A.M.P. de Jesusb, A.S. Ribeirob, et al: Theoretical and Applied Fracture
Mechanics, Vol. 50 (2008) No. 2, p. 81-91.
Advanced Materials Research Vol 1102 (2015) pp 27-32 Submitted: 2015-01-04
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.27

Emission Factors for Dioxin for Coal burning, Secondary Smelting and
Alloying of Aluminum, and Cement Kiln Processes
Yu-Yen Cheng1,a, Mei-Fang Lu2,b,Jim Juimin Lin1,c
1
Department of Safety, Health, and Environmental Engineering, National Kaohsiung First
University of Science and Technology, Nanzih, Kaohsiung City, 811, Taiwan
2
Institute of Engineering Science and Technology, National Kaohsiung First University of Science
and Technology, Nanzih, Kaohsiung City, 811, Taiwan
a
u0213814@nkfust.edu.tw, bemily@estc.tw, cenvelin@nkfust.edu.tw

Keywords: Emission factor, dioxin, coal burning, secondary smelting and alloying of aluminum,
cement Kiln

Abstract. While conducting research for dioxin emission factor for stationary source emission, it is
found that some factors come from overestimation. The cause of discrepancy for estimating process
comes often from lacking understanding of the process. The result of this study for secondary
aluminum refining emission factors indicates that if raw materials are scraps, the factors averaged at
541 ng I-TEQ/Ton-raw materials. If waste aluminum is used as the raw material, factors averaged at
1338 ng I-TEQ/Ton-raw material. The factors based on site sampling for coal-fired power
generation process is 24.84-549.62 ng I-TEQ/Ton-fuel. It varies according to coal sources.
Emission factor for cement producing process is 95.4-102.66 ng I-TEQ/Ton-raw material. Due to
fewer differences in operating traits, raw materials and fuel application for cement producing
process, emission factors have smaller differences. The reliability for emission variables is
relatively higher.

Introduction
Since dioxin was first discovered in flue gas and fly ash of urban municipal solid waste incinerator,
it has become a significant subject for numerous countries because its importance to toxicology and
its damage to living body. The main source of dioxin in the environment is human activity including
waste incinerating, thermal power generation, energy industry, metal smelting and chemical
engineering process. In developed countries such as US, UK, Italy, and Japan, the major dioxin
emission sources are often urban waste incinerators [1, 2]. Although individual conditions vary in
different nations, contribution of metal smelting cannot be neglected.
Research result in Europe pointed out that the major source for dioxin was urban waste
incinerator. The second biggest was welting furnace. Along with other important dioxin sources,
these dioxin emission sources accounted for 90% of annual dioxin emission in Europe. In UK,
urban waste incinerator is an important dioxin emission source. Its contribution is at 30-56% of the
total emission. As to metal smelting, including welting furnace, iron and non-iron metal smelting, it
also accounts for a considerable amount. Its contribution is at 15-26% of the total emission [1].
In Lombardy, a highly industrialized area in northern Italy with population at 9 million, its main
dioxin emission source is urban waste incinerator, which accounts for 32% of the total emission.
Electric arc furnace comes as the second at 26%. Other main emission sources are diesel vehicles
(10%), household wood burning (7%), secondary aluminum refinery (9%) and cement
manufacturing (5%). These six dioxin emission sources account for over 90% of the total emission
[2].
In Japan, the main dioxin emission source is waste incinerator including urban waste incinerator
and industry waste incinerator. Their contribution accounts for approximately 87% of the total
emission. Electric arc furnace comes at the second. Other important dioxin emission source is
welting furnace (3.2%), secondary zinc refinery (1.2%) and secondly aluminum refinery (1.6%).
28 Metal, Ceramic and Composite Materials

Emission estimation is the foundation for air pollution control and management. Establishing
emission factor can help to estimate emission amount. It can also be applied in emulating estimation
of air quality format, emission amount estimation for stationary pollution source permit application,
formulation of air quality administrative strategy and calculation of emission amount for air
pollution control fee. This study will aim at stationary pollution sources such as coal-fired boiler in
electricity and cogeneration profession, secondary aluminum refinery and cement kiln to construct
its professional dioxin emission factors to provide future reference for processing control strategies
formulation.

Method and Material


Dioxin Emission from Secondary Aluminum Refinery Process. In terms of procession,
secondary refinery for non-iron metal is mainly to fuse, smelt and cast refuse again to produce new
metal ingot. When refuse may contain impure items such as plastic, paste, and some solvents, it is
highly possible that dioxin will appear during incomplete burning. As secondary alluminin refinery
profession mainly uses recycled aluminum as raw materials that may contain oil, plastic, and other
organic ingredients. It is highly possible to produce dioxin during incomplete burning. The process
of aluminum refining includes two stages: pre-cleaning and smelting. Both may produce dioxin.
Dioxin Emission from Coal-fired Electricity Generating Process. Coal-fired boiler is boiler
using fire coal. As fire coal is burned in furnace to release energy to heat water or other organic heat
carrier such as heat transfer oil et al. to a certain degree or pressure, the boiler will function as a heat
energy power device. Fire coal boiler procession mainly uses coal. If the burning involves other
materials, dioxin and other heavy metal damages may occur in addition to four traditional pollution
items (Par, SOX, NOX, VOCs).
Dioxin Emission from Cement Procession. Rotary kiln is the heart for cement production. It is
direct heating for air and solid bodies. Most cement kilns use RDF [3]. Rotary kiln uses brown coal
which is complemented by waste tiles. It may produce dioxin. First take coal out of the coal dump
and transfer it via belt conveyor to coal mill to dry and grind to fine powder to increase burning
efficiency. Then coal powder will be delivered to coal storage before it is thrown respectively into
kiln head and preheater for burning. On the other hand, waste tiles will be transferred via weighing
system to preheater as complementary burning material to meet the needs of rotary kiln.
Sampling and Analysis of Dioxin. This study adopts methodology of NIEA A810.13B announced
by Taiwan Environmental Protection Administration (TWEPA) to proceed with on the spot testing.
This study authorized national certification body to conduct sampling and analysis of dioxin.

Results and Discussion


Dioxin Emission Factor for Secondary Aluminum Refinery Process. To compare factories A, B,
and C, their flue air dioxin emission factors (raw materials based) are shown in Figure 1. The result
indicates that emission factors are between 540.9-1349.45 ng I-TEQ/Ton-raw materials. Figure 2
shows emission factors of factory A, B, and C based on product. The result indicates that emission
factors are between 555.1-1565.186 ng I-TEQ/Ton-product (722.65558.72 ng
I-TEQ/Ton-product).
Table 1 is the emission factors for secondary aluminum refining. Removal efficiency of dioxin of
bag filter is assumed to be about at 85%, which is used to estimate emission factor before control
enforced.
It is obtained from the foregoing outcome that:
1. Factory A is different from factory B and C in terms of emission factor because factory A uses
aluminum waste and alloy as raw materials. They are mostly factory scraps used for
aluminum refining. As to factory B and C, they mostly used waste aluminum purchased. It
often contains paste, oil, plastic, robbery and other organic non-metal miscellaneous, which
are main cause of dioxin generation.
Advanced Materials Research Vol. 1102 29

2. Research result pointed out that emission factor for secondary aluminum refining is 2650 ng
I-TEQ/T-raw materials [4]. The discrepancy between that and this study is relatively
significant. The dioxin emission factor is estimated based on the composition analysis of raw
materials used, therefore the reliability is relatively lower.

1800
1600
Emission factor, ng I-TEQ/Ton-raw materials

1600

Emission factor, ng I-TEQ/Ton-product


1400
1400
1200
1200
1000
1000

800
800

600 600

400 400

200 200

0
0
A B C
A B C
Factory
Factory

Figure 1. Dioxin Emission Factor (raw materials Figure 2. Dioxin Emission Factor (product
based) for Secondary Aluminum Refinery Process based)for Secondary Aluminum Refinery
Process

Table 1. Emission factors for secondary aluminum refining obtained in this study
Raw Air pollution Emission factor Emission factor
Factor
materials control device (before control) (after control)
540.93
aluminum (ng I-TEQ/Ton-raw ---
A waste and none materials )
alloy 555.11
---
(ng I-TEQ /Ton-product)
8996.35 1349.45
(ng I-TEQ/Ton-raw (ng I-TEQ/Ton-raw
waste pulse jet materials ) materials )
B
aluminum baghouse 1565.18
10434.57
(ng I-TEQ
(ng I-TEQ /Ton-product)
/Ton-product)
8844.75 1326.71
(ng I-TEQ/Ton-raw (ng I-TEQ/Ton-raw
waste pulse jet materials ) materials )
C
aluminum baghouse 1474.06
9827.11
(ng I-TEQ
(ng I-TEQ /Ton-product)
/Ton-product)

Dioxin Emission factor for Coal-fired Power Generation Process. To compare factories A, B,
and C, their flue exhaust dioxin emission factors (based on fuel) are shown in Figure 3. The result
indicates that emission factors are between 24.84-549.62 ng I-TEQ/Ton-fuel (363.98294.14 ng
I-TEQ/Ton-fuel). Figure 4 is the emission factors of factories A, B, and C based on product. The
result indicates that emission factors are between 9.163-205.56 ng I-TEQ/Ton-product
(136.9355110.75 ng I-TEQ/Ton-product).
30 Metal, Ceramic and Composite Materials

Table 2 is the emission factors of coal-fired power generation process. The efficiency to destroy
dioxin in gas phase via SCR is at about 99%. It can be used to estimate emission factors before
control enforced. It is discovered from the foregoing outcome that:
1. Factory C and factory A or B have larger differences in terms of emission factors because
factory C uses mostly soft coal from Indonesia. The amount used in factory A and B is larger
and mostly Australian soft coal. Chlorine content and ash phase in Indonesian soft coal are
lower those in Australian soft coal. It is the cause of discrepancy for dioxin concentration and
emission factor.
2. Result from previous study indicates that power generation process using coal-fired with
emission factor of 114 ng I-TEQ/T-fuel has smaller discrepancy from the study. The test is
done within one factory. All operation unit sampling of flue exhaust and ash sampling are
estimated via mass-energy balance. It is different from this study which targets three factories
for flue gas sampling to calculate the emission factor. Dioxin emission factor for coal-fired
power plant is 620 ng I-TEQ/ton from result of previous study. That factor is obtained based
on site sampling and is comparable with the result of this study.

600 250
Emission factor, ng I-TEQ/Ton-raw materials

Emission factor, ng I-TEQ/Ton-product


500
200

400
150

300

100
200

50
100

0 0
A B C A B C

Plant Plant

sFigure 3. Dioxin Emission Factor (raw materials Figure 4. Dioxin Emission Factor (product
based) for Coal-fired Power Generation Process. based) for Coal-fired Power Generation
Process.
Table 2. Emission factors for coal-fired power generation obtained in this study
Raw Air pollution Emission factor Emission factor
Factor
materials control device (before control) (after control)
54962.03 540.62
(ng I-TEQ/Ton-raw (ng I-TEQ/Ton-raw
materials ) materials )
A soft coal
196.0
19609.13
(ng I-TEQ
(ng I-TEQ /Ton-product)
/Ton-product)
51750.23 517.50
ESP
(ng I-TEQ/Ton-raw (ng I-TEQ/Ton-raw
+
materials ) materials )
B soft coal FGD(wet)
205.56
+ 20556.33
(ng I-TEQ
SCR (ng I-TEQ /Ton-product)
/Ton-product)
2484.43 24.84
(ng I-TEQ/Ton-raw (ng I-TEQ/Ton-raw
materials ) materials )
C soft coal
9.16
24.84
(ng I-TEQ
(ng I-TEQ /Ton-product)
/Ton-product)
Advanced Materials Research Vol. 1102 31

Dioxin Emission factor for Cement Producing Process. To compare factory A and B, their flue
exhaust dioxin emission factors (based on raw materials) are shown in Figure 5. The result shows
that emission factors are between 95.4-102.66 ng I-TEQ/Ton-raw materials (66.055.07 ng
I-TEQ/Ton- raw materials). Figure 6 is the emission factors for factory A and B based on product.
The result indicates that emission factors are between 149.89-1177.13ng I-TEQ/Ton-product
(10919.26 ng I-TEQ/Ton-product).

120 200
Emission factor, ng I-TEQ/Ton-raw materials

180

Emission factor, ng I-TEQ/Ton-product


100
160

140
80
120

60 100

80
40
60

40
20
20

0 0
A B A B

Factory Factory

sFigure 5. Dioxin Emission Factor (raw materials Figure 6. Dioxin Emission Factor (product
based) for Cement Producing Process. based) for Cement Producing Process.

Table 1. Emission factors for cement producing obtained in this study


Air pollution Emission factor Emission factor
Factor
control device (before control) (after control)
513.32 102.66
(ng I-TEQ/Ton-raw materials ) (ng I-TEQ/Ton-raw materials )
A
885.65 177.13
(ng I-TEQ /Ton-product) (ng I-TEQ /Ton-product)
ESP
477.41 95.48
(ng I-TEQ/Ton-raw materials ) (ng I-TEQ/Ton-raw materials )
B
749.46 149.89
(ng I-TEQ /Ton-product) (ng I-TEQ /Ton-product)

Table 3 is the emission factor for cement process. The efficiency to eliminate dioxin via
electrostatic precipitators is at about 80%. It can be used to estimate emission variables before
control enforced. It is discovered from the foregoing outcome that:
1. Factory A and B has no considerable difference in terms of emission factors. It is evident that
operation trait and raw materials for cement producing have no considerable difference in
terms of fuel. Therefore, the emission factors estimated have no considerable difference.
2. Study by USEPA [5] indicates that cement kiln using waste tiles at a factor of 10.7 ng
I-TEQ/T-raw materials has larger difference from the result of this study. The contents of that
study mainly estimates by cement kiln using waste tiles without taking fuel and raw materials
into account while estimating. The source of the factor, moreover, is not based on site
sampling. Therefore, the reliability is relatively lower.
3. Study result pointed out that cement kiln using derived fuel has the emission factor at 2000 ng
I-TEQ/T-product [5]. The data is not estimated by site sampling. Therefore, the reliability is
relatively lower either.
32 Metal, Ceramic and Composite Materials

Conclusions
While conducting research for dioxin emission factor for stationary source emission, it is found
that some factors come from overestimation. The cause of discrepancy for estimating process comes
often from lacking understanding of the process. Thus improper processing factors are quoted. As a
result, during sampling of emission flue gas, errors in factor quoted and factor estimation are
relatively serious.
The result of this study for secondary aluminum refining emission factors indicates that if raw
materials are scraps, the factors averaged at 541 ng I-TEQ/Ton-raw materials. If waste aluminum is
used as the raw material, factors averaged at 1338 ng I-TEQ/Ton-raw material. The factors based on
site sampling for coal-fired power generation process is 24.84-549.62 ng I-TEQ/Ton-fuel. It varies
according to coal sources. Emission factor for cement producing process is 95.4-102.66 ng
I-TEQ/Ton-raw material. Due to fewer differences in operating traits, raw materials and fuel
application for cement producing process, emission factors have smaller differences. The reliability
for emission variables is relatively higher.

References
[1]. R.E. Alcock, R. Gemmill, K.C. Jones: Chemosphere Vol. 38 (1999), pp. 759-770.
[2]. S. Caserini, A.M. Monguzzi: Chemosphere Vol. 48 (2002), pp. 779-786.
[3]. Y. Li, R. Zhang: Procedia Environmental Sciences Vol. (16) (2012), pp. 633-640.
[4]. T. Ba, M. Zheng, B. Zhang, W. Liu, K. Xiao, L. Zhang: Chemosphere Vol. 75 (2009), pp.
1731178.
[5]. USEPA: EPA-454/R-97-003 (1997).
Advanced Materials Research Vol 1102 (2015) pp 33-36 Submitted: 2015-01-14
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.33

Effect of shot peening coverage on fatigue strengths of steels and


aluminum alloy

Junji Sakamoto1,a, Yong-Sung Lee2,b and Seong-Kyun Cheong2,c


1
Center for Risk Management and Safety Sciences, Yokohama National University, 79-5 Tokiwadai,
Hodogaya-ku, Yokohama-shi, Kanagawa, 240-8501, Japan
2
Department of Mechanical and Automotive Engineering, Seoul National University of Science and
Technology, 172 Gongreung-dong, Nowon-gu, Seoul, 139-743, South Korea
a
sakamoto@ynu.ac.jp, bs1lys@hanafos.com, cskjung@seoultech.ac.kr

Keywords: Shot peening, Coverage, Fatigue strength, Steels, Aluminum alloy

Abstract.

The relations between shot peening coverage and fatigue strength in the quenched-tempered and
annealed medium-carbon steels and aluminum alloy A7075-T6 were experimentally studied using a
rotating bending fatigue testing machine. It was found that the fatigue strengths of the annealed steel
and aluminum alloy were increased by shot peening. On the other hand, the fatigue strength of the
quenched-tempered steel were not increased by shot peening. Same shot peening condition causes the
positive effect and the negative effect depending on the type of material. Moreover, the valuable
coverage might depend on the type of material. More detailed work is necessary to resolve the issue.

Introduction
Shot peening (SP) is an effective and economical surface treatment method for improving the fatigue
strength of metallic materials by inducing compressive residual stress and hardening the layer near the
surface. In general, the SP effect is simply evaluated by main two parameters: coverage and peening
intensity. However, one of the problems with using coverage is that, the coverage is dominated by the
size of indentation and therefore coverage has no relation with the compressive residual stress and
hardening. Thus, it is unclear that coverage can be treated as the main parameter of shot peening.
The aim of this study is to clarify the applicability of coverage as main parameter of shot peening. In a
previous paper, we reported that the valuable coverage for improvement of the fatigue strength in
round bar of annealed medium-carbon steel (A steel) was 280-560% [1].
In this paper, the relations between shot peening coverage and fatigue strength in other materials
(quenched-tempered steel (QT steel) and aluminum alloy A7075-T6) were experimentally
investigated using a rotating bending fatigue testing machine and round-bar specimens.

Experimental methods
Materials and fatigue testing. The materials used in this study were steel and aluminum alloy
A7075-T6. Tables 1 list the chemical compositions of the tested materials. The annealed steel (A
steel) was obtained using a heat-treated at 845 C for 30 min and subsequently cooled in the furnace
[1]. The quenched-tempered steel (QT steel) was obtained using a heat-treated at 845 C for 1 h and
subsequently water-cooled and thereafter at 600 C for 1 h and subsequently cooled in the furnace. Fig.
1 shows the shape and dimensions of the fatigue test specimen. Fatigue tests were carried out using an
Ono-type rotating bending fatigue machine, in which the stress ratio was always -1, at room
temperature and atmosphere at a frequency of 30 Hz. The fatigue strength was defined as the
maximum stress amplitude under which the specimen endured 107 cycles.
34 Metal, Ceramic and Composite Materials

Table 1 Chemical compositions of the tested materials (mass%)


Material Chemical composition (mass%)
C Si Mn S Al Cr Ni Cu Fe
Medium-carbon steel
0.45 0.77 0.54 0.32 1.32 0.15 0.33 0.23 Bal.
Si Fe Cu Mn Mg Cr Zn Ti+Zr Al
A7075
0.15 0.29 1.6 0.14 2.4 0.19 5.7 0.25 Bal.

Fig. 1. Shape and dimensions of fatigue specimen.

Shot peening. Shot peening was performed under the processing parameters (Table 2) using
impeller type shot peening machine. Fig. 2 shows the surface states of the specimens after shot
peening for 10 seconds. In each case in Fig. 2, optical microscopy images (left side) are shown
together with image coverage, C%, converted pictures (right side). The average of C of the
quenched-tempered medium-carbon steel and aluminum alloy A7075-T6 were 16% and 33%,
respectively. The coverage prediction equation using one experimental data is as follows [2].

Cn = 1- (1-C1)n (1)

Where C1 is the coverage after 1 cycle (10 seconds in this study), Cn is the coverage after n cycles, and
n is the number of cycles. Table 3 shows the relations between shot peening exposure time and
coverage in the steels and A7075-T6. 98% surface coverage is considered as full coverage in this
study according to SAE J2277 [3]. Moreover, 200% coverage is defined as peening twice the
exposure time required achieving full coverage. The exposure time given full coverage of the
quenched-tempered medium-carbon steel and aluminum alloy A7075-T6 are 3 min 50 sec and 1 min
40 sec, respectively. The exposure times which were selected for the fatigue testing are indicated as
the symbols * in Table 3.

Table 2 Shot peening parameters


Shot ball Shot Almen
Material
Material Shape Diameter velocity Intensity
QT steel 70 m/s 0.62 mm A
A steel [1] Steel Spherical 0.8 mm 70 m/s 0.62 mm A
A7075-T6 30 m/s 0.24 mm A
Advanced Materials Research Vol. 1102 35

Fig. 2. Optical microscopy images of surfaces of specimens after shot peening for 10 sec.: (a)
Quenched-Tempered medium-carbon steel; (b) Aluminum alloy A7075-T6.

Table 3 Relations between exposure time and coverage in the steels and aluminum alloy
Exposure time
Material
10 sec 1 min 40 sec 2min 50 sec 3 min 50 sec 4 min 8 min 16 min
QT steel 16% 83% 95% 98% 105% 210%* 420%*
A steel [1] 21% 91% 98% 135% 140%* 280%* 560%*
A7075-T6 33%
98% 170% 230% 240%* 480%* 960%*

Experimental value obtained by observation.

Results and discussion


Fig. 3 shows the effect of shot peening coverage on fatigue strengths of QT steel, A steel [1], and
A7075-T6. In case of the A steel and A7075-T6, the fatigue strengths were increased by SP. On the
other hand, in case of the QT steel, the fatigue strength was not increased by SP. This indicates that
same shot peening condition causes the positive effect and the negative effect depending on the type
of material. The valuable range of coverage of the A steel for improving the fatigue strength in short
time was 280-560% [1]. In case of the A7075-T6, the fatigue strength of the 210% coverage
specimens are same as that of the 420% coverage specimens. This suggests that the valuable coverage
might depend on the type of material. More detailed work is necessary to resolve the issue. The
authors would report that in further work.
36 Metal, Ceramic and Composite Materials

Quenched-Tempered steel
Annealed steel [1]
A7075-T6

800
R = -1 Solid: broken
700 Open: unbroken
Stress amplitude, a [MPa]

600

500

400

300

200

100

0
0 200 400 600
Coverage, C [%]
Fig. 3. Effect of shot peening coverage on fatigue strengths of QT steel, A steel [1], and A7075-T6.

Summary
The relations between shot peening coverage and fatigue strength in the quenched-tempered and
annealed medium-carbon steels and aluminum alloy A7075-T6 were experimentally studied using
rotating bending fatigue test. It was found that the fatigue strengths of the annealed steel and
aluminum alloy were increased by shot peening. On the other hand, the fatigue strength of the
quenched-tempered steel were not increased by shot peening. Same shot peening condition causes the
positive effect and the negative effect depending on the type of material. Moreover, the valuable
coverage might depend on the type of material. More detailed work is necessary to resolve the issue.

References
[1] J. Sakamoto, Y.S. Lee, S.K. Cheong, Effect of shot peening coverage on fatigue limit in round
bar of annealed medium carbon steel, Journal of Mechanical Science and Technology 28(9)
(2014) 3555-3560.
[2] D. Kirk, An evolutionary guide to shot peening intensity measurement, The Shot Peener, Spring
(2012) 24-32.
[3] SAE Surface Enhancement Division (2009) SAE Standard J2277-Shot Peening Coverage
Determination.
Advanced Materials Research Vol 1102 (2015) pp 37-44 Submitted: 2015-01-17
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.37

Patch designing for damaged metallic structure repaired with adhesively


bonded composite patch
Zhang Ding1,a, Ding Wenyong1,b, Su Feng1, Jin Shijiao1
1
Naval Aeronautical and Astronautical University, Qingdao Branch, Qingdao, Shandong, China,
266041
a
zd21century@hotmail.com, b422627203@qq.com

Key words: Composite Patch Adhesively Bonded Repair, 3D Finite Element Model, Patch
designing, SIF

Abstract: Composite patch adhesively bonded repair is an advanced repairing method for damaged
metallic structure, and patch designing is important before repairing, which decides final repairing
effect. Based on 3D FEM, the model of aluminum plate with a central through crack adhesively
bonded repaired with composite patch was established, which can take into account residual thermal
stress and bending deformation caused by differences of thermal expansion coefficient. SIF was
calculated using virtual crack closure technology and compared to evaluate patch designing. The
results show that for through centrally cracked aluminum plate, rectangle boron/epoxy patch is the
best, and more plies whose fiber direction is parallel to loading direction are, better repairing can be
got.

Introduction
For metallic structures, corrosion damage and fatigue crack are the most common. Adhesively
bonded repair with composite patch is advanced, which can improve stress distribution in the
damaged area, and restore its function and strength partly or mostly[1,2]. So it is more and more
regarded and interested in science and engineering.
Patch designing, including several elements such as composite material, shape, geometrical
parameters, ply direction .etc, is crucial to composite patch adhesively bonded repair, because
rational design can make the most effective repairing. Lots of researches have been done on patch
designing[3-7], but they almost made several predigestions. The most common and important one is
residual thermal stress caused by difference of thermal expansion coefficient between metal and
composite and bending deformation induced by residual thermal stress are ignored, which have
some influences on repairing effect.
In this paper, the model of aluminum plate with a central through crack adhesively bonded
repaired with composite patch is established via three dimension finite element technology, which
can take into account residual thermal stress and bending deformation caused by it. Basing on the
model, different patch design is analyzed and evaluated via comparing stress intensity factor at
crack tip, which is computed with virtual crack closure technology.

1 3D FE modeling
3D FE model of aluminum plate with a central through crack adhesively bonded repaired with
rectangle composite patch is established using ABAQUS. Dimensions of aluminum plate are
180mm120mm3mm. The central through crack is 14mm. Composite patch is anisotropic,
which has 6 plies, and each ply is 0.1mm. So its total dimensions are 76mm38mm0.6mm. The
fiber direction of all plies is 0, which is parallel to loading direction. In order to evaluate different
38 Metal, Ceramic and Composite Materials

patch design, material properties are set in following sections. Length and width of adhesive are the
same as those of composite patch, and its thickness is 0.1mm. Properties of aluminum plate and
adhesive are shown in Table1-Table2.

Table1 Properties of aluminum plate


Module Yield strength Tensile strength Elongation percentage
68GPa 353Mpa 475 Mpa 18.1%

Table2 Properties of adhesive


Module Tensile strength
2.6 GPa 85 Mpa

Aluminum plate, adhesive and composite patch are meshed with C3D8R, COH3D8 and SC8R.
Noticeably, because of differences of thermal expansion coefficient among several materials,
residual thermal stress would be produced during curing, and then bending deformation is caused.
In addition, single-sided repair makes offset of neutral axis, and when loading, bending deformation
forms as well. But those two bending deformations are reverse, just shown in Fig. 1. Of course, they
have an influence on repairing effect. In order to simulate accurately, residual thermal stress and
bending deformation caused by it are calculated in curing process shown in Fig. 2 as the first step.
And then load = 100 Mpa is applied at the end surface. Because of symmetry, 1/4 model is
established, which is shown in Fig.3. After fist step analyzing, residual thermal stress and bending
deformation would be produced just like Fig.4.

Fig.1 Curvature due to (a) thermal residual stress and (b) neutral axis offset

Fig.2 Curing process


Advanced Materials Research Vol. 1102 39

Fig.3 FE model of centrally through cracked aluminum plate repaired with composite patch

Fig.4 Residual thermal stress and bending deformation during curing

2 Material designing
First, material designing is evaluated. It is considered that the centrally through cracked
aluminum plate is adhesively bonded repaired with boron/epoxy, graphite/epoxy and glass/epoxy
composite patch respectively. Material properties are shown in Table3. After modeling and
computing, SIF at the crack tip both on the repaired side and the unrepaired side are calculated with
VCCT, and results are shown in Fig.5.

Table3 Properties of composites


glass/epoxy graphite/epoxy boron/epoxy
E11 ( GPa ) 50 134 208

E22 ( GPa ) 14.5 10.3 25.4

E33 ( GPa ) 14.5 10.3 25.4

G12 ( GPa ) 2.56 5.5 7.2

G13 ( GPa ) 2.56 5.5 7.2

G23 ( GPa ) 2.24 3.2 4.9


12 0.33 0.33 0.1677
13 0.33 0.33 0.1677
23 0.33 0.53 0.035
1 (Fiber direction) 4.310-6 -0.710-6 4.510-6
40 Metal, Ceramic and Composite Materials

Fig.5 SIF at crack tip on both sides of aluminum plate

It can be observed that after repairing, SIF on the repaired side is reduced to a great extent, which
can be ignored. And by comparison, after repairing with boron/epoxy, SIF on both sides are the
least. Then those with graphite/epoxy repairing get larger. The most are SIF in the condition of
glass/epoxy repairing. The reason is that among all three materials, the stiffness of boron/epoxy is
maximal, and its thermal expansion coefficient is approximate to that of aluminum, which can
support more load for cracked aluminum plate and produce less residual thermal stress and bending
deformation. Finally, minimal SIF both on the repaired side and on the unrepaired side are made. So
it can be concluded that best repairing can be got when using boron/epoxy.

3 Shape designing
Besides rectangular composite patch, circular, elliptical, and filleting rectangular patch are also
introduced to evaluate shape designing. On the basis of Section2, boron/epoxy is used to model.
Dimensions of composite patch are shown in Table4. No matter what shape composite patch is, it
has 6 plies, and each ply is 0.1mm, whose fiber direction is 0, parallel to loading direction.
Adhesive shape is identical to that of patch. Its thickness is 0.1mm. Other parameters are the same
as those in Section1. 3D FE models of repaired structure with circular, elliptical, and filleting
rectangular patch are shown in Fig.6-Fig.8 respectively. Because of symmetry, 1/4 model is
established. After analyzing, SIF at the crack tip on both sides are calculated with VCCT, and
results are shown in Fig.9.

Table4 Dimensions of composite patch


Dimensions
Circular patch Radius 19mm
Elliptical patch Major axis 76mm Minor axis 38mm
Rectangular patch Length 76mm Width 38mm
Filleting rectangular patch Length 76mm Width 38mm Radius of corner 3mm
Advanced Materials Research Vol. 1102 41

Fig.6 Finite element model for adhesively bonded repair with circular patch

Fig.7 Finite element model for adhesively bonded repair with elliptical patch

Fig.8 Finite element model for adhesively bonded repair with filleting rectangle patch
42 Metal, Ceramic and Composite Materials

Fig.9 SIF at crack tip on both sides of aluminum plate

It also can be observed that after repairing, SIF on the repaired side is reduced to a great extent,
which can be ignored. By comparison, that with circular patch repairing is the least. Then that with
elliptical patch, rectangle patch and filleting rectangle gets incremental gradually. The increasing
amplitude is 45.94%, 61% and 69% respectively. Accordingly, on the unrepaired side, SIF with
circular patch repairing is the most. Then that with elliptical patch, filleting rectangle patch and
rectangle patch turns decremental gradually. The decreasing amplitude is 1.66%, 2.34% and 3.86%
respectively. Mentioned above, during curing, bending deformation is caused due to differences of
thermal expansion coefficient and residual thermal stress it induced. When loading on the end
surface, reverse bending moment is formed. Bending deformation opposite to that in curing step is
produced. And final bending deformation is shown in Fig.10, in which the repaired side is convex.
Obviously, the curvature of circular patch repaired structure is the least serious, then elliptical patch
repaired structure, filleting rectangle patch repaired structure and rectangle patch repaired structure
flexes more and more. So above result is made. There is no significance to analyzing shape
designing with SIF on the repaired side because it is so little. And SIF on the unrepaired side is used
for this purpose. Through comparing, it is easy to reason out that rectangle patch can get best
repairing.

Fig.10 Bending deformation in aluminum plate


Advanced Materials Research Vol. 1102 43

4 Ply direction designing


Ply direction designing is also analyzed. Rectangle boron/epoxy composite patch is adopted,
which contains 8 plies and each ply is 0.1mm. Ply direction is [0]8 , [(0 / 90)2 ]s , [ 45 / 0 / 90]s ,

[( 45)2 ]s respectively. Other parameters are the same as those in Section1. Similarly, because of
symmetry, 1/4 model is established. Because after repairing, SIF on the repaired side is too little to
be helpful to evaluate ply direction designing. After analyzing, only SIF on the unrepaired side is
calculated, which is shown in Fig.11.

Fig.11 SIF at crack tip on the unrepaired sides

It is explicit that with the number of 0 plies added, SIF on the unrepaired side goes down. In
composite, fiber bears much more load than matrix. And in 0 plies, fiber direction is parallel to
loading direction. So composite patch can support more load for cracked aluminum plate. Finally,
SIF is reduced.

5 Conclusions
Via 3D FEM, model of aluminum plate with a central through crack single-sided adhesively
bonded repaired with composite patch is established, which can consider the influence of bending
deformations caused by residual thermal stress and offset of neutral axis. Using this model, patch
designing is analyzed with SIF calculated with virtual crack closure technology. Following
conclusions can be drawn.
(1) Comparing to graphite/epoxy and glass/epoxy, boron/epoxy is the best choice for its high
stiffness and close thermal expansion coefficient to aluminum which can bear more load and
introduce less residual thermal stress and bending deformation. So in material designing, besides
stiffness, matching of thermal expansion coefficient is also important.
(2) With the most serious bending deformation of the repaired structure caused by residual
thermal stress and offset of neutral axis, in which the repaired side is convex, rectangle composite
patch is the most suitable to aluminum plate with a central through crack.
(3) Because fiber supports much more load than matrix, more plies whose fiber direction is
parallel to loading direction are, more load composite patch can carry for damaged metallic
structure.
44 Metal, Ceramic and Composite Materials

References
[1] Alan Baker. Bonded composite repair of fatigue-cracked primary aircraft structure[J].
Composite Structure, 1999, 47: 431-443
[2] Naveen Rastogi, Som R.Soni, Jason J.Denney. Analysis of composite patch repaired metallic
structures: an overview[R]. AIAA-98-1883
[3] Jiao Liang, Li Yan. Research on optimization of patch shape for bonding repair of aircraft
skin[J]. Adhesion, 2006, 27(4): 42-44. In Chinese
[4] Ching-Hwei Chue, Thomas Jin-Chee Liu. The effects of laminated composite patch with
different stacking sequences on bonded repair[J]. Composites engineering, 1995, 5(2): 223-230
[5] K. Kaddouri, D. Ouinas, B. Bachir Bouiadjra. FE analysis of the behaviour of octagonal bonded
composite repair in aircraft structures[J]. Computational materials science, 2008, 43: 1109-1111
[6] D. Ouinas, B. Bachir Bouiadjra, B. Serier, M. SaidBekkouche. Comparison of the effectiveness
of boron/epoxy and graphite/epoxy patches for repaired cracks emanating from a semicircular notch
edge[J]. Composite structures, 2007, 80: 514-522
[7] William P. Schonberg, Paresh Kumar, Madhukar Singh. Bonded composite patch design for
aircraft structures exhibiting cracking and corrosion[R]. AIAA-2002-1724
Advanced Materials Research Vol 1102 (2015) pp 45-49 Submitted: 2015-02-02
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.45

Application of the Betti number to prior austenite grain size analysis of


repeatedly quenched steel based on the Homology method
Koshiro Mizobe1,a, Katsuyuki Kida1,b, Wakana Matsuda1,c,
and Kazuaki Nakane2,d
1
University of Toyama, 3190, Gofuku, Toyama City, Toyama Prefecture, 930-8555, Japan
2
Osaka University, 1-7, Yamadaoka, Suita City, Osaka Prefecture, 565-0871, Japan
a
kmizobe@eng.u-toyama.ac.jp, b kida@eng.u-toyama.ac.jp, c s1170476@ems.u-toyama.ac.jp,
d
knakane2008@gmail.com

Keywords: Homology image analysis, Prior austenite grain size, Repeated quenching, Refinement,
Bearing steel (JIS, SUJ2)

Abstract
Grain refinement has a strong influence on material strength. Therefore, the automatic evaluation
method of the grain refinement level is useful. In order to develop a grain size analysis system, we
applied the homology theory to prior austenite grain analysis. We prepared repeatedly quenched
high-carbon high-chromium steel samples (JIS-SUJ2) and evaluated their microstructures in order to
investigate the relation between the Betti numbers (b0 and b1) and prior austenite grains. We found
that the Betti number (b1) is strongly correlated to the number of grains.

Introduction
It is a well known fact that repeated heating has a strong influence on material microstructure
size. Grange developed a method of producing ultrafine grained steel and investigated the relation
between repeated heating and material strength [1, 2]. In particular, later researchers found that
refining the prior austenite grain (PAG) size improves the material strength in high-carbon
high-chromium steels (JIS-SUJ2). Our research group has investigated the repeated heating of high
carbon steels [3-5]. We applied the Homology theory to steel structure analysis [7] and classified
three sizes of repeatedly quenched PAGs [8].
In this study, pictures at three levels of magnification were evaluated using the Homology
method in order to improve the reliability of this method.

Experimental procedure
Specimen. We prepared JIS-SUJ2 bar specimens which were similar to those used in our previous
work [6]. The features are as follows. The chemical composition of the SUJ2 bars in weight percent
was: 0.99C, 1.38Cr, 0.4Mn, 0.26Si, 0.11Cu, 0.02Mo, 0.019P, 0.08Ni, 0.007S, 0.002Ti, 80N, 5O (N,
O: ppm). The microstructure were ferrite and spheroidized cementite.
Figure 1 shows a schematic illustration of the heat treatment process. Specimens were furnace
heated for 40 minutes at 850, quenched and subsequently tempered at 180 for 60 minutes.
Specimens quenched once, twice and three times are referred to as Q1T1, Q2T1 and Q3T1,
respectively. The hardness values after the heat treatment were 772HV (+27, 39) for samples
quenched once, 791HV (+13, 22) for samples quenched twice and 808HV (+14, 27) for samples
quenched three times.
After tempering, the samples were polished, and then they were etched by immersion for 5
minutes in a solution of saturated picric acid (Van Gieson solution) consisting of 2 ml of Teepol and 3
drops of hydrochloric acid. The sample surfaces were lightly polished for between 10 and 30 seconds
using Buehler 1 m colloidal silica. The structure of the etched surfaces was observed using a laser
confocal microscope (Keyence VK-9700). Table 1 shows the observation samples in detail. We
prepared four photographs for each of six observation conditions. Average of the Betti numbers was
calculated based on these four photographs and the typical examples are shown in Figs. 2 and 5.
46 Metal, Ceramic and Composite Materials

850
Oil quenching
40 min. 180
Air cooling
60 min.
(a) Quenched once
850 850
Oil quenching Oil quenching 180
40 min. 40 min. Air cooling
60 min.
(b) Quenched twice
850 850 850
Oil quenching Oil quenching Oil quenching 180
40 min. 40 min. 40 min. Air cooling
60 min.
(c) Quenched three times
Fig. 1 Schematic illustration of heat treatment process [6].

Table 1 Observation samples in detail


Observation area
Figure No. Q1T1 Q2T1 Q3T1
(m2)
Fig. 2 6.6103 4 4 4
4
5.910 4 NA NA
4
Fig. 5 2.310 NA 4 NA
3
6.610 NA NA 4

Homology image analysis. Homology is a mathematical method used to replace a quantitative


description of a figures characteristics with an algebraic formula [9-12]. While the structure images
of materials have a variety of shapes, the Betti number is independent of their shapes. In this study, we
used ChomP programs [13, 14] in order to calculate the Betti numbers of the grain structures.

Results and discussion


Observation. We observed microstructures in order to investigate the relation between the number of
times a sample was quenched and the Betti numbers. Figure 2 shows the microstructures of repeatedly
quenched samples whose observation areas are 6.6103 m2. These specimens were etched by picral
and the black lines represent the PAG boundaries. Figure 3 shows the PAG outlines which were
drawn based on Fig. 2. In the case of Q1T1, the PAGs were big and evenly distributed. In Q2T1, some
big PAGs and some colonies including several small refined PAGs were unevenly distributed. The
PAGs in Q3T1 were evenly distributed and most of them were smaller than those in both the Q1T1
and Q2T1 samples. These features are what we found in our previous work [15].
Figure 4 shows the distribution of the Betti numbers (b0 and b1) in 6.6103 m2 areas. The Betti
numbers (b0 and b1) were calculated based on four photographs of areas of the same size. The average
of the b0 numbers were less than 10 and not related to the number of times a sample was quenched.
However, the average of the b1 numbers did increase with the quenching times. These results were
similar to what we obtained in our previous work [16].
Advanced Materials Research Vol. 1102 47

Fig. 2 Microstructure of the repeatedly quenched samples observed in 6.6103 m2 area.

Fig. 3 Outlines of the PAGs observed in 6.6103 m2 area.

250
b1
The Betti numbers (b0 and b1)

b0
200

150

100

50

0
1 2 3
Quenchingtimes
Quenching times

Fig. 4 Distribution of the Betti numbers (b0 and b1) in 6.6103 m2 area.

In order to investigate the effect of the number of grains on the Betti numbers (b0 and b1), we
prepared photographs in three sizes. Figure 5 shows microstructures of repeatedly quenched samples.
The photograph area of Q1T1 is 5.9104 m2, Q2T1 is 2.3104 m2 and Q3T1 is 6.6103 m2. We
selected these observation areas in order to calcuate the Betti numbers in the area which have almost
the same number of PAGs. Figure 6 shows the PAG outlines which were drawn based on Fig. 5.
There are a lot of small PAGs in each pictures.
Figure 7 shows the distribution of the Betti numbers (b0). It can be seen that the b0 numbers are
not increased by the number of times the sample quenched, however, in Fig. 8, the distribution of the
Betti numbers (b0) per 100 m2 did are increased by the number of times the sample quenched. It is
concluded that the Betti number (b1) is strongly correlated to the number of quenched times.
48 Metal, Ceramic and Composite Materials

Fig. 5 Microstructure of the repeatedly quenched samples


(5.910 m2 in Q1T1, 2.3104 m2 in Q2T1 and 6.6103 m2 in Q3T1)
4

Fig. 6 Outlines of the PAGs


(5.9104 m2 in Q1T1, 2.3104 m2 in Q2T1 and 6.6103 m2 in Q3T1)

20 5
4 2 4 2
Observation area 5.910 m Observation area 5.910 m
4 2
4 2 5.910 m
5.910 m 4 4 2
4 2 Observation area 2.310 m
15 Observation area 2.310 m
2

3 2
6.610 m
The betti number ( b0)

b1 number per 100 m

3 2
6.610 m
3
10
2

5
1

0 0
1 2 3 1 2 3
Quenching times Quenching times

Fig. 7 Distribution of the Betti number (b0) Fig. 8 Distribution of b1 number per 100 m2

Conclusion
Bearing steel samples (JIS-SUJ2) were quenched once, twice and three times. The Betti numbers of
their prior austenite grains were observed by a numerical Homology method. We concluded that the
Betti number (b1) per unit area is strongly correlated to number of grains.

Acknowledgement
A part of this research was supported by the Grant-in-Aid for JSPS Fellows, Grant Number 25-4761.
Advanced Materials Research Vol. 1102 49

References
[1] R. A. Grange, E. R. Shackelford, U.S. Patent 3,178, 324, (1965)
[2] K. Mizobe, E. C. Santos, T. Honda, H. Koike, K. Kida and T. Shibukawa, Observation of
non-metallic inclusions on repeatedly quenched SAE52100 bearing steel fracture surfaces,
International Journal of Materials and Product Technology, Vol.44, No. 3/4, pp.227-239, (2012)
[3] H. Koike, E. C. Santos, K. Kida, T. Honda and J. Rozwadowska, Effect of Repeated Induction
Heating on Fatigue Crack Propagation in SAE 52100 Bearing Steel, Advanced Material Research,
Vol. 217-218, pp. 1266-1271, (2011)
[4] E. C. Santos, K. Kida, T. Honda, J. Rozwadowska, K. Houri and K. Hashimoto, Influence of
Cyclic Heat Treatment and Oxygen Content on the Rolling Contact Fatigue of JIS SUJ2 Bearings,
Advanced Material Research, Vol. 217-278, pp. 982-987, (2011)
[5] M. Koga, E. C. Santos, T. Honda, K. Kida and T. Shibukawa, Investigation of wear in
induction-heated AISI E 52100 steel bars under reciprocating motion, International Journal of
Materials and Product Technology Vol. 44, No. 3/4, pp. 240-251, (2012)
[6] K. Mizobe, T. Honda, H. Koike, E. C. Santos, T. Shibukawa and K. Kida, Relationship between
repeatedly quenching and fisheye cracks around TiN and Al2O3 inclusions in high carbon bearing
steel, Material Research Innovations, Vol. 18 (Supplement 1), pp. S60-S65, (2014)
[7] K. Nakane, K. Kida, K. MizobeHomology Analysis of Prior Austenite Grain Size of SAE52100
Bearing Steel Processed by Cyclic Heat Treatment, Advanced Materials Research, Vol. 813, pp.
116-119, (2013)
[8] K. Nakane, E. C. Santos, T. Honda, K. Mizobe and K. Kida, Homology Analysis of Structure of
high carbon bearing steel, Effect of Repeated quenching on Prior Austenite Grain Size, Materials
Research Innovations, Vol. 18 (Supplement 1), pp. 33-37, (2014)
[9] M. Arkowiz, Introduction to homotopy theory (universitext), London, Springer Press. (2011)
[10] A. Dold, Lectures on algebraic topology in Classics in mathematics, Berlin, Springer-Verlag,
(1995)
[11] A. Hatcher, Algebraic topology, Cambridge, Cambridge University Press, (2001)
[12] S. V. Matveev, Lectures on algebraic topology in EMS Series of lectures in mathematics,
Zurich, European Mathematical Society, (2006)
[13] CHomP (Computational Homology Project), http://chomp.rutger-s.edu/
[14] T. Kaczynski, K. Mischaikow and M. Mrozek, Computational homology, Applied
Mathematical Sciences, Vol. 157, Heidelberg, Springer-Verlag, (2004)
[15] K. Mizobe, W. Matsuda, M. Matsushita, T. Shibukawa and K. Kida, Effect of twice quenching
on prior austenite grains and rotating bending fatigue cracks in SUJ2 steel, Applied Mechanics
and Materials, Vol. 620, pp. 443-448, (2014)
[16] K. Kida, K. Mizobe, R. Arai and K. Nakane, The Betti number of prior austenite grain of
repeated quenching bearing steel (JIS, SUJ2), Advanced Materials Research, in press.
Advanced Materials Research Vol 1102 (2015) pp 51-54 Submitted: 2014-12-31
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.51

Effect of Heat Treatment on the Mechanical Properties and Interface


structure of 3-ply Ti/Cu/Ti Clad Composite
Yong Keun Kim 1,a, Pyung Woo Shin2,b, Sun Ig Hong*1,c
1
Department of Advanced Materials Engineering, Chungnam National University,
Daejeon, 305-764, South Korea
2.
Department of Metallurgy and Materials Eng, Changwon National University, #9 Sarimdong,
Changwon, Kyungnam 641-773, South Korea
a
yongkeunkim84@gmail.com, bpwshin@chanwon.ac.kr, c,*sihong@cnu.ac.kr

Keywords: Titanium, Copper, Heat-treatment, Clad, Composite, Interface, Intermetallics

Abstract. The effect of heat treatment on the mechanical properties and interface structure of 3-ply
Ti/Cu/Ti clad composite was investigated. Strength decreased and the ductility increased with
increase of heat treatment temperature. The gradual increase of hardness at 700 and 800
indicates the growth of intermetallic compounds at the interface. No visible intermetallic compound
formation was observed up to 400. The intermetallic layer grew very rapidly above 600 and its
thickness reached ~10m after heat treatment at 800. The absence of cracks emanating from the
corners of the indentation mark indicates that intermetallics in Ti/Cu/Ti clad are ductile enough to
accommodate the micro-plastic flow from indentation. Four intermetallic layers at the interface
were confirmed to be Cu4Ti, Cu3Ti2, CuTi and CuTi2 based on the EDS spectra, XRD and phase
diagram analyses.

Introduction
The clad metals consisting of two or more metal layers have been developed because of their
unique combination of properties such as high strength, high conductivity and excellent corrosion
resistance [17]. The properties of clad composite material are determined by the selection of
component materials to be joined and the stacking structure of different materials with various
thicknesses and the interface structure and the properties between different materials [35]. Clad
composite material with different metals and alloys along with various combination of properties
have been used in various industrial fields [1-7]. For example, titanium has high strength and
corrosion resistance and copper has excellent electrical conductivity. Ti/Cu clad material has been
used effectively as an electrode material for chemical processing in corrosive environments such as
chlor-alkali, electrolysis, metal recovery, and plating facilities. Titanium cladding can be used to
increase both the mechanical strength and the corrosion resistance of the clad composite [8]. The
aim of this study was to investigate the effects of heat treatment on mechanical properties and
microstructure of 3-ply Ti/Cu/Ti clad metal fabricated by roll-bonding.

Experimental method
In order to investigate the effect of heat treatment on the mechanical properties of the warm-roll-
bonded and cold rolled 3-ply Ti/Cu/Ti clad composites, heat-treatment was carried out between
200~800 for 2 hours. Hardness and mechanical testing of heat-treated Ti/Cu/Ti clad composites
were performed using Vickers hardness tester and Universal Testing Machine respectively.
Microstructure of specimens was observed by field emission scanning electron microscope
(FESEM) equipped with energy dispersive X-ray spectroscopy (EDX). In order to identify the thin
intermetallic layers at the interface more accurately, energy dispersive X-ray spectroscopy (EDX)
and X-ray diffraction (XRD) analyses were carried out.
52 Metal, Ceramic and Composite Materials

Result and Discussion


Fig. 1 (a) shows the variation of strength and ductility of Ti/Cu/Ti clad composite with heat
treatment temperature. Tensile strength and ductility of cold-rolled specimen was observed to be
562MPa and 10%, respectively. Ti/Cu/Ti clad composite after heat treatment at 800 exhibited
the tensile strength of 234MPa and 48% ductility. As expected, the strength decreased and the
ductility increased with increase of heat treatment temperature. Fig. 1(b) shows the variation of
Vickers micro hardness of Ti, Cu and the interface as a function of heat treatment temperature. The
drops of hardness value in copper at 400 and in Ti at 480 are attributed to the recrystallization
of Cu and Ti. The hardness of interface region also decreased with the decrease of hardness in Ti
and Cu, reflecting the overall softening in Cu and Ti matrix. One interesting observation in Fig. 1(b),
however, is the gradual increase of hardness at the interface. The gradual increase of hardness at
700 and 800 indicates the growth of intermetallic compounds at the interface.

Fig. 1. Variation of strength and ductility (a) and Vickers micro hardness (b) of Ti/Cu/Ti clad composite as a
function of heat treatment temperature.

.
Fig. 2. Vickers micro hardness indentation mark at the interface of Ti/Cu/Ti clad composite; as-cold-rolled (a)
and heat treated at 600 (b), 700 (c) and 800 (d).

The highest hardness value (~163 Hv) at the interface after heat treatment at 800, although it is
greater than that at the interface of as-cold-rolled clad composite, is still lower than the hardness
values (~220Hv) of Ti matrix before recrystallization. The highest hardness value at the interface in
Fig. 1 is far smaller than that (~400 Hv) of the interface observed in Cu/Al/Cu clad composite [9].
The smaller hardness at the interface of Ti/Cu/Ti clad composite suggests that the intermetallics is
not as brittle as those observed in Cu/A/Cu [6,9] and Mg/Al/STS [2,5]. Fig. 2 shows the Vickers
micro hardness indentation mark at the interface of Ti/Cu/Ti clad composite; as-cold-rolled (a) and
heat treated at 600 (b), 700 (c) and 800 (d). No interfacial cracks were induced by the
indentation in the as-cold-rolled and heat-treated Ti/Cu/Ti clad composites, indicating the sound
interfacial bonding. In the heat-treated Ti/Cu/Ti clad composite at 800 with the appreciable
intermetallic layer formation, no emanating cracks from the corners of the indentation mark, which
Advanced Materials Research Vol. 1102 53

is typically observed in the clad with brittle intermetallic layers, were observed. The absence of
cracks from the corners of the indentation mark in Fig. 2(d) indicates that intermetallics in Ti/Cu/Ti
clad are ductile enough to accommodate the micro-plastic flow from indentation.

Fig. 3. FESEM images of the near-interface region of Ti/Cu/Ti clad composite with heat treatment from 200 up
to 800

Fig. 3 shows the FESEM images of the near-interface region of Ti/Cu/Ti clad composite with
heat treatment from 200 up to 800.The FESEM images in Fig. 3 clearly revealed that there is
no visible intermetallic compound formation up to 400. Above 450, a very thin intermetallic
compound layer was observed. The intermetallic layer grew very rapidly above 600 and its
thickness reached ~10m after heat treatment at 800. Fig. 4 shows the FESEM image of Ti/Cu
interface and EDX spectra from interface intermetallic layers of Ti/Cu/Ti clad composite heat
treated at 800. At the interface between Cu and Ti, 4 intermetallic layers are clearly visible. The
phase diagram shows the presence of 5 intermetallics Cu4Ti, Cu3Ti2, Cu4Ti3, CuTi, CuTi2, at or
below 800 [10]. An intermetallic phase Cu2Ti was found to be stable above 1120, which is not
likely formed by heat treatment up to 800. The atomic ratio obtained from EDS spectra analyses
in Fig. 4 reveals that the layer in contact with Cu matrix is Cu4Ti and next two layers are Cu3Ti2 and
CuTi. EDS spectra was not taken from the very thin intermetallic layer between Ti and CuTi
because it is not likely to be accurate. But the comparison of three layers from Cu to Ti in Fig. 4
with the intermetallic phases in phase diagram [9] strongly suggest that the thin layer is CuTi2. The
presence of Cu4Ti, Cu3Ti2, CuTi and CuTi2 was confirmed by XRD analysis.

Fig. 4. FESEM image of Ti/Cu interface and EDX spectra from interface intermetallic layers of Ti/Cu/Ti clad
composite heat treated at 800.
54 Metal, Ceramic and Composite Materials

Summary
As a result of the study on the tensile properties and microstructure of Ti/Cu/Ti clad composite,
the following conclusions are obtained:
[1] Strength decreased and the ductility increased with increase of heat treatment temperature. The
drops of hardness value in copper at 400 and in Ti at 480 are attributed to the
recrystallization of Cu and Ti.
[2] The gradual increase of hardness at 700 and 800 indicates the growth of intermetallic
compounds at the interface. The absence of cracks from the corners of the indentation mark
indicates that intermetallics in Ti/Cu/Ti clad are ductile enough to accommodate the micro-
plastic flow from indentation.
[3] Four intermetallic layers was suggested to be Cu4Ti, Cu3Ti2, CuTi and CuTi2 based on the EDS
spectra, XRD and phase diagram analyses.

Acknowledgement
This work was supported by the 2nd phase of the Fundamental R&D Programs for Core
Technology of Materials funded by Ministry of Trade, Industry and Energy (2014-2015).

References
[1] J.S. Ha, S.I. Hong, Mater. Des. 51(2013)293299.
[2] I. K. Kim, Hong SI. Metall Mater Trans. A 44 (2013) 3890-3900.
[3] I. K. Kim and S. I. Hong. Mater. Des. 57 (2014) 625-33.
[4] K.S. Lee, D.H. Yoon, H.K. Kim, Y.N. Kwon, Mater. Sci. Eng. A 556 (2012) 319330.
[5] I.K. Kim, S.I. Hong, Mater. Des. 49 (2013) 935944.
[6] I.K. Kim, S.I. Hong, Mater. Des. 47 (2013) 590598.
[7] H. Kim, S. I. Hong, Materials and Design 67 (2015) 42-49
[8] F. Xu, J. C. Fredette, R. A. Holt, R. B. Rogge, D. Pickard and L. Tuck, Journal of Neutron
Research, 15 (2007) 259266.
[9] I.K. Kim, J.S. Ha, S.I. Hong, Korean J. Met. Mater. 50 (2012) 939-948.
[10] H. Okamoto, Journal of Phase Equilibria 23 (2002) 549-550.
Advanced Materials Research Vol 1102 (2015) pp 55-58 Submitted: 2014-12-28
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.55

Effect of heat treatment on galvanic corrosion of Cu/Al/Cu clad soaked


in 3.5% NaCl brine solution
Gyeong Tae Kang, Sun Ig Hong*
Department of Advanced Materials Engineering, Chungnam National University, Daejeon, Korea.
*corresponding author
k_372090v@naver.com, *sihong@cnu.ac.kr

Keywords: galvanic corrosion, copper, aluminum, clad, brine, heat treatment, intermetallic.

Abstract. The galvanic corrosion of Cu/Al/Cu clad in 3.5% NaCl brine solution has been studied.
No visible intermetallic compounds were found at the Cu/Al interface after roll-bonding, But
Intermetallic layer grew up to 40m after heat treatment 500 for 10hours. The galvanic corrosion
occurred significantly in the Al region in contact with Cu. The depth of the dissolved Al gradually
decreased with distance away from the Cu/Al interface. Both IDR width and IDR depth increased
steadily with increase of soaking time. The IDR width and depth of as-roll-bonded Cu/Al/Cu is far
greater than those of heat-treated Cu/Al/Cu at 500 for 10hours, suggesting the presence of
intermetallic compounds influences the galvanic corrosion behavior.

Introduction
Copper/aluminum clad composites have been widely studied because of their advantages associated
with high conductivity, low density and price competitiveness over copper and copper alloys [1-5].
Cu/Al clad sheet can almost reduce 3550% in weight, with the electrical and thermal conductivity
equivalent to those of some copper alloys [6] and Cu/Al clad composite is frequently used for
conductor wires and bus-bar conductor joint [1, 2]. Cu/Al is a potential alternative to pure copper
for automotive applications in power cables, including battery cables and power leads to electronic
components because it offers a lightweight alternative that provides superior conductivity.
Not only interfacial mechanical locking but also mechanochemical bonding was suggested to work
at the contact surface during Cu and Al joining [2,7,8]. Thin intermetallic compounds were found to
be formed even after low temperature roll-bonding [2]. The brittle intermetallic compound layers
have been reported to increase at elevated temperatures deteriorating the electrical and mechanical
reliability of Cu/Al clad composites[1-5]. Interfacial structure and properties of Cu/Al clad
composites have been studied by many investigators[1-5,7,9]. However, the investigations on the
effect of interfacial intermetallic on the corrosion properties are still few. The objective of this study
is to examine the influence of intermetallic compounds on galvanic coupling between copper and
aluminum.

Experimental
In this study, Cu/Al/Cu clad was fabricated using rolling process. Materials used in this study were
OFHC copper(99.9 % purity, 0.002%O, 0.015%P, 0.002%Zn) and 1060 aluminum of the
commercial purity (99.5% purity, 0.20% Si, 0.25% Fe, 0.05%Cu). The total thickness of clad
composite plate was 50mm and that of Cu and Al layer was 5 mm and 40mm respectively. In order
to investigate the effect of heat treatment on corrosion properties of the extruded Cu/Al/Cu
composite plate, the 3-layered composites were heat-treated 500 for 10 hours. Corrosion
properties of the Cu/Al/Cu composites were examined soaking in 3.5wt% NaCl brine solution for
24, 48 and 100 hours at room temperature. Interfacial structure and morphology of Cu/Al/Cu clad
metal were examined by a scanning electron-microscope(SEM). For chemical analyses of
intermetallics and the corrosion reaction products, energy dispersive X-ray spectroscopy(EDS) was
used.
56 Metal, Ceramic and Composite Materials

Result and Discussion

Intermetallic reaction layer was reported to increase with temperature and time and the layer with a
thickness of about 2 m was observed to be formed after annealing at 300 for 3hours in
Cu/Al/Cu layered composite [1]. With the increase of annealing temperature to 400, and 500,
the reaction layer continues to thicken to 12m, and 28 m respectively [1]. Fig. 1 shows the
interface region of as-roll-bonded Cu/Al/Cu clad after soaking in 3.5% NaCl brine for (a)24, (b)48,
(c)100 hours. In Fig. 1, Al in contact with Cu was dissolved away with increase of soaking time and
the depth of the dissolved Al gradually decreased with distance away from the interface. The less
noble Al is the anode and tends to undergo accelerated corrosion, while the more Cu will tend to
experience reduced corrosion effects. The region with the Al dissolution is defined as the interfacial
dissolution region (IDR). The width from the interface to the point where no dissolution occurs is
defined as IDR width, WI and the depth from the top surface to the bottom of IDR at the interface is
defined as IDR depth. Except in this zone, aluminum was not corroded. Both IDR width and IDR
depth increased steadily with increase of soaking time.

Fig. 1 SEM images of Cu/Al interface of as-roll-bonded Cu/Al/Cu soaked in 3.5% NaCl brine for
(a)24, (b)48, (c)100 hours. 3.5% NaCl brine

Fig. 2 SEM images of Cu/Al interface heat-treated Cu/Al/Cu (at 500 for 10 hours) soaked in for
(a)24, (b)48, (c)100 hours.

Fig. 2 shows the Cu/Al interface heat-treated Cu/Al/Cu (at 500 for 10 hours) soaked in 3.5%
NaCl brine for (a)24, (b)48, (c)100 hours. The intermetallic compound layers grew up to 40m after
10 hr heat-treatment. The galvanic coupling between Cu and Al can be disturbed in the presence of
intermetallics, which may influence the galvanic corrosion behavior. Indeed, the dissolution rate of
Al in 3.5% NaCl brine solution decreased significantly as shown in Fig. 2. Fig. 3 displays the
variation of IDR width and depth as a function of soaking time in brine solution. As shown in Fig. 3,
IDR width and depth of increased with soaking time. It should be noted that IDR width and depth of
as-roll-bonded Cu/Al/Cu is far greater than those of heat-treated Cu/Al/Cu at 500 for 10hours.
The IDR width reached up to 196m after soaking for 100hours as-roll-bonded Cu/Al/Cu. In the
heat-treated Cu/Al/Cu at 500 for 10hours, the IDR width dropped to 54m after soaking for
100hours. The IDR depths of as-roll-bonded and heat-treated Cu/Al/Cu were observed to be 77m
and 12m, respectively, after soaking for 100hours.
Advanced Materials Research Vol. 1102 57

Fig. 3 The (a)thickness and (b)depth of Interfacial dissolution region

Fig. 4 SEM image and EDS spectra from regions marked by numbers 1, 2 and 3 of as-roll-bonded
Cu/Al/Cu clad soaked in 3.5%NaCl brine for 48hours.

Fig. 4 shows the SEM image and EDS spectra from regions marked by numbers 1, 2 and 3 of as-
roll-bonded Cu/Al/Cu clad soaked in 3.5%NaCl brine for 48hours. The region 1 with negligible
corrosion damage in this figure is copper. However, there is some indication of corrosion
deterioration in copper close to Cu/Al interface [9,10]. The region 2 shows the interfacial
dissolution region in less noble Al. There are numerous holes presumed to be caused by the fall-out
Si particles [1] during Al dissolution. Kim and Hong [1] observed numerous Si particles in Al side
of Cu/Al clad. The presence of these Si-containing particles could induce the micro-galvanic
reaction between Al matrix and Si [11] and small pits and Si fall-out can occur by dissolution of Al
at Al/Si interface. EDS spectra from region 2 shows the presence of mainly Al. EDS spectra from
region 3 shows presence of Al, O and Si. Strong aluminum and oxygen peaks suggest that the
corrosion products could be aluminum hydroxide Al(OH)3 or hydrated Al2O3. The close
examination of the corrosion products in region 3 revealed the presence cracked layers on the Al
surface and small particles. The presence of Si peak is attributed to the separation of Si particles
from Al matrix due to micro galvanic reaction.
58 Metal, Ceramic and Composite Materials

Conclusion
In this study, the influence of intermetallic compounds on galvanic coupling between Cu and Al in
Cu/Al/Cu clad was examined. The following conclusions were reached;
1. No visible intermetallic compounds were observed in the as-roll-bonded Cu/Al clad
plates. However, Intermetallic compounds grew up to 40m after heat treatment 500
for 10hours.
2. The galvanic corrosion occurred significantly in the Al region in contact with Cu. The
depth of the dissolved Al gradually decreased with distance away from the Cu/Al
interface. Both IDR width and IDR depth increased steadily with increase of soaking time.
3. The IDR width and depth of as-roll-bonded Cu/Al/Cu is far greater than those of heat-
treated Cu/Al/Cu at 500 for 10hours, suggesting the presence of intermetallic
compounds influences the galvanic corrosion behavior.
4. Strong aluminum and oxygen peaks suggest that the corrosion products could be
Aluminum Al(OH)3 or hydrated Al2O3. The presence of these Si-containing particles
could induce the micro-galvanic reaction between Al matrix and Si.

Acknowledgement
This work was supported by the 2nd phase of the Fundamental R&D Programs for Core Technology
of Materials funded by Ministry of Trade, Industry and Energy (2014-2015).

References
[1] I. K. Kim and S. I. Hong. Mater Des. 47 (2013) 590-598.
[2] I. K. Kim, Hong SI. Metall Mater Trans. A 44 (2013) 3890-3900.
[3] I. K. Kim and S. I. Hong. Mater Des. 57 (2014) 625-631.
[4] H. Kim, S. I. Hong, Mater Des. 67 (2015) 42-49.
[5] I. K. Kim, J. S. Ha and S. I. Hong. Korean J Met Mater. 50 (2012) 939-948.
[6] Y. C. Choi, H. S. Kim, S. I. Hong, Metals Mater Inter. 15 (2009) 733-739.
[7] S. I. Hong and M. A. Hill, Mater Sci Eng A. 264 (1999) 151-158.
[8] C. Y. Chen, H. L. Chen, W. S. Hwang. Mater Trans. 47 (2006) 1232-1239.
[9] J. B. Jorcin, C. Blanc, N. Pebere. J Electro Soc. 155 (2008) 46-51.
[10] K. Habib. Opt Lasers Eng. 31 (1999) 13-20.
[11] M. K. Chung, Y. S. Choi, J.G. Kim, J.C. Lee, Mater Sci Eng A. 366 (2004) 282-291.
Advanced Materials Research Vol 1102 (2015) pp 59-63 Submitted: 2015-01-15
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.59

Relation between the Betti number of fatigue fracture surfaces and


stress intensity factors of low carbon steel (JIS, S45C)
Masayuki Ishida1,a, Katsuyuki Kida1,b *, Koshiro Mizobe1,c
and Kazuaki Nakane2,d
1
University of Toyama, 3190, Gofuku, Toyama City, Toyama Prefecture, 930-8555, Japan
2
Osaka University, 1-7, Yamadaoka, Suita City, Osaka Prefecture, 565-0871, Japan
a
m1471205@ems.u-toyama.ac.jp, b kida@eng.u-toyama.ac.jp, c kmizobe@eng.u-toyama.ac.jp
d
knakane2008@gmail.com, * corresponding author

Keywords: Compact tension specimen, Fatigue fracture, Homology, Low carbon steel (S45C)

Abstract
The study of fracture surfaces of materials is very important to understand failures in engineering
elements, especially in structural products. However, it is not easy to evaluate fatigue fracture
surfaces by using ordinary image analysis techniques. In this paper, we propose a new analyzing
method to evaluate fatigue fracture surfaces from a view point of Homology.

Introduction
Fatigue failure of machine components is caused by cyclic load. Fatigue fracture surfaces record a lot
of information. However, quantitative evaluation of its fracture surfaces has not been enough carried
out. Nakane et al. [1-5] introduced a new method based on the homology theory to analyze the steel
structures. In our former research [5], we analyzed the steel structures, and concluded that two kinds
of the Betti numbers allowed us to evaluate the structures quantitatively. In the present work, fracture
surfaces were observed by an optical microscope. After the fatigue testing, Homology method is
applied to the micrographs in order to evaluate the fracture surfaces. Relation between the Betti
numbers and stress intensity factors is discussed.

Experimental procedure
The compact tension (CT) specimen dimensions were 48 mm (width of specimen measured from
center line of applied load, W) 12 mm (thickness, t) 57.6 mm (height). Figure 1 shows the
schematic illustration of the specimen. The material of the specimen was JIS S45C low carbon steel.
Table 1 shows the chemical compositions of the material. The specimen was heat treated at 1113K for
60 minutes and cooled in a furnace. The hardness was about 218 HV. The hardness was measured
using a standard Vickers hardness tester (AVK-CO, Akashi, Japan) under a 9.8 N weight during 15 s.
The CT specimen was tested using a hydraulic servosystem. Fatigue test were carried out at a
constant frequency of 5 Hz, under a stress ratio (R) of 0.1. The maximum load of 15.5 kN. The initial
value of the stress intensity factor range was 27.4 MPam1/2. The stress intensity factors of the cracks
were calculated using J. E. Srawley equations, basing on boundary collocation method. The stress
intensity factor K and its variation during fatigue tests were calculated using the following equation in
[6, 7]:
2+ 0.886 + 4.64 13.32 + 14.72 5.6
=
. 1
1

Here P = the applied load, t = thickness, W = width of specimen measured from center line of applied
load, = , = crack length from load line, and 0.2 1.

Table 1 Chemical composition of S45C: mass%.


C Si Mn P S Ni Cr Cu
0.46 0.18 0.77 0.011 0.018 0.2 0.18 0.01
60 Metal, Ceramic and Composite Materials

t W

Fig. 1 Dimensions of CT spencimen.

Experimental results and discussions


Observation results. Fatigue fracture surfaces were observed using an LCM (Laser confocal
microscope, Keyence VK-9700, Japan). Distances of the observation surfaces from the slit tip were 5
mm, 7 mm, 10 mm and 13 mm. Figure 2 shows example of fatigue fracture surfaces 5mm away from
the slit tip. Figures 2(a) and (b) are optical micrograph and height data. Three different areas for each
distance from the slit tip were observed. Figures 3, 4 and 5 are examples for 7 mm, 10 mm and 13 mm
away from the slit tip, respectively. (a) is optical photo and (b) is height data.

100m
(a) (b)
Fig. 2 Fatigue fracture surface 5 mm away from the slit tip: Crack growth
(a) Optical micrograph; (b) height data. direction

100m
(a) (b)
Fig. 3 Fatigue facture surface 7 mm away from the slit tip:
(a) Optical micrograph; (b) height data.
Advanced Materials Research Vol. 1102 61

100m
(a) (b)
Fig. 4 Fatigue facture surface 10 mm away from the slit tip:
(a) Optical micrograph; (b) height data.

100m
(a) (b)
Fig. 5 Fatigue facture surface 13 mm away from the slit tip:
(a) Optical micrograph; (b) height data.

Homology analysis
In this paper, we used the same method to analyze the fracture surface based on out former paper for
prior austenite grains in JIS SUJ2. [8] It is noted that the Betti number is topologically invariable, i.e.,
while the images of fracture surfaces have a variety of shapes, the Betti number is independent of their
shapes. This method was firstly introduced to detect lesion area of medical samples. [9]
To use ChomP analytical method to the images, we binarized height data with the mean value.
The results are shown in Figure 6. The binarized images allow creating simplicial complex from the
images. Then we can obtain the infomation of homology such as the Betti numbers. Table 2 shows the
results of the ChomP calulation. The values b0 and b1 represent the numbers of red areas and their
connection numbers, reaspectively.
From Table 2, it is found that the values b1 are strongly correlated to distance of the surfaces
from the slit tip. The Betti numbers, b0 and b1 decrease as distance from the slit tip. Figure 7 shows
changes in one of the Betti numbers, b1 during the fatigue tests. Comparing the all Betti numbers b1 of
all distances, it was found that the Betti number b1 decreased with stress intensity factor. This means
we can evaluate fatigue and crack growth from the view point of Homology.

Conclusion
In the present work, fracture surfaces were observed by an optical microscope and a laser confocal
microscope. After fatigue testing, Homology method is applied to the three dimensional micrographs
in order to evaluate the fracture surfaces. The Betti numbers, b0 and b1 decreases with distance from
the slit tip on the fracture surfaces, and especially the Betti number b1 has a strong colleration to stress
intensity factor. This means we can evaluate fatigue and crack growth from the view point of
Homology.
62 Metal, Ceramic and Composite Materials

(a) (b)

(c) (d)
Fig. 6 Binarized images:
(a) The position 5mm away from the slit tip (b0 = 52, b1 = 32); (b) the position 7mm away from the slit
tip (b0 = 39, b1 = 16); (c) the position 10mm away from the slit tip (b0 = 29, b1 = 18); (d) the
position 13mm away from the slit tip (b0 = 37, b1 = 13).

Table 2 Homology analysis results.


(a) Betti numbers, b0
Observation Distance from notch root [mm]
No. 5 7 10 13
1 52 32 27 25
2 50 39 19 37
3 55 50 29 36
average 52.3 40.3 25.0 32.7

(b) Betti numbers, b1


Observation Distance from notch root [mm]
No. 5 7 10 13
1 32 28 13 14
2 24 16 18 13
3 47 26 18 13
average 34.3 23.3 16.3 13.3
Advanced Materials Research Vol. 1102 63

50

Betti numbers(b1)

25

20 30 40 50 60
Stress intensity factor range [MPam]
Figure 7 Relation betwqeen stress intensity factor range and the Betti number, b1.

Acknowledgement
A part of this research was supported by the Grants-in-Aid for Scientific Research, JSPS (KAKENHI,
Scientific Research (B), Grant Number 26310209).

References
[1] K. Nakane, K. Mizobe, E. C. Santos and K. Kida, The Betti number of prior austenite structure of
repeated quenching bearing steels (JIS, SUJ2), Apple Mechanics and Materials, Vol. 307, pp.
409-414. (2013)
[2] K. Nakane, K. Mizobe, E. C. Santos and K. Kida, The Quantization of the structure of fisheyes via
homology method, Apple Mechanics and Materials, Vol. 372, pp. 273-276. (2013)
[3] K. Nakane, K. Kida, T. Honda, K. Mizobe and E. C. Santos, Influence of repeated quenching on
bearing steel martensitic structure investigated by homology, Apple Mechanics and Materials, Vol.
372, pp. 270-272. (2013)
[4]K. Nakane, K. Kida and K. Mizobe, Homoogy analysis of austenite grain size of SAE52100
bearing steel processed by cyclic heat treatment, Apple Mechanics and Materials, Vol. 813, pp.
116-119. (2013)
[5] K. Nakane, E. C. Santos, T. Honda, K. Mizobe and K. Kida, Homology analysis of structure of
high carbon bearing steel: effect of repeated quenching on prior austenite grain size, Materials
Research Innovations, Vol. 18 (Supplement 1), pp33-37. (2014)
[6] J. E. Srawley, Wide range stress intensity factor expressions for ASTM E 399 standard fracture
toughness specimens, Int Journ of Fracture, 12 (1976)
[7] J. C. Newman, Jr., Stress Analysis of the Compact Specimen Including the Effects of Pin Loading,
fracture Amalysis, ASTM STP 560, American Society for Testing and Materials, pp. 105-121. (2013)
[8] Katsuyuki Kida, Koshiro Mizobe, Ryosuke Araiand Kazuaki Nakane, The Betti number of prior
austenite grain of repeated quenching bearing steels (JIS, SUJ2), Advanced Materials Research, Vol.
1082, pp 191-196. (2015)
[9] K. Nakane, Y. Tsuchihashi and N. Matsuura, A simple mathematical model utilizing topological
invariants for automatic detection of tumor areas in digital tissue images, Diagn Pathol. Vol. 8 (Suppl
1), S27. (2013)
CHAPTER 2:

Ceramic Materials and Processing of Ceramics


Advanced Materials Research Vol 1102 (2015) pp 67-71 Submitted: 2014-12-29
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.67

High Thermal Stability of Mo/Si3N4/Mo/Si3N4/SiO2 Multilayer Solar


Selective Coatings

Ruihua Yang 1,a, Jinyang Liu 1,2,b*, Limei Lin1,c, Fachun Lai 1,2,d*, Yan Qu1,e
and Weifeng Zheng1,2,f
1
College of Physics and Energy, Fujian Normal University, Fuzhou 350108, China
2
Key Laboratory of Quantum Manipulation and New Energy Materials, Fuzhou 350108, China
a b* c d*
rhykenvy@qq.com, jyliu@finu.edu.cn, linlm@fjnu.edu.cn, laifc@fjnu.edu.cn,
e
quyan@fjnu.edu.cn, fzhengwf3@fjnu.edu.cn

Keywords: Thermal stability, Solar selective coating, Multilayer, High temperature.

Abstract. In terms of good optical properties and high thermal stability, Mo/Si3N4/Mo/Si3N4/SiO2
coatings based on metal/dielectric multilayer structure were adapted to the solar selective coating at
high operating temperatures. The coatings exhibited high solar absorptance in the range of 0.924 ~
0.936 and low thermal emittance of 0.114 ~ 0.118. The coatings deposited on quartz substrates were
thermally stable up to 625 C in air for 2 h, while they were degraded at 650 C from the
characterization of the absorptance and emittance. The degradation of the coatings was mainly due to
the oxidation of molybdenum in air, which was confirmed by Raman spectroscopy. Compared with
the thermal stability in air, the coatings were much more stable in vacuum under high temperature.
The remarkable thermal stability of the Mo/Si3N4/Mo/Si3N4/SiO2 coatings in air and in vacuum
makes them suitable to be applied at high temperature applications.

Introduction
Solar selective coating has a significant influence on solar-thermal conversion. To increase the
energy conversion efficiency, desirable solar selective coatings should have a lower reflection in the
solar spectrum region and a higher reflection in the infrared wavelength region [1]. Moreover, it
should be thermally stable at high temperatures (400C), since efficiency increases with operating
temperature in concentrated solar power (CSP) [1]. For this purpose, some solar selective coatings for
high temperature applications have been studied. Graded W-Al2O3 coating using sputtering displayed
a good thermal stability in vacuum up to 580 C [2]. Multilayer Mo/Al2O3 coating was thermally
stable in air up to 550 C [3]. However, high temperature annealing of solar selective coating may
induce diffusion, oxidation and reaction between layers. These modifications result in poor optical
properties. Therefore, it is of great interest to search the materials exhibiting good thermal stability
and optical properties in CSP applications [1].
Si3N4 is a promising material to be widely applied to high temperature components due to its
chemical inertness, low thermal expansion coefficient and good oxidation resistance [4]. But it is
rarely reported that Si3N4 is used as a candidate for the multilayer interference solar selective coatings.
Mo has high melting point, good high temperature strength and nitriding resistance [1,5]. Based on
these properties, the multilayer interference stack of Mo/Si3N4/Mo/Si3N4/SiO2 was adapted to the
solar selective coating in this paper. The additional of SiO2 layer was used to increase the efficiency
further due to antireflective effect. Thermal stability in air and in vacuum of the coatings was studied.

Experimental Details
The Mo/Si3N4/Mo/Si3N4/SiO2 solar absorber coatings were deposited on quartz substrates (25
25 0.5 mm3) using magnetron sputtering method in vacuum chamber, which was pumped down to
4.0 10-4 Pa before the deposition. The metal Mo films were deposited from direct current sputtering
68 Metal, Ceramic and Composite Materials

of Mo target (4N purity) in Ar plasma at 80 W, and the sputtering pressure was kept at 0.5 Pa. The
dielectric Si3N4 and SiO2 films were prepared from the reactive radio-frequency sputtering of Si target
(4N purity) in Ar + N2 plasma and Ar + O2 plasma with nitrogen flow rate of 7 sccm and oxygen flow
rate of 5 sccm, respectively. The sputtering pressure was kept at 0.3 Pa for the SiO2 and Si3N4 films,
and the flow rate of argon was 40 sccm for all layers.
The reflectance of the coatings was measured by a Perkin-Elmer Lambda 950 spectrophotometer
and a Perkin Elmer Spectrum One infrared spectrophotometer. The solar absorptance () and thermal
emittance () were calculated from the measured reflectance data in the wavelength range of 0.3 ~ 2.5
m and 2.5 ~ 25 m, respectively [6]. The cross-section morphology of the coating was observed by a
JSM-6700F scanning electron microscopy (SEM). In order to test the thermal stability, the coatings of
Mo/Si3N4/Mo/Si3N4/SiO2 were annealed in air at different temperatures for 2 h. The furnace
temperature was elevated at a heating rate of 3 C/min and kept at the anneal temperature for 2 h.
Subsequently, the furnace was cooled down to the room temperature at a rate of 3 C/min. The thermal
stability of the coatings in vacuum (8 10-4 Pa) was also studied.

Results and Discussion


Multilayer interference stack of Mo/Si3N4/Mo/Si3N4/SiO2 was adapted to the solar selective
coating. Fig. 1 shows the cross-sectional micrograph of the coating deposited on quartz substrate, and
the thicknesses of each layer from substrate to surface are about 200 nm, 60 nm, 10 nm, 50 nm and 60
nm, respectively. The optimized Mo/Si3N4/Mo/Si3N4/SiO2 coating exhibited high absorptance of
0.924 ~ 0.936 and low thermal emittance of 0.114 ~ 0.118. In order to test the thermal stability, the
Mo/Si3N4/Mo/Si3N4/SiO2 coatings were annealed at different temperatures (400 C ~ 650 C) in air
for 2 h. The corresponding absorptance and emittance values of the coatings are given in Table 1.
There is no significant change in the absorptance and emittance of the coating annealed at 400 C,
Even after being annealed at 625 C, slightly change in the absorptance ( = -0.012) and emittance

Fig.1 The cross-sectional SEM micrograph of the coating of Mo/Si3N4/Mo/Si3N4/SiO2 in the


as-deposited state.
Advanced Materials Research Vol. 1102 69

Fig.2 The reflectance spectra of the coatings of Mo/Si3N4/Mo/Si3N4/SiO2 in the as-deposited state,
annealed at 625 C and 650 C in air.
Table 1 Effect of 2 h annealing in air on the absorptance and emittance of the
multilayer coatings of Mo/Si3N4/Mo/Si3N4/SiO2.
annealing absorptance () emittance ()
temperature as-deposited annealed as-deposited annealed
400 C 0.931 0.934 0.003 0.118 0.122 0.004
500 C 0.932 0.937 0.005 0.116 0.125 0.009
600 C 0.936 0.931 -0.005 0.115 0.129 0.014
625 C 0.934 0.922 -0.012 0.115 0.134 0.019
650 C 0.936 0.873 -0.063 0.117 0.181 0.064
( = 0.019) was observed. However, the
absorptance decreased ( = -0.063) and the
emittance increased ( = 0.064) significantly
after being annealed up to 650 C. The reflectance
increased in the solar spectrum region and
decreased in the infrared wavelength region
resulting in the poor solar selectivity of the
coating annealed up to 650 C in air, as shown in
Fig. 2. This indicates that the
Mo/Si3N4/Mo/Si3N4/SiO2 coating has good
thermal stability in air up to 625 C. These results
were also confirmed by Raman spectroscopy
measurements on the annealed coatings.
The Raman spectra of the coatings of
Mo/Si3N4/Mo/Si3N4/SiO2 in the as-deposited
state and annealed up to 650 C in air for 2 h are
shown in Fig. 3. There was no change in the
nature of the Raman spectra even after being
annealed at 625 C. However, at 650 C, the
molybdenum oxides are appeared. In detail, the
bands centered at 203 228 353 364 495 and 573
cm-1 correspond to MoO2 [7], the bands
centered at 393, 818, 848, 958 and 995 cm-1 Fig.3 The Raman spectra of the coatings of
correspond to MoO3 [7,8], and the band Mo/Si3N4/Mo/Si3N4/SiO2 in the as-deposited
centered at 973 cm-1 corresponds to Mo5O14 [9]. state and annealed at different temperatures in
These molybdenum oxide phases indicated that air.
the degradation of the multilayer coating of Mo/Si3N4/Mo/Si3N4/SiO2 was mainly due to the
oxidation of molybdenum in air.
70 Metal, Ceramic and Composite Materials

In order to study whether the multilayer coatings of Mo/Si3N4/Mo/Si3N4/SiO2 were stable at


higher annealing temperatures, they were annealed in vacuum at temperatures in the range of 400 C ~
800 C for 2 h. The corresponding absorptance and emittance values of the coatings after being
annealed are given in Table 2. Compared with the thermal stability in air, the coatings were much
more stable in vacuum under high temperature. It is evident from the table that there is no significant
change in the absorptance ( = 0.009) but a slight increase in the emittance ( = 0.024) as a result of
air annealing at 800 C. This shows that the Mo/Si3N4/Mo/Si3N4/SiO2 coating was thermally stable up
to 800 C in vacuum. The high thermal stability of the coatings is mainly attributed to high melting
point and good nitriding resistance of molybdenum, diffusion barrier of Si3N4 for molybdenum and
stable microstructures.
Table 2 Effect of 2 h annealing in vacuum on the absorptance and emittance of the
multilayer coatings of Mo/Si3N4/Mo/Si3N4/SiO2.
annealing absorptance () emittance ()
temperature as-deposited annealed as-deposited annealed
400 C 0.928 0.931 0.003 0.117 0.121 0.004
500 C 0.924 0.930 0.006 0.119 0.124 0.005
600 C 0.924 0.928 0.004 0.117 0.125 0.008
650 C 0.925 0.933 0.008 0.118 0.127 0.009
700 C 0.924 0.931 0.007 0.116 0.131 0.015
750 C 0.928 0.934 0.006 0.116 0.133 0.017
800 C 0.927 0.936 0.009 0.114 0.138 0.024

Summary
Multilayer solar selective coatings of Mo/Si3N4/Mo/Si3N4/SiO2 were deposited on quartz
substrates using magnetron sputtering system. The coatings exhibited high solar absorptance ( =
0.924 ~ 0.936) and low thermal emittance ( = 0.114 ~ 0.118). The thermal stability study showed that
the multilayer Mo/Si3N4/Mo/Si3N4/SiO2 coating was thermally stable up to 625 C in air. At higher
temperature, the optical property was degraded, which was attributed to the formation of
molybdenum oxide phases in air. The multilayer coating was also thermally stable up to 800 C in
vacuum. This indicates that the Mo/Si3N4/Mo/Si3N4/SiO2 coating is useful for high temperature solar
selective applications.

Acknowledgments
This work was financially supported by the Natural Science Foundation of China (No. 11074041,
11374052), the Natural Science Foundation of Fujian Province of China (2012J01256, 2013J01174),
and the Science and Technology Project from Education Department of Fujian Province of China
(JB13023).

References
[1] C.E. Kennedy, Progress to develop an advanced solar-selective coating, NREL/CD-550-52709,
National Renewable Energy Laboratory, Florida, 2008.
[2] A. Antonaia, A. Castaldo, M.L. Addonizio, S. Esposito, Stability of W-Al2O3 cermet based solar
coating for receiver tube operating at high temperature, Sol. Energ. Mat. Sol. C. 94 (2010) 1604-1611.
[3] J.A. Thornton, A.S. Penfold, L.J. Lamb, Sputter-deposited Al2O3/Mo/Al2O3 selective absorber
coatings, Thin Solid Films, 72 (1980) 101-110.
Advanced Materials Research Vol. 1102 71

[4] H. Klemm, Silicon nitride for high-temperature applications, J. Am. Ceram. Soc. 93 (2010)
1501-1522.
[5] Q.-C. Zhang, Recent progress in high-temperature solar selective coatings, Sol. Energ. Mat. Sol. C.
62 (2000) 63-74.
[6] Y. Wu, W. Zheng, L. Lin, Y. Qu, F. Lai, Colored solar selective absorbing coatings with metal Ti
and dielectric AlN multilayer structure, Sol. Energ. Mat. Sol. C. 115 (2013) 145-150.
[7] M.A. Camacho-Lpez, L. Escobar-Alarcn, M. Picquart, R. Arroyo, G. Crdoba, E.
Haro-Poniatowski, Micro-Raman study of the m-MoO2 to -MoO3 transformation induced by
cw-laser irradiation, Opt. Mater. 33 (2011) 480-484.
[8] K. Ajito, L.A. Nagahara, D.A. Tryk, K. Hashimoto, A. Fujishima, Study of the photochromic
properties of amorphous MoO3 films using Raman microscopy, J. Phy. Chem. 99 (1995)
16383-16388.
[9] G. Mestl, MoVW mixed metal oxides catalysts for acrylic acid production:from industrial
catalysts to model studies, Top. Catal. 38 (2006) 69-82.
Advanced Materials Research Vol 1102 (2015) pp 73-76 Submitted: 2015-01-07
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.73

Validation of a Novel Approach for Co2/N2 Gas Separations by means of


a Hybrid Ceramic Membrane

Ngozi Claribelle Nwogu, Mohammed Nasir kajama & Edward. Gobina*

Centre for Process Integration and Membrane Technology, School of Engineering,


The Robert Gordon University, Riverside East Garthdee Road, Aberdeen AB10 7GJ. United
Kingdom
*
Corresponding Author: Tel: +441224262348: email: e.gobina@rgu.ac.uk

Keywords: Global warming, carbon dioxide, inorganic ceramic membrane, gas separation &
separation factor

Abstract: Global warming has been documented as one of the worlds foremost ecological
concerns. Although it is difficult to totally end human related global warming, there is a possibility
to alleviate these effects through a wide spectrum of options. One such possibility is the reduction
of atmospheric carbon dioxide emissions, a major greenhouse gas widely thought to be responsible
for global warming. This paper therefore looks at an experimental validation of gas separation by
means of a high selective membrane for CO2 recovery applications. Analysis of the results obtained
is in good agreement with literature experimental data. Additional results show that CO2 selectivity
factor is reasonable for capture of CO2 from N2 as a key constituent in a flue gas stream.

Introduction

While capture of carbon dioxide with amine solvents is the most established technology, another
potential and efficient candidate is gas separation membranes. This technology has been improved
further by recent developments in materials process engineering resulting in the innovation of novel
materials with the thermal and mechanical stability required for most gas separations. In particular,
inorganic gas separation membranes are examined. Membrane technology is a viable option to
conventional separation methods since it offers cheaper capital investment and is relatively less
energy consuming. As a result, there is high demand for this technology in environmentally
demanding processes leading to an upshot in membrane technological market. To achieve
optimization in membrane separation systems performance in an economical manner, the
development of a reliable way of dealing with the design of membrane technology is gaining great
attention [1]. An accurate description of the process behaviour in the membrane separation process
is an important factor in order to reduce some technical risks which could be encountered especially
in relation to traditional separation technique. In addition, designing of process models for
membrane gas separation in particular is vital and requires an extremely thorough and careful
approach. Thus an accurate and dependable simulation of the entire system can be employed for the
design of the separation process. Subsequently, efforts made towards development of a detailed
model for membrane gas separation has not gained enough grounds at the moment and not very
readily available in most published literatures, although a limited number of unit models exists in
some literatures [2] [3] [4].

Mathematical Model

Figure 1 shows binary gas permeation through a hybrid ceramic membrane operating in a one
directional flow (concurrent) and cross flow mode. It is also necessary to observe that the
concentration of the permeate exiting the membrane coated surface, YI is not the same as the bulk
permeate leaving the porous layer, Yi. However both flows are equal at the end of the membrane
[5].
74 Metal, Ceramic and Composite Materials

The model is based on a binary gas mixture whereby the feed gas is on the coated side of the
asymmetric membrane. Combination of permeate fluxes of different components does not take
place inside the porous supporting layer of the membrane. The resistance of gas flow (feed) of the
support layer as well as diffusion of flow through the membrane pore path is negligible as a result
of the high level permeate influx. Membrane distortion under pressure is also negligible. All
operating conditions in the simulation were used for experimental validation. The main objective of
the model is to ascertain and validate the degree of CO2 recovery occurrence using a single stage
membrane simulation model.

Fig. 1: Binary gas permeation through a Nanostructured Ceramic membrane

Experimental methods.

The commercial ceramic membrane used was purchased from Ceramiques Techniques et
Industrielles (CTI SA), France and made up of 77% -alumina + 23% TiO2 with an average pore
diameter of 6000nm. The membrane has 20 mm and 25 mm internal and outer diameter
respectively, and an effective penetrable length of 318 mm. The feed pressure applied was between
1 to 5 bar. Gas permeation experiments were carried out to investigate and test CO2 and N2 gas
mixtures permeation behaviours as well as separation performance process under relevant operating
conditions. Experimental values were used as input data in the membrane simulation. The
selectivity of the membrane was assumed so as to enable comparison of experimental and model
parameters.
For a single stage membrane process, applicable parameters used in this work including operational
conditions are pressure, temperature and gas flow rates. Others are selectivity, permeability,
membrane surface areas and inlet feed composition. Other variables that contribute to the
membrane performance are the membrane selectivity also called separation factor usually
predicted and calculated from the permeability ratio of pure gases [6]. An estimation of CO2
selectivity to N2 can be calculated from the ratio of CO2 permeance to that of N2 using the formula
below:


Selectivity, =

Results and Discussions


Experimental values were obtained using ceramic membranes module parameters at simulation
operating conditions as well as its selectivity while that of literature was obtained from results using
polymeric membrane for CO2 removal from flue gas [7].
Advanced Materials Research Vol. 1102 75

Tables 1- 3 show the simulation, experimental and literature values respectively. Figure 2 depicts a
comparative analysis of CO2/N2 selectivity to CO2 removal based on experimental and literature
values. In this case, all model parameters were kept constant with varying polymeric membrane
selectivity inputs. For both scenarios, at a relatively low selectivity, the trend shows that more CO2
was recovered with our experimental input compared to that of literature values having higher
selectivity but lower permeability properties being characteristic features of organic membrane
while the ceramic membrane exhibited high permeability and enhanced CO2 recovery characteristic.

Table 1: Model Values


Simulation benchmark

CO2/N2 Selectivity CO2 Percentage Recovery


10 35.5

Table 2: Experimental Values


Experimental Values
CO2/N2 selectivity CO2 Percentage Recovery
0.75 4.66
1.52 11.13
2.43 18.02
3.48 25.31
3.83 27.96

Table 3: Literature Values


Literature Values
CO2/N2 selectivity CO2 Percentage Recovery
19.3 0.7615
21 1.7993
24.2 1.2361
17.1 0.9882
29.1 1.5884
76 Metal, Ceramic and Composite Materials

Comparative Analysis of CO2/N2 Selectivity , CO2


recovery Relation: An Experimental Validation
CO2 Percentage Recovery
40
35
30 Literature [ Ref.
25 53]
(%)

20
Experimental
15 [This work]
10
5 Model[Reference
case]
0
0 10 20 30 40
CO2/N2 Selectivity

Fig. 2: Comparative analysis of CO2/N2 selectivity and CO2 removal: Experimental Validation

Conclusion

A hybrid inorganic ceramic membrane has been prepared by immersion in a silica solution forming
an asymmetric structure for CO2/N2 gas separations. The experimental results obtained from the
permeation study indicate a good agreement between the experimental, literature and model. This
therefore validates the experimental and literature results by the model which ultimately can now be
used for a range of parametric studies for fuel gas separation applications for CO2 recovery.

Nomenclature

FF = Flow rate of gas component in the feed side (l/min)


FR = Flow rate of gas component in the retentate side (l/min)
FP = Flow rate of gas component in the permeate side (l/min)
XF = Mole fraction of gas component in the feed side (%)
XR = Mole fraction of gas component in the retentate side (%)
XP = Mole fraction of gas component in the permeate side (%)

References
[1] ElHalwagi MM. Synthesis of reverseosmosis networks for waste reduction. AIChE Journal.
1992; 38(8):1185-1198.
[2] Krovvidi K, Kovvali A, Vemury S, Khan A. Approximate solutions for gas permeators
separating binary mixtures. Journal of Membrane Science. 1992; 66(2):103-118.
[3] Qi R, Henson MA. Modeling of spiral-wound permeators for multicomponent gas separations.
Industrial & Engineering Chemistry Research. 1997; 36(6):2320-2331.
[4] Coker D, Freeman B, Fleming G. Modeling multicomponent gas separation using hollowfiber
membrane contactors. AIChE Journal. 1998; 44(6):1289-1302.
[5] Pan C. Gas separation by permeators with highflux asymmetric membranes. AIChE Journal.
1983; 29(4):545-552.
[6] Bounaceur R, Lape N, Roizard D, Vallieres C, Favre E. Membrane processes for post-
combustion carbon dioxide capture: a parametric study. Energy. 2006; 31(14):2556-2570.
[7] Liu L, Chakma A, Feng X. Preparation of hollow fiber poly (ether block amide)/polysulfone
composite membranes for separation of carbon dioxide from nitrogen. Chemical Engineering
Journal. 2004; 105(1):43-51.
CHAPTER 3:

Composites, Advanced Materials and Nano


Materials
Advanced Materials Research Vol 1102 (2015) pp 79-82 Submitted: 2014-12-25
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.79

Baddeleyite Type Monoclinic Zirconium Oxide Nanocrystals Formation

R. Yuvakkumar1,a and Sun Ig Hong1,b*


1
Department of Nanomaterials Engineering, Chungnam National University, Daejeon, 305-764,
South Korea
a b
yuvakkumar@gmail.com, sihong@cnu.ac.kr
*Corresponding author

Keywords: Baddeleyite, monoclinic, ZrO2, nanocrystals

Abstract. We report successful synthesis of baddeleyite type monoclinic zirconium oxide


nanocrystals formation. The product mixture of zirconium incubated at room temperature for 7 days
were thoroughly investigated employing X-ray diffraction, Raman spectroscopy and transmission
electron microscopy studies. XRD and Raman studies revealed the formation of baddeleyite type
monoclinic zirconium oxide nanocrystals. TEM studies revealed the nanocrystal formation with size
ranging from 100 nm to 200 nm.

Introduction
Zirconium oxide nanostructures have been widely studied in the areas of anti-ferroelectric [1],
quenching [2], optical, electrical and mechanical [3], photoluminescence [4], adsorption [5] and
micro filtration membrane applications [6]. Recently, Jiang and his co-workers studied preparation
and photocatalytic performance of zirconium oxide nanotubes fabricated with anodization process
[7]. Pradhan and his co-workers investigated the effect of zirconium oxide nanopowder on the
thermal, chemical and gas barrier properties of starch [8]. Selvaraj and his co-workers studied
selective atomic layer deposition of zirconia on copper patterned silicon substrates using ethanol as
oxygen source as well as copper reductant [9]. Liu and his co-workers investigated preparation of
nano t-zirconium oxide particle by the integrated process of high gravity field and hydrothermal
crystallization [10]. However, green synthesis of nanostructures with sustainable materials
development has been inevitable in emerging societal issues [11-14]. Therefore, in the present study
green synthesis of baddeleyite type monoclinic zirconium oxide nanocrystals formation has been
explored.

Materials and methods


Zirconium oxide nanocrystals were prepared through a simple and eco-friendly green
synthesis method. 0.1 M of 50 mL zirconium (IV) nitrate [ZrO(NO3)2xH2O] and 10 mL of rambutan
extract were mixed drop wise under magnetic stirring at 80C for 2 h and then incubated at room
temperature for 7 day to form zirconium-ellagate complex formation. Zirconium-ellagate complex
formed after adequate time of incubation was washed, filtered and then dried prior to analysis. White
zirconium oxide crystals was obtained after washing the product alternately and repeatedly with water
and absolute acetone. Zirconium oxide nanocrystals was obtained on direct decomposition of the
zirconium-ellagate complexes in muffle furnace at 450C for 1 hour in static air atmosphere. Phase
and crystalline nature were identified by X-ray powder diffraction patterns (Bruker AXS, Germany)
employing CuK as radiation (1.5406 A) source. High resolution (Dispersive Raman microscope,
Horiba Jobin Yvon; LabRAM HR UV/vis/NIR) Raman spectroscopy (excitation, 514 nm) was used
to confirm zirconium oxide nanocrystals. Transmission electron microscopic images were obtained
using a transmission electron microscope (CM 200; Philips, USA).
80 Metal, Ceramic and Composite Materials

Results and discussion


The present study reported a detailed structural and lattice vibration studies employing XRD
and Raman techniques along with microstructure observation using transmission electron
microscopy. In general, several reports were already performed employing XRD and Raman
techniques to describe the structure and properties of zirconium oxide. However, in the present study,
X-ray diffraction in range 1080 and Raman spectra over a frequency range of 80800 cm-1 were
measured systematically for obtained zirconium oxide nanocrystals of the size ranging from 100 nm
to 200 nm at room temperature. 7 days of reaction product from the mixture of 0.1M
ZrO(NO3)2.xH2O and 10 ml extract was calcinated in muffle furnace at 450C for 1 hour in static air
atmosphere. The obtained product structure and the approximate crystallite size were conventionally
determined employing X-ray diffraction studies. In general, zirconium oxide (ZrO2) is a useful
functional material in high temperature optical and electronic applications. Moreover, zirconium
oxide is also a typical material in structural phase transitions because of the existence of several
zirconium oxide polymorphs. Fig. 1a shows the XRD pattern of the obtained product. The XRD
results clearly revealed that the obtained product contained baddeleyite type single phase monoclinic
zirconium oxide (JCPDF card file 01-070-2491) (Fig. 1a) with lattice constants a=5.14890,
b=5.21180, c=5.31470, =90, =99 and =90. Baddeleyite type is a rare zirconium oxide mineral
(ZrO2 or zirconia) occurring in a variety of monoclinic prismatic crystal forms. The obtained XRD
result clearly revealed mono phase monoclinic structure of the obtained product. Baddeleyite has the
group symmetry P21/c with four zirconium oxide in the unit cell. The average crystallite size of the
obtained zirconium oxide nanocrystals was calculated using the X-ray diffraction pattern employing
the Debye Scherer formula D = k/(cos), where is the wavelength of the X-ray radiation, k is a
constant taken as 0.898, is the diffraction angle and is the full width at half maximum [15-19]. The
calculated average crystallite size was approximately 62 nm which is similar to the crystal size
observed from TEM studies. The obtained XRD data clearly revealed that the product obtained
employing green synthesis method is effective for the synthesis of baddeleyite type monoclinic
zirconium oxide nanocrystals.
(-111)

(C) (a)
(111)
Intensity (a.u)

(020)
(110)

(220)
(022)
(-211)
(200)

(202)
(-202)

(311)
(311)
(001)

(310)
(021)

(222)
(041)

10 20 30 40 50 60 70 80
2 (deg)
Advanced Materials Research Vol. 1102 81

Ag (b)
Ag Ag
Raman Intesity (a.u)

Bg
Ag AgBg 2Bg 2Ag
Bg A Bg Ag
g

Bg

100 200 300 400 500 -1 600 700 800


Raman Shift (cm )

Figure 1. (a) XRD, (b) Raman (c) TEM studies of baddeleyite type monoclinic ZrO2 nanocrystals

The obtained product vibrational properties were studied employing Raman spectroscopy.
According to the group theory, out of 36 zirconium oxide lattice vibrational modes, Raman active
modes contains 9Ag and 9Bg and infrared active mode contains 8Au and 7Bu modes and remaining
others contains acoustic modes. Monoclinic zirconium oxide is belonging to the space group C52h
(P21/c) with four unit cell dimensions. A similar result was observed in the obtained XRD results.
The expected 36 lattice vibrational modes are Gmono=9Ag+9Au+9Bg+9Bu. Owing to the above group
theory, in the present study, the obtained fourteen Raman peaks such as 104 (Ag), 180 (Ag), 192 (Ag),
222 (Bg), 308 (Ag), 335 (Bg), 349 (Ag), 381 (Bg), 475 (Ag), 501 (Bg), 537 (Bg), 559 (Ag), 615 (2Bg) and
639 cm-1 (2Ag) were exactly correlated to the baddeleyite type monoclinic zirconium oxide
nanocrystals as shown in Fig. 1b. Raman results revealed no other vibrational mode at 148 cm-1 or 266
cm-1 which suggested the absence of tetragonal zirconium oxide phase in the obtained product. The
obtained Raman results further revealed mono phase monoclinic structure of the obtained product. It
is noteworthy that the obtained Raman spectra of zirconium oxide nanocrystal in the region 100800
cm-1 exactly matches with space group C with respect to the mode position, linewidth and intensity.
The obtained Raman result clearly revealed that the product at 7 days of incubation were consisted of
single monoclinic zirconium oxide phase structure which were in good agreement with the obtained
XRD results.

A transmission electron microscope study was further employed to confirm the size and
morphology of 7 days incubated product. Fig. 1c shows representative transmission electron
microscopy (TEM) image with the scale of 200 nm. TEM images clearly revealed the agglomerated
nanocrystal formation with size ranging from 100 nm to 200 nm (Fig. 1c). In the present study, we
systematically measured XRD, Raman and TEM studies of 7 days incubated product mixture
calcinated at 450C to confirm the structure and phase formation of zirconium oxide nanocrystals.
82 Metal, Ceramic and Composite Materials

Both XRD and Raman, and TEM studies revealed the single phase structure and vibration
characteristics of baddeleyite type monoclinic zirconium oxide nanocrystals formation.

Summary
We report successful synthesis of baddeleyite type monoclinic zirconium oxide nanocrystals
formation. Combined XRD, Raman and TEM studies results correlated the formation of baddeleyite
type monoclinic zirconium oxide nanocrystals.

Acknowledgment
This work was supported by a grant from the Fundamental R & D Program for Industrial
Technology (2014), funded through the Ministry of trade, Industry and Energy and by a grant from the
Basic Atomic Energy Research Institute Program, through the National Research Foundation of
Korea (NRF) (2014).

References
[1] S.E. Reyes-Lillo, K.F. Garrity, K.M. Rabe, Phys. Rev. B: Condens. Matter 90 (2014) 140103.
[2] H. Torabmostaedi, T. Zhang, J. Therm. Spray Technol. 20 (2014) 15p.
[3] V.S. Dang, M. Banerjee, H. Zhu, N.B. Srinivasan, H. Parala, J. Pfetzing-Micklich, A.D.
Wieck, A. Devi, Chem. Vap. Deposition (2014) DOI: 10.1002/cvde.201407124.
[4] Q. Chen, Y. Chang, C. Shao, J. Zhang, J. Chen, M. Wang, Y. Long, J. Mater. Sci. Technol. 66
(2014) 74-80.
[5] C. Wang, Y. Le, B. Cheng, Ceram. Int. 40 (2014) 10847-10856.
[6] Q. Zhang, R. Xu, P. Xu, R. Chen, Q. He, J. Zhong, X. Gu, Desalin. 256 (2014) 71-74.
[7] W. Jiang, J. He, J. Zhong, J. Lu, S. Yuan, B. Liang, Appl. Surf. Sci. 307 (2014) 407413.
[8] G.C. Pradhan, S. Dash, S.K. Swain, Mater. Sci. Semicond. Process. 23 (2014) 115121.
[9] S.K. Selvaraj, J. Parulekar, C.G. Takoudis, J. Vac. Sci. Technol., A 32 (2014) 010601.
[10] S.W. Liu, J.C. Zhou, R.X. Liu, Adv. Mat. Res. 881-883 (2014) 933-939.
[11] R. Yuvakkumar, J. Suresh, A. Joseph Nathanael, M. Sundrarajan, S.I. Hong, Mater. Sci. Eng.,
C 41 (2014) 17-27.
[12] R. Yuvakkumar, J. Suresh, A. Joseph Nathanael, M. Sundrarajan, S.I. Hong, Mat. Lett. 128
(2014) 170-174.
[13] R. Yuvakkumar, J. Suresh, B. Saravanakumar, A. Joseph Nathanael, S.I. Hong, V. Rajendran,
Spectrochim. Acta Part A 137 (2015) 250-258.
[14] R. Yuvakkumar, A. Joseph Nathanael, S.I. Hong, RSC Adv. 4 (2014) 44495-44499.
[15] R. Yuvakkumar, V. Elango, V. Rajendran, N. Kannan, Synth. React. Inorg. Met.-Org. Chem.
41 (2011) 309-314.
[16] R. Yuvakkumar, V. Elango, V. Rajendran, N. Kannan, J. Exp. Nanosci. 9 (2014) 272-281.
[17] R. Yuvakkumar, V. Elango, V. Rajendran, N. Kannan, Digest J. Nanomaterials and
Biostructures 6 (2011) 1771-1776.
[18] R. Yuvakkumar, A. Joseph Nathanael, V. Rajendran, S.I. Hong, J. Sol-Gel Sci. Technol. 72
(2014) 198205. 250-258.
[19] R. Yuvakkumar, P. Peranantham, A. Joseph Nathanael, D. Nataraj, D. Mangalaraj, S.I. Hong,
J. Nanosci. Nanotechnol. 15 (2015) 2523-2530.
Advanced Materials Research Vol 1102 (2015) pp 83-86 Submitted: 2014-12-30
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.83

Effects of selenization conditions on the properties of Cu-In-Ga-Se thin


films prepared by Pulsed laser deposition

Lei Tang 1, a, Shuiyuan Chen 1, b,*, Huiqin Zhang 1, c and Zhigao Huang 1, d,*
1
College of Physics and Energy, Fujian Normal University, Fuzhou, 350007, China
a b c d
602173240@qq.com, sychen@fjnu.edu.cn, 707606108@qq.com, zghuang@fjnu.edu.cn
* Corresponding Author

Keywords: Cu-In-Ga-Se films, Pulsed laser deposition, Selenization annealing, Optical properties

Abstract. In this paper, Pulsed laser deposition (PLD) is attempted to be an alternative synthesis
method for Cu-Ga-In (CIGS) thin films. Cu-In-Ga prefabricated metal films were grown on the quartz
substrates by PLD method. Then the CIGS films with different Se content were synthesized by the
post-selenization(Se)-annealing processes. The structure, element component proportion and optical
properties of the CIGS films were investigated. The experimental results indicate that the CIGS films
can be well obtained by the PLD method with Se-annealing. The performances of the CIGS films are
much affect by the Se-annealing processes, including Se-annealing temperature, annealing time, and
the way of Se-annealing. The CIGS films prepared by PLD method with Se-annealing present fine
visible light absorption properties.

Introduction
Polycrystalline thin-film chalcopyrite Cu(In,Ga)Se2(CIGS) is currently used as an absorber layer
for high efficiency photovoltaic (PV) solar cells. CIGS based chalcopyrite compounds are attracting
considerable interests as light absorbing materials for thin film solar cells for their inherent
advantages of high absorption coefficient, adjustable bandgap, strong ability to resist radiation, and
high stability under high energy irradiation [1-2]. CIGS absorber is the core part of the whole solar
cells. Therefore, it is very important to obtain optimum processes of CIGS thin film. The main
preparation methods for thin film CIGS solar cells, such as co-evaporation, sputtering, and chemical
vapor deposition, have been well investigated and obtained much development [3-7].
As we know, the element content proportion much affects the light absorptive properties of the
CIGS solar cell. Therefore, proportion control is important in the synthesis progress of the CIGS thin
film. Pulsed Laser Deposition (PLD) method is currently widely used in composite thin film
preparation Owing to the simple operation and lots advantages. An advantage of the method is that the
element proportion of the film is much the same as that of the target. Compared with other coating
methods, its preparation of parameters is flexibly adjustable and it can make high resolution and
purity film in a high vacuum environment [8-11]. In this paper, Pulsed Laser Deposition, an
alternative synthesis method for CIGS thin film, is attempted.
Here, we deposit the Cu-In-Ga metal prefabricated layer by PLD method, and then synthesis CIGS
thin film by the post-selenization(Se)-annealing processes. The effects of selenization conditions on
the properties of CIGS thin films are investigated.

Experimental details
Choosing the high purity In target (99.9 %) and CuGa alloy target (99.9 %, the Ga account for 30 %
of the total mass), we first deposit the In on the quartz substrate and then sputter the CuGa alloy target
to complete the preparation of the In-Cu-Ga prefabricated metal film by adjusting the laser pulse
counts and sputtering energy (In and CuGa layer are deposited at 250 and 330 mJ/pulse energy,
respectively) in 10-4 Pa. After that, the In-Cu-Ga prefabricated metal film is selenized annealed in the
selenization furnace. The structure, component and optical properties of the CIGS film were analyzed
84 Metal, Ceramic and Composite Materials

by MiniFlex-II X-ray diffractometer (XRD), Scanning Electron Microscopy(SEM), Energy


Dispersive Spectrometer(EDS), and Perkin Elmer Lambda 950.

Results and Discussion


We explored the best preparation conditions to synthesis high-quality CIGS thin film. In the
preparation process, Se-annealing condition is vital. We perform different synthesis process to
investigate the effect of Se-annealing condition on the properties of the films. The CIG prefabricated
metal films are annealed in selenium atmosphere for 60 minutes at 250 C. Then the above thin films
are divided into two groups for next process. The first group thin films are further annealed at
different temperature (400 C, 450C and 500 C,) without selenium atmosphere for 30 minutes (the
samples are named as A1, B1, C1, respectively). The second group thin films are further annealed in
selenium atmosphere at 450 C with different annealing time (30, 40 and 50 minutes, the samples are
named as A2, B2, C2, respectively).
Fig. 1(a) shows the XRD patterns of the first group samples (A1, B1, C1). From the figure, several
diffraction peaks along (112), (211), (220/204), (116/312) directions are observed, which were well
matched for CIGS phase indexed from standard PDF. The results indicated that the preparation
method described above (PLD + Se-annealed) can obtain fine CIGS thin film. Besides, some impure
phases are dispersed in A1 sample annealed at 500 C, indicating the excessive anneal temperature.
Fig. 1(b) presents the XRD patterns of the second group samples (A2, B2, C2). Compared to the first
group samples, we can notice that the second group thin films possess better crystallization, indicating
that Se-annealing is important to obtain high-quality CIGS film. Furthermore, C2 sample possesses
the best structure property, which presents that more Se-annealing time is beneficial for obtaining fine
CIGS film.

Fig. 1 XRD patterns of CIGS film at different annealing processes.


According to literature, fine CIGS solar cell requires suitable proportion, i.e., the atomic ratio of
Cu/(In+Ga) may be from 0.69 to 0.98, Ga/(In+Ga) from 0.21 to 0.38, Se/(Cu+In+Ga) from 0.95 to
1.10. We perform EDS measurement for further understanding the element composition of as
prepared CIGS films. Fig. 2 presents the typical EDS result for sample B2 and C2, Except for Cu, In,
Ga and Se element, no other elements can be indexed in the EDS results, which is in accord with the
XRD results. Further analysis shows that the contents of Cu, In, Ga and Se elements for CIGS
absorption layers locate in the atomic ratios mentioned above, indicates that the samples possess the
basic conditions to be solar cell candidates.
Advanced Materials Research Vol. 1102 85

Fig. 2 EDS analysis of the sample B2 and C2.


Fig. 3 exhibits the SEM images of the sample C1and C2. From the results of the SEM we can see
that the two samples present compact structure and fine crystalline. The average grain size of C1
sample is larger than that of C2 sample, for that the higher annealing temperature is beneficial for the
crystallization of the CIGS films, and therefore reduces grain boundary trap states, which may
improve the open-circuit voltage (Voc) of the CIGS thin film solar cells.

(a) (b)

Fig. 3 SEM morphologies of (a) sample C1 and (b) sample C2.


Fig. 4 (a) shows the ultraviolet-visible absorption spectra for the second group thin films (A2, B2
and C2). Due to the optical interference by multiple reflections in the CIGS absorption layer, the three
samples possess higher level optical absorption between 350 nm and 700 nm, which almost locates in
the visible light region. The absorbance properties of A2 sample is the best among the three samples.
Fig. 4(b) presents the optical absorption coefficient () vs. photon energy (h) of the A2, B2 and C2
samples. The three samples show different photon energy absorption behavior, indicating different
band gap energy (BGE) of the three CIGS thin films. These results indicates that the CIGS thin films
prepared by PLD method and Se-annealing display fine optical properties, and are suitable for
applications in thin film solar cells.

Fig. 4 Optical properties of the A2, B2, C2 samples.


(a) Absorption spectra. (b) The photon energy dependence of the absorption coefficient.
86 Metal, Ceramic and Composite Materials

Summary
In summary, the CIGS absorption layer thin films have been successfully synthesized by PLD
method and Se-annealing. The structure, component and optical properties of the CIGS films have
been measured. The experimental results indicate that the CIGS thin films show fine visible
absorption properties, indicating that the preparation method is suitable for synthesizing CIG thin film
for solar cell.

Acknowledgements
This work is supported by the National Natural Science Foundation of China (11004039,
11004031) and Natural Science Foundation and Education Office Foundation of Fujian Province
(2012J01003, JA12054)

References
[1] M.Y. Chiang, S.H. Chang, C.Y. Chen, F.W. Yuan, H.Y. Tuan, Quanternary CuIn(S1-xSex)2
nanocrystals: Facile heating-up synthesis band gap tuning, and gram-scale production,
J.Phys.Chem.C 115 (2011) 1592-1599.
[2] J.F. Guillemoles, L. Kronik, D. Cahen, U. Rau, A. Jasenek, H.W. Schock, Stability issue of
Cu(In,Ga)Se2-based solar cells, J. Phys. Chem. B 104 (2000) 4849-62.
[3] M. Kaelin, D. Rudmann, F. Kurdesau. CIS and CIGS layers from selenized nanoparticle
precursors, Thin Solid Films 58 (2003) 431-432.
[4] R. Bock, J. Schmidt, R. Brendel, Effective passivation of highly aluminum doped p type silicon
surfaces using amorphous silicon, Appl. Phys. Lett. 91 (2007) 248.
[5] D.B. Mitzi, M. Yuan, W. Liu, A high efficiency solution deposited thin film photovoltaic device,
Adv. Mater. 20 (2008) 3657.
[6] L. Li, N. Coates, D. Moses, Solution processed inorganic solar cell based on in situ sunthesis and
film deposition of CuInS2 nanocrystals. J. Am. Chem. Soc. 132 (2010) 22.
[7] M.G. Panthani, V. Akhavan, B. Goodfellow, Synthesis of CuInS2, CuInSe2, and
Cu(InxGa1-x)Se2(CIGS) nanocrystal inks for printable photovoltaics. J. Am. Chem. Soc. 130 (2008)
16770.
[8] P.K. Mishra, Y.K. Gautam, J.N. Prasad, A.K. Choudhary, Ramesh Chandra Effect of Working
Pressure on Structural, Electrical and Optical Properties of CIGS Thin Film Deposited by PLD, Adv.
Mater. Res. 768 (2014) 70-74.
[9] S.C. Chen, D.H. Hsieh, H.Jiang, Y. K. Liao, F. I. Lai, C. H. Chen, Growth and characterization of
Cu (In, Ga) Se2 thin films by nanosecond and femtosecond pulsed laser deposition. Nanoscale Res.
Lett. 9 (2014) 280.
[10] M. G.Tsai, H. T. Tung, I. G. Chen, C. C. Chen, Y. F. Wu, X. Qi, C. W. Cheng, Annealing Effect
on the Properties of Cu (In0. 7Ga0. 3) Se2 Thin Films Grown by Femtosecond Pulsed Laser Deposition,
J. Am. Ceram. Soc. 96 (2013) 2419-2423.
[11] Y. Zhao, H. Li, X. J. Liu, L. L. Guan, Y. L. Li, J. Sun, N. Xu, Enhancing the light absorbance of
polymer solar cells by introducing pulsed laser-deposited CuIn0. 8Ga0. 2Se2 nanoparticles,
Nanoscale Res. Lett. 9 (2014) 1-6.
Advanced Materials Research Vol 1102 (2015) pp 87-90 Submitted: 2015-01-04
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.87

Effect of Nano-oxide Addition on Mechanical Properties of


Alumina-Mullite-Zirconia Composites
Anurat Poowancum and Sukasem Watcharamaisakul*
School of Ceramic Engineering, Institute of Engineering
Suranaree University of Technology, Nakhon Ratchasima 30000 Thailand
e-mail*:sukasem@sut.ac.th

Keywords: Alumina-Mullite-Zirconia Composite, Oxide Nanoparticles, Fracture toughness

Abstract

This work reports on the effect of CeO2 and Cr2O3 addition on the mechanical properties of
Al2O3-mullite-ZrO2 composites. The amount of additives added was 1.5wt%CeO2, 1.5wt%Cr2O3
and the mixture of 1.5wt%CeO2 and 1.5 wt%Cr2O3. Bulk composites have been sintered by using
pressure-less sintering method. Phases, microstructure, density and porosity of sintered composites
were characterized. Flexural strength and fracture toughness were also determined. The results from
mechanical tests showed that the highest flexural strength was obtained with the mixture of CeO2
and Cr2O3 addition, while the highest fracture toughness was obtained with CeO2 addition.

Introduction

Alumina ceramic is one of the most widely used material for engineering applications
because of its exceptional mechanical properties such as high hardness, high strength and good
chemical stability but rather weak in fracture toughness [1]. Many attempts have been made to
improve the fracture toughness of alumina ceramic by using various types of additives [2]. The
addition of zircon into Al2O3 system has been studied by Claussen and Jahn and was found to
improve fracture toughness and flexural strength by the formation of Al2O3-mullite-ZrO2 (AMZ)
composite [3]. Many research works were also reported positively on the effect of CeO2 or Cr2O3
addition on mechanical properties of zirconia-mullite system [4], but none has been reported on the
effect of these additives for AMZ system.The purpose of the present work is to study the effect of
CeO2, Cr2O3 and the mixture of CeO2 and Cr2O3 additions in an attempt to improve the mechanical
properties mainly flexural strength and fracture toughness of AMZ composite.

Experimental

The major starting materials were 4 m-Al2O3 powder (AM-21, Sumitomo Chemical Co.,
Japan), 150 nm zircon powder, 1.5 wt% CeO2 with d50 = 120 nm and 1.5 wt% Cr2O3 with d50 = 100
nm were used as the additives. The sample code and batch compositions are given in Table1. Each
batch was ground and mixed with distilled water in an attritor mill at 500 rpm for 6 hr using
zirconia balls. The mixtures were prefomed by dry pressing and then isostatically pressed at 120
MPa with PVA as a binder. All green compacts were sintered in air for 2 hr at 1575oC. Sintered
composites were cut, ground with a diamond wheel and polished using 3m and 1m diamond
suspensions, respectively. The thermal etching was performed at a temperature 100oC below
sintering point for 1 hr before the microstructure analysis by scanning electron microscopy (SEM).
Specimens for mechanical test were cut into a bar shape with dimensions of 4mm x 4mm x 30mm.
the strength was measured using a three-point bending method using an Instron 5565 universal
testing machine. The span was 20 mm and cross head speed was 0.5 mm/min. the fracture
toughness were determine by Vickers indentations techniques on the polished surface, using applied
loads 5 kgf. The fracture toughness calculated according to following equation:
88 Metal, Ceramic and Composite Materials

KIC = 0.016 (E / H )
1/ 2
(P / C ) 3/ 2

Where E is the elastic modulus, H is Vickers hardness, P is the indentation load and c is half of
average of the crack length. The phase development was analyzed by X-Ray diffractometer
(Bruker-D5005, Germany) with Ni filtered, CuK radiation. The microstructure of the specimens
was observed by SEM (JEOL-JSM6400, Japan) after thermal etching. Thermal etching was
performed on the polished specimens at a temperature 100oC below sintering temperature for 1 h.

Table 1 Batch composition and sample codes for different batch


No. Sample code Al2O3 ZrSiO4 CeO2 Cr2O3
(wt.%) (wt.%) (wt.%) (wt.%)
0 A 100 - - -
I AMZ 80 20 - -
II AMZ-Ce 78.8 19.7 1.5 -
III AMZ-Cr 78.8 19.7 - 1.5
IV AMZ-CrCe 77.6 19.4 1.5 1.5

Results and discussion

By adding zircon in alumina it was found strength and toughness of sample I was increased
and apparent porosity was decreased illustrate in Table 2, Fig1 and Fig 2.

Table 2 Physical properties of composites


Properties Batch No.
0 I II III IV
3
Bulk density (g/cm ) 3.44 3.66 3.76 3.61 3.77
True density (g/cm3) 3.94 3.86 3.94 3.89 3.96
Apparent porosity (%) 7 3.74 2.26 4.23 2.08
Fracture toughne ss (M Pam )
0 .5

550.00 8
Fle xural s tre ngth (M pa)

500.00
7
450.00
400.00 6
350.00 5
300.00
4
250.00
200.00 3
150.00 2
100.00 1
50.00
0.00
0
Pure Alumina 80/20 CeO2 Cr2O3 CeO2+Cr2O3 P ure Alumina 80/20 CeO2 Cr2O3 CeO2+Cr2O3

Fig1. Flexural Strength of composite Fig2. Fracture toughness of composite


contain different composition contain different composition

Because of two reason, first reason is, zircon promote sinterability of alumina demonstrate
by apparent porosity decrease when add zircon. And second reason is, mullite and zirconia were
appearing in the samples which contain zircon (illustrate in Fig3.).
Advanced Materials Research Vol. 1102 89

A:Al2O3
M:mullite
m:m-ZrO2
t :t-ZrO2

Fig3. XRD patterns for different batch after sintering at 1575oC for 2 h.

Mullite and zirconia are important factor for strength and fracture toughness. The sample
which use CeO2 as the additive show high strength and highest fracture toughness because of, Wen-
Cheng report, the densification rate of zirconia mullite composite was prepared by reaction sintering
can be enhanced by the addition of CeO2. And that illustrate in Table2, AMZ-Ce (batch No.II) has
low apparent porosity but even if, the apparent porosity of AMZ-Ce is lower than AMZ, but
strength of AMZ-Ce is still lower than AMZ. Due to the grain size of AMZ-Ce is large (show in
Fig4.). The influence of large grain size was reducing the strength but enhance the fracture
toughness. It was reported by Annamaria in 2003, D.Casellas in 2003, is in the sample which large
grain size have high fracture toughness due to effect of crack deflection and grain bridging
mechanism. That is the result for AMZ-Ce show high strength and fracture toughness.
When Cr2O3 was added in AMZ strength decrease due to Cr2O3 reduce sinterability of
AMZ. The effect of Cr2O3 on sinterability of Al2O3 was reported by Kato et al., the sintering of
alumina is retarded by Cr2O3 and the retarding effect of Cr2O3 was confirmed by Gerasimov et al. In
this study, the main phase of AMZ is Al2O3 so, it possibility as AMZ was reduced sinterability by
add Cr2O3. The evidence was illustrate in Table2, the apparent porosity of AMZ (batch No.I) was
increased by add Cr2O3 (Batch No.III). The maximum value of strength was observed with the
batches containing both CeO2 and Cr2O3 as additives. Ryshkewitch, Smothers and Reyholds was
report, grain size of Al2O3 was repressed by Cr2O3. So when we use both CeO2 and Cr2O3 as
additives the result are, the CeO2 play the role on promote sinterability, the grain growth of AMZ
was repressed by Cr2O3. Therefore the grain size of sample in the batch AMZ-CrCe was smaller
than AMZ-Ce and the sample AMZ-CrCe was lowest apparent porosity, illustrate in Fig4, Table2
respectively. However the toughness of AMZ-CrCe was lower than AMZ-Ce due to, the grain size
of AMZ-CrCe was smaller than AMZ-Ce.
90 Metal, Ceramic and Composite Materials

Fig.4. SEM micrographs of Alumina-Mullite-Zirconia composite


(a) AMZ, (b) AMZ-CE, (c) AMZ-Cr, (d) AMZ-CrCe

Conclusion

The mechanical properties of alumina such as flexural strength, fracture toughness can
improve by add a few zircon in alumina. The additives were influence on mechanical properties of
alumina-mullite-zirconia composites. The maximum value of toughness was observed with the
batches containing CeO2 as additive, the sample was added both CeO2 and Cr2O3 as the additives
show highest strength.

References
[1] A. Celli, A.T.L. Esposito and C. Palmonari, J. Euro.Cer.Soc. 23(2003) 469-479.
[2] C. Aksel, Cer.Inter., 29(2002) 311-316.
[3] D. Casellas, M.M. Nagl, L. Llanes and M. Anglada, J.Mat.Process. Tech. 143-144(2003) 148-
152.
[4] D.H. Riu, Y.M. Kong and H.E. Kim, J. Euro.Cer.Soc. 20(2000), 1475-1481.
[5] H. Schneider, K.Okada and J.A.Pask Mullite and Mullite Ceramics, John Wiley & Sons Ltd.,
England, (1994)
Advanced Materials Research Vol 1102 (2015) pp 91-94 Submitted: 2015-01-05
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.91

Hydrogen Separation Using Inorganic Membranes

Mohammed Nasir Kajama1, Ngozi Claribelle Nwogu1 and Edward Gobina1*


1
Centre for Process Integration and Membrane Technology, (CPIMT), School of Engineering, The
Robert Gordon University, Aberdeen, AB10 7GJ. United Kingdom.

m.n.kajama@rgu.ac.uk, n.c.nwogu@rgu.ac.uk, e.gobina@rgu.ac.uk*

Keywords: alumina membrane; gas permeation; hydrogen separation.

Abstract

Gas permeation of hydrogen (H2) and nitrogen (N2) were obtained from 30 and 6000 nm pore
diameter tubular commercial alumina ceramic membranes at 0.05 to 1.00 bar and 298 K. Flow rates
of up to 3.279 and 2.296 l/min were obtained for H2 and N2 respectively. The ratio of H2/N2 flow
rates were used to calculate H2/N2 selectivity. The experimental H2/N2 selectivities obtained were
1.85 and 1.43 for the 30 and 6000 nm respectively.

Introduction
Almost 80 percent of the global energy demand comes from fossil fuels such as natural gas, coal
among others. Unlike using fossil fuels, hydrogen when combusted produces only water as
byproduct [1]. The world energy consumption is forecasted to rise by 56 percent from 2010 to 2040
mainly from coal and natural gas [2]. Therefore, it is advantageous to substitute hydrogen with the
current fossil fuels because it is widely accepted as a clean energy carrier in for instance fuel cell
systems. These could help to address the problems linked to energy security which includes air
pollution and global climate change. The need for hydrogen as a source of renewable energy will be
enhanced in the coming years due to its demand for raw material processing in the chemical
industry as well as home heating [3]. The state of the art membrane technology has emerged as an
attractive substitute in the chemical industry, particularly for hydrogen separation from process
streams.
Membranes can be classified into inorganic and organic/inorganic (hybrid) systems. The organic
ones are further divided into biological and polymeric constituents, while the inorganic membranes
can be divided into metallic and ceramic (porous and non-porous) membranes [4]. The applications
of inorganic membranes have received considerable attention during the past three decades in a
wide array of industrial operations for gas separation due to their thermal, chemical and mechanical
stability when compared with polymeric membranes. This includes hydrogen separation/recovery
for fuel cell application to generate electricity [5], carbon dioxide separation from fuel gas [6],
hydrogen sulphide separation from natural gas [7] and recovery of methane from bio-gas [5].
Inorganic membranes are environmentally benign, cost-effective and energy efficient.
According to the International Union of Pure and Applied Chemistry (IUPAC), porous membranes
are classified as; Micropores 0.5 - 2 nm where separation is based on molecular sieving mechanism,
mesopores 2 - 50 nm where Knudsen flow mechanism is the dominant flow but multilayer flow
and/or capillary condensation and viscous flow can also take place, and macropores > 50 nm where
there is no separation and the flow mechanism is basically influenced by viscous flow [8-10]. In all
cases, some considerations on productivity and separation selectivity, mechanical integrity, and
membranes durability at the operating conditions needs to be taken into account against cost issues
[4], because the importance of each of these requirements differs with their applications.
Permeability and selectivity are the main basic properties of a membrane.
92 Metal, Ceramic and Composite Materials

The main aim of the present paper is to report experiments undertaken using a tubular commercial
alumina membrane to understand the transport through inorganic membranes for hydrogen
separation from natural gas.

Experimental
Commercially available -alumina mesoporous ceramic membranes (77% alumina and 23% TiO2)
tubes 368 mm long with 7 and 10 mm internal and outer diameter where obtained from Ceramiques
Techniques et Industrielles (CTI SA) France. The ceramic membranes had a 30 and 6000 nm pore
diameter.
Figure 1 shows the permeation system. Each membrane was sealed in a stainless-steel reactor by
graphite O-rings which are high temperature resistant. The feed pressure used was between 0.05 to
1.0 bar and was measured by a pressure transducer. At the permeate, the permeation flux was
measured using a flowmeter, and a thermometer to record the required temperature. The permeate
side was fully opened to the atmosphere. The temperature was maintained at 298 K. The permeate
pressure was always the atmospheric one.

Pressure Gauge 4
Pressure Gauge 3

Valve 4 Feed
Membrane Valve 3

Stainless Steel Membrane Reactor

Main Feed
Pressure Gauge 2
Pressure Gauge 1
Digital Gas
Flow Meter
Retentate Exit
Valve 1
H2
N2

Valve 2
Permeate to the Flow meter

Figure 1: Permeation test experimental system.

Results and discussion


Figure 2 shows the H2 and N2 single gas flow rates as a function of feed pressure through a 30 and
6000 nm commercial alumina membrane was studied at 298 K. H2 flow rates on 6000 nm
permeated faster than that of 30 nm, and a similar trend can be seen for N2 flow rates (Figure 3).
This result indicates that the higher the pore size, the higher the permeation rate.

Figure 2. H2 flow rate as a function of feed pressure for 30 and 6000 nm at 298 K.
Advanced Materials Research Vol. 1102 93

Figure 3. N2 flow rate as a function of feed pressure for 30 and 6000 nm at 298 K.
Figure 4 shows a comparison of the H2 and N2 flow rates as a function of feed pressure for the 30
nm pore membrane. It can be seen that hydrogen had the higher flow rate due to its lower molecular
weight compared with nitrogen which indicates knudsen flow contribution. Table 1 shows H2 and
N2 flow rates and H2/N2 selectivities from 30 and 6000 nm commercial alumina membrane at 1 bar
and 298 K. Experimental gas selectivities were calculated as the ratio of H2 and N2 flow rates. The
experimental H2/N2 selectivities obtained were 1.85 and 1.43 from 30 and 6000 nm respectively. An
increase in selectivity was obtained from 30 nm pore diameter membrane. This is a clear indication
that the smaller the pore diameter, the higher the selectivity that would be achieved. Consequently,
microporous membranes can be employed in order to enhance higher selectivities than those of
knudsen.

Figure 4. H2 and N2 flow rate as a function of feed pressure for 30 nm at 298 K.

Table 1. Hydrogen and nitrogen flow rates and H2/N2 at 1 bar and 298 K
Membrane pore Flow rate (l/min) Selectivity, (H2/N2)
diameter (nm)
H2 N2 Theoretical Experimental

30 2.554 1.384 3.74 1.85

6000 3.279 2.296 3.74 1.43


94 Metal, Ceramic and Composite Materials

Coclusions

A simple but effective system using commercial alumuna mesoporous membrane has been designed
to measure and compare the permeation of single gases (H2 and N2) at 0.05 to 1.0 bar feed pressure
and 298 K. Higher flow rates were obtained for H2 and N2. The ratio of H2 and N2 flow rates used to
calculate H2/N2 selectivity.

Acknowledgement

The authors gratefully acknowledge Petroleum Technology Development Fund (PTDF) Nigeria for
funding this research.

References
[1] A. Sushil, F. Sandun, Hydrogen membrane separation techniques, Ind. Eng. Chem. Res., 45
(2006) 875-881.
[2] http://www.eia.gov/forecasts/ieo/world.cfm. (Accessed on 31/03/2014).
[3] H.A. Meinema, R.W.J. Dirrix, H. W. Brinkman, R.A. Terpstra, J. Jekerle, P.H. Kosters, Ceramic
membranes for gas separation Recent developments and state of the art, Interceram, 54 (2005) 86-
91.
[4] G.Q. Lu, J.C. Diniz da Costa, M. Duke, S. Giessler, R. Socolow, R.H. Williams, T. Kreutz,
Inorganic membranes for hydrogen production and purification: A critical review and perspective,
Journal of Colloid and Interface Science, 314 (2007) 589-603.
[5] M.R. Othman, H. Mukhtar, A.L. Ahmad, Gas permeation characteristics across nano-Porous
inorganic membranes, IIUM Engineering Journal, 5 (2004) 17-33.
[6] D.R. Luebke, H.W. Pennline, C.R. Myers, Surface selective membranes for carbon dioxide
separation. Conference: 22nd Annual International Pittsburgh Coal Conference, Pittsburgh, PA,
Sept. (2005) 12-15.
[7] M.N. Kajama, H. Shehu, E. Gobina, Purification of Gases Using Nanoporous Inorganic
Membranes, International Journal of Scientific Engineering and Technology, 3 (2014) 1156-1159.
[8] L. Zhang, I. Park, K. Shqau, W.S. Winston Ho, H. Verweij, Supported inorganic membranes,
Promises & Challenges, 61 (2009) 61-71.
[9] A.L. Ahmad, N.N.N. Mustafa, Sol-gel synthesized of nanocomposite palladium-alumina
ceramic membrane for H2 permeability: preparation and characterisation, International Journal of
Hydrogen Energy, 32 (2007) 2010-2021.
[10] http://ufdcimages.uflib.ufl.edu/AA/00/00/03/83/00141/AA00000383_00141_00062.pdf. (Acc-
essed on 30/10/2014).
Advanced Materials Research Vol 1102 (2015) pp 95-98 Submitted: 2015-01-16
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.95

Magnesium phosphate/ polybutylene succinate composite for bone


repair
Quanxiang Wang1,a, Xieping Dong2,b, Xiaofeng Tang1,c, Yingyang Wu1,d,
Yueting Ding1,e, Baoqing Yu3,f, Jie Chen1,g, Jie Wei1*,h
1
Key Laboratory for Ultrafine Materials of Ministry of Education, East China University of Science
and Technology, Shanghai 200237, China
2
Department of Orthopaedic Surgery, Jiangxi People's Hospital, Nanchang 330006, China
3
Department of Orthopaedics, Changhai Hospital, Second Military Medical University, Shanghai
200433, China
a
email: leo823449871@163.com, email: jxzhyxh@163.com, cemail: 356180032@qq.com,
b

d
email: 13262918095@163.com, eemail: dyt901208@163.com, femail: bqy680314@qq.com,
g
email: 1060197245@qq.com, hemail:jiewei7860@sina.com

Keywords: Magnesium phosphate, PBS, Biomaterial.

Abstract. MP/PBS composite was fabricated by melt blending method. The weight loss ratio of
MP/PBS composite reached 35.43 wt.% after immersion into phosphate buffered saline solution for
12 weeks, and the pH change from 7.40 to 7.72. The relative proliferation rate of cells on the
materials increased with time and the proliferation of the cells on the MP/PBS composite was
significantly higher than PBS. The results reveal MP/PBS composite has good biocompatibility,
which satisfied the relevant criterion on biomaterials.

Introduction
Magnesium (Mg) is the fourth major cations of the human body, it was reported that divalent
cation of Mg plays critical roles in bone remodeling and skeletal tissue development of the human
body [1]. Recently, Mg-containing bioactive materials are receiving considerable attentions for
bone tissue engineering applications [2]. Polybutylene succinate (PBS), one of biodegradable
aliphatic polyesters, is an ideal candidate due to its good biocompatibility, easy-processing ability,
and non-toxic degradation products. However, some drawbacks have limited the wider applications
of PBS, like slow degradation rate, hydrophobicity and biological inert [3]. In order to overcome
these limitations, in this study, magnesium phosphate (MP)/PBS composite was fabricated and
characterized.
Materials and methods
Preparation of MP/PBS composite. MP was prepared as previously reported and PBS was
purchased from Shanghai Showa Highpolymer Co. LTD. MP/PBS composite with MP content of
40 wt.% were fabricated by melt blending method. The MP and PBS were dried at 80 for 12h,
which were blended together with Haake torque rheometer at 60 rpm, 130 for about 10min. Then
the disc-shaped composite samples (12 2 mm) were prepared by plate vulcanization machine at
130 and 10MPa (pressure). As a control, PBS disc-shaped samples (12 2 mm) were prepared
using same method. The composite were characterized with X-ray diffractometry (XRD).
Degradation in phosphate buffered saline. The in vitro degradability of the MP/PBS
composite was determined by testing the weight loss ratio of the samples in phosphate buffered
saline solution. Briefly, the samples were weighed (Wi), immersed into phosphate buffered saline
solution (pH=7.4), and incubated at 37 with constant shaking at 100 rpm for 12 weeks. At
96 Metal, Ceramic and Composite Materials

preselected time point, the samples were removed from the solution and rinsed gently with
deionized water. The samples were dried at 37 until the water volatilized totally, and then
weighed (Wf). The weight loss was calculated according to the following equation:
Weight loss (%) = (Wi Wf)/Wi100%
The surface morphology of MP/PBS composite samples after immersing into phosphate
buffered saline solution for 12 weeks were observed by SEM. The pH value of the phosphate
buffered saline solution containing the MP/PBS composite was measured using a pH meter at
different time.
Cell proliferation. Cell proliferation was determined using cell counting kit-8 (CCK-8) assay
at 1, 3 and 7 days of culture using a standard procedure. Briefly, the MP/PBS composite specimens
(12 2 mm) were sonicated in ethanol and sterilized using ultra-violet light. MC3T3-E1 cells
with a density of 3104 cells/mL were seeded on the samples, which were placed into a 24-well
plate. At each time point, the samples were gently rinsed with phosphate-buffered saline three times
and then transferred to a new 24-well plate. A total of 50 L of CCK-8 solution was added to each
well and empty wells containing DMEM were set up as a blank control. After incubated for 3 hours,
10 L of the supernatant was transferred into a 96-well plate and read at 450 nm using a microplate
reader with 620 nm as a reference wavelength. The mean absorbance values/optical density (OD)
obtained from the blank control were subtracted from the ODs of the test groups.

Results and discussions


b

BM
P
S/
P
P
B
S
1
0

2
0

3
0

4
0
2
T
h
e
t
a
(
d
e
g
)

Fig.1 Digital photo (a) and XRD pattern (b) of PBS and MP/PBS composite
Digital photo and XRD. Fig.1a is digital photo of PBS and MP/PBS composite. The phase
composition of PBS and MP/PBS composite were characterized by XRD as shown in Fig.1b. It can
be seen that the peaks at 2=12.2, 13.5, 21.1, 23.2, 29.6, 31.4 and 34.7are attributed to MP,
while two strong peaks around 20 belong to PBS. All the characteristic peaks appear in MP/PBS
composite in the XRD pattern.
Advanced Materials Research Vol. 1102 97

5
0

ab
ab

7
.
8
4
0

)
%
(
s

7
.
6
s
3
0

o H
L p
t
2
0

7
.
4
h
g
i
e
W
1
0

7
.
2
0

5m

6d

1
0
1
1
0

1
0

1
2
t
i
m
e
(
w
e
e
k
)

T
i
e
(
a
y
)
Fig.2 Weight loss ratio (A) and pH change (B) of PBS (a) and MP/PBS composite (b) in phosphate
buffered saline solution
Degradability and pH change. Fig.2A presents the weight loss ratio of PBS and MP/PBS
composite immersed in phosphate buffered saline solution versus time. The weight loss ratio of
PBS and MP/PBS composite increased with time, and the weight loss ratio of PBS reached 17.95
wt.% after immersion into solution for 12 weeks. While the weight loss ratio of MP/PBS composite
reached 35.43 wt.% after immersion into solution for 12 weeks. The results showed that the
MP/PBS composite degraded better in phosphate buffered saline solution.
Fig.2B shows the pH change of the solution after PBS and MP/PBS composite immersed into
phosphate buffered saline solution for different time. It is found that the pH of the solution for PBS
reduced from 7.40 (initial) to 7.30 (at 10 days) while the pH for MP/PBS composite increased from
7.40 (initial) to 7.72 (at 10 days). The increasing of pH of the solution containing MP/PBS
composite might be correlated to the dissolution of alkaline ions (Mg2+) from the MP.

(a) (b)

Fig.3 SEM images of morphology of PBS (a) and MP/PBS composite (b) after immersed in
phosphate buffered saline solution for 12 weeks.
Fig.3 shows the surface morphology of PBS and MP/PBS composite after immersion in
phosphate buffered saline solution for 12 weeks. MP/PBS composite surface exhibited more surface
roughness than PBS, and this increase the surface area of composite, which means MP accelerated
the degradation of PBS.
98 Metal, Ceramic and Composite Materials

3
.
0

ab
2
.
5
2
.
0
1
.
5
D
O

1
.
0
0
.
5
0
.
0

3T

7
1

C
u
l
t
u
r
e
i
m
e
(
d
a
y
s
)
Fig.4 Proliferation of MC3T3-E1 cells cultured on PBS (a) and MP/PBS composite (b) at
different time evaluated by CCK-8 assay.
Cell proliferation. Fig.4 shows the proliferation of MC3T3-E1 cells on the materials
evaluated by CCK-8 assay. The relative proliferation rate of cells on the materials increased with
time and the proliferation of the cells on the MP/PBS composite was significantly higher than PBS.

Conclusions
MP/PBS composite was fabricated in this study. The results revealed that the MP/PBS composite
had a better degradability than PBS, and the pH change of MP/PBS composite was appropriate.
Through the CCK-8 assay, the proliferation of cells on the MP/PBS composite was significantly
higher than PBS. These results indicated that the MP/PBS composite had potential for being used as
a good kind of biomaterial for bone repair.

Acknowledgements
The authors are grateful for the financial support from National Key Technology R&D Program of
the Ministry of Science and Technology in the 12th Five Year Plan of China (2013BAI05B10), and
Key Medical Program of Science and Technology Development of Shanghai (No.12441903600,
12441902802).

References
[1] Z. S. Seyedraoufi and S. Mirdamadi, Effects of pulse electrodeposition parameters and alkali
treatment on the properties of nano hydroxyapatite coating on porous MgZn scaffold for bone
tissue engineering application, Materials Chemistry and Physics. 148 (2014) 519-527.
[2] X. Ye, S. Cai, Y. Dou, G. Xu, Bioactive glassceramic coating for enhancing the in vitro
corrosion resistance of biodegradable Mg alloy, Applied Surface Science. 259 (2012) 799-805.
[3] M. L. Alves da Silva, A. Crawford, J. M. Mundy, Chitosan/polyester-based scaffolds for
cartilage tissue engineering: Assessment of extracellular matrix formation, Acta Biomaterialia. 6
(2010) 1149-1157.
Advanced Materials Research Vol 1102 (2015) pp 99-102 Submitted: 2015-01-20
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.99

Characterisation of Inorganic Composite Ceramic Membrane for Lactic


Acid Esterification Processes

Edidiong Okon1,a, Habiba Shehu2,b and Edward Gobina3,c*


1,2,3,
Center for Process Integration and Membrane Technology (CPIMT), School of Engineering.
The Robert Gordon University, Aberdeen, United Kingdom, AB10 7GJ.
E-mail: ae.p.okon@rgu.ac.uk, bh.shehu@rgu.ac.uk & *Corresponding Author: ce.gobina@rgu.ac.uk.
Phone No.: +44(0)1224262348.

Keywords: Composite membrane, Nitrogen adsorption, Gas transport, Esterification and


Characterisation.

Abstract. The use of inorganic composite membranes in chemical industries has received a lot of
attention more recently due to a number of exceptional advantages including thermal stability and
robustness. Inorganic membranes can selectively remove water from the reaction mixture during
esterification reactions in order to enhance product formation. The characterisation of inorganic
composite membranes used in this work including the determination of the pore diameter and
specific surface area was performed using Liquid Nitrogen adsorption at 77 K. The membrane was
modified once. The permeation test for the single gases including carbon dioxide (CO2), helium
(He), nitrogen (N2) and argon (Ar) through the inorganic composite ceramic membrane was carried
out at the gauge pressure range of 0.10 1.00 bar and at the temperature of 393 K. The order of the
gas molecular weight was He < N2 < CO2 < Ar. The BET surface area of the dip-coated silica
membrane showed a type IV isotherm characteristic of mesoporous structure with hysteresis. The
BJH curve of the silica-membrane was in accordance with mesoporous classification.

Introduction
Lactate ester is obtained from bio-based feedstock in which its separation is achieved from the
esterification of ethanol and lactic acid generated from biomass raw materials. This solvent can
replace the use of environmentally-damaging toxic and halogenated solvents such as chloroform
and carcinogenic methylene chloride [1]. The use of inorganic ceramic membrane to selectively
eliminate water from the reaction product during esterification of lactic acid is another important
application that has attracted a lot of attention [2]. Inorganic membranes have found significant
potential applications in high temperature and high pressure separation, filtration and catalytic
membrane reactor due to its unique structure, chemical and thermal stabilities [3]. The ability of a
membrane to allow the separation of mixtures is based on two parameters including selectivity and
permeability of the membrane material. Selectivity is defined as the ability of a membrane to
separate two different species, while permeability relates to the flux through the membrane with
respect to the thickness and driving force of the membrane [4]. The reduction of pore diameter of
the inorganic membranes can be achieved using different methods such as sol-gel, chemical
deposition and sintering methods [3]. The gas transports through inorganic membranes are
classified into five different classes namely: capillary condensation, Knudsen diffusion, surface
diffusion, viscous flow and molecular sieving mechanisms. However, these mechanisms also
depend on the chemical interaction between the permeating gas molecules, the membrane material
and also pore size distribution of the membrane [5]. In Knudsen flow mechanism, gas separation
depends on the inverse square root of the gas molecular weight. Viscous flow mechanism occurs
when there is a defect or a crack on the structure of the membrane and as such, the gas molecule can
move across the membrane without any separation taking place. Molecular sieving mechanism
usually occurs in microporous or submicroporous systems whereby separation occurs by the
diffusion of smaller molecules. Surface diffusion depends on both the mobility and adsorption of
the membrane. In solution diffusion, the gas molecule dissolves into the membrane material and
diffuses to the permeate side [4].
100 Metal, Ceramic and Composite Materials

Experimental

Permeation Tests
The permeation tests of single gases (helium, nitrogen, carbon dioxide, and argon) were carried out
at 120 oC to determine the gas transport mechanism based on the procedure developed by Gobina
(2006) [6]. The schematic diagram of the permeation test setup is shown in figure 1.

Fig 1. Schematic diagram of gas permeation setup [7].

Nitrogen Adsorption
Figure 2 shows the schematic diagram of the Liquid Nitrogen adsorption instrument. The sample
was weighed before and after the analysis. Prior to the analysis, the sample was degased with
helium and nitrogen gases in order to remove impurities before the analysis and also for measuring
the total volume of the sample. The degassing temperature was set at 65 oC for 2 hrs prior to the
sample analysis. The liquid nitrogen temperature was 77 K. The surface areas of both the internal
and external pores of the membrane support were determined using the Brunauer-Emmett-Teller
(BET) isotherm while the pore size distribution of the sample was obtained using the Barrette-
Joyner-Halenda (BJH) curve. A similar procedure to that of Vospernik et al. (2004) [8] was
employed in the analysis with some modification in the temperature. The sample cell was 27.7g
while the membrane + sample cell was 34.8g. The sample was analysed at 350 oC for 6 hrs after the
degassing process.

Fig. 2 Schematic diagram of the liquid nitrogen adsorption.

Results and discussion

Figure 3 shows the plot of the gas permeance (molm-2s-1Pa-1) against the inverse square root of the
gas molecular weight (mol/g) at 393 K.
Advanced Materials Research Vol. 1102 101

0.0000035

Gas permeance (molm-2s-1Pa-1)


0.000003 Ar

0.0000025
CO2
0.000002
N2
0.0000015
0.000001 He
0.0000005
0
0 0.1 0.2 0.3 0.4 0.5 0.6
1/MW

Fig. 3 Gas permenace (molm-2s-1Pa-1) against inverse square root of the gas molecular weight
(mol/g) at 393 K.
From figure 3 it was found that Ar with the molecular weight of 40g/mol showed a higher
permeance in contrast to CO2 with the molecular weight of 44g/mol. For the gas flow to be
described by Knudsen mechanism of transport, it was expected that the result in figure 3 could have
been a straight line graph starting from the origin with CO2 (44g/mol), Ar (40g/mol), N2 (g/mol) and
He (4g/mol) point passing through the best fit line. However, it was suggested that Knudsen flow
mechanism was responsible for CO2 and N2 gas flow, whereas He and Ar were controlled by
another flow mechanism at 393K. The decreasing order of the gas molecular weight is given as He
< N2 < Ar < CO2.
The graph of volume (cc/g) was plotted the relative pressure (P/Po) to obtained the BET and BJH
curves for the surface area and the pore diameter analysis of the dip-coated membrane. Figure 4 and
5 shows the BET and BJH curves for the dip-coated silica membrane.

Fig. 4 Nitrogen adsorption isotherm for dip-coated silica membrane at 77 K.

Fig. 5 BJH curve for dip-coated membrane at 77 K.


102 Metal, Ceramic and Composite Materials

From the result obtained in Figure 4, it was observed that the silica membrane showed a type IV
isotherm with hysteresis indicating a mesoporous layer which is classified in the range of 2 50nm.
The surface area of the ceramic membrane was found to be 5.306m2/g while the pore diameter of
the membrane was 4.180nm. The result obtained from the BJH curve in Figure 5 also confirmed
that the membrane pore diameter (4.180nm) was in accordance with the mesoporous classification
range which will be compared in subsequent analysis.

Conclusion
The inverse square root of the gas molecular weight in relation to the gauge pressure did not show
any evidence of Knudsen mechanism of gas transport. The Knudsen flow mechanism was
responsible for CO2 and N2 gas flow which were found to be close to the best fit line, whereas He
and Ar gas flow was controlled by another flow mechanism at 393K. The surface are of the dip-
coated membrane possess a type IV isotherm with hysteresis indicating a mesoporous layer in the
range of 2-50nm. The result for pore diameter of the silica membrane corresponded with the
mesoporous membrane classification.

References

[1] Pereira CS, Rodrigues AE. Ethyl Lactate Main Properties, Production Processes, and
Applications. Alternative Solvents for Natural Products Extraction. : Springer; 2014. p. 107-125.

[2] Asthana N, Kolah A, Vu DT, Lira CT, Miller DJ. A continuous reactive separation process for
ethyl lactate formation. Organic process research & development. 2005; 9(5):599-607.

[3] Cameron M, Gartland I, Smith J, Diaz S, George S. Atomic layer deposition of SiO2 and TiO2
in alumina tubular membranes: pore reduction and effect of surface species on gas transport.
Langmuir. 2000; 16(19):7435-7444.

[4] Zornoza B, Casado C, Navajas A. Chapter 11 - Advances in Hydrogen Separation and


Purification with Membrane Technology. In: Diguez LMGAM, editor. Renewable Hydrogen
Technologies. Amsterdam: Elsevier; 2013. p. 245-268.

[5] Lee H, Yamauchi H, Suda H, Haraya K. Influence of adsorption on the gas permeation
performances in the mesoporous alumina ceramic membrane. Separation and purification
technology. 2006; 49(1):49-55.

[6] Gobina, E. Apparatus and Method for separating gases. United state patent. Patent
No.7,048,778B2. 2006. Robert Gordon University, Aberdeen, UK.

[7] Edidiong O., Habiba S and Edward G. Novel Application of Gas Transport Properties with
ceramic membrane for VOC Emission and Lactic Esterification. 2014; 2(2): 37-44. Robert Gordon
University, Aberdeen, UK.

[8] Vospernik M, Pintar A, Beri G, Batista J, Levec J. Potentials of Ceramic Membranes as


Catalytic Three-Phase Reactors. Chemical Engineering Research and Design. 2004; 82(5):659-666.
Advanced Materials Research Vol 1102 (2015) pp 103-106 Submitted: 2015-01-21
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.103

Fatigue Life of Arenga Pinnata/Epoxy Composites

Abdul Hakim Abdullah1,a and Muhd Faiz Mat2,b


1
Faculty of Mechanical Engineering, University of Technology MARA, 40450 Shah Alam, Selangor
Darul Ehsan, Malaysia
a
hakimnen@yahoo.com and bmuhdfaiz7062@salam.uitm.edu.my

Keywords: Fibre reinforced composites, Arenga Pinnata fibres, fatigue

Abstract. In this study, fatigue properties of unidirectional Arenga Pinnata fibre reinforced epoxy
composites have been investigated under tension-tension fatigue loading. Composites were made at
25% by Arenga Pinnata fibre weight contents using hand lay-out technique. Tensile test was
performed prior fatigue loading evaluation. The composites where then were cycled between 90%,
80%, 70% measured from ultimate tensile stress (UTS), with 10 Hz of frequency and stress ratio of
R at 0.1. Tensile results indicated that the composites resist to deformation and this reaction denotes
that the composites are a brittle material. Fatigue life of composite was improving as the cycled
stress decreases. The fatigue resistance were calculated by employing analytical expression and
compared with relevant fatigue data.

1. Introduction
Fibre reinforced polymer (FRP) composites is one of the emerging materials since the World
War 2. The uses of glass, aramid, carbon, Kevlar as FRP has become prominent in the application
such as construction, automobile, defence system and aircraft [1]. These materials provide good
mechanical properties for desired applications. However, these materials are technically not
environmental friendly because they are not easily to dispose at end of their life and may become
problems in the future. Instead, natural fibres have become popular by many researchers and
engineers throughout the world in FRP composites as long the strength is not critical factor. It is
well known that natural fibres provide sustainability, eco-friendly and more importantly, remain
low in cost [2]. More and more automobile makers embrace natural fibres as their parts or
components in the vehicles.
Arenga Pinnata or famously known as Ijok for local Malaysian is an abandoned natural
fibres, growing wild in the secondary forest. For centuries, this fibre has been used in traditional
application such as paint brush, filter, and door mat chair/sofa cushion [3]. The introduction Arenga
Pinnata reinforced into polymer composites could bring future development in green materials.
The studies of Arenga Pinnata and its composites have been reported by many authors in the
past [3-5]. However, none of the subjects deals with the fatigue evaluation or performances of this
composite material. Fatigue failure generally is difficult to predict for a material and it is even
harder for composites materials. The incompatibility issues between hydrophilic natural fibres and
the hydrophobic polymer matrix could yield inferior fatigue responses [5].
This research investigates the fatigue properties of Arenga Pinnata fibre reinforced epoxy
composite. The fatigue data is intended to be preliminary and exploratory has been described by
the ASTM D3479 that describes critical overview of the fatigue studies.

Methodology

Specimen preparation. The raw fibres were harvested from its tree in Kuala Pilah, Negeri
Sembilan, Malaysia. The raw fibre was cleaned and all the dirt was removed to ensure cleanliness
of specimens. At the same time, an aluminium mould was prepared using hand saw and drill
machine. The rectangular shape mould parameters are 180 mm wide, 280 mm length pattern and 5
mm thickness.
104 Metal, Ceramic and Composite Materials

The weight ratio of the composite specimen was 10% fibre to 90% epoxy resins. The hand
layout technique was used for composites fabrication. The fibres were aligned carefully inside the
mould cavity. The weight ratio of the composite specimen is 10% fibre to 90% epoxy respectively.
Next, the mixture of epoxy resin and hardener were then poured inside the mould cavity. Using a
brush, the layer of mixtures was distributed evenly inside the mould. The mould cavity was then
closed for curing process overnight. Finally, the cured composite plate was removed for cutting
process.
The specimens preparation was in accordance to ASTM D3039 standards. The standard
was required for tensile test and pre-determined fatigue test description, the ASTM D3479.
Coupons having dimension of 5 mm and the width is 25 mm with the specimen length being 20mm;
longer than support span, shown in Fig. 2.

Tensile Test. Tensile test was carried out by using Instron 8802 dynamic testing machine prior to
fatigue evaluation. The end coupons were attached on the hydraulic grip before the tensile were
loaded a constant rate 2 mm per minute up to rapture. The tensile tests measured the Youngs
modulus and Ultimate Tensile Strength (UTS).

Fatigue Test. The fatigue test was done in accordance to ASTM D3479 that describes tension-
tension fatigue loading evaluation. The repetitive maximum and the minimum stress was measured
from percentage UTS of the composites, 90% UTS, 80% UTS and 70% UTS respectively. The test
subject was then repeated for 3 times and ultimately, 9 specimens were plotted on S-N curves for
the purpose of preliminary and exploratory data scheme.
The ends coupons were held rigidly between a fixed beam and moving beam (the
crosshead). A load cell (sensor) is used to measure the stress that builds up in the material as its
length in increased by moving the crosshead. A low frequency of 5 Hz was set and 0.1 stress ratio
was applied for amplitude calculation. The number of cycles to failure of the specimens and time
needed for each specimen to break were recorded.

Results and Discussion

Tensile Data. Mechanical properties as result of tensile test were list in Table 1. All the means were
calculated according to normal distribution. The UTS and Youngs modulus were 25 MPa and 1.7
GPa respectively. The experimental values are compared with previous studies; Arenga Pinnata
reinforced high impact polystyrene (HIPS) composites and Arenga Pinnata/Epoxy composites [4,6].
The results indicate that the experimental values are likely almost identical to the Arenga
Pinnata/HIPS specimens unlike aging Arenga Pinnata/Epoxy composites and thus, the experimental
data have a good agreement. The variations of the tensile properties are attributed to many aspects
such as manufacturing technique, natural fibre variables, matrix formulation and handling
conditions.

Table 1. The tensile results of experimental and previous reference data

Materials Fibre Ratios UTS (MPa) Young's Modulus (GPa) References


Arenga Pinnata/Epoxy 25 20.1 1.7 Experimental
Arenga Pinnata/HIPS 20 23.2 1.6 [4]
Arenga Pinnata/HIPS 30 19.2 1.7 [4]
Arenga Pinnata/Epoxy 10 20 4.2 [6]

Fatigue Test Results. A stress level versus fatigue life graph or S-N curve from experiment is
shown in Figure 4 below. All the data are scattered on the huge scale corresponding to 9 specimens.
The 90% UTS specimens failed between 130,900 and 340,010 cycles. The reduce stress level up to
80% UTS shows further delay of failure. It is found that the cycle to failure varies between
Advanced Materials Research Vol. 1102 105

234,9091,140,120 cycles. Finally, the 70% UTS specimens indicates the first point of failure is
778,099 cycles. The next two dotted are run-out specimens, in this case the specimens were not
failing and test was stop since the fatigue test time consuming and costly. The survived specimens
are stopped at 1,034,000 and 1,230,900 cycles respectively. The experiment data indicates that the
fatigue life is improves as the stress level decreases.

20

18
Stress (MPa)

90% UTS
16 80% UTS

70% UTS

14

12
1x104 1x105 1x106 1x107

Cycles To Failure (Nf)

Fig. 2: S-N curve of Arenga Pinnata fibre reinforced epoxy composites. Arrows indicates the
specimens were run-out.

However, the fatigue life of Arenga Pinnata composites could describes something meaningful by
normalizing the stress level. Table 2 showed comparison fatigue life of composites taken from
references data for both natural fibres and man-made fibres reinforced polymer composites this
include kenaf and glass fibre respectively. While most of the references data fall within four digits
numbers, the experimental data however shifted into higher six digits numbers in between 70%
UTS and 90% UTS test setup. The increasing fatigue life on given similar normalized stress
indicates that the fatigue life of Arenga Pinnata is greater even the specimens were cycled at high
fatigue loadings.
Although there is no further investigation deals with the fatigue damage of Arenga Pinnata
composites, it is perhaps the extension of fatigue life of the Arenga Pinnata composites can be
described by simple mechanics theory. It is acknowledged that the fatigue failures of polymer
composites are not related to plasticity as opposed to the metals [7]. Often, natural fibres themselves
could perform viscoelastic behaviour when undergoing deformation, in regards of the cell wall [8].
The elasticity properties on the other hand help the composites to delay its fatigue failure as been
shown in stress-strain curves in Figure 3.

Table 2. The tensile results of experimental and previous reference data

UTS (Cycle to Failure)


Material Reference
90% 80% 70% 60% 50%
130,900 234,909 778,099 - -
Arenga Pinnata/Epoxy 236,182 458,018 1,034,000* - - Experiment
340,010 1,140,120 1,230,900* - -
Kenaf/Epoxy 55 44 1209 75,950 550,010 [9]
17 88 213 1,056 3,492
Glass/Epoxy [10]
21 108 261 1,376 4,841
106 Metal, Ceramic and Composite Materials

Conclusion

Fatigue life of Arenga Pinnata reinforced epoxy composites was successfully investigated. The
fatigue life of composites improves as the stresses decreasing, with greater number of cycles were
recorded as compared to reference data.

Acknowledgement

The authors would like to thank the Research Management Institute (RMI) of Universiti Teknologi
MARA (UiTM) and Ministry of Education, Malaysia for financial support and facilitating this
project support to this study through Research Acculturation Grant Scheme (RAGS) awards 600-
RMI/RAGS 5/3 (47/2013) and 600-RMI/RAGS 5/3 (41/2012).

References

[1] Bryan Harris, Engineering Composite Materials, The Institute of Materials, London, 1999.
[2] A.H. Abdullah, A. Khalina, Aidy Ali. Effects of fiber volume fraction on unidirectional
kenaf/epoxy composites: The transition region, Polym-Plast Technol. 50 (2011) 1362-1366.
[3] M.F.S. Khalid, A.H. Abdullah. Storage modulus capacity of untreated aged arenga pinnata
fibre-reinforced epoxy composite, Applied Mechanics and Materials. 393 (2013) 171-176.
[4] S.M. Sapuan, D. Bachtiar. Mechanical properties of sugar palm fibre reinforced high impact
polystyrene composites, Procedia Chemistry. 4 (2012) 101-106.
[5] D. Bachtiar, S.M. Sapuan, M.M. Hamdan. The effect of alkaline treatment on tensile properties
of sugar palm fibre reinforced epoxy composites, Mater Design. 29 (2008) 1285-1290.
[6] Aidy Ali, A.B. Sanuddin, E. Saifuliwan, The effect of aging on Arenga pinnata fiber-reinforced
epoxy composite, Mater Design. 31 (2010) 3550-3554.
[7] B. Claude, An engineering point of view about fatigue of polymer matrix composite materials,
Int. J. Fatigue, 28 (2006) 1094-1099.
[8] M. Hughes, J. Carpenter, C. Hill. Deformation and fracture behavior of flax fibre reinforced
polymer matrix composites, J. Mater. Sci. 42 (2007) 2499-2511.
[9] A.H. Abdullah, S.K. Alias, N. Jenal, K. Abdan, Aidy Ali. Fatigue Behavior of kenaf fibre
reinforced epoxy composites, Engineering Journal. 16 (2012)105-114.
[10] M.H.R. Jen, C.H. Lee. Strength and life in thermoplastic composite laminates under static and
fatigue loads. Part 1: experimental. Int. J. Fatigue. 20 (1998) 605-615.
Advanced Materials Research Vol 1102 (2015) pp 107-112 Submitted: 2015-01-19
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.107

Preparation and Electrical Properties of Graphene/PEK-c Films

Fenghui Shi 1,2,a , Jia Li2,b , Baoyan Zhang1,2,c ,


Xueshi Qiu3,d, Liangyong Gu3,e
1
AVIC Beijing Institute of Aeronautical Materials, Beijing 100095, China
2
AVIC composite Co. ltd, Beijing 101300, China
3
Chengdu aircraft design institute, Chengdu 610041, China
a b c
fenghuishi@126.com, youxianxvxv@126.com, zhangbaoyan0916@163.com,
d e
qiu_xueshi@sina.com, buaa32050124@sina.com

Keywords: graphene, PEK-c, films, electrical properties.

Abstract. This article shows us a way to disperse graphene with PEK-c and a method to make
graphene/PEK-c films with great surface conductivity. The characterization of grapheme was
analyzed by scanning electron microscope (SEM) as well as chloroform without dispersant was
identified to be the suitable system to disperse graphene. Graphene/PEK-c films are fabricated in a
large scale using coating method. The conductivity of the films were tested by surface resistance tester
as well as the surface resistance of graphene/PEK-c film with 50%wt graphene is about 50/sq.
Micro cracks were observed by SEM images that were the main reasons to explain the performance
differences between real and theoretical conductivity.

Introduction
Grapheme with extraordinary high elastic modulus and excellent electrical conductivity has
attracted many scientists since it was first prepared in 2004[1]. Indeed, graphene might even replace
silicon as the electronic material of choice in the future. Large-scale preparation of paper-like
graphene films has aroused much attention for their unique mechanical and electrical properties [2-4].
Fitting graphene film with composite materials is one of the directions for future research [5].
Phenolphthalein polyether ketone (PEK-c) has good film forming characteristics are now widely used
in the aerospace industry for the excellent thermal stability and high modulus. Disperse graphene with
PEK-c in solvent and prepare the graphene/PEK-c film later can give the way to improve surface
conductivity properties of the composites by bonding the films with composites on surface,
meanwhile avoid spending a lot of cost to improve the conductive properties of resins[6]. This article
gives a way to disperse graphene with PEK-c and a method to make graphene/PEK-c film with great
surface conductivity.The films hold a great potential for applications in the fields of nanoelectronics,
sensors, transparent electrodes, supercapacitors, and nanocomposites [7].

Methods
Materials. The grapheme in the experiment was purchased from Beijing qingdajiguang Technology
Development Co., Ltd, Beijing, China. Other solvents and reagents were of analytical reagent grade
and used.
Characterization and evaluation of grapheme. Scanning electron microscope (SEM) image (Fig.1)
was taken by S-4800 microscope (Hitachi, Ltd., Tokyo, Japan). The morphology of the films was
observed from the image.
Dispersion of graphene in solvents. Obvious aggregates can be found in graphene sample so that
means we must give enough attention to its dispersion. In order to choose a suitable dispersion
solvent, we studied the dispersion of graphene in different solvents include organic solvents and
inorganic solvents. Combined with relevant literatures and functional group of graphene, three kinds
108 Metal, Ceramic and Composite Materials

of commonly used solvents were chosen, respectively are chloroform, absolute ethyl alcohol and
deionized water. Ultrasonic dispersion in aqueous solution about 30min were produced when 5ml
chloroform, absolute ethyl alcohol and deionized water were added respectively into three bottles
with 0.05g graphene sample. Three kinds of dispersants were also chosen to improve the dispersion
effect of graphene for further use. We studied the influence of different dispersants on the dispersion
of graphene after the suitable solvent was chosen. The experiments were carried out following Table 1.
Each sample was ultrasonic dispersed in aqueous solution for about 30min.
Table 1 The influence of different dispersants
on the dispersion of graphene experimental group
Serial number Graphene Dispersant Solvent
1 None
2 9076
0.05g Chloroform
3 903
4 525
Preparation of Graphene/PEK-c film. A large size graphene /PEK-c film with a certain tension can
be prepared by disperse the graphene in PEK-c which has good film forming ability. Coating method
was used to form graphene/PEK-c film after graphene and PEK-c dispersed in chloroform with
certain proportion according to Table 2. The experimental steps are as follows: mixed the system well
and put it in an ultrasonic dispersion instrument for 30min, then poured the viscous solution out of the
beaker on a smooth glass plate which had been placed on the experimental platform, cylindrical
thread roll was used to daub the solution flatness and the graphene/PEK-c film was taken off from the
glass as the solvent evaporated.

Table 2 Graphene /PEK-c films preparation experiments


Serial number Solvent PEK-c Solvent Solvent content
1 0.05g 0.95g 5%wt
2 0.05g 0.45g Chloroform 10%wt
3 0.10g 0.20g (Solution concentration 33%wt
4 0.05g 0.075g is 10%wt) 40%wt
5 0.05g 0.05g 50%wt
Characterization of Graphene/PEK-c film. Conductive performance was detected by surface
resistance tester as well as the results and theories were discussed by a series of tests such as scanning
electron microscope (SEM) image, et.al.

Results and discussion


From the SEM image (a) as shown in Figure 1, it is observed that the graphene have different forms
such as sheet and floc. Graphene layers with obvious thickness can prove the existence of multi-layer
graphene while the size of obtained graphene sheet is approximately 3-5m. Single sheets of
graphene can also be observed as the form of curls. Larger floccule graphene can be seen in image (b)
due to the agglomerate of a large number of curled graphene, therefore we should take it seriously in
later use of the graphene.

Fig.1 SEM images of graphene


Advanced Materials Research Vol. 1102 109

The aggregation of graphene sheets was observed in our experiments (bottle b and c in Fig. 2). The
graphene cant be dispersed well in ethyl alcohol and deionized water due to its strong hydrophobicity
and molecular structure. Chloroform dispersible graphene sheets were obtained at the same time,
which are well-dispersed homogeneous dispersion for more than one day (bottle a in Fig. 2). We
choose chloroform as the most suitable dispersing solvent for graphene. Then we studied the
influence of different dispersants on the dispersion of graphene use chloroform as the solvent
following Table 1. The digital picture of graphene sheets dispersed in chloroform with three
dispersants (9076, 903 and 525) was shown in Figure 3. Graphene with dispersant 9076 (bottle 2) and
with none dispersant (bottle 1) can achieve well dispersion in chloroform, while dispersant 903(bottle
3) and 525(bottle 4) can make graphene agglomerate. The dispersing effects of bottle 1 and 2 have no
obvious change after standing for one day, meanwhile taking the purity into account chloroform
without dispersant was identified to be the final system to disperse graphene.

Fig.2 Dispersion of graphene in solvents


(Solvents: (a)chloroform, (b)absolute ethyl alcohol, (c)deionized water)

1 2 3 4

Fig.3 Dispersion of graphene in chloroform with dispersant


(Dispersants: (1)None, (2)9076, (3)903, (d)525)

Graphene/PEK-c films with different graphene content were obtained as depicted in Figure 4
which showed fine distribution and uniform color of the films. Excitingly, the morphology of the
graphene films can be perfectly maintained, which suggests that this facile and novel method is
suitable for the large-scale production of graphene films. It has certain mechanical strength when the
content of graphene is no more than 40%wt, meanwhile the mechanical strength will fall rapidly
accompany with the increase of graphene. With the adding of graphene, the color of films gradually
deepens to black, but it is easily shattered when moving with the content of graphene is 50%wt like
picture (d) in Figure 4.
110 Metal, Ceramic and Composite Materials

Fig.4 Graphene/PEK-c films with different graphene content


(Graphene content: (a)5%wt, (b)10%wt, (c)33%wt, (d)40%wt, (e)50%wt)

To obtain the true value of the conductivity of the graphene/PEK-c films, we made the films
flattening and use surface resistance tester to detect resistances (graphic (a) in Figure 5). A rapid
increase in the conductivity of graphene/PEK-c films take place along with the content increase of
graphene when the conductive filler forms an infinite network of connected paths through the
insulating matrix. The test results which can be seen in graphic (b) of Figure 5 showed us the
conductive properties of the films is very poor when the graphene content are 5%wt and 10%wt, but a
certain conductivity system can achieved when graphene percentage reached 50%wt although it is far
away from theoretical value. The surface resistance of graphene/PEK-c film with 50%wt graphene is
about 50/sq.

(a) (b)
Fig.5 Surface resistance measurement of Graphene/PEK-c films

Graphene/PEK-c films (graphene percentage were 5%wt, 10%wt, 33%wt and 50%wt) were
studied by scanning electron microscope (SEM) observation showed in Figure 6. We can achieve the
conclusion from images as shown in Figure 6 that graphene can well-dispersed in graphene/PEK-c
films, as well as introducing impurities were observed by energy spectrum analysis to prove the
inexistence of obvious agglomeration of graphene in Figure 7. Banded gullies can be found in films
(graphic (3) in Figure 8) with 33%wt graphene content due to the use of cylindrical thread roll in the
roll coating process. For further analysis of this film, magnification SEM images as shown in Figure 6
was taken and graphene with low level in gullies can be seen, which do the major contribution to low
conductivity.
Advanced Materials Research Vol. 1102 111

(1) (2)

(3) (4)
Fig.6 SEM images of Graphene/PEK-c films
(Graphene content: (1)5%wt, (2)10%wt, (3)33%wt, (4)50%wt)

Fig.7 Energy spectrum analysis of agglomerations

Fig.8 SEM images of Graphene/PEK-c films with 33%wt graphene

For the graphene/PEK-c films with 50%wt graphene, the percolation threshold is, to the best of our
knowledge, about two times lower than reported. Such a high percolation threshold is evidently due to
the extremely high aspect ratio of the graphene sheets and their excellent homogeneous dispersion in
the system. As we know, graphene is the best conductivity (about 10-6 cm) material in the world,
the conductivity of graphene/PEK-c films with 50%wt graphene is far away from theoretical value.
To study the reason of this phenomenon, magnification SEM images were taken to analyze the film
carefully in Figure 9.The enlargement figure show us micro cracks spread all over the film and
percolation threshold was cutted by those cracks, it is the main reason to explain the case. But there
have obviously evidences to show us it differ from the case of graphene/PEK-c films with 33%wt
graphene. Though micro cracks exist in large quantities, we can also find a large amount of graphene
in connection part and form the pathway. Film with 50%wt graphene has better conductive properties
than film with 33%wt graphene, however the existence of micro cracks reduced mechanical
properties and do great responsible for the performance differences between real and theoretical
conductivity.
112 Metal, Ceramic and Composite Materials

Fig.9 SEM images of Graphene/PEK-c films with 50%wt graphene

Summary
In summary, the characterization of grapheme was analyzed by scanning electron microscope
(SEM) as well as chloroform without dispersant was identified to be the suitable system to disperse
graphene. We have demonstrated that graphene/PEK-c films are fabricated in a large scale using
coating method. The conductivity of the films was tested by surface resistance tester as well as the
results and theories were discussed by a series of tests such as scanning electron microscope (SEM)
image, et.al. The surface resistance of graphene/PEK-c film with 50%wt graphene is about 50/sq
which is far away from theoretical value. Micro cracks were observed by SEM images that were the
main reasons to explain the performance differences between real and theoretical conductivity. The
films hold a great potential for applications in the fields of nanoelectronics, sensors, transparent
electrodes, supercapacitors, and nanocomposites.

References
[1] Novoseiov K S, Geim A K, Morozov S V, et a1. Electric field in atomically thin carbon films, J.
Science. 306(2004):666-669.
[2] He L, Yao L, Sun J, Wu W, Yang J, et al. Insite mineralization of amorphous calcium carbonate
particles: a facile and efficient approach to fabricate poly(L-lactic acid) based hybrids, J. Polym
Degrad Stabil. 96(2011)11871193.
[3] Fuentes-Cabrera M, Rhodes B, Fowlkes J, et al. Molecular dynamics study of the dewetting of
copper on graphite and graphene: implications for nanoscale selfassembly, J. Phys Rev E.
83(2011):041603.
[4] Geim A. Graphene: status and prospects, J. Science. 324(2009):15301534.
[5] Sasha Stankovich, Dmitriy A. Dikin, Geoffrey H. B. Dommett. Graphene-based composite
materials, J. Nature. 20(2006):282-286.
[6] Rungang Gao, Nantao Hu, Zhi Yang, et al. Paper-like graphene-Ag composite films with
enhanced mechanical and electrical properties, J. Nanoscale Research Letters. 8(2013):32.
[7] Sheng-Zhen Zu, Ding Zhou, Bao-Hang Han. Supramolecular Surface Modification and
Dispersion of Graphene in Water and Organic Solvents, J. Journal of Nanoscience and
Nanotechnology. 13(2013): 946953.
Advanced Materials Research Vol 1102 (2015) pp 113-116 Submitted: 2015-02-02
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.113

Estimation for the Si-O structures based on the Homology concept


Kazuaki Nakane1,a and Istvan Halasz2, b
1
Osaka University, 1-7, Yamadaoka, Suita City, Osaka Prefecture, 565-0871, Japan
2
PQ Corporation, Research and Development Center, 280 Cedar Grove Road, Conshohocken, PA
19428 (USA)
a
k-nakane@sahs.med.osaka-u.ac.jp, bistvan.halasz@pqcorp.com

Keywords: Homology, image analysis for molecular structure, silica gel

Abstract. The first atomically resolved transmission electron microscopic (TEM) pictures of silica
gels showed visible differences between their molecular densities (MD), which might be the
molecular foundation of their distinctly different macroscopic properties, like porosity, surface area,
fragility, etc. [1, 2]. Thus, learning to control MD promises the first time to add rational chemical
design to the largely empirical synthesis methods of silica gels with desirable physical properties.
However there is no known method for the quantitative comparison of molecular densities of these
amorphous materials. We report in this paper a new numerical image-analyzing method, utilizing the
mathematical theory of homology [3], which allows to measure quantitatively the molecular density
of silica gels from their high resolution TEM pictures.

Introduction
Silica gels are widely used adsorbents, fillers, coatings, catalyst supports, etc. materials in a variety of
technologies [4, 5]. These versatile materials are also considered for a number of new optical,
electronics, pharmaceutical, and other advanced applications [6-8]. A variety of synthesis and post
synthesis methods have been empirically developed to adjust their morphology, porosity, surface
area, mechanical and thermal stability, etc. parameters to suitable values for specific uses. It is
naturally desirable to understand the underlying molecular chemistry of these methods and utilizing
this knowledge for improved control in silica polymerization.
Yet, still very little is known even about the molecular constitution of silica gels. According to the
general view they are made of indistinguishably complex polymer siloxane rings and/or chains hence
little effort has been made until very recently to elucidating real molecular structures [6, 9]. However,
in the course of the past five years or so, suddenly several laboratories initiated experimental and
computer modeling studies to explore the gelling process and gel structures at the molecular level
[10-12]. Our own research groups relied mainly on 29Si NMR, FTIR and Raman spectroscopic data
to characterize the molecular constitutions of variously made silica gels [1, 2, 13-18]. In the course of
these studies we identified distinct molecular differences between acid and base set gels polymerized
from either aqueous alkaline silicate solutions or water free tetraethyl orthosilicate, TEOS,
monomers.
Very recently we also managed to obtain atomically resolved TEM images about some of these gels,
which allowed the first time to visualize some spectroscopically predicted differences between such
siloxane polymers [1, 2]. In addition, these high resolution molecular images also revealed visible
differences in the molecular density of acid and base set gels, which provided a plausible qualitative
explanation for the different fragility and porosity of these materials. However the conscious design
of these physical properties from the density of polysiloxane structures would require quantitative
numerical correlations between the molecular densities and the macroscopically measured physical
properties. One impediment to establishing such correlation is that no numerical measure has been
known for the molecular density of silica gels. The random atomic arrangement in these amorphous
materials makes this a very complicated mathematical task.
We show in this paper that this complex task can be tackled by using a new image analysis method,
which was originally developed for detecting cancer lesion areas in microscopic tissue images [3].
114 Metal, Ceramic and Composite Materials

This method is based on the mathematical theory of homology [19] and the binarization of visualized
figures into quantitative numbers.

Experimental methods
The silica gels applied for these microscopic studies were prepared by saturating some water rinsed,
dry Amberlite XAD7HP porous raisin beads (average diameter ~ 300 m) from Rohm & Haas (today
Dow Chemical) with TEOS and submerging the saturated beads into pH ~ 1 aqueous HCl (acid set gel
T2-A) or pH ~ 11 aqueous NH4OH solution (base set gel T2-B). After water-washing and air-drying
them at 80 oC, they were calcined at 560 oC to burn out their organic content. Further details on the
synthesis and properties of these gels have been published elsewhere [17, 18].
Transmission electron microscopy (TEM) tests were carried out on a FEI Titan3 G2 60-300
instrument operating in high resolution imaging mode at 300.0 kV.
Method of the computational image analysis
Our recently developed image analysis method is based on the mathematical theory of homology,
a modern mathematical concept [19]. Homology can be used among others for quantitatively
evaluating the degree of image contacts. We employed this concept first to measure the degree of
contacts in a cancerous cell, because it was found that loss of contacts can inhibit cancer. In a similar
manner, one can see contacts between [SiO4] tetrahedra in the high resolution TEM pictures of silica
gels. It is plausible to claim that the lower is the number of contacts the lower is their atomic density.
It is also reasonable to assume that the bright spots in the TEM pictures are the oxygen atoms of
[SiO4] tetrahedra reflecting the scanning electrons. The brighter they are the closer they should be to
the sample surface, while the totally black spots are at depth to where the electrons cannot penetrate.
Thus, by capturing the number of contacts for different grey layers in the picture, we can obtain a
reliable description about the average number of contacts in the picture. This can be considered to be
representative on the bulk sample since the slice for the TEM measurement was sliced from the bulk.
When the structure of an observation area is finer than that of other area the former Betti number b1
is larger than in the later area. Consequently higher b1 corresponds to lower molecular density.
Likewise, higher b0 corresponds to higher molecular density. To calculate the Betti numbers, we
binarize the images using the CHomP [20] algoritm. Because binarized images are mathematical
objects, this transformation allows to compute the b0 and b1 values.
This method has two major advantages over other image analyses: i) it is topologically invariant
because the Betti numbers represent quantities, which do not change by continuous transformation;
this is extremely important in reliable pattern recognition on non-uniform structures like our
amorphous gels; ii) uses averaged unit area, which is critical with random patterns; with our method
the user can select any area from the picture to represent the overall material.
Results and conclusion
Fig. 1 (a) and (b) show two equally sized (~ 16 nm2) areas from the TEM pictures of the acid set
T2-A and base set T2-B sample. One can easily visualize that T2-A has a much looser, open structure
than T2-B, which presumably causes that the beads of the former gel have smooth while the latter
ones highly fragmented surfaces [2].
These pictures were binarized in 200 steps and we recorded the values at each 25 step to
distinguish the small differences in the gray scale of the pictures. The connectivity is tested after each
step. The computed numerical values of b1 and b0 at each recorded binarizing parameters are shown
in Table 1. Their graphical comparison illustrates in Fig. 2 that b0, the Betti number characterizing
the connectivity of [SiO4] tetrahedra represented by white spots in the TEM pictures, gradually
decrease as the images were taken deeper and deeper from the surface until only isolated white spots
remain at the depth from where the electron beams are not reflected any more, ie., all light gray
connections disappear from between the white spots, hence b0 remains steady.
Advanced Materials Research Vol. 1102 115

(a) (b) (c) (d)


Fig. 1 TEM pictures of (a) acid set T-2A and (b) base set T-2B silica gels. (c) (d)binarized images
(parameter=125 )

Table 1. Computed 1-dimensional Betti numbers at different digitalization steps of the TEM pictures
of samples T2-A and T2-B (see Fig. 1)

b0 Betti numbers b1 Betti numbers


Binarization step # T2-B T2-A T2-B T2-A
50 275 75 2 2
75 190 40 4 12
100 110 30 6 29
125 40 20 52 38
150 25 20 137 45
175 20 20 162 59
200 20 20 150 63

Fig. 2 Graphical illustration of changes in the [SiO4] connectivity as represented by the b0 Betti
numbers computed at various digitalization steps at different depths characterized by gray scale
changes. Comparative changes in the b1 Betty numbers characterizing the non-reflecting empty
spaces around the [SiO4] units at various depths.
The binarized pictures in Fig. 1/c and 1/d represent this final stage for the two silica gels at the 200th
binarizing step. The larger b0 values for T2-B express higher connectivity, i.e., higher molecular
density, than those for T2-A. This agrees well with our qualitative visual observation. Fig. 2 also
illustrates that in contrast to b0, the b1 Betti numbers were low at the beginning of binarization and
substantially increase as the connectivity of [SiO4] tetrahedra (white spots) is decreasing. The lower
b1 of T2-A is also in agreement with its larger molecular density. Note that the numerical values of b0
and b1 were independently calculated from each other, thus they independently quantify the [SiO4]
connectivity in these amorphous structures. As Fig. 2 illustrates, these two Betti numbers cross each
116 Metal, Ceramic and Composite Materials

other near the half of the digitalization steps. As Fig. 1/e and 1/f indicate, the corresponding
digitalized images represent quite well the corresponding TEM pictures (1/a and 1/b). Hence it is
reasonable to select these Betti numbers to represent an average molecular density for such
amorphous materials.
Thus the measure of molecular density of the acid set T2-A sample is 27 b0 unit/16 nm2 = 1.69 b0/nm2
and that for the base set T2-B sample is 55 b0 unit/16 nm2 = 3.44 b0/nm2. Computing similar values
for many more silica gels will allow the quantitative correlation between their molecular density
characterized structures and the macroscopically measured physical parameters.

Acknowledgments
This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research (B) Grant
Number 26310209.

References
[1] Halasz, I., Kierys, A., Goworek, J., J. Colloid and Interface Science, 2015, 441, 65.
[2] Kierys, A., Rawski, M., Goworek, J., Halasz, I., Advanced Materials Letters 2015, 6, 40-46.
[3] Nakane K, Tsuchihashi Y and N. Matsuura; A Simple Mathematical Model Utilizing a
Topological Invariant for Automatic Detection of Tumor Areas in Digital Tissue Images.
Diagnostic Pathology 2013, 8 (Suppl 1).
[4] Halasz, I. (editor); Silica and Silicates in Modern Catalysis; Transworld Research Network,
Kerala, India, 2010.
[5] Patterson, R. E. Surfactant Sci. Ser. 2006, 131, 779.
[6] Wu, Y.; Xue, P.; Kang, Y.; M. Hui; K. M. Analytical Chemistry 2013, 85, 3166.
[7] Morais, E. C.; Correa, G. G.; Brambilla, R.; Radtke, C.; Baibich, I. M.; dos Santos, J. H. Z.
Colloids and Surfaces, B: Biointerfaces 2013, 103, 422.
[8] Warren, S. C.; Perkins, M. R.; Adams, A. M.; Kamperman, M.; Burns, A. A.; Arora, H.; Herz, E.;
Teeraporn Suteewong, T.; Sai, H.; Li, Z.; Werner, J.; Song, J.; Werner-Zwanziger, U.; Zwanziger,
J. W.; Grtzel, M.; DiSalvo, F. J.; Wiesner, U. Nature Materials 2012, 1, 460.
[9] J. Livage, J., Stud. Surf. Sci. Catal. 1994, 85, 1-42.
[10] Sandoval-Diaz, L. E., Coy-Barrera, E. D., Trujillo, C. A., React. Kinet. Mech. Cat. 2012, 105,
335.
[11] Norstrm, J., Nilsson, E., P. Jarvol, M. Nayeri, A. Palmqvist, J. Bergenholtz, A. Matic, J. Colloid
Interface Sci. 2011, 356, 37-45.
[12] Hould, N. D., Foster, A., Lobo,R. F. Microporous Mesoporous Materials 2011, 142, 104-115.
[13] I. Halasz, M. Agarwal, R. Li, N. Miller, in Characterization of Porous Solids, Ed. by S. Kaskel, P.
L. Llewellyn, F. Rodriguez-Reinoso, N. A. Seaton, RSC Publishing, Cambridge, UK, 2009, 318,
416-423.
[14] I. Halasz, M. Agarwal, R. E. Patterson, Stud. Surf. Sci. Catal. 2010, 175, 209-216.
[15] A. Kierys, M. Dziadosz, J. Goworek, J. Colloid Interf. Sci. 2010, 349, 361-365.
[16] R. Zaleski, A. Kierys, M. Grochowicz, M. Dziadosz, J. Goworek, J. Colloid Interf. Sci., 2011,
358, 268-276.
[17] Halasz, I., Kierys, A., Goworek, J., Liu, H., Patterson, R. E., J. Phys. Chem. C 2011, 115, 24788.
[18] Zaleski, R., Kierys, A., Dziadosz, M., Goworek, J., Halasz, I., RSC Advances 2012, 2, 3729.
[19] Hibi, T., Commutative Algebra and Combinatorial Logic. Springer-Verlag, Tokyo, Inc, 1995.
[20] CHomP [http://chomp.rutgers.edu/projects/]
CHAPTER 4:

Mechanical Behavior and Characterization of


Materials, Technologies in Materials Engineering
Advanced Materials Research Vol 1102 (2015) pp 119-123 Submitted: 2014-12-25
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.119

Penetration Performance of Shaped Charge Jets using Insensitive


Munition: Simulative and Experimental Studies
Mukhtar Ahmed1, a, Zheng-xiang Huang2, b, A.Q.Malik2, b, Xu-dong ZU2, b
Qiang-qiang Xiao2, b, Munir Ahmed2, b
1
National University of Science and Technology, NUST, Islamabad, Pakistan
2
Nanjing University of Science and Technology, NJUST, Nanjing-Jiangsu, China
1, 2
Present Address: School of mechanical Engineering, Nanjing University of Science and
Technology, Nanjing-Jiangsu, China
a
mag_jee@yahoo.com, baqmalik46@yahoo.com, bhuangyue@njust.edu.cn,
b
zuxudong9902@njust.edu.cn, bxiao_wawj@njust.edu.cn, bmunirjt@yahoo.com

Key Words: Shaped Charge, Liner, Jet formation, Jet Penetration, Penetration Depth, and Flash
X-Ray Radiography.

Abstract. The simulation and experimental studies of the Jets obtained from different conical liners
from a charge having 56mm-diameter is presented in this paper. An OFHC copper liner of thickness
1 mm and 54 mm-diameters is used for jet formation analysis. Charge ignition is made by using
Point detonation method. The simulation results are obtained from Ansysis Autodyn 2-D hydrocode.
Experiments are conducted for a charge having 600 conical liner. Flash X-ray Radiography is used
for image capturing. The performance capacity of standard shaped charge is predicted by its
penetration against steel target. It is observed that kinetic energy of the jet decreases as angle
increases, also total energy of the explosive increases and hence the Ratio of the K.E of the jet to
the total energy of the explosive decreases as the angle increases. Simulation results are compared
with the experimental results. It is observed that simulation results are in favorable agreement
within acceptable range of error with the experimental results.

Introduction:
Shaped Charge is one of the most deadly devices in armour weapon technology that has better
penetration potential against armoured targets [1]. When jet is much less compressible than the
target, the penetration of the jet is strongly affected by the compressibility of the target [2-3]. The
incompressible Hydrodynamic Penetration theory over predicts the penetration of jets into water [4,
5]. From simulations it is predicted that jet tip velocities of copper remains same for jet formation
and penetration in air [6]. The swift development of computer technology has made it possible to
evaluate huge stresses and strains which arise during jet formation. Such techniques are utilized
when a large number of parameters involved affect the results. The explicit dynamic Hydrocode
AUTODYN predicts the parametric changes occurring during the jet penetrate mechanism and
shows graphically the jet penetration mechanism. The main focus is to achieve the longest, non
fragmented high energy jet to carry, maximum possible kinetic energy at the target. Due to high
velocity impact the shaped charge jets are the most lethal bullying to tanks and other armored
vehicles, as the penetration depth of the jet can overcome the armor thickness of these vehicles [7].
In the present paper we consider, jet formation and its penetration against steel target at a standoff
distance of 80 mm. Experiments are performed using high speed diagnostic techniques. Simulation
and experimental results are compared.
120 Metal, Ceramic and Composite Materials

Shaped Charge Numerical Simulations


The behavior of the copper jet formed into air and its penetration through steel target can be
simulated with AUTODYN software [8]. The Charge diameter is kept 56mm for simplification of
the protection calculation and cost factor. The shaped charge consists of a copper liner of diameter
54mm with a thickness of 1mm. A 2-dimensional axial symmetric multi-material Euler solver is
selected for explosive and liner. The steel target is modeled by using Lagrange solver. To overcome
the interaction problem between Euler materials and Lagrange material, an Euler/Lagrange coupling
with automatic (polygon type) is used. The geometry of the explosive (8701) is 70 mm length and
56 mm in diameter. Along x-axis, the mesh size is 0.2 mm. For y-axis the size of mesh for 0-20 mm
is 0.2 and for 20-40 mm it is 0.4mm. The free space is filled up with still air having density
0.001225 mg/mm3 and internal energy 2.0664105 micro-joules. The charge is initiated by using
point detonation method. The AutoCAD drawing and the cross- sectional view of shaped charge
used in simulation is shown in figure 1.
70 Explosive
60

56

Liner

Fig. 1 Design of Standard Shaped Charge and cross-sectional view of Shaped Charge used in the Simulation.
The model consists of material constitutive model, strength model and Equation of state. The
material models for explosive, liner, target and air are listed in Table 1.
Table 1. Material Model
Part Name Material Equation of State Strength Model
Explosive 8701 JWL None
Liner Cu-OFHC Gruneisen Steinburg Guinan
Target Steel-45 Shock Johnson Cook
Still Air Air Ideal-Gas None
The 45-Steel is used as target material with geometric strain and with an erosion strain factor of
2.0. The explosive 8701 is modeled as a high explosive by using Jones Wilkins Lee Equation of
State, which in most generalized form is given by Equation 1
w R 1V w R 2V wE
P = A 1 e + B 1 e + (1)
R 1V R 2V V

where A, B, R1, R2 and w are the input parameters, P, E and V represent the pressure released by
explosive element, the internal energy per unit volume and the relative volume respectively. The CJ
parameters of 8701 are density, =1.7 g/cm3, Velocity of detonation D = 8.3 km/sec and C-J
pressure PCJ = 30 GPa. The other parameters 8701 explosive for JWL equation of state are listed in
Table 2. The Gruneisen equation of state is selected to predict the response of liner under the action
of detonation wave. The material parameters of OFHC-Copper for Gruneisen equation of state are
listed in table 3. The target is made of 45-steel. The material parameters of shock equation of state
used for target are listed in table 4.
Table 2. Material Parameters for 8701 (Explosive) in JWL equation of State
A[GPa] B[GPa] R1 R2 w E
852 18.0 4.6 1.3 0.3 8.5e+06
Advanced Materials Research Vol. 1102 121

Table 3. Shock Parameters for OFHC-Copper (Liner)


[g/cm3] G C1[Km/sec] C2 S1 S2

8.93 2.02 3.94 0.0 1.489 0.0


Table 4. Material Parameters for 45-Steel (Target)
[g/cm3] G C1[Km/sec] C2 S1 S2
7.83 2.17 4.56 0.0 1.33 0.0

Simulation Results
Simulations are performed by using Ansysis Autodyn 12.1. The jet tip velocity and Energy ratio
for liners with angle from 48o-72o with a step of 6o is predicted. The behavior of jet tip velocity and
energy ratio ( K.E of the jet/T.E of the explosive) with increasing obliquity is shown in figures
3(a ,b). The jet tip velocity measured after 30 microseconds is 6.9 km/sec.

Fig. 3(a) Jet tip velocity analysis with increasing angle Fig. 3(b) Energy Ratio analysis with increasing angle

Experimental Set up
Flash X-Ray Radiography (FXR) setup
To radiograph the jet formation and penetration of the jet in the target, FXR experiment method
was used. The shaped charge was hanged vertically and at a standoff distance of 80mm. The shaped
charge and other armor were hanged by a cotton rope. Passive detonation technology was utilized in
the experiments to visualize the jet deformation accurately and to avoid the discrepancies
attributable to detonator. Two 450 KV flash x-ray tubes were installed for registration of the jet
after interaction. The protected cassette films were positioned to capture the fragments after
interaction. The pictorial representation and schematic of the FXR setup is shown in figure 4(a,b)

Fig. 4(a) FXR Experimental Setup Fig. 4(b) FXR Schematic Setup
0
The jet emerging from a shaped charge at 60 conical liner is investigated. The jet tip velocity and
jet formation parameters were obtained from AUDTDYN-2D finite difference Eulerian code. In
order to verify the simulated results, a series of experiments were performed to calculate the jet
velocity. The formation of jet is x-rayed using FXR technique. The jet tip velocity calculated from
122 Metal, Ceramic and Composite Materials

data obtained by using FXR technique is 6.51 km/sec. The results obtained from FXR technique are
shown in figure 5(a,b).

Fig. 5(a) The FXR image for jet penetration after 30s Fig. 5(b) The FXR image for jet penetration after 50s
zt 562.32 301.92
v jt = = =6.51 m m s
k t 2 ( 50 30 )
zw 1 2 3 .9 0 7 6 .3 2
v jw = = = 1 .1 8 9 5 m m s
k t 2 (5 0 3 0 )
Experimental set up for Depth of Penetration (DOP)
The penetration performance of the shaped is determined by its penetration against target made of
45-steel. The charge is placed at a standoff distance of 80 mm from the target. The target is made of
three cylindrical blocks having 100 mm height and 100 mm diameter.. The experimental set up for
the Depth of penetration before experiment and shape of the penetrated target after the experiment
is shown in figure 6 (a ,b).

Fig. 6(a) Experimental set up for DOP (Before experiment) Fig. 6(b) Shape of the target (After Experiment)
The holes produced by the shaped charge jet were measured. The shape of the hole at the
entrance and exit of three targets is shown in figure 7 (a,b,c,d,e) respectively.

(a) (b) (c) (d) (e)


Fig. 7 The hole of steel targe:(a). First target entrance hole (35mm), (b) First target exit hole(10mm),
(c)Second target entrance hole (10mm), (d) Second target exit hole ( 5m), (e) the hole of the third target (
1mm)

Results and Discussion


The finite difference model of the shaped charge is based on Eulerian frame work.
Euler/Lagrange coupling is used between Euler materials and Lagrange materials. The simulation
and numerical results for jet tip velocity and depth of penetration are in good agreement with the
Advanced Materials Research Vol. 1102 123

experimental results. The simulation model and simulation results can be used to characterize the jet
formation mechanism. From simulation it is observed that the jet tip velocity decreases with
increase of liner obliquity. The same behavior is observed for the ratio between the kinetic energy of
the jet and total energy of the explosive. The jet velocity calculated from simulation is a bit higher
than measured from experiments, but it is still within acceptable range of error.

Conclusion
Jet tip velocity and energy ratio decreases with increase of liner angle. When angle is greater than
60o, the energy ratio and Jet velocity decrease slowly. Therefore the angle should be less than 60o
for the shaped charge to have maximum depth of penetration for armour application. The shaped
charge presented in this paper can be used for metal cutting, cutting holes in the barriers for fire
services. Also it can be utilized as HEAT round for penetration against armored vehicle, bunkers,
rocks, tunneling and as a useful device for penetration in weapon technology. Being light weight
shaped charges have proved most lethal device in the main battlefield as well.

References:
[1] Xu-dong Zu, Zheng-xiang Huang, Wen Zhai, Xin Jia, Performance of Kevlar Fiber-Reinforced
Rubber Composite Armour against Shaped-Charge Jet Penetration, Latin American Journal of
Solids and Structures, Vol. 11 (2014).
[2] Walter, W.P., Zukas, J.A., Fundamentals of Shaped Charges, Wiley, New York, 1989.
[3] Held, M., Backofen, J.E., Penetration of Shaped Charge into water, Proceedings of the 12th
International Symposium on Ballistics, San Antonio, Texas. Vol. II (1990) 30-40.
[4] Chick, M.C., Frey, R.B., Trimble, J.J., Bines, A., Jet Penetration in Plexiglas, Proceedings of
8th International Symposium on Ballistics, Orlando, Florida. Vol. II (1994) 21-26.
[5] Lee, E.S., Oh, K.H, Song, S.Y., Penetration of Particulated Shaped Charge Jet into Water,
Proceedings of SPIE, High Speed Photography and Photonics, 21st International Congress, Ung
Kim. Vol. 2513 (1995) 975-981.
[6] Lee, S., Kim, E.P., Kim, Y., Lee, S.H., Hong, M.H., Noh, J.W., Penetration Performance of
Tungsten-Copper Shaped Charge Liner, Proceedings of 22nd International Symposium on
Ballistics, Vancouver BC, Canada. (2005) 437-443.
[7] Xin Jia, Zheng-Xiang Huang, Xu-Dong Zu, Xiao-Hui Gu, Qiang-qiang Xiao, Theoretical
Analysis the Disturbance of the Shaped charge Jet Penetrating a Rubber Fabric Woven
composite, International Journal of Impact Engineering. 65 (2014) 69-78.
[8] Xiangrong Zhang, Cheng Wu, Fenglei Huang, Penetration of Shaped Charge Jets with
Tungsten-Copper and Copper Liners at Same Explosive-to-Liner mass Ratio in Water, Shock
Waves. 20 (2010) 263-267.
[9] AUTODYN Explicit Software for Nonlinear Dynamics, v12.1 (2011).
[10] Jian-ping FU, Zhi-gang CHEN, Xiu-cheng HOU, Shu-qiang Li, Shou-cang Li,Jing-wen
WANG, Simulation and Experimental Investigation of Jetting Penetrator Charge at Large
Stand-off Distance, Defence Technology 9 (2013) 91-97.
Advanced Materials Research Vol 1102 (2015) pp 125-128 Submitted: 2015-01-07
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.125

Encapsulation of Lanthanide -diketone Hybrids in Organic Oligomer

Poly(ethylene glycol) 400

HuaTi Li1, 2, a; Dong Liu1; Ying Li1, b * ; JieLin Wang1 ,1,2 Xia Wang1
1
School of Material Science and Engineering, University of Shanghai for Science and Technology,
Shanghai 200093, P. R. China
2
School of Environment and Architecture, University of Shanghai for Science and Technology,
Shanghai 200093, P. R. China
a
1464533674@qq.com, bliying@usst.edu.cn

Keywords: luminescence, hybrid polymeric material, Eu3+\ Tb3+, TTA

Abstract
The organic ligand 2-Thenoyltrifluoroacetone (TTA) which was selected as molecular bridge for
sensitization of europium and terbium ions were modified by 3-(triethoxysilyl)-propylisocyanate
(TEPIC) to obtain the hybrid precursor TTA-Si. Then the solution of Eu(NO3)3 and Tb(NO3)3 were
added in the presence of tetraethylorthosilicate (TEOS). The binary lanthanide organic/inorganic
hybrid material were obtained. The ternary hybrid materials were obtained by adding the solution of
Eu(NO3)3 and Tb(NO3)3 and Poly (ethylene glycol) (PEG400). We investigated the thermal stability
and luminescence properties of hybrids and found that the ternary hybrid materials exhibit better
thermal stability and stronger emission intensity. Furthermore, compared with the binary
mesoporous material Eu-(TTA-Si)3 and Tb-(TTA-Si)3, the ternary mesoporous material
Eu-(TTA-Si)3-PEG and Tb-(TTA-Si)3-PEG exhibits the characteristic emission of the Eu3+ and
Tb3+ ion with a higher luminescence intensity, suggesting that the introduction of PEG is of benefit
for the sensitization of Eu3+ and Tb3+ luminescence, by replacing H2O groups that can quench the
luminescence of Eu3+ and Tb3+ ions.

Introduction
In recent years, the interests in photoluminescence properties of rare earth compounds have been
fascinating researched.[1]What we have known was lanthanide hybrid materials have superior
mechanical properties than the pure molecular lanthanide complexes. In some circumstances, the
color purity and emission quantum efficiency could be increased remarkable because the light
energy transferred more efficiently between the ligand and the center lanthanide ions. [2]According
to the literature, the trivalent lanthanide series are characterized by a gradual filling of the 4f
orbitals. Because the partially filled 4f shell is well shielded from its environment by the closed 5s2
and 5p6 shells, the lanthanide luminescence have specific properties. [3] When the ligand lowest
triplet energy level is higher than the central ion energy level in the complex, it occur energy of
lowest triplet energy in ligand transfer to rare earth ions low energy level intramolecular to produce
central ion characteristics fluorescence. [4]Thereby certain organic ligands such as -diketones ,
heterocyclic, and aromatic carboxylic acids derivatives are usually used to coordinate with the rare
earth ions from vibrational coupling and increase the light absorption cross section by the antenna
effect. [5]As we know,certain polymers containing long organic chains can remedy the defect as
the second ligand. [6] Whats more, the second ligand also has some advantages, on the one hand,
126 Metal, Ceramic and Composite Materials

the second ligand expand the scope of the conjugated bond so that it is conducive to the energy
transfer; on the other hand, Binary complexes contain the coordinated water will produce
fluorescence quenching. The second ligand may be diverted position of the coordinated water
molecule. It is proved that the hybrid materials with polymers possess high molecular weight have
excellent effective properties. [7]
In this paper, we report on the lanthanide ions Eu3+ and Tb3+ coordinated with the inorganic
polymeric chains(Si-O-Si) through the modification of the 2-Thenoyltrifluoroacetone (TTA) with
3-(triethoxysilyl)propyl isocyanate (TEPIC). In addition, polyethylene glycol (PEG 400) is selected
as the second ligand after the early theoretical study and calculation. For the luminescence
properties and the thermal stability, we compared the binary hybrid Eu-(TTA-Si)3 and
Tb-(TTA-Si)3 with the ternary hybrid Eu-(TTA-Si)3-PEG and Tb-(TTA-Si)3-PEG in detail. The
results indicate both the photophysical properties and the thermal stability were improved markedly.
Experimental Section
2.Chemicals
Tetraethoxysilane (TEOS, Aldrich), and 3-(triethoxysilyl)-propy-lisocyanate ( TEPIC, Aldrich)
were used as received. RE (RE=europium and terbium) chlorate ethanol solution as obtained by
dissolving Eu2O3 and Tb4O7 into concentrated hydrochloric acid. Tetrahydrofuran (THF) was used
after desiccation with anhydrous calcium chloride.The two ligands 2-Thenoyltrifluoroacetone (TTA)
and poly(ethylene glycol) (PEG 400 for the polymer with the molecular weight about 380-430 )
were purchased from Sinopharm Goup Co. Ltd. All other reagent are analytically pure.

2.2 Synthetic Procedures.


2.2.1Synthesis of the precursor TTA-Si
2-Thenoyltrifluoroacetone (TTA)(1 mmol,0.222g) was dissolved in 20 ml of Tetrahydrofuran
(THF) completely under nitrogen atmosphere purging. and 2mmol NaH(0.046g) was added into the
solution at the same time stirring at temperature 65 . After 2 h reflux reaction, 2 mmol
3-(triethoxysilyl)-propyl-lisocyanate (TEPIC)(0.2214g) dissolved in THF was dropwised into the
solution. The mixture was refluxed and heated at 65 in a covered ask for approximately 12 h at
the nitrogen atmosphere. The solvent was removed in revolved and evaporated to remove the
solvent. The residue gave a light yellow viscous liquid.

2.2.2Synthesis of Binary Hybrid Material Eu-TTA-Si and Tb-TTA-Si .


The precursors TTA-Si was dissolved in 15 ml N,N-dimethyl formamide (DMF) solvent, and
while being stirred, europium and terbium chlorate ethanol solution was dropwised into solution.
After 2h, an appropriate amount of tetraethoxysilane (TEOS) was added to the mixed solution. To
promote the hydrolysis reaction, a drop of hydrochloric acid wad dropped in. The molar ratio of
Eu(NO3)36H2O /TTA-Si/TEOS/H2O = 1: 3: 6: 24. After stirred the mixture at room temperature for
6h. The predicted structure of Eu-TTA-Si and Tb-TTA-Si were obtained as outlined in Figure 1.

2.2.3Synthesis of Ternary Hybrid Material Eu(TTA-Si)3PEG and Tb(TTA-Si)3PEG


The synthesis procedure for Eu(TTA-Si)3PEG and Tb(TTA-Si)3PEG were similar to that of
Eu-TTA-Si and Tb-TTA-Si except that the mixed ethanol solution of Eu(NO3)3 and Tb(NO3)3
were replaced by Eu(NO3)3 and Tb(NO3)3 and PEG. The resulting material was dried at 60 in the
vacuum overnight.
Advanced Materials Research Vol. 1102 127

3.Results and Discussion


3.1 Fourier Transform Infrared Spectra
The IR spectra of the 2-Thenoyltrifluoroacetone (TTA), the precursor TTA-Si, and
3-(triethoxysilyl)-propy-lisocyanate (TEPIC) are shown in Figure1. In the spectra of TTA-Si, it can
be seen that the sharp peak at 3028 cm-1 of TTA had be replaced by the strong peaks which
corresponds to three methylene groups of TEPIC at 2974 cm-1 of TTA-Si. In addition, in spectra of
TTA-Si, the characteristic peaks of Si-C are observer at 1185 cm-1 and Si-O's is at 1063 cm-1.
Furthermore, the band centered at 3371 cm-1 corresponds to the stretching vibration of -NH- groups.
Peak at 1637 cm-1 is attributed to the absorption of the -CONH- group. New peaks at 1713 cm-1 and
1637 cm-1 are the absorption peak of C=O. All of these indicate that TEPIC has been successfully
grafted onto TTA.

1524 1185 1063


2993 1637
1713
3371
2268
T T A -S i
Absorbence%

3399
3028
TTA

2974 2287
T E P IC

4000 3500 3000 2500 2000 1500 1000 500


-1
W avelen gth (cm )

Fig 1. IR spectra for TTA, TEPIC, and hybrid material TTA-Si


3.2 Thermogravimetric Analysis.
Thermogtavimetry trace of ternary material Eu(TTA-Si)3PEG is shown in Figure 2. From the
Figure we can observe that the sample shows three main steps in weight loss. The first step of
weight loss is about 12.1% for Eu(TAA-Si)3PEG from 0-250 . It can be due to desorption of
physically adsorbed water and residual solvent, without any decomposition of the chemical bonds.
The second step of weight loss (about 26.2%) between 250 and 450 is attributed to the first
decomposition of organic ingredients, it is the loss of TTA and PEG. Finally, the weight loss from
450 to 650 is assigned to the further decomposition of organic ingredients (TTA andPEG).
From the TGA curve of the hybrid system, it can be concluded that the hybrid material has
relatively thermal stability from room temperature to 250 .

100

1
90

80
2
70
Weight %

60
3 Eu(TTA-Si)3PEG
50

40

30

20
100 200 300 400 500 600 700 800

Temperature ()

Fig 2. Thermogravimetry trace (TG) of Eu(TTA-Si)3PEG


128 Metal, Ceramic and Composite Materials

3.3 Photoluminescence Properties


Fluorescence properties of hybrid materials containing Eu3+ Figure 3 shows excitation and
emission spectra of binary and ternary hybrid material. The excitation spectra of these materials
were all obtained by monitoring the strongest emission wavelength of the Eu3+ at 613 nm. In the
excitation spectra, the broad absorption band can be observed at 250-350nm which is attributed to
the *electronic transition in ligand. Absorption peak of binary hybrid material Eu-(TTA-Si)3 is
located at 374nm. The ternarys peak is located at 325nm.
In the emission spectra, characteristic Eu3+ emissions are observed. It is clearly seen that
peaks assigned to the 5D07FJ (J=0, 1, 2, 3, 4) transitions at 578, 590, 613, 650 and 700 nm
between 450nm and 700nm. Among these emission peaks of the hybrids, the most striking red
fluorescence signals (5D07F2) of the induced electric dipole transition at about 613 nm are all
stronger than the orange emission intensities of the magnetic dipole transition. As a result, the
efficient intermolecular energy transfer between the organic ligands and the central Eu3+ ion has
occurred.

EX 325 Eu-(TTA-Si)3 5 7
Eu-(TTA-Si)3-PEG
EM D0- F2 Eu-(TTA-Si)3-PEG
Eu-(TTA-Si)3
374
Relative Intensity / a.u.

Eu(TTA-Si)3PEG
Relative Intensity / a.u.

Eu-(TTA-Si)3-PEG

Eu(TTA-Si)3

5 7
D0- F1

5 7 5 7
5
D0- F0
7 D0- F3 D0- F4

Eu-TTA-Si

250 300 350 400 450


450 500 550 600 650 700 750 800
Wavelength / nm Wavelength / nm

Fig 3. Excitation and emission spectra for of Eu-TTA-Si and Eu(TTA-Si)3PEG


Acknowledgements
This work was supported by the National Natural Science Foundation of China (21101107,
51173107), the Hujiang Foundation of China (B14006).

References
[1]X. N. Fan ,G. L. Xiahou,Synthesis,Characterization and Anitumor Activity of Rare Earth Eu3+
and Dy3+ Complexes with Quercetin, Biomedical Engineering and biotechnology(iCBEB),
48(2012)567-569.
[2] K.Binnemans,Lanthanide-Based Luminescent Hybrid Materials.Chem.Rev, 109,4283 (2009)
[3]O. V. Borshchev , S. A. Ponomarenko, Branched oligothiophene silanes with the efficient
nonradiative energy transfer between the fragments,Russian Chemical Bulletin. 59(2010) 797-805
[4]H. Lujan-Upton ,Y. Okamoto ,Use of Terbium(III) as a Probe for the Ion-Binding Properties of
Tactic Polyacids and Triacidic Model Compounds.journal of Fluorescence, 8(1998)355-362
[5] V. Bekiari , P. Lianos ,Strongly Luminescent Poly(ethylene glycol) -2,2-bipyridine Lanthanide
Ion Complexes, AdV. Mater., 10(1998) 1455
[6] H. Y.Chen , R. D. Archer ,Synthesis and Characterization of Linear Luminescent Schiff-Base
Polyelectrolytes with Europium(III) in the Backbone , Macromolecules, 29 (1996) 1957
[7] X. F. Qiao , B. Yan, Assembly, Characterization, and Photoluminescence of Hybrids Containing
Europium(III) Complexes Covalently Bonded to Inorganic SiO Networks/Organic Polymers by
Modified -Diketone, J. Phys. Chem. B, 112(2008) 14742-14750.
Advanced Materials Research Vol 1102 (2015) pp 129-134 Submitted: 2015-01-14
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.129

Relationship between solid lubricant layer and friction coefficient of PPS


races-PTFE retainer hybrid thrust bearings under dry condition
Xiaochen SHI1,a, Koshiro MIZOBE1,b,
Yuji KASHIMA2,c, and Katsuyuki KIDA1,d
1
University of Toyama, 3190, Gofuku, Toyama, 930-8555, Japan.
2
Kashima Plastic Bearings Co., 2-9-1 Himesato, Nishiyodogawa-ku, Osaka, 555-0025, Japan.
a
d1378001@ems.u-toyama.ac.jp, bkmizobe@eng.u-toyama.ac.jp,
c
mail@kashima-kagaku.com, dkida@eng.u-toyama.ac.jp.

Key words: PPS, PTFE, polymer bearing, rolling contact fatigue, friction coefficient, EPMA.

Abstract: Due to the advantage on low coefficient and self-lubricant, PPS races-PTFE retainer
hybrid polymer bearings can offer better service performance under certain condition. In our
previous study, the relationship between friction coefficient and rotation speed was investigated,
and formation of black layer was found. In the present study, EPMA was introduced to investigate
the component and thickness of the black layers. Based on the observations, the relationship
between friction coefficient, solid lubricant layer and thrust load was also discussed.

Introduction
With the development of high polymer chemistry, polymer bearings have been widely used for
application where corrosion, chemistry resistance and non-lubricant (dry) are required. Many
studies on polymer bearings have been published nowadays. However, it is still hard to estimate
service life of polymer bearings because of failure theory which is different from metal bearings.
Under water-lubricant condition, it was found that flaking failure occurred due to surface crack
propagation. However, under dry condition, the main reason of failure was found to be wear loss,
which is related to friction.
In our previous study [1] about hybrid PPS bearings under dry condition, relationship between
friction coefficient and rotation speed was investigated. From RCF (Rolling Contact Fatigue) tests
under a thrust load of 400N, a preliminary conclusion that friction coefficient decreases with
rotation speed was made. At the same time, formation of black layer, generating from PTFE, was
found on the surface of PPS races after the test. It was believed the decrease of friction coefficient is
connected with the black layer.
In the present study, observations of the black layers were made using an EPMA (Electro Probe
Micro-analyzer) to investigate component and thickness of the layers. The friendship between
friction coefficient and thrust load was also discussed in this paper briefly.
130 Metal, Ceramic and Composite Materials

Test procedure
Material. For the present study, races of hybrid PPS bearings were made of unreinforced PPS
(TECHTRON PPS), and retainers were made of composite PTFE (Du Pont-Mitsui
Fluorochemicals 1191-J). PPS (Polyphenylenesulfide) is one of semi-crystalline thermoplastic
polymers with excellent mechanical properties. PTFE (Polytetrafluoroethylene) is well known as
Teflon by DuPont Co. Composite PTFE contains 75 % pure PTFE and 25% carbon graphite. The
chemical structural formulas of materials are shown in Fig. 1. The material mechanical properties
are shown in Table 1.
Specimen. As shown in Fig. 2, bearing sets used in this study consisted of two races and a
retainer sandwiched between them. As there is no specific design standard for polymer bearings, the
geometry of the bearing specimens was designed based on the JIS B 1513 #51305 standard for
metal bearings. The outer diameter was 52mm, the inner diameter was 25mm and the pitch circle
diameter was 38.5mm. In each bearing retainer, nine 9.525 mm-diameter Al2O3 balls were held.
RCF tests. The RCF tests were conducted using a thrust type machine shown in Fig. 3a. The
setting of the bearing is shown in Fig. 3b. The rotation speeds of the tests were 450, 500, 550 and
600 rpm. The thrust loads were 400N, 450N and 500N. The total number of rotating cycles was
2.88105. The number of 2.88105 was defined according to 8 hours at 600 rpm.
Friction coefficient measurement. As shown in Fig. 3, thrust load L (N) on bearing holder is
applied through a lever, at the end of which weights are introduced. The bearing rotates by a driving
motor. Load cell was used to measure the driving force F (N). The calculation of friction coefficient
was done by using the following equation based on the equilibrium of forces:
=(FR)/(Lr) (1)
Here R is the distance from center of bearing to the load cell (mm), and r is the pitch radius
(mm).
Observation of solid lubricant layer. Two methods were introduced to make observations of
solid lubricant layers in the test. In the first method, friction coefficient was measured during the
test, and tests were paused manually at an increment of measurement cycle or a failure occurs, then
photos of top races were taken. In the other observation method, an EPMA (JEOL JXA8230,
Japan) was introduced. The sections of the top bearing races were prepared after the tests at cycles
of 2.88105. Using the EPMA, component and thickness of solid lubricant layers were investigated.

Table 1. Material properties of PPS and PTFE.


Melting Glass transition Specific Tensile Bending Bending elastic
point, point, density, strength, strength, modulus,
3
C C g/cm MPa MPa MPa
PPS 278 90 1.35 83 142 3900
PTFE 327 130 2.10 18 9.41 1166
Advanced Materials Research Vol. 1102 131

Fig. 1 Chemical structural formulas of PTFE and PPS.

Fig. 2 PPS race and PTFE retainer.

(a) (b) (c)


Fig. 3 (a) RCF test machine, (b) Specimen setting on test machine,
(c) Schematic illustration of friction coefficient measurement system (top view).
Results
Thrust load and friction coefficient. The relationship among friction coefficient, rotation speed
and thrust load is shown in Fig 4. It was found that around 105 cycles, friction coefficient changed
irregularly, then they stabilized. In the test under a thrust load of 450N, with the increase of rotation
speed, friction coefficient decreased. The same feature was also partly observed in the tests under
thrust loads of 400N and 500N. From the observations of top race at rotation speed of 450rpm and
thrust load of 400N, solid lubricant layer was not confirmed, which is one possible reason for its
lower friction coefficient. And it also shows the significant effect of solid lubricant during wear
process. In the tests under thrust loads of 400N and 450N, friction coefficients barely changed. In
the tests under 500N, the friction coefficient became larger than those of 400N and 450N.
Considering the data scattering, we believed that friction coefficient tends to increase with thrust
load.
132 Metal, Ceramic and Composite Materials

Observation of solid lubricant layer using an EPMA. The sections of top races under thrust
loads of 500N and 400N were observed using EPMA. Before EPMA analysis, two possible results
about the component of solid lubricant layer were considered, either PTFE or carbon. The main
elements as investigation objects were carbon, fluorine and sulphur. However, from the results of
the EPMA analysis, we found that the contents of carbon and fluorine were less than we expected.
At the same time, content of sulphur was apparently different from that of PPS substrate. In Fig. 5
and Table 2, the thicknesses of the sulphur layers and values of friction coefficient at the end of the
tests are shown. It is found that the thicker the layer, the smaller the friction coefficient is. Only in
the test under a thrust load of 500N and rotation speed of 600rpm, the feature of thickness of
sulphur layer and friction coefficient was different from others.
Observation of top races of hybrid PPS bearings during the test. From the real time
measurement of friction coefficient, we found that generally the value of friction coefficient
fluctuated from cycles of 0.5105 to 1.5105, and plateaued after. Fig. 6 shows the photos of top
races during the test under a thrust load of 450N and at rotation speed of 500rpm. It is indicated that
the black lubricant layer was formed during this period. Although we believed the increase of solid
lubricant layer thickness leads to the decrease of friction coefficient, the rapid raise of friction
coefficient appeared around cycles of 1.5105. This means the change of friction coefficient during
the period when solid lubricant layer grows is more complicated than influence of layer thickness.

Conclusion
In the present study, EPMA was introduced to investigate the component and thickness of solid
lubricant layers. The friendship between friction coefficient, solid lubricant layer and thrust load
were also discussed. While unsolved problems about the black solid lubricant remained, from the
test results, we concluded that
1. friction coefficient decreases with the rotation speed;
2. friction coefficient increases with the thrust load;
3. the thicker the sulphur layer, the smaller the friction coefficient is;
4. black layer grows when irregular change of friction coefficient.

Fig. 4 Friction coefficient of RCF tests.


Advanced Materials Research Vol. 1102 133

(a) Thrust load of 400N.

(b)Thrust load of 500N.


Fig. 5 Sulphur layer of the sections of race tracks observed by EPMA.
134 Metal, Ceramic and Composite Materials

Table 2. Relationship between thickness of sulphur and friction coefficient.


Test condition Thickness of sulphur layer (m) Friction coefficient
400N 450rpm 62.5 0.016
400N 500rpm 18.2 0.024
400N 550rpm 11.4 0.029
400N 600rpm 41.7 0.021

Test condition Thickness of sulphur layer (m) Friction coefficient


500N 450rpm 34.1 0.035
500N 500rpm 68.2 0.030
500N 550rpm 113.6 0.020
500N 600rpm 22.7 0.025

0.05

0.04
Friction coefficient

0.03

0.02

0.01

0
0 0.5 1 1.5 2 2.5 3 3.5
5
Number of cycles [10 ]

Fig. 6 Photos of top race during test under a thrust load of 450N and at rotation speed of 500rpm.

Reference
[1] Xiao Chen Shi, Koshiro Mizobe, Yuji Kashima, Katsuyuki Kida, Effect of Rotation Speeds on
Friction Coefficients of PPS Race-PTFE Retainer Hybrid Polymer Thrust Bearings under Dry
Contact, Applied Mechanics and Materials, Volume 418(2013), pp. 205-208.
[2] Koshiro Mizobe, Takashi Honda, Hitonobu Koike, Edson Costa Santos, Yuji Kashima and
Katsuyuki Kida, Effect of PTFE retainer on friction coefficient in polymer thrust bearings under dry
contact, Advanced Materials Research (Trance Tech Publications, ISSN: 1022-6680), Vol.
683(2013), pp 90-93.
[3] Koshiro Mizobe, Masahiro Takamiya, Takashi Honda, Hitonobu Koike, Edson Costa Santos,
Yuji Kashima and Katsuyuki Kida, Fourier Transform Infrared Spectroscopy for Wear Debris
Adhesion on PEEK Bearing Surface, Advanced Materials Research, Vol. 307(2013), pp 372-376.
Advanced Materials Research Vol 1102 (2015) pp 135-138 Submitted: 2015-01-16
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.135

An image analyzing method by a homology concept for fracture surfaces


Akihiro Takiyama1, a, Kazuaki Nakane2,b and Katsuyuki Kida3,c

1
Hokkaido University, Nishi 7 kita 15 Kita Ward, Sapporo, Hokkaido 060-8638, Japan
2
Osaka University, 1-7, Yamadaoka, Suita City, Osaka Prefecture, 565-0871, Japan
3
University of Toyama, 3190, Gofuku, Toyama City, Toyama Prefecture, 930-8555, Japan
a
takiyama@med.hokudai.ac.jp, bk-nakane@sahs.med.osaka-u.ac.jp,
c
kida@eng.u-toyama.ac.jp

Keywords: homology method, image analysis, fracture surface

Abstract. From early 20th, the homology have been studied as a branch of mathematics. It has been
one of a major topic in mathematics and applied in many scientific fields. The homology can be
regarded as a mathematical concept of quantifying the contact degree. Because it belongs to modern
mathematics, advanced knowledge of mathematics are required to understand this concept.
Fortunately, since the definition of the homology is very simple in two-dimensional case, we can
easily apply this to image analyzing methods.
Fracture surface is composed of a process that cracks grow. If these cracks contact one another,
rupture occurs. Progress of the fracture would vary due to the effect of external force. We identify the
rapid and the fatigue fracture surface by a homology method. By analyzing the state of fracture via a
homology method, it is believed that the cause of fracture can be estimated.

Introduction
The technology to construct skyscrapers and bridges has been improved with the modern society
development. At the same time, because a number of improvements have been made, ships and
aircrafts can now be operated in very safely. However, accidents due to fatigue failure are depriving
many lives not only to important social infrastructure. To investigate the cause of the accidents, a
research field called fractography has been developed. They pigeonhole the fracture morphology and
analyze them. In particular, by determing the rapid fracture area and the fatigue fracture area, they
estimate the cause of the accidents.
Fracture surface is, by the repetition of vibration and stress amplitude, formed in the area of plastic
deformation occurring where the stress concentration parts are expanded. Since the direction where
each crack progresses is unstable, it is quite difficult to express the expansion of the fracture surface
by differential equations. Moreover, because the fracture sarfaces are very complex, the ordinary
image analysis methods do not work, effectively. Identification of the fracture surfaces is performed
by observation of the trained technicians. In this way, the results may differ depending on technicians.
Moreover a large amount of images can not be processed. The quantification by objective indicators
that using computers is a big challenge.
In this note, we propose a new image analysis method based on the homology. By using the concept
of the homology, we quantify the contact degree of cracks and try to identify fracture images.

Homology method
The homology is a certain general procedure to associate a sequence of abelian groups with a given
figures [1-3]. Especially, it treats the properties and quantities which do not change with respect to the
continuous deformations. For example, the number of the connected components and the number of
the area surrounded by the boundary do not chage in continous transformation. To proceed this
procedure, we convert figures to simple ones, such as points, lines and triangles. We call each of them
simplex. The union of simplexes is called (simplicial) complex.
136 Metal, Ceramic and Composite Materials

Here, we explain the concept of a homology. We start by recalling some basic notions which are
suitable for numerical computation. An abstract simplicial complex is a pair (V, ), where V is a finite
set and is a set of nonempty finite subsets of V such that if X and YX then Y. We can obtain
a geometric realization of an abstract simplicial comlex, which suffices the definition of geometric
simplicial complexes. A geometric simplicial complex K in Rn is a finite set of simlexes such that
every subsimplex of a simplex of K is a simplex of K and the intersection of any two simplexes of K is,
if it is not empty, a subsimplex of each of them. We can also write a geometric simplicial complex as
an abstract one. Furthermore, we can define the notion of a geometric polyhedral complex, including
a cubical complex which is important for computational homology. Here, for simplicity, we consider
homology groups of a simplicial complex.

Let K be an n-dimensional simplicial complex and C(K) be the chain complex of K:


0 Cn(K) Cq+1(K) Cq(K) Cq-1(K) C0(K) 0.
The following two sets:
Zq(K) = Ker q = {c Cq(K) | q(c) = 0 }
Bq(K) = Im q+1 = {c Cq(K) | c' Ck+1(K), c = q+1(c') }
are subgroups of Cq(K). The quotient group:
Hq(K) = Zq(K)/Bq(K)
is the q-th homology group of simplicial complex K. This means intuitionally a set of q-simplexes
which have no boundary and are not boundaries of (q+1)-simplexes. By the fundamental theorem of
Abelian groups, the q-th homology group Hq(K) is represented by a direct sum of infinite cyclic
groups Z or cyclic groups Zi of order i :
Hq(K) Z Z Z Z1 Z2 Zs ( denotes direct sum),
where i divides i+1 for all i. The number of infinite cyclic groups Z is the q-th Betti number of K,
denoted by q or bq.

The definition of the homology seems to be very difficult, however, in two dimensional case, it is
very simple. The 0-dimensional Betti number b0 is the number of connected components and the
1-dimensional Betti number b1 is the number of the surrounded areas. Fig. 1 is a schematic
illustration. As the connection lines increase, the number of surrounded area is increased. We can see
the b1 is strongly related to the contact degree.

Fig. 1 Schematic illustration of network. The connection line is added, new surrounded area appears.
Namely, if the number of b1 is large, the connection of among the points are strong.
In this note, by counting the b1, we estimate the contact degree of unit area.

The image analysing method applying the homology concept was firstly intoroduced to the detection
of the cancer lesion area from the degital pathological images [4]. Even in a normal area, the form of
tissues are slightly different. This makes it difficult to apply image recognization methods. Instead of
recognizing the shape, we estimate the degree of cell density. Because of the loss of the contact
Advanced Materials Research Vol. 1102 137

inhibition, one of the character of cancer, the cancer areas are included in the high cell density area.
Not only to detect the lesion area, but also this method has been applied many fields [5-11].
Here, we apply this idea to images of fracture surfaces. In this case, the number of the surrounded
areas by the cracks is the 1-dimensional Betti number b1. And we calculate the Betti numbers in the
unit area.

(a) (b)
Fig. 2 Schematic illustration of fracture structures.
The number of hexagon in a unit area expresses fineness of the structures.
(a) Fine structure (b0=1, b1=45) (b) Coarse structure (b0=2, b1= 6).

(a) (b) (c)


Fig. 3 Typical fracture images.
(a) Fatigue fracture (b) Fatigue fracture with small burden (c) rapid fracture

Numrerical results
We calculate the the Betti numbers b0 and b1 for each images. To proceed our numerical
computations, we binarized the images of fracture surface. Because binarized images can be regard as
a complex, we can calculate the Betti numbers. Here we have used ChomP [12] to calculate the Betti
numbers.

Table 1 The results of numerical calculation.


(a) (b) (c)
b0 2290 2466 2144
b1 262 164 123
b1/b0 0.114 0.066 0.057

From table 1, the 1-dimensional Betti number b1 differ from one another. This implies that the
fracture situation is quantified by b1.
138 Metal, Ceramic and Composite Materials

Discussion and conclusion


The homology is a concept which belongs to modern mathematics. By replacing the figures in algebra,
we have been studying the properties of the figures via the homology. This concept has been applied
to many sicentific fiels, such as cosmic physics and quantum mechanics. Because advanced
knowledge of mathematics are requied to understand the concept of the homology, it has not applied
to the practial use. Since the homology is very simple in two-dimensional case, we can use this to
analyze the complex images such as cancer tissues and fracture surfaces.
Fracture surface would be coarse when the amplitude of vibration is large, conversely it would be fine
when it is small. By binarizing the images of fracture surface, we can calculate the Betti numbers.
From the numerical results, we can quantify fracture images by the homology.
However, the relationship between the fracture conditions and the Betti numbers has not been clear,
yet. In the future, we would like to conduct experiments with different conditions of fracture and
analyze the relationship of the Betti numbers with the fracture surface in detail.

Acknowledgement
This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research (B) Grant
Number 26310209.

References
[1] Hibi T: Commutative Algebra and Combinatorial Logic. Springer-Verlag, Tokyo,Inc, 1995.
[2] Hibi T: Enumeration Mathematics, Asakura Publishing Co., Ltd., 1997.
[3] D. Kozlov, Combinatorial Algebraic Topology, Springer, 2008. T. Kaczynski, K.Mischaikow, and
M. Mrozek, Computational Homology, Springer, 2004.
[4] Nakane K, Tsuchihashi Y: A Simple Mathematical Model Utilizing a TopologicalInvariant for
Automatic Detection of Tumor Areas in Digital Tissue Images. Diagnostic Pathology 2013, 8
(Suppl 1).
[5] Nakane K, Honda T, Santos E. C., Kida K: Image Analysis for the Structures of Grains via
Homology Methods. Proc. New Methods of Damage And Failure Analysis of Structural Parts:
2012: 271-277.
[6] Nakane K, Mizobe K, Santos E. C., Kida K: Topological difference of grain composition in the
WMZ (Weld Metal Zone) in low carbon steel Plates (JIS-SS400). Advanced Materials Research
(Trance Tech Publications, ISSN: 1022-6680), 2013, Vol. 566: 399-405.
[7] Nakane K, Kida K, Mizobe K: Homology analysis of prior austenite grain size of SAE52100
bearing steel processed by cyclic heat treatment. Advanced Materials Research, 2013, Vol. 813:
116-119.
[8] Nakane K, Mizobe K, Kida K: Homology estimate of grain size measurement based on the JIS
samples. Applied Mechanics and Materials, 2013, Vol. 372: 116-119.
[9] Nakane K, Kida K, Honda T, Mizobe K: Influence of repeated quenching on bearing steel
martensitic structure investigated by homology. Applied Mechanics and Materials, 2013, Vol.
372: 270-272.
[10] Nakane K, Mizobe K, Santos E. C., Kida K: Quantitative estimates of repeatedly quenched high
carbon bearing steel. Applied Mechanics and Materials, 2013, Vol. 372: 273-276.
[11] Nakane K, Santos E. C., Honda T, Mizobe K, Kida K: Homology Analysis of Structure of high
carbon bearing steel: Effect of Repeated quenching on Prior Austenite Grain Size. Materials
Research Innovations, 2014, Vol. 18: 33-37.
[12] CHomP [http://chomp.rutgers.edu/projects/]
Advanced Materials Research Vol 1102 (2015) pp 139-142 Submitted: 2015-01-21
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.139

Dynamic Mechanical Thermal Analysis of Arenga Pinnata Fibre


Reinforced Epoxy Composites: Effects of Alkaline Treatment
Abdul Hakim Abdullah1,a Mohamad Syafiq Abdul Khadir1,b
and Nik Roselina Nik Roseley1,c
1
Faculty of Mechanical Engineering, University of Technology MARA, 40450 Shah Alam, Selangor
Darul Ehsan, Malaysia
a
hakimnen@yahoo.com, bsyafiqk89@gmail.com and cr_roselina@hotmail.com

Keywords: Composites, Arenga Pinnata fibre, alkaline treatment, Dynamic mechanical thermal
properties

Abstract: In present investigations, dynamic mechanical thermal behaviours of Arenga Pinnata


fibres prior and after alkaline treatment were studied. The alkaline treatments were applied on the
Arenga Pinnata fibres by immersed in the alkaline solution, 6% Sodium Hydroxide (NaOH). Using
hand lay-out technique, composites were fabricated at 20% and 40% by Arenga Pinnata fibres
weight contents. The thermal behaviours of both untreated and treated composites were determined
by employing Dynamic Mechanical Analysis (DMTA). The results shows that the treated fibres
owned better results of Storage Modulus (E), Loss Modulus (E) and Tan Delta temperatures
ranges from 0oC to 80oC.

Introduction
Arenga Pinnata, Sugar Palm or famously known as Ijok for local Malaysian, are growing wildly
in tropical forest and usually close to human settlements. The raw fibres are found in hairy black,
think fibrous trunk and it is covered with dense leaves, which is white colour and underside. In fact,
the sugar pump tree can grow up until 12 meter high [1].
Natural fibres reinforced in polymer composite materials can provide significant function and as
well sustainable benefits, ranging from acceptable mechanical strength, cost saving features and
biodegradable composites. Various fibres treatment was done such as alkaline and seawater play
important role in strengthening fibre-matrix adhesion in composites [2,3].
Dynamic Mechanical Analysis (DMTA) is widely used as a technique to find out the interfacial
characteristic of heterogeneous polymer systems under temperature dependent dynamic properties
[4]. It helps researcher to insight interaction between polymer matrix and fibre reinforcement in the
glass transition region. Tan is a measurement of the damping of the material meanwhile The
Storage Modulus is the intimation of the stiffness of the material. Loss modulus (E), loss factor or
the mechanical damping (tan ) and the dynamic storage modulus (E) are among the characteristics
that can be determined by using the DMTA technique.
However, no study appears to have reported of the effect alkaline treatment towards the DMTA
properties and how this affects to the performances. The mechanical responses of these fibres
behave differently under various conditions such as temperatures rising and unlikely remain less
understood. In present investigation, the DMTA response of treated and untreated Arenga Pinnata
fibre-reinforced epoxy composite was investigated, with differing level of fibres loading of 20%
and 40% respectively. It is perhaps the results could be beneficial in the study of natural fibre
reinforced composites.

Methodology
Fibre Preparation and Alkaline Treatment. The Arenga Pinnata fibres were harvested directly
from its own tree in Negeri Sembilan, Malaysia. The Arenga Pinnata fibres were immersed in the
6% of Sodium Hydroxide (NaOH) for 6 hours in a bucket as the alkaline treatment procedure. The
total amount of NaOH tablet needed was 12 g in order to get 6% NaOH since the all the fibres need
140 Metal, Ceramic and Composite Materials

to be immersed in 5 liter of NaOH solution. While the untreated fibre skipped the process, both
treated and untreated fibres were then dried in the oven at 800C for 60 minutes for moisture
removal. Lastly, all the fibres were stored in sealed plastic bags in similar purpose, to avoid any
moisture absorption effects.
Specimen Fabrication. The technique that been used to fabricate the composite specimen is
hand lay-up process. Each layer of Arenga Pinnata fibres were arranged in the mould, which are
different between 4 parameters, 40% Treated Arenga Pinnata (TAP), 40% Untreated Arenga
Pinnata (UAP), 20% Treated Arenga Pinnata (TAP) and lastly 20% Untreated Arenga Pinnata
(UAP). Then, the matrix was poured onto the fibre and it was distributed uniformly in order to
ensure all fibre part will be covered by the matrix and the best composite specimen can be
produced. The completed composite in the mould was placed and being pressurized. Finally, the
mould was left for 24 hours for curing process.
DMTA Test. The DMTA properties were determined by using Perkin Elmer D8000 Analyzer.
Samples in dimension of 10 mm in width, 3 mm of thickness and 30 mm in length was insert in the
machine cavity, employing a simple cantilevered mode. The heating process was maintained
2C/min, starting from 0C to 80C across the band. The process was repeated 5 times for better
results.
Results And Discussion
Storage Modulus. From Figure 1(a), it can be simplify that in terms of composition of fibre, the
highest amount of the fibre had greater strength compared than the composite with less fibre in the
composite. It can be proved between 20% AP and 40% AP for both treated fibre and untreated
fibre. Refer to H. Anuar and co-worker [6], whenever the increasing of the fibre used, the strength
also enhanced because of the inert surface layer, which caused the interaction between fibre and
fibre was gone. At the same time, in term of alkaline treatment, the treated fibre contributes tend to
have higher value of E as compared to the specimen with same composition fibre but not undergo
the alkaline treatment. Except for 20% for TAP and UAP that likely to have a small improvement of
E, the 40% TAP gained significant benefits of alkaline treatment. This suggests the alkaline
treatment could improve the storage modulus.
Meanwhile for the temperature, the highest value of Storage Modulus, E will drop first
compared than the lowest value of E. Firstly was 40% TAP, which was drop first at 5 oC and
followed by 58oC for the next specimen, 40% UAP. Simultaneously, for the next two specimens
gave the same temperature drop, 60oC. It showed that the specimen with high strength, the
resistance to the temperature was low. It can be seen from the graph trend. The line for the
specimen with the best strength value will drop first and followed by the other in descending order.
Above the 60oC, most of the composites suffer lows thermal properties, as indicates by Storage
Modulus that nearly exceed zero values. This suggests the alkaline treatments are not viable at high
operating temperature.
Loss Modulus. Figure 1(b) shows the loss modulus results of all tested specimens. The highest
value of Loss Modulus, E was obtained from the specimen 20% UAP, with the value was 9.7 GPa
and followed by 20% TAP. Its value from the graph was 8.8 GPa. The next value was 6.8 GPa,
which was obtained from 40% UAP and lastly followed by 40% TAP and its reading was 6 GPa.
The tabulation of Loss Modulus, E were opposite with the trend of Storage Modulus, E. When
there were reduction in the value for E, the result of E will increase. According to Anuar and co-
worker, the possible reason for this trend was because of there was more dominant for fibre-fibre
interaction. Because of that, the effectiveness to transfer stress form matrix to fibre was declining
[6]. Indirectly, the value of Loss Modulus were decreasing by the increasing the value of Loss
Modulus.
Advanced Materials Research Vol. 1102 141

80 12
Storage Modulus, E' (GPa)

70

Loss Modulus, E" (GPa)


10
60
8
50
40 6
30 4
20
2
10
0 0
0 20 40 60 80 0.0 20.0 40.0 60.0 80.0
o
Temperature, T ( C)
Temperature, T (oC)
(a) (b)
1.4
1.2
1.0 Legend
Tan Delta

0.8 20 % Untreated AP
0.6 20%Treated AP
0.4 40 % Untreated AP
40 %Treated AP
0.2
0.0
0.0 20.0 40.0 60.0 80.0
Temperature, T (oC)
(c)

Figure 1: DMTA Results of (a) Storage Modulus, (b) Loss Modulus and (C) Tan Delta
Tan Delta. Figure 1(c) shows the Tan Delta results of all tested specimens. The highest value of
the peak graph occurred on 20 % UAP specimen and the value was 1.13 and followed by the
specimen 20 % TAP with 0.93. The next value was 0.87 when the test conducted on 40 % UAP and
ended with the specimen 40 % TAP, 0.68. In temperature aspect, the result pattern similar to the
pattern for Loss Modulus. The lowest temperature was obtained from 40 % TAP specimen with the
value was 64oC and followed by 40 % UAP, 69oC. For 20 % AP, either treated or untreated both
share the same number for the temperature.
From all of these values, the trend that can be seen that whenever the presence of fibre in the
specimen increased, the value of Tan will be reduce. It can be shown when 40% Arenga Pinnata
fibre was used in composite; the value was low compared than the value when the fibre was used
only 20%. Meanwhile, in terms of alkaline treated, the treated fibre had the low value of Tan
compared than the value from untreated fibre. The Tan Delta often relates to the damping
characterises, a measure of the energy dissipation of a material. The behavior of increasing damping
peak in the tan delta curves is attributed by decreases of amount of matrix incorporated with fibres
[7].This suggests the treated fibres have gain benefits improving their mechanical properties.

Summary
In this study, effects of alkaline treatment towards Dynamic Mechanical Analysis are successfully
investigated. It can be concluded that the alkaline treatment of 20% and 40% Arenga Pinnata
gained such benefits in DMTA as the results of Storage Modulus, Loss Modulus and Tan Delta.
However, their thermal performances are not affected by the alkaline treatment and as well fibre
compositions.
142 Metal, Ceramic and Composite Materials

Acknowledgment
The authors would like to thank the Research Management Institute (RMI) of Universiti Teknologi
MARA (UiTM), Malaysia for financial support and facilitating this project support to this study
through grants awards of 600-RMI/RAGS 5/3 (41/2012) and 600-RMI/RAGS 5/3 (47/2013).

References
[1] M.F.S. Khalid and A.H. Abdullah, Storage Modulus Capacity of Untreated Aged Arenga
pinnata Fibre-Reinforced Epoxy Composite, Appl Mech Mater 393, (2013) 171-176.
[2] A.H. Abdullah, S.K. Alias and S. Abdul Latip, S. Yunus, Storage Modulus Analysis of Kenaf
Fibre Reinforced Epoxy Composites, Appl Mech Mater 393 (2013) 167-170.
[3] A.H. Abdullah, A. Khalina and Aidy Ali, Effects Of Fiber Volume Fraction nn Unidirectional
Kenaf/Epoxy Composites: The Transition Region, Polymer-Plastics Technology and Engineering
50 (2011) 1362-1366.
[4] M. Vasumathi and Vela Murali. Mechanical Behaviour of Chemically Treated Reshira-Epoxy
Composite at Cryogenic Temperatures, Appl Mech Mater, 488-489 (2012) 718-723.
[5] A. Koyun, E. Ahlatcioglu, D.C. Erdogan and Y.K. Ipek. (2012), The DMA, TG/DTA and FT-
IR characteristics of Wood Polymer Composite. Adv Mat Res. 445 (2012) 219-224.
[6[ H.Anuar, S.H.Ahmad, R.Rasid, A.Ahmad, W.N. Wan Busu. Mechanical Properties and
Dynamic Mechanical Analysis of Thermoplastic-Natural-Rubber-Reinforced Short Carbon Fibre
and Kenaf Fibre Hybric Composite. Journal of Applied Polymer Science, 107 (2007),4045-4050.
[7] M. Idicula, S.K. Malhotra, K. Joseph and S. Thomas. Dynamic Mechanical Analysis of
Randomly Oriented Intimately Mixed Short Banana/Sisal Hybrid Fibre Reinforced Polyester
Composites. Composites Science and Technology, 65 (2005)1077-1087.
Advanced Materials Research Vol 1102 (2015) pp 143-147 Submitted: 2015-01-28
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.143

Fabrication of high-aspect-ratio structural change microregions in


silicon carbide by femtosecond Bessel beams
Liangliang Zhao1,a, Feng Wang1,b *, Jun Xie1,c , Weiwei Zhao1,d
1
Laser Micro/Nano-Fabrication Laboratory, School of Mechanical Engineering, Beijing Institute of
Technology, 100081, PR China
a
zll_hope_future@126.com, bwangfeng01@tsinghua.org.cn, cxjun0220@126.com,
d
zhaoweiweibit@126.com.
Keywords: Femtosecond laser, High aspect ratio, Microstructure fabrication, Bessel beams.
Abstract:
This letter presents the morphology of femtosecond Bessel beams induced high-aspect-ratio
structural change regions in bulk silicon carbide. An axion is engaged in transforming Gaussians
beams to Bessel beams, which are then focused on the surface or below the surface of the sample by
combination of a plano-convex lens and a microscope objective. The sample is scanned by the
focused femtosecond Bessel beams at the preset patterns. Through this method, the high-aspect-ratio
uniform laser induced structure change regions have been produced and the highest respect ratio can
reach 206 with the depth of 330 m, the width of 1.6 m in optimized conditions of appropriate
focusing position and pulse energy. This result is attributed to uniform energy distribution in the long
propagation distance of Bessel beams with nondiffracting. This technique will have great potential
applications to make high-aspect-ratio microgrooves in wide-gap and transparent materials.
Introduction
Femtosecond laser pulses are promising tools for machining variable materials, including metals
[1], semiconductors [2], dielectric [3] and polymers [4], which has stimulated extensive research
interests in recent years. In machining some microstructures in all materials, especially wide-gap
materials, femtosecond laser machining offers its unique advantages, such as the minimized recast,
micro-cracks, heat affect zones and capacity of three dimensional processing, which is considered as a
non-thermal, non-linear and non-equilibrium process due to ultra-short pulse durations and ultra-high
power densities [5-6].
Among these microstructures, microgrooves, especially high-aspect-ratio (HAR) microgrooves
have been more and more widely used in micro-fluid system, micro flow sensor and micro-electro-
mechanical systems (MEMS) [7-8], such as silicon trench isolation and silicon trench capacitor and
so on, and they will make the performance of devices improved greatly. Some studies about
fabrication HAR microgrooves in silicon using femtosecond laser have been reported [9-11], but the
silicon HAR microgrooves are limited in the harsh environment. According to the investigation,
silicon carbide (SiC) substrate is the most ideal candidate to overcome the shortcomings due to its
excellent physical and chemical properties [12-14]. Considering SiC is transparent to 800 nm laser, an
800nm femtosecond laser is able to induce structural changes, which can be removed by the
appropriate etching technique, in SiC.
In this paper, the uniform and HAR laser induced structure change (LISC) regions are produced
in SiC by femtosecond Bessel beams scanning irradiation. The sample is scanned by the focused
femtosecond Bessel beams at the preset patterns to produce the LISC regions, which can be removed
subsequently by appropriate etching technique to fabricate the HAR microgrooves. An optical
microscope and a scanning electron microscope (SEM) are used to characterize the morphology. The
highest respect ratio of the LISC regions can reach 206 with the depth of 330m, the width of 1.6 m.
The result is attributed to that the field depth of Bessel beams is far larger than that of Gaussian beams
and the laser energy is deposited over extended propagation lengths of Bessel beams uniformly
[15-17]. This technique will have great potential applications to make HAR microgrooves in
transparent and wide-gap materials.
144 Metal, Ceramic and Composite Materials

Experimental details
Fig.1 shows the schematic of the experiment setup of the fabrication of LISC regions in SiC with
femtosecond Bessel beams irradiation in ambient air. An amplified Ti: sapphire femtosecond laser
system (Spectra Physics Spitfire) is used to generate femtosecond laser pulses with 50 fs pulse
duration, 800 nm laser wavelength, 1 KHz repetition, Gaussian-type intensity distribution. By
combining a half-wave plate with a polarizer, the energy of the laser pulses can be continuously
adjusted. An electromechanical shutter is used to control the irradiation time. An axion was used to
convert the Gaussian beams to Bessel beams [18].
The 6H-SiC wafers <0001>with the thickness of 330m is first cleaned in de-ionized water and
alcohol in ultrasonic cleaner and then is fixed on a computer controlled six-axis piezo stage
(PIM-840.5DG) and the positioning accuracy is 1m in the x and y directions and 0.5 m in the z
direction. A plano-convex lens (f=150mm) and a 20 microscope objective with the numerical
aperture (NA) of 0.45 are combined to focus the femtosecond laser to the surface of SiC. The
scanning direction is parallel to the laser polarization direction and the length of each scanning line is
adjusted to be 1000m. The separation distance of each two scanning lines is set 100m. The CCD
cameral is used to observe the machining process in real time and provide a feedback to adjust the
experimental conditions. To observe the morphology of the fabricated LISC regions, the cross section
is characterized by an optical microscope and a scanning electron microscope (SEM).

Fig.1 The schematic of the experiment setup of the fabrication of LISC regions in SiC with femtosecond Bessel beams
direct irradiation in ambient air. The scanning direction is along y-axis.
Experiment results and discussion
The full wave at half maximum (FWHM) of Bessel beams intensity distribution is larger than
that of Gaussian beams when the beam waist of the Gaussian is the same as the width of the central
lobe of the Bessel beam. In view of the extended depth of the field of Bessel beams, the best focusing
positon of the femtosecond laser is need to guarantee the field covering the sample as much as
possible. The curves in the Fig.2 illustrate the dependence of LISC regions depth on the focusing
position. The pulse energy is 15 J and the scanning speed is 10m/s. It can be seen the depth is the
highest of 330m (equal to the thickness of the sample) when the femtosecond laser is focused on
position of Z=+300. The FWHM is proved to be larger than 330m, which is consistent with the
theoretical calculation results of our experimental conditions : FWHM along z-direction is 812m.
Advanced Materials Research Vol. 1102 145

Fig.2 The dependence of the depth of LISC regions on the focusing position, the pulse energy is 15 J, the scanning speed
is 10m/s, the focusing position is Z=+50, +100, +150, +200, +250, +300mrespectively. The scale bar is 50 m.

In order to improve the aspect ratio of the LISC regions, the focusing positon of z=+300m is
engaged in the following experiments. The LISC regions are produced at ambient air. The scanning
speed is set 10m/s and the pulse energy of the laser was set at 5J10J15J20J25J. Figure
3 depicts the cross-sectional optical and SEM images of the LISC regions fabricated at different pulse
energies. It can be seen from the fig.3 (a), the LISC regions are produced at the irradiated zones along
femtosecond laser propagation direction at the different pulse energies. Fig.3 (b) shows that some
irregular structures at the LISC regions. The width of the uniform LISC region at the pulse of 20J is
2.5m, while the depth is 330mthat is to say that the aspect ratio is 132, which is much larger than
the produced LISC regions using femtosecond Gaussian beams irradiation.

Fig.3 Cross-sectional optical (a) and SEM (b) images of the LISC regions fabricated at different pulse energies of 5J
10J15J20J25J. The scale bar in (a) is 50 m, and in (b) is 10 m.

The formation of LISC regions is attributed to the bond breaking by intense femtosecond laser
induced multi-photon ionization [19]. The hexagonal crystal structure of 6H-SiC is destroyed by laser
induced-heat and high-pressure, leading to the formation of dangling bonds, which will pull oxygen
element into the material. The unique properties of Bessel beams, which allow uniform energy
distribution in the long propagation distance, are a key factor to the high-aspect-ratio and uniform
morphology of the LISC regions. The LISC regions can be etched by the appropriate etching
technique due to their less stable chemical property.
To investigate the dependence of the aspect ratio of LISC regions on the scanning speed, the
scanning speed of 10m/s, 15m/s, 20m/s, 30m/s, 40m/s, 50m/s are used to produce the LISC
regions at the pulse energy of 20J. The increase of the scanning speed leads to the increase of the
aspect ratio of the LISC regions, which is shown in fig.4, with the depth constant at 330m and the
decrease of the width. The decrease of the width is a result of the decreasing pulse energy deposited on
unit size of the materials with the increase of the scanning speed. In view of the above, the aspect
ratio of LISC regions can be improved by increasing the scanning speed properly.
146 Metal, Ceramic and Composite Materials

Fig.4 The dependence of the aspect ratio of LISC regions on the scanning speed, the pulse energy is 20Jthe scanning
speed are 10m/s, 15m/s, 20m/s, 30m/s, 40m/s, 50m/s, respectively.
Summary
In conclusion, the uniform and HAR LISC regions have been fabricated by femtosecond Bessel
beams scanning irradiation. The highest respect ratio of the LISC regions can reach 206 with the depth
of 330m, and the width of 1.6 m. The formation of LISC regions is attributed to the bond breaking
by intense femtosecond laser induced multi-photon ionization. The uniformity and high aspect ratio
are a result of the uniform energy distribution in the long propagation distance of femtosecond Bessel
beams, which are nondiffracting. Due to the less stable chemical property of LISC regions, they can
be subsequently removed by the appropriate etching technique to form the HAR mirogrooves. This
technique will have great potential applications to make HAR microgrooves in wide-gap and
transparent materials.
References
[1] A. Y. Vorobyev and C. L. Guo, "Colorizing metals with femtosecond laser pulses", Appl. Phys.
Lett. 92, 041914(2008)
[2] A. Borowiec and H. K. Haugen, "Sub-wavelength ripple formation on the surfaces of compound
semiconductors irradiated with femtosecond laser pulses", Appl. Phys. Lett. 82, 4462 (2003)
[3] C. Hnatovsky, R. S. Taylor, P. P. Rajeev, E. Simova, V. R.Bhardwaj, D. M. Rayner and P. B.
Corkum, "Pulse duration dependence of femtosecond-laser-fabricated nano-gratings in fused
silica", Appl. Phys. Lett. 87, 014104 (2005)
[4] K. C. Vishnubhatla, J. Clark, G. Lanzani, R. Ramponi, R. Osellame, and T. Virgili, "Ultrafast
optofluidic gain switch based on conjugated polymer in femtosecond laser fabricated
microchannels", Appl. Phys. Lett. 94, 041123 (2009)
[5] Y. Iga, T. Ishizuka, W. Watanabe, K. Itoh, Y. Li and J. Nishii, "Characterization of
micro-channels fabricated by in-water ablation of femtosecond laser pulses," Jpn. J. Appl. Phys.
43 4207 (2004).
[6] Q. Sun, A. Saliminia, F. Thberge, R. alle and S. Leang, "Microchannel fabrication in silica
glass by femtosecond laser pulses with different central wavelengths," J. Micromech.
Microeng.18, 035039 (2008).
[7] W. Noell, P. A. Clerc, L. Dellmann, B. Guldimann, H. P. Herzig, O. Manzardo, C. R. Marxer, K.
J. Weible, R,Dndliker, and N. Rooij, IEEE J. Sel. Top. Quantum Electron. 8, 148 (2002).
[8] W. Ong, J. Kee, A. Ajay, N. Ranganathan, K. Tang, and L. Yobas, "Buried microfluidic channel
for integrated patch-clamping assay", Appl. Phys. Lett. 89(9), 093902 (2006).
[9] Y. Ma, H. T. Shi, J. H. Si, H. Ren, T. Chen, F. Chen and X. Hou, "High-aspect-ratio grooves
fabricated in silicon by a single pass of femtosecond laser pulses," J. Appl. Phys. 111, 093102
(2012).
[10] A. Pan, J .H. Si, T. Chen, Y. C. Ma, F. Chen, and X. Hou, "Fabrication of high-aspect-ratio
grooves in silicon using femtosecond laser irradiation and oxygen-dependent acid etching", Opt.
Express 21, 16657 (2013).
[11] Y. Ma, A. Pan, J. H. Si ,T. Chen, F. Chen, X. Hou, "A simple method for fabrication of
high-aspect-ratio all-silicon grooves," Appl. Surf. Sci. 284, 372 (2013).
Advanced Materials Research Vol. 1102 147

[12] D. J. Young, J. G. Du, C. A. Zorman, and H. K. Wen, "High-temperature single-crystal 3C-SiC


capacitive pressure sensor," IEEE Sens. J. 4, 464 (2004).
[13] C. A. Zorman, S. Rajgopal, X. A. Fu, R. Jezeski, J. Melzak and M. Mehregany, "Deposition of
polycrystalline 3C-SiC films on 100 mm diameter Si (100) wafers in a large-volume LPCVD
furnace," Electrochem. Solid-State Lett. 5, 10 (2002).
[14] M. stling, S. M.Koo, C. M. Zetterling, S.Khartsev, A. Grishin, "Ferroelectric thin films on
silicon carbide for next-generation nonvolatile memory and sensor devices," Thin solid films 469,
444 (2004).
[15] M. K. Bhuyan, F. Courvoisier, P. A. Lacourt, "High aspect ratio nanochannel machining using
single shot femtosecond Bessel beams." Appl. Phys. Lett. 97, 081102 (2010).
[16] M. K. Bhuyan, F. Courvoisier, H. S. Phing, "Laser micro- and nanostructuring using femtosecond
Bessel beams," Eur. Phys. J. Special Topics 199, 101 (2011).
[17] J. Durnin, J. J. Miceli, and J. H. Eberly, "Holographic generation of diffraction-free beams," Phys.
Rev. Lett. 58, 1499 (1987)
[18] F. Courvoisier, P. A. Lacourt, M. Jacquot, M. K. Bhuyan, "Surface nanoprocessing with
nondiffracting femtosecond Bessel beams." Opt. Lett. 34, 3163 (2009).
[19] K. Vanthanh, Y. C. Ma, J. H. Si, T. Chen, F. Chen, X. Hou, "Fabrication of Mmicro-grooves in
silicon carbide using femtosecond laser irradiation and acid etching," Chin. Phys. Lett. 3, 037901
(2014).
Advanced Materials Research Vol 1102 (2015) pp 149-153 Submitted: 2015-02-03
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.149

Numerical investigation of the damage behavior of S355 EBW by


cohesive zone modeling
H. Y. Tu, U. Weber, S. Schmauder
Institute for Materials Testing, Materials Science and Strength of Materials (IMWF) University of
Stuttgart, Pfaffenwaldring 32, D-70569 Stuttgart, Germany
haoyun.tu@imwf.uni-stuttgart.de, ulrich.weber@imwf.uni-stuttgart.de
siegfried.schmauder@imwf.uni-stuttgart.de

Keywords: Cohesive zone model, electron beam welded joints, crack propagation, finite element
modeling

Abstract In this paper, the cohesive zone model is used to study the fracture behavior of an
electron beam welded (EBW) steel joint. Mechanical properties of different weld regions are
derived from the tensile test results of flat specimens, which are obtained from the respective weld
regions. Based on the tensile test of notched round specimens, the cohesive strength T0 can be fixed.
With the fixed T0 value, the cohesive model is applied to compact tension (C(T)) specimens with
the initial crack located at different positions of weldment with different cohesive energy values 0.
Numerical simulations are compared with the experimental results in the form of force vs. Crack
Opening Displacement (COD) curves as well as fracture resistance (JR) curves.

Introduction

The failure of the weldments always is the hot spot for scientific research as the fracture behavior of
the welded joints influences the crack growth in structures, which affects the service lifetime of
welded components. With the development of the finite element method, a large number of
publications appeared on the topic of fracture behavior of welded joints with different damage
models [1-3]. Introduced by Dugdale [4] and Barenblatt [5], the cohesive model has many
applications on ductile or brittle fracture behavior of a number of materials [6-10]. The damage of
the cohesive zone is described by a traction-separation law which is described by the cohesive
strength T0, the critical separation 0 and the cohesive energy 0. In this paper, the cohesive zone
model is adopted to study fracture of welded S355 EBW joints.

Experimental investigations

A butt joint obtained from two S355NL plates with the thickness of 60 mm was investigated in the
experiments. After the hardness test, the dimensions of the fusion zone (FZ) and the heat affected
zone (HAZ) are found to be 2.8 mm and 3.1 mm, respectively. The mechanical properties of
different weld regions of S355 EBW joints are obtained from flat specimens along the weld line of
which the gauge length is 50 mm. These stress-strain curves are used for finite element model input
data as shown in literature [3].

Fracture toughness tests of S355 electron beam welded joints were performed using compact ten-
sion (C(T)) specimens with a thickness of B=25 mm, a net thickness of Bn=20 mm and the width of
specimen is W=50 mm. The F-COD curves of compact tension (C(T)) specimens with initial crack
located in the base material (C(T)-BM), in the center of FZ (C(T)-FZ) and at the interface between
the FZ and HAZ (C(T)-HAZ) are shown in Fig. 1(a). For a C(T) specimen with the crack in the FZ,
the specimen suddenly ruptures, showing, thus, a rather brittle fracture behavior. The JR curves for a
C(T)-BM and a C(T)-HAZ specimen are presented in Fig. 1(b).
150 Metal, Ceramic and Composite Materials

50 500
C(T)-BM
40 C(T)-HAZ
400

J-Integral (N/mm)
F (KN) 30 300

20 200

C(T)-BM
10 100
C(T)-FZ
C(T)-HAZ
0 0
0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
(a) COD (mm) (b) a (mm)

Fig. 1: Experimental (a) Force vs. Crack Opening Displacement (COD) and (b) fracture resistance
JR curves of compact tension (C(T)) specimens with the initial crack located in the BM, in the
center of the FZ and at the interface between the FZ and the HAZ, respectively.

Numerical calculation

Before the application of the cohesive model investigating the fracture behavior of S355 welded
joints, the cohesive parameters should be fixed first. According to the discussion of Cornec and
Scheider [11], for a mode I situation, the cohesive parameter T0 is equal to the projection of the
applied force on a plane perpendicular to the specimen cross section. The notched round specimens
extracted from the BM and HAZ are used for the determination of T0. The comparison of axial
stress versus the diameter reduction curve from the FE simulation and the experiment as well as the
maximum true axial stress in the center of the specimen can be found in Fig. 2. The simulated axial
stress versus the diameter reduction curve coincides with the experimental one until crack initiation,
where the experimental curve drops suddenly. At this point, the maximum value of stress
distribution over the cross section of the specimen is determined from the simulation and set equal
to the cohesive stress T0. After the comparison of simulation and experiment, the numerically
derived tractions for the BM and the HAZ are shown in Fig 2(a): T0=1180 MPa and Fig. 2(b):
T0=1350 MPa respectively. For the FZ, as the notched round specimen from the FZ is not available,
the T0 for the FZ is assumed roughly as three times of the yield stress of the FZ according to
literature data [12], which leads to T0=1600 MPa.
1800 1800
Engineering axial stress (MPa)
Engineering axial stress (MPa)

S355 BM--notched round specimen 1600 S355 HAZ--notched round specimen


1600

1400 1400
T0=1180MPa T0=1350 MPa
1200 1200

1000 1000

800 800

600 600

400 400
test: eng. axial stress test: eng. axial stress
200 el-pl FEM: eng. axial stress 200 el-pl FEM: eng. axial stress
el-pl FEM: axial stress in the centre el-pl FEM: axial stress in the centre
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
(a) D(mm) (b) D (mm)

Fig. 2: Determination of the cohesive stress T0 for (a) S355 BM and for (b) S355 HAZ.

For the C(T)-BM and C(T)-FZ specimen, because the structure shows symmetry with respect to the
crack plane, only half of the C(T) specimen is modeled, loading is defined on the loading point
(Red point) by the displacement, the finite element mesh and boundary conditions are shown in Fig.
3. Due to the inhomogeneous material behavior with respect to the crack plane of the C(T)
specimen, the whole C(T) structure is used for the simulation. As for the ductile behavior of BM
and HAZ, trapezoid traction-separation law is used to study the fracture behavior of the C(T)-BM
and C(T)-HAZ specimens. After calculation, good agreement between the numerical and
experimental results can be obtained in terms of F-COD and JR curves when T0=1180 MPa, 0=23.8
Advanced Materials Research Vol. 1102 151

N/mm for the BM and T0=1350 MPa, 0=22.5 N/mm for the HAZ are adopted, as shown in Figs.
4(a)-(b) and Figs. 5(a)-(b). For the C(T)-FZ, as the FZ shows a more brittle behavior, a linear
decreasing traction-separation law is required for the cohesive model. The numerical F-COD curve
matches the experimental one well before the sudden rupture, showing the cohesive model can
predict the F-COD curve of the C(T)-FZ specimen well, as can be seen in Fig. 6.

Loading direction

Initial crack length Symmetry plane


Symmetry plane Cohesive element layer
Fig. 3: Finite element mesh and boundary conditions for the C(T)-BM and C(T)-FZ.
50 500
Trapezoidal softening, 0=1180 MPa Trapezoidal softening, 0=1180 MPa

40 400
J-Integral (N/mm)

30 300
F (KN)

20 experiment--BM 200
Cohesive--0=23.8 N/mm experiment--BM
10 100 Cohesive--0=23.8 N/mm

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
(a) COD (mm) (b) a (mm)

Fig. 4: Comparison of experimental and numerical (a) force vs. Crack Opening Displacement
(COD) curves, and (b) fracture resistance curves for C(T)-BM.
50 400
Trapezoidal softening (TSL) Trapezoidal softening, T0=1350 MPa
350
40
300
J-Integral (N/mm)

30 250
F(KN)

200
20
150

10 experiment--HAZ 100
elastic-plastic experiment--HAZ
cohesive--T0=1350 MPa, 0=22.5 N/mm 50
cohesive--0=22.5 N/mm
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
(a) COD (mm) (b) a (mm)

Fig. 5: Comparison of experimental and numerical (a) force vs. Crack Opening Displacement
(COD) curves, and (b) fracture resistance curves for C(T)-HAZ.
152 Metal, Ceramic and Composite Materials

60
T0=1600 MPa
50

40

F (KN)
30

20

10 experiment--FZ
Cohesive--0=70 N/mm
0
0 1 2 3 4 5 6 7
COD (mm)
Fig. 6: Comparison of experimental and numerical force vs. Crack Opening Displacement (COD)
curves for C(T)-FZ.

Conclusions

Crack propagation was studied for S355 EBW joints using the cohesive model. Three different C(T)
specimens, i.e., the C(T)-BM, the C(T)-FZ and the C(T)-HAZ were investigated. Based on the axial
stress versus diameter reduction curve of notched round specimens, the cohesive strength T0 is
fixed. When a trapezoidal traction-separation law is used, the cohesive model can predict the F-
COD and JR curves of the C(T)-BM and the C(T)-HAZ well. The cohesive model can also predict
F-COD curves of C(T)-FZ specimens before the sudden rupture when the linear softening TSL is
used. All in all, the investigations of the fracture behavior of S355 EBW joints with the cohesive
model confirm that the cohesive model is able to predict the crack propagation of the homogenous
BM and the inhomogeneous welded joints well.

Acknowledgements

The research has been financially supported by the Deutsch Forschungsgemeinschaft (DFG) under
Schm 746/134-1. This financial support is gratefully acknowledged.

Reference

[1] P. Ngre et al.. Numerical simulation of crack extension in aluminium welds. Comp Mater Sci
28 (2003), pp. 723-731.
[2] A. Nonn, W. Dahl, W. Bleck. Numerical modelling of damage behaviour of laser-hybrid welds.
Engng Fract Mech 75 (2008), pp. 3251-3263.
[3] H. Y. Tu, S. Schmauder, U. Weber, Y. Rudnik, V. Ploshikhin. Simulation of the damage
behavior of electron beam welded joints with the Rousselier model. Engng Fract Mech 103
(2013), 153-161.
[4] D. S. Dugdale, Yielding of steel sheets containing slits. J Mech Phys Solids 8 (1960), pp. 100-
104.
[5] G. I. Barenblatt. The mathematical theory of equilibrium cracks in brittle fracture. Adv Appl
Mechanics 7 (1962), pp. 55-129.
[6] A. Hillerborg et al.. Analysis of crack formation and crack growth in concrete by means of
fracture mechanics and finite elements. Cem Conc Research 6 (1976), pp. 773782.
[7] A. Needleman. A continuum model for void nucleation by inclusion debonding. J Appl Mech
ASME 54 (1987), pp. 525531.
[8] A. Needleman. An analysis of decohesion along an imperfect interface. I J Frac 42 (1990), pp.
21-40.
Advanced Materials Research Vol. 1102 153

[9] V. Tvergaard, J. W. Hutchinson. The relation between crack growth resistance and fracture
process parameters in elastic-plastic solids. J Mech Phys Solids 40 (1992), pp. 13771397.
[10] I. Scheider. Bruchmechanische Bewertung von Laserschweissverbindungen durch numerische
Rissfortschrittsimulation mit dem Kohsivzonenmodell. PhD thesis, TU Hamburg-Harburg,
Geesthacht, 2001.
[11] A. Cornec et al.. On the practical application of the cohesive model. Engng Fract Mech 70
(2003), pp. 1963-1987.
[12] K.-H. Schwalbe, I. Scheider, A. Cornec. SIAM CM09 - The SIAM method for application
cohesive models of the damage behaviour of engineering materials and structures. GKSS report,
2009/1, Geesthacht, 2009.
Advanced Materials Research Vol 1102 (2015) pp 155-159 Submitted: 2015-01-30
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.155

Determining the True Strength of the Material of Fiberglass Thick Rings


When Stretched with Half-disks
TURUSOV Robert A.1, a, KUPERMAN Alexandr M.1,a,
ANDREEV Vladimir I.1,b
1
Moscow State University of Civil Engineering, 26 Jaroslavskoje shosse,
Moscow, 129337, Russian Federation,
a
rob-turusov@yandex.ru, b asv@mgsu.ru

Keywords: fiberglass plastics, wet winding, unidirectional rings, strength, elastic properties,
Hooks law

Abstract. The article presents the results of a study of elastic-strength properties of unidirectional
rings of fiberglass in tension them with half-disks at a variation in a wide range of diameter and
thickness of the rings. The analysis of the results clarified the mechanisms of destruction
fibreglasses and the possibility of expanding the application of the method.

Introduction
Unidirectional fibreglasses ring samples were prepared by the so-called "wet" winding glass
fibers brand VMP magnesium aluminosilicate composition (E = 95 GPa = 3.5 GPa) average
diameter of 10 microns as a tourniquet with linear density of 400 tex manufactured using a lubricant
N76 containing in its composition coupling agent AGM-9. Used as a binder widespread epoxy risen
EDT-10.

Experimental technique
During experiments the mandrel represents the sets of half-disks [1] of various diameters D: 85,
155, 220 and 300 mm, with internally machined annular grooves in them by the cylindrical surface
of the depth of 10 mm and a width of 5 mm (Fig. 1). Before winding thin plastic film cover inner
surface of the groove. This created
antiadhesive film layer, and reduces the
friction ring about the metal mandrel by
subsequent tensile tests. Before winding
thin plastic film cover inner surface of the
groove. The fiber placement across the
width of the sample was carried out using
the cam spreader which provides a
winding pitch of from 1 to 2 mm per
revolution. The inner diameter D1 of the
ring samples obtained as a result was,
respectively, 65, 135, 200 and 280 mm.
Maximum thickness of the samples was
determined by complete filling of the Figure 1. Experimental apparatus
groove, and then varied each subsequent
time halving the time of winding. Thus obtained samples of nominal thickness h were: 10; 5; 2.5;
1.25 and 0.63 mm. The curing schedule of the binder was slowly (over 4 h) heating the samples to
160C, holding at this temperature for 6 hours and natural cooling in the Heating Cabinets. The
content of fibers, binder and porosity of the samples were determined, based on the weight ratio of
156 Metal, Ceramic and Composite Materials

the components and their density. The fiber content was 50-60% by volume, the porosity of - less
than 3%.
Tests ring samples were carried out in tension on a universal testing machine "Instron 1196" with
a maximum load of 250 kN. For fastening half-disks were manufactured special tractions excluding
the loss of stability half-disks. The loading was performed at a constant speed of movement of the
lower clamp 10 mm/min. During the test chart recorded: force P - absolute increment of sample
length l, which allows calculating the chart - and correspondingly elastic-strength properties of
the composite. During experiments, it was tested at least three samples of each size.
Typical averaged tension diagrams - of rings obtained when tested using the half-disks
depending on the thickness and diameter of the samples are shown in Fig. 2.

Fig. 2. Typical stress-strain diagrams - fiber in ring samples


) inner diameter of 135 mm and of thickness 0.63 mm (1), 1.25 mm (2), 2.5 mm (3),
5 mm (4) & 10 mm(5);
b) thickness of 1.25 mm and inner diameter of 65 mm (1), 135 mm (2), 200 mm (3) and 280 mm (4)
Fig. 3 and 4 (a, b, c) shows the dependence of the elastic and strength characteristics of
unidirectional fiberglass ring samples tested using the half-disks of diameter and thickness of
specimens. As the Young's modulus so the strength of rings substantially reduced with decreasing
diameter and increasing the thickness of the rings. As an example, increasing the thickness of about
15 times these parameters are reduced by 2 times for rings 280 mm diameter and 3 times - for a ring
diameter of 65 mm. The reasons for this are defined inhomogeneous stress-strain state of rings in
the test.
Rebuilding the dependence of the elastic modulus E and strength of the samples on the
relative thickness /Rav all experimental data fall on a common curve (Fig. 3,b and 4,b) for rings of
different diameters. In the semi-logarithmic coordinates E lg (/Rav) and lg( / Rav ) become
linear dependence (Fig. 3,c and 4,c) and expressed by the following analytical formulas:

fib 65.0 lg( / Rav ), GPa , fib 1.8 lg( / Rav ), GPa .

Naturally, these relationships are valid only in the investigated range of the relative thicknesses
of the ring samples. If we extrapolate the straight, it may happen that, for large thicknesses of the
ring will have zero strength and stiffness, which is certainly not true. Yet, on the basis of the data
obtained can be counted obtained elastic-strength properties of the samples to the conditions of the
test in a more or less homogeneous field strains and stresses.
If we assume that such is observed at / Rav = 0.01, then
E1/E2 = K = 2 / lg( 2 / Rav ) ,
Advanced Materials Research Vol. 1102 157

where the index 1 refers to the calculation, and the index 2 - to the measured characteristics, K -
correction factor, the values of which are shown in Fig. 5.

Fig.3. The dependence of the elastic modulus of the Fig.4. Dependence of the strength of the fibers on
fibers on the samples thickness the sample thickness
1 - 65 mm (), 2 - 135 mm ( ), 3 - 200 mm () 4 - 280 mm ().
The horizontal dashed lines - passport Young's modulus and strength of the filaments.

In [2] proposed an amendment to the strength of


rings, which take into account their bending in the
area of half-disks connectors. On the basis of the
beam theory, the correction is expressed by the
following formula:

= /(2Rav).

More rigorous calculation using the theory of


elasticity of anisotropic bodies shows that the
The dependence of the correction factor K of
correction should be increased by 10-12% [3,4].
the relative thickness of the samples / Rav
The only feature which is independent of the
geometric dimensions of the rings when tested using
158 Metal, Ceramic and Composite Materials

Fig. 7. Dependence of the elastic modulus of the


Fig. 6. The dependence of the maximum
fibers in the composition of the composite on the
deformation of rings on different diameter. logarithm of the relative thickness of lg(/Rav)
determined by different methods (1 - stretching half-
Notations are the same as in Fig. 3 and 4
disks, 2 - acoustic resonance method)

the half-disks is the ultimate deformation of samples that can be seen from Fig. 6. This apparently
explains the fact that the destruction begins from the inner layers of the ring, the most heavily
loaded [5-7]. At the moment of achievement in their extreme characteristics of the material
destruction spread over the section of the ring like an avalanche.
Independent determination of the values of the elastic modulus of the Rings was also carried out
using acoustical technique by measuring the resonance frequency of the sample. The results
presented in Fig. 7, indicate that under uniform stress state modulus does not depend on dimensions
of the samples.
Knowing the value of the breaking strain measured for rings of different diameters and
thicknesses in experiments on stretching them with a half-disks (Fig. 6) 2.7% and measuring
the elastic modulus of using acoustical technique (E = 59.2 GPa) You can reliably determine the
average strength of the material from the Hooke's law:
0.027 59 1.60GPa .

Conclusions
As seen from the results of the work shows the possibility of the test using half-disks fibreglasses
ring samples in a wide range of diameters and thicknesses of the samples and the calculation of their
elastic-strength properties when used independent methods for the determination of the elastic
modulus of the samples. The authors believe that we have to turn to the scale factor when we do not
understand the reasons depending on the size of the samples and do not have the correct testing
procedures. Here is a test methodology that allows you to correctly assess the strength of the
material reinforced winding fiberglass composed of thick rings and thereby eliminate the need for
recourse to the concept of the scale factor.

Acknowledgement
This work was supported by the Ministry of education and science of Russia under grant number
7.2122.2014/.
Advanced Materials Research Vol. 1102 159

References
[1] A.M. Kuperman, R.A. Turusov, Study of relaxation properties of reinforced plastics for half-
disks tension of ring specimens. Mechanics of Composite Materials, Riga Vol. 48, No 3 (2012)
[2] Kuperman A. M., Turusov R. A., Gorenberg A.Ya. Study of elastic and strength properties of
hybrid and gradient polymer composites. Composites: mechanics, computations, applications, an
international journal. v1.i4.50, pp. 361-373
[3] S.P. Timoshenko, Theory of oscillations in engineering. oscow, GTTI, 1932, 344 p.
[4] Andreev V.I. Some problems and methods mechanics of inhomogeneous bodies., Moscow,
ASV, 2002, 286 p.
[5] E.S.Zelenski, A.M.Kuperman, L.V.Puchkov, S.L. Bazhenov, The effect of the preparation and
testing conditions of laboratory unidirectional FRP- S.L. spesimens on the strength measurement
results. Proc.ICCM-VII. Peking, (1989) 495 p.
[6] V.I. Andreev, R.A.Turusov, Thermal strength of adhesion bond, Applied Mechanics and
Materials, Vols. 670-671 (2014) pp. 153-157
[7] R.A.Turusov, L.I. Manevich, Contact_Layer Method in Adhesive Mechanics: Adhesive Strength
during Normal Detachment. Polymer Science, Series D. Glues and Sealing Materials, Vol. 3, No. 3,
(2010) pp. 159169
Advanced Materials Research Vol 1102 (2015) pp 161-168 Submitted: 2015-01-04
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.161

The effects of Ca(OH)2 on the steam gasification process of acid-washed


and raw lignite
Chao Shuaia, Song Hub,*, Jun Xiangc, Sheng Sud, Lushi Sune, Yi Wangf
State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan
430074, China
a
shuaichao@hust.edu.cn, bhusong_hust@163.com, cxiangjun@hust.edu.cn,
d
susheng@hust.edu.cn, esunls_hust@163.com, falenwang@hust.edu.cn

Keywords: lignite; Ca(OH)2; steam gasification; tar reforming; WGS reaction

Abstract. The effect of Ca(OH)2 prepared by the wet-mixing method on acid-washed and raw
lignite steam gasification process at 800C were analyzed in a spout-fluid bed reactor. The addition
of Ca(OH)2 on acid-washed lignite obviously increased the coal conversion, gas yield and reduced
the tar yield, and the Ca(OH)2 addition on raw lignite significantly reduced the tar yield but just
slightly increased the coal conversion and gas yield. As a catalyst, Ca(OH)2 can promote the tar
steam reforming in the gasification process. Based on GC-MS data, it can be deduced that Ca(OH)2
has different catalytic activity on the steam reforming of tar with different molecular structures.
Aliphatic hydrocarbons, Class 2 and Class 5 tars were clearly catalytic reformed, while the
concentration of Class 4 tars increased. The CO concentration decreased and the CO2 and H2
concentration increased with the addition of Ca(OH)2 on both acid-washed and raw lignites, which
was mainly caused by the catalytic effect of Ca(OH)2 on WGS reaction. The effect of Ca(OH)2
addition on acid-washed lignite was more significant than on raw lignite, and no adsorption of CO2
was detected in the process by using active carbon as a model compound.
1. Introduction
Coal is the most abundant and geographically most widely distributed fossil fuel in the world,
and gasification with steam has been considered as one of the promising clean coal technologies to
provide energy and feedstock. Lignite is particularly suitable for the utilization by gasification
technology for its high gasification reactivity [1].
The current steam gasification process mostly carries out at very high temperature, above
1000C, but higher temperature gets lower energy efficiency. Catalysts are needed to accelerate the
gasification process and lower the reaction temperature simultaneously [2], and calcium based
catalysts (CaO and Ca(OH)2) have been got abroad attention for they can be derived from limestone
or dolomite. Ohtsuka et al. [3] studied the steam gasification reactivity of 16 coals in the presence of
Ca(OH)2 prepared by wet-mixing method in a thermogravity (TG), the result showed that the
calcium greatly enhanced the gasification rate. CaCO3 added by ion-exchange or kneading method
also showed considerable catalytic activity [4]. But it should be noted that the calcium based
catalysts added by physical mixing method found to have insignificant catalysis on gasification
process [4-5].
CaO is also generally used as a CO2 sorbent in the gasifier to capture CO2 as soon as it was
produced [6]. Beside this, CaO is an effective catalyst on WGS reaction, which can also improve
the hydrogen concentration [7-8]. Xu et al. [9] found that CaO acted as CO2 sorbent at 722C and
acted as catalyst at 795C.
162 Metal, Ceramic and Composite Materials

Interestingly, the calcium based catalysts can also promote the steam reforming of tar, which
can poison catalyst and block downstream equipment. Lots of studies observed that CaO was an
effective catalyst for decreasing the tar yield [10-12], but only limited researches studied the effect of
CaO on tar compounds with different molecular structures [13-14]. So its still necessary to get an
in-depth analysis for tar decomposition mechanism by calcium based catalysts.
The calcium based catalysts were mostly added by physical mixing method or used as bed
materials in these studies, resulted in limited catalytic effect on char gasification, litter work studied
their integrated effects. In the present work, Ca(OH)2 was loaded on two acid-washed and raw
lignites by wet-mixing method (similar to kneading), to investigate the its performance on the lignite
steam gasification in several aspects, including the char steam gasification rate, tar composition and
gas composition, furthermore, to compare the difference of Ca(OH)2 addition on acid-washed
lignite and raw lignite.
2. Experiment
2.1. Materials
Two kinds of typical Chinese lignite obtained from Inner Mongolia, Huolinhe (HLH) and
Ximeng (XM), and active carbon (AC) were sieved to 150-250 m for experiments. Proximate
analysis and ultimate analysis of the lignite were carried out using TG-2000 and EL-2 analyzer,
respectively. The results are given in Table 1.

Table 1- Proximate and ultimate analyses of HLH and XM lignite


Lignite Ultimate analysis (wt%, air-dry basis) Proximate analysis (wt%, air-dry basis)
C H N S Oa Moisture Ash Volatile Fixed carbon
HLH 42.34 2.52 0.90 0.41 20.78 10.34 22.71 37.12 29.83
XM 57.47 4.14 0.75 1.34 16.22 6.88 13.19 34.73 45.20
a By difference.

2.2. Sample preparation


In order to distinguish the difference of the added Ca(OH)2 and the inherent minerals on the
lignite gasification, acid washing was performed. The acid washed lignite was prepared by washing
the lignite with aqueous solution of HCl (6M, 150ml per 10g coal), and stirred for 24h at 50C. Then
it was filtered, washed with distilled water and dried at 105C for 24h. The most common catalytic
metals (K, Na, Ca, Mg and Fe) [15] in the ash of raw and acid-washed (AW) lignite were measured
by ICP-MS (Agilent, 1100 LC/MSD Trap), the results are shown in Table 2.
Table 2 - Contents of the main catalytic metal in the samples
Metal analysis (wt%, dry basis)
Sample
Na Mg Fe K Ca
HLH lignite raw 0.494 0.194 1.616 0.284 1.218
AW 0.077 0.061 0.181 0.060 0.120
XM lignite raw 0.409 0.258 2.821 0.441 1.495
AW 0.088 0.071 0.327 0.072 0.137
Ca(OH)2 (analytical pure) was added into the acid-washed or raw lignite by wet-mixing method
(denoted as AW-Ca and Raw-Ca, respectively): Ca(OH)2 was mixed with the sample in deionized
water and dried at 378K for 24h. The Ca added was 10 wt% as a metal. Ca(OH)2 was added into
active carbon by wet-mixing method, and 9.95wt% Ca was loaded on it (denoted as AC-Ca).
Advanced Materials Research Vol. 1102 163

2.3. Experimental Procedure


A schematic diagram of the experimental used in this study is presented in Fig. 1. The
experiments were conducted at atmospheric pressure and 800C, using a spout-fluid bed quartz
reactor. The inner diameter and the spout nozzle of the quartz reactor was 26mm and 6mm,
respectively. 6.60.05g/h water was evaporated to steam in the steam generator, and carried with
1501.5ml/min high purity nitrogen. When the reactor reached the experimental temperature,
20.01g sample (10.01g for active carbon sample) was fed into the reactor. After the reaction
proceeding for 10 min, the steam was switched off and the sample was held in the nitrogen
atmosphere for 5 min to sweep the steam and the product gas. Then the sample was rapidly cooled in
the nitrogen by moving the reactor out of the furnace. The weights of the apparatus were carefully
measured before and after each experiment.
The unreacted water and tar were captured by four traps connected in series, the first one was
empty mainly to capture water, the last three traps filled with a mixture of HPLC-grade chloroform
and methanol to capture tar. The calculation of tar content was described in detail by Min et al. [16].
The mass balance ratios (product/reactant) of all the experiments were between 84 % and 95 %.
5 6

7
8

3 4 10

11
N2
2

1 8

12 13 14 15

Fig. 1 Schematic diagram of experimental apparatus. 1, injection pump; 2, steam generator; 3, mass flow
controller; 4, temperature controller; 5, feeder; 6, ground glass joint; 7, silica wool filter; 8, heater bands; 9, reactor; 10,
quartz fruits; 11, furnace; 12, ice bath; 13, dry ice bath; 14, gas-washing bottle; 15, gas bag.

2.4. The quantification and characterization of products


The conversions of the samples were measured by a TG (Netsch STA449 F3). Gaseous products
were analyzed using an Agilent GC equipped with a molecular sieve column, a Plot U column and a
TCD detector. Components of tars were identified by GC/MS using an Agilent 7890A gas
chromatograph coupled with Agilent 5795C mass detector.

3. Results and discussion


3.1 Products yields
3.1.1. Coal conversion
The coal conversion is the percentage of the diminished weight to the initial weight of the coal
(dry-ash&catalyst-free basis). Fig. 2(a) shows the effect of Ca(OH)2 on the coal conversion of HLH
and XM at 800C. As shown in Fig. 2, after the addition of Ca(OH)2 on acid-washed lignite, the
char conversion of HLH and XM increased by 30.4% and 28.0%, respectively, which reveals that
Ca(OH)2 prepared by wet-mixing method has a good catalytic effect on the steam gasification of
both lignites.
164 Metal, Ceramic and Composite Materials

Comparing with the acid-washed lignite, it can be found that the conversion of raw HLH and
XM lignite was 19.59% and 24.59% higher, respectively, which shows the catalytic effect of
inherent minerals. By adding Ca(OH)2 on raw lignite, the conversion of HLH and XM increased by
6.37% and 5.61%, respectively, it suggests that the Ca(OH)2 loading on raw lignite still has catalytic
effect on the char conversion, however, the performance was decreased, which may be limited by
the amount of catalyst loading.
90 10 2.4
(a) HLH (b) HLH (c) HLH
80 XM XM XM
2.0
70 8

60 1.6
Conversion (%)

Gas yields (L)


Tar yield (%)
6
50
1.2
40
4
30 0.8
20 2
0.4
10
0 0 0.0
AW AW+Ca Raw Raw+Ca AW AW+Ca Raw Raw+Ca AW AW+Ca Raw Raw+Ca
Sample Sample Sample

Fig. 2 - Effect of Ca(OH)2 on (a) coal conversion, (b) tar yield and (c) gas yield at 800 C (dry ash&catalyst-free).
3.1.2. Tar yields
The effect of Ca(OH)2 on tar yields of HLH and XM is shown in Fig. 2(b). It was found that,
after the addition of Ca(OH)2 on acid-washed lignite, the tar yield of HLH and XM reduced by
62.3 % and 73.8 %, respectively. It is thought that the O2- ions which form the active sites on CaO
(decomposed from Ca(OH)2) have spatially diffuse electron clouds which can overlap with the
orbitals of the incoming molecules and disrupt the stability of the p-electron cloud of tar compounds
and leading to the splitting of the aromatic rings [14].
Comparing with the acid-washed lignite, the tar yield of raw HLH and XM lignite was reduced
by 16.38% and 48.14%, respectively, which was caused by the catalytic effect of inherent minerals
[17]. The tar decrease of XM was much higher than that of HLH, for the Mg, Fe, K and Ca contents
was much higher in XM.
By adding Ca(OH)2 on raw lignite, the tar yields of HLH and XM reduce by 82.51% and
85.53%, respectively, which reveals that Ca(OH)2 has significantly catalytic effect on tar steam
reforming.
3.1.3 Gas yields
Fig. 2(c) shows the effect of Ca(OH)2 on the gas yield of HLH and XM. As expected, the gas
yields of acid-washed HLH and XM both increased with the addition of Ca(OH)2, which was
consistent with the decrease of char and tar yields discussed above.
3.2. Catalytic reforming of tar
Tar compounds are generally divided into five classes based on their condensation behavior and
water solubility in downstream processes [13]: Class 1 (cant be detected by GC), Class 2 (tars
containing heteroatoms, highly water soluble compounds), Class 3 (light hydrocarbons with single
ring), Class 4 (two and three ring compounds), Class 5 (four to six ring compounds).
From the tar compositions detected by GC/MS, the aliphatic hydrocarbons (AH) were found,
which were also detected by other researchers [10, 18]. AH belongs to none of the five classes, which
was treated as an independent class in the present work.
Advanced Materials Research Vol. 1102 165

Class2 Class3 Class4 Class5 HA


90 90
(a) (b)
80 80

70 70

Concentration (%)
Concentration(%)

60 60

50 50
40 40

30 30
20 20

10 10
0 0
AW AW+ Ca Raw Raw+Ca AW AW+ Ca Raw Raw+Ca
Sample Sample

Fig. 3 Effect of Ca(OH)2 on the tar composition at 800C. (a) HLH; (b) XM.
Fig. 3 shows the effect of Ca(OH)2 on the tar composition at 800C. The AH tars were just found
in acid-washed sample of HLH, and completely reformed in the presence of Ca(OH)2 or in raw
sample, which was consistent with Jia et al.s [10] conclusion that AH tars were easily to be
reformed.
After loading Ca(OH)2 on acid-washed lignite, it was observed that the concentration of Class
2 tars of HLH completely converted, which of XM also dramatically decreased from 11.40% to
1.03%. And Class 5 tars of HLH decreased from 8.99% to 5.78%, whereas that of XM decreased
from 17.59% to 8.50%.
However, Class 4 tars were increased obviously for both lignites by adding Ca(OH)2, this
phenomenon was also found by other researchers [19], for the concentration of naphthalenes
(naphthalene and its derivations) were very high and very difficult to convert [20].
The concentration of Class 3 tars of HLH increased by adding Ca(OH)2, while that of XM
decreased. Class 3 tars could be decomposed by catalytic reforming or cracking and formed by the
decomposition of Class 4 and 5 tars at the same time [14], so the concentration of Class 3 tars was
largely related to the original tar compounds derived from different lignite.
The addition of Ca(OH)2 on raw lignite also showed similar variation trend for different
classes of tars.
3.3. Gas composition
3.3.1 The mechanism of Ca(OH)2 on gas concentration of active carbon
Steam gasification of lignite is a very complex process and char steam gasification is the key
reaction in the process. Thus, active carbon was selected as a model compound to study char steam
gasification, the important reactions involved are given below:
Char steam reaction:C+H2OCO+H2 (1)
Boudouard reaction: C+CO22CO (2)
WGS reaction: C+H2OCO2+H2 (3)
Carbonation reaction:CaO+CO2CaCO3 (4)
With the addition of Ca(OH)2 on active carbon, the H2 concentration increased from 51.19% to
60.24% at 800C, which may be explained by the adsorption of CO2 or the catalytic effect of
Ca(OH)2 on WGS reaction.
166 Metal, Ceramic and Composite Materials

3.3.1.1. Absorption of CO2


CaO can absorb CO2 though carbonation reaction (Eq. 4), in the atmospheric pressure, the
absorption of CO2 is favored to occur when the CO2 concentration is higher than ~ 20% at 800C
[21]. The average CO2 concentration of AC-Ca sample at 800C estimated to be 5.36%, which was
much lower than the equilibrium partial pressure, so little CO2 would be absorbed.
From Eqs. 1 - 3, it can be derived from these reactions that the ratio of (CO+2CO2)/H2 should
be 1 (the element oxygen and hydrogen in the product gas both come from H2O, 1 and 2 mol of H2
will be produced when producing 1 mol of CO and CO2, respectively), but if CO2 was absorbed (Eq.
4), the ratio would be lower than 1.
Table 4 shows the ratio of (CO+2CO2)/H2 at 700C and 800C. As shown in Table 4, it was
found that most of the ratio were close to 1 except that of AC-Ca sample at 700C, indicating that the
absorption of CO2 (Eq. 4) obviously occurred just at 700C and little CO2 was absorbed at 800C.
The observation was in agreement with the results of the thermodynamic equilibrium theory.
Table 4 The ratio of (CO+2CO2)/H2 in the gas composition of active carbon gasification
Sample AC-700C AC-Ca-700C AC-800C AC-Ca-800C
Ratio 1.07 0.82 1.02 0.97
3.3.1.2. Catalyze the WGS reaction
The WGS reaction (Eq. 3) leads to an increase in H2 and CO2 concentration, whilst a decrease
in CO concentration. It was already found that there was little adsorption of CO2 at 800C in the
present work, thus the catalytic effect on WGS reaction was thought to be the only one responsible
for the increase of H2 concentration at these temperatures.
3.3.2 The effects of Ca(OH)2 on gas composition
Fig. 4 shows the effect of Ca(OH)2 on gas concentration. The H2/CO ratio of acid-washed HLH
and XM lignites was 1.04 and 1.03, respectively, and with the addition of Ca(OH)2, that of HLH
and XM was 1.82, and 1.81, respectively. The average CO2 concentration of HLH and XM at 800C
in the reactor was much lower than the corresponding equilibrium partial pressure of CO2, thus little
CO2 would be absorbed. And the main reason causing the increase of H2/CO ratio should be the
catalytic effect of Ca(OH)2 on WGS reaction, and the increase of CO2 concentration when adding
Ca(OH)2 also supported this. The CH4 concentration decreased with the addition of Ca(OH)2, for
the the reforming of CH4 by Ca(OH)2.
Comparing with the acid-washed lignite, the CO and CH4 concentration decreased, whereas
the CO2 and H2 concentration increased, the variation was similar to the effect of Ca(OH)2 on
acid-washed lignite. But it can be seen that the CO2 and H2 concentration of raw lignite was higher
than that of AW-Ca lignite, while the CO concentration was lower, which suggests that the catalytic
effect of inherent minerals was higher than added Ca(OH)2.
By loading Ca(OH)2 on raw lignite, it was found that the CO and CH4 concentration decreased,
whereas the CO2 and H2 concentration increased, the H2/CO ratio of raw HLH and XM lignites was
2.58 and 2.06, respectively, and with the addition of Ca(OH)2, that of HLH and XM was 3.02, and
2.16, respectively. It suggests that the addition of Ca(OH)2 on raw lignite enhanced the performance
of WGS reaction, however, the effect was much lower comparing with the addition of Ca(OH)2 on
acid-washed lignite.
Advanced Materials Research Vol. 1102 167

70 70
(a) CO CO2 CH4 H2 (b) CO CO2 CH4 H2
60 60

50 50
Concentration (%)

Concontration (%)
40 40

30 30

20 20

10 10

0 0
AW AW+Ca Raw Raw+Ca AW AW+Ca Raw Raw+Ca
Sample Sample

Fig. 4 Effect of Ca(OH)2 on the gas composition at 800C. (a) HLH; (b) XM.

4. Conclusion
The work analyzed the effects of Ca(OH)2 addition by wet-mixing method on the steam
gasification of two acid-washed and raw lignites in a spout-fluid bed reactor. The following
conclusions can be got:
 The addition of Ca(OH)2 on acid-washed lignite greatly promoted the char conversion, tar
reforming and gas yield; and the addition of Ca(OH)2 on raw lignite just increased the coal
conversion and gas yield a little, while significantly increased the tar reforming.
 For acid-washed and raw lignites, the concentrations of aliphatic hydrocarbons, Class 2, 5 tars
decreased with the addition of Ca(OH)2, whereas Class 4 tars increased largely due to the low
reforming reactivity of the naphthalene.
 The Ca(OH)2 addition increased the H2/CO ratio of acid-washed lignite a lot, but just slightly
increased the H2/ratio of raw lignite. The increase of H2/CO mainly caused by the catalytic
effect of Ca(OH)2 on WGS reaction, not the adsorption of CO2.

Acknowledgment
This work was supported by the National Science Foundation of China (NSFC) (No.21176098,
51176062).

References
[1] S. Kajitani, H.-L. Tay, S. Zhang, C.-Z. Li, Mechanisms and kinetic modelling of steam
gasification of brown coal in the presence of volatilechar interactions, Fuel. 103 (2013) 7-13.
[2] Y. K. Kim, J. I. Park, D. Jung, J. Miyawaki, S. H. Yoon, I. Mochida, Low-temperature catalytic
conversion of lignite: 1. Steam gasification using potassium carbonate supported on perovskite oxide,
Journal of Industrial and Engineering Chemistry. 20 (2014) 216-221.
[3] Y. Ohtsuka, K. Asami, Steam gasification of coals with calcium hydroxide, Energy & Fuels. 9
(1995) 1038-1042.
[4] Y. Ohtsuka, K. Asami, Ion-Exchanged Calcium from Calcium Carbonate and Low-Rank Coals:
High Catalytic Activity in Steam Gasification, Energy & Fuels. 10 (1996) 431-435.
[5] N. Kayembe, A. H. Pulsifer, Kinetics and catalysis of the reaction of coal char and steam, Fuel.
55 (1976) 211-216.
168 Metal, Ceramic and Composite Materials

[6] C. C. Dean, J. Blamey, N. H. Florin, M. J. Al-Jeboori, P. S. Fennell, The calcium looping cycle
for CO2 capture from power generation, cement manufacture and hydrogen production, Chemical
Engineering Research and Design. 89 (2011) 836-855.
[7] A. K. Dalai, E. Sasaoka, H. Hikita, D. Ferdous, Catalytic Gasification of Sawdust Derived from
Various Biomass, Energy & Fuels. 17 (2003) 1456-1463.
[8] L. Wei, S. Xu, L. Zhang, C. Liu, H. Zhu, S. Liu, Steam gasification of biomass for hydrogen-rich
gas in a free-fall reactor, International Journal of Hydrogen Energy. 32 (2007) 24-31.
[9] G. Xu, T. Murakami, T. Suda, S. Kusama, T. Fujimori, Distinctive Effects of CaO Additive on
Atmospheric Gasification of Biomass at Different Temperatures, Industrial & Engineering Chemistry
Research. 44 (2005) 5864-5868.
[10] Y. Jia, J. Huang, Y. Wang, Effects of Calcium Oxide on the Cracking of Coal Tar in the Freeboard
of a Fluidized Bed, Energy & Fuels. 18 (2004) 1625-1632.
[11] D. Sutton, B. Kelleher, J. R. H. Ross, Review of literature on catalysts for biomass gasification,
Fuel Processing Technology. 73 (2001) 155-173.
[12] L. Devi, K. J. Ptasinski, F. J. J. G. Janssen, A review of the primary measures for tar elimination in
biomass gasification processes, Biomass and Bioenergy. 24 (2003) 125-140.
[13] J. Udomsirichakorn, P. Basu, P. A. Salam, B. Acharya, Effect of CaO on tar reforming to
hydrogen-enriched gas with in-process CO2 capture in a bubbling fluidized bed biomass steam
gasifier, International Journal of Hydrogen Energy. 38 (2013) 14495-14504.
[14] C. A. Jordan, G. Akay, Effect of CaO on tar production and dew point depression during
gasification of fuel cane bagasse in a novel downdraft gasifier, Fuel Processing Technology. 106
(2013) 654-660.
[15] G. Skodras, G. P. Sakellaropoulos, Mineral matter effects in lignite gasification, Fuel Processing
Technology. 7778 (2002) 151-158.
[16] Z. Min, M. Asadullah, P. Yimsiri, S. Zhang, H. Wu, C.-Z. Li, Catalytic reforming of tar during
gasification. Part I. Steam reforming of biomass tar using ilmenite as a catalyst, Fuel. 90 (2011)
1847-1854.
[17] D. M. Quyn, H. Wu, J.-i. Hayashi, C.-Z. Li, Volatilisation and catalytic effects of alkali and
alkaline earth metallic species during the pyrolysis and gasification of Victorian brown coal. Part IV.
Catalytic effects of NaCl and ion-exchangeable Na in coal on char reactivity, Fuel. 82 (2003)
587-593.
[18] T. Zhu, S. Zhang, J. Huang, Y. Wang, Effect of calcium oxide on pyrolysis of coal in a fluidized
bed, Fuel Processing Technology. 64 (2000) 271-284.
[19] L. Devi, K. J. Ptasinski, F. J. J. G. Janssen, S. V. B. van Paasen, P. C. A. Bergman, J. H. A. Kiel,
Catalytic decomposition of biomass tars: use of dolomite and untreated olivine, Renewable Energy.
30 (2005) 565-587.
[20] R. Coll, J. Salvad, X. Farriol, D. Montan, Steam reforming model compounds of biomass
gasification tars: conversion at different operating conditions and tendency towards coke formation,
Fuel Processing Technology. 74 (2001) 19-31.
[21] N. H. Florin, A. T. Harris, Enhanced hydrogen production from biomass with in situ carbon
dioxide capture using calcium oxide sorbents, Chemical Engineering Science. 63 (2008) 287-316.
Advanced Materials Research Vol 1102 (2015) pp 169-172 Submitted: 2015-01-24
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.169

The analysis of the effect of surface stresses at nanoscale


Ou Zhiying1,a, Wu Yawen1,b
1
School of Science, Lanzhou University of Technology, Lanzhou, China
a
zhiyingou@163.com, bwuyawen32@126.com

Keywords: Surface energy; Effect; Nanoscale; Cosine function.

Abstract: Based on classical elasticity theory, the effects of surface stresses on the nano-sized
contact problem in an elastic half-plane which contains a nano-cylindrical hole are analyzed.
Meanwhile, the effects of surface energy of the contact nanosized surface are considered. The
complex variable function method is applied to derive the fundamental solution of the contact
problem. As example, the deformation induced by a distributed traction of cosine function on the
nanosized surface is analyzed in detail. The results tell some interesting characteristics in contact
mechanics, which are different from those in classical elasticity theory.

Introduction
The study of the mechanical behavior of nanosized inhomogeneous materials containing
inclusions or holes has attracted great interest in materials field, solid state physics, and
nanomechanics. Based on the studying about the potential applications of nanostructured materials,
researchers recently have considered the effect of surface energy on the deformation field of elastic
materials containing nano-inclusions or nano-holes by using the theory of surface elasticity [1-8].
Jammes et al. [9] used semi-analytical method to solve the problem that the half-plane contains
multiply, circular nanoinclusions. Zhao and Rajapaski [10] derived the fundamental solution of an
elastic layer bonded to a rigid substrate with surface effects. However, none of them examined the
effect of surface stress on the deformation behavior of the nanosized surface of an elastic half-plane
containing an inhomogeneity.
The purpose of this paper is to analyze the effect of surface stress on the deformation behavior
of an elastic half-plane containing a cylindrical hole near the surface of the elastic half-plane. The
elastic half-plane is applied to a uniform, distributed loading over a segment of its surface and a
surface tension at the cavity boundary.
1. Mathematical Formulation
Considering a semi-infinite, isotropic elastic half-plane, we refer to a Cartesian coordinate
system, as shown in Fig. 1. The elastic half-plane contains a nanosized cylindrical hole of radius r
0
and the problem of the surface effect at the cavity boundary are considered. The distance of the
r
holes centre to the surface of the elastic half-plane is h ; and the surface loading of P is
subjected to the straight surface of the elastic half-plane in the portion of b x b , symmetric to
the y -axis. The shear modulus and Poissons ratio of the elastic half-plane are and , respectively.
The surface energy is 0 .

Fig.1 Half-plane with circular cavity


170 Metal, Ceramic and Composite Materials

2. Principle of Superposition
The problem can be solved from the two problems denoted by I and II below through the
principle of superposition [4]. The superscripts of I and II refer to the field variables associated with
the solution of sections I and II, respectively.

ij = ij + ij , ui = ui + ui . (1)

2.1. Section I: Homogeneous half-plane under a surface distributed cosine pressure


In the case where there is a uniform distributed pressure P acting over the region x b , we
can obtain the stresses [5] as

p y 1 sin ( a + a ) sin ( a + a )
xx ( x, y ) = + y
cos x e d
, (2)
0 + 1 + 1 1

p y + 1 sin ( a + a ) sin ( a + a )
yy ( x, y ) =

y
+ cos x e d , (3)
0 + 1 + 1 1

p y s in ( a + a ) s in ( a + a )

(x, y ) = + s in x e
y
d , (4)
xy
0 + 1 +1 1

0 (1 )
where = . The corresponding displacements are expressed as

p y + 2 1 sin ( a + a ) sin ( a + a ) , (5)


u ( x, y ) =
2 ( + 1)
y
+ sin x e d + C3
+1 1
0

p0 2 (1 ) + y sin ( a + a ) sin ( a + a ) . (6)


w ( x, y ) = ( + 1) + y
cos x e d + C4
2 0
+ 1 1

Assuming that the point of ( 0, 0 ) has no displacement in the x direction, one has C3 = 0 .
On the contact surface y = 0 , the stress and displacements are given by

p 1 sin ( a + a ) sin ( a + a ) , (7)


yy ( x, 0 ) =

+ cos x d
0 + 1 +1 1

2.2. Section II: a cylindrical hole embedded in the elastic half-plane


To solve the problem, the proper complex potential based on the Laurent series introduced as
below

= ( ) = ak k + bk k , = ( ) = ck k + d k k , (8)
k =0 k =1 k =0 k =1

where ak , bk , ck , d k are coefficients to be determined.


The stress functions ( ) and ( ) can be easily found to be

( ) 2i 2 , ( ) i 2 ,
= 2i (1 + 2 ) + 2i + = 3i (1 + 2 ) + 2i 2 + i 2 +
2i (9)
+
L L 2
Advanced Materials Research Vol. 1102 171

where L = 2 h 0 . Then, we obtain as


r (1 2 )(1 4 )

4 zh (1 + 2 )(1 2 ) h 2 (1 + 2 )(1 2 )
2 2

( z ) =
, (10)
L 4i L
h (1
2 2 2
) + z (1 + 2 2 2
) h (1
2
)
2 2
+ z (1 +
2
)
2 2

2
4h (1 2 ) 3iz (1 + 2 ) h 2 (1 2 ) + 5 z 2 (1 + 2 ) .
2 2 2
(11)
( z ) = 1 + (1 + 2 ) L 2 zh 2 (1 2 )
2
i 2
h 2 (1 2 ) + z 2 (1 + 2 )
2 2
h (1
2 2 2
) + z (1 + 2 2 2
)

Letting y = 0 , the surface stresses are simplified to

x2 (1+ 2 ) h2 (1 2 ) h2 (1 2 ) 3x2 (1+ 2 )


2 2 2 2
, (12)

( x,0) = 8L h (1+ 2
)(1 ) 2 2
2
8L h (1+ 3 2
)(1 )
2 4
3
h2 (1 2 )2 + x2 (1+ 2 )2 h2 (1 2 )2 + x2 (1+ 2 )2
xx

( x, 0 ) 3h 2 (1 2 ) x 2 (1 + 2 )
2 2
yy
= x (1 + 2 2 2
)
8 Lh (1 + 2 )(1 2 )
2 2
h 2 (1 2 )2 + x 2 (1 + 2 )2

h 2 (1 2 ) 3x 2 (1 + 2 ) x 2 (1 + 2 ) h 2 (1 2 )
2 2 2 2
, (13)
+ h (1
2
)
2 2
3
+ 2
h 2 (1 )
2 2
+ x 2 (1 + )
2 2 h 2 (1 )
2 2
+ x 2 (1 + 2 2
)

x h2 (1 2 )2 h2 (1 2 )2 + x2 (1+ 2 )2
( x,0) h2 (1 2 ) + x2 (1+ 2 ) 2xh (1 2 ) (1+ 2 )
2 2 2 2
xy
= 4 + 8x
4Lx (1+ 2 )
3 3 3
h2 (1 2 )2 + x2 (1+ 2 )2 h2 (1 2 )2 + x2 (1+ 2 )2 (1+ 2 )3

x 4 (1 + 2 ) + 5 x 2 h 2 (1 + 2 ) (1 2 ) 2h 4 (1 + 2 ) + x 4 (1 + 2 ) x 2 h 2 (1 2 ) .
2 2 2 4 4 2

+ (14)
3 3
h 2 (1 2 )2 + x 2 (1 + 2 )2 h 2 (1 2 )2 + x 2 (1 + 2 )2

3. Numerical Calculation
It is instructive to examine the influence of the surface stress on the stresses or displacements
on the contact nanosized surface and compare them with those in classical elasticity theory. In what
0
follows, we set P = cos x , = 26.3MPa , = 0.4 , 0 = 0.1J / m 2 , h / r = 1.5 and b / r = 3 .
r
xx and yy depend on the parameter of s / b . Thus the stresses obtained from the
superposition are functions of the size of the cylindrical hole and the nanosized surface stress on the
surface of the half-plane, where we set x / r = 1 . According to Wang gangfeng [5], the absolute
values of s / b for the metals are about 0.0 to 2.0 , where the solution of s / b = 0 is consistent
with the classical elastic result. It is also found from Fig. 2 and Fig. 3 that the normal and the tangential
stresses change smoothly, as is different from the lower two stresses which predicted by the classical elasticity.
15 10
s/a=0 s/b=0
s/a=1 s/b=1

5
10

0
xx/P

yy/P

5
-5

0 -10
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
r r

Fig. 2 Distribution of normal stress Fig. 3 Distribution of tangential stress


172 Metal, Ceramic and Composite Materials

15
80 r=x
x/r=1 r=x/2
x/r=2 10 r=x/3
x/r=3 r=x/4
60 x/r=4 5

yy/P
40
xx/P

-5

-10
20

-15

0 -20
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
r r

Fig. 4 Distribution of normal stress on the Fig. 5 Distribution of tangential stress on the
surface of the different coordinates surface of the different coordinates
Fig. 4 and Fig. 5 show the various stresses on the surface for s / b = 1 with different
coordinates of the half-plane. It can be seen that the stresses drastically reduced when r is small
enough. In the Fig. 5, the normal stress increases slowly with the growth of r around r = 0.2 ,
while the tangential stress reduces slowly around r = 0.3 .

Summary
Using the theory of the surface elasticity and the complex variable method, the effect of the
surface stress on the deformation of an elastic half-plane containing a cylindrical hole which
subjected a radial stress was analyzed when the nanosized surface of the elastic half-plane was
subjected to a surface loading. It is found that the surface elasticity theory illuminates some
interesting characteristics of contact problems at nanoscale, which are distinctly different from the
classical solutions of elasticity without surface effect. Therefore, the effects of surface stresses
should be accounted for.

References
[1] Z. Y. Ou, S. D. Pang, Fundamental solutions to Hertzian contact problems at nanoscale. Acta
Mech. 224 (2013) 109-121.
[2] A.Verruijt, Deformations of an elastic half plane with a circular cavity. Int. J. Solids Struct. 35
(1998) 2795-2804.
[3] Z.Y. Ou, G.F. Wang, T. J. Wang, Effect of residual surface tension on the stress concentration
around a nanosized spheroidal cavity. Int. J. Engine Res. 46 (2008) 475485.
[4] K. Amir, Miri, Effect of surface stress on the deformation of an elastic half-plane containing a
nano-cylindrical hole under a surface loading. J. Compos. Tech. 8 (2011) 1-6.
[5] G. F. Wang, X. Q. Feng, Effects of surface stresses on contact problems at nanoscale. J. Appl.
Phys. 101 (2007) 013510-013510-6.
[6] G. F. Wang, X. Q. Feng, Effects of the surface elasticity and residual surface tension on the
natural frequency of microbeams. Appl. Phys. Lett.90 (2009) 231904-231904-3.
[7] M. E. Gurtin, A general theory of curved deformable interfaces in solids at equilibrium. Philos.
Mag. 78 (1998) 10931109.
[8] V. B. Shenoy, Size-dependent rigidities of nanosized torsional elements. Int. J .Solids. Struct. 39
(2002) 4039-4052.
[9] M. Jammes, S. G. Mogilevskaya, S. L. Crouch, Multiple circular nano-inhomogeneities and/or
nano-pores in one of two joined isotropic elastic half-planes. Eng. Anal. Bound. Elem. 32 (2009)
233248.
[10] X. J. Zhao, Rajapakse, Analytical solutions for a surface-loaded isotropic elastic layer with
surface energy effects. Int. J. Eng. Sci. 47 (2009) 14331444.
Advanced Materials Research Vol 1102 (2015) pp 173-182 Submitted: 2014-12-25
(2015) Trans Tech Publications, Switzerland Accepted: 2015-02-04
doi:10.4028/www.scientific.net/AMR.1102.173

EXPERIMENTAL STUDY OF THE DEFORMATION BEHAVIOUR AND


MECHANICAL PROPERTIES OF FRESH REINFORCED BAMBOO

L.Gyansah1, a and A. M. Abd El-Aty 2, b


1, 2
Shenyang National Lab. for Materials Science
1
Institute of Metal Research, Chinese Academy of Sciences, Shenyang, China
a
lawrence14b@imr.ac.cn or gyansahaclass@yahoo.com,
b
abdelaty14b@imr.ac.cn or abdelaty85@gmail.com

KEYWORDS: Reinforced Fresh Bamboo, Failure Stress to Weight ratio, Strength to Volume Ratio,
Load to Height Ratio, Specific Compressive Strength.

Abstract. This paper investigates the deformation behavior of reinforced fresh bamboo subjected to
static-compressive loading condition. Crushing strength test was performed using Uniaxial
Compression Machine with maximum loading capacity of 1500 kN. The data is plotted as strength
to weight ratio versus height, strength to volume ratio versus height, failure stress versus height,
load versus time of failure and load to height ratio versus time of failure. Result shows that increase
in height of the fresh reinforced bamboo increases the strength and vice versa. The moisture content
also has significant influence on the strength of the fresh reinforced bamboo. The moisture content
reduced the strength of the fresh reinforced bamboo by 21.8 %. This percentage is really significant
in structural analysis since it can cause catastrophic failure to structures. Concrete is significant in
reinforcing bamboo structures for a better strength. The crushing stresses for fresh reinforced
bamboo of heights 250 , 210 ,170 , 130 and 90 mm were found to be 26.09 , 25.52 , 24.85 , 21.97 ,
21.86 MPa respectively. The failure stress of the bamboo is about say 1.26 M times the weight of
the bamboo per square meter. The specific compressive strength which explains the bending
moment per kilogram of the bamboo specimen also increases with increases in height. Failure stress
to weight ratio decreases with increases with the height of the bamboo. Strength to volume ratio
also decreases with increases with the height from the bamboo.

Introduction

Bamboo is a woody perennial evergreen plant that is actually part of the true grass family. Although
they can grow to towering heights, bamboo is not actually considered a tree. Bamboo belongs to the
family Poaceae, subfamily Bambusoideae, tribe Bambuseae. Bamboo along with being one of the
tallest grasses in the world is also the fastest growing plant in the world, capable of growing 3 to 4
feet per day. Bamboo mainly consists of the roots, culm and leaves. The bamboo culms are the main
components that can be used as building materials and it is one of the oldest used by humankind.
The culms consist of cellulose fibers which run the length of the culm carrying nutrients between
roots and leaves and the light coloured lignin. Bamboo is considered a composite material because
it consists of cellulose fibers imbedded in a lignin matrix [1]. It is light, strong and easy to grow. In
spite of these advantages, it is widely perceived as a temporary, poor mans material. However
bamboo has been used widely for household products and extended to industrial applications due to
advances in processing technology and increased market demand. In Asian and African countries,
bamboo has been used for household utilities and also widely used in building applications, such as
flooring, ceiling, walls, windows, doors, fences, housing roofs, trusses, rafters and purlins etc.
Bamboo's resistance to stretching and its ability to support weight are at least double that of other
kinds of wood, making bamboo an ideal material for houses, scaffolding, supporting pillars, and
work sheds [2]. L. Gyansah, A.S. Akinwonmi and M. Affam published the crushing strength of
174 Metal, Ceramic and Composite Materials

bamboo and the Fracture Behaviour of Fresh Bamboo under Uniaxial Compressive Loading
Condition [3]. In the paper, the crushing stresses for fresh bamboo for a height of 250, 210, 170,
130, and 90 mm were 51.3093, 71.1447, 74.5867, 79.5905 and 85.2036 MPa, respectively. This
reveals that bamboo has strength much stronger than other wooden structures and even concretes.
Again it is a renewable natural resource, environmentally safe, stronger and has a longer life span
than wood. In addition to the above, bamboo is economically cheaper than wood and reinforced
steel rods. Bamboo is unique in that it is strong in both tension and compression. While tensile
strength remains the same throughout the age of the bamboo plant, compressive strength increases
as it gets older [4]. L. Gyansah and S. Kwofie, also published data on the effect of stress
concentration on the performance of bamboo using the notched and the un-notched specimen [5]. In
their research, it was obvious that the higher the notched angle within a bamboo, the lower the
crushing loads. The variation in the notched angles also had significant effect on the crushing time.
A notched-angle of 20, 30, 60, 80 and 90 gave a crushing load of 8.0, 5.55, 3.33, 1.33 and 0.93 kN
respectively and a crushing time of 42.46, 35.78, 21.89, 18.02 and 10.30 seconds, respectively,
explaining the fact that the crushing time or time of failure of the bamboo decreases with increases
in the notched angle.

For this research work, the reinforcement to be used in the bamboo is concrete instead of the normal
steel rods that are widely known in application in the construction industry. The importance of
concrete in modern society cannot be underestimated. Look around you and you will find concrete
structures everywhere in buildings, roads, bridges, and dams. There is no escaping the impact
concrete makes in everyday life. Concrete is a composite material which is made up of filler and
binder. The binder (cement paste) glues the filler together to form a synthetic conglomerate. The
constituents used for the binder are cement and water, while the filler can be fine or coarse
aggregate. Cement and water are the two ingredients that do all the chemical reaction process. The
gravel and sand give strength. The proportions of each ingredient (cement, sand, gravel and water)
will determine how strong the concrete will be [6]. Concrete essentially has a high level of
compressive strength, while the tensile strength of concrete is relatively very weak. As concrete can
crack under its own weight, it needs to be reinforced. It is generally reinforced using steel bars or
fiber and iron mesh. To reduce the tensile strength of concrete, it is also pre-stressed with the use of
steel cables. The deciding factor for strength is also inherently related to the proportion and ratio of
water and cement, the type of cement used and the strength of used aggregate. The correct
proportion of ingredients, placement, and curing are needed in order for these properties to be
optimal. Good-quality concrete has many advantages that add to its popularity. First, it is
economical when inputs are readily available. Concrete's long life and relatively low maintenance
requirements increase its economic benefits. Concrete is not as likely to rot, corrode, or decay as
other building materials. Concrete has the ability to be molded or cast into almost any desired
shape. Building of the molds and casting can occur on the work-site which reduces costs. Concrete
is a non-combustible material which makes it fire-safe and able to withstand high temperatures. It is
resistant to wind, water, rodents, and insects. Hence, concrete is often used for storm shelters.
Concrete does have some limitations despite its numerous advantages. Concrete has a relatively low
tensile strength (compared to other building materials), low ductility, low strength-to-weight ratio,
and is susceptible to cracking [7]. Despite all the numerous benefits of the applications of bamboo
and concrete, accidents do occur with the use of this material as well as its general constructional
usage. There are also certain limitations of the use of bamboo in construction. The starchy interior is
attractive to insects. In addition, because bamboo has a slick waterproof coating, it cannot be
painted. However, this coating allows bamboo to be used as water pipes. To determine the
feasibility of again using bamboo to reinforce concrete, the problem is, however, that bamboo soaks
up the water in the concrete, causing the bamboo to swell then shrink, the process of which can
break the concrete. In addition, adhesion between the bamboo and the concrete is very poor [8].
The relative density of bamboo ranges between 0.5 and 0.8 g/cm3 (oven-dry weight). This value
increases from the central parts to the peripheral parts of the culm and from the bottom to the top
Advanced Materials Research Vol. 1102 175

[9]. The studies on bamboos mechanical properties are commonly based on laboratory tests of the
strength of bamboo (tensile strength, bending strength, compression strength, shear strength and
modulus of elasticity) [10, 11].These tests show remarkable differing values when changing
species, ages, moisture content, locations, soil and climatic conditions. The variation of mechanical
properties is similar to wood, but even more remarkable. This is one of the reasons why bamboo is
difficult to be used in the construction industry. However bamboo can be used as a plant in agro-
forestry and in ornamental horticulture and also as an engineering material under various categories.
The durability of bamboo depends strongly on the preservative treatment methods in accordance
with basic requirements: its chemical composition should not have any effect on the bamboo fiber
and once injected it must not be washed out by rain or humidity. The preservative can be applied
using simple systems such as leave transpiration, immersion, and impregnation [12]. A Composite
in engineering sense is any materials that have been physically assembled to form one single bulk
without physical blending to foam a homogeneous material. The resulting material would still have
components identifiable as the constituent of the different materials. One of the advantages of
composite is that two or more materials could be combined to take advantage of the good
characteristics of each of the materials. Composites are made up of individual materials referred to
as constituent materials. There are two categories of constituent materials: matrix and
reinforcement. At least one portion of each type is required. The matrix material which in this case
is bamboo surrounds and supports the reinforcement materials (concrete) by maintaining their
relative positions. The reinforcements impart their special mechanical and physical properties to
enhance the matrix properties (bamboo).A synergism produces material properties unavailable from
the individual constituent materials, while the wide variety of matrix and strengthening materials
allows the designer of the product or structure to choose an optimum combination. Engineered
composite materials must be formed to shape. The matrix material can be introduced to the
reinforcement before or after the reinforcement material is placed into the mould cavity or onto the
mould surface. The matrix material experiences a melding event, after which the part shape is
essentially set. Depending upon the nature of the matrix material, this melding event can occur in
various ways such as chemical polymerization or solidification from the melted state. The physical
properties of composite materials are generally not isotropic (independent of direction of applied
force) in nature, but rather are typically orthotropic (different depending on the direction of the
applied force or load). For instance, the stiffness of a composite panel will often depend upon the
orientation of the applied forces and/or moments. Panel stiffness is also dependent on the design of
the panel. For instance, the fibre reinforcement and matrix used the method of panel build,
thermoset versus thermoplastic, type of weave, and orientation of fibre axis to the primary force. In
contrast, isotropic materials (for example, aluminium or steel), in standard wrought forms, typically
have the same stiffness regardless of the directional orientation of the applied forces and/or
moments [13]. The relationship between forces/moments and strains/curvatures for an isotropic
material can be described with the following material properties: Young's Modulus, the Shear
Modulus and the Poisson's Ratio, in relatively simple mathematical relationships. For the
anisotropic material, it requires the mathematics of a second order tensor and up to 21 material
property constants [14]. Therefore for the purposes of design and reliability, it is imperative to study
and understand the behaviour of reinforced fresh bamboo subjected to uni-axial compressive
loading. In this project, emphasis will be on fresh reinforced bamboo (composite material)
subjected to Uni-axial compressive loading using the Uni-axial compression testing machine. The
research is also conducted to study the behaviour of reinforced bamboo under uni-axial compressive
loading and to investigate into the effect of variation of length on the crushing strength of the
composite material.
176 Metal, Ceramic and Composite Materials

Fig. 2.4 Bambusa Vulgaris

Materials and Methods

Materials

The type of bamboo specie used for this research project is Bambusa Vulgaris. This is because it is
the most common type of bamboo specie in Africa and especially at the W.V.S. Tubman
Universitys campus. It can grow up to a height of 12 m, and a thickness of 8 cm. This type of
bamboo comprises the following; 5 % starch, 2 % deoxidized saccharide, 3 % fat and 5.5 % protein.
The bamboo specimen involved both fresh reinforced with concrete with different lengths. The
materials for concrete reinforcement within the bamboo specimen were cement, fine aggregate
(sand), coarse aggregates (gravels) and water, measured in their right proportions. The size of the
fine aggregate consisted of sand which is able to pass through a 5 mm BS sieve at the Mechanical
Laboratory-W.V.S. Tubman University and Shenyang National Lab. for Materials Science, whereas
the coarse aggregate was also graded using sieve complying to BS 410:1976, the methods of test
being outlined in BS 812: part 1:1975. Fig. 3.1 shows the sectional view of the composite material.

Fig. 3.1 Sectional View of the Composite Material (i.e. Reinforced Bamboo Specimen)

Sample Preparation

Fresh bamboos as received from the forest were free from insect infestation. They were then dried
for two weeks at a temperature range of 30-33 and cut into the needed lengths such as 250 mm,
210 mm, 170 mm, 130 mm and 90 mm with a cross-cut saw (i.e. interval of 40 mm). The dried
bamboo specimens were then polished with P 1200, P 600, P 400, P 240, P 180 Abrasive paper.
These specimens were carefully cut such that, the node lies at the centre of the height. The external
and internal diameters Do and Di (i.e. Fig. 3.1 Schematic Diagram of Bamboo) were measured
using a micrometer screw gauge and found to be between 82 to 86 mm and also between 72 to 74
mm respectively. The specimens were finally weighed on a scale as shown in Fig. 3.2. (a).
Advanced Materials Research Vol. 1102 177

Fig. 3.2.(a) Fresh Reinforced Bamboo on a Scale

Uni-Axial Compressive Test


This test was carried out using the equipment known as the Uni-axial Compression Machine at
room temperature. The experiment involves placing the specimens on the lower platen of the
compressive testing machine with maximum loading capacity of 1500 kN and crushing them to
know the compressive load. The fresh bamboo specimens were weighed in order to take their initial
weight as mass of dried bamboo (WD) and mass of fresh bamboo (WF) respectively before filling
with concrete. This is necessary to help in the determination of the moisture content of the
specimens. The bamboo samples were then filled with concrete paste of ratio 1:3:6 and allowed for
a week to dry. The weight of the dried concrete were weighed again to know the overall weight of
the reinforced fresh bamboo; with (W1 ) as weight of reinforced fresh bamboo. Fig.4 shows
reinforced fresh bamboo sample under crushing strength test.

Fig. 4 Fresh Reinforced Bamboo under Compressive Test

Moisture Content Analysis

This was performed to know the effect of moisture content on the compressive strength of the fresh
reinforced bamboo. All the specimens used for this analysis were from the same culm, and thus
were weighed to record their initial weights before filling with concrete. According to Head [15],
the moisture content formula is given as:

Moisture Content (MC %) = 100 % (3.1). Where, MC % is the moisture content of fresh
reinforced bamboo specimen, WF is the Mass of fresh reinforced bamboo, and WD is the Mass of
dried reinforced fresh bamboo.

Results and Discussion


Experimental results are shown in table 4.1, table 4.2. and table 4.3. Five specimens of the same
height were crushed and for the sake of accuracy of results, average loads of failure were calculated.
178 Metal, Ceramic and Composite Materials

For instance, for a height of 250 mm, five specimens of height 250 mm were crushed and average
load of failure was calculated. A similar procedure was followed for the heights of specimen of
210,170, 130 and 90 mm for fresh reinforced bamboo specimen. From table 4.1 it could be seen that
the strength (load of failure) increases with increases in the height or the strength increases as the
height increases. This is due to the fact that the concrete gave additional strength to the bamboo.
Therefore, concrete is significant in reinforcing bamboo structures for a better strength. The
moisture content also has significant influence on the strength of the fresh reinforced bamboo.
Using Head, K.A., the moisture content reduced the strength of the fresh reinforced bamboo by 21.8
%. This percentage is really significant in strength analysis since it can cause catastrophic failure to
structures. From table 4.3, it is clear that the average failure stresses also increases with increases in
height of the specimens [15].
All the failure stresses/crushing stresses were computed using equations (4.4) and (4.5). According
to Gere and Timoshenko [14], the stress that caused the reinforced bamboo specimens to fail is the
total compressive stress ( ):

= (4.3)
Where,

Stress due to oncrete (c) = (4.4)

Stress due to Bamboo (B) = (4.5)

where, (E) = Modulus of Elasticity, (EC) = Modulus of Elasticity For Concrete = 17 GPa, (EB)=
Modulus of Elasticity For Bamboo = 11 GPa, (PL ) = Load of failure, (AB ) = Cross-Sectional Area
of Bamboo Samples, (AC ) = Cross-Sectional Area of Concrete Samples.
The failure stress of the bamboo is about say 1.26 M times the weight of the bamboo per square
meter (see table 4.3).
Table 4.1 Results for Fresh Reinforced Bamboo Specimen

Specimen Length External Internal Load of Mass of Mass of Volume of the Density of
Type (mm) Diameter - Diameter-Di (mm) Failure (kN) Bamboo(WF) Reinforced Reinforced Reinforced
Do (mm) (g) Bamboo Bamboo (m3) Bamboo
(W1)(g) (kg/m3)

Fresh 250.0 86.00 74.00 84.00 516.90 2830.30 1.4510-3 1948.51


-3
Fresh 250.0 86.00 74.00 83.00 521.54 2785.98 1.4510 1920.69
-3
Fresh 250.0 86.00 74.00 85.00 512.65 2635.56 1.4510 1817.24
Fresh 250.0 86.00 74.00 82.00 519.40 2819.02 1.4510-3 1944.14
-3
Fresh 250.0 86.00 74.00 84.00 514.29 2677.43 1.4510 1846.21
-3
Average 250.0 86.00 74.00 83.60 516.96 2749.66 1.4510 1895.36

Fresh 210.0 85.00 73.00 80.00 420.60 2187.50 1.1910-3 1837.82


Fresh 210.0 85.00 73.00 82.00 432.07 2157.89 1.1910-3 1812.61
Fresh 210.0 85.00 73.00 79.00 418.67 2179.43 1.1910-3 1831.09
Fresh 210.0 85.00 73.00 80.00 422.60 2189.51 1.1910-3 1839.50
Fresh 210.0 85.00 73.00 81.00 415.89 2121.99 1.1910-3 1782.35
Average 210.0 85.00 74.00 80.40 421.97 2167.26 1.1910-3 1820.67

Fresh 170.0 84.00 72.00 76.00 366.6 1757.00 9.4210-4 1864.74


Fresh 170.0 84.00 72.00 77.00 371.3 1755.32 9.4210-4 1863.06
Fresh 170.0 84.00 72.00 72.00 362.7 1723.89 9.4210-4 1829.09
Fresh 170.0 84.00 72.00 75.00 364.9 1745.67 9.4210-4 1877.92
Advanced Materials Research Vol. 1102 179

Fresh 170.0 84.00 72.00 79.00 357.2 1769.42 9.4210-4 1877.92


-4
Average 170.0 84.00 72.00 75.80 364.54 1750.26 9.4210 1862.55

Fresh 130.0 84.00 72.00 68.00 257.90 1314.60 7.2110-4 1823.67


-4
Fresh 130.0 84.00 72.00 65.00 258.20 1344.21 7.2110 1864.10
Fresh 130.0 84.00 72.00 63.00 261.40 1320.32 7.2110-4 1830.80
-4
Fresh 130.0 84.00 72.00 70.00 253.70 1322.45 7.2110 1833.60
-4
Fresh 130.0 84.00 72.00 69.00 259.01 1312.43 7.2110 1819.70
Average 130.0 84.00 72.00 67.00 258.05 1322.80 7.2110-4 1834.37

Fresh 90.0 82.00 71.00 60.00 193.70 926.10 4.7510-4 1948.03


-4
Fresh 90.0 82.00 71.00 64.00 194.60 932.14 4.7510 1962.11
Fresh 90.0 82.00 71.00 59.00 192.50 919.79 4.7510-4 1934.74
-4
Fresh 90.0 82.00 71.00 64.00 190.90 937.12 4.7510 1972.63
-4
Fresh 90.0 82.00 71.00 67.00 193.50 916.66 4.7510 1928.42
-4
Average 90.0 82.00 71.00 62.80 193.04 926.36 4.7510 1949.19

Table 4.2 Summary Results for Fresh Reinforced Bamboo-part 1


Height External Internal Mass of Mass of Mass of Average Time of
of Fresh Diameter Diameter Only Reinforced Only Load of Failure
Bamboo (mm) (mm) Bamboo Bamboo Concrete Failure (Seconds)
(mm) (g) (W1)(g) (g) (kN)

250.00 86.00 74.00 516.96 2830.30 2313.34 83.60 57:62


210.00 85.00 74.00 421.97 2187.50 1765.53 80.40 48:92
170.00 84.00 72.00 364.54 1757.20 1392.66 75.80 39:77
130.00 84.00 72.00 258.05 1314.60 1056.55 67.00 35:02
90.00 81.00 71.00 193.04 926.10 733.06 62.80 29:71

Table 4.3 Results for Fresh Reinforced Bamboo-part 2

Height External Internal Area of Area of Average Average Time of Failure Load Specific strength to
(mm) Diameter Diameter Concrete Bamboo Load of Failure Failure Stress To Compres- volume
(mm) (mm) (Ac) (AB) Failure Stress (Seconds) To Height sive ratio
(mm2) (mm2) (kN) ( MPa ) Weight Ratio Strength (MPa/m3)
Ratio (KN/m) (KNm/kg)
(1/m2)

250.00 86.00 74.00 4300.84 1507.96 83.60 26.09 57:62 0.97 M 334.40 13.717 17993.10
210.00 85.00 74.00 4300.84 1373.66 80.40 25.52 48:92 1.20 M 382.86 13.602 21445.38
170.00 84.00 72.00 4071.50 1470.27 75.80 24.85 39:77 1.45 M 445.90 13.342 26380.04
130.00 84.00 72.00 4071.50 1470.27 67.00 21.97 35:02 1.70 M 515.38 11.977 30471.57
90.00 82.00 71.00 3959.19 1193.81 62.80 21.86 29:71 2.40 M 688.89 11.215 46021.05
180 Metal, Ceramic and Composite Materials

Fig.4.4 A Graph of Load versus Time of Failure Fig.4.7 A Graph of Crushing Stress versus Time of Failure

Fig.4.6 Failure Stress to Weight Ratio versus Height Fig.4.5 Specific Compressive Strength versus Height

Fig.4.9 Strength to Volume Ratio versus Height Fig.4.8 Load to Height Ratio versus Time
Advanced Materials Research Vol. 1102 181

From Fig. 4.4 and Fig.4.7, it is clear that the average loads and crushing stresses increase as the
height of the bamboo and time of failure increase respectively. At a height of 250 mm the average
load of failure for fresh reinforced bamboo is 83.60 kN.
The specific compressive strength which explains the bending moment per kilogram of the
reinforced bamboo specimen also increases with increases in height. This is shown in Fig. 4.5.
Failure stress to weight ratio (i.e. Fig.4.6) decreases with increases with the height of the bamboo.
This makes emphasis on the stress per unit area of the bamboo specimen. The failure stress of the
bamboo is about say 1.26 M times the weight of the bamboo per square meter. This indicates
clearly that short bamboo has greater strength than longer bamboo. Load to height ratio also
decreases with increase with the height of the bamboo. This explains the load of failure per meter of
the bamboo specimen (see Fig.4.8). The strength of bamboo has significance in its varying heights.
Strength to volume ratio also decreases with increases with the height from the bamboo. It signifies
stress per meter cube of the bamboo specimen.

Conclusions
The following conclusions were arrived at after the crushing tests experiment and moisture content
analyses were performed.

Increase in height of the reinforced bamboo increases the strength and vice versa.
The failure stress of the bamboo is about say 1.26 M times the weight of the bamboo per
square meter.
The specific compressive strength which explains the bending moment per kilogram of the
bamboo specimen also increases with increases in height.
Failure stress to weight ratio decreases with increases with the height of the bamboo.
The moisture content also has significant influence on the strength of the fresh reinforced
bamboo.
The moisture content reduced the strength of the fresh reinforced bamboo by 21.8 %.This
percentage is really significant in structural analysis since it can cause catastrophic failure to
structures.
Strength to volume ratio also decreases with increases with the height from the bamboo.
That means stress per meter cube of the bamboo specimen.
Concrete is significant in reinforcing bamboo structures for a better strength.
The crushing stresses for fresh reinforced bamboo of heights 250 , 210 ,170 , 130 and 90
mm were found to be 26.09 , 25.52 , 24.85 , 21.97 , 21.86 MPa respectively.

Recommendations
It is recommended that:

Construction companies and industries in Ghana, Liberia, China and Africa in general
should be encouraged to use bamboo as an alternative to timber for structural purposes in
order to reduce the depletion of our forest reserve.
Reinforced fresh bamboo has significant strength and is hence, recommended for
construction industry.

Acknowledgement
Authors are also thankful to the management of the Mechanical laboratory of the W.V.S. Tubman
University, Liberia and Shenyang National Lab. for Materials Science, Institute of Metal Research,
Chinese Academy of Sciences, China.
182 Metal, Ceramic and Composite Materials

References
[1] Kishen, J., Ghosh, D.P., and Rehman, M.A. Studies on moisture content, shrinkage, swelling
and intersection point of mature (Dendrocalamus strictus) male bamboo. Indian Forest Rec.,
Vol. 1: (1956), pp. 1 - 30.
[2] Liese, W. and G. Weiner., Modifications of bamboo culm structures due to ageing and
wounding. The Linnean Society, London, (1997), pp. 313 - 322.
[3] L. Gyansah, A.S. Akinwonmi and M. Affam, The Fracture Behaviour of Fresh Bamboo
Under Uniaxial Compressive Loading Condition. Research Journal of Applied Sciences,
Engineering and Technology, 2(8): , (2010), 720-726.
[4] Limaye, V.D., Strength of bamboo (Dendrocalamus strictus). Ind. For. Rec., Vol. 78,
(1952), pp. 558 - 575.
[5] L. Gyansah and S. Kwofie, Investigation into the Performance of Bamboo Using the
Notched and the Un-Notched Specimen. Research Journal of Applied Sciences, Engineering
and Technology, 3(04): (2011), 245-251.
[6] hudley, R., oncrete, Construction Technology, Second SI Edition, ELBS with Longman
(Publishers) Ltd, Vol. 1, (1994), pp. 20-24.
[7] Swamy, R.N., Sustainable oncrete for Infrastructure Regeneration and Reconstruction,
International Conference on Sustainable Construction into the next Millennium
Environmentally friendly and innovative Cement Based Materials, Joao Pessoa, Brazil,
(2000), pp. 15-44.
[8] Maheshwari, S. and Satpathy, K., Pulp and papermaking characteristics of nodes, internodes
and culm of bamboo Dendrocalamus Strictus, IPPTA, Vol. 25, (1988), pp. 15 19.
[9] Zhou, F. . Studies on physical and mechanical properties of bamboo woods,
Nanjing Technology College Forest Production, Vol. 2, (1981), pp. 1 32.
[10] Espiloy, Z. B. Physico-Mechanical properties and Anatomical relationships of some
Philippine bamboos. Proceedings of International Bamboo Workshop Hangzhou, China,
Singapore, P.R, Vol.1, (1987), pp. 257 - 264.
[11] Seema, J. and Kumar, R. Mechanical behaviour of bamboo and bamboo composite, Journal
of Material Sci., (1992), pp. 4598 4604.
[12] Ghavami, K. and Rodrigues, . S., Engineering Materials and Components with Plants,
CIB Symposium, Construction & Environment, Theory into Practice Proc., So Paulo,
Brazil, CD-ROM, ISBN 85-88142-01-5, Global Seven Editor, (2000), pp.1-16.
[13] Lakkad, S. . and Patel, J.M. Mechanical Properties of bamboo, a natural omposite, Fiber
Science Technology, Vol. 14, (1980), pp. 319 - 322.
[14] Gere, J.M. and Timoshenko, S.P. Mechanics of Machines, Third SI Edition, S. Thornes
(Publishers) Ltd, Cheltenham, (1998), pp.761, ISBN 0-7487-4084-8.
[15] Head, K.A., Manual of Soil Laboratory Testing, Soil Classification and Compaction Test.
2nd Edn., Pentech Press,London, (1992), pp: 68-76.
Keywords Index

3D Finite Element Model 37


6N01 Aluminum Alloy 23 D
Dioxin 27
A Dynamic Mechanical Thermal 139
Alkaline Treatment 139 Properties
Alloying of Aluminum 27 Dynamic Recrystallization 17
Alumina Membrane 91
Alumina-Mullite-Zirconia 87 E
Composite
Effect 169
Aluminum 7, 55
Elastic Properties 155
Aluminum Alloy 33
Electrical Properties 107
Arenga Pinnata Fibre 139
Electron Beam Welded Joints 149
Arenga Pinnata Fibres 103
Emission Factor 27
Austenite Stainless Steel 17
EPMA 129
Esterification 99
B Eu3+\Tb3+ 125
Baddeleyite 79
Bearing Steel (JIS, SUJ2) 45 F
Bessel Beams 143
Failure Stress to Weight Ratio 173
Biomaterial 95
Fatigue 103
Brine 55
Fatigue Fracture 59
Fatigue Strength 33
C Femtosecond Laser 143
Ca(OH)2 161 Fiberglass Plastics 155
Carbon Dioxide 73 Fibre Reinforced Composites 103
Cement Kiln 27 Films 107
Characterisation 99 Finite Element Modeling 149
Clad 7, 51, 55 Flash X-Ray Radiography 119
Co Doping 11 Fracture Surface 135
Coal Burning 27 Fracture Toughness 87
Cohesive Zone Model 149 Friction Coefficient 129
CoMn2O4 11 Fuel Cladding Tube 3
Compact Tension Specimen 59
Composite 51 G
Composite Membrane 99
Galvanic Corrosion 55
Composite Patch Adhesively 37
Bonded Repair Gas Permeation 91
Composites 139 Gas Separation 73
Compression Test 17 Gas Transport 99
Copper 7, 51, 55 Global Warming 73
Cosine Function 169 Graphene 107
Coverage 33
Crack Propagation 149 H
Creep Properties 3 Heat Treatment 3, 7, 51, 55
Cu-In-Ga-Se Films 83 High Aspect Ratio 143
184 Metal, Ceramic and Composite Materials

High Temperature 67
Homology 59, 113 P
Homology Image Analysis 45
Patch Designing 37
Homology Method 135
PBS 95
Hooks Law 155
PEK-c 107
Hot Deformation Behavior 17
Penetration Depth 119
Hybrid Polymeric Material 125
Polymer Bearing 129
Hydrogen Separation 91
PPS 129
Prior Austenite Grain Size 45
I PTFE 129
Image Analysis 135 Pulsed Laser Deposition 83
Image Analysis for Molecular 113
Structure
R
Inorganic Ceramic Membrane 73
Interface 51 Refinement 45
Intermetallic 55 Reinforced Fresh Bamboo 173
Intermetallics 51 Repeated Quenching 45
Rolling Contact Fatigue 129

J
S
Jet Formation 119
Jet Penetration 119 SCR 11
Secondary Smelting 27
Selenization Annealing 83
L
Separation Factor 73
Lignite 161 Shaped Charge 119
Liner 119 Shot Peening 33
Load to Height Ratio 173 SIF 37
Low Carbon Steel (S45C) 59 Silica Gel 113
Luminescence 125 Solar Selective Coating 67
Specific Compressive Strength 173
M Steam Gasification 161
Magnesium Phosphate 95 Steels 33
Mechanical Properties 23 Strength 17, 155
Microstructure 23 Strength to Volume Ratio 173
Microstructure Fabrication 143 SUJ2) 45
MnOx/TiO2 11 Surface Energy 169
Monoclinic 79
Multilayer 67 T
Tar Reforming 161
N Tension Test 7
Nanocrystals 79 Thermal Stability 67
Nanoscale 169 Titanium 51
Nitrogen Adsorption 99 Toughness 87
TTA 125
O
Optical Properties 83
U
Orthogonal Design 23 Unidirectional Rings 155
Oxide Nanoparticles 87
Advanced Materials Research Vol. 1102 185

W
Wet Winding 155
WGS Reaction 161

Z
Zr-1Nb-0.12O 3
ZrO2 79
Authors Index

A
Abd El-Aty, A.M. 173 K
Abdul Khadir, M.S. 139 Kajama, M.N. 73, 91
Abdullah, A.H. 103, 139 Kang, G.T. 55
Ahmed, M. 119, 119 Kashima, Y. 129
Andreev, V.I. 155 Kida, K. 45, 59, 129,
135
Kim, I.K. 7
C
Kim, I.W. 3
Chen, J. 95 Kim, W.N. 7
Chen, S.Y. 83 Kim, Y.K. 51
Cheng, Y.Y. 27 Kuperman, A.M. 155
Cheong, S.K. 33
L
D
Lai, F.C. 67
Ding, W.Y. 37 Lee, Y.S. 33
Ding, Y.T. 95 Li, B. 23
Dong, X.P. 95 Li, D.F. 23
Li, H.T. 125
F Li, J. 107
Fang, D. 11 Li, Y. 125
Lin, J.J.M. 27
G Lin, L.M. 67
Liu, D. 125
Gobina, E. 73, 91, 99
Liu, J.Y. 67
Gu, L.Y. 107
Lu, M.F. 27
Gyansah, L. 173

M
H
Malik, A.Q. 119
Halasz, I. 113
Mat, M.F. 103
He, F. 11
Matsuda, W. 45
Hong, S.I. 3, 7, 17, 51,
55, 79 Mizobe, K. 45, 59, 129
Hu, H. 11
Hu, S. 161 N
Huang, Z.G. 83 Nakane, K. 45, 59, 113,
Huang, Z.X. 119 135
Nik Roseley, N.R. 139
I Nwogu, N.C. 73, 91

Ishida, M. 59
O
J Oh, H.K. 17
Okon, E. 99
Jin, S.J. 37
Ou, Z.Y. 169
Advanced Materials Research Vol. 1102 187

Xie, J.L. 11
P Xu, F.L. 23
Park, S.M. 17
Poowancum, A. 87 Y
Yang, R.H. 67
Q Yu, B.Q. 95
Yuvakkumar, R. 79
Qiu, X.S. 107
Qu, Y. 67
Z
S Zhang, B.Y. 107
Zhang, D. 37
Sakamoto, J. 33
Zhang, H.Q. 83
Schmauder, S. 149
Zhao, L.L. 143
Shehu, H. 99
Zhao, W.W. 143
Shi, F.H. 107
Zheng, W.F. 67
Shi, X.C. 129
Zhou, C.H. 23
Shin, P.W. 51
Zu, X.D. 119
Shuai, C. 161
Su, F. 37
Su, S. 161
Sun, L.S. 161

T
Takiyama, A. 135
Tang, L. 83
Tang, X.F. 95
Tu, H.Y. 149
Turusov, R.A. 155

W
Wang, F. 143
Wang, H.B. 23
Wang, J.L. 125
Wang, Q.X. 95
Wang, X. 125
Wang, X.M. 23
Wang, Y. 161
Watcharamaisakul, S. 87
Weber, U. 149
Wei, J. 95
Wu, J.J. 23
Wu, Y.W. 169
Wu, Y.Y. 95

X
Xiang, J. 161
Xiao, Q.Q. 119
Xie, J. 143